Advances in Geophysics ~ Estuarine Physics and Chemistry-Studies in Long Island Sound, Vol. 22

ADVANCES IN

GEOPHYSICS

VOLUME 22

Contributors to This Volume ROBERTC. ALLER L. K. BENNINGER HENRYJ. BOKUNIEWICZ J. K. COCHRAN ROBERTB. GORDON RICHARDJ. MCCAFFREY JOHN THOMSON K. K. TUREKIAN

Advances in

GEOPHYSICS VOLUME 22

Estuarine Physics and Chemistry: Studies in Long Island Sound Edited by

BARRY SALTZMAN Department of Geology and Geophysics Yale University New Haven, Connecticut

1980

Academic Press A Subsidiary of Harcourt Brace Jovanovich, Publishers

New York

London Toronto Sydney San Francisco

COPYRIGHT @ 1980, BY ACADEMIC PRESS,INC. ALL RIGHTS RESERVED. NO PART OF THIS PUBLICATION MAY BE REPRODUCED OR TRANSMITTED IN ANY FORM OR BY ANY MEANS, ELECTRONIC OR MECHANICAL, INCLUDING PHOTOCOPY, RECORDING, OR ANY INFORMATION STORAGE AND RETRIEVAL SYSTEM, WITHOUT PERMISSION IN WRITING FROM THE PUBLISHER.

ACADEMIC PRESS, INC.

111 Fifth Avenue, New York, New York 10003

United Kingdom Edition published by ACADEMIC PRESS, INC. (LONDON)LTD. 24/28 Oval Road, London N W l 7 D X

LIBRARY OF CONGRESS

CATALOG CARD

NUMBER: 52-12266

ISBN 0-12-018822-8 PRlNTED 1N THE UNlTED STATES OF AMERICA

80 81 82 83

98 7 6 5 4 3 2 1

CONTENTS LIST OF CONTRIBUTORS ................................................................. FOREWORD.................................................................................. PREFACE.....................................................................................

ix xi xiii

The Sedimentary System of Long Island Sound

ROBERTB. GORDON 1. Introduction ............................................................................. 2 . Geological History .................................................................... 3 . Sea Level ................................................................................ 4 . Physical Oceanography .............................................................. 5. Sedimentation .......................................................................... 6. Further Research ...................................................................... References ...............................................................................

1 2 12 20

25 33

35

Storm and Tidal Energy in Long Island Sound

HENRYJ . BOKUNIEWICZ AND ROBERTB. GORDON 1. Introduction ............................................................................. 2. Tidal Energy ............................................................................ 3 . Storm Energy ........................................................................... 4 . Water Level Deviations ............................................................. 5. Conclusions ............................................................................. Appendix I . Formulation of the Energy Balance in an Embayment .... Appendix I1. Estimate of Tidal Dissipation of All of Long Island Sound .................................................................. References ...............................................................................

41 43 48 55

60 61

65 67

Sediment Transport and Deposition in Long Island Sound

.

HENRYJ BOKUNIEWICZ AND ROBERT B. GORDON 1. Introduction

............................................................................. 2 . Power Sources ......................................................................... 3. Sediment Sources ..................................................................... 4 . Sediment Transport and Bottom Stability ...................................... 5 . Sediment Deposition and Distribution ........................................... 6. Comparison with Other Estuaries ................................................ References ............................................................................... V

69 70 84 87 95 99 104

vi

CONTENTS

Sand Transport at the Floor of Long Island Sound

HENRYJ . BOKUNIEWICZ

1. Introduction

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

2 . Background ............................................................................. 3 . Long Island Sound .................................................................... 4. Sediment Transport ................................................................... 5 . Formation of the Transition Zone ................................................ 6 . Discussion ............................................................................... 7 . Summary and Conclusions ..........................................................

References

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

107 107 110 113 116 122 124 126

The Sources and Sinks of Nuclides in Long Island Sound

K . K . TUREKIAN. J . K . COCHRAN. L . K . BENNINGER. AND ROBERTC . ALLER 1. Introduction ............................................................................. 129 2. Sources of Trace Metals Delivered to Long Island Sound ................. 131 3 . The Distribution of Trace Metals in Long Island Sound Sediments .... 137 4. Trace-Metal Distributions in Mussels and Oysters: An Index of the Composition of Suspended Particles ................................................... 142 5. Processes Affecting the Deposition and Accumulation of Trace Metals in Long Island Sound Sediments .................................................. 147 6. Processes Affecting the Vertical Distribution of Nuclides in the Sediment Pile ........................................................................... 153 161 7 . Summary ................................................................................. References ............................................................................... 163

A Record of the Accumulation of Sediment and Trace Metals in a Connecticut Salt Marsh

RICHARDJ . MCCAFFREY AND JOHNTHOMSON 1. Introduction ............................................................................. 2 Experimental Methods and Results .............................................. 3 . Discussion ............................................................................... 4. Summary and Conclusions ..........................................................

.

References

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

165 169 189 227 229

CONTENTS

vii

Diagenetic Processes Near the Sediment-Water Interface of Long Island Sound .I . Decomposition and Nutrient Element Geochemistry (S. N. P)

ROBERTC . ALLER 1. Introduction ............................................................................. 2 . Location of Study and Station Description .................................... 3 . Sampling ................................................................................. 4 . Methods .................................................................................. 5 . Results .................................................................................... 6. Discussion ............................................................................... 7 . Summary ................................................................................. Appendix A . Macrofauna (>1 mm) Sieved from Flux-Core Boxes ..... Appendix B . Box-Core and Gravity-Core Data from Long Island Sound .................................................................. Appendix C . Flux-Core Data ...................................................... List of Symbols ........................................................................ References ...............................................................................

238 238 250 252 257 272 317 320 322 340 343 343

Diagenetic Processes Near the Sediment-Water Interface of Long Island Sound.11. Fe and Mn

ROBERTC . ALLER 1. 2. 3. 4.

Introduction ............................................................................. Location and Description of Study Area ....................................... Sampling ................................................................................. Methods .................................................................................. 5 . Results .................................................................................... 6. Discussion ............................................................................... 7 . Summary ................................................................................. List of Symbols ........................................................................ References ...............................................................................

351 352 353 353 355 367 406 409 410

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

417

Index

This Page Intentionally Left Blank

LIST OF CONTRIBUTORS Numbers in parentheses indicate the pages on which the authors' contributions begin.

ROBERT C. ALLER,* Department of Geology and Geophysics, Yale University, New Haven, Connecticut 06520 (129, 237, 351) L. K. BENNINGER,~ Department of Geology and Geophysics, Yale University, New Haven, Connecticut 06520 (129) HENRYJ. BOKUNIEWICZ,Marine Sciences Research Center, State University of New York, Stony Brook, New York 11794 (41, 69, 107) J. K. COCHRAN, Department of Geology and Geophysics, Yale University, New Haven, Connecticut 06520 (129)

ROBERT B. GORDON,Department of Geology and Geophysics, Yale University, New Haven, Connecticut 06520 (1, 41, 69)

RICHARDJ. MCCAFFREY,'Department of Geology and Geophysics, Yale University, New Haven, Connecticut 06520 (165) JOHNTHOMSON,'Department of Geology and Geophysics, Yale University, New Haven, Connecticut 06520 (165)

K. K. TUREKIAN, Department of Geology and Geophysics, Yale University, New Haven, Connecticut 06520 (129)

* Present address: Department of Geophysical Sciences, The University of Chicago, Chicago, Illinois 60637. Present address: Department of Geology, University of North Carolina, Chapel Hill, North Carolina 27514. * Present address: Graduate School of Oceanography, University of Rhode Island, Kingston, Rhode Island 02881. Present address: Institute of Oceanographic Sciences, Wormley, Godalming, Surrey GU8 SU8, England. ix

This Page Intentionally Left Blank

FOREWORD Occasionally, a complete volume of Advances in Geophysics will be devoted to a single topic of special interest. This volume, dealing with some of the important physical, chemical, and biological processes occurring in estuaries, does so. The works included here have an additional common feature: all are based on studies made in Long Island Sound, which poses problems prototypal of those encountered in many estuarine settings. We are still a long way from establishing the kind of general quantitative theory of the estuarine variables that is necessary for the understanding and effective management of these coastal environments. The articles in this volume expose the rich variety of phenomena and interactions that will have to be included in such a theory:Although particular emphasis is placed here on the transport, physicochemical structure, and evolution of the bottom sediments, the relationship of these factors to their broader geological and hydrodynamical contexts is also considered and clarified. BARRYSALTZMAN

xi

This Page Intentionally Left Blank

PREFACE Estuaries are usually classified in terms of the characteristics of their circulation, broadly speaking, as “salt wedge,” and “partially mixed.” Estuaries differ in many other ways-for example, in the nature and amount of sediment they transport and the composition and diversity of the biological communities they support. No two estuaries, no matter how close in affinity they may be according to hydrologic classification, are ever identical. Long Island Sound is partially mixed, but its estuarine circulation is due to a “freshwater” source that is actually the brackish water of New York Harbor-water that receives large quantities of wastes from New York City. The two major rivers entering the Sound are the Connecticut and the Housatonic. Their hydrologic importance is local, but they are the principal sources of the “natural” sediment entering the Sound. The eastern passes of the Sound are the major source of its tidal flow, although the western pass is also of local importance. The energy of the tides and estuarine circulation has a major influence on sediments and animal communities and is harnessed most dramatically at times of strong wind stress. The atmosphere also is a courier of pollutants to the Sound, mainly from New York City and the industrial area southwest of it. The chemicals transferred through the atmosphere leave their imprints on the materials deposited in the Sound, together with those transported hydrographically. We see Long Island Sound as an “urban” estuary occupying a basin-like setting, protected from the highly energetic encounters with the open ocean, and therefore capable of retaining a memory of estuarine processes in its deposits over its entire existence. This collection of articles represents a part of the work that has been going on in the Department of Geology and Geophysics at Yale over the past 15 years. It builds on the classic work done there in earlier years by members of the Bingham Laboratory, notably the group associated with Gordon Riley. Our work has been aimed at understanding the fluxes of materials to and within the Sound as influenced by natural forces and the activities of people. The approach has been both chemical and physical. The articles herein refer to a considerably larger body of research at Yale and at other institutions. In particular the work of our Yale colleagues Robert A. Berner and Donald C. Rhoads and their research groups must be cited. Our four research groups have pursued investigations of Long Island Sound with perspectives and methodologies characteristic of each group, but the net effect on our understanding of the system has been far

xiii

xiv

PREFACE

greater than a simple summing of the individual efforts. The references cited in the contributions of this cluster clearly attest to this. We feel that we now have a much better understanding of how one estuary functions as a whole. This understanding yields a sense of satisfaction, but does not leave us in a state of complacency; much remains to be learned about the complexities of Long Island Sound, let alone about the other estuaries of the world. Nonetheless, we believe that many of the methods we have used to examine the Sound will prove useful in studies elsewhere. Funding for these studies has come from several sources. These include government agencies, particularly the Department of Energy (and its antecedents), the U.S. Army Corps of Engineers, and the National Science Foundation. In addition, the United Illuminating Company provided financial support as part of its interest in the environmental consequences of the construction of a new power plant adjacent to New Haven Harbor. These and other benefactors are cited in each contribution as appropriate. KARLK. TUREKIAN ROBERTB. GORDON

ADVANCES IN

GEOPHYSICS

VOLUME 22

This Page Intentionally Left Blank

THE SEDIMENTARY SYSTEM OF LONG ISLAND SOUND ROBERTB. GORDON Deparfment of Geology and Geophysics Yale University New Haven, Connecricuf

I. 2. 3. 4. 5. 6.

Introduction . . . . . Geological History . . Sea L e v e l . . . . . . Physical Oceanography Sedimentation . . . . Further Research . . . References . . . . . .

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

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

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

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

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

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

1

2 12 20 25 33 35

1 . INTRODUCTION

Water-transported products of erosion from the land surface (both solutes and particulates) reach the sea by passing through estuaries. The flow of water from a source area through a river system to an estuary is governed by the balance of the downhill component of the gravitational force and the friction between the fluid and the bed of the stream. Sediment moves intermittently with the fluid flow; most is in motion during periods of high discharge. A wider range of physical, chemical, and biological phenomena are active in an estuary than in a river, and the processes of sediment transport and deposition become both more diverse and complicated. These include periodic water motions due to the tide and to waves, both those generated in the estuary and the swell from the sea. Salt is encountered and may have both chemical and physical effects on the form of the erosion products. Estuarine circulation results from the density difference between fresh and salt water. The length scales of the flow are increased in the estuary and geostrophic effects begin to be important. There may be large lateral flows and sharp boundaries between different water masses in the estuary. Sediments entering an estuary encounter marine animals which ingest them and transform their physical form; in shallow water, interaction with marine plants may facilitate sediment deposition. As a result of all these processes the sediment transported by a river may be processed into a new form within the estuary, the distribution between elements in solution and those adsorbed on solid particles may be changed, some materials may be stored tem1 ADVANCES IN GEOPHYSICS. VOLUME

22

Copyright 8 1980 by Academic Press. Inc. All rights of reproduction in any form resewed, ISBN 0-12-018822-8

2

ROBERT B. GORDON

porarily or permanently in the estuary, and what is exported may be of different size, shape, or composition than that supplied by the river. The number of physical processes that control the behavior of erosion products in an estuary is large and most of these processes are sensitive to the boundary conditions that obtain in each individual estuary. Description of them is likely to be of only parochial interest and the difficulties in the way of generalization are great. There are satisfactory schemes for the hydrographic classification of estuaries (Dyer, 1973, Chapter 2), but none as yet for the classification of estuarine sedimentary systems. In fact, the full scope of the interaction among physical, chemical, and biological processes in an estuary has hardly been examined. Such an examination is one objective of the articles in this volume. The selection of Long Island Sound for this purpose is due partly to geographical convenience; nevertheless, the Sound does display many estuarine characteristics in relatively simple form and is in many ways typical of the estuaries on glaciated terrain that surround the North Atlantic. Often these are the estuaries that have suffered most from the impacts of industrial society. We believe that many of the specific results obtained for Long Island Sound, and most of the methods developed to study it, will find application in the study of other estuaries. The quantitative description of the processes that control the passage of the products of erosion through the estuary will provide a basis for the efficient intercomparison of other estuaries, i.e., the development of an estuary classification scheme. This article is a review of previous research on the geology and related physical oceanography of Long Island Sound. 2. GEOLOGICAL HISTORY

The basin occupied by Long Island Sound is a product of the period of prolonged erosion of eastern North America that occupied the late Mesozoic and the Cenozoic eras. Because the onshore geological record consists of erosion surfaces, and the products of this erosion are now submerged on the continental shelf, reconstruction of a detailed geological history is not possible. However, the main events are well established. When the Atlantic Ocean began to open, about 180 MY BP, the Southern New England land surface had high relief, as shown by block faulting of sediments in the Connecticut Valley, which persisted into the Jurassic. Between the mid-Jurassic and early Cretaceous this relief was reduced to no more than about 100 m (McMaster and Ashraf, 1973, and other references given therein). The eroded surface produced then is now the Fall Zone surface (Flint, 1963) and its northern bound is the Fall Line.

--

THE SEDIMENTARY SYSTEM OF LONG ISLAND SOUND

3

The Fall Zone surface defines the top of the bedrock beneath Long Island Sound. A simple sequence of events that will account for the formation of the basin now occupied by the Sound is illustrated in the series of sketches in Fig. I , After opening of the Atlantic Ocean (Fig. la), cooling of the continental margin resulted in its subsidence (Fig. lb) at a rate that decreased as the continent moved away from the mid-ocean ridge. Additional subsidence was caused by the weight of the sediment accumu-

',

....... ....:...*..: ..:,;:;. .......::..

. I : : .

%.

(f)

2 MY bp

I.

I

1

(0)

today

.*$.... ......

Now England '/ I T - -$ .r:X+ ,,. Upland Surfacr Fall Zonr fau

F a . I . Schematic diagram of the evolution of the basin of Long Island Sound showing: (a) Start of opening of Atlantic Ocean: (b) subsidence of the continental margin due to cooling as the continent moves away from the mid-ocean ridge; (c) accumulation of sediment on the continental margin causing further subsidence; (d) reduction in sea level due to worldwide reduction in the rate of sea-floor spreading: (e) erosion of coastal plain sediments due to lowered base level of rivers: (0 glaciation: and (9) present configuration. (See text for sources.)

4

ROBERT B. GORDON

lating on the continental shelf (Watts and Ryan, 1976). This contribution to the subsidence can be described as resulting from downward, elastic flexure of the continental crust about a hinge line, now identified as the Fall Line. The mean downward tilt of the Fall Zone surface today is 9.4 x lo-'. If this developed over the full 170 Myr available, the average tilt rate would be only 5 x lo-" yr-I, which can be easily accounted for by cooling of the continental margin and deflection under the accumulating sediment load. [Tilting rates along the east coast of North America are observed to be as great as 2 x yr-' at present (Brown, 1978); this suggests that fluctuating rather than slow, steady tilts have contributed to the long-term average tilt rate.] Since the downwarping of the continental margin did not require uplift inland of the Fall Line, erosion rates on much of the land surface supplying sediment to the continental margin must have remained relatively low. Menard (1961) estimated that 7.8 x lo6 km3 of rock must have been removed from the Appalachians over 125 Myr to account for the sediment now on the continental terrace and rise and on the abyssal plains off the east coast of North America. The mean sediment yield required to produce this material is -0.2 kg/(m2 yr). (For comparison, this is about the same as the sediment yield of the Missouri River drainage basin today.) Matthews (1975) has used the more extensive data on sediment thicknesses off the Atlantic Coast now available to estimate that the sediment yield of eastern North America over the past 60 Myr was 0.012 kg/(m2 yr) for the northern half of the coast and 0.067 for the southern half. These sediment yields can be attained with a land surface relief of a few hundred meters under temperate climate conditions and so are consistent with the hypothesis that both the elevation and relief of most of the land surface have remained moderate since the opening of the Atlantic Ocean. Continued erosion of the New England upland and deposition on the continental margin resulted in the configuration bedrock, sediment, and sea surface shown in Fig. lc. Sediment cover extended at least to the Fall Line (Sharp, 1929, p. 38; Johnson, 1931, p. 39), while the shore line was not far removed. According to Pitman (1978) the shoreline began to move rapidly seaward starting at about 65 MY BP because of an acceleration in the rate of fall of the sea-surface elevation caused by diminished world average sea-floor spreading rates. The base level for the rivers on the New England upland surface was lowered and the coastal plain sediments covering the upper part of the old Fall Zone surface werre removed by erosion, as shown in Figs. Id and e, to form a cuesta at the south side of the new valley. As the coastal plain sediments were removed, the rivers from New England again flowed over the Fall Zone surface. According to McMaster and Ashraf (1973) they in part returned to their old, pre-

THE SEDIMENTARY SYSTEM OF LONG ISLAND SOUND

5

Cretaceous valleys. These authors find that the major valleys on the landward end of the Fall Zone surface join the valleys on the adjacent Fall Zone surface still covered by coastal plain sediments. River valleys on the Connecticut highlands usually follow the rock formations of lowest erosion resistance (Flint, 1963) and it may be supposed that the same is true offshore on the Fall Zone surface. The removal of the coastal plain sediments was probably accomplished by drainage into an easterly flowing river (Dana, 1890), which turned south near the present eastern end of Long Island (see Fig. 2 ) . Rivers crossing the Fall zone surfacejoined this stream from the north. Sediment from the south was delivered by streams flowing down the steep face of the coastal plain cuesta. Excavation of sediment to form the valley to the cuesta as mapped in Fig. 2 required denudation at a mean rate of about 8 x lo7 kg/yr. For comparison, about this much sediment is carried now by streams entering Long Island Sound west of the Connecticut River. One difficulty with this hypothesis for the denudation of the bedrock surface under the Sound is that there are deep basins north of the mapped edge of the cuesta that could not have been excavated by the proposed river system. These are shown by the shaded areas in Fig. 2 . They are thought to be formed on the Fall Zone surface but they may be closed off by outliers of Coastal Plain sediments. Some river valleys on the Fall Zone surface have been overdeepened by subsequent glacial erosion (the Quinnipiac River valley both south and north of New Haven contains basins up to 250 m deep, for example), but it is unlikely that the deep areas on the Fall Zone surface could have been formed this way since the shape of the basins is not elongated in the direction of ice flow. More detailed mapping of the topography of the Fall Zone surface under Long Island may help resolve this problem. The geological history of the region for the past several million years has been dominated by successive continental glaciations and their accompanying changes in sea level. Deposits formed during both pre-Wisconsinan and Wisconsinan glaciations have been identified on Martha’s Vineyard (Kaye, 1964), but there is only limited evidence of the location of the southern edges of any of these ice sheets other than the Wisconsinan. At many localities in Connecticut two distinct till layers are found. The older rests directly on bedrock, is compact and weathered at its upper surface, and is thought to be of early Wisconsinan-Illinoian age (J. P. Schafer, personal communication, 1978; Pessl and Schafer, 1968). Older till generally forms the cores of drumlins in southern New England (Schafer and Hartshorn, 1965). Several islands in Long Island Sound (including Falkner Island, 8 km offshore) are composed of older till (R. F. Flint, personal communication, 1975). On Long Island, the Mannetto and

THE SEDIMENTARY SYSTEM OF LONG ISLAND SOUND

7

Jameco formations were probably deposited by a pre-Wisconsin glaciation. Thus, it is likely that at least one ice sheet older than Late Wisconsinan covered the basin now occupied by Long Island Sound. Some of the Pleistocene deposits in the basin of the Sound may be pre-Wisconsin in age, but nothing definite can be said because no deep borings have been made. The late Wisconsinan ice sheet extended to the Ronkonkoma end moraine on Long Island (see Fig. 3a). The end moraines on Long Island, and their westward extensions, are not independently dated, but were formed before about 16,000 yr BP (Schafer, 1979). If the profile of the Antarctic ice sheet is fitted to an end point on the Ronkonkoma moraine, the ice thickness over southern Connecticut is found to be in excess of 1 km. Evidence of glacial erosion and deposition is found on the tops of the highest mountains throughout New England, proving a substantial ice thickness. No satisfactory estimate of the total amount of glacial erosion of southern New England has been made yet because the total volume of glacial drift has not been measured. Maps of the bedrock surface of Connecticut show that the erosion was concentrated along the valleys of existing rivers. In places these have been overdeepened by more than 100 m. Schafer and Hartshorn (1965) erstimate that perhaps as much as 20 m of rock was removed from the southern New England surface by glacial erosion. For rock of density 2.4 Mg/m3, this is erosion of 5.2 x lo4 kg/ m2. If it was removed in four episodes of glaciation each lasting 50,000 yr, the erosion rate was 0.26 kg/(mzyr), not very different from Menard's estimate of the long-term, mean denudation rate of the Appalachians. The deglaciation of southern New England is believed to have taken place by stagnation-zone retreat (Koteff, 1974) rather than by either melting down in place or by an active ice margin building recess moraines. According to this model a band of stagnant ice is supplied with rock debris by active ice advancing against its up-ice end. This debris is converted by the action of meltwater to ice-contact drift on and adjacent to the stagnant ice and to outwash deposits further away from the margin. A characteristic unit of stratified drift-a valley train-results. Small end moraines may form at the boundary between the live and stagnant ice (Schafer and Hartshorn, 1965). When the stagnation zone retreats to a new location, a new stratified drift unit is initiated. Former ice margins are thus marked by small end moraines or the heads of stratified drift units. The locations of the dated organic remains most useful in establishing the chronology of the deglaciation of the Long Island Sound region are shown in Fig. 3b. The most important of these is the date of 14,240 yr BP, the oldest from a series of dates on deposits in Rogers Lake shown

m

FIG.3(a). Contours show the depth in meters below sea level of the top surface of the glacial drift (principally ice contact drift and outwash) in the basin of Long Island Sound. End moraines are shown in solid black; the Elmhurst Moraine (EM) mapped by Newman (1977) is shown by the heavy dashed line. Dotted areas are shoreside deposits of outwash and ice contact drift. The abbreviations are: C1, Captain Islands; NI, Norwalk Islands: MM, Madison moraine; LM, Ledyard moraine; OSM, Old Saybrook moraine; CM, Charlestown moraine; HHM, Harbor Hill moraine; RM, Ronkonkoma moraine. Locations of end moraines in eastern Connecticut based on Flint and Gebert (1976); field work now in progress may show some revisions necessary (Schafer, 1979, also personal communication).

t4’030‘ (b)

5900 vr bs

-

FIG.3(b). Location of sources of dated organic material used to establish the chronology of deglaciation (crosses) and sea-level chang (squares). Depths are measured from mean high water (i.e.,marsh surface). Triangles mark the saddle points on the sills bounding the ba! of the central Sound. Abbreviations: TB, Totoket bog; RL, Rogers Lake; BI, Block Island; MS, Mattituck sill. The Bloom and Stuiver s level curve was established for location B-S. Dates based on Davis (1%5) with additional data from Schaffel (1971) and Newman (197 Other data sources are given in the text.

10

ROBERT B . GORDON

to be younger than all nearby till (Stuiver et al., 1963; Davis, 1969). Two lines of minor moraines, and their offshore extensions, have been identified in southern Connecticut by Flint and Gebert (1976). These are the Madison and Old Saybrook moraines shown in Fig. 3a. (The Ledyard moraine does not reach Long Island Sound, but may correlate with the Madison moraine.) They are interpreted as the boundary between live and dead ice. The Old Saybrook moraine is south of Rogers Lake and must have been formed, therefore, before 14,240 yr BP. The position and orientation of the moraine line suggests that all the central and eastern Sound must have been clear of ice by this date. Extensive sub-bottom acoustic reflection profiling through the eastern Sound has failed to reveal any trace of submerged or buried end moraine segments between the Harbor Hill moraine and the Old Saybrook moraine (J. A. Gebert, personal communication, 1976). Thus, the ice retreat across this part of the Long Island Sound basin must have been more nearly continuous than that across southern Connecticut. The Captain and Norwalk Islands (see Fig. 3a) were examined by Flint and Gebert and identified as end-moraine segments. They may correlate with the Lordship outwash and the Madison moraine (J. P. Schafer, personal communication, 1979). Additional moraine segments have been mapped on the western end of Long Island by Newman (1977). During retreat of the ice sheet across Long Island Sound large quantities of ice-contact drift and outwash sand must have been released by the melting ice. Such deposits can be studied in detail in southern Connecticut. Their extent along the coast is shown in Fig. 3a. Beneath Long Island Sound they form the principal horizon on which subsequent lacustrine and marine sediments were deposited. The ice-contact drift and outwash sand are good acoustic reflectors and have been mapped in some detail by means of acoustic reflection profiling (Grim et af., 1970; Bokuniewicz el al., 1976). The slopes of the outwash surface along the shore of central Long Island Sound and of the nearby valley trains are listed in Table I. Also listed in the table for comparison are the slopes of the Fall Zone surface and of the New England upland surface. All these slopes will have been increased subsequent to deglaciation by tilting of the land surface due to viscous rebound after removal of the ice load. An estimate of this increase in slope can be made from the tilt of the shorelines of glacial Lake Hitchcock, which occupied much of the valley of the Connecticut River and was drained about 10,700 yr BP (Flint, 1956). We will assume that the regional tilt due to rebound since that time is the same as the 0.8 x 10-’-rad tilt (Emerson, 1898; Jahns and Willard, 1942; Koteff, 1968) of the lake shorelines. Hence, the present slopes would have to be reduced by 0.8 x

11

THE SEDIMENTARY SYSTEM OF LONG ISLAND SOUND

TABLEI. MEANSLOPES OF OUTWASH SURFACE ALONG CENTRAL LONGISLANDSOUND SHORE AND NEARBY VALLEY TRAINS Surface Fall Zone surface (Flint, 1963) New England upland (based on Barrel], 1920) Lake Hitchcock shoreline (Emerson, 1898) Outwash surface (based on topographic map of the Guilford, Conn. quadrangle) Valley trains (lower ends where concave up): Mill River (based on Lougee, 1938, Plate XII) West River (based on Flint, 1971, Fig. 4) Farm River (based on Flint, 1964, Fig. 5) Quinnipiac River (based on Flint, 1965, Fig. 7) Submerged outwash sands: South of New Haven (based on Fig. 3 of Bokuniewicz ef nl., 1976) South of shore East Haven to Madison extending to 6.8 km , offshore (based on Fig. 6 of Bokuniewicz et ~ l . 1976)

Slope 9.4 x 4.4 0.82 3.8

(rad)

2.0 2.8 2.2 1.1 3.2 2.9

rad to find the slopes at the time of deglaciation. The present slope of the outwash surface on land is slightly greater than the slopes of the valley trains and the offshore extensions of the outwash. Although the slopes of the surfaces of the outwash sands on land and offshore are nearly the same where they have been compared in the eastcentral Sound region, the two surfaces are not continuous. Comparison of the elevations of these surfaces is complicated by the need to refer the marine surveys to the same datum as that used in mapping the land surface. Absolute elevation control was not maintained in the available bathymetric surveys and it is estimated that there may be an error of 1-2 m in matching the data of on-land and marine surveys. Figure 4 shows the configuration of the surface of outwash deposits near the shoreline at one location in Madison, Connecticut (C. Sullivan, personal communication, 1977). The projection of the onshore land surface seaward is 7.6 m above the surface of the now-submerged sands offshore. This difference is large compared to errors that may result from establishment of the survey datum. This elevation difference is interpreted as the amount of the outwash that has been removed by wave erosion during the rise in sea level that followed deglaciation. The eroded sand was probably incorporated in the marine mud that was being deposited simultaneously in the deeper water further offshore. On the basis of the evidencejust presented, the sand deposits extending from 6 to 10 km out from the Connecticut shore are interpreted as continuations of the onshore deposits of outwash sand and it is hypothesized that they were formed at the time when the ice margin was north of the

12

ROBERT B. GORDON

FIG.4. Cross section through glacial outwash sands (dotted) along the Connecticut shore (at longitude 72"34'W). Marine mud deposited on the outwash is shown shaded. (Based on data obtained by C. Sullivan, personal communication, 1977.)

Madison and Old Saybrook moraines. The volume of the outwash sand deposit can be estimated from the altitude and inclination of the Fall Zone and outwash surfaces. It is 2.1 x lo5 m3 per meter of shoreline (including that subsequently removed by marine erosion). The time for the ice to move across the Sound was not more than 4000 yr (and probably was much less). The outwash extends about one-quarter of the way across the Sound. If it were formed in 1000 yr and the source area for the rock debris were 100 km long, the requisite erosion rate is 3.5 kg/(m2 yr), which is very large compared to the rates discussed earlier. It could be reduced by perhaps a factor of 4 by allowing a longer formation time or a larger source area. For comparison, note that Boulton (1974) measured an abrasion rate on basalt of 2.7 kg/(m2 yr) under ice 40 m thick flowing at a speed of 9.6 m/yr. This suggests that the glacial erosion rates are either much higher during deglaciation than during glacial advance, or that the long-term, average glacial erosion rates for southern New England have been estimated too low. The former hypothesis is favored on the basis of Boulton's (1974) demonstration that glacial erosion rates are greatest for intermediate ice thickness. 3. SEALEVEL At the maximum of the late Wisconsinan glaciation, sea level was much reduced and the sea margin was well out on the continental shelf. The subsequent rise of the sea relative to the land determined the marine history of Long Island Sound. Local field data reveal only the more recent parts of the sea level history of the region, so reliance must be placed on

THE SEDIMENTARY SYSTEM OF LONG ISLAND SOUND

13

geophysical arguments, or historical data from other localities, to define the sea level in the Sound directly after deglaciation began. Changes in sea level are the result of several physical effects. Melting of the ice sheets increases the volume of water in the ocean basins. The volume of the basins changes as isostatic compensation adjusts the elevation of the sea floor to the increased water load. Gravitational attraction between the ice and the sea tilts the water surface. The elevation of the land surface changes in response to the removal of the ice load. Each of these effects must be known if a sea level curve is to be computed for a given locality. A self-consistent theory of sea level changes that takes into account these factors (except the local gravitational attraction) and allows for the spherical shape of the earth was developed by Cathles (1975). The problem has been studied further by Farrell and Clark (1976) and by Peltier and Andrews (19761, who reach similar conclusions to those of Cathles about the distribution of viscosity within the earth that will match the theory of glacial rebound to the available field data. In order to compute a sea level curve for Long Island Sound we use the meltwater curve obtained by Cathles from geological evidence for the extent of the ict sheets as a function of time during deglaciation and his Model I ocean basin volume adjustment. These combine to give eustatic sea level curve C in Fig. 5 . Morner’s (1969) eustatic curve M, based on the interpretation of extensive field data in Scandanavia, is shown for comparison. To construct a curve of land elevation, we use local sea level data and a physical model to interpolate between the available data points. Where the sediment supply is adequate (as it is in Long Island Sound), salt marsh surfaces grow up to the level of mean high water (McCaffrey and Thomson, this volume). Radiometrically dated salt marsh peat taken from borings therefore gives sea level at former times. The most extensive data of this type in the area are from the Hammock River Marsh (B-S in Fig. 3b); Bloom and Stuiver (1963) established a sea level curve extending back 7000 yr for this locality. Additional data points are available for the Quinnipiac River (Upson et al., 1964),the south pier of the Throgs Neck Bridge and a bore in Flushing Bay (Newman, 1977), and one deep boring made in Long Island Sound off Manhasset Neck (Schaffel, 1971). (Locations are shown in Fig. 3b.) There are no data for ages greater than 12,300 yr. To get a starting point for a land elevation curve, we use the total uplift expected for the latitude of Long Island Sound from Cathles’ model 2 curve for the total ultimate uplift expected. Curve R in Fig. 5 starts at this elevation. The solid part of the curve is constructed from the data of Bloom and Stuiver; the two branches show the likely error in the radiometric dating. It is likely that the initial portion of the land

14

ROBERT B. GORDON

c

1

I6

I2

4

8

rear

bp

FIG.5. The eustatic sea level curves proposed by Cathles (1979, C, and Morner (1969), M. compared with land elevation curves R derived as explained in the text. The curve R-25 is drawn 25 m below R .

uplift curve is influenced by the elastic deflection of the crust near the ice margin (Walcott, 1970). The forebulge caused by this deflection is expected to follow the retreating ice with little delay and to cause an initially rapid rise in the land surface. The dashed portion of curve R is estimated on the basis of the available data points referenced earlier, an elastic deflection at the ice margin of 10 m, and the slowest physically reasonable viscous rebound from an initial crustal depression of 85 m. The dash-dot curve is drawn with no allowance for elastic deflection. The present configuration of the surface of the glacial drift in Long Island Sound is a deep central basin bounded by sills on the east and west. The submergence history of the Sound depends on the elevations of the lowest points on these sills relative to the sea level curve. On the Mattituck sill (to the east) this elevation is now -25 m. Sand is now being transported from east to west across the Mattituck sill and it is possible that the sill is now at a higher elevation than it was immediately after retreat of the ice. (More detailed acoustic reflection profile studies of the internal structure of the sill may answer this question.) The lowest point on the sill to the west (which has not yet been surveyed in as much detail) is higher than - 20 m. (The locations of the saddle points on the eastern

THE SEDIMENTARY SYSTEM OF LONG ISLAND SOUND

15

and western sills are shown by the triangles in the map, Fig. 3b.) It is likely, therefore, that the sea entered the central part of the Sound from the east when the sill depth dropped below sea level. The curve R-25 in Fig. 5 shows the elevation of the lowest point on the Mattituck sill relative to sea level. Until 8000 yr BP it was probably above the sea surface and the central part of the Long Island Sound basin would have been occupied by a lake. Because of the steep rise of the sea surface elevation curve for this time, error in the estimate of the sill depth will have a relatively small effect on the estimated submergence data. However, the possibility of an earlier submergence, around 12,000 yr BP, cannot be ruled out with the data now available. It would certainly have occurred if there were no elastic forebulge. Such an earlier submergence may have been temporary. The western end of the Sound was submerged by 10,000-12,000 yr BP, as shown by Newman’s and Schaffel’s data, but the high ground separating the western and central parts of the Sound (western triangle in Fig. 3b) probably remained about sea level until after 8000 yr BP. Several water level recorders have been operated nearly continuously in Long Island Sound over the past 40 years. It should be possible to use their records to determine the submergence rate of the coast. Water level depends on the tide, the season of the year, on the passage of storm centers, local wind stress, and changes in the elevation of the recorder such as might be due to subsidence of the ground. It is generally considered that comparison of yearly averages eliminates all but the last of these effects from the record. [Kaye and Stuckey (1973) show how the tidal range is affected by the 18.6-yr tidal period and suggest that there is also an effect on mean sea level. If there is, it is very small.] The yearly mean sea level for New London, computed from the National Ocean Survey tide gauge records, is shown in Fig. 6. [It is generally similar to that published by Hicks and Crosby (1974).] For comparison, the mean sea level for the five winter months is also shown. The rise is about 3 mm/yr. (A much longer record is available for New York City and this shows a generally steady rise of sea level there since 1930.) If inferences about the rebound of the land surface or eustatic changes in sea level are to be drawn from these data, it is necessary to be sure that all influence of meteorological events on the elevation of the water surface has been removed by the averaging. Several aspects of the data in Fig. 6 indicate that this is not the case. The variability of the winter-months curve is greater than that of the yearly average curve. The rate of sea level rise for the winter months (3.2 mm/yr) is different than that for the yearly mean (3.0 mm/yr). Several of the high and low water levels are for years in which there as an unusually large amount of storm activity. The data on water level changes due to storms presented in the next article in this volume,

16

ROBERT B. GORDON

c -1

1-61

40-41

50-51

60-61

70-71

YEAR

FIG.6. Annual mean sea level (solid circles) and mean sea level for the winter months (November-March, open circles) computed from tide gauge records taken at New London.

p. 41, show that a small change in the proportion of offshore and onshore winds in a year can be expected to have an effect on the mean water level comparable to the amplitude of the variability of the mean sea level curve in Fig. 6. It is likely that meteorological events are not adequately eliminated from tide gauge records by yearly averaging and that changes in the annual mean water level may be due to changes in storm-track distributions as well as to changes in eustatic sea level and the elevation of the land surface. Brown (1978) has shown that there is a large discrepancy between the tilting of the east coast of the U. S. revealed by water level and geodetic surveys. Tide gauge data may not be as reliable an indicator of changes in the elevation of the land surface as are precise leveling measurements. The surfaces of salt marshes are generally considered to grow up to the elevation of mean high water (Chapman, 1960). Different species of salt marsh plants are tolerant to different amounts of immersion and so grow at different elevations on the marsh. Adams (1963) has shown that narrow elevation zones of marsh plant species occur along the North Carolina coast. A series oflevelingmeasurementswas done on a salt marsh surface in Stony Creek, Connecticut, to see if a similar relationship could be established for the marshes of Long Island Sound. The marsh studied is located near the tide gauge at the Yale Field Station (station 2773a in Vol. I1 of the Admiralty Tide Tables, where the tidal characteristics are sum-

THE SEDIMENTARY SYSTEM OF LONG ISLAND S O U N D

17

marked). Figure 7 shows the range of elevations in which each of the principal plant species on this marsh is found. There is some overlap of the ranges, but it is clear that the high marsh species mark well the elevation of mean high-water springs, which is also the average elevation of the largest extent of nearly flat marsh surface. This is well below the highest astronomical tide. The marsh surface is also flooded by storm tides. The amount of tidal immersion for each elevation range can be read off the diagram in Fig. 8, which is computed for the tidal constants for this station. If the marsh surface is keeping up with changes in sea level, there should be measurable changes in its elevation over a span of a few years. Harrison and Bloom (1977) describe how these elevation changes are detected by measuring the depth of burial of marker beds placed on the

I

--MHWS

--MHWN

-0

M SL

FIG.7. Ranges of elevation in which different plant species occur on a salt marsh near Guilford, Connecticut. Datum is mean sea level (MSL). The tidal ranges shown are the highest astronomical tide (HAT), mean high-water springs (MHWS). and mean high-water neaps (MHWN).

18

ROBERT B. GORDON

-2

0

6

12 IMMERSION TIME (hr)

18

FIG.8. Average immersion time as a function of elevation for the salt marsh of Fig. 7.

marsh surface each year. They find 10-yr average burial rates ranging from 2 mm/yr near the mouth of the Connecticut River to 6 mm/yr on the Connecticut shore of central Long Island Sound. These rates are larger than those revealed by other types of evidence. The marsh surface cannot grow above the level at which submersion by sea water prevents invasion by upland plant species. McCaffrey and Thomson (this volume) show that there is no self-compaction of the salt marsh peat. Thus, a burial rate in excess of the rate of rise of sea level cannot be sustained. Since the rate of peat formation along Long Island Sound is determined by the supply of nutrients (Steever et al., 1976) and, perhaps, by the amount of storm energy dissipation, the burial rate of the marker beds used by Harrison and Bloom may be variable on a time scale long compared to the 10-yr sampling interval, and that the observed high burial rates are a follow-on to a previous interval of low rates. The burial measurements should be continued over a longer time span to resolve this question. Sea level changes are the dominant factor influencing the evolution of salt marshes on the coast along Long Island Sound. The salt marsh surface grows upward with the rise of sea level, but the marsh can survive only if it receives protection from erosion of its seaward edge by waves. Thus, most marshes are found behind protective barriers and, if this protection is lost for any reason, destruction of the marsh by wave action follows. Figure 9 illustrates a marsh that is evolving in this way. Extensive salt marsh extends up the valleys of the rivers shown in the map, but the face

24

THE SEDIMENTARY SYSTEM OF LONG ISLAND SOUND

19

of the marsh exposed to the sea is being cut back by erosion due to waves. The marsh formed on the surface of an extensive plain of outwash sands bounded laterally by bedrock and end-moraine segments. It is inferred that the marsh once extended out to location B (bottom sampling shows submerged salt marsh peat between A and B) and that there was a protective barrier of sand extending from the end-moraine segment nearly to the bedrock outcrop, where the river mouth was located. Some additional protection may have been derived from another end-moraine segment now reduced by wave erosion to the boulder field shown on the map. Rise of sea level evidently caused these barriers to be overtopped and their remnants are the sand deposit seaward of B and the boulder field shown in the cross section and map of Fig. 9. It is estimated from

PEAT SAND END MORAINE BEDROCK BOULDER

4. 4. 4.

FIG.9. Map and section showing evolution of the salt marsh at Guiiford Harbor, Connecticut. When sea level was several meters lower, a protective barrier of sand extended westward past B and the marsh formed behind this barrier. When sea level rose above the barrier, the face of the marsh began to retreat under the attack of waves leaving the erosion surface between B and A. Retreat of the exposed marsh face is continuing while the marsh surface simultaneously grows upward with rising sea level.

20

R O B E R T B. GORDON

the sea level curve that this may have happened about 1000 years ago. In that time span the marsh face has retreated from B to A, where erosion is still active. Destruction of some unprotected marsh faces around the shore of the Sound at a comparable rate is expected as long as sea level continues to rise. 4. PHYSICAL OCEANOGRAPHY

Study of the tides of Long Island Sound began early in the nineteenth century and by 1932, when Le Lacheur and Sammons published their summary volume, a substantial body of data had been collected. It was recognized that the period of longitudinal oscillation of the Sound is close to 12.4 hr and that the tidal characteristics of the Sound are those of a resonant basin. Thus, the amplitude of the tidal height increases, and the speed of the tidal stream decreases, to the west; the greatest speed of the stream occurs when the water surface is nearly level and the time of high and low water is nearly synchronous throughout the Sound. Harmonic constants for the tide height and the speed of the tidal stream have been determined for a number of additional stations in the Sound since 1932 and we summarize these in Table 11. The dominant constituent is the lunar semidiurnal tide, M2. The amplitude of M 2 for the tide height increases by about a factor of 3 between the eastern and western ends of the Sound, while the amplitude of the stream decreases by almost as much. At the far-western end, shallow water components become relatively large (Bowman, 1976b). A calculation of the tidal exchange between successive sections across the axis of the Sound is given by Koppelman et al. (1976). A refined model of the tidal oscillation in Long Island Sound must allow for geostrophic and frictional forces. Rotary tides are observed in the center of the Sound, as illustrated by the current-meter record shown in Fig. 10. Each arrow in this diagram is the average velocity vector for a 20-min time interval. The anticlockwise rotation is in the direction expected for the effect of the earth’s rotation on a resonant basin (Doodson and Warburg, 1941). A theory of the damped resonant oscillation of the Sound with a linearized representation of the friction was published by Redfield (1950) and was further developed by Ippen and Harleman (1966). The average tidal dissipation calculated from Redfield’s model at mean tidal range is 460 MW. It is shown on p. 66 of the next article in this volume that an estimate of the dissipation based on direct measurements of water height and speed is 455 MW, that the Q-’ of the tidal oscillation is 0.32, and that the average specific dissipation is 0.060 W/mz. Direct measurements of the specific dissipation have been made for only a few

TABLE11. PRINCIPAL HARMONIC CONSTANTS OF THE TIDEAND TIDALSTREAM FOR LONGISLAND SOUND'

M2 Name

Lat . 41'21' 41"IO' 4176' 41"15' 41"17' 40"57' 41'10' 40°48'

0 1

g

H

(deg)

(m)

0.07 0.08

133 144

0.10

144 150

0.05 0.05 0.06 0.06 0.06

Long. Source' ~

New Londonb Plum Island Saybrook Jetty Hoadley Pt. New Havenb Port Jeffersonb Bridgeport Willetts Point

KI

s 2

72'06' 72"12' 72'21' 72'44' 72%' 73'05' 73"Il' 73"47'

(I) (I) (I) (1,2) (1,2) (I) (1)

(I)

~

273 287 307 319 325 328 32 1 334

~~

0.36 0.37 0.51 0.79 0.95

0.94 O.%

1.12

278 295 318 330 345 348 345 353

0.07 0.07 0.09 0.14 0.18 0.16 0. I6 0.20

104 111 104 I10 1I3 123 117 120

0.08 0.10 0.09 0.09 0.10

142 147 147

0.07 0.06

0.06

150

Stream

M2 g

Station name (73- 15) (72-2) (72-5) (EN-A)

Lat. 41%' 41W7' 41W' 41"OO'

Long. Source 72"32' 72"53' 72"58' 73"26'

(2) (2) (2) (3)

(deg) 47 55

KI

s 2

uo

(mdsec)

50

400 228 154

56

150

g

(deg) 66 66 61 74

uo

(mdsec) 71 40 27 34

g

(deg) 190

227 174 342

01

uo

(mrn/sec) 28 19 35 6

g

(deg) 220 267 214 93

uo

(mrnkec) 17 11

21 0.4

'References: ( I ) Admiralty Tide Tables, Vol. I1 (1978); (2) Gordon and Pilbeam (1975); (3) Bokuniewicz et a / . (1977). Tide gauge within harbor; may not be representative of conditions on the open coast. Notation: g is the phase angle and H and UOthe amplitudes of the tidal constituents; see the Admiralty Tide Tables for descriptions of these constituents.

22

ROBERT B. GORDON no

FIG.10. Velocity vectors recorded for successive 20-min time intervals by a current meter 2 m above the bottom near the geometrical center of Long Island Sound. The rotary character of the tide and the net drift of bottom water due to the estuarine circulation are shown.

other localities and are listed in Table 111; comparison shows that, while Long Island Sound is a tidally dominated estuary, it is one of intermediate specific dissipation. The tide advances up the Connecticut and the Housatonic Rivers as a progressive wave (Le Lacheur and Sammons, 1932). Tides and circulation in the Thames River are described by Tolderlund (1975). The East River is a tidal straight with most of its tidal prism derived from New York Harbor (Bowman, 1976a). Long Island Sound is connected to the sea by three passes at its eastern end and by the East River (through New York Harbor) at its western end. The principal source of fresh water entering the Sound is the Connecticut River, which enters near the eastern end, as shown in Fig. 11. Thus the Sound does not have the conventional configuration of an estuary with TABLEIll. SPECIFIC TIDALDISSIPATION Chandeleur Sound (Hart and Murray, 1978) Narragansett Bay (Levine and Kenyon, 1975) Long Island Sound Irish Sea (Taylor, 1919) Bay of Fundy (McLellan, 1958)

0.006 (Wlm’) 0.03 0.06 1.3 1.9

THE SEDIMENTARY SYSTEM OF LONG ISLAND SOUND

T

TI

23

1.

FIG.1 I . Entry of fresh water (excluding that from the Hudson River) and sewage effluent into Long Island Sound, by sections. (After Bowman, 1975.)

a river at its head end. Nevertheless, there is a well-developed estuarine circulation with less-saline water flowing eastward at the surface and more-saline water flowing westward at the bottom. This was first shown by Riley (1952, 1956, 1967) in his comprehensive study of the physical oceanography of the Sound. The water flow due to the estuarine circulation deduced from the salt balance is 25,000 m3/sec in the eastern Sound and 3500 m3/sec in the western end. Although it is a tidal strait rather than a river, the East River plays an important role in the maintenance of the estuarine circulation in the Sound because there is a net export of salt at the average rate of 1.2 x lo4 kg/sec through it (Bowman, 1975). This

24

ROBERT B. GORDON

maintains the longitudinal salinity gradient throughout the Sound. A small, additional residual flow is due to nonlinear tidal effects (Ianiello, 1977). Wilson (1976) has used a one-dimensional model to compute the gravitational circulation resulting from the density gradient determined by the salinity and temperature distributions and the friction due to tidal stream turbulence. He finds good agreement with the circulation deduced by Riley. The presence of this circulation is confirmed by drifter returns (Gross and Bumpus, 1972; Paskausky and Murphy, 1976) and currentmeter observations (Gordon and Pilbeam, 1975). However, Paskausky and Murphy’s inference that bottom-water flow does not penetrate into the central part of the Sound in the summer is not confirmed by other sources of data. Layers of surface and bottom water separated by a well-defined interface are present in the central part of Long Island Sound during much of the year. This interface is shown by the temperature and salinity profiles reproduced in Fig. 12. The Sound is large enough that lateral effects in the circulation are quite important. For example, the boundary between surface and bottom water is tilted upward to the north in response to the Coriolis force, while in the shallow water along both shores tidal mixing is strong enough to eliminate the interface between surface and bottom water. Bottom water flows laterally into these shoreside mixing zones, causing upwelling along the north and south coasts of the Sound (Gordon and Pilbeam, 1975). Sharp frontal boundaries between water masses are also present in the waters of Long Island Sound. One of the most persistent forms between river and sound water off the mouth of the Connecticut River during periods of high discharge or ebb flow (Garvine, 1974, 1975, 1977; Garvine and Monk, 1974). [At low river discharge salt penetrates several kilometers up the Connecticut River, which is itself a small estuary (Meade, 1966).] Bowman and Esaias (1977) have found a front separating the waters of Smithtown Bay, Long Island, from the waters of the central Sound. They suggest that instabilities at the front are responsible for the periodic injection of patches of high concentrations of phytoplankton into the Sound. In addition to receiving fresh water, large quantities of wastes and sediment are inserted into the Sound. The injection of sewage wastes and, for comparison, the inflow of fresh water, along the axis of the Sound is shown in Fig. 11 (after Bowman, 1975). New York City is the dominant source of wastes and, because they are inserted where the natural circulation is weakest, they represent the main cause of environmental degradation of the Sound at the present time. Bowman (1976a) has shown how this problem could be resolved by the construction of tidal locks across the East River. The principal source of sediment entering the

THE SEDIMENTARY SYSTEM OF LONG ISLAND SOUND

25

FIG.12. Temperature and salinity gradients measured near the geometrical center of Long Island Sound in the summer. The elevation above the bottom is Z . The bottom water layer is colder and more saline.

Sound is the Connecticut River, which drains most of central New England and enters the Sound near its eastern end. This is an area of strong mixing (because of a high level of tide-stream turbulence) and sediment brought down the river is mixed into the bottom water that subsequently flows into the central Sound. The extensive deposits of silt-clay sediment in the central basin of the Sound are formed, therefore, as a consequence of the estuarine circulation. 5. SEDIMENTATION

The mechanics of sediment transport and deposition in Long Island Sound are discussed in some detail in subsequent articles in this volume. Only an overview of the principal results is presented here.

26

ROBERT B. GORDON

The sediments of the Sound consist principally of glacial sands and marine muds. A margin of sand is found along both the Connecticut and the Long Island shore except for some localities where the shore consists of bedrock (see Fig. 3). The marine mud, deposited since the sea reentered the Sound, occupies most of the central basin. If lake sediments exist in the Sound, they are now buried under deposits of the marine mud. Long cores that would permit their detection have never been collected in the Sound. Throughout most of the Sound the lateral boundary between the mud and the sand is quite sharp, as shown on the map on p. 97 of this volume, but toward the eastern end the transition from muddy to sandy bottom is gradual. This is because sand carried into the central Sound by the combined action of the tidal stream and the estuarine circulation is being incorporated into the accumulating marine mud. A quantitative model of this transition zone is presented later in this volume. The volume of the marine sediment in Long Island Sound has been measured by acoustic-reflection profiling (Bokuniewicz et al., 1976). The method works well in the Sound because the upper surface of the glacial drift (the “sub-bottom”) is a strong acoustic reflector and the marine mud is sufficiently gas free to be acoustically transparent. Illustrations of acoustic-reflection profiles showing deposits of marine mud over outwash sand and over a submerged end-moraine segment are shown in Fig. 13. All available reflection profile data were used to construct the contour map of the surface of the glacial sands shown in Fig. 3a. The volume of silt-clay sediment contained in the marine mud is estimated by subtracting the volume of the contained sand (based on the analysis of the sand content of cores taken throughout the Sound) from the measured total volume of sediment. The volume of silt-clay sediment in the Sound is 5.3 x lo9 m3 and, since the density is about 0.8 Mg/m3, the mass is about 4.2 x 1OI2 kg. A small amount of sediment in the Sound is of biogenic origin and some originates from erosion of the north shore of Long Island, but most is supplied by the rivers that drain into the Sound. The Connecticut River is the principal source of fresh water (71% of the drainage basin area) and since all the rivers entering the Sound flow over generally similar terrain, it is likely that the sediment contribution of each is approximately proportional to its discharge. The present rate of supply of sediment to the Sound per unit area of existing mud bottom calculated from available data on the sediment yield of the rivers entering the Sound is 0.26 kg/(m2yr). The sediment yield of the Connecticut River is low compared to that of most rivers because much of its drainage basin is on glacial till, which is very erosion resistant. The principal source of sediment entering the river is bank collapse where the river flows over the beds of old glacial

FIG. 13. Structure of the bottom of Long Island Sound revealed by acoustic reflection profiles made with 7-kHzacoustic pulses. (Upper echo is produced by a 200-kHz echo sounder.) (a) Section of end moraine capped by boulders and almost buried by marine mud. (b) Thick deposit of marine mud in central Long Island Sound on top of outwash sand with reflector above thought to be surface of lacustrine deposits. (c) Sand-to-mud transition zone in central Long lsland Sound. In all records each division on the vertical scale is 600 mm.

FIG. 13b. See p. 27 for legend.

M

FIG.13c. See p. 27 for legend.

30

ROBERT B. GORDON

lakes (Gordon, 1979). This source is insensitive to land use and it is expected that the sediment yield has not changed much during the marine regime of the Sound. We will assume that the sediment supply rate over the past 8000 yr is about the same as the present rate, 0.26 kg/(m2 yr). The average, long-term mean rate of sedimentation in the Sound computed from the amount of accumulated silt-clay sediment is 0.29 kgl(m2 yr). It follows that the trapping efficiency of the Sound for sediment delivered by rivers must be nearly loo%, since sediment is accumulating at a rate slightly greater than the rate at which it is being delivered to the Sound by the rivers. Further evidence of high trapping efficiency is obtained from the recently completed radiometric determination of the sedimentation rate in the Sound by Benoit et al. (1979). The sediment accumulation rate calculated from their data is -0.4 kg (mZ yr), which exceeds the rate at which sediment is supplied by the rivers entering the Sound. The principal sources of error in the determination of the trapping efficiency of the Sound are the possibility that shore erosion may contribute more sediment to the Sound than our estimate and the possibility that the sea entered the Sound at an earlier date than that inferred from the data presented in Fig. 5 . The sediment-trapping efficiency of Long Island Sound is high because of the combined effects of several factors. First, the volumetric capacity of the Sound to store sediment is large compared to the sediment yield of the rivers entering it. The large capacity is a consequence of the deepening of the Sound basin by both preglacial erosion and by glacial overdeepening. The low sediment yield of the rivers is a consequence of the erosion-resistant character of the glaciated terrain of central New England. For comparison, estuaries along the southeast coast of the U.S. are much less likely to have such a large sediment storage capacity because of the high sediment yield of their drainage basins. They were filled nearly to capacity as they were flooded by the postglacial rise of sea level. An important factor in keeping the sediment storage capacity of Long Island Sound large is that sea level is rising faster than sediment is accumulating in the estuary. Where the rise of sea level is small, or the coast is emergent, the storage capacity of an estuary is much more likely to be exceeded as, for example, in the estuaries of the Lune and the Mersey on the west coast of the U. K. Finally, the specific dissipation in the Sound (0.06 W/ m2) is low enough to permit accumulation of sediment on the existing mud bottom. In estuaries having a much higher specific dissipation, such as the Bristol Channel of the U. K. (-6 W/m2),delivered sediment is not retained on the bottom. Sediment trapped in Long Island Sound is not necessarily isolated from contact with the ambient water or the waters of the continental shelf. The

THE SEDIMENTARY SYSTEM OF LONG ISLAND SOUND

31

specific dissipation in the Sound is large enough to regularly resuspend some of the trapped sediment, and, as is shown on p. 93, the amount of resuspension may be substantially increased during storms. The material resuspended in the water column consists of pelletized silt-clay sediment produced by the benthic animals of the Sound. This pelletized material forms a mantle about 10-mm thick on top of cohesive mud deposits. The amount of resuspension depends both on the level of the specific dissipation and on the properties of the pelletized material. The critical erosion velocity of the pellets shows seasonal variation in response to the seasonal changes in the production of organic binding by bacteria in the sediments (Rhoads et al., 1979). The data available on both the properties and the excitation of the sediment are not as yet adequate to permit a quantitative description of the amount of resuspension. Observations show that the mantle of pelletized material is fully excited into the water column only infrequently. New silt-clay sediment entering the Sound is processed to pellets by the benthic animals almost at once. The residence time of the silt-clay sediment in the pelletized mantle is about 10 years (see p. 98), after which it is converted into cohesive mud (or muddy sand) that is not subject to resuspension into the water column and is effectively isolated from the ambient water. The characteristics of the sedimentary regime of the Sound have several important consequences for the management of its marine resources. The first is that the sedimentary system is not likely to be much altered by any of the ordinary range of engineering works-dredging, construction of dams on the tributary rivers, or a bridge over the Sound-likely to be undertaken in the area. However, construction of tidal locks across the East River, as suggested by Bowman, would be extraordinary because the resultant alteration of the salinity of the Sound would change the benthic animal population responsible for the maintenance of the pelletized layer of sediment on the bottom. Materials entering the Sound attached to silt-clay sediment particles (such as heavy metals) are likely to be retained in the Sound and not exported to the sea. The Sound has a very large capacity to store these materials, but during the first 10 years or so of storage there will be frequent direct contact with the ambient water because of resuspension. Resuspended sediment may intermix, through tidal exchange, with suspended sediment in the waters of the continental shelf. Qualitative data are lacking. Akpati (1974) has examined the composition of sediment in Fishers Island Sound (which is actually the eastern extremity of Long Island Sound) and finds that material from both the adjacent land and from offshore sources is present, although in what proportions remains unknown. The pelletized mantle subject to resuspension

32

ROBERT B. GORDON

in the deep water of the Sound is also a reservoir of sediment available for accumulation in dredged channels. Sawhney and Frink (1979) have shown that the clay mineralogy of the sediments in New Haven Harbor is like that of the sediment reservoir in the central Sound and unlike that of the sediment carried by the rivers entering the harbor. Resuspended sediment is exchanged with that in harbors by the lateral circulation of the Sound and the local estuarine circulation of the harbor. Thus, continued sediment accumulation in the harbors is expected as long as the pelletized mantle of sediment in the central Sound exists. Once the silt-clay sediment is incorporated into the permanent mud bottom of the Sound, subsequent contact with the water above will be limited to molecular diffusion through the interstitial water. Contaminated sediment released from a point source in the Sound will be rapidly dispersed throughout the mantle of pelletized material by tidal mixing, so that large, local concentration gradients are not expected to persist around sediment disposal sites (Bokuniewicz and Gordon, 1979). However, regional gradients in the heavy metal content of the sediment are present. The capacity of the Sound to store and retain contaminants attached to sediment particles is a consequence of its high trapping efficiency. Determinations of the trapping efficiency of other estuaries are not generally available, although for some, such as the Bristol Channel, it is near zero, so that comparison of this aspect of the Sound with other environments is not yet possible. In the context of the longer span of geological time, the present sedimentary processes in the Sound are anomalous. Before glaciation, sediment was removed from the land surface at average rates comparable to the present-day sediment yield of drainage basins of moderate relief and elevation. The debris produced was transported to the continental shelf and margins. During the late Cenozoic glaciations there were periods of rapid denudation in the land surface, with rates which may have been as high as 3.5 kg/(m2yr). The sedimentation rate in glacial Lake Hitchcock, for example, was at least three times greater than the present rate in Long Island Sound (Gordon, 1979). In the interglacial periods the denudation rate became very small because the debris produced during the periods of rapid erosion was not cleared from the land surface. The present is one of those periods. Most of the sediment transported by the river system draining into Long Island Sound is debris produced by the late Wisconsin ice and left behind as the ice retreated. Most of this material is quite erosion-resistantand the regional denudation rate is small in consequence. The sediment yield of the streams draining into Long Island Sound is remarkably small by any standard. Because of the high trapping efficiency of the Sound, this material is not reaching the sea and probably will not

THE SEDIMENTARY SYSTEM OF LONG ISLAND SOUND

33

until sea level begins to fall at some time in the future. Because of its depth and configuration, Long Island Sound will probably be one of the last estuaries on the east coast of North America to lose its trapping efficiency during a fall of sea level. 6. FURTHER RESEARCH

A number of problems have arisen in our discussion of the evolution and operation of the sedimentary system of Long Island Sound that remain unresolved. Further field work and analysis are needed on these. We summarize below some of the more interesting ones. The origin of many features of the drainage system of Southern New England remains obscure. A record of an earlier stage of the development of this system is preserved beneath the sediments in the Sound. This record could be revealed by acoustic reflection profiling and used to establish the relationship of the ancient to the present drainage system. It may be possible to use microtopography determined from the reflection profiles to identify the continuation of bedrock formations offshore beneath the sediments in the Sound; this would permit the determination of the relation of the ancient drainage pattern to lithology. It would be of particular interest to trace the path of the ancestral Housatonic River across the Sound basin and to find where it passes under Long Island. These data should help resolve the question of how the overdeepened basins on the bedrock surface beneath the Sound were formed. It is likely that the older till found in Connecticut extends underneath the Sound; it may be continuous with some of the tills found on Long Island. Because it is highly compacted, the older till is expected to have a relatively high sound speed. It should show up as a distinct acoustic horizon and it should be possible to trace it under the Sound by acoustic methods. Data on the distribution of the older till would be most helpful in the interpretation of the early glacial history of the Long Island Sound basin. The thickness of marine sediments in the Sound has been measured by reflection profiling. A corresponding set of data on the thickness of the glacial drift could be used to determine glacial erosion rates in Southern New England. Since acoustic penetration to the rockhead can be attained throughout the Sound, there should be no technical difficulty in obtaining the requisite data. It should be possible to differentiate between the compact tills and outwash materials on the basis of their acoustic properties. Some horizons can be sampled by following them to outcrop, but some number of deep drill holds will undoubtedly be required to properly iden-

34

ROBERT B. GORDON

tify other horizons. A related problem that can be approached by the same methods is the internal structure of the Mattituck sill. It is important to determine whether or not the sill depth has been built up by the accretion of tidally transported sand or has remained nearly unchanged since deglaciation, since the sill depth determines the date of the return of the sea to the Sound. Reflection-profiler surveys should be made to better locate the sill at the western end of the Long Island Sound basin and to determine whether or not its altitude has been altered by marine processes. The history of sea level change in the Sound is reasonably well established for the past 5000 years, but there are substantial uncertainties about it before that time. The recovery of organic remains suitable for dating from deep bores within the Sound are needed to resolve this question. Because of its particular geographical situation, the Sound is an ideal place to test theories about rebound of the crust near an ice margin. It is likely that a description of regional rebound could be assembled by utilizing, in addition to the data from the Long Island Sound area, published information on the tilting of glacial lakes (such as Lake Hitchcock). Additional information could be obtained from careful leveling to obtain the inclination of the lake bottoms; excellent exposures are available in Southern New England. These data may also help to resolve the apparent discrepancy between the relative uplift rates east and west of the Connecticut Valley revealed by sea level and geodetic data. It is particularly important to determine whether or not there was appreciable elastic rebound due to a forebulge at the margin of the retreating ice sheet since, if this rapid rebound was small, the marine history deduced for the Sound will require substantial modification. The evolution of salt marshes along the Connecticut and Long Island shores continues to be of practical as well as academic concern; much useful data could be obtained at relatively low cost by continuing marker bed experiments (such as those initiated by Bloom), by assembling historical data on the geography of the marshes, and by making repeated surveys of the marsh margins. It would also be of interest to determine the absolute elevation of marsh surfaces with respect to mean sea level and the tidal range (such as presented in Fig. 7) for localities at the eastern and western ends of the Sound. The principal uncertainty in the determination of the amount of sediment entering the Sound arises from a lack of information about the composition of the materials being eroded from the north shore or Long Island. Field work and sampling along this shore are needed. Before it is incorporated into the permanent bottom of the Sound, sediment is available for resuspension for a number of years. The amount and frequency of resuspension depends on both physical and biological factors, which can

THE SEDIMENTARY SYSTEM OF LONG ISLAND SOUND

35

be studied individually in the laboratory. However, there are no long-term data of sufficiently high sampling density to permit description of the amount of resuspension of the muddy bottom that takes place in the Sound. The physical and biological factors that control the susceptibility of muddy sediment in the Sound to resuspension are sufficiently complex, and their interaction sufficiently uncertain, that one could not have confidence in an evaluation of the stability of the muddy bottom based on laboratory data alone; good quality field observations are necessary. Resuspended sediment is available for exchange with sediment in Block Island Sound or the continental shelf, but the amount of the intermixingwhich is important in determining how efficiently the Sound traps effluent materials attached to the sediment particles-is not known at all. Extensive observations of sediment concentrations and water movements in the eastern end of the Sound would be needed to answer this question. Available data show that Long Island Sound retains nearly all of the sediment delivered to it, its trapping efficiency is near 100%. The factors responsible for this are the size and depth of the basin, the magnitude of the specific energy dissipation, and the rapid biological processing of the sediment delivered by the rivers. If this identification of the factors responsible for the high trapping efficiency is correct, it should be possible to use an evaluation of these same factors to determine the trapping efficiency of other estuaries. Measurements of the sediment mass balance of other representative estuaries is needed to test these ideas. ACKNOWLEDGMENT 1 thank Henry Bokuniewicz, Walter Newman, and J. P. Schafer for valuable comments on topics discussed in this chapter.

REFER E N cE s Adams, D. A. (1963). Factors influencing vascular plant zonation in North Carolina salt marshes. Ecology 44,445-456. Akpati, B. N. (1974). Mineral composition and sediments in eastern Long Island Sound, New York. Marit. Sediments 10, 19-30. Barrel], J. (1920). The Piedmont terraces of the northern Appalachians. Am. J . Sci. [4] 49, 227-258. 327-362. 407-428. Benoit, G. J., Turekian, K. K.. and Benninger, L. K. (1979). Radiocarbon dating of a core from Long Island Sound. Estiiarine Coastal Mar. Sci. 9, 171-180. Bloom, A. J., and Stuiver, M. (1963). Submergence of the Connecticut coast. Science 139, 332-334. Bokuniewicz, H. J., and Gordon, R. B. (1979). Containment of particulate wastes at openwater disposal sites. In “Ocean Dumping and Marine Pollution” (H. D. Palmer and M. G. Gross, eds.), pp. 109-129. Dowden, Hutchinson L Ross, Inc., Stroudsburg, Pennsylvania.

36

ROBERT B. GORDON

Bokuniewicz, H. J., Gebert, J., and Gordon, R. B. (1976). Sediment mass balance of a large estuary, Long Island Sound. Estuarine Coastal Mar. Sci. 4, 523-536. Bokuniewicz, H. J., Dowling, M., Gebert, J., Gordon, R., Kaminsky, P., Pilbeam, C., and Tuttle, C., (1977). “Aquatic Disposal Field Investigations Eatons Neck Disposal Site Long Island Sound Appendix A: Investigation of the Hydraulic Regime and the Physical Characteristics of Bottom Sedimentation,” Tech. Rep. D-77-6. U.S. Army Engineer Waterways Experiment Station, Vicksburg, Mississippi. Boulton, G. S. (1974). Processes and patterns of glacial erosion. I n “Glacial Geomorphology” (D. R. Coates, ed.), pp. 41-87, State University of New York at Binghamton. Bowman, M. J. (1975). Pollution prediction model for Long Island Sound. Proc. Ocean Eng., 3rd, 1975 pp. 1084-1103. Bowman, M. J. (1976a). Tidal locks across the East River: An engineering solution to the rehabilitation of Western Long Island Sound. I n “Estuarine Processes” (M. Wiley, ed.), Vol. 1, pp. 28-43. Academic Press, New York. Bowman, M. J. (1976b). The tides of the East River, New York. J. Geophys. Res. 81, 1609- 1616.

Bowman, M. J., and Esaias, W. E. (1977). Coastal jets, fronts and phytoplankton patchiness. 8th Liege Colloq. Ocean Hydrodyn., 1977 pp. 255-268. Brown, L. D. (1978). Recent vertical crustal movement along the coast of the United States. Tectonophysics 44, 205-23 1 . Cathles, L. M., 111 (1975). “The Viscosity of the Earth’s Mantle.” Princeton Univ. Press, Princeton, New Jersey. Chapman, V. J. (1960). “Salt Marshes and Salt Deserts of the World.” Wiley (Interscience), New York. Dana, J. D. (1890). Long Island Sound in the Quaternary era, with observations on the submarine Hudson River channel. Am. J . Sci. [3]40,425-437. Davis, M. B. (1%5). Phytogeography and palynology of the northeastern United States. In “The Quaternary of the United States” (H. E. Wright, Jr. and D. G. Frey, eds.), pp. 377-401. Princeton Univ. Press, Princeton, New Jersey. Davis, M. B. (1969). Climatic changes in southern Connecticut recorded by pollen deposition at Rogers Lake. Ecology 50,409-422. Doodson, A. T., and Warburg, H. D. (1941). “Admiralty Manual of Tides.” HM Stationery Ofice, London. Dyer, R. R. (3973). “Estuaries, A Physical Introduction.” Wiley. New York. Emerson, B. K. (1898). Geology of Old Hampshire County, Massachusetts. U.S . Geol. Surv., Monogr. 29, 1-790. Farrell, W. E., and Clark, J. A. (1976). On postglacial sea level. Geophys. J. 46, 647-667. Flint, R. F. (1956). New radiocarbon dates and late-Pleistocene stratigraphy. Am. J. Sci. [5] 254, 265-287.

Flint, R. F. (1%3). Altitude, lithology and the fall zone in Connecticut. J. Geol. 71,683-697. Flint, R. F. (1964). The surficial geology of the Branford Quadrangle. Conn., State Geol. Nat. Hist. Surv., Quadrangle Rep. 14. Flint, R. F. (1965). The surficial geology of the New Haven and Woodmont Quadrangles. Conn., State Geol. Nat. Hist. Surv., Quadrangle Rep. 18. Flint, R. F. (1971). The surficial geology of the Guilford and Clinton Quadrangles. Conn., State Geol. Nat. Hist. Surv., Quadrangle Rep. 28. Flint, R. F., and Gebert, J. A. (1976). Latest Laurentide ice sheet: New evidence from southern New England. Geol. SOC.Am. Buli. 87, 182-188. Garvine, R. W. (1974). Physical features of the Connecticut River outflow during high discharge. J . Geophys. Res. 79, 831-846.

THE SEDIMENTARY SYSTEM OF LONG ISLAND SOUND

37

Garvine, R. W. (1975). The distribution of temperature and salinity in the Connecticut River estuary. J. Geophys. Res. 80, 1176-1 182. Garvine, R. W. (1977). Observations of the motion field of the Connecticut River plume. J. Geophys. Res. 82, 441-454. Garvine, R. W., and Monk, J. D. (1974). Frontal structure of a river plume. J. Geophys. Res. 79, 2251-2259. Gordon, R. B. (1979). Denudation rate of central New England determined from estuarine circulation. Am. J. Sci. 279, 632-642. Gordon, R. B., and Pilbeam, C. C. (1975). Circulation in central Long Island Sound. J . Geophys. Res. 80, 414-422. Grim, M. S., Drake, C. L., and Heirtzler, J. R. (1970). Sub-bottom study of Long Island Sound. Geol. SOC.Am. Bull. 81,649-666. Gross, M. G., and Bumpus, D. F. (1%9). Residual drift of near-bottom waters in Long Island Sound. Limnol. Oceanogr. 17,636-638. Hamson, E. Z., and Bloom, A. L. (1977). Sedimentation rate on tidal salt marshes in Connecticut. J. Sediment. Petrol. 47, 1484-1490. Hart, W. E., and Murray, S. P. (1978). Energy balance and wind effects in a shallow sound. J. Geophys. Res. 83, 4097-4106. Hicks, S. D., and Crosby, J. E. (1974). Trends and variability of yearly mean sea level. NOAA Tech. Mem. NOS 13, 1-14. laniello, J. (1977). Nonlinearly induced residual currents in tidally dominated estuaries. Ph.D. Thesis, University of Connecticut, Storrs. Ippen. A. T., and Harleman. D. R. F. (1966).Tidal dynamics of estuaries. In “Estuary and Coastline Hydrodynamics” (A. T. Ippen, ed.), pp. 493-545. McGraw-Hill, New York. Jahns, R. H., and Willard, M. E. (1942). Late Pleistocene and Recent deposits in the Connecticut Valley, Massachusetts. Am. J. Sci. [5] 240, 161-191, 265-287. Johnson, D. (1931). “Stream Sculpture on the Atlantic Slope.” Columbia Univ. Press, New York. Kaye, C. A. (1964). Outline of Pleistocene geology of Martha’s Vineyard, Massachusetts. U.S . , Geol. Surv., Prof. Pap. 401-C. Kaye, C. A., and Stuckey, G . W. (1973). Nodal tidal cycle of 18.6 Yr. Geology 1, 141-144. Koppelman, L. E., Weyl, P. K., Gross, M. G., and Davies, D. S. (1976). “The Urban Sea: Long Island Sound.” Praeger, New York. Koteff, C. (1968). Postglacial tilt in Southern New England. Geol. SOC. Am., Spec. Pap. 101 (abstr.). Koteff, C. (1974). The morphological sequence concept and deglaciation of southern New England. In “Glacial Geomorphology” (D. R. Coates, ed.), pp. 121-144. State University of New York at Binghamton. Le Lecheur. E. A., and Sammons. J. C. (1932). Tides and currents in Long lsland Sound and Block Island Sound. U.S . Coast Geodetic Surv., Spec. Publ. 174. Levine, E. R., and Kenyon, K. E. (1975). The tidal energetics of Narragansett Bay. J. Geophys. Res. 80, 1683-1688. Lougee, R. J. (1938). Physiography of the Quinnipiac-Farmington Lowland in Connecticut. Colby Coll. Monog. No. 7. pp. 1-64. McLellan, H. J. (1958). Energy consideration in the Bay of Fundy System. J. Fish. Res. Board Can. 15, 115-134. McMaster, R. L., and Ashraf, A. (1973). Subbottom basement drainage system of inner continental shelf off southern New England. Geol. SOC.Am. Bull. 84, 187-190. Mathews, W. H.(1975). Cenozoic erosion and erosion surfaces of eastern North America. Am. J . Sci. [5] 275, 818-824. Meade, R. H. (1966). Salinity variations in the Connecticut River. Water Res. 2, 567-579.

38

ROBERT 9. GORDON

Menard, H. W. (1961). Some rates of regional erosion. J. Geol. 69, 154-161. Morner, N. (1969). The late Quaternary history of the Kattegatt Sea and the Swedish west coast, deglaciation, shoreline displacement, chronology, isostasy and eustasy. Arsb., Sver. Geol. Unders., Ser. C63, No. 3, 404-453. Newman, W. S. (1977). Late Quaternary paleoenvironmental reconstruction: Some considerations from northwestern Long Island, New York. Ann. N . Y. Acad. Sci. 288, 545-570. Oldale, R. N., and Uchupi, E. (1970). The glaciated shelf of northeastern United States. U . S . , Geol. Surv., Prof. Pap. 700, B167-8173. Paskausky, D. F., and Murphy, D. L. (1976). Seasonal variation of residual drift in Long Island Sound. Estuarine Coastal Mar. Sci. 4, 413-522. Peltier, W. R., and Andrews, J . T. (1976). Glacial-isostatic adjustment. 1. The forward problem. Geophys. J . 46,605-646. Pessl, F., and Schafer, J. P. (1968). Two-till problem in Naugatuck-Torrington area, Western Connecticut. In “Guidebook for Fieldtrips in Connecticut” (P. M. Orville, ed.), Guidebook No. 2. State Geological and Natural History Survey of Connecticut, Hartford. Pitman, W. C., 111 (1978). Relationships between eustacy and stratigraphic sequences on passive margins. Geol. SOC.Am. Bull. 89, 1389-1403. Redfield, A. C. (1950). The analysis of tidal phenomena in narrow embayments. Pap. Phys. Oceanogr. Meteorol. 11, 1-35. Rhoads, D. C., Yingst, J. Y., and Ullman. W. J. (1979). Seafloor stability in central Long Island Sound. Part I. Temporal changes in the erodibility of fine-grained sediment. In “Estuarine Interactions,” (M. L. Wiley, ed.), pp. 221-244. Academic Press, New York. Schafer, J. P. (1979). The late Wisconsinan Laurentide ice sheet in New England. Geol. SOC. Am., Abstr. Programs 11, 52. Riley, G. A. (1956). Oceanography of Long Island Sound 1952-1954. 11. Physical oceanography. Bull. Bingham Oceanogr. Collect. 15, 15-46. Riley, G. A. (1967). Aspects of oceanography of Long Island Sound. 11. Transport and mixing processes in Long Island Sound. Bull. Bingharn Oceanogr. Collect. 19, 35-61. Sawhney, B. L., and Frink, C. R. (1979). Clay minerals as indicators of sediment source in tidal estuaries of Long Island Sound. C/ays, Clay Miner. (in press). Schafer, J. P. (1979). The late Wisconsinan Laurentide ice sheet in New England. Geol. SOC. Am., Absir. Programs 11,52. Schafer, J. P., and Hartshorn, J. H. (1965). The Quaternary of New England. I n “The Quaternary of the United States” (H. E. Wright, Jr. and D. G. Frey, eds.), pp. 113-128. Princeton Univ. Press, Princeton, New Jersey. Schaffel, S. (1971). Reconstruction of late glacial and post-glacial events in Long Island Sound. Ph.D. Thesis, New York University. Sharp, H. S. (1929). The physical history of the Connecticut shoreline. Conn., Stare Geol. Nut. Hisi. Surv. Bull. 46, 1-97. Steever, E. 2..Warren, R. S., and Niering, W. A. (1976). Tidal energy subsidy and standing crop production of Spartina alterniflora. Estuarine Coasial Mar. Sci. 4,473-478. Stuiver, M., Deevey, E. S., Jr., and Rouse J. (1963). Yale natural radiocarbon measurements. VIII. Radiocarbon 5, 312-341. Taylor, G. I. (1919). Tidal friction in the lrish Sea. Phiios. Trans. R. SOC.London, Ser. A 220, 1-33. Tolderlund, D. S. (1975). “Ecological Study of the Thames River Estuary (Conn.),” Rep. No. RDCGA575. U. S. Coast Guard Academy. Upson, J. E., Leopold, E. B. and Rubin, M. (1964). Postglacial change of sea level in New Haven Harbor, Connecticut. Am. J . Sci. [ 5 ] 262, 121-132.

THE SEDIMENTARY SYSTEM OF LONG ISLAND SOUND

39

Walcott, R. J. (1970). Isostatic response to loading of the crust in Canada. Can. J . Earth Sci. 7 , 7 16-727. Watts, A. B . , and Ryan, W. B. F. (1976). Flexure of the lithosphere and continental margin basins. Tectonophysics 36, 25-44. Wilson, R. E. (1976). Gravitational circuiation in Long Island Sound. Estuarine Coastal Mar. Sci. 4, 443-453.

This Page Intentionally Left Blank

STORM AND TIDAL ENERGY IN LONG ISLAND SOUND HENRY J. BOKUNIEWICZ Marine Sciences Research Center State University of New York Stony Brook, New York

AND

ROBERT B. GORDON Department of Geology and Geophysics Yale University New Haven, Connecticut

1.

2. 3. 4. 5.

Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Tidal Energy. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Storm Energy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Water Level Deviations . . . . . . . . . . . . . . . . . . . . . . . . . . . . Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Appendix 1. Formulation of the Energy Balance in an Embayment . . . . . . . . . Appendix 11. Estimate of Tidal Dissipation of All of LIS . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. .

41 43 48 55 60 61 65 67

1. INTRODUCTION

Power is used in an estuary to mix fresh and salt water, to resuspend bottom sediment, to transport sand and mud, and to maintain turbulence in the tidal stream. The requisite mechanical energy may be derived from the oceanic tide, the gravitational potential of the moon, the inflow of fresh water, or the wind stress acting on the water surface of either the estuary or the adjacent ocean. If the dominant power source for an estuary is tidal, the temporal variability of mixing and of sediment resuspension and transport should be regular; it should be possible to analyze their variation in terms of tidal constituents in much the same way that water level changes are analyzed. This is not possible when other power sources are important. Some alternative means of describing the time variation of power supplied to the estuary is then required for the analysis of estuarine mixing and transport processes, or before a program of observation can be designed to measure these reliably. Thus, determination of the temporal variation of the power supplied to an estuary can be of 41 ADVANCES IN GEOPHYSICS, VOLUME

22

Copyright 0 1980 by Academic Press, Inc. All ri&s of reproduction in any form reserved.

ISBN 0-12-018822-8

4,.JOa

74.00'

7 b w

I

73-00

I

72-30'

I

7200'

FIG.1. Map of Long Island and Block Island Sounds showing the locationsof tide gauges,anemometers,and current meters. The tidegaugeswere at New London (NL), New Haven (NH), Bridgeport (Bpt), Port Jefferson (PJ), New Rochelle (NR), Montauk (M), and Sandy Hook (SH). Newport (Np) is 67 km east of NL and not shown on the map. Current meters were operated at locations J, D, S, X, and Y; a water level recorder was also operated at J. Stratford Point Is St. Anemometer locations are shown by open circles. Power calculations are done for the section between A and B.

STORM AND TIDAL ENERGY IN LONG ISLAND SOUND

43

practical as well as scientific interest. Here we address this problem for Long Island Sound (LIS). The Sound (Fig. 1) is a large estuary in which the dominant source of power is thought to be the tide. However, observations of sediment resuspension and transport in water much too deep for the bottom to be affected by waves generated by the wind, show temporal variability that cannot be accounted for in terms of tidal constituents (see following article, this volume). Significant mechanical power must be entering the Sound from other sources. Long Island Sound receives fresh water principally from the Connecticut River. The resultant estuarine circulation was first described by Riley (1952, 1956) and has been calculated in a one-dimensional model by Wilson (1976). The circulation and mixing in the central Sound has been measured by Gordon and Pilbeam (1975). Because the natural period of longitudinal oscillation of LIS is very near 12.4 hr, the Sound has a resonant co-oscillating tide with an amplitude in the western end about four times that at the eastern entrance. Tidal currents in the Race reach speeds up to 4 knots. The tidal currents and water heights vary by about a factor of 2 between neap and spring tide. The Sound is subject to frequent winter storms and occasional hurricanes.

2. TIDALENERGY The first quantitative analysis of the tide in LIS was made by Redfield (1950), who represented the energy dissipation by linearized friction. The tidal oscillation used in his model has a node at longitude 72" 45'W and an apparent reflecting surface at 73" 30'W, about two-thirds of the way along the longitudinal axis of the Sound. The average rate at which energy is dissipated by Redfield's resonant co-oscillating tide, calculated according to the method derived by Ippen and Harleman (1966), is 460 MW at the mean tidal range. The relative rate of energy dissipation in an oscillating system may be described by the quality factor Q, which is 2n times the maximum energy stored per tidal cycle divided by the energy dissipated per cycle. (Q is also 2n times the number N of cycles required to dissipate the energy stored in the system.) For LIS, maximum energy stored in a tidal oscillation of mean range is 12.3 x lo6 MJ and Q is 3.8. The tidal oscillation is heavily damped ( N = 0.6). The only data used in Redfield's analysis are tide height observations. If current measurements are also available, the power P passing over any section across the Sound perpendicular to the current direction can be calculated by the method derived by Taylor (1919). If D is the water depth

44

HENRY J. BOKUNIEWICZ AND ROBERT B. GORDON

at mean sea level, h the height of the water level above D, and v the velocity normal to the section taken positive for inward flow, then (2.1)

P

=

pgDs(hv)

where p is the density, g the acceleration due to gravity, and s the length of the section. It is assumed that h 6 D in the derivation of Eq. (2.1). When the water level changes and the current velocities are solely due to the tide, we set P = P,, which is the tidal power crossing the section. The symbol (hv) indicates that the product is averaged over an appropriate time interval. In Taylor’s original analysis the averaging of h v was done over a semidiurnal tidal cycle. Because of the diurnal inequality and the spring-neap variation in tidal amplitude, there is a net gain (or loss) of energy from cycle to cycle as the tidal amplitude changes. To allow for this, we use an averaging period of 15 days in calculating the tidal energy flux. It is recognized that cross terms in (hv)due to lower frequency tidal constituents may not be negligible, but longer spans of data were not at hand. Simultaneous measurements of v and h were made at location J (Fig. 1) on the north-south section A-B through the middle of LIS. The measurements covered a 16-day period during which the winds were calm. The water depth was measured with a Bass model WG-100 wave gauge each hour and the current, with a General Oceanics model 20-10 current meter. Both instruments were fitted with quartz-crystal clocks so that timing errors were small, The current meter was set 2 m above the bottom on a taut mooring. The north-south component of the current at this location is only 15% of the east-west component, and the estuarine circulation carries bottom water westward into the Sound at a speed of about 5 cm/ sec; there is a corresponding outward flow of less-saline surface water (Gordon and Pilbeam, 1975). The east-west component of the measured velocity is used for v. Since in the calculation of (hv) there must be no net flow of water across the north-south section over an integral number of tidal cycles, this mean flow component was removed from the current meter record. There is no significant flow of tidal energy across any other boundary of the Sound to the west of section A-B so Pt is the tidal power passing into the Sound west of the section. If h and v measured at location J are representative of the entire section A-B, the energy crossing the section and entering LIS west of 72” 53’W in time T is the sum pgDs Z hv At, where At is the interval between successive, simultaneous measurements of h and v , and the summation is over all At in T. This sum is shown in Fig. 2 as a function of T. The indicated tidal power (i-e., the slope of the curve in Fig. 2) is larger at springs and smaller at neaps, as expected. The average observed power is 53 MW. The decrease in energy

STORM AND TIDAL ENERGY IN LONG ISLAND SOUND

45

10

1

N

-0

w

ou-

h

P

%,'

w

40-

20 0 (

100

200 .

- ~.

300

J

400

T (hr)

FIG.2. The total apparent tidal energy-crossing section A-B as calculated from h and v measured at J during calm weather. Curve L is the energy used in work done on the moon by the water within section A-B.

after T = 270 hr in Fig. 2 is unexpected, since there is no obvious source from which power could flow seaward across section A-B for times longer than a semidiurnal tidal cycle. The negative slope of the energy curve could arise from errors in the measurement of the phase difference between h and v , from release of energy stored in low-frequency components of the tidal oscillation, or from spatial variation in the flux of tidal power across section A-B. The data needed to check these possibilities are not at hand. However, the second seems quite unlikely, the first and third, about equally likely. Improvement in the measurement of the phase difference between h and v, if possible, would require a very long run of data because of the relatively large fluctuating component of v (see preceding article, this volume). Lateral variation of power along the section is discussed below. To find the actual P, for section A-B from the results presented in Fig. 2, allowance must be made for: (1) the variation of the amplitude and phase of v through the water column at location J; (2) the lateral variation of the magnitude of h and v across the section; and (3) the change in the phase of h and w across the section.

The variation of the magnitude and phase of v with height is due to friction at the bottom and the estuarine circulation. Data from a vertical current

46

HENRY J. BOKUNIEWICZ AND ROBERT B. GORDON

meter array at location S were used to evaluate (hv)for the upper and lower half of the water column. Current meter records for locations S, X, and Y spaced along the section were used to calculate (hv) at these locations. The records are not simultaneous and no concurrent measurements of h are available. It was necessary to compute h from the tidal constituents at J because tide height data for the southern part of section A-B are not available and tidal intervals for the shore of Long Island published in tide tables do not yield a consistent pattern that can be used for interpolation. The phase of the tide measured at J and that predicted for location A, the New Haven Harbor entrance, is nearly the same and is, therefore, assumed to be constant across the section. Le Lacheur and Sammons (1932) report that the tidal range is only 5% less on the Long Island shore than on the Connecticut shore. On the basis of these computations it is estimated that the P, shown in Fig. 2 should be decreased by 20% to get the actual energy flow across the section. When this correction is applied, the average, net tidal power through section A-B becomes 43 MW. Variation in the lateral distribution of the tidal power flow across the section between spring and neap conditions is not ruled out and may be the cause of the negative slope in the energy curve in Fig. 2. Part of the energy flowing into western LIS across section A-B is used in mixing surface and bottom water, part is used in work done on the moon, and the rest is dissipated by friction. The power required for the mixing of surface and bottom water in the western half of LIS is estimated by the method of McLellan (1958) to be about 1 MW, or 2% of the mean tidal power. The work done on the moon is substantially larger, however. The method of calculating this work was derived by Taylor (1919). Taylor’s method has been widely used, but it has recently been shown to be incorrect (Garrett, 1975). The work done on a volume of water of surface area A inside a boundary of length S and mean water depth D by

FIG.3. Definition of the phase angle used in Eq. (2.3).

STORM AND TIDAL ENERGY IN LONG ISLAND SOUND

47

TABLEI. TIDALPOWERCHARACTERISTICS OF WESTERNLIS AVERAGED OVER THE SPRING-NEAP PERIOD Measured tidal power over section A-B as measured at location J Corrected tidal power for section A-B Power used in mixing surface and bottom water Power used in work done on the moon Power dissipated through friction

53 MW

43 1 17

25

the moon’s attraction during two complete lunar semidiurnal tides is (2.2)

Em

=

pDS(sl~)- PA

J

h dfl

where n is the gravitational potential of the moon, the integration is over all changes in in two semidiurnal tidal cycles, and the average is also taken over two cycles. We show how this expression can be used to calculate the mechanical energy balance of an embayment in Appendix I. If the tide height and tidal stream speeds can be represented by sinusoidal terms, Eq. (2.2) reduces to MR’ (2.3) E m = 3pG 7cos2d cos’ B($DSvoP sin 24$, - h A H sin 2 4 ~ ~ ) Dm where G is the gravitational constant, R the Earth’s radius, D , the radius of the moon’s orbit, M the mass of the moon, 0 the latitude, 2H the tidal range, P the tidal period, and d the declination of the moon. The angle cb0, defined in Fig. 3 , is measured from the time interval between the moon’s meridian passage and high water; 4; is measured from the meridian passage to slack water. High water occurs before the moon’s meridian passage inLIS so c $ > ~ 0 and, in the western Sound, the second term in Eq. (2.3) is larger than the first and work is done on the moon by the tide. High water is nearly simultaneous and the variation of tidal amplitude is small throughout LIS west of section A-B. Calculation of Emto the requisite accuracy may then be made with the aid of tide tables and the nautical almanac. This calculation was done numerically for the period in which the data in Fig. 2 were obtained and is shown as curve L; over the spring-neap cycle Em averages - 17 MW or 40% of the energy flow across section A-B. This leaves 25 MW as the average tidal power dissipated by friction in the western half of LIS. The results are summarized in Table I. Data that would permit computation of the influx of tidal power to all of LIS are not at hand, but an estimate can be made. It is given in Appendix 11.

48

HENRY J. BOKUNIEWICZ AND ROBERT B. GORDON

3. STORMENERGY

Storms passing over southernNewEngland cause water level deviations at the shore ranging up to -1 m (Miller, 1958). These deviations result principally from set-up, i.e., sloping of the sea surface due to wind stress. Easterly storm winds also generate strong alongshore currents in the waters of the continental shelf; the pressure gradient due to the set-up and the alongshore current are found to be nearly balanced (Beardsley and Butman, 1974). Miller found that the water level deviation at any given observing station can be separated into a regional component (due to sea surface tilt) and a local component, which is influenced by the configuration of the land near the site of observation. This implies that energy from a storm reaches the shore zone from both local winds and regional winds blowing over the waters of the continental shelf. In order to evaluate the changes in the energy flow into LIS due to storms we first examine records for one major winter storm. The “northeaster” of 15-16 December 1972 is chosen because simultaneously recorded water height, current meter, and wind velocity data are available for it. Deviations of observed water levels from predicted tidal heights (“residuals,” Ah) at New Haven (NH), New London (NL), and Newport (Np) were calculated for an eight-week period starting in mid-December 1972 by a regression method using 18 tidal constituents. Additional water level residuals were supplied to us by J. Ianiello (personal communication, 1975) for the days of the storm of 15-16 December. Two current meters set 2 m above the bottom and located 3 km apart were in operation at D and S in Fig. 2 nearly on section “A-B,” throughout the study period. These meters were in water sufficiently deep to be unaffected by waves at the water surface. The recorded velocities were resolved into E-W (u,) and N-S (u,) components. Each of these components was then divided into mean, tidal, and fluctuatingparts with the aid of the regression analysis. Wind data are available from the U. S. National Weather Service station on Stratford Point, from several Coast Guard Stations around the Sound, and, in more detail, from a weather tower at apower plant adjacent to New Haven Harbor. The locations of all observing stations are shown in Fig. 1 . The residual water levels at NH, NL, and Np, the residual currents (the sum of the mean and fluctuating components), and the square of the E-W and N-S components of the wind speed measured at Stratford (assumed proportional to the components of the wind stress on the water surface, a,, a,) are shown for the period of the storm in Fig. 4. The water levels at NH, NL, and Np rise while the wind has a strong easterly component; the subsequent fall in level begins when oxreverses. The times of the maximum and zero Ah for these three water level stations are nearly

49

STORM AND TIDAL ENERGY IN LONG ISLAND SOUND

--

WATER LEVEL RESIDUAL I

1000-(WIND SPEED^

--N-S

bE

N -

-----

'

New Haven New London Newport

500-ig N-

a W

w n

o=-,

v)

-

'F

5OO-ggIn 1000

14

I

I

I

15

I

I

16 DECEMBER 1972

I

17

I I

18

FIG.4. Departure of observed from predicted water level at NH, NL, and Np; the mean flow recorded 2 m above the bottom at location D, and the relative wind stress as measured at NH during a winter storm, December 1972.

coincident, but the magnitude of Ah increases to the westward. Cross correlations were computed for the water level residuals at NH, NL, and Np for a 15-day period beginning 15 December 1972. The results are summarized in the following tabulation and show that there is a very close correlation between the water level deviations at these three stations: Harbor

Correlation

NL vs. Np NH vs. Np

0.96 0.83

50

HENRY J. BOKUNIEWICZ AND ROBERT B. GORDON

The time required for a shallow-water wave to travel from Newport to New Haven is about 2.5 hr, from New London to New Haven, about 1.3 hr. Since the greatest Ah occurs earlier at New Haven than at New London, the rise in water level in LIS is not a surge advancing as a progressive wave from the sea. Evidence of a storm surge in LIS is found only for very intense, rapidly moving storm systems, such as the 1938 hurricane (Redfield and Miller, 1957). The water level residuals observed at successive times during the 15-16 December 1972 storm are shown in Fig. 5 for all observing stations. Station locations are projected onto a line running along the axis of LIS (direction 075 true from Throgs Neck) to construct the abscissa of the graphs. Wind speeds and directions are shown on an adjacent set of maps. The residual water levels at Bridgeport and Port Jefferson, nearly opposite each other across the Sound, are almost the same throughout the storm. Hence, there is little change in water level across the Sound and the levels shown in Fig. 5 define the longitudinal slope of the water surface throughout the study area. During the period of easterly wind, the water level throughout the region is raised and the water surface slopes upward to the west. The water level residuals at Montauk Point and Sandy Hook, on the open ocean, are also shown in Fig. 5. The dashed line connecting them represents the slope of the sea surface in the direction 075 outside of LIS and Block Island Sound (BIS); it is nearly the same as the slope inside the two Sounds. The water level and wind data suggest that the following sequence of events occurred during the December 1972 storm: easterly winds set water on the continental shelf in motion towards the west, increasing water levels along the coast and tilting the sea surface upward, as observed for similar storms by Beardsley and Butman (1974). Similar changes in water level and surface slope occur in BIS and LIS. (There is also an increase in water level along the coast due to the reduced barometric pressure, but this is small compared to the change due to wind stress; no correction for local barometric pressure has been made in calculating the residual water levels.) The movement of the storm center, and the resulting change in wind speed and direction, are sufficiently slow that approximate balance between wind stress and surface slope is maintained. The inflow of water to LIS is seen in the current meter records (Fig. 4) as a residual flow to the west (u, < 0) during the period of water level rise. The observed residual flow across section A-B accounts for the observed increase in water volume within LIS westward of the section to within the accuracy of measurement. The water level at the extreme western end of LIS is influenced by the net flow through the East River, but the volume involved is negligible compared to that entering through the eastern passages to

STORM AND TIDAL ENERGY IN LONG ISLAND SOUND

51

BIS and the Atlantic Ocean. When the alongshore component of the wind reverses direction, the excess water level begins to fall and, as the intensity of the west wind increases, the surface slope both inside and outside the two Sounds reverses. An outward net flow of water from LIS is then indicated by the current meters. The mean tidal prism of LIS is 5.47 x lo9 m3. The greatest excess volume of water in the Sound during the 15-16 December storm is 3.4 X lo9 m3 or 62% of the mean tidal prism. The potential energy of the excess water at maximum 6h is 18 x lo6 MJ; the potential energy at the top of the tidal oscillation of mean range is 12.3 x lo6 MJ. The power-crossing section A-B during the storm can be evaluated from the current meter records from locations D and S and tide gauge data, but the quality of the data is not as high as that shown in Fig. 2 for two reasons. First, mechanical clocks were used in the current meters and interpolated corrections for their rates are required. Second, the available water level data are from a tide gauge within New Haven Harbor (rather than one at a current meter site), and a correction for the phase and amplitude difference of the tide between this location and that of the current meters is required. The tide height was corrected by a factor of 1.3, determined by a direct comparison of simultaneous water level measurements at the New Haven harbor tide gage and location J. The phase of the water level observed at NH was then corrected so that the tidal power dissipation during a calm period of about 25 hr duration that occurred about four days after spring tides on both the December 1972 and the October 1975 records, was the same on both records. This required that the phase angle be corrected by a factor of 0.78, which corresponds to a shift of about 1". The lunar work rate increased by about 3% as a result of this phase correction. Figure 6 shows the power-crossing section A-B as measured at locations J and D after the power used in lunar work and in fresh water-salt water mixing has been removed from each record. Both curves represent energy that must be dissipated by friction, curve 1 for calm conditions and curve 2 for time that includes the storm. The positions of the curves on the time axis have been shifted so that the times of spring tides coincide. (Maximum spring tides occur 130 hr after the start of both curves.) The slopes of the two curves are nearly identical for the time after the storm, which shows that the corrections to h based on the 24-hr calm periods, as described earlier, are reasonable. The agreement of the slopes also suggests that the decrease in the total energy near the ends of both curves is not due to measurement errors, but results from inadequate correction for the lateral variation of the energy flux across section A-B. During the stormy period, the first 85 hr of the record, the power crossing

FIG. 5. Left-hand graphs show the water level residual along the central axis of LIS at successive times during the December 1972 storm; wind velocities are shown on the righthand side. The dashed line shows the water level outside of the Sound.

STORM AND TIDAL ENERGY IN LONG ISLAND SOUND

IWO

3

2

-<-- -

-

*-

.

.

,200

,

1400

/ /

F' NL

3 2

s r

- - - %---

u

c 201ac,

53

54

HENRY J. BOKUNIEWICZ A N D ROBERT B. GORDON

60

-

0

100

200

300

400

Time (hr)

FIG. 6. Power-crossing section A-B during calm weather (curve I ) and during a major winter storm (curve 2) after the lunar work has been removed from each curve. The curves have been shifted on the time axis so that the times of spring tides coincide.

section A-B that must be dissipated by friction is 124 MW and the difference between the two energy curves increases at a mean rate of 68 MW. The total additional energy that must be dissipated by friction in the Sound west of section A-B during the storm is about 21 x lo6 MJ, which makes the total dissipation 184% of that during calm weather. During the stormy period water velocities near the Sound floor are observed to be greater than the normal tidal velocities for only brief intervals of time. If the increased dissipation is to be described in terms of mean flow speeds, there must be an increase in the drag coefficient at the bottom. For the dissipation rate to increase from 56 to 124 MW, CDwould have to increase by a factor of 2.2. Changes in bottom roughness during stormy periods may contribute to an increase in CD.However, upon the initiation of stormy conditions, the dissipation increases rapidly compared to the time required to alter the configuration of bed features, such as sand waves, large enough to produce form drag. Another mechanism must operate. An alternative suggestion is found in the observation that wind stress increases the intensity of tidal stream velocity fluctuations in LIS (Bokuniewicz et al., 1975). Histograms of the fluctuating components of u, and u, are found to broaden; during the 14-19 December 1972 storm period the standard deviation (SD) of u, reaches 7 cm sec-I, whereas under calm conditions it is 1 cm sec-'. Observations of the SD of u, have not been made, but an increase comparable to that for u, and

STORM AND TIDAL ENERGY IN LONG ISLAND SOUND

55

u, is anticipated. A steeper velocity gradient near the bottom and greater energy dissipation then results. Increased power dissipation during storms will have a direct effect on bottom processes in the Sound. For example, if the rate of transport of sediment is proportional to the power dissipation, transport rates will be substantially increased during stormy periods even though the water is much too deep for the bottom to be affected by waves on the surface. Resuspension of silt-clay-size sediment will also be increased and greater "stress" will be applied to benthic animal communities. The storm analyzed in Fig. 6 would be considered a major winter storm, but its duration is only about 85 hr. To assess the importance of increased power dissipation during storms on sedimentary processes and the total energy budget of the Sound, measures of the total amount of storm-inducedpower dissipation during a year (or series of years) and of the frequency of occurrence of storms of different magnitudes are needed. Long-term measurements of v required for the caldulation of (hv) have not been made, but records of h over many years are available for some National Ocean Survey tide gauges. Since the mechanism by which the internal friction of the tidal oscillation is increased by wind stress has not been identified, there is no theoretical basis for establishing a relation between the magnitude of Ah and the storm-induced power inflow to LIS. An extended set of observations that would permit the relation to be established empirically is not available either. Some assumption has to be made and we suggest that the additional energy dissipated during a storm be taken as proportional to Ah, the greatest water level residual during the storm. This is equivalent to assuming that the additional energy dissipated is proportional to the greatest additional potential energy in the water column due to the rise in water level or, since most winter storms have about the same duration, that the increase in the internal friction of the tidal oscillation is proportional to the water level residual. With this assumption the tide gauge records can be used to estimate the temporal variation in the release of storm-induced energy in the waters of the Sound.

4. WATER LEVELDEVIATIONS

The water level data in Fig. 5 show that Ah at any place in the Sound (except at the extreme western end) is proportional to Ah at New London. Then, the analysis of water level deviations need be done only for the New London tide gauge. The hourly water levels at this station for the years 1938-1975 were obtained from the U.S. National Ocean Survey. The tidal component of the water level was removed by a regression

56

HENRY J. BOKUNIEWICZ AND ROBERT B. GORDON

calculation using 18 tidal constituents (Dronkers, 1964). A separate regression was used for each year’s data. The water level residuals were then plotted so that the dates of particular events could be determined. Examples are shown in Fig. 7 for a winter and a summer month. The characteristic rise and subsequent fall in Ah associated with passing storm systems is easily recognized. The dates and amplitudes of Ah were read from the graphs and arranged in order of decreasing amplitude. These data were then analyzed by the methods used to characterize the frequency of floods on a river (Lindsley and Franzini, 1972), that is, Ah is plotted against the recurrence interval calculated according to an extreme value probability distribution, as shown in Fig. 8. A linear relation, such as is usually found for flood records, also obtains for the water level deviations due to winter storms. Hurricanes do not follow the same distribution, as expected. Figure 7 can be used to obtain the recurrence interval of any particular storm; for the 14 December 1972 storm examined about it is 15 months. It is estimated, then, that for a “10-yr” storm, the

-I

1.0

February 1960

-

0

5

10

New London

15

20

25

30

0

E

August 1960

FIG. 7. Water level residuals for a winter and a summer month at NL.

Days

57

STORM AND TIDAL ENERGY IN LONG ISLAND SOUND

-

I

I

I

I

I

2.0

A

Tidal Deviations at New London

1

-

-

.

A

A = Hurricane

A

E

Y

G 1.0

-.

-

I

I

I

I

5

10

50

100

I

-

500 1000

FIG. 8. Recurrence intervals for water level residuals calculated from a 38-yr water level record at NL.

power dissipation level in the western Sound would be raised by 86 MW; the total dissipaiion would be 372% above that of the mean, calm weather tidal oscillation. For each month of the 38-yr record the SD of Ah and the mean water level was computed. The averages of these for each month in the record interval are plotted in Fig. 9 in order to show the seasonal change. Mean sea level changes by 0.11 m during the year. There is a particularly rapid drop in water level between November and December and a more gradual rise in the spring. This does not coincide with the seasonal change of LIS water temperature and is, therefore, thought to be a response to changes occurring outside of the Sound. The curve for SD of Ah, Fig. 9a, shows that there is a large seasonal variation of the monthly mean storm-induced energy dissipated in LIS. This storm energy is high in December-March, low in June-September, and goes through a very rapid change in the intervening months. An estimate of the power dissipation levels in the western Sound can be made from the analysis in the preceding section. It was observed that when the tidal deviation at New London was 1 m, the additional power entering the western Sound was 68 MW. If, as as-

t t , , , , , , , , , , ,

' S O N D J

1

1

1

1

F

1

M

1

A

1

M

J

1

J

1

A

1

1

1

New London mean sea level, 1938-1974

€

J

F

M

A

M

f f

4

J J A MONTHS

S

f

O

N

D

FIG.9. (a) The mean SD of the water level residuals for each month in a 38-yr water level record from NL (heavy line) and its SD (light lines); (b) mean sea level for each month computed from the same 38-yr water level record.

STORM AND TIDAL ENERGY IN LONG ISLAND SOUND

59

sumed previously, the excess power is proportional to Ah, then the mean excess power can be computed for any month from the SD of Ah. For an average winter month the total average power dissipation is 39 MW and for July, when the excess due to storms is a minimum, it is 30 MW. The monthly mean winter power level is, therefore, 30% above the mean summer level. A corresponding increase in, for example, monthly mean sediment transport rates in the winter months is expected. The variability of the monthly mean SD of Ah is also much greater in the winter, as is shown by its own SD (the thin lines in Fig. 9a). In the 38 years studied the highest monthly mean power dissipation was 47 MW, a level of 156% above the mean summer power. Hurricanes may cause very large Ah, but the short duration and infrequent occurrence of these storms causes them to make only a small contribution to the SD of Ah. The pattern of seasonal variation in the monthly mean SD of Ah is well defined in the 38-yr average (Fig. 9a), but there is a wide variation from

FIG.10. Mean SD of the water level residuals for five winter and five summer months for each year of the tide-gauge record from NL.

60

HENRY J. BOKUNIEWICZ AND ROBERT B. GORDON

year to year in the amount of energy dissipated in LIS in any given month or season. The average SD of Ah for the winter months (November-March) and the summer months (May-September) for each year of the 38-yr record is plotted in Fig. 10. There is no significant long-term trend in either the winter or summer storm-induced energy dissipation, but the variation of energy dissipation between different years can be very large. In the interval 1950-1956 the variation was unusually small; the excess energy was particularly large in the winters of 1945-1946, 1958-1959, and 1972-1973. These variations are expected to correlate with variations in the annual sediment transport and, perhaps, benthic animal populations in LIS. 5 . CONCLUSIONS

The tide is the principal source of the power used in the estuarine processes of mixing and sediment transport in LIS. (Processes powered by waves are confined to a zone of shallow water around the margins, as described in the following article.) Work done by the gravitational attraction of the moon on the water in the Sound is a significant part of the energy budget. In the western half of LIS work is done by the water on the moon, but, for the Sound as a whole, the moon does work on the water. Power not used in lunar work and in mixing fresh and salt water is dissipated by friction and results in a large damping of the tidal oscillation. The temporal variation of the dissipation is large. Part of this variation is due to long-period tidal constituents, is regular, and could be predicted if sufficiently long data runs of h and v were available. However, the internal friction of the Sound is found to be strongly influenced by the local wind stress; under storm conditions Q-' may be more than doubled. The resultant increase in the tidal dissipation causes increased resuspension and transport of sediment on the bottom of the Sound even though the water is much too deep for the bottom to be directly affected by waves on the water surface. The data base that is available for determination of the tidal dissipation in LIS is quite limited. However, the available data do provide a basis for designing observation programs for quantities such as bottom stability or bed-load and suspended-load transport rates that are controlled by the power dissipation. They also provide a basis for estimating long-term transport from the relatively short runs of data that are available. These predictions can be made more quantitative when the time scales of the variability of the forcing can be compared to the time scales of the system response. For example, it is thought that hurricanes are quite unimportant

61

STORM AND TIDAL ENERGY IN LONG ISLAND SOUND

in the long-term sediment transport in the deep waters of the Sound because of their short duration, but some transport processes may be activated only under the short but intense excitation that they provide. Certain additional measurements are clearly needed in the Sound. An array of current meters and water level recorders should be operated on a cross section so that the lateral variation of the power flux can be reliably determined. Longer data runs for h and v are needed to adequately define the mean tidal dissipation. Simultaneous measurements of sediment resuspension and’transport and of the tidal dissipation could be used to test the hypothesized interrelation between these quantities.

APPENDIXI. FORMULATiON OF THE ENERGYBALANCEIN AN

EMBAYMENT Consider an embayment as shown in Fig. Al. The boundary S is a section across the embayment that contains an area A. The boundary B is the shoreline. The governing equations for the vertically integrated current vector d and the elevation of the water above mean sea level h , are

aa + f x

p-

(A. 1)

at

pd + p g V ( h - he)

+E =

0

and

ah + V - (pD&) = 0 pat where p is the density of water, D the mean water depth, g the acceleration of gravity, f the Coriolis parameter, P the friction, and he is the level of the equilibrium tide; i.e., pg Vh, is the tide generating force (see Garrett, 1975). An expression for the energy balance within the embayment may be derived by adding (A.l) D d and (A.2) X (gh),

-

(A.3)

?!!at + V . (pgDdh) - pgDd

*

Vh,

+ pDd

*

#

=

0

where E is the energy density per unit area; E = $p 1 u I2D + fpgh2. Since pgDh V h , = V . (pDdgh,) - pgh, V - ( D d ) , using (A.2), an alternate form of (A.3) is

-

aE ah (A.4) - + V (pgDlih) - V * (pDdgh,) - pgh, - + pDd at at

*

P

=

0

62

HENRY J. BOKUNIEWICZ AND ROBERT B. GORDON

FIG. A l . Idealized embayment used in the derivation of the equations for the energy dissipation.

The total rate of change of energy within that part of the embayment contained by the section S is the integral of (A.4) over A. Taking A as the outward unit vector everywhere normal to S and B, li * A = 0 everywhere along the shoreline B . The energy balance may be written as

:+

-da J A

Dpli.Edu = -

I A

Is

pgDhi2.fi ds +

Is

pgDh,G*A ds

To simplify this expressiom, let v be the current velocity normal to S , let Iv( be constant along S, and v positive inward so that when water is flowing into the embayment, the energy in the embayment is increasing. The energy balance then becomes (A.6)

J

A

at

da +

I,

Dpli . E da = pgDShv

- pgDSh ev

ah + S, pgh e % da

Over a long time (many tidal cycles) the total energy within the embayment neither increases nor decreases. We indicate the long time average

STORM AND TIDAL ENERGY IN LONG ISLAND SOUND

63

by angular brackets, F

(A.7)

=

(pgDShv) - (pgDSh.v)

$ d.)

+

where F = (JA Dpi P du) is the average rate of energy dissipation due to friction. The lunar gravitational potential R is equal to - (gh,) (Lamb, 1945, p. 359), so an alternate form of (A.7) is

The first term on the right is the average energy flux through the section into the embayment. This term was originally derived by Taylor (1919). The third term was also used by Taylor; since the linear dimension of the embayment is small so that R does not vary appreciably throughout the volume, he wrote this term as -rn(dR/dt) where rn is the total mass of water contained by the sections [see also Jeffreys, 1970, 08.07, Eq. (7), p. 31 11. As Garrett (1975) has shown, Taylor and subsequent investigators neglect the second term and, for the usual case where the scale of h e is much greater than the size of the embayment, the magnitude of the sum of the second and third terms is much less than the magnitude of either term separately. The present study evaluated each of the three terms from measurements of the water level h i and the component of the current velocity perpendicular to the section vi. These terms were evaluated as follows: CI pghS(hv) was calculated as -pghS hivi At T i

C.

CZ

phS(Rv) was calculated as -phS RiviAf T i

(A.9)

-

I,pa!$

C3

RAWhi+ I - hi- 111

da was calculated as - A p

T

where T is the record length, At is the time between successive measurements, and C,,Cz,and C,are correction coefficients that are explained in the text. The approximation used for dhldt assumes that the water level within the section is the same everywhere. While this is, of course, not strictly true, it is nonetheless a reasonable approximation in western LIS. The lunar gravitational potential is (A.10)

R = jG(MR/D,)(A -

COS'

+)

64

HENRY J. BOKUNIEWICZ AND ROBERT 9. GORDON

where G is the gravitational constant, M the mass of the moon, D , the radius of the moon’s orbit, R the radius of the Earth, and Cp the angle between the line joining the centers of the Earth and moon and that radius of the Earth that passes through the point on the Earth’s surface being considered [Lamb, 1945, Appendix to Chapter VIII, Eq. (2), p. 3581. If Cp is the latitude through the center of LIS, and LHA and d are the local hour angle and declination of the moon, respectively, then by the law of cosines for spherical triangles, (A. 11)

cos Cp = sin 8 sin d

+ cos 8 cos d cos LHA

In estimating the work done against the moon’s gravity, Taylor derived the special case where d = 0. The general expression will be presented subsequently. The calculations discussed in this article were done using the equations just given [i.e., (A.8)-(Av1l)]. It is instructive, however, to examine a simple analytical representation of the tidal flow in the Sound. Let the Sound be represented by a closed channel. Flow into and out of the channel takes place only across one face. The channel is, of course, fixed on the Earth rotating in the lunar gravitational field. The water level in the channel is given by

h = ho cos 2(x

(A. 12)

+

$0)

where x is the angle between the longitude through the center of the Sound and the longitude through the sublunar point, ho is the amplitude of the tide, and Cp,, is defined in the text (Fig. 3). Likewise, the tidal velocity is given by (A. 13)

v

=

-v,, sin 2(x

+ 44)

where v,,is the maximum tidal velocity. The sign of v is chosen to ensure that the water level is rising when the current flow is into the Sound, i.e., when v > 0. When v is directed into the Sound the energy in the Sound is increasing. We substitute (A.10), (A.12), and (A.13) into (A.8); the time averages can be evaluated analytically over two complete tidal cycles. Since the moon’s declination varies slowly, it may be assumed to remain constant over the two tidal cycles:

F = -fpgDShovo sin[2(& - +0)1 r

1

STORM AND TIDAL ENERGY IN LONG ISLAND SOUND

65

FIG. A2. Water-height change h , east-west velocity v , and equilibrium tide he for location J on 29 October 1975.

where T is the tidal period. The third term on the right in this expression is equivalent to that used by Taylor to calculate the lunar work rate. Taylor, however, evaluated the special case where d = 0. Along the section through LIS used in this study, ho is =90 cm and vo = 30 cmlsec. At the beginning of the records made at Station J, 29 October 1975, +o = 19.6"and, since slack tide occurs about 10 min. after high water, &, = 17.2'. The water level, current velocity, and the equilibrium tide over two complete tidal cycles are shown in Fig. A2. For this simplified representation, after two complete tidal cycles the Sound returns to its initial condition with no net gain or loss of energy, Evaluating (A.14) with the suggested quantities results in an average frictional dissipation rate of 90 MW. APPENDIX11. ESTIMATE OF TIDALDISSIPATION OF ALL OF LIs An estimate of the dissipation of energy by friction in all of LIS can be made if it is assumed that the only difference in the tidal friction between the eastern and western parts of the Sound is that due to the increased speed of the tidal stream. The power used in work against friction

66

HENRY J. BOKUNIEWICZ AND ROBERT B. GORDON

TABLEAl. TIDALPOWERCHARACTERISTICS OF LIS ~~

Mean for springneap period (MW)

Characteristic Measured tidal power over section A-B measured at location J Corrected tidal power for section A-B Power used in mixing surface and bottom water, western LIS Power applied to moon, western LIS Frictional dissipation, western LIS Estimated frictional dissipation in LIS Estimated power applied to moon in LIS Power entering LIS from the sea

At mean tidal range (MW)

53

109

43 1

87 1

17 25 192 - 85

17 70 538 - 85

109

455

is expected (for constant drag coefficient) to be proportional to v3. The magnitude of v over the whole Sound is shown in the tidal stream diagrams published by the U. S. National Ocean Survey. These data were contoured and used to find the ratio of the mean values of v3 in the western Sound and in the whole Sound. The ratio is 0.13 so, if the dissipation in the western Sound is 25 MW, it must be 192 MW in the whole Sound. The lunar work is found from Eq. (2.3). When it is evaluated for the entire Sound the second term changes from - 83 to - 166 MW, primarily because of the increase in area, whereas the first term increases from 60 to 251 MW because of the higher values of vo in the eastern Sound. For all of LIS work is done on the water by the moon at the rate of +85 MW. The power used in mixing fresh and salt water is about 2 MW. Since 192 MW is being dissipated by friction, the inflow of power to the Sound from the ocean outside must be only 109 MW. This is an average value over the spring-neap cycle. If these computations are repeated for the dissipation at mean tidal range, the inflow of power is found to be 455 MW, which may be compared with the 460 MW estimated from Redfield’s model for mean tidal range. The computations are summarized in Table Al, ACKNOWLEDGMENTS We thank Carol C. Pilbeam and Jack laniello for assistance with the computation of tidal residuals and Malcolm Bowman and Robert Hall for helpful discussions on the computation of tidal power. Many of the data used were obtained with the aid of Jeffrey Gebert, Peter Kaminsky, and Matthew Reed during studies made for the United Illuminating Co. of New acknowledges Haven and for the U.S. Army Corps of Engineers. One of the authors (H.B.)

STORM AND TIDAL ENERGY IN LONG ISLAND SOUND

67

the support of the University Awards Committee of the Research Foundation of the State University of New York.

REFERENCES Beardsly, R. C., and Butman, B. (1974) Circulation on the New England Continental Shelf: Response to strong winter storms. Geophys. Res. Lett. 1, 181-184. Bokuniewicz, H. J., Gordon, R. B., and Pilbeam, C. C. (1975). Stress on the bottom of an estuary. Nature (London) 257, 575-576. Dronkers, J. J. (1964). “Tidal Computations in Rivers and Coastal Waters.” North-Holland Publ., Amsterdam. Garrett, C. (1975). Tides in gulfs. Deep-sea Res. 22, 23-35. Gordon, R. B., and Pilbeam, C. C. (1975). Circulation in central Long Island Sound. J. Geophys. Res. 80, 414-422. Ippen, A, T., and Harleman, D. R. F. (1966). Tidal dynamics of estuaries. I n “Estuary and Coastline Hydrodynamics” (A. T. Ippen, ed.), pp. 493-545. McGraw-Hill, New York, Jeffreys, H. (1970). “The Earth,” 5th ed. Cambridge Univ. Press, London and New York. Lamb, H. (1945). “Hydrodynamics,” 6th ed. Dover, New York. Le Lacheur, E. A., and Sammons, J. C. (1932). Tides and currents in Long Island and Block Island Sounds. U.S . Coast Geodetic Surv., Spec. Publ. 174. Lindsley, R. K., and Franzini, J. B. (1972). “Water Resources Engineering.” McGraw-Hill, New York. McLellan, H. J. (1958). Energy considerations in the Bay of Fundy System. J . Fish. Res. Board Can. 15, 115-134. Miller, A. R. (1958). The effects of winds on water levels on the New England coast. Limnol. Oceanogr. 3, 1-14. Redfield, A. C. (1950). The analysis of tidal phenomena in narrow embayments. Papers Phys. Oceanogr. Meteorol. 11, 1-35. Redfield, A. C., and Miller, A. R. (1957). Water levels accompanying Atlantic coast hurricanes. Meteorol. Monogr. 2, 1-21. Riley, G. A. (1952) Hydrography of Long Island and Block Island Sounds. Bull. Bingham Oceanogr. Collect. 13, 5-39. Riley, G. A. (1956). Oceanography of Long Island Sound, 1952-1954, 2. Physical oceanography. Bull. Bingham Oceanogr. Collect. 15, 15-46. Taylor, G. I. (1919). Tidal friction in the Irish Sea. Philos. Trans. R . SOC.London, Ser. A 220, 1-33.

Wilson, R. E. (1976). Gravitational circulation in Long Island Sound. Estuarine Coastal Mar. Sci. 4, 443-453.

This Page Intentionally Left Blank

SEDIMENT TRANSPORT AND DEPOSITION IN LONG ISLAND SOUND HENRY J . BOKUNIEWICZ Marine Sciences Research Center State Universiry of New York Stony Brook New York

.

AND

ROBERTB . GORDON Department of Geology and Geophysics Yale University New Haven . Connecticut

1 . Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2. Power Sources . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.1 Tide . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.2 Waves . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.3. Currents . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.4. River Flow . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.5. Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 . Sediment Sources . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.1. Shoreside Erosion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.2. Exchange with the Continental Shelf . . . . . . . . . . . . . . . . . . . . . 3.3. River Discharge . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4 . Sediment Transport and Bottom Stability . . . . . . . . . . . . . . . . . . . . . 4.1. Sand Transport . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.2. Mud Transport . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5 . Sediment Deposition and Distribution . . . . . . . . . . . . . . . . . . . . . . 5.1. Rate of Deposition of Mud . . . . . . . . . . . . . . . . . . . . . . . . . 5.2. The Sediment Mass Balance . . . . . . . . . . . . . . . . . . . . . . . . 6 . Comparison with Other Estuaries . . . . . . . . . . . . . . . . . . . . . . . . 6.1. Energy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.2. Sediments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Appendix . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

69 70 70 70 75 82 82 84 84 85 86

87 89 91 95 97 99 99 99 101 103 104

1. INTRODUCTION

Sediment may enter an estuary from rivers. from erosion of the shoreline. from the air. or from the sea. Some of it is converted to a different physical form within the estuary; some of it is stored. and some may be exported to the ocean . The power required to move sediment within the estuary comes from the tide. the wind. and the inflow of fresh water . Our 69

.

ADVANCES IN GEOPHYSICS VOLUME 22

Copyright 0 1980 by Academic Press. lnc. All rights of reproduction in any form reserved. ISBN &12-018822-8

70

HENRY J . BOKUNIEWICZ AND ROBERT B. GORDON

object here is to develop a quantitative description of the processes of sediment transport and deposition for one large estuary and to put the results in a form that will facilitate comparison with other estuaries. Long Island Sound (Fig. 1) is typical of the larger estuaries found around the North Atlantic above 40"N latitude. Study of the tidal characteristics of the Sound began in 1835 (Le Lacheur and Sammons, 1932), but the first study of its physical oceanography was made by Riley (1952, 1956). Since that time the Sound has attracted the attention of scientists of many disciplines so that much of the basic data needed for analysis of sedimentary processes has been accumulated. We begin by finding the amount of power from the tide, wind, and river flow available to drive sedimentary processes in the Sound. The temporal variability of these power sources is characterized. Next we identify the sediment sources and the processes by which sediment received by the estuary is altered in form. Sediment transport and the stability of sediment deposits is related to the power dissipated in the estuary. Deposition rates and the sediment mass balance of the estuary are examined. Finally, a set of parameters that characterize the estuarine sedimentary system are defined and evaluated. It is suggested that these parameters may be used to compare the sedimentation characteristics of different estuaries.

2. POWERSOURCES 2.1. Tide

Power used in the transport of sediment through Long Island Sound is derived from the tide, the wind, and the inflow of fresh water. Of these, the tide is the most important. The natural period of oscillation of the water in the Sound is near 12.4 hr so that there is a large, resonant, cooscillating tide. The oscillation is, however, heavily damped by tidal friction (Redfield, 1950). We described measurements of the tidal power in Long Island Sound in the previous article (this volume). The internal W, where S W is the energy friction of the tidal oscillation is Q - ' = 6 W / ~ T dissipated per tidal cycle and W is the maximum stored energy of the tidal oscillation. At mean tidal range, Q - ' is about 0.31, which means that the damping of the oscillation is large. The total mean dissipation of tidal power in the Sound is about 192 MW and the mean dissipation per unit area is 0.060 W/m2. 2.2 Waves Work done by the wind blowing over the surface of Long Island Sound is used to generate waves, to set the surface layer of water in motion,

FIG.1. Map of Long Island Sound showing contours of water depth, the Mattituck Sill (MS), Falkner Island (FI), Cable and Anchor Reef (CAR), Eatons Neck (EN), and the Race (R).

72

HENRY J . BOKUNIEWICZ AND ROBERT B. GORDON

and to tilt the water surface. The Sound is almost completely surrounded by land and estimates of the sea state resulting from various wind conditions can be made by the methods developed tq evaluate waves generated in large reservoirs (Saville et al., 1962). Linsley and Franzini (1972) present these results in the following equations for the period t (sec), length A (m), and height z (m) of fully developed waves: =

0 57rj0.44~0.28

A = i.56t2

=0.064~1.~3.47

Here V (m/sec) is the wind speed at a height of 9 m above the water surface, and F (km) is the effective fetch. (The “effective fetch” is the average fetch projected onto the wind direction over a 90” arc, which is centered on the wind arrow.) The maximum horizontal component of the water-particle velocity at the bottom is, according to first-order wave theory, Um =

7FZ

t sinh 2n(d/A)

where d is the water depth. These relations may be used to evaluate urn throughout Long Island Sound for various wind conditions. At the center of the Sound, F = 48 km; the greatest value of F that can occur in the Sound is 56 km. In this range of effective fetch values, urn is quite insensitive to the actual value of F. The water depths at which u rn = 1 , 10, and 25 cmlsec for various wind speeds over Long Island Sound have been computed from the relations presented previously and are shown in Fig. 2. At any given wind speed V, a time T is required to raise a fully developed sea and for u, to reach the values given by the curves in Fig. 2. As V increases, T decreases slowly, as shown in Fig. 2. Large wind speeds over Long Island Sound are due to intense, fast-moving storm centers and, so, are of short duration. The length of time that the wind may be expected to blow at different speeds over the Sound was determined from a 50-day continuous wind record made at Falkner Island Light House during the winter of 1975-1976. The anemometer was at a height of 30 m; because of the elevation and exposed location, wind speeds at this site are about double those recorded at nearby shore stations. For each storm, the length of time that the wind exceeded successive 2.24 d s e c speed increments was measured. The averages of these durations are shown as curve D in Fig. 2; there is a rapid decrease in duration when wind speeds exceed I 1 m/ sec. The duration curve D crosses the curve showing the time required to raise a fully developed sea T at 13 m/sec. Stronger winds, when they do occur over the Sound, are of insufficient duration to excite the sea state they would otherwise raise. When the characteristic sea state for

SEDIMENT TRANSPORT AND DEPOSITION

73

Fic. 2. Maximum horizontal particle speed at the bottom in water of depth d due to waves on the surface produced by wind of speed V blowing over fetch F. Also shown is the time T required to generate fully developed seas and the duration D of winds of various speeds observed on Long Island Sound.

a given wind speed is not fully developed, urn remains below the values given in Fig. 2. When urnappreciably increases the water speed over the bottom above that due to the tidal stream, waves raised by the wind are expected to contribute to the excitation of sediment. In Long Island Sound the amplitude of the tidal stream oscillation is -25 cm/sec; if u , L 25 cm/sec, wave excitation of sediment dominates tidal excitation, whereas if urn 5 10 cm/sec, it is expected to be relatively unimportant. [Madsen (1976) shows that the bottom shear stress due to wave-induced flow is larger than that due to a steady current of the same magnitude. It has not been demonstrated that this generalization applies in the presence of the turbulence generated by a strong tidal stream; if it does, the wave-affected zone would extend to deeper water.] If we call that part of the bottom where it is possible for urn 2 10 c d s e c the “wave-affected zone,” then this zone extends out to water depths of about 18 m. It is unlikely that the bottom of the Sound in depths greater than -18 m will ever be significantly affected by the particle velocity due to waves on the surface, because winds of the requisite speed do not last long enough to raise a fully developed sea.

74

HENRY J . BOKUNIEWICZ AND ROBERT B. GORDON

To test these estimates of bottom disturbance by waves, a wave recorder was placed on Cable and Anchor Reef, where the water is 12.5 m deep, during the winter months of 1975. Water pressure was recorded continuously for 3 min each hour. Cable and Anchor Reef is surrounded by deep water and there is a long, open fetch to the east. The recorder is sensitive to pressure changes equivalent to 2 cm of water; the period and amplitude of all recorded signals greater than this were read out and analyzed. The wavelength of the wave responsible for a pressure fluctuation Sp is found by graphical solution of the equation: gt *-

27r

h tanh (2ndh)

and the wave height, from z = 2(Sp/tg) cosh (27rdlX)

where Sp is the amplitude of the pressure fluctuation due to waves. Because of hydraulic attenuation, the smallest wave height that can be detected is strongly dependent on the water depth; at a 12.5-m depth only waves with t > 3.5 sec can be recorded. The wave recorder was on station at Cable and Anchor Reef during the stormiest part of the year. Several major storms occurred during this period but water pressure fluctuations were detected only when the wind was blowing between NE and SE, the only direction with a relatively large fetch. Gale-force winds from other directions did not cause detectable pressure fluctuations at a 12.5-m depth; although they raise a steep sea, it is a sea with a relatively short wavelength. Particle speeds on the bottom calculated from the wave record are illustrated in Fig. 3 for the duration of one winter gale. The square of the wind speed, proportional to the wind stress, and the average value of u , over each 3-min recording period are shown for the period of time that the wind was blowing from the quadrant between NE and SE. After 0600 hr, the wind had backed into the NNW and, although there was no further diminution of wind speed, urnbecame zero. During the storm the wind speed persisted at 14 m/sec for a time (-5 hr) sufficient to raise a fully developed sea. The predicted u r nis 32 cm/sec; this level of urn is reached or exceeded for -4 hr during the storm. The highest reported wind speed reported was 17.4 m/sec; the corresponding predicted u , is 64 cdsec; the highest observed is 42 cm/sec. Thus, particle velocities at the bottom due to waves during this storm are generally in accord with the predictions that would be made with the aid of Fig. 2. The part of the bottom of Long Island Sound that is expected to be directly affected by wind-raised waves is generally that on the shoreward

75

SEDIMENT TRANSPORT AND DEPOSITION

,F\,A

I

--0-- urn Wind

\

E

3

250

- 200 - 150

;Ot

20

1

F'

10

-

100

4

en

1z 0

I""

't

0 08

12

16

20 HOUR

24

04

08

4-15March 1975

FIG.3. Maximum horizontal water particle speed at the bottom on Cable and Anchor Reef, as calculated from wave-recorder records, and the square of the wind speed measured at Eatons Neck during a winter storm.

side of the 20-m contour line, Fig. 1 . Disturbance of the bottom by waves will be greatest for easterly winds, since these have the longest fetch. In the deeper parts of the Sound particle speeds due to waves at the surface are small compared to the tidal stream. (Since the tidal stream is rotary, there is never a time when the current due to the tide is zero.) In water deeper than -20 m the bottom should remain undisturbed by storm-raised waves.

2.3 Currents Currents in the Sound are due to the tidal stream, to the estuarine circulation, and to wind stress acting on the water surface. Systematic surveys of the currents in the Sound have been made from time to time by the U.S. National Ocean Survey. Current meters have been placed in grid-pattern arrays for time intervals sufficiently long to reveal the principal tidal constituents of the current. Data obtained this way were used by G. A. Riley (1952, 1956) to describe the estuarine circulation of the Sound. The utility of these meter records in the study of sediment transport is limited because the observations were all made during the

76

HENRY J. BOKUNIEWICZ AND ROBERT B. GORDON

calm weather of the summer months, covered time spans short compared to the variability of the estuarine circulation, and had a sampling interval large compared to important fluctuations in water speed. Very few of the measurements were made close to the bottom. In a study of sediment stability and transport in the Sound we have obtained current meter data at 60 different locations over a total recording time of 57,672 hr. Records from two of the localities extend over 4 yr. All meters were set 2 m above the bottom on taut moorings except for one experiment done with a vertical array. Two types of meters have been used: type “H” records the mean speed and direction for successive 20-min intervals and type “G” yields instantaneous readings of speed and direction at a sampling interval that can be varied from 1 to 30 min. When it is desired to examine the current for only a few tidal periods, successive flow vectors can be displayed in a graph (as in Fig. 5 ) . When the flow over times long compared to the tidal period is of interest, the tidal constituents of the flow can be removed, to a good approximation, by finding the resultant current for successive 12.4-hr intervals. A “resultant vector diagram” shows the long-term, nontidal components of the flow that can be separated by harmonic analysis of the current meter data. Water near the surface of the Sound is easily set in motion and follows the direction of the wind stress closely. Figure 4 shows the resultant flow vectors (successive 12.4-hr sums of the recorded velocities) obtained from a vertical current meter array. The meters were spaced about equally through the water column. The resultant flow vectors for the upper half of the water column match the corresponding resultant wind vectors, whereas the flow in the lower half shows no detectable response to the wind stress. We reported earlier (Gordon and Pilbeam, 1975) that variations in the resultant flow vectors of the bottom water in Long Island Sound do not correlate with local wind stress on the Sound and that much of the variation may originatein processes occurring in BlockIsland Sound or on the continental shelf. A similar result has been reported by Elliott (1978)for the Potomac estuary. He found that local forcing could account for only 55% of the variance in the nontidal components of the flow. The lack of correlation between bottom water movement and wind stress has been confirmed by all subsequent current meter records, with one exception. During the passage of hurricane “Belle” over Long Island Sound in August 1976 a perturbation of bottom water flow directly related to wind stress was observed. The upper part of Fig. 5 shows successive 20min water displacement vectors recorded over a tidal cycle during calm weather near the geometrical center of the Sound. The current meter was 2 m above the bottom. The lower part of the figure shows the corre-

SEDIMENT TRANSPORT AND DEPOSITION

-20

- 10

77

0 10 c W k r n E 4

Surface

Currant

I

I

K) m h e c

10 cmhrs

I

FIG.4. Each line segment in the upper diagram shows the net flow of water over a 12.4hr tidal period. The current meters were at elevations of 0.88d (A), 0.46d (B),0.13d (C), and 0.08d (D),where d = 24 m is the water depth. The lower diagrams show resultant current vectors obtained from meter A (the one nearest the surface) and, for comparison, the resultant surface wind for the same time intervals. The surface water follows the wind stress closely; the flow near the bottom is unrelated to the wind stress. The meters were set 7.4 km north of Eatons Neck.

sponding vectors recorded during the hurricane. Wind vectors are also shown. The flow of bottom water is perturbed for about 5 hr during the time of greatest wind speeds starting at 2240 hours on the 9th. The ebb current is stopped and then reversed by the wind blowing from the northeast at speeds up to 23 m/sec. When the wind direction reverses, at about 0120 hr, the ebb flow of the bottom water is greatly accelerated. Thus,

78

HENRY .I. BOKUNIEWICZ AND ROBERT B. GORDON

km .-3

2120-09-

10

-4

km

-3

-2

-I

0

0 Wlnd,'Belle' +I

FIG.5 . Water displacement vectors measured 2 m above the bottom near the center of Long Island Sound during hurricane "Belle." The record starts at 1600RT09 August 1976. Vectors are identified by hour and day. A corresponding set of displacement vectors measured under calm conditions at the same location are shown above. Wind vectors measured during the hurricane are shown at the bottom of the diagram.

out of 57,672 hr of current meter recording of bottom-water flow, windrelated perturbation was observed for 5 hr, or 0.009% of the time. In the deeper parts of the Sound, below the wave-affected zone, the wind has no direct effect on water flow, either through wave-induced current or circulation caused by the wind stress at the surface. Wind-driven circulation is present in the upper part of the water column (above about middepth) along the central axis of the Sound. Current meter measurements have not been made in the wave-affected zone, but it is expected that wind-driven circulation extends to the bottom there. We have previously reported that, although most storms do not alter the bottom-water resultant flow vectors, the probability of large water speeds over the bottom is increased (Bokuniewicz et al., 1975b)when the wind stress is high. We now present further evidence on this effect. A periodogram calculated for a current meter record obtained in the western end of Long Island Sound is presented in Fig. 6. Three tidal peaks are

79

SEDIMENT TRANSPORT AND DEPOSITION

evident; aside from these, energy is distributed over a wide range of frequencies with no well-defined breaks or gaps. Hence, in analyzing the current meter data it is convenient to separate the observed flow into three components, the mean flow into three components, the mean flow over the duration of the record, the tidal constituents, and the residual flow, which we will call the fluctuating component. The tidal components are removed by a regression analysis over 15 or 29 days using 18 sinusoids of the tide in the Sound. Throughout the Sound the direction of the greatest tidal velocity is nearly E-W; this is taken as the direction of the x axis for computation. The fluctuating components of the E-W and N-S velocities, u’ and v’, for the current record of Fig. 6 are shown in Fig. 7. The average amplitude of the tidal stream at this station is U 20 cm/ sec. Figure 7 shows that u’ is an appreciable fraction of U throughout the record interval and that u’ = U on a number of occasions. When such

-

4.0

Time interval = 20 min

2500 observations h

2.4

I

Semidiurnal

a u

Quarter -diurnal

‘N 0

8

)I

0.8

‘N

E

0

.-C

Sixth-diurnal

h

.-cIn

5

-0.8

TI

:

2,

c

W Y

m

-2.4

3

-4.0

I I

- 3.2

- 2.4

-1.6

-0.8

0.0

Log (Frequency in CPH) FIG6. Periodogram calculated from a current meter record made in western Long Island Sound (7.4 km north of Eatons Neck) showing the dominant semidiurnal tide and two shallow water constituents (CPH, cycles/hr).

81

SEDIMENT TRANSPORT AND DEPOSITION

-<0.5

f0. u = 3.88

0.I s - 2 0

,

0

,

I

,

10

,

,

20

crn /sec

(a 1

-10

10

0

l&J!L.-

20

u=3.88

0.1

-20

-10

0

10

W (b) Velocity, cm/sec

20

E

FIG.8. (a) Polar histogram of the fluctuating velocity component of the data in Fig. 7. Shading indicated the frequency of occurrence of different velocities. (b) Distributions of the north-south and east-west fluctuating velocity compoonent. The normalized probability density function P is in units of oh',where u is the standard deviation and u' the fluctuating velocity component.

Fig. 9. The best correspondence between the form of the two records is obtained when u' is plotted with a 3-hr lag time. Work done on the water surface by the wind causes an increase in the level of the fluctuating component of the velocity near the bottom within a matter of hours even though the water is too deep to be directly affected by the wind (wind-

100

N W

-c E

W

50

1 7

eE

-

N

z

0

October 1974

FIG9. Square of the wind speed ( w 2 ,histogram) and mean square fluctuating velocity ( u ' ~measured ) 2 m above the bottom near the center of Long Island Sound. The wind-speed histogram is shifted to lead the current by 3 hr.

82

HENRY J. BOKUNIEWICZ AND ROBERT B. GORDON

driven net flow is found only in the upper third of the water column, as shown in Fig. 4). The mechanism by which local wind stress increases the amplitude of the fluctuating component of the velocity at the bottom has not been found. However, the resultant effects on the bottom of the Sound are large. We show in the previous article that the tidal dissipation may be more than doubled during periods of high local wind stress. Thus, it is clear that even in the deepest waters of the Sound, local storms can cause large changes in the stress on the bottom. It is expected that the stability of sediment deposits may be affected by these events and, since storms are more frequent and intense during the winter months, that there will be a strong seasonal variation in the disturbance of the Sound bottom. 2.4. River Flow

Vertical and horizontal salinity gradients in Long Island Sound cause a large estuarine circulation (G. A. Riley, 1952, 1956; Wilson, 1976; Gordon and Pilbeam, 1975). This circulation must be maintained against the friction due to the tidal stream turbulance; the power required is derived from the inflow of fresh water and amounts to -5% of the tidal power dissipation. Mixing of the low-density surface water with the higher density bottom water also requires power, but this is negligibly small compared to the power dissipated by friction. River flow entering Long Island Sound maintains the longitudinal salinity gradient which, in turn, drives the estuarine circulation. This circulation is responsible for the advective transport of sediment in the Sound. We have made observations of the mean flow of bottom water over long time intervals at several places in the Sound. Figure 10, for example, shows resultant flow vectors for bottom water near the western end of the Sound over an interval of 5 months. Westward flow of bottom water due to the estuarine circulation is the dominant movement, but the direction and magnitude of the flow is variable over long time scales. This variation is largely unrelated to local wind stress, as reported previously, and is also unrelated to river discharge. 2.5. Summary

Waves cause significant water velocities over the bottom of Long Island Sound within a "wave-affected zone" that is confined to a shoreside area where the water depth is less than -18 m. Wind-driven currents occur only in the upper third of the water column throughout the deep water of the Sound. The tide is the dominant source of power for bottom pro-

4

84

HENRY J. BOKUNIEWICZ AND ROBERT B. GORDON

TABLEI . POWER CHARACTERISTICS OF LONGISLAND SOUNDO Tide (tranquil weather) Wave power (on 150 km of shoreline, major winter storm) Dissipation beneath wave-affected zone: Average winter month Extreme winter month “10-year” storm Power used to maintain estuarine circulation Power used in vertical mixing L1

100 (%)

328 145

182 800 9 1

Expressed as percent of mean tidal dissipation of 192 MW.

cesses at water depths greater than those of the wave-affected zone. The distribution of the fluctuation component of the water velocity over the bottom is broadened under conditions of high local wind stress and (as shown in the following article) the tidal dissipation is increased. Lowfrequency components of the velocity fluctuations-which affect the longterm flow of bottom water-are of significant magnitude relative to the estuarine circulation, but are unrelated to local wind stress or river discharge. We discuss the characterization of temporal variation of bottom stress in the preceding article. The characteristic power parameters for bottom processes in the Sound are summarized in Table I.

3. SEDIMENT SOURCES Sediment may reach an estuary from rivers, erosion of the shoreline, by airborne transport, by biological production in the estuarine waters, and by inward transport of material from the sea floor outside. Benninger (1976) has shown that airborne transport is a negligible source of sediment in Long Island Sound and that the contribution of biological production is small compared to other sources. 3.1. Shoreside Erosion

Erosion of the shore under attack of waves is sufficiently rapid at many places around Long Island Sound to be a matter of concern in the management of shoreside property. The nature of the materials which constitute the Long Island Sound shoreline is summarized in Table 11. Silt-clay material that may be released upon erosion is present in the stratified drift, the interglacial deposits, and end moraines. End moraines account for only 17% of the erodible shoreline and this percentage has

SEDIMENT TRANSPORT AND DEPOSITION

85

not been significantly different in the past, since only a very small volume of eroded end moraine segments is found on the bottom (or subbottom) of the Sound (Flint and Gebert, 1976). To make a significant contribution to the accumulating mud in Long Island Sound, the shoreline composed of end moraine would have to be eroding at a rate an order of magnitude greater than that which is observed. Rapid erosion of the cliffs on the north shore of Long Island over the past 90 yr has been deduced from comparison of old and new maps (D. S. Davies, E. W. Axelrod, and J. S. O’Connor, undated report, Marine Sciences Research Center, State University of New York, Stony Brook, New York). Because the areas of greatest erosion are quite localized and the composition of the material making up the cliffs has not been systematically sampled, a reliable estimate of the amount of silt-clay sediment entering the Sound from this source cannot be made now. There is a strong possibility that it is significant but nevertheless smaller than the river input.

3.2. Exchange with the Continental Shelf Strong estuarine circulation at the eastern end of Long Island Sound results in the outflow of surface water and inflow of bottom water through the eastern passes at the rate of about 19,000 m3/sec (Riley, 1956), 41 times the fresh water inflow rate. Because these flows are so large, relatively small differences in the concentration of suspended sediment in the surface and bottom water will have a large effect on the calculated net flux through the eastern passes into the ,Sound. Regular measurements of the sediment concentration in these waters have been made by Bohlen TABLE11. NATURE OF MATERIALS FORMING THE SHORE OF LONGISLAND SOUND* Material Bedrock Outwash, stratified drift and interglacial deposits End moraine Beach sand Marsh

Area (km)

Amount

109 209

26 50

53 40 6

13 10

(96)

1

Source: published maps (Anonymous, 1967) and on-site inspections (J. Gebert, personal communication, 1976).

86

HENRY J. BOKUNIEWICZ AND ROBERT B. GORDON

(1975), who finds that the concentrations are much influenced by storms. Thus it appears that much of the material in suspension is locally activated by processes producing resuspension of material already on the bottom of the Sound. Consequently, a short-term movement of suspended sediments both in and out of the Sound with the ebb and flood of the tide are expected to occur. In the presence of this large and variable flux, the detection of the average, long-term, net flux through the eastern passes is a difficult matter and it seems unlikely that resources to permit measurements of the requisite frequency and duration will be available in the near future. Observations made by Benninger (1976) indicate a small inward net flux, but the data are not sufficient to fix its magnitude. 3.3. River Discharge Rivers are the major source of the mud that is accumulating in Long Island Sound. Two difficulties are encountered in establishing the amount of material they supply. The first is the great variation in the load of sediment carried (with most of the total transport of solids occurring in a few periods of very high discharge); stream-load data over a long period of time are required to establish a reliable average flux of sediment. Second, to get the total sediment input, the flux over the past 8000 yr is needed; it may not be possible to make a reliable estimate from modern stream-load measurements. The Connecticut River accounts for 71% of the fresh water entering Long Island Sound; it is only for this river that quantitative data on the sediment load are available. An early study of the sediment flux carried by New England rivers was made by Dole and Stabler (1909). Their estimate for the Connecticut River is 0.8 x 10' kg/yr. More recent data (Benninger, 1976) show that the concentration of suspended solids becomes so large during times of very high discharge that, for example, over the year May 1973-April 1974, nearly half the total flux of solids was passed during the 17 days the discharge exceeded four times the average flow. Since maximum discharge conditions vary greatly from year to year and continuous records of the load of suspended solids have been kept for only a short time, the average annual flux of solids is not yet known to good accuracy. Benninger's estimate based on data from the sampling station at the Enfield Dam is 3 x 10' kglyr; an estimate made by combining data from upstream sampling stations (Anonymous, 1975) is 5 x 10' kg/ yr. Benninger has shown that essentially all this material enters Long Island Sound because the area of growing marsh in the lower reaches of the river is too small to accumulate a significant fraction of this flux.

SEDIMENT TRANSPORT AND DEPOSITION

87

Geochemical data show that overland flow of water into rivers of the Susquehanna system in the northeast U. S. is negligible compared to that which enters as interflow and from the groundwater (Lewis, 1976). The same is probably true for the rivers draining into Long Island Sound and it is likely that most of the sediment they carry is derived from bank erosion (Gordon, 1979). Observations of the slumping and recession of the river banks show the requisite amount of erosion to be occurring along the Connecticut River and its tributaries (Anonymous, 1974). Since little sediment is trapped behind dams on the Connecticut River (Anonymous, 1975), the turbulence in the stream must be sufficient to keep the silt-clay fraction of the eroded soil in suspension until it reaches Long Island Sound. Bank erosion should be much less sensitive to agricultural use of the land than overland erosion so that the present sediment load of the Connecticut River may be reasonably representative of the load over the past 8000 years. This may be contrasted with Meade’s (1969) estimate that the sediment load carried by streams in the southeastern U. S. is now about four times more than it was before 1620. Suspended load data comparable to those for the Connecticut River are not available for the other streams entering Long Island Sound. All these have much smaller discharges and it is estimated that they contribute only about an additional 10% of the Connecticut River input to the Sound. 4. SEDIMENT TRANSPORT AND BOTTOM STABILITY

Bokuniewicz (1976) has shown that the most satisfactory two-component descriptions of the sediment in Long Island Sound is in terms of sand (particle size >70 Fm) and mud (all smaller particles). Sand is transported close to the bottom, probably as bed load most of the time, whereas mud is excited throughout the water column by tidal stream turbulence. Sand grains are not cohesive, but mud particles may adhere to each other because of electrostatic attraction or the presence of organic adhesive agents . It was estimated from wave data that only in a zone around the margins of the Sound where the water is less than 18 m deep are the particle velocities of waves a significant fraction of the tidal stream speed. Direct evidence of excitation of sediment by waves in this zone is found in turbidity measurements and in the structure of surficial sediment layers. For example, Fig. 1 1 shows a turbidity track made from deep to shallow water in an area where the bottom is mud and the tidal stream weak. There is resuspension of mud through the water column where the depth is less than about half the wavelength of the waves present at the time the track

80 -

P

v

9

*

s 70 82

10-

[r

a

W

-

td

s

> z w

- a

r_ 0

w cn 2 0 -

w tw

-

-

I

I I

a

-

-

W

0

10-

200 m

FIG. 11. Optical transmittance at a depth of I m measured along the track shown in profile below. Resuspension of sediment from the bottom bv waves results in lowered transmittance in shallow water

SEDIMENT TRANSPORT AND DEPOSITION

89

was made. Evidence confirming the estimated bounds of the wave-affected zone was obtained by R. C. Aller (personal communication, 1976) after passage of hurricane “Belle” in August 1976. This hurricane passed over the Sound near longitude 73” 10’ and caused wind speeds up to 41 d s e c . Box cores were collected by Aller along a transect across the Sound after the storm. His radiographs of these cores show that the sediment on the bottom was overturned to a depth greater than 10 mm shoreward of the 12-m isobath, but no evidence of disturbance was found in the samples from deeper water. Much of the bottom of the Sound within the wave-affected zone is sandy. In some localities where the shore is bedrock, deposits of muddy sediment form up to the surf zone. This is possible because mud is trapped in mats of algae that mantle the bottom near the shore. Muddy material also accumulates in the dredged shipping channels of the harbors along the Connecticut shore at an average rate of 4 m d y r , as computed from records of maintenance dredging (Bokuniewicz and Gordon, 1979). The surface area and volume of all of these deposits is small compared to the area and volume of the muddy sediment below the wave-affected zone. The rest of this article deals mostly with these deep mud deposits. 4.1. Sand Transport

The transport of sand in the Sound may be represented as a two-dimensional process in which sand moves as a thin layer over the bottom. Dispersion of sand from a small source in an estuary can be observed by use of sand grains marked either by irradiation or by chemical coatings. Sand fluxes over a wide area can be measured by observing the motion of large bed forms, such as megaripplies (commonly called “sand waves”). Methods of measurement are summarized by Hubbell (1964), who derived the basic equation relating sediment flux to migration rate. He points out that if grains eroded from the upstream sand-wave face are not deposited on the slip face, they do not contribute to the sand-wave advance. An extensive set of field and laboratory tests of the use of bedform migration to determine bed-load transport in rivers was made by Simons et a / . (1965). They derive the relation q b = (1 - p)Vs(fh) -I-C I

for the volume rate of bed-load transport per unit width q b , in terms of the porosity of the sand bed p, the speed of bed-form advance V , , and the sand-wave height h. The part of the sediment being transported that does not enter into sand-wave advance is C It is assumed that V , and

90

HENRY J . BOKUNIEWICZ AND ROBERT B. GORDON

h are constant across the channel, that p is constant, and that the sand waves are triangular in shape. Excellent agreement with observations was found when C was measured independently. In an estuary it is expected that C l will be small if the wavelength of the waves is large. When the lateral constraints imposed by a stream channel are removed, it is likely that a sand wave will not remain straight and at constant height while advancing a large distance. However, small displacements of large waves can be observed by precision bathymetric surveys and used to determine q b . This method has been used to determine the sand flux over the Mattituck still in eastern Long Island Sound (Bokuniewicz et ai., 1977); waves 1-m high and 25-m long were observed to advance at a rate of about 1 m/week. The corresponding sand flux is 1.0 m2/sec. The method has been refined by constructing a time series of detailed topographic maps of a 200 X 600-m section of a sand-wave field on the Mattituck sill (K. Zimmerman, personal communication, 1976). Successive cross sections of one-large sand wave are shown in Fig. 12. The crest advanced 6.7 m in 148 days and the sand flux was 0.7 m2/ sec. The westward flux of sand over the Mattituck sill is due to the excitation of the sand by the tidal stream and advection by the estuarine circulation (Bokuniewicz et al., 1977). According to this model, the sand

FIG.12. Observed displacement of a large sand wave on the Mattituck sill over 148 days.

SEDIMENT TRANSPORT AND DEPOSITION

91

flux should remain about constant during the year, but the increased amplitude of velocity fluctuations and associated increase in tidal dissipation due to storms will increase the amount of sand excited, and there should be a seasonal variation in q s even though the water is too deep for the bottom to be affected by waves. Continuation of the sand-wave measurements over a period of a year or more would permit detection of this effect.

4.2. Mud Transport The mineral constituents of mud from the bottom of Long Island Sound are silt- and clay-size particles. However, only a small fraction of the material collected in a fine-mesh net towed through the waters of the Sound consists of these individual mineral particles (Benninger, 1976). Instead, much larger aggregates consisting of mineral grains bound by organic matter dominate the collection. These particles are fecal pellets produced by benthic animals. Because of their large size the settling speed of these aggregates is much greater than those of silt or clay (Rhoads, 1 974). Figure 13 is an in situ photograph of a cross section through the sediment-water interface of the mud bottom of Long Island Sound. The uppermost layer of sediment is seen to be composed of pellets (size about 100-500 pm). Additional pellets, resolved as individual particles in the photograph, may be seen in the water above the bottom; a cloud of pellets is being swept out of an animal burrow on the right side of the photograph. Photographs and hardness measurements (Bokuniewicz et al., 1975a) made throughout Long Island Sound show that wherever the bottom is mud, or mud containing up to about 80% sand, it is blanketed by a layer of pellets about 10 mm thick. The average total volume of this mantle of pellets is estimated to be 2 x lo7 m3. [Data obtained by Rhoads et al. (1977) show that there is a seasonal variation in this volume.] The dry density of the layer of pellets is 0.33 Mg/m3 (Aller and Cochran, 1976); from the density and volume the mass of silt and clay held in the pellets is calculated to be about 13 times the average annual supply of sediment to the Sound. Benninger (1976) reports that fresh, unprocessed mineral material is found in the suspended sediment of the central Sound only when the river discharge is unusually large and the activity of the benthic animals very low. The capacity of the benthic animals to porocess sediment and produce pellets is, therefore, large compared to the rate of supply of fresh sediment. Animal processing rates for deposit feeders are on the order of 100 cm3 per animal per year (Rhoads, 1974). Animal dens-

FIG.13. Photograph of the cross section of the muddy bottom of Long Island Sound taken with the interface camera developed by Rhoads and Cande (1971). The length of the sediment-water interface spanned in the picture is 0.15 m.

SEDIMENT TRANSPORT AND DEPOSITION

93

ities are well above 1 x 104/m2on the muddy bottom of the Sound and they should be able to rework the annual supply of sediment many times in the course of a year. In fact, measurements of 234Thprofiles at the sediment-water interface show that animals such as Yoldia limatula and Nucula annulata feeding below the sediment-water interface in Long Island Sound reprocess the layer of pellets about once each month (Aller and Cochran, 1976). The evidence just presented shows that new silt-clay-size sediment entering Long Island Sound is rapidly incorporated into the surfcial layer of pellets that mantles the mud bottom. This layer is the source of material which is resuspended into the water above. The amount of suspended sediment in the water column over the mud bottom of the Sound is found to range from 0.01 to 1 kg/m2 under tidal excitation and to 2 kg/mz in major storms. The amount of material contained in the pellet layer is -3.3 kg/m2. The mantle of pellets on the bottom is fully excited into the water column only under extreme storm conditions. Otherwise it serves as a source of material available for resuspension into the water column. Newly supplied sediment must pass through the mantle of pellets before it is incorporated into the permanent sediment below. Growth of the mud bottom can occur only through this incorporation step, i.e., the conversion of pellets to cohesive sediments. It is hypothesized that this conversion occurs at the underside of the layer of pellets by the destruction of the organic binder between the mineral grains. Although the mechanism by which this is accomplished is an important step in the process of permanent sediment formation in Long Island Sound, it has, as yet, received little attention. Since a layer of unbound particles is almost always present on the muddy bottom of the Sound, and the formation of the permanent sediment (i.e., the sediment not subject to excitation by water movements) occurs below this layer, it follows that the detailed mechanics and local patterns of mud transport by currents in Long Island Sound have no direct bearing on the sedimentation process. In dealing with sediment formation it suffices to know that mud-size minerals are everywhere supplied faster than they can be incorporated into the permanent bottom sediment. Transport of mud in Long Island Sound requires, first, resuspension of pellets from the surficial layer on the bottom and then the advection and diffusion of the resuspended material. To describe the transport of suspended sediment quantitatively, the sediment concentration gradient in the vertical direction through the water column and the flux of material into (or out of) the bottom layer of unbound sediment must be known. Both quantities depend on the distribution of velocity fluctuations in the tidal stream and so, even in deep water, are sensitive to weather condi-

94

HENRY J. BOKUNIEWICZ AND ROBERT B. GORDON

tions. This is illustrated by the data displayed in Fig. 14, which show the integrated sediment content of the water column near the center of Long Island Sound during a period of gale-force winds. The fluctuating component of the current speed u’, measured at the observing site is also shown. A large increase in the amount of resuspended sediment is observed beginning at about 1500 hours, about 2 hours after u’ started to increase rapidly. (The measurements were terminated when the observing ship was forced to leave its station.) A method of calculating the total amount of sediment in the water column at any time has not yet been worked out and is a limitation on transport calculations. However. advective transport of the resuspended sediment can be evaluated for Long Island Sound because the velocities of the tidal stream and estuarine circulation are fairly well known. Average lateral diffusion coefficients have been estimated from the measured distribution of the fluctuating component of velocity measured by current meters (see tabulation). These Estimated Lateral Eflective Diffusion Coefficients

Summer Winter

K

=

13 m2/sec (E-W) 21

K = 12 m2/sec (N-S) 43

coefficients are less than those calculated from the salinity distribution (Riley, 1952) because they are based only on the fluctuating velocity component, i.e., velocity variations with periods of about an hour or less.

)O

1600

2400

HOUR (EST) FIG 14. Average fluctuating current speed (solid line) measured 2 m above the bottom near the center of Long Island Sound, and the total amount of suspended sediment in the water column.

SEDIMENT TRANSPORT AND DEPOSITION

95

They are greater in the winter months because of the broadening of the distributions of u' and v' under storm conditions ( u and v are E-W and N-S velocity components). In Long Island Sound the pellets at the sediment-water interface undergo frequent resuspension before their incorporation into the permanent sediment. Dispersion of resuspended pellets by both advection and diffusion is rapid. New silt-clay-size mineral matter introduced into the Sound will be rapidly distributed throughout the layer of pellets mantling the mud buttom. Chemical species adsorbed on silt-clay particles will be similarly dispersed.

5. SEDIMENT DEPOSITION AND DISTRIBUTION If the sources of supply, initial and boundary conditions, and the transport and deposition mechanisms are adequately defined, the resultant distribution of marine sediment can be deduced. This can be done qualitatively for Long Island Sound; some of the results can be put in quantitative form. The initial conditions at the start of the marine regime in the Sound at 8000-yr BP are shown in Fig. 15. Lacustrine mud had accumulated on part of the bottom of the lake formed behind the Mattituck sill. The sill itself, and the lake floor surrounding the lacustrine mud, was formed of deposits of glacial outwash. Most of this outwash, already submerged in fresh water, was not subject to reworking under the advancing surf zone. Small

FIG. 15. Boundaries of the lake that occupied the basin of Long Island Sound before the sea reached the elevation of the sadle on the Mattituck sill. The Connecticut River (CR) did not flow into this lake.

96

HENRY J. BOKUNIEWICZ AND ROBERT B. GORDON

end-moraine segments were exposed to the surf as sea level continued to rise; with the fines removed these became lag deposits of boulders, now appearing as shoals. Under marine conditions the major transport of outwash sand was accomplished, not by the surf, but by tidal energy and the estuarine circulation causing westward movement of sand over the Mattituck sill, transport evident today in the migration of the sandwave field that covers the sill. Since the strength and pattern of the tide and the estuarine circulation should have changed little since 8000-yr BP, steady sand migration into the central Sound is expected to have prevailed during this period. Throughout Long Island Sound the amplitude of the tidal stream exceeds the critical erosion velocity for both the mud-size range of sediment and the mantling layer of pellets on the mud bottom. Permanent sediment is formed at the bottom of the layer of pellets, but this layer is present only on the existing mud bottom. Any pellets deposited on a sandy bottom during slack water are swept away by the subsequent tidal flow. Thus, for mud to accrete on the bottom of the Sound, there must be existing mud deposits that serve to nucleate the deposition process. In the absence of any such nuclei, unbound sediment would accumulate until some other mechanism of deposition is initiated. This might be the formation of stable mud during times of nearly slack water, as described by Terwindt and Breusers (1972). In Long Island Sound the existing deposits of lacustral sediments were available to serve as nuclei for the deposition of the marine mud. Since the initial mud bottom area was much smaller than it is now while the sediment supply rate has remained about the same, the concentration of unbound sediment must have been greater in the early marine period of the Sound than it is now. The principal processes controlling the formation of muddy sediment in Long Island Sound are the upward growth of the mud bottom (with lateral spreading at its margins) at a rate controlled by the rate of incorporation of pellets into the permanent sediment and the westward transport of sand under the combined effect of tidal excitation and est-uarine circulation. A quantitative model of the transport and binding of sand has been developed by Bokuniewicz (1976, and this volume), who shows how the sand concentration gradients observed in the Sound can be calculated from the above initial conditions and the observed currents. According to hismodel, steady-stateconditionsare attainedquickly(inabout 30years); the variation of sand content through the sediment column should, therefore, be small. He predicts that during periods of storminess the sand transport is increased and that the thin layers of increased sand fraction found in cores of the mud bottom are probably due to changes in the

97

SEDIMENT TRANSPORT AND DEPOSITION

frequency of occurrence of storms over a period of years rather than to individual storms. 5.1. Rate of Deposition of Mud

The average rate at which silt and clay have been incorporated into the bottom of the Sound over the past 8000 years has been calculated from the measured marine sediment thickness, the sand content of the sediment, and the duration of the period of marine submergence (J. A. Gebert, personal communication, 1976). The calculated incorporation rates are shown in Fig. 16. There are large variations in this average rate over the area where muddy bottom is being formed. The largest rates are found just to the west of both the Mattituck sill and Stratford Shoal, where the transport of sand onto the accreting mud bottom in response to tidal excitation and estuarine circulation is largest. Low average rates of incorporation are found in the deepest water of the central basin. No explanation of this variation is found in the hydrographic characteristics of the region, as was suggested would be the case earlier. In calculating the average accumulation rate, it is assumed that marine sedimentation begins as soon as the bottom is covered by sea water. If marine sediment deposits were nucleated at only a few places in the Sound, the apparent accumulation rate would be greatest at those places because of the long time actually available for accumulation. In fact, large accumulations of marine

units

73-30'

i ~ ' g r n / r n yr ~

73-15

73.00'

TP.5'

72.30

FIG 16. The average rate of accumulation of marine mud on the bottom of Long Island Sound over the past 8OOO yr (J. Gebert, personal communication, 1976).

98

HENRY J. BOKUNIEWICZ AND ROBERT B. GORDON

sediment are found over those areas of the Sound bottom where lake sediments appear to be present, an observation in accord with the idea that existing deposits of lacustral mud were the nuclei for the subsequent deposition of marine material. It is unlikely that this is the only factor responsible for the variations of deposition rate shown in Fig. 16. It does not explain why the average accumulation rate is greater on the flanks than at the bottom of the central topographic trough of the Sound, for example. Areas of high average incorporation rate may be places that support large populations of sediment-processing benthic animals. Incorporation rates are greater in the areas of the mud bottom with relatively large sand content because the greater bottom hardness that results from the relatively large influx of sand creates a move favorable habitat. These observations suggest that the incorporation process is controlled by animal activity, but independent evidence on this point is lacking. TABLE111. SEDIMENTATION PARAMETERS FOR LONGISLAND SOUND Fraction of Sound bottom covered by mud Sediment supply rate from rivers: By weight Volume per unit area of permanent mud bottom Mass per unit area of mud bottom Sediment present in the Sound: Total volume of marine sediment Volume of lake sediment Marine mud: total volume mass (at p = 0.8 Mg/m3) Sedimentation rates, mean mud-incorporation rate: Volume per unit area of mud bottom Mass per unit area of mud bottom Reservoirs of unbound sediment: Average amount of unbound sediment: Total volume Volume per unit area of mud bottom Mass of mud contained (at p = 0.33 Mgl m') Mass per unit area of mud bottom Masslmass supplied by rivers per year Benthic processing rate Mud-supply ratelmud-processing rate Trapping efficiency Rate of sea level risehate of growth of mud bottom

56% 4.7 x 10' kglyr 0.33 mm/yr 0.26 kgl(m2 yr) 1.0 x 4.9 x 5.3 x 4.2 x

10" m3 10' m3 lo9 m3 lot2kg

0.37 mm/yr 0.29 kg/(m2 yr) 2 x 107m3 6 mm 6 x lo9 kg 3.3 kg/m2 13 yr 2.4 X 10" kg/yr 0.020 100% 9

SEDIMENT TRANSPORT AND DEPOSITION

99

5.2. The Sediment Mass Balance The processes of sediment transport and deposition are believed to have operated continuously with little change in rate since the start of the present marine regime in Long Island Sound at 8000-yr BP. The volume of marine sediment that has accumulated should be equal to the integrated supply of solid materials, which has been supplied principally by the rivers. The integrated flux of riverine sediment supplied to the Sound over the pasf 8000 yr at the present supply rate is nearly equal to the mass of mud which has accumulated in that time (Bokuniewicz et al., 1976; Gordon, 1979). Thus, it is likely that Long Island Sound traps all of the sediment delivered to it and may be accumulating some additional mud-size material from the continental shelf. (Note that this does not rule out the possibility of exchange of sediment between the Sound and the shelf.) At the present time the rate of upward growth of the mud bottom of the Sound is less than the rate of rise of sea level so that water depths in the Sound are not shoaling. Characteristic parameters of the Long Island Sound sedimentary system are listed in Table 111.

6. COMPARISON WITH OTHERESTUARIES

Schemes for the classification and intercomparison of the hydraulic characteristics of different estuaries have been devised (see, for example, Dyer, 1973, Chapter 2) and used extensively. Systematic intercomparisons of estuarine sediment systems have not been made, in part because the requisite data have not been gathered. The task of collecting these data is eased once the critical characteristics needed for intercomparison are established. It is suggested that a scheme for the intercomparison of estuarines can be based on the quantities listed in Tables I and IV. 6.1. Energy

A useful way of describing the forcing of estuarine sedimentary processes is in terms of the specific dissipation (wattdsquare meter), which will be a function of both time and location throughout the estuary. Direct, systematic measurements of the specific power are not likely to be available for many (if any) estuaries, so estimates of the specific power must be based on the characteristics of the forcing mechanisms. The most im-

100

H E N R Y J. BOKUNIEWICZ AND ROBERT B. GORDON

portant of these are waves at the water surface, river flow, wind-driven currents, and the tide. A first step toward a description of the forcing of estuarine sedimentary processes is determination of the relative importance of each power source for the estuary as a whole and for different times and places within the estuary. The specific dissipation due to wave power is strongly dependent on water depth and, therefore, will have sharply defined bounds in most estuaries. It is determined by the depth, the available fetch, and the intensity of the winds having sufficient duration to raise a fully developed sea. For Long Island Sound the wave-dominated zone is that in water shallower than 18 m; this constitutes 54% of the total area of the Sound. Within the wave-dominated zone the particle motion due to waves at the water surface is more effective in exciting sediment from the bottom than other causes of water movement. Large quantities of sediment may be set in motion by the waves and relatively small currents can then effect substantial transport of the material so excited. An example of an estuary in which wave-excited sediment is an important fraction of the total sediment available for estuarine processes is the Tay, where wave erosion followed by overland flow on bare mudflats exposed on the ebb of the tide results in large sediment concentrations in the water of the estuary (Buller et al., 1975). The specific dissipation due to river power is most likely to be important near the head end of an estuary, but in some estuaries where the discharge is very large and the tide weak, power from the inflow of fresh water may dominate throughout. The specific dissipation due to the fresh water flow is y dSa, where z i is the mean flow speed, S the slope of the water surface, d the depth, and y the unit weight of water. Long Island Sound has no significant area where the specific dissipation due to fresh water inflow is dominant. In the estuary of the Connecticut River it is expected that river power will be a significant fraction of the tidal power when the river is in spate, but detailed calculations have not been done. In addition to raising waves on the water surface, winds will set the surface layers of water in motion in the direction of the wind stress or, if the water is sufficiently shallow, set up a circulation pattern extending to the bottom. Pickard and Rodgers (1959) have shown, for example, how an up-estuary wind can set the surface layer of water in the Knight Inlet (B.C.) in motion against the estuarine circulation. Elliott (1978) has demonstrated the importance of wind stress in determining the circulation in the Potomac estuary. To have much influence on estuarine sedimentary processes, however, the wind-driven circulation must penetrate to the bottom, which is likely to happen only in relatively shallow estuaries. In

SEDIMENT TRANSPORT AND DEPOSITION

101

Long Island Sound wind-driven water movements of local origin do not influence the water flow over the deeper parts where the deposits of muddy sediment are located. Their effects in the wave-affected zone have not been investigated. When the tide is the principal source of the specific dissipation, as is the case in the deeper parts of Long Island Sound, it would be expected that estuarine sedimentary processes would be driven with a regular temporal variation determined by the tidal constituents. The data presented here and in the prece,ding article show that this is not the case, that the internal friction of the tidal oscillation is sensitive to the local wind stress. Large changes in the specific dissipation occur consequently, and this is responsible for much of the variability in the resuspension of the muddy bottom of the Sound. The phenomenon has not been studied in other estuaries but it is likely to be of general occurrence. Thus, because all of the principal estuarine power sources are subject to large temporal variation in most localities, a steady, regular estuarine sedimentary regime must be most exceptional, if it does occur at all. 6.2. Sediments

Sediment may be carried into an estuary by rivers, by erosion of the estuary shore, from the sea, or may be deposited in the estuary by the wind. Sediment may be produced within the estuary by biological processes. In some estuaries the introduction of sewage and industrial wastes may add to the sediment supply. The relative importance of these different sources varies greatly between estuaries and is one element to be included in a classification scheme. In Long Island Sound the supply of sediment from rivers dominates all other sources. (There remains the possibility of a significant supply from the erosion of the north shore of Long Island yet to be proved or disproved.) Examples of other estuaries where different sources of sediment are important are Chesapeake Bay, where shoreside erosion is an important sediment source (Schubel, 1971), and the Potomac, where samples of the resuspended river sediment contained only 10% mineral material and the balance was of biogenic origin (Meade, 1972, p. 95). Mud erosion, transport, and deposition occur principally over the existing mud bottom. Hence, it is useful to express the rate of supply of mud to the estuary in terms of input per unit area of mud bottom. The physical form assumed by the sediment in the mud-size range depends on the ratio of the sediment supply rate to the rate of processing by benthic animals, a rate subject to both short- and long-term fluctuations as the

102

HENRY J. BOKUNIEWICZ AND ROBERT B. GORDON

animal populations wax and wane. In Long Island Sound this ratio is small, with the consequence that nearly all the unbound sediment present is in the form of pellets produced by benthic animals. In many estuaries, fluid mud layers, layers of high density fluid with high concentrations of silt and clay particles, are found. The properties of these fluid muds are described by Einstein and Krone (1962). Fluid mud is expected to occur when the sediment supply-to-processing ratio is large. This may occur because the input of mud is large, the mud-bottom area fraction is small, or the population density of benthic animals is low. Estuaries having large quantities of fluid mud are the Chao Phya (Allersma et al., 1966), the Thames (inglis and Allen, 1957), and the Severn (Kirby and Parker, 1975). In the Chao Phya the rate of supply of new mud is unusually large, in the Severn the area of existing mud bottom is very small, whereas in the Thames the processing rate is thought to be low because of the heavy stress that has been placed on the animal populations by toxic effluents. Fluid mud is also found locally in many harbors. For example, the inner part of New Haven Harbor is virtually devoid of benthic animals (Rhoads, 1975) and the dredged ship channel is found to fill rapidly with fluid mud. The second ratio that determines the nature of the suspendible sediment in an estuary is the sediment supply rate relative to the rate of formation of permanent mud bottom (i.e., the incorporation rate). When this is > I , unincorporated sediment accumulates in the estuary. When this happens, the percentage of new mineral material in the layer of pellets may increase or the total volume of pellets may increase. This accumulation cannot continue indefinitely. An increase in the volume of pellets may reduce the sediment net supply rate (by increased sediment export to the sea) or the incorporation rate may increase as the concentration of granules increases. If the sediment supply and incorporation rates are independent of each other, a stable concentration of pellets could not be expected to obtain. Several processes may intervene to provide long-term stability of the thickness of the mantle of pellets. For example, incorporation may be accelerated if benthic animals increase the contact area between granules and permanent mud when the mantle of granules becomes thicker. Alternatively, incorporation may be subject to interruption due to excitation of granules by storms and these interruptions may be less effective as the concentration of granules increases. Evidence is lacking and this is a subject which deserves further attention. Long island Sound has substantial capacity to contain sediment delivered to it without alteration of its hydraulic regime. This capacity is increasing, and has increased throughout the present marine period, because the rate of rise of sea level is greatet- than the rate of growth of the

SEDIMENT TRANSPORT AND DEPOSITION

103

mud deposits on the bottom. If an estuary is initially shallow, has a large sediment supply rate, or is on a coast where sea level is falling, its capacity to store sediment may be fully utilized. Such an estuary is expected to export all of the sediment that is delivered to it. It will also be sensitive to any alteration of its configuration, such as might be produced by dredging or the stabilization of channels by training works. Meade (1972, p, 107) has suggested that the salt intrusion and the inflow of sediment in the Savanna River from the sea was much less before the navigation channel was dredged. Dredging in the Mersey was found to cause rapid shoaling of some channels rather than the deepening expected (McDowell and O’Connor, 1977, p, 257). In both cases it seems that the storage capacity of the estuary was fully utilized by existing sediment deposits. It is expected that retention of sediment entering these estuaries from their rivers now must be very small. If an estuary is to retain sediment delivered to it, necessary conditions are that the specific dissipation level be low enough to permit the sediment to enter the deposits on the estuary bottom and that there be sufficient capacity to accommodate the sediment without alteration of the hydraulic regime in such a way as to increase the specific dissipation. The amount of sediment retained can range between 0 and 100% among different estuaries. In those where the sediment is retained, some fraction of the deposited material will be in permanent deposits that remain undisturbed and the rest may be periodically resuspended in the water column. These characteristics, described by the parameters enumerated previously, will determine how effectively any given estuary transmits sedimentary materials, and substances that may be adsorbed on sediment particles, to the sea.

ACKNOWLEDGMENTS We thank Carol Pilbeam and Robert Meade for valuable discussion. Many of the data reported here were obtained with the aid of J. A. Gebert, P. Kaminsky, M. Reed, and C. Tuttle in the course of studies done for the United Illuminating Company of New Haven and the U . S. Army Corps of Engineers.

APPENDIX

Shape, volume, and tidal characteristics of Long Island Sound useful in calculations are summarized in Table A1 . The description is based on a division of the Sound into seven segments of 20 km in length and one, the westernmost, of 8.9 km in length.

104

HENRY J . BOKUNIEWICZ A N D ROBERT B. GORDON

TABLEA l .

GEOMETRIC AND TIDAL CHARACTERISTICS OF LONG ISLAND SOUNDO

Segment number Segment bounds

I

I1 Ill IV V VI VII VIII

72"02'7-72"17'0 72"17'0-72"31'4 72"31'4-72"45'7 72"45'7-73"OO'O 73°00'0-73014'2 73"14'2-73"28'4 73'28'4-73'42'6 73"42'6-73"47'4

A

W

263.7 399.8 620.7 662.2 518.1 387.3 261.0 87.1

13.18 19.99 31.04 33.11 25.91 19.36 13.05 9.81

d 2

d l

22.28 17.27 16.68 15.71 16.28 14.17 1 I .30 8.81

VI

22.70 17.85 17.47 16.63 17.27 15.21 12.37 10.03

v2

5.875 5.985 6.904 7.136 10.353 10.843 10.403 11.012 8.434 8.948 5.488 5.891 2.949 3.228 0.776 0.874

Ah1

Ah2

0.84 1.16 1.58 1.84 1.98 2.08 2.14 2.24

1.05

1.47 1.94 2.24 2.42 2.50 2.55 2.65

" Notation: A , segment area (km2), total area = 3199.9km2; W ,mean segment width (km);

2 1 ,mean segment depth at mean low water (MLW) (m); d 2 , mean segment depth at mean

sea level (MSL) (m); V I ,segment volume at MLW = lo9 m3, total volume = 51.182 x lo9 m3; v 2 , segment volume at MSL = lo9 m3, total volume = 53.917 x lo9 m3; Ah ,, mean tidal range (m); and A h 2 , spring tidal range (m). Minimum Cross-Section Areas of Passes (at ML W )

East River Plum Gut and Race East Point, Fishers Is. to Watch Hill

9.98 x lo3 m2 2.96 x los m2 2.87 x lo4 m2

REFERENCES Aller, R. C., and Cochran, J. K. (1976).234Th/238Udisequilibrium in near-shore sediment: Particle reworking and diagenetic time scales. Earth Planer. Sci. Lett. 29, 37-50. Allersma, E., Hoekstra, A. J., and Bijker, E. W. (1966).Transport patterns in the Chao Phya estuary. 10th Conf. Coastal Eng. Vol. 1, pp. 632-650. Anonymous (1967). "Engineering Geology of the Northeast Comdor, Washington, D.C., to Boston, Massachusetts, Coastal plain and surficial geology." U . S . , Geol. Sur. Misc. Geol. Invest. Map 1-514-B. Anonymous ( 1974)."Connecticut River Basin Bank Erosion Study" (Reconnaissance Report). New England River Basins Commission, Boston, Massachusetts. Anonymous (1975)."Connecticut River Basin Program, Part Ill of Phase I, Water Quality Reconnaissance for the Connecticut River Supplemental Study." U. S. Environmental Protection Agency Region 1. Benninger, L. K. (1976).The uranium-series radionuclides as tracers of geochemical processes in Long Island Sound. Ph.D. Thesis, Yale University, New Haven, Connecticut. Bohlen, W. F. (1975). An investigation of suspended material concentrations in Eastern Long Island Sound. J . Geophys. Res. 80, 5089-5100. Bokuniewicz, H. J. (1976).Estuarine sediment flux evaluated in Long Island Sound. Ph.D. Thesis, Yale University, New Haven, Connecticut. Bokuniewicz, H. J., and Gordon, R. B. (1979).Containment of particulate wastes at openwater disposal sites. I n "Ocean Dumping and Marine Pollution" (H. Palmer and G. Gross, eds.), pp. 109-129.Dowden, Hutchinson, Ross, Inc., Stroudsburg, Pennsylvania.

SEDIMENT TRANSPORT AND DEPOSITION

105

Bokuniewicz, H. J., Gordon, R. B., and Rhoads, D. C. (1975a). Mechanical properties of the sediment-water interface. Mar. Geo:. 18, 263-278. Bokuniewicz, H. J . , Gordon, R. B., and Pilbeam, C. C. (1975b). Stress on the bottom of an estuary. Nutiire (London) 257, 575-576. Bokuniewicz, H. J.. Gebert, J., and Gordon, R. B. (1976). Sediment mass balance o f a large estuary (Long Island Sound). Estuarine Coastal Mar. Sci. 4, 523-536. Bokuniewicz. H. J . , Gordon, R. B., and Kastens, K. (1977). Form and migration of sand waves in a large estuary. Mar. Geol. 24, 185-199. Buller, A. T., Green, C. D., and McManus, J. (1975). Dynamics and sedimentation: The Tay in comparison with other estuaries. In “Nearshore Sediment Dynamics and Sedimentation” (J. Hails and A. Carr, eds.), pp. 201-249. Wiley, New York. Dole, R. B., and Stabler, H. (1909). Denudation. U.S., Geol. Surv., Water-Supply Pap. 234, 78-93. Dyer, K. R. (1973). “Estuaries: A Physical Introduction.” Wiley, New York. Einstein, H. A., and Krone, R. B.(1962). Experiments to determine modes of cohesive sediment transport in water. J. Geophys. Res. 67, 1451-1461. Elliott, A. J. (1978). Observations of the meteorologically induced circulation in the Potomac estuary. Estuarine Coastal Mar. Sci. 6 , 285-299. Flint, R . F.. and Gebert, J. A. (1976). Latest Laurentide ice sheet: New evidence from southern New England. Geol. SOC. Am. Bull. 87, 182-188. Gordon, R. B. (1979). Denudation rate of Central New England determined from estuarine sedimentation. Am. J. Sci. 279, 632-642. Gordon, R. B., and Pilbeam, C. C. (1975). Circulation in central Long Island Sound. J. Geophys. Res. 80, 414-422. Hubbell. D. W. (1964). Apparatus and technique for measuring bedload. U.S.,Geol. Surv., Water-Supply Pap. 1748. Inglis, C . C.. and Allen, F. H. (1957). The regimen of the Thames estuary as affected by currents, salinities and river flow. Pror. Inst. Civ. Eng. 7, 827-868. Kirby, R.. and Parker, W. R. (1975). Sediment dynamics in the Severn Estuary: A background for studies of the effects of a barrage. In “An Environmental Appraisal of the Severn Barrage” (T. Shaw. ed.), pp. 35-46. Univ. of Bristol Press, United Kingdom. Le Lacheur, E. A., and Sammons, J. C. (1932). Tides and currents in Long Island and Block Island Sounds. U . S. Coast Geodetic Surv., Spec. Publ. 174. Lewis, D. M. (1976). The geochemistry of manganese, iron, uranium, lead-210 and major ions in the Susquehanna River. Ph.D. Thesis, Yale University, New Haven, Connecticut. Linsley, R. K., and Franzini, J. B. (1972). “Water Resources Engineering,” 2nd ed. McGraw-Hill, New York. McDowell, D. M., and O’Connor, B. A. (1977). “Hydraulic Behavior of Estuaries.” Wiley, New York. Madsen, 0. S. (1976). Wave climate of the continental margin: Elements of its mathematical description. In “Marine Sediment Transport and Environmental Management” (D. J. Stanley and D. J. P. Swift, eds.), pp. 65-87. Wiley, New York. Meade, R. H. (1969). Errors in using modern stream load data to estimate natural rates of denudation. Bull. Geol. SOC.A m . 80, 1265-1274. Meade, R. H. (1072). Transport and deposition of sediments in estuaries. Mem., Geol. SOC. Am. 133. Pickard, G. L., and Rodgers, K. (1959). Current measurements in Knight Inlet, British Columbia. J. Fish. Res. Board Can. 16, 635-678.

106

HENRY J . BOKUNIEWICZ A N D ROBERT B. GORDON

Redfield, A. C. (1950). The analysis of tidal phenomena in narrow embayments. Papers Phys. Oceanogr. Meteorol. 11, 1-35. Rhoads, D. C. (1974). Organism-sediment relations on the muddy sea floor. Oceanogr. Mar. Biol. 12, 263-300. Rhoads, D. C. (1975). “Benthic Monitoring Study, New Haven Harbor; Base Line Studies 1974.” Report submitted to the United Illuminating Co. of New Haven, Connecticut. Rhoads, D. C., and Cande, S. (1971). Sediment profile camera for the study of organismsediment relations. Limnol. Oceanogr. 16, 110-1 14. Rhoads, D. C., Aller, R. C., and Goldhaber, M. B. (1977). The influence of colonizing benthos on physical properties and chemical diagenesis of the estuarine sea floor. Belle W. Baruch Libr. Mar. Sci. 6 . Riley, G. A. (1956). Oceanography of Long Island Sound, 1952-1954. 2. Physical oceanography. Bull. Bingham Oceanogr. Collect. 15, 15-46. Riley, G. A. (1952). Hydrography of Long Island and Block Island Sounds. Bull. Bingham Oceanogr. Collect. 13, 5-39. Saville, T. S., McClendon, E. W., and Cochran, A. S. (1962). Freeboard allowances for waters in inland reservoirs. J . Waterways Harbors Div. Am. SOC.Civ. Eng. May, pp. 93-124. Schubel, J. R. (1971) “The Estuarine Environment.” Am. Geol. Inst., Short Course Lect. Notes, Washington, D.C. Simons, D. B., Richardson, E. V.,and Nordin, D. F., Jr. (1965). Bedload equation for ripples and dunes. U.S. Geol. Surv.,Prof. Pap. 462-H. Terwindt, J. E. J., and Brewers, H. N. C. (1972). Experiments on the origin of flaser lenticular and sand-clay alternating bedding. Sedimentology 19, 85-98. Wilson, R. (1976). Gravitational circulation in Long Island Sound. Esruarine Coastal Mar. Sci. 4, 443-453.

SAND TRANSPORT AT THE FLOOR OF LONG ISLAND SOUND* HENRYJ. BOKUNIEWICZ Marine Sciences Research Center State University of New York Stony Brook, New York

I. 2. 3. 4. 5. 6.

7.

Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Background . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Long Island Sound . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Sediment Transport . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Formation of the Transition Zone . . . . . . . . . . . . . . . . . . . . . . . . Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Summary and Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

107 107 110 113 116 122 124 126

1. INTRODUCTION

If we think of an estuary as a machine for handling sediment particles, then the products of this machine are the sediment deposits that line the estuary floor. Estuarine sedimentary processes undergo not only tidal and seasonal cycles, but also are disturbed occasionally by major storms or floods. Because of this extreme variability, it is very difficult to accurately predict the ultimate fate of sediment particles from measurements of sedimentary processes over short periods of time. The sedimentary deposits, however, are the results of these many and variable transport processes that have been active over long periods of time. Sediment deposits are thus manifestations of the long-term behavior of the estuarine sedimentary system, and, as such, they may be useful in evaluating the relative importance of tides or winds in controlling the sediment budget of the coastal zone. The patterns of sedimentation may be helpful, for example, in deciding whether or not more material is moved episodically during storms than is transported routinely by the tides. This article deals with the distribution of sand and mud in a large estuary (Long Island Sound) and discusses the processes that control the sand-mud transitions.

2. BACKGROUND In many coastal areas the regional sedimentary patterns seem to be controlled by the tides. The correlation between the strength of the tide * Contribution No. 231 of the Marine Sciences Research Center. 107 ADVANCES IN GEOPHYSICS, VOLUME

22

Copyright 8 1980 by Academic Press. Inc: All rights of reproduction in any form reserved. ISBN 0-12-018822-8

108

HENRY J. BOKUNIEWICZ

and the resulting grain-size distribution in the sediments has been recognized in, for example, Barataria Bay, Louisiana (Krumbein and Aberdeen, 1937), Buzzards Bay, Massachusetts (Moore, 1963), the Northumberland Straits (Krank, 19761, and the English Channel (Stride, 1963; Pingree and Maddock, 1977). It has been the goal of students of marine sedimentary processes to interpret such patterns in the marine environment with quantitative models of the transport phenomena. These investigations were hampered, however, by inadequacies in the existing, fluvial sediment transport formulas to handle unsteady flows. When faced with the vagaries of sedimentary processes in the coastal sea, investigators evolved new concepts of sediment transportation from the long tradition of fluvial studies. One of the most fruitful developments was the concept of the unidirectional transport of particles by oscillatory or intermittent currents. The onshore movement of fine-grained suspended sediment has been discussed for the U. S. Atlantic coast by Meade (1969) and Hathaway (1972), and for the Dutch Wadden Sea by Postma (1961). The unidirectional transport of sand by tidal currents has been documented in a variety of coastal environments by the form and migration of bed forms (Guilcher, 1967; Salsman et al., 1966; Jones et al., 1965; Belderson and Kenyon, 1969; Caston and Stride, 1970; Ludwick, 1972; Stride, 1974; Bokuniewicz et al., 1977). Several mechanisms have been discussed by which the tides produce a net sediment flux in one direction. One such mechanism is referred to as the “lag effect” and depends upon the fact that the sediment load of a current does not adjust itself instantaneously to changes in the current speed, but rather lags behind the current variations. As explained by Van Straaten and Kuenen (1958), a net movement of sediment results from this behavior when it occurs in tidal streams that vary in strength from place to place. The maximum tidal current speeds typically decrease into an estuary and the lag effect produces a net transport of sediment from regions of swifter tides to areas of slower tides. The lag effect is relatively more important for fine-grained suspended sediment than for sand because changes in the fine-grained sediment load usually requires particles to travel through the entire water column by settling or turbulent diffusion. The coarser fraction settles rapidly and is typically transported near the sea floor. As a result, the lag times are greater for fine-grained particles. The lag times for sand grains may usually be neglected, and the oneway motion of sand depends upon another mechanism that is due to the nonlinear relationship between the sediment flux and the speed of the transporting current. The sediment flux goes as the current speed to some power, which is typically found to be about 3 (Bagnold, 1966; Colby,

SAND TRANSPORT AT THE FLOOR OF LONG ISLAND SOUND

109

1964; Nordin, 1975). Small differences in the speeds of the ebb and flood tides at the same place, therefore, produce large net sediment fluxes in the direction of the swifter current. Tidal asymmetries may result from the superposition of a small residual current on the tidal stream (e.g., Stride, 1963; Meade, 1969). This mechanism is important in estuaries where the tides are superimposed on an estuarine circulation. Tidal asymmetry may also be enhanced by the geometry of the estuary. Shoals at the mouths of many estuaries typically deliniate ebb-dominated or flood-dominated channels and the net transport of sediment is partitioned accordingly (Swift and McMullen, 1968; Kline, 1970; Ludwick, 1972). Postma (1961) also discusses the net transport of sediment resulting from the distortion of a tidal wave moving into shoaling water. Another type of asymmetry in the transport of sediment is produced by the enhancement of the current temporarily by wind-induced flows. Much, if not most, of the motion of sediment may take place during infrequent storms. There is an increasing recognition of the importance of occasional storms or floods in controlling the sediment budget of the coastal zone. For example, more sediment was transported in the Chesapeake Bay during a few days of flooding than during many years of normal river flow (Schubel, 1974; Nichols, 1977). The development of predictive formulas for marine sediment transport is plagued not only by the usual problems, such as predicting the onset of grain motion, but also with problems arising from the great irregularity of marine currents. Many field and laboratory experiments have been successfully applied to the first type of problems (e.g., Sternberg, 1972; Miller et al., 1977; Young and Southard, 1978). Advances are also being made on the second type of problem as, for example, in the study of drag coefficients in natural tidal flows (Ludwick, 1975a; Kachel and Sternberg, 1971). Nevertheless, few attempts have been made to calculate the rates of sediment transport from observations of the near-bottom current velocities (McCave, 1971; Ludwick, 1975b; Gadd et al., 1978). Investigations of marine sediment transport have also been extended to both shorter and longer time periods than are normally considered when applying deterministic formulas. On the shorter time scale, it has been recognized that whereas the deterministic formulas deal with timeaveraged quantities, the entrainment of sediment actually occurs intermittently during brief periods of very high bottom shears. The importance of these events (called “bursting phenomena”) has been discussed by Sutherland (1967), Gordon (1975), McCave (1970), and Jackson (1976). Toward the other end of the spectrum, detailed, deterministic models are inappropriate for describing sediment motion over very large areas and long periods of time. Swift et al. (1972) have presented a statistical, or

110

HENRY J. BOKUNIEWICZ

diffusion model for the study of unsteady transport over large areas of the sea floor. This article will be a discussion of the transport of sand in Long Island Sound in terms of some of these developing elements in marine sediment transport. 3. LONGISLAND SOUND

Long Island Sound (Fig. 1) occupies a trough between the bedrock surface that dips southward from the fall line in Connecticut and a wedge of Cretaceous coastal plain sediments underlying Long Island. The Wisconsin glaciation left much of this region covered with a blanket of glacial drift. The deposition of marine mud in the Sound is part of a regional pattern of onshore transport of fine-grained suspended sediment over the Atlantic shelf (Meade, 1969; Hathaway, 1972). The accumulation of marine muds in the Sound has been investigated by Bokuniewicz et al. (1976). The Sound was established as an arm of the sea about 8000 yr BP and, when this happened, fine-grained sediment particles from the shelf waters and the rivers of New England began to be transported into the Sound and incorporated in deposits of marine mud. The floor of the central and western Sound is the top of a layer of mud many meters thick. The general character of the surficial sediments in Long Island Sound is known from the published data of Buzas (1965), McCrone et al. (1961), Bokuniewicz et al. (1976), and the reports prepared by Donohue and Tucker (1970), Ali and Feldhausen (1975), and the U. S . National Marine Fisheries Service (1974). The floor of the eastern Sound is composed of reworked, glacial outwash sand, and the sandy bottom is also found along the exposed shoreline. A map of the sand content of the bottom sediments has been prepared by Schubel and Wise (1979) from the published and additional unpublished data. This map (Fig. 2) represents over 400 analyses. Sand covers about 44% of the Sound floor and is the dominant component over about 75%. The major feature of the sand distribution is the transition from the sand floor of the eastern Sound to the nearly sand-free silts of the central Sound. The transition zone occurs over a distance of about 20 km. Within the transition zone the sand content increases uniformly to the east. Contours of constant sand fraction run approximately north and south in this region. The eastern boundary of the transition zone is a sand ridge (Mattituck Sill, Fig. 1) that crosses the Sound from north to south. The U. S . National Marine Fisheries Service (1974) has

41.k

--

CONNECTICUT

LONG ISLAND II I II I l l 1 1 1 1 I I

73.45'

,&

...

20'I-

30'

11111

15'

IIIIIIIII

73-00'

IIIIIIIIIIIII 1 1 1 1 I I 1 1 1 1 1 1

45'

30'

IIII1I11111111 11111

15'

72-00'

FIG. 1 . Long Island Sound. The bathymetry is contoured in meters and the location of Mattituck Sill is shown by the shaded area.

SAND TRANSPORT AT THE FLOOR OF LONG ISLAND SOUND

I13

collected 43 bottom samples within the sand-mud transition zone and has determined the grain-size distribution of these samples as the weight percent in each of 10 phi-interval categories. For the present study, these data were subjected to principal components analysis, as suggested by Davis (1970). The results are given in Table I. Over 50% of the variance among samples was accounted for by a single parameter. This parameter is essentially the weight percentage of sand particles between 0.062 mm and 1 mm in diameter. The correlation coefficient between the new parameter and the percent sand is 0.94. This analysis supports the hypothesis suggested by Buzas (1965) that the sediment of Long Island Sound is a mixture of two populations of particles-sand and silt. The distribution of sediment types might be effectively described by describing the distribution of sand.

4. SEDIMENT TRANSPORT Both sand and silt are regularly transported by the currents in Long Island Sound. The Sound is a large estuary and the flow is dominated by a strong, semidiurnal tide. The set of the tide is along the long axis of the Sound, which runs approximately east and west. The tidal currents may exceed 250 cdsec through the eastern passages but are typically between 20 and 50 cm/sec in the central and western Sound (Gordon and Pilbeam, 1975). These tidal currents are superimposed on a weak estuarine cirTABLEI" ~

+-interval category

~

~

Principal components 1

2

3

4

5

6

7

8 ~

-3--2 -2--1 - 1-0 0- 1 1-2 2-3 3-4 4-5 5-6 6-7

Cumulative % of trace:

0.00 0.00 0.01 0.21 0.55 0.41 -0.12 -0.59 -0.29 -0.20

-0.01 -0.01 -0.02 -0.23 -0.52 0.66 0.45 -0.14 0.10 -0.06

0.00 0.01 0.01 0.03 0.00 0.07 0.11 0.82 0.13 -0.51 -0.48 -0.08 0.80-0.12 -0.17 -0.20 -0.22 -0.02 -0.15 -0.00

0.03 -0.33 0.24 -0.73 0.00 -0.63 -0.11 0.04 -0.00 -0.52 -0.33 0.50 -0.20 0.24 -0.10 0.03 -0.06 0.16 -0.11 0.06 0.15 -0.10 0.05 -0.16 0.04 0.03 0.15-0.13 -0.65 0.14 -0.03 0.08 0.49 0.22 -0.69 -0.23 0.51 0.13 0.54 0.39

9

~

10

_

0.46 -0.29 0.13 0.75 0.23 -0.56 0.32 0.10 0.31 0.07 0.31 0.07 0.29 0.06 0.32 0.06 0.17 0.07 0.45 0.09

-

51.45 81.99 94.44 98.53 99.63 99.83 99.91 99.98 99.99 100.0

Analysis performed on data from National Marine Fisheries Service (1974).

_

114

HENRY J . BOKUNIEWICZ

culation (Gordon and Pilbeam, 1975; Wilson, 1976). Saline bottom water flows westward into the Sound at net speeds of 5-10 c d s e c . The measured flow velocities are also characterized by large amplitude fluctuations. The fluctuating velocity component has been described by Bokuniewicz et al. (1975). The perturbations have been measured with periods ranging from minutes to several days. The amplitude of the fluctuations does not correlate with the phase of the tide, as observed elsewhere by Gordon (1979, but increases in amplitude during windy periods. During storms the mean flow velocity is little affected, but the shortperiod fluctuations increase in intensity (Bokuniewicz and Gordon, this volume, pp. 75-82). Much of the Sound is sufficiently deep that the bottom remains undisturbed by wind-generated waves; even in deep water, however, the Sound floor may be disturbed by increased, water velocity fluctuations generated during storms. The fluctuations significantly enhance tidal resuspension (Bokuniewicz and Gordon, this volume, p. 94). The water column of the central Sound may contain over 200 mg/cm2 of suspended material (Bokuniewicz and Gordon, this volume, p. 94); high concentrations are attributed to frequent resuspension of the silt floor. Using sediment traps, McCall (1977) has measured resuspension rates of from 85 to 292 mg/cm2-day. The Sound floor in this area is a featureless plain, and cross-sectional photographs at the sediment-water interface show a layer of fecal pellets about 1 cm thick covering the bottom. [The cross-sectional photographs were taken with the interface camera developed by Rhoads and Cande (1971).] The dry density of this layer is 0.33 gm/cm3 (Aller and Cochran, 1976), and the resuspended material is drawn from this layer of fecal pellets. Typically, the tides are capable of resuspending a layer several millimeters thick. The maximum turbidity level was observed during a gale (Bokuniewicz and Gordon, this volume, p. 94), and the calculated suspended sediment concentration corresponds to a resuspension of about 6 mm of the layer of granules. Whereas the top of this layer is resuspended every tidal cycle, the transition of this material into the permanent mud deposit occurs at the bottom of the layer, presumably by a combination of physical and chemical disagglomeration. In the central Sound, marine mud is accumulating at a rate of about 1 mm/yr (Bokuniewicz et al., 1976). When the sand content of the sediment rises above 90%, sand waves appear on the Sound floor (Bokuniewicz et al., 1977). Sand waves that cover Mattituck Sill are composed of fine sand with a mean grain diameter of about 0.3 mm. Divers have observed sand grains moving over the bottom in this area and have found ripples superimposed upon the sand waves. The rate of migration of the sand waves has been documented at two locations on the sill by Bokuniewicz et al. (1977) and Karen Zim-

SAND TRANSPORT AT THE FLOOR OF LONG ISLAND SOUND

115

merman (personal communication, 1976). Changes in the position of selected sand waves were seen in repeated bathymetric surveys over periods ranging from weeks to months. From these observations the net sand flux across the sites was calculated; assuming a bed porosity of 0.6, the net sand flux was about 1.0 gm/m-sec to the west. The net westward sand flux has been attributed to the superposition of an estuarine circulation on the strong tidal currents. The way in which this situation results in the net, one-way transport of sand may be made more clear with the aid of a simple representation of the transporting currents and the resulting sediment flux. The currents in Long Island Sound may be decomposed into three components, the long-term mean flow component, the periodic, tidal component, and random variations in the flow. Over short times the bottom currents may be approximated well as a one-dimensional flow in the east-west direction with the tidal component represented by a simple sinusoid so that (4.1)

6 = li

+ uo sin(ot) + u’

where uo is the maximum tidal velocity during the period of interest, o is the semidiurnal tidal frequency, li is the mean flow over several tidal cycles, and u’ stands for the nonperiodic variations in the flow. The flow velocity 6 is an average velocity near the sea floor that will be used to predict the flux of sediment. Before proceeding, it is appropriate to discuss briefly the period over which the averaging is done and, conversely, the range of velocity fluctuations to be included in u ’ . Velocity fluctuations in coastal waters may be divided into two groups, those governed by internally defined time and length scales and those characterized by externally imposed scales. Fluctuations controlled by internal scales are expected to be inherent in all geophysical flows and to be definable by a small number of flow parameters, such as the friction velocity or some average velocity. Only if this is true will laboratory investigations and the theory of sediment transport be applicable in the field, because then the number and intensity of intermittent, high velocities that actually move the sediment grains are adequately represented by some flow parameter, say the mean velocity. The amount of sediment moved and the rate at which it moves is also a function of the characteristic flow parameter. At some scale, however, the characteristics of the fluctuating velocity become dependent upon the size of the body of water and the duration or periodicity of the driving forces. The amount of sediment moved by these externally controlled velocity variations must be calculated directly since they cannot be implicitly represented in general sediment transport formulas. The tilde over the u on the left side of (4.1)

116

HENRY J. BOKUNIEWICZ

indicates, therefore, an average over the period of the longest internally controlled fluctuation. This period of time may be estimated as follows. The largest internally controlled eddies in an estuary cannot be larger than the smallest dimension of the estuary, which is usually the water depth D , and the characteristic speed of these eddies is expected to be on the order of one-tenth the maximum tidal speed uo (Tennekes and Lumley, 1972). Hence, the division between internally and externally controlled fluctuations should occur at periods of about dlu’ where u’ 0 . 1 ~In ~ .a tidal estuary, the maximum tidal speeds are usually on the order of decimeters per second and the water depth is typically about 10 m. As a result, li is an average velocity over an interval of about lo3 sec. Many useful transport formulas describe the sediment flux as proportional to some measure of the fluid power as, for example, the near-bottom flow velocity cubed (Colby, 1964; Bagnold, 1966; Inman et al., 1966; Sternberg, 1972; Nordin, 1975). The proportionality constant must be evaluated in order to calculate the absolute sediment flux. In time-varying flows, the difficulties involved in predicting the proportionality constant make this approach unsuitable for general application (Sternberg, 1968; Ludwick, 1975a). To illustrate how the net flux of sand arises from the superposition of current components, however, the assumption that the sediment flux is proportional to the near-bottom velocity cubed is adequate. Thus, cubing (4.1) and averaging over several tidal cycles, we obtain an expression proportional to the average sediment flux j ;

-

(4.2)

-

j a C ( l . 5 ~+ ~u2 + 3u”)

In deriving this expression, U ’ was assumed to be symmetrically distributed about zero and to be statistically independent of the phase of the tide. For the currents near the floor of Long Island Sound, these characteristics of u’ have been demonstrated with field data (Bokuniewicz et al., 1975). The first term of (4.12) dominates the sum, and the relative magnitudes of the respective terms are 10: 1 : 1 . As a result, the net sand flux is approximately proportional to 1.5 a u f .The direction of the average sand flux is controlled by the sign of the mean velocity although the magnitude of the sand flux is dominated by the amplitude of the tidal velocity. 5 . FORMATION OF THE TRANSITION ZONE

The sand-mud transition zone may be described in terms of the net flux of sand over the accreting deposits of marine mud. Under excitation by the currents sand moves westwardly from the eastern Sound by advection and turbulent diffusion until it is incorporated into the accreting

SAND TRANSPORT AT THE FLOOR OF LONG ISLAND SOUND

117

mud deposit and immobilized. Anywhere within the transition zone the amount of sand in transport is ch where h is the thickness of the surficial layer of sediment that is regularly disturbed by the currents and c is the concentration of sand in this layer. The thickness of the layer of mobile sediment is typically several millimeters but may extend to over a centimeter during major storms (Bokuniewicz and Gordon, this volume, p. 93). The concentration of sand within the layer of mobile sediment is governed by a mass-balance in a control volume hdxdy:

(5.1)

aclat h dx dy = -dUc/dx h dx dy -dVcI@ h dx dy

lo rh

-

awclaz dz dx d y

where U , V , and Ware the instantaneous velocities of sand grains in the

x, y , and z directions. For simplicity, h is a constant and will be assumed to be 1 cm. A variable h may be handled mathematically, but in general, this cannot be done without loss of the simplicity of an analytical solution. The sand concentration, c is the instantaneous concentration of sand in the mobile layer, regardless of whether or not that sand is in motion at any particular time. In applying the mass-balance equation to the east-west, sand-mud transition zone, the gardients in the y direction (north-south) will be neglected because the contours of constant sand fraction run approximately north and south in this area. The instantaneous sand-grain velocities may be written as the sum of three components; the mean value over many tidal cycles, the tidal component, and a fluctuating component, (5.2)

u = u + % + U'

where % is composed of functions of time with tidal periods. Likewise, c may also be decomposed into mean, periodic, and fluctuating components.

(5.3)

c

= E

+ (e + c'

Since the typical sedimentation rate is less than about a millimeter per year, the mean components of U and c represent averages over about 10 yr. Sediment deeper than h is here assumed to be part of the permanent mud deposit; the fraction of this sediment that in fact may be mobilized by occasional very severe storms or bioturbation is neglected. The formation of the mud deposit occurs at a rate of - Wo and W, is the velocity at which sand grains pass into the permanent deposit of marine mud and are removed from transport.

118

HENRY J. BOKUNIEWICZ

Substituting (5.2) and (5.3) into the mass-balance equation and taking the time average over many tidal cycles (i-e., years), a simplified massbalance equation may be written to represent the steady-state situation (5.4)

k a2ddX2 -

u a dax

- c auiax - g c =

o

where g = - Wdh and k is an effective eddy-diffusion coefficient. The eddy-diffusion coefficient represents the fluctuating and periodic transport of sand [i.e., k d2CIdx2 is assigned to replace a(U’c’)laxand d(%%)l ax; the brackets indicate time averages]. This concept has successfully been used for prediction of estuarine mixing processes (e.g., Bowden, 1963). In obtaining (5.4), terms involving the derivatives of the correlation between periodic and fluctuating quantities have been neglected. Specifically, d(U’%)lax = d(c’%)/ax = 0. As mentioned earlier, there is a lack of correlation between the fluctuating and periodic water velocity components. Because the variations c’ and U’are due to fluctuations in the water flow, it may not be unreasonable to assume that (c’%) and (U’%) are zero. The difficulty in establishing this condition arises because only (c) is measurable. A similar problem occurs in calculating a salinity balance for an estuary where the assumption is also made that the timeaveraged cross-products of periodic and fluctuating quantities are zero (e.g., Dyer, 1974). The boundary conditions are chosen to approximate the present situation where sand is supplied to the central Sound from Mattituck Sill which remains mud free; E = 1 at x = 0 and C remains finite as x becomes large. When the coefficients U , k , and g are constants, the solution is (5.5)

t = exp(x(UI2k - [(U/2kI2 + g/kl”2))

The mean sand concentration and the sedimentation rate are measurable; the rate of sand transport over the Sill is also known, but the sand transport velocity and the sand diffusion coefficient need to be evaluated. To apply the mass balance to the Sound, the details of the sand distribution were studied in a small area along the north shore. A series of short cores were taken along the tracks shown in Fig. 3. The top 20 cm of these samples was wet sieved and the shell removed with dilute HCI. The material that passes a I-mm screen but is retained on a 0.064-mm screen is defined as sand. The volume of material retained on the fine screen was measured after a standard settling time (1 hr) and was compared to the initial volume of sediment to define the sand-mud ratio. The weight-percent sand was found to be about 1.2 times the sand-mud ratio. Harbors along the north shore in the study area have mud floors with a lower sand content than the offshore sediments. South of the harbors, however, the curves of constant sand fraction run nearly north and south

SAND TRANSPORT AT THE FLOOR OF LONG ISLAND SOUND

119

73.00'

S

E

3

0

'

I I I I I I I I 1 I1 I I I l I I I 1 1 I I I 1 1 1 I l l FIG.3. Tracks along which samples were taken in order to determine the sandhud ratio in the northern Sound. New Haven and Sachem Head are indicated by NH and SH, respectively.

so that the variation of sand content may be represented by a single east-west section. The mean water depth decreases to the west so that the muddy floor is at a higher elevation than the sand to the east. This is the reverse of the usual estuarine structure where a shallow sandy margin surrounds a deeper mud bottom. The sedimentation rate is taken to be 1 mm/yr in the study area. This corresponds to a value of g equal sec - I . to 3 x An eddy-diffusion coefficient k can be determined independently of the advection velocity U from measured sand-concentrationgradients where the advective flux of sand is parallel to the contours of constant sand fraction. This would occur over the onshore-offshore sand-mud transitions. The onshore-offshore transitions were examined along tracks 1-4 (Fig. 3). Transitions along these tracks are abrupt; the composition of the Sound floor changes from nearly silt-free sand to silt with as little as 10% sand over a distance of a few kilometers (Fig. 4). If along these transects U = 0, fitting the steady-state solution of the mass-balance equation to these data requires k to be 70 cm2/sec(Fig. 4). If there is some advective component of the sand flux onshore, this value of k is a maximum estimate. Because onshore gradients in the sand content along transects 1-4 are nearly identical, there is no measurable variation in the value of k in this area. For the present illustration, the diffusion will be assumed to be isotropic. The fluctuating water velocities contributing to the diffusion of sand are isotropic, but in regions where d(%%)ldx is not negligible, this simplifying assumption is only a crude approximation. If % and % are

120

HENRY J . BOKUNIEWICZ

Track 60

-

I. 2 0 4 0

0

,0403 20

0

I

I

2

I

3

km

FIG.4. The north-south (onshore-offshore) variation in sand content in northern Long Island Sound. Track locations are shown in Fig. 3; north is to the right of this graph. The solid line is the predicted steady-state sand content with U = 0, k = 70 cm2/sec, g = 3.0 x 10-9/sec.

symmetric functions with equal periods but d 2 out of phase, then (%%) = 0. Even if (%%) is not zero, however, its lateral rate of change, d(%%)/dx, may still be negligible. This is likely to be the case when the

sand concentration gradient is approximately linear. In the Sound this gradient is nearly linear over the distance of the tidal excursion, except at the western limit of the transition zone. With the assumption of a constant and isotropic value of k, therefore, the poorest predictions are expected for the western end of the transition zone. In the Sound, however, the model will prove to be least sensitive to the choice of k, and a more detailed study of this parameter will not be undertaken here. With an estimate of k, the first estimate of U may be made by applying the model to a section through the near-shore, east-west transition zone (Fig. 5). Assuming U is a constant, the value of the effective, sand-grain velocity must be 3.5 x cm/sec in order to approximate the measured sand-concentrationgradient. This solution is shown as the thin line in Fig. 5 and it requires an advective sand flux that is about six times the diffusive sand flux. The total predicted sand flux is 420 mg/m-sec into the transition zone. This value is comparable to the rate of sand transport measured over the sand-wave field. The agreement between these two values is significant since the fluxes compared are independently determined by very different methods. Because the rate of sand transport measured over a period of months is similar to the longer term average, the sand distribution is apparently the result of a nearly continuous motion of sand grains every tidal cycle. The difference between the measured sand flux and the calculated value is about 0.6 gm/m-sec. If the measured flux is indeed typical of the longer term rate of transport, then this difference

SAND TRANSPORT AT THE FLOOR OF LONG ISLAND SOUND

121

is the rate at which sand is accumulating on Mattituck Sill or transported laterally out of the study area. The Sill is about 104 m wide and, assuming a sediment porosity of 0.6, the rate of sand accumulaton must be less than 2 mm/year. The mathematical representation of the sand content in the transition zone is improved by allowing i/ to change from place to place. If the variation of C with distance is prescribed to fit the data (the heavy line in Fig. 5 ) , then the required variation in U may be calculated by a finite difference scheme. When this is done, U is predicted to decrease almost linearly to the west as shown in the inset of Fig. 5 . The “hump” in the sand concentration near the seventeenth kilometer corresponds to the place where the mean sand-grain advection velocity reverses direction. As shown earlier, the net sand flux was expected to be in the direction of the mean water velocity. Analysis of the published tidal stream data by Riley (1956) does show that the inflow of bottom water on the northern side of the Sound diminishes westward and stops somewhere between Sachem Head and New Haven Harbor (Fig. 3). Current meter observations also suggest that the westward flow of bottom water diminishes and disappears in this region as bottom water is diverted into a shoreside mixing zone (Gordon anJ Pilbeam, 1975). Further to the south, where the inflow of bottom water continues westwardly , the transition zone widens as might be expected.

80

-

-

9 60-

1

I

20

10 DISTANCE

0

( km 1

FIG.5. The east-west variation in sand content in northern Long Island Sound. The thin line is the predicted sand content with U = 3.5 x c d s e c , k = 70 cm2/sec, and g = 3.0 x lO-’/sec.

122

HENRY J . BOKUNIEWICZ

The decrease in the mean advection velocity is due, at least in part, to a decrease in the mean water velocity. The change could also be a manifestation of a decrease in the maximum tidal velocity andlor an increase in the critical erosion velocity due to decreasing mean grain size. [The maximum tidal velocity does decrease by about 12% (4 cdsec) along the section.] Both of these phenomena would help to reduce the time that sand grains actually spend in motion and to lower the mean velocity of transport with a consequent decrease in the width of the transition zone. The width of the zone will also be decreased by an increase in the sedimentation rate or a decrease in the diffusion coefficient. 6 . DISCUSSION

This simple mass balance is useful because it provides a systematic structure for discussing the dispersion of sand in Long Island Sound. Using this formulation, the evolution of the sand distribution in the Sound may be considered under both constant and variable conditions. To examine the results of changing conditions, however, the time-dependent equivalent of 5 must be solved. If U , g , and k are constant, analytical solutions may be found for a variety of useful initial conditions (Bokuniewicz, 1976, Appendices 2 and 3). One of the simplest starting situations would be that the central Sound floor was sand free and bounded by the sandy bottom on the east. This condition might approximate the early sedimentary history of the region as described by Bokuniewicz et al., (1976). Before the Sound became an arm of the sea, a large glacial lake probably occupied the central Sound. When contact between the sea and the central Sound was established, the sand ridge which is now Mattituck Sill began to supply sand to presumably fine-grained lacustrine sediment to the west. Because sea level had to rise above the lowest point on the ridge in order to flood the interior of the Sound, when this did happen, the geometry of the ancient Sound must have been similar to its present shape. The strength of the tides, the circulation, and, consequently, the sand flux would also be similar to their present-day values. If this was the case than a total of 8 x lo9 m3 of sand has been swept into the central Sound over the last 8000 yr. In addition, it is likely that the sedimentation rate of marine mud has also been constant (Gordon, 1979). As a result, once the transition zone was established, it most likely has persisted in much the same form as it is found today. For this case, the concentration of sand at any time and location is

SAND TRANSPORT AT THE FLOOR OF LONG ISLAND SOUND

123

given by:

C = 0.5 exp(gx/Zk){exp[-x(B/k)”*] erfc[(x/2(kt)”*- (Bt)”q

+ exp[~(B/k)”~I erfc[x/Z(kt)’” + ( ~ r ) ~ ’ ~ ] ) where B = g + U2/4kand t is the time. With these values for all parameters previously obtained, the predicted sand distribution approaches the steady-state distribution in about 30 years. A 30-year period is a negligible compared to the lifetime of the Sound. As a result, the sand concentration of the sediments in the transition zone should not be expected to vary with depth, but there are no reliable data to show any systematic change in the sand concentration with depth in the transition zone since the time when marine conditions were established in the Sound. If the distribution of sand in the transition zone has indeed persisted unchanged, then the total amount of sand contained there may be calculated from the surficial distribution of sand and the thickness of marine sediments. This value comes out to be about 4 x lo9 m3 (J. Gebert, personal communication), which is only about half of the estimated volume that has been supplied by the floor of the eastern Sound. The deficit may represent that volume of sand that has been deposited on Mattituck Sill. It seems likely, therefore, that the elevation of the Sill has been raised several meters. Although the gross features of the transition zone are expected to have remained constant, slight variations in the sand supply or the sedimentation rate may be represented as irregularities in the sand concentration with depth in the sediment. In cores 40-100 cm in length there are changes in the sand content of up to 15% from centimeter to centimeter. These variations may represent fluctuations in the sediment fluxes over periods of decades. In some cores there is also found a trend toward higher sand contents nearer the sediment-water interface. The difference in sand concentration between the top and bottom of the core may be 10-20%. If such a trend is verified in longer cores throughout the transition zone it would indicate slowly increasing sand fluxes and/or decreasing sedimentation rates. Such changes could be due to an increase in the frequency or severity of storms. Strong winds tend to increase the intensity of random watervelocity fluctuations near the Sound floor but produce only a small change in the mean flow (Bokuniewicz and Gordon, this volume, p. 84). In terms of the mass-balance equation, the diffusion coefficient for mobile sand might be expected to increase during periods of high winds. The influence of such events on the sand distribution was examined with a tifne-de-

124

HENRY J . BOKUNIEWICZ

pendent mass-balance equation as follows. Beginning with the normal steady-state sand distribution, the diffusion coefficient k was increased to lo6cm’hec, which is approximately the horizontal eddy-diffusion coefficient in the water. This is, therefore, an extreme upper limit to the turbulent diffusion coefficient for mobile sand. The “storm” was allowed to persist for 2 days and the change in the surficial sand content at any location was found to be less than 6%. After 30 days, the change was as much as 25% in some places. The sand distribution relaxes back to the steady-state distribution within a few years after the disturbance stops. The accreting deposit of marine mud would preserve some of the storminduced sand concentrations in the immobile sediment. Subsequent bioturbation would reduce the measured effect of the storm in increasing or decreasing the sand concentration preserved in a core (Berger and Heath, 1968), nevertheless, the changes in sand content induced by occasional periods of high sand diffusion may result in the fine-scale variations in sand content that are observed in short cores. A single storm probably would not measurably effect the large-scale sand distribution because it is capable of only a 6% change in the sand content. A series of 30 or 40 events, however, that might occur over a few winter seasons would be enough to significantly alter the distribution. 7. SUMMARY A N D CONCLUSIONS In this article a simplified mass balance has been used to describe the net transport of sand over an accreting mud bottom. The combination of these two sedimentary processes controls the transition from sand to mud on the floor of the Sound. The distribution of sand may be described with three parameters: an advection velocity of sand grains, an eddy-diffusion coefficient for mobile sand, and a rate of accumulation of marine mud. (Only the ratios of these quantities are needed if the distribution is in a steady state.) The motion of sand is thereby represented with both a deterministic part and a statistical part. The net, one-way advection of sand is the result of the superposition of an estuarine circulation on the tidal stream, and unpredictable variations in the rate of sand transport are represented as an eddy-diffusion process. Sand is immobilized when it is incorporated into the permanent deposit of marine mud. In applying this model to the Sound, a number of simplifications were warranted. The tides in the central Sound set approximately east and west, the net inflow of bottom water is to the west, and the sand content of the bottom sediment is approximately constant laterally across the

SAND TRANSPORT AT THE FLOOR OF LONG ISLAND SOUND

125

Sound. For these reasons the transition zone was described using a onedimensional form of the mass-balance equation. The muddy floor of the Sound is blanketed with a layer of readily resuspendable agglomerates. This layer is about 1 cm thick. Any sand in this layer was assumed to be regularly set in motion by the currents, while sediment grains below this layer remained immobile. The formation of a permanent mud deposit occurs at the base of this layer at the rate of about 1 mm/year. This value was taken as the speed with which sand grains are removed from the zone of mobile sediment. It was found that the sand-mud transition zone could be adequately represented by allowing the sedimentary processes to proceed at constant rates. When this was done, the calculated sand flux agrees well with the sand flux that has been measured over the sand-wave field in the eastern Sound. Since the tides and the estuarine circulation control the sand fluxes, the resulting distribution of sand approaches a steady state very quickly. It is likely that the sand-mud transition was established soon after the Sound became an arm of the sea and has persisted unchanged to the present day. As a corollary to this hypothesis, it would seem that Mattituck Sill has been accreting over the lifetime of the Sound. Small variations in the sand content preserved in short cores are probably due to the perturbation of the sand distribution by a series of winter storms rather than individual storm events. A better description of the sand distribution can be obtained by allowing the net speed of sand transport to decrease into the Sound. This is consistent with what is known about the current velocities in the region. More detailed models could also be devised to account for variations in the sedimentation rate, the thickness of the layer of mobile sediment, and the effective diffusion coefficient for mobile sand. Another refinement would include the lateral transport of sand that undoubtedly does occur. The further development of mathematical representations of estuarine processes should proceed simultaneously with investigations of both specific sedimentary processes and regional sedimentary systems. For the model proposed here some of the specific processes that deserve attention in the future include the processes that control the rate of formation of marine mud at the base of the surficial layer of agglomerates and the relationship between the eddy-diffusion coefficient for sand transport and fluctuations in the water velocity. The study of specific processes tell us little about the long-term manifestations of these processes. For this there is the need to develop comprehensive descriptions of estuarine sedimentary systems and to begin to contrast and compare sediment budgets in different coastal areas.

126

HENRY J. BOKUNIEWICZ

ACKNOWLEDGMENTS 1 wish to thank Robert Gordon for suggesting this study. I have had the benefit of many helpful discussions with him and also C. C. Pilbeam and B. Saltzman of Yale University. M. Reed, P. Kaminsky, and J. Gebert assisted with the field work. This research was financed by a grant from the Geological Society of America. I would also like to thank the National Marine Fisheries Service and J. R. Schubel and W. Wise for making sediment data available for this study.

REFERENCES Ali, S. A,, and Feldhausen, P. H. (1975). Sedimentary environmental analysis of Long Island Sound, USA with multivariate statistics. I.M.G. Sess., 9th Congr. In?. Sedimentat., 1975. Aller, R. C., and Cochran, J. K. (1976). 234Thf38U disequilibrium in near-shore sediment: Particle reworking and diagenetic time scales, Earth and Planet. Sci. Lett. 29, 37-50. Bagnold, R. A. (1966). An approach to the sediment transport problem from general physics. U.S., Geol. Surv., Prof. Pap. 422-1, 11-137. Belderson, R. H., and Kenyon, N. H. (1969). Direct illustration of one-way sand transport by tidal currents, J. Sediment. Petrol. 39, 1249-1250. Berger, W. H., and Heath, G. R. (1968). Vertical mixing in pelagic sediments, J. Mar. Res. 29,134-143. Bokuniewicz, H. J. (1976). Estuarine sediment flux evaluated in Long Island Sound. Ph.D. Dissertation, Yale University, New Haven, Connecticut. Bokuniewicz, H. J . , Gordon, R. B., and Pilbeam, C. C. (1975). Stress on the bottom of an estuary. Nature (London) 257, 575-577. Bokuniewicz, H. J., Gebert, J., and Gordon, R. B. (1976). Sediment mass balance in a large estuary. Estuarine Coast. Mar. Sci. 4, 523-536. Bokuniewicz, H. J., Gordon, R. B., and Kastens, K. (1977). Form and migration of sand waves in a large estuary. Mar. Geol. 24, 185-199. Bowden, K. F. (1963). The mixing process in a tidal estuary. In?. J. Air Water Pollut. 7, 343-356. Buzas, M. A. (1965). The distribution and abundance of foraminifera in Long Island Sound. Smithson. Misc. Collect. 149, No. 1, 1-88. Caston. V. N. D. and Stride, A. H.(1970). Tidal sand movement between some linear sand banks in the North Sea off north-east Norfolk. Mar. Geol. 9, M38-M47. Colby, B. R. (1964). Discharge of sands and mean velocity relationships in sand-bed streams. U S . , Geol. Surv., Prof. Pap. 462-A, 1-47. Davis, J. C. (1970). Information contained in sediment-size analyses. Int. J . Math. Geol. 2, 105-112. Donohue, J. J., and Tucker, F. B. (1970). “Marine Mineral Identification Survey of Coastal Connecticut.” United Aircraft Research Laboratories Rep. No. 5-970660-1. Dyer, K. R. (1974). The salt balance in stratified estuarines. Estuarine Coastal Mar. Sci. 2, 273-282. Gadd, P. E., Lavelle, J. W., and Swift, D. J. P. (1978). Estimates of sand transport on the New York shelf using near-bottom current meter observations. J. Sediment. Petrol. 48, 239-252. Gordon, C. M. (1975). Sediment entrainment and suspension in a turbulent tidal flow. Mar. Geol. 18, M57-M64.

SAND TRANSPORT AT THE FLOOR OF LONG ISLAND SOUND

127

Gordon R. B. (1979). Denudation rate of central New England determined from estuarine sedimentation. Am. J. Sci. 279, 632-642. Gordon, R. B., and Pilbeam, C. C. (1975). Circulation in central Long Island Sound. J . Geophys. Res. 80,414-422. Guilcher, A. (1967). Origin of sediments in estuaries. I n “Estuaries” (G. H. Lauff, ed.), Publ. No. 83, pp. 149-157. Am. Assoc. Adv. Sci., Washington, D.C. Hathaway, J. C. (1972). Regional clay mineral facies in estuaries and the continental margin of the United States east coast. Mem.. Geol. SOC.Am. 133, 293-315. Inman, D. L., Weing, G. C., and Corliss, J. B. (1966). Coastal dunes of Geurrero Negro, Baja California, Mexico. Geol. SOC.Am. Bull. 77, 787-802. Jackson, R . G. (1976). Sedimentologicai and fluid-dynamic implications of the turbulent bursting phenomenon in geophysical flows. J. Fluid Mech. 77, 531-560. Jones, N. S., Kainjand, J. M., and Stride, A. H. (1965). The movement of sand waves on Warts Bank, Isle of Man. Mar. Geol. 3, 329-336. Kachel, N. B., and Sternberg, R. N. (1971). Transport of bedload as ripples during ebb current. Mar. Geol. 19, 229-244. Klein, G. (1970). Depositional and dispersal dynamics of intertidal sand bars. J. Sediment. Petrol. 40, 1095-1 127. Krank, K. (1976). Tidal current control of sediment distribution in Northumberland Strait, Maritime Provinces. J. Sediment. Petrol. 42, 596-601. Krumbein, W. C., and Aberdeen, E. (1937). The sediments of Barataria Bay. J. Sediment. Petrol. I, 3- 17. Ludwick, J. C. (1972). Migration of tidal sand waves in the Chesapeake Bay entrance. In “Shelf Sediment Transport” (D. J. P. Swift, D. B. Duane, and 0. H. Pilkey, eds.), pp. 377-410. Dowden, Hutchinson & Ross, Inc., Stroudsburg, Pennsylvania. Ludwick, J. C. (1975a). Variations in the boundary drag coefficient in the tidal entrance to Chesapeake Bay, Virginia. Mar. Geol. 19, 19-28. Ludwick, J. C. (1975b). Tidal currents, sediment transport and sandbanks in Chesapeake Bay entrance, Virginia. I n “Estuarine Research” (L. E. Cronin, ed.), Vol. 2, pp. 253-265. Academic Press, New York McCall, P. L. (1977). Community patterns and adaptive strategies of the infaunal benthos of Long Island Sound. J. Mar. Res. 35, 221-266. McCave, 1. N. (1970). Deposition of fine-grained suspended sediment from tidal currents. J. Geophys. Res.75. 4151-4159. McCave, I. N. (1971). Sand waves in the North Sea off the coast of Holland. Mar. Geol. 10, 199-225. McCrone, A. W., Ellis, B. F., and Charmatz, R. (1961). Preliminary observation on Long Island Sound sediments. Trans. N. Y. Acad. Sci. [2] 24, 119-129. Meade, R. H. (1969). Landward transport of bottom sediments in estuaries of the Atlantic Coastal Plain. J . Sediment. Petrol. 39, 222-234. Miller, M. C., McCave, I. N., and Komar, P. D. (1977). Threshold of sediment motion under unidirectional currents. Sedimentology 24, 507-527. Moore, J. R. (1963). Bottom sediment studies, Buzzards Bay, Massachusetts. J . Sedimenr. Petrol. 33, 511-558. National Marine Fisheries Service (1974). “Environmental Baselines in Long Island Sound 1972-1973,” Informal Rep. No, 42, 2 vols. Middle Atlantic Coastal Fisheries Center (unpublished). Nichols, M. M. (1977). Response and recovery of an estuary following a river flood. J. Sediment. Petrol. 47, 1176-1 186.

128

HENRY J. BOKUNIEWICZ

Nordin, C. F. (1975).Tidal data used to calibrate mathematical models for sediment transport in unsteady flows. EOS, Trans. Am. Geophys. Union 46, 981 (abstr.). Pingree, R. D., and Maddock, L. (1977). Tidal residuals in the English Channel. J. Mar. Biol. Assoc. U.K . 57, 339-354. Postma, H. (1961). Transport and accumulation of suspended matter in the Dutch Wadden Sea. Neth. J. Sea Res. 1, 148-190. Rhoads, D., and Cande, S . (1971). Sediment profile camera for in situ study of organismsediment relations. Limnol. Oceanogr. 16, 110-1 14. Riley, G. A. (1956). Oceanography of Long Island Sound, 1952-1954. 2. Physical oceanography. Bull. Bingham Oceangr. Collect. 15, 15-46. Salsman, G. G., Tolbert, W. H., and Villars, R. G. (1966). Sand ridge migration in St. Andrews Bay, Florida. Mar. Geol. 4, 11-19. Schubel, J. R. (1974). Effects of tropical storm Agnes on the suspended solids of the Northern Chesapeake Bay. In “Suspended Solids in Water” (R. J. Gibbs, ed.), pp. 113-132. Plenum, New York. Schubel, J. R., and Wise, W. M. (1979). “A Dredged Material Management Plan for Long Island Sound,” report to New York Sea Grant Institute, Albany, New York. Marine Sciences Research Center, State University of New York (in preparation). Sternberg, R. W. (1968). Friction factors in tidal channels with differing bed roughness. Mar. Geol. 6, 243-260. Sternberg, R. W. (1972). Predicting initial motion and bedload transport. In “Shelf Sediment Transport” (D. J. P. Swift, D. B. Duane, and 0. H. Pilkey, eds.), pp. 61-82. Dowden, Hutchinson & Ross, Inc., Stroudsburg, Pennsylvania. Stride, A. H. (1963). Current-swept sea floors near the southern half of Great Britain. Q. J. Geol. SOC.London 119, 175-197. Stride, A. H. (1974). Indications of long term, tidal control of net sand loss or gain by European coasts. Estuarine Coastal Mar. Sci. 2, 27-36. Sutherland, A. J. (1967). Proposed mechanism for sediment entrainment by turbulent flows. J. Geophys. Res. 72, 6183-6194. Swift, D. J. P., and McMullen, R. M. (1968). Preliminary studies of intertidal sand bodies in the Minas Basin, Bay of Fundy, Nova Scotia. Can. J. Earth Sci. 5 , 175-183. Swift, D. J. P., Ludwick, J. C., and Bochmer, W. R. (1972). Shelf sediment transport: A probability model. I n “Shelf Sediment Transport” (D. J. P. Swift, D. B. Duane, and 0. H. Pilkey, eds.), pp. 195-223. Dowden, Hutchinson & Ross, Inc., Stroudsburg, Pennsylvania. Tennekes, H., and Lumley, J. L. (1972). “A First Course in Turbulence.” MIT Press, Cambridge, Massachusetts. Van Straaten, L. M. J. U., and Kuenen, P. H. (1958). Tidal action as a cause of clay accumulation. J. Sediment. Petrol. 28, 406-413. Wilson, R. (1976). Gravitational circulation in Long Island Sound. Estuarine Coastal Mar. Sci. 4, 443-453. Young, R. A., and Southard, J. B. (1978). Erosion of fine-grained sediments; sea-floor and laboratory experiments. Geol. SOC. Am. Bull. 89, 663-672.

THE SOURCES AND SINKS OF NUCLIDES IN LONG ISLAND SOUND K. K. TUREKIAN, J. K. COCHRAN,L. K. BENNINGER,* AND ROBERT.c. ALLERt Depurtment of Geology and Geophysics Yule University New Haven, Connecricirt

1. 2.

3. 4.

5.

Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . Sources of Trace Metals Delivered to Long Island Sound . . . . . . . . . 2.1. Stream Supply . . . . . . . . . . . . . . . . . . . . . . . . . 2.2. Sewer Outfalls: New Haven Harbor . . . . . . . . . . . . . . . . 2.3. Atmospheric Supply: The Record in a Salt Marsh . . . . . . . . . . . The Distribution of Trace Metals in Long Island Sound Sediments . . . . . . Trace-Metal Distributions in Mussels and Oysters: An Index of the Composition of Suspended Particles . . . . . . . . . . . . . . . . . . . . . . . . 4.1. Oysters . . . . . . . . . . . . . . . . , . . . . . . . . . . . 4.2. Mussels . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.3. The Cause of the Observed Trace-Metal Distributions . . . . . . . . . Processes Affecting the Deposition and Accumulation of Trace Metals in Long Island Sound Sediments . . . . . . . . . . . . . . . . . . . . . 5.1. Water Column Scavenging . . . . . . . . . . . . . . . . . . . 5.2. Horizontal Distribution . . . . . . . . . . , . . . . . . . . . . . Processes Affecting the Vertical Distribution of Nuclides in the Sediment Pile . 6.1. Establishing Chronologies: Time Scales of Accumulation and Bioturbation 6.2. Determination the Final Repository of Metals Introduced into Long Island 6.3. Major Sources of Metals Delivered to Long Island Sound . . . . . . . Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . I

6.

7.

. . . . . .

.

.

,

,

.

. . .

129 131 131 133 136 137

.. .

. . . . . . , . . . . . ., . . ,

142 142 144 144

. . , ..... ... . . . . . . . . . . . . Sound . . . . . . . . . ... . ... .

147 149 150 153 153 157 159 161 163

.

. . , . . . , . . . . . .. .

. .

I. INTRODUCTION There is little doubt that the sedimentary record in most, if not all, estuaries is now indelibly marked by the products of human activity. Indeed, even deep-sea deposits have begun to respond to this influence. The concerns about possibly deleterious effects of this impingement of man on his oceanic environment have brought renewed interest to the problem of the fate of chemical species in the estuarine environment, * Present address: Department of Geology, University of North Carolina, Chapel Hill, North Carolina 275 14. t Present address: Department of Geophysical Sciences, The University of Chicago, Chicago, Illinois 60637. I29 ADVANCES IN GEOPHYSICS, VOLUME

22

Copyright 0 1980 by Academic Press. Inc. All rights of reproduction in any form reserved. ISBN 0-12-018822-8

130

K. K.TUREKIAN et al.

The estuarine regions of the world are the sites of encounter of riverborne materials from the continents with the ocean. The estuary, however, is not a passive way station, but a combination tollgate and indoctrination center that extracts its fees from the transients and imposes a new way of life on the survivors of the passage. Long Island Sound is an estuary that also has the properties of a large protected settling basin. In such a system many of the processes affecting the behavior of material injected into the coastal zone can be followed more directly than in some other systems more responsive to large-scale and seasonal variables. Smaller systems such as “simple” river estuaries can typically accommodate very limited sediment accumulation and the meager record is highly susceptible to loss during floods or coastal storms; highly exposed coastal regions like the New York Bight do not confine inputs in a predictable way. In this article we summarize what we have learned about the behavior of metals in the Long Island Sound. Many of the insights can be transferred to other estuaries and should be useful in understanding the effects of human activities impinging on such systems. As reviewed in the first article in this volume, by Gordon, Long Island Sound came into existence as sea level rose after the end of the Wisconsin glacial age. The initially fresh-water reservoir was transformed by the rising sea into the present estuarine Long Island Sound about 8000 years ago. At that time a change in sediment properties is inferred to have resulted and this is now seen by sonic profiling as a strong reflector. Using the depth to this reflector (dated at 8000 yr BP by the sea level curve as discussed earlier), Bokuniewicz et al. (1976) calculated an average sediment accumulation rate in the mud and silt area covering 64% of the present Sound of 0.33 mm yr-’. Benninger (1976) showed that this could all be the result of mud and silt supply by the Connecticut River if the present-day sediment flux existed throughout the history of the Connecticut River and the Sound. Imprinted on this naturally accumulating sediment are the additions of materials resulting from human activity. These include dredge spoils, which are of local importance, and metals, organic matter, and other substances introduced from a number of sources including the atmosphere, industrially polluted rivers, and sewage outfalls. The subsequent behavior of metals transported to the coastal zone is determined by several factors of which the efficiency of scavenging from the water column, horizontal redistribution during deposition, and vertical redistribution after deposition are perhaps the most important. The first of these depends on the concentration and reactivity of the suspended particles in the water column. The last two result from tidal action, storms,

NUCLIDES IN LONG ISLAND SOUND

131

and estuarine circulation all affecting particles suspended in the water column, on the one hand, and physical and biological mixing of particles in the sediment column on the other. Using diagnostic natural radionuclides it is possible to identify the mechanisms and time scales of the processes. Members of the uranium decay series, particularly 234Th(halflife, 24 days) and 210Pb(half-life, 22 years), have proven useful in this pursuit. In addition, man-made radionuclides such as the plutonium isotopes and I3'Cs also have proven valuable tracers because of their sharply defined dates of introduction into the environment. Carbon-14 occupies a unique position in this constellation of useful radionuclides since its origins are both natural (cosmogenic) and man-made. The ''C/'2C ratio can be used not only as a time indicator, but as a tracer of the sources and fates of organic carbon as well. 2. SOURCES OF TRACEMETALS DELIVERED TO LONGISLAND SOUND The silt and clay composing the sediments of the central basin of Long Island Sound are derived ultimately from the weathering and erosion of the rocks of New England. As solid phases transported by streams they bear a burden of trace elements characteristic of their mineralogy and weathering history. This process has been going on since the Sound was formed about 8000 years ago and provides the background material on which is imposed the recent increase in trace-metal supply due to human activity. The primary methods of supply are by streams, sewer outfalls, and atmospheric transport. Maintenance or construction-related dredging of metal-contaminated sediments results in the transport of the material to other locations in the Sound where the dredge spoil is dumped. The identification of such dumping has been made in the New York Bight using trace metals and organic content as well as other indicators (Gross, 1976; Carmody et al., 1973), but as we shall see, other factors operate in Long Island Sound to compromise preservation of a local identity. 2.1. Stream Supply The trace-metal concentrations in streams are controlled not only by input from the weathering of rocks and from aerosols, but also by the chemical reactions occurring within the streams. The evidence from studies utilizing "OPb as a tracer for heavy-metal behavior in streams indicates rapid scavenging by particles associated with the flowing water (Bennin-

132

K. K.TUREKIAN et a / .

ger et al., 1975; Lewis, 1977). Surfaces of organic debris and manganese and iron-oxide coatings appear to be the primary agents (Gibbs, 1967). Competing with this process is the formation of dissolved organic-chelated compounds. Much of the dissolved iron found in streams, for instance, is in this form (Sholkovitz, 1976). A study of Connecticut streams (Turekian, 1971) indicated that the trace metals, cobalt and silver, are maintained at low concentrations in solution as the result of the scavenging action of suspended particles. Even where acid industrial wastes are dumped into the stream, as in the Naugatuck River, which joins the Housatonic River, suspended particles act to lower the dissolved concentrations. We infer from studies involving the behavior of ‘“Pb in the Susquehanna River and of Co and Ag in the major Connecticut rivers that the dissolved trace-metal concentration is maintained at low levels in stream water and thus the primary mode of transportation to the estuarine zone is via particles. The work at Yale University on the Quinnipiac River, a river carrying effluents from the major metal industries of Meriden and Wallingford and entering into New Haven Harbor, supports this inference. Figure 1 shows the distribution of total silver in Quinnipiac River waters in 1965 and demonstrates an increase in concentration through Wallingford. In Wallingford, the >0.45-pm fraction (associated with particles) adds to the 3 pg/l delivered from the “uncontaminated” reservoirs. (Because of the possibility of atmospheric transport of metal contaminants, it is not likely that any nearby reservoir is truly uncontaminated.) Figure 2 shows that the bottom sediments of the system, sampled in 1973, are strongly impacted by the trace-metal injections from industry. Although the silver concentration is very high in the sediments close to the source of industrial contamination, this effect is strongly attenuated downstream. Indeed, as we shall see, the effect is not discernible in New Haven Harbor where other sources predominate. This implies that most of the metals are retained behind the dams and a relatively small fraction escapes to impact the estuary. About 15 km to the west of New Haven Harbor the Housatonic River with its heavily polluted tributary, the Naugatuck River, empties into the Sound. (The confluence occurs below the last dam on the Housatonic.) This river supplies a significant amount of trace metals to the adjacent part of Long Island Sound, mainly in particulate form (Turekian, 1971). This contrasts sharply with the Quinniapiac River and demonstrates that the construction of dams is certainly one important factor in inhibiting transfer of metal-polluted sediments to the estuarine zone. The Connecticut River, although the most important river draining into Long Island Sound, seems to be least important in the transport of metals

133

FIG.1. Silver in the Quinnipiac River (Connecticut) system in 1965 (previously unpublished Yale University data). All concentrations determined on unfiltered samples except where indicated. The measurements were made by emission spectrography after silver-free sodium chloride was added and the solution freeze-dried.

from human activities to the Sound. This is no doubt due in part to the extensive damming along the course of the river and in part to the minimal amount of metal fabrication along its length.

2.2. Sewer Outfalls: New Haven Harbor Applequist et al. (1972) showed that the mercury concentration in the sediments of New Haven Harbor varied in relation to distance from the several sewage-treatment plants discharging into the harbor (Fig. 3). As is the case with most of the older New England cities, storm sewers are combined with sanitary sewers and the effluent is processed through the sewage-treatment plants. During periods of high discharge associated with large storms, the treatment plant is bypassed and the unprocessed effluent

134

K. K. TUREKIAN et a / .

d BROAD BROOK RES.

FIG.2. The concentrations of silver, lead, and copper in sediments of the Quinnipiac River (Connecticut) system (previously unpublished Yale University data). Analyses made by emission spectrography. Concentrations in parts per million.

is debouched directly into the harbor, resulting in organic carbon and metal enrichment of the sediment around the outfalls. In an unpublished report from Yale (Turekian et al., 1972), it is shown that a relation exists between high concentrations of Pb, Zn, Cu, and organic matter in sediments and proximity to a sewer outfall on the eastern shore of New Haven Harbor. A more cursory survey of the sediments on the west side of the harbor showed the same thing (unpublished results). Thus the pattern established by our mercury study, we believe, can be extended to the other elements associated with sewage sludge. Figure 4 is the representation of the relationship between Zn concentration and the weight lost on ignition (a rough expression of the combination of water loss from clay minerals and the degradation of organic matter) for a traverse up the New Haven Harbor shipping channel. This cuts across the two sewer outfall regions discussed earlier (see Fig. 3), and the strong correlation of Zn and the volatile solids concentration reflects their influence.

NUCLIDES IN LONG ISLAND SOUND

135

FIG.3. The distribution of mercury (in ppm) in the tops of sediment cores raised from New Haven harbor. The dominant control on mercury concentration is proximity to sewagetreatment plant outfalls, marked by arrows. (After Applequist el a/., 1972.)

"

0

1

2

3

4

5

6

7

8

9

0

% VOLATLE SOLIDS FIG.4. The relation of zinc concentration to volatile solids (mainly organic matter with some adsorbed water in clays) in sediments from the New Haven harbor channel. Data from the U.S. Corps of Engineers files (New Haven Harbor Project: Report on Environmental Sampling and Testing, 1972).

136

K. K. TUREKIAN et a/.

2.3. Atmospheric Supply: The Record in a Salt Marsh

Salt marshes are a common feature of the Long Island Sound coast. Where they remain protected from wave erosion their surface is an index of high tide. If coastal submergence occurs over time, as has been the case in New England for at least the last 100 years, the protected marsh grows upward to maintain its surface at high tide and provides a record of previous environmental conditions. The surface of a marsh is exposed to the atmosphere most of the time. The highest point of the tidal cycle immerses the surface only about 5% of the time. Unlike marshes in other parts of the east coast, Long Island Sound marshes are not dominated by detrital sediment, but are constructed of the fibrous framework of the marsh vegetation. Because of this, burrowing by organisms does not appear to perturb the sedimentary record as occurs in the muddy sediments at the bottom of Long Island Sound (see later). As the marsh grows upward in response to the rising sea level, each layer should preserve a record reflecting the depositional environment of the time. Changes in detritus supply from streams, for example, should be recorded by sediment trapped in the fibrous framework. Similarly, the atmospheric flux records of metals delivered to the marsh surface should be maintained in the layers. McCaffrey (1977) and McCaffrey and Thomson (this volume) have shown that the 210Pbchronology from a vertical 'IoPb profile in a Connecticut salt marsh agrees with tide-gauge data (Fig. 5). This shows that during the past 100 years the sea level has been rising relative to the Connecticut coast. The agreement between the '"Pb and tide-gauge data is especially striking because the rate of coastal submergence has not been constant over the past 100 years. In addition, these researchers showed that the calculated Z'oPbflux, as determined by the standing crop of unsupported "OPb in the salt marsh, equaled the atmospheric flux as determined for New Haven by Benninger (1978). This implies that: (1) the trace-metal distribution vertically in the salt marsh reflects the changing flux over time and that no vertical migration of the trace metals is expected by diffusion or biological activity; and (2) that there should be an atmospheric flux of trace metals recorded in the growing salt marsh. Indeed, the calculated fluxes of Cu, Pb, and Zn (Fig. 6) appear to be almost solely atmospheric as the predicted fluxes of these metals are in agreement with the estimated atmospheric fluxes (Table I). The implication from both *IOPband trace-metal data is that the marsh surface, exposed above the sea surface most of the time, behaves like an atmospheric collector and can be used to monitor the changing atmospheric flux of trace metals over time.

137

NUCLIDES IN LONG ISLAND SOUND

A G E (yr) 40

0 0 -

+.

I20

80

160

200 '1

c

**

-? c E

0

10

-

20

-

-.

Smoothed Tide-Gouge Record Salt-Marsh Pb2"

Record

-

....-.

-

1

W

> 0)

d (u

b

z

0 W

m

-

r

c

4

W

0

ji/_

, ;

e4% e49

*

,

, , ; ,

,

e8

,

, ; ,

\B8

,

, , ; ,

\B8

,

I

,

,

4b0

\

i

YEAR

FIG. 5. Depth versus time curves for the Farm River (Connecticut) salt marsh based on "'Pb dating with a comparison with the tide-gauge record for New York City. (From McCaffrey and Thomson, this volume.)

3. THEDISTRIBUTION O F TRACEMETALS IN LONGISLAND S O U N D SEDIMENTS

Greig et af. (1977)have recently made a detailed study of the distribution of a number of trace metals (Sb, Cd, Co, Cr, Cu, Pb, Mn, Ni, Ag, Sc, Zn) in the top 4 cm of Long Island Sound sediments collected using a Smith-McIntyre grab sampler. The 4-cm sampling fortuitously represents, to within a centimeter, the rapidly reworked portion of the sediments as determined using 234Th(Aller and Cochran, 1976). Figures 7-9 show concentration maps for Cu, Zn, and Pb constructed from the data of Greig et al. (1977). The primary control on the trace-metal concentrations is the grain size of the sediment. This can be seen by comparing the trace-metal maps with a grain-size distribution map (Fig. 10) for the Sound. The sand-rich sediments have the lowest trace-metal content. There is, however, an im-

138

K . K.TUREKIAN et

0

1972

4

8

12

4

al.

12

8

0

4

8

I

1912

a 1892

a

Zn

1852

1792 FIG. 6. The “excess” flux of copper, zinc, and lead as a function of time at the Farm River (Connecticut) salt marsh. (From McCaffrey and Thomson, this volume.)

portant second-order effect related to the coastal sources of trace metals. Sediments adjacent to Throgs Neck, the Housatonic River (to the west), and New Haven Harbor are higher in trace metals than other sediments of the same grain size. These three areas are heavily impacted either by sewer outfalls or direct injection of industrial sewage along a contiguous channel (as in the Naugatuck-Housatonic system). A number of cores collected from central Long Island Sound have been analyzed for trace metals as a function of sediment depth (Thomson et al., 1975; Turekian, 1979; Benninger et al., 1979). They show roughly the same patterns for Cu, Zn, and Pb (Fig. 11): a roughly exponential decrease in concentration with depth. At greater depths there are occasional peaks of high concentrations. TABLEI. CALCULATED EXCESS METALFLUXTO THE SURFACE OF THE FARMRIVERSALT MARSHCOMPARED TO MEASURED ATMOSPHERICDEPOSITION RATESAT SELECTED SITES

Site and date of collection Branford, Connecticut, salt marsh (1972) New York City (1969-1970) Nantucket, Massachusetts (1966- 1967)

Cu 8

f

9.8 5.6

Zn

2

12 f 3 32 7.6

Pb

Source

7 +- 4 McCaffrey (1977); and McCaffrey and Thomson (this volume) 35 Volchok and Bogen (1971) 8.5 Lazrus et a/. (1970)

NUCLIDES IN LONG ISLAND SOUND

139

COPPER pglgrn DRY SEDlMM

FIG. 7. Map of copper concentrations in surface sediments of Long Island Sound. (Constructed from the data of Greig et al., 1977.)

ZINC pglgm DRY SEDIMENT

FIG.8. Map of zinc concentrations in surface sediments of Long Island Sound. (Constructed from the data of Greig et al., 1977.) 1

LEAD pg/gm DRY SEDIMENT

m

'Do

wm

0-m

n


FIG. 9. Map of lead concentrations in surface sediments of Long Island Sound. (Constructed from the data of Greig et al., 1977.)

I40

K. K. TUREKIAN et d.

MUD(%<62pm)

5 30-70 0
t

1148 0

5’ MEP 102375

FIG. ll(a). See p. 141 for legend.

141

NUCLIDES IN LONG ISLAND SOUND

pg/gm ASH(500OC) 20

0

0

-l.E i":wo 6 0 Y

,

4 0 -' O : lE R( '<

80

40

60 T

cu

80

Kx)

Ix,

140

160

'

180

I r'

I

xx,

1

FIG. 1 1 . The distribution of trace metals in cores from Long Island Sound. (a) Location map of.cores described in this article. (b) Core 1148 data from Thomson et al., 1975. (c) Core NWC-102975 data from Benninger et al. (1979). (d) Core 1036 data from Turekian (1979).

142

K. K. TUREKIAN et

d.

As Fig. 11 shows, there is some variability in the integrated trace-element contents of cores from central Long Island Sound. These sediments are not located near intense sources of trace metals nor are there major grain-size differences. As we will show later, the differences in both metal concentrations and inventories can be related primarily to the intensity and depth of biological mixing of the sediment column. 4. TRACE-METAL DISTRIBUTIONS IN MUSSELS AND OYSTERS: AN

INDEX OF THE COMPOSITION OF SUSPENDED PARTICLES Mussels and oysters (epifauna) are filter feeders which attach to hard surfaces. Their isolation from the sediment means that their trace-metal compositions are likely to be reflective, primarily, of the suspended material in the water. In this section, therefore, the data available on tracemetal concentrations in mussels and oysters from the Connecticut shore are reviewed. 4.1. Oysters

Feng and Ruddy (1974) made a detailed study of the composition of the soft tissue of oysters (Crassostrea virginica) harvested along the Con-

FIG.12. Zinc concentrations in the soft tissues of oysters as a function of location, along the Connecticut coast, and time, starting in June 1972 and ending in March 1974. (Plotted from the data of Feng and Ruddy, 1974.)

143

NUCLIDES IN LONG ISLAND SOUND r

I

I

1

I

I

I

-

looBRlDoEpoRT

90-

-

-

20

-

-

NOANK

NORW

J-S

72

0-D

n

J-M

P J

73

73

J-S 73

0-D

73

-

J-M 74

FIG. 13. Cadmium concentrations in the soft tissues of oysters as a function of location, along the Connecticut coast, and time, starting in June 1972 and ending in March 1974. (Plotted from the data of Feng and Ruddy, 1974.)

necticut coast. A single stock of oysters obtained as yearlings were distributed among six stations: (1) Norwalk Harbor at the Northeast Utilities Company pier, (2) Bridgeport at the Pleasure Beach Bridge, (3) the Housatonic River below Devon, (4) New Haven Harbor at the Coast Guard Station finger pier, ( 5 ) New London Harbor at the U.S. Navy Underwater Systems Center pier, and (6) Noank at the University of Connecticut Marine Sciences Institute pier. The stock was then sampled periodically between June 1972 and April 1974 and tissue analyzed for Cd, Cu, Hg, Mn, and Zn. The oyster tissue did not vary significantly in the concentration of these elements from the native oysters also analyzed. The highest values for all of the trace elements except mercury were found at the Bridgeport and Housatonic sites. Figures 12-14 show the changes in composition with time, at each of the six locations, for Zn, Cd, and Cu, respectively. There is clearly a marked increase from the summer of June 1972 to the winter of 1974 for the Housatonic-Bridgeport region for all metals and a marked increase for zinc for all other locations except Norwalk, which seems to have gone through a maximum in the winter or spring of 1973.

144

K. K. TUREKIAN et a!.

4 500 4000 3500

k

3000 2500

2000 I500

1000

J-S

0-D

72

72

J-M 73

A-J 73

J-S 73

0-0 73

J-M

74

FIG. 14. Copper concentrations in the soft tissues of oysters as a function of location, along the Connecticut coast, and time, starting in June 1972 and ending in March 1974.

4.2. Mussels

Trace metals in Long Island Sound mussels collected in 1975 and 1976 were determined by Curran et al. (1980). A map of the sampling locations is shown in Fig. 15. [Additional samples were collected at other sites and analyzed. These are in part reported in Turekian (1979) and are included in this study.] The geographic variations of trace metals in native mussels (Figs. 16-20) show the same patterns as the oysters, although the concentrations are considerably lower in the mussels. The pattern holds for all the trace metals analyzed including Pb and Ni as well as Zn, Cd, and Cu. In addition to the high values associated with the Housatonic-Bridgeport area, a region of high metal concentration in the mussels is found in the area around Throgs Neck (an area not included in the oyster study).

4.3. The Cause of the Observed Trace-Metal Distribution Obviously both oyster and mussel tissue compositions are influenced by the trace-metal content of the particles they ingest. There should then be a relationship between the chemical properties of the particles of the surrounding water and the sediments and the compositions of the tissues.

NUCLIDES IN LONG ISLAND SOUND

NEW YORK

CONNECTICUT

I

145

3 ISLAND

FIG.15. Location map of mussel sampling sites: MC, Morris Cove; UI, United Illuminating Company, New Haven harbor generating plant site; OB, Oyster Bay; EN, Eaton’s Neck; PJ, Port Jefferson; WC, west side of Connecticut River mouth; EC, east side of Connecticut River mouth.

Consequently, the chemical composition of the ingestible particles could be inferred from two environmental indicators: the composition of the sediments at the sediment-water interface, and the composition of the bulk water (including the fine-grained particles) associated with the or-

I

2

34

5

6 7 89

MCUD

11 W EC

12

FIG.16. Copper concentrations of soft tissues of mussels at sites designated in Fig. 15. The bar labeled LI represents the range of values from the Long Island north shore sites OB, EN, and PJ. The locations designated WB, H, NH, and C are the Whitestone Bridge, Housatonic River, New Haven Harbor, and Connecticut River as shown in Fig. 15. (Data from sources described in text.)

K. K. TUREKIAN et a / .

146

I'

I

" ' I

'I'

I I

I

i i

I

C

I-

w 3.

(LII I

I

I1

I

I

2

34

5

6 7 89

I I (

It I

I I

I

I

_

tz

II W E

MUID

FIG. 17. Zinc concentrations of soft tissues of mussels in Long Island Sound. (See legends of Figs. 15 and 16 for key.)

$7

H

C

1

1 1

E l Q

NH

mt I

I

I I

I

I I I I

Ill

2

34

5

6 7 89

MUD

I

I I

I1 W E

t

(y)

12

F ~ G18. . Cadmium concentrations of soft tissues of mussels in Long Island Sound. (See legends of Figs. 15 and 16 for key.)

I

2

34

5

6 7 89

MUD

II W EC

12

FIG.19. Lead concentrations of soft tissues of mussels in Long Island Sound. (See legends of Figs. 15 and 16 for key.)

147

NUCLIDES IN LONG ISLAND SOUND I

I I

I

.. w

H

a 1

II I

I I I I

N

1

II

I

C

H

E "tI

2

34

5

6 7 89

1 I I MCUID

1

I I

II W EC

I

P

FIG.20. Nickel concentrations of soft tissues of mussels in Long Island Sound. (See legends of Figs. 15 and 16 for key.) Considering the trace-metal maps for the top 4 cm of Long Island Sound sediment cores (Figs. 7-9), there is a marked similarity between the areas of high metal concentrations in the sediments and high concentrations in the mussels and oysters. Similarly, a comparison of the nickel concentration in mussels (Fig. 20) with the coastal distribution of Ni in Long Island Sound water (Fig. 21) also shows a marked correlation. [Nickel is the only element determined in the mussel study that has also been extensively determined in water samples from along the Connecticut coast (Turekian, 1971).] We conclude that the primary source of the metals found in elevated levels in the soft tissues of mussels and oysters is the suspended organic-rich debris in the Sound. This is accentuated where a significant source of metal-bearing organic-rich particles from human activities is introduced by direct supply or secondary resuspension. Therefore, a strong correlation exists between high metal concentrations in all components of the coastal system: water, sediment, and organisms, and the proximity of polluted fresh-water stream and sewer discharges. 5 . PROCESSES AFFECTING THE DEPOSITION AND ACCUMULATION OF

TRACEMETALSIN LONGISLAND SOUND SEDIMENTS In the previous sections the case was made for two major classes of trace-metal impingement on Long Island Sound. One type is the supply by polluted streams and sewer outfalls, which, on the basis of the distributions of trace metals in the sediments and near-shore suspensionfeeding bivalves, was inferred to be predominantly in the form of particles. The other is atmospheric supply, some part of which presumably is in

NUCLIDES IN LONG ISLAND SOUND

149

dissolvable form. Whatever the source, two processes act in the estuary to determine the ultimate distributions of the metals. The association with suspended particles in the Sound will remove metals from the dissolved state and the movement and, ultimately, the accumulation of this suspended material will, to a large degree, determine their final repositories, although some diagenetic redistribution may also occur. In order to follow these processes the natural radionuclides, 234Th(24 days), "OPb (22 yr), and 'Be (54 days) provide the best prospects for tracing both the rate of removal of nuclides to the sediment surface and the processes governing their ultimate horizontal and vertical distributions. The properties of each of these nuclides are listed in Table 11. 5.1. Water Column Scavenging

Benninger et al. (1975) and Benninger (1978) showed that the 210Pb concentration of unfiltered Long Island Sound water was correlated with the amount of suspended matter in the water sample and that the intercept at zero concentration of particles yielded a zero value for "'Pb (Fig. 22). This indicated that virtually no "'Pb was dissolved in the water despite the continuous supply from the atmosphere. Clearly the residence time for "'Pb in Long Island Sound water must be short relative to final removal to the sediment pile, since high 210Pbvalues are found in the surface waters of the open ocean where the residence time has been determined to be about 1 yr (Nozaki et al., 1976). We depend on other tracers to determine more precisely the residence time of highly adsorbed chemical species in Long Island Sound relative

TABLE11. THEPROPERTIES OF NATURALRADIONUCLIDES USEFULIN DETERMINING METALPATHWAYS IN LONG ISLAND SOUND Method of production

Nuclide

Half-life

234Th

Decay of dissolved 23RUwhose concentration is directly proportional to salinity 54 days Cosmic rays on atmospheric atoms; delivery by precipitation 22 yr Predominantly supplied from the atmosphere where it is produced by the decay of 222Rn;delivery by precipitation

'Be 2'0Pb

24 days

150

K. K. TUREKIAN et d.

0

4 8 12 16 Suspended sediment ( x

20

24

gm/kg) FIG.22. Total *"Pb concentration in surface water of Long Island Sound versus suspended matter concentration. (After Benninger et al., 1975, and Benninger, 1978.)

to removal to the sediments. The most useful nuclides are 234Thand 'Be since they have sufficiently short half-lives to be used for the assay of scavenging on a short time scale. Aaboe et al. (1980) have shown that the residence time of 'Be in Long Island Sound is less than 10 days. Aller and Cochran (1976) and Aller et al. (1980) using the 234Th/238U in Long Island Sound water and the standing crop of unsupported 234Thin Long Island Sound sediments indicate that thorium has a residence time of less than 10 days and probably as low as 1 day. This is compatible with the observations of Kaufman et a / . (1980) on New York Bight apex waters. We conclude that the residence times relative to permanent removal to sediments of all metals behaving like Pb, Th, or Be during adsorption are short in Long Island Sound, and thus the metals will be transported to the sediments virtually irreversibly, although there may be preferred repositories such as tidal mud flats for some of the metals. 5.2. Horizontal Distribution Aller et al. (1980) have used the distribution of 234Thin the sediments of Long Island Sound to show that there is rapid homogenization of the

NUCLIDES IN LONG ISLAND SOUND

151

incohesive fine-grained fraction deposited at the sediment-water interface. Figure 23 shows that in the top 0-1 cm of the sediment pile there is a close correlation between unsupported 234Th(produced in the water column by the decay of 238Uand adsorbed) and 232Th,which is an indicator of the fine-grained fraction. Moreovx, Aller et al. (1980) showed that there is no strong relationship between the 234Thinventory in the sediments and the 234Thproduction in the immediately overlying water. The latter should be directly related to the depth of water and salinity, which together determine the total production rate of 234Thfrom 238U. Instead, the 234Thinventory (limited mainly to the top 5 cm of the sediment pile) is largely determined by the amount of fine-grained component present at each location, and this fine-grained component is homogenized over the range of water depths on a time scale rapid compared with the halflife of 234Th(24 days). Because a trace-metal concentration distribution map shows a texture that is not related exclusively to grain-size, but also to point sources of pollution along the Connecticut coast (Figs. 7-9), this means that the net

FIG.23. Excess (unsupported)234Thconcentrationversus total 232Thconcentrationin the 0-1-cm depth layer of cores from Long Island Sound. (From Aller et al., 1980.)

152

K. K.TUREKIAN et a / .

homogenization process cannot be instantaneous or Sound wide. It probably works most effectively perpendicular to the long axis of the Sound, since tidal mixing does not lead to efficient homogenization along the long axis. The maintenance of gradients across the Sound axis implies either that the rate of supply of the metal from the coastal source is sufficiently high to survive the homogenization process or that different depths of biological mixing alter the trace-metal concentration in the surface sediments. The results of Curran et al. (1980) on the Z'oPoP'oPbcomposition of the soft tissue of mussels collected west-east along the long axis of the Sound (Fig. 24) may speak to the efficiency of homogenization along the tidal axis. The mussels from the western part of the Sound have 2'0Po/210Pb activity ratios generally less than 10 (with one set of high values at Rye, New York). Starting at Milford Point, east of the Housatonic River mouth, the ratio increases eastward toward the open ocean from about 12 to 36. A similar increase in this ratio from shore to open ocean was observed for plankton by Turekian et a/. (1974) and the same effect may be the cause of this trend in the mussels. This effect in both cases could be due to the dilution of low 210Po1210Pb sediment detritus with high 210Po/2'0Pb biogenic debris. Whatever the reasons for this gradient, it does show that homogenization of the suspended particles sensed by the mussels is not complete along the axis of the Sound. In summary, we believe that "reactive" trace elements and other substances are adsorbed very rapidly in Long Island Sound, with the mean

1 1

i

FIG.24, The 2'0Pof"Pb activity ratio in soft tissue of mussels from the Connecticut coast (see Fig. 15 for locations). (After Curran et al., 1980.)

NUCLIDES IN LONG ISLAND SOUND

153

life relative to removal to the sediment pile of the order of 1-10 days. We also believe that homogenization of the fine-grained fraction need not be basin wide, but may depend on the strength of the sources of the trace metals as well as the nonisotropic effects of tidal mixing. 6. PROCESSES AFFECTING THE VERTICAL DISTRIBUTION OF NUCLIDESIN THE SEDIMENT PILE 6.1. Establishing Chronologies: Time Scales of Accumulation and Bioturbation

Figure 1 1 shows a number of trace-element profiles in sediment cores raised from Long Island Sound. The increase in concentrations of the trace metals as the sediment-water interface is approached from below is interpreted as primarily the result of human activity in mobilizing these elements in greater and greater amounts since the industrial revolution. Establishing a chronology for metal input to the Sound is not straightforward, however. In sediments deposited from anoxic or quasi-anoxic waters, such as some of the Gulf of California basins, certain fjords, and Santa Barbara basin, there is no macrofauna disturbing the sediment. In such environments the sedimentary pile is an uncomplicated record of the changing properties of the sequentially added materials, and *"Pb has proven to be an invaluable tool to confirm or establish a chronology. Once established, this chronology acts as a firm measure of the year by year changes in trace-metal supply to that basin. In coastal sedimentary basins (like Long Island Sound) that remain oxygenated at the sediment-water interface for most of the time, a lush macrofauna exists. The burrowing and related activities of these organisms strongly influence the vertical distribution of sediments and generally act to confound the simple chronology of accumulation. Although it is true that below the zone of biological mixing (or "bioturbation") time-averaged sedimentary properties are retained, the length of the intervals averaged may be longer than the resolution required for interpreting relatively recent events. The depths and rates of particle mixing in Long Island Sound sediments have been studied by Aller and Cochran (1976), Benninger et al. (1979), Aller et al. (1980), and Krishnaswami et al. (1980). Depth profiles of both 234Thand 'Be show that the top 4-5 cm of Long Island Sound sediments must be mixed rapidly in order to distribute these short-lived nuclides throughout this zone. The mixing process can be quantitatively described

154

K. K. TUREKIAN et a/.

by analogizing it to eddy diffusion. Then the steady-state distribution of a radioactive tracer in the sediments is given by the following equation: a2N dN Dg--S-AN = 0 az2 az where D Bis the particle-mixing coefficient, N is the number of atoms of nuclide of interest, S is the sediment accumulation rate, A is the decay constant, and z is the depth in the sediment column. [Porosity and the mixing coefficient are assumed to be constant in Eq. (6.1).] The effect of S on the solution to Eq. (6.1) depends upon its value relative to the rate of decay of the nuclide of interest. For short-lived nuclides like 234Th and 7Be, sediment accumulation is negligible over several half-lives and their distributions in the sediment column are governed by mixing and radioactive decay. For '"Pb (see later) this is not the case, and both mixing and sedimentation effects must be considered. Values of D, in surficial sediments of Long Island Sound have been determined from the distribution of 234Th(Aller et al., 1980); DB ranges (3-10) x cm2 sec-I. The fact that these values pertain to the upper few centimeters of sediment is shown from depth distributions of longer lived tracers. Such an example is shown in Fig. 25, which compares '"Pb profiles from two stations in the central Sound. In both cases, the "OPb activity is nearly constant in an upper zone of 4-10 cm and decreases quasi-exponentially below. One interpretation of the decrease in '"Pb activities is that it is due to sediment accumulation alone. Rates of 0.1-0.6 cm yr-' are calculated from the profiles. However, this explanation cannot be valid if the average rate of accumulation of 0.033 cm yr-' for the central Sound (Bokuniewicz et al., 1976; see Gordon, this volume) is taken as typical of present accumulation rates. Indeed, the presence of Pu in the zone of "OPb decrease in one of these cores (NWC-102975:Benninger et al., 1979) suggests that both sediment accumulation and mixing are important in producing the observed 'I0Pb distributions. Benninger et al. (1979) used cm' the Pu profile in core NWC-102975 to calculate a DB of -2 x sec-' below the surficial mixed zone (Fig. 26). This is at least an order of magnitude less than values obtained for the top few centimeters using 234Th. Using the calculated mixing coefficients for a rapidly mixed surficial layer and a more slowly mixed deep layer, Benninger et al., (1979) were able to reproduce the observed "OPb profile. On the basis of this analysis, the interpretation of exponential decrease in 'IOPb Long Island Sound sediments as reflecting a sediment accumulation rate (see e.g., Thomson et al., 1975) is not valid, and the effects of mixing may be dominant. Extending this concept of particle mixing by organisms further, it is

NUCLIDES IN LONG ISLAND SOUND

155

FIG.25. The 2’0Pb“rates” of accumulation for cores at NWC-102975 and DEEP-102375 (see Fig. 1la for locations). The highest “rate” also has the highest standing crop of *“Pb. The 210Pbversus depth plot, which formally yields a “rate,” is inferred to be due primarily to biological mixing. (Data from Benninger ef al., 1979 and unpublished results obtained at Yale University.)

possible to imagine mixing rates continuing to decrease with depth in the sediment column. Figure 27 shows the effect such a pattern will have on depth profiles of nuclides with different half-lives where mixing rates are assumed constant over discrete sediment layers. In addition to eddy diffusion-like mixing of sediment, Fig. 27 also shows the existence of singlemixing events by deep burrowers that bypass concentration gradients and result in tracer input below a zone of continuous mixing. This is an explanation of the “spikes” of metal and ’‘OPb concentrations observed by Thomson et al. (1975) and Benninger et al. (1979) at depths well below the beginnings of the anthropogenic trace-metal imprint on Long Island Sound. After a long enough period of time, there will be a distribution of trace metals and long-lived man-made radionuclides like plutonium and the bomb-produced component of the ‘4C burden to yield a mixing coefficient for long-term, deep-particle mixing. The fact that this deep bioturbation rate constant is considerably smaller

156

K. K.TUREKIAN et a/.

02

c/c. 0.4 0.6

0.8

1.0

0.2

c/c.

0.4

0.6 -0.8

____

I!

-

D.2 x lo-' cm2/sec

FIG. 26. Approximate best fits to normalized 239.240Pudata (filled circles) assuming Pu transport in layer 2 by particle "diffusion" only. Co, concentration (dpdcm') in depth interval 0-2 cm. Layer 2 begins at (a) x = LI = 2 cm and (b) x = LI = 3 cm; in (b) the Pu data are recalculated on the assumption that the Pu concentration is uniform over 0-3 cm. For both (a) and (b) particle diffusion is considered to stop at x = Lz = 10 cm. Solid and broken curves represent two different boundary conditions for the top of layer 2. Solid inventory in 0-10 curve: C, const. = Co,0 < x < LI for all t 2 0. Broken curves: 239.240Pu cm of core NWC-102975 contained in 0 < x < LI at t = 0, with no further addition of Pu. t = 0 is taken to be 1965, so that the time available for diffusive transport is 10 yr. (From Benninger et al., 1979.)

than the near-interface values can be seen by comparing the I4Clong-term rate of accumulation with the expected geophysically determined rate. Such a study was made by Benoit et al. (1979) on the same core analyzed by Benninger et al. (1979). Figure 28 shows the I4Cdistribution with depth in the organic fraction of the core, sampled with the aid of x-radiography to avoid obvious low density and recent burrows. The samples showed no *loPbor trace-metal spikes and therefore would not be expected to show any bomb I4C spikes. Another factor, however, rules out the possibility of observing too high a spike even if near-surface sediments were transported into the deep burrows; the I4C-rich planktonic material is metabolized rapidly in the strongly bioturbated zone of the upper 5 cm of the sediment pile. [Benoit et al. incorrectly estimated a residence time of this carbon source of about 24 days in their paper. The correct residence time is about two years (Turekian et al., 1980).] The carbon preserved in the sediment column is refractory organic matter probably mainly derived from land as evidenced by its 2300 yr BPage at the time of deposition. The I4C rate of accumulation determined for this core is about 0.07 cm yr-I, which is not very different from the geophysically estimated rate for this location of 0.06 cm yr-'. Nevertheless, the figure is higher than the rate of 50.05 cm yr-' calculated from "OPb and plutonium systematics (Benninger et al., 1979) and may indicate that the rate of accumulation

NUCLIDES IN LONG ISLAND SOUND

-+

*

I

c-

E

5 -

I

In

c ._

I I

C, Dc <
I

5 -

I

0

I

n

'I

157

i I

i

i

i

FIG. 27. Schematic representation of depth profiles of excess radionuclides (1,2,3) in a mixed deposit. Decay constants are assumed to decrease in the order A, B A2 B A3, and mixing ("particle diffusion") coefficients (DA,DB.Dc,DD)decrease with increasing depth (zones A, B, C, D) in the sediment. The concentration profiles are continuous over depth intervals where mixing is rapid on the time scale of radioactive decay and discontinuous below, except that discontinuities may occur at depths where the mixing regime changes (e.g., curve 2). Mixing and sediment accumulation both influence the shapes of the profiles in the continuous segments, except in zone D which is unmixed. (From Benninger et a / . , 1979.)

is lower than that calculated by ignoring the effect of deep bioturbation on the l4C profile. 6.2. Determination of the Final Repository of Metals Introduced into Long Island Sound

An additional effect of the rate and depth of long-term mixing is seen in the 210Pbprofiles in Fig. 25: As the rate and depth of mixing increase so does the integrated amount of '"Pb in the core. This is also seen by plotting the integrated '"Pb activity as a function of the apparent sediment accumulation (largely dominated by mixing) in cores from a traverse from New Haven to Long Island (Fig. 29). The geophysically determined sediment accumulation rates do not vary by more than a factor of 2, but the '"Pb apparent rates vary by a factor of at least 6. Although it is possible that the 210Pbdata represent true differences in local sediment accumu-

158

K. K. TUREKIAN et a / .

T

CE

\

7

t3.(

\ FIG. 28. The I4C distribution with depth in organic fraction of core NWC 102975. (From Benoit et a / . , 1979.)

N

5

\

k

0

aQ

2

u)

In

0)

0

a 1

2

0

r 0) C

-c> I

0‘

I

Q2

I

I

0.4

I

I

I

0.6

‘lOpb%ediment accumulation rate“ (cm/yr)

FIG. 29. The relationship between excess (unsupported) *‘‘Pb inventory and the formal “sediment accumulation rate” [TTM 1 and TTM 2 are cores described in Thomson et al. (1975); NWC is NWC-102975 described by Benninger et nl. (1979); DEEP is from unpublished Yale University data displayed in Fig. 25 and FOAM is from unpublished data obtained at Yale by E. Dion.]

NUCLIDES IN LONG ISLAND SOUND

159

lation rates obscured by the geophysical survey, the more likely reason for this disparity appears to lie in the spatial variability in rates of deep burrowing. Excavation of deep burrows and their subsequent infilling exchanges 210Pb-and metal-poor older particles for 'IoPb- and metal-rich surficial, younger particles. Rapid turnover of the upper 4-5 cm of the sediment column by resuspension and lateral mixing of fine-grained sediments provides a continuing source of "OPb- and metal-rich particles. Slower turnover of the deeper sediment horizons results in storage of these particles in the areas of the bottom where deep burrowing is intense. These variations in mixing depth and intensity are consistent with the kinds of macrofauna dominating along the transect (McCall, 1977; Aller, this volume). Another expression of this is found in the copper distribution in the upper parts of cores along the transect (Fig. 30). Sampling at 1-cm intervals shows that the copper distribution with depth differs across the Sound. FOAM-I, which shows little or no deep bioturbation, shows a sharp decrease in Cu and a small average concentration, whereas DEEP1, at the other extreme, is marked by a consistently high concentration of Cu at all depths sampled, indicating deeper bioturbation. The two NWC cores occupy intermediate positions. From these insights we conclude that the total burden in the sediment pile of a substance introduced at the sediment-water interface over a period of time will depend on the rate and depth of mixing. Rapid mixing to great depths will yield a higher total burden of the substance introduced isotropically into the Sound than mixing limited to very shallow depths in the sediment pile. Thus, the final repositories of metals in Long Island Sound depend not only on their association with the fine-grained fraction and the extent of its horizontal homogenization, but also on the spatial variability in depth and rate of vertical particle mixing that govern the total integrated metal content in the sediment column. 6.3. Major Sources of Metals Delivered to Long Island Sound From our knowledge of what controls the distribution of *'OPbin Long Island Sound sediments we can test the following question: How much of the trace-metal content of Long Island Sound sediments may be explained by an atmospheric source and how much by supply from injections along the coast by sewer outfalls and polluted streams? To answer the question, we integrate the total excess "OPb in a core and compare it with the integrated excess metal content. Benninger (1978) has shown that the 2roPbcontent in Long Island Sound sediment is due predominantly to atmospheric supply and that there is no loss from the Sound. We then

3

0

pD

0

0

--

n

n

~ " . ,-.-.---

.-

-

Y)

'.

-.

A

0

2

,

0

z

z

i., *. . . .....

. ..,

----

I

CU

4

z

8

5 a 0 LL

I

-

z

Y

I

-

..... 0

L

$

Y Q

161

NUCLIDES IN LONG ISLAND SOUND

TABLE111. TOTALEXCESSMETALTO EXCESS"'Pb STANDING-CROPRATIOSFOR SALTMARSHAND LONGISLAND SOUND DEPOSITS Total excess metal (pg cm-') to "'Pb standing crop (dpm cm-') Location Farm River salt marsh (Branford, Connecticut) Long Island Sound sediment cores Station NWC Station TTM-2

CuP'OPb

ZnI2"Pb

Pb?"Pb

13

19

13

34 17

33 32

12 11

compare the ratio of integrated excess trace-metal content to integrated 210Pbcontent found in sediment cores to the same ratio found by McCaffrey (1977) and McCaffrey and Thomson (this volume) in the Farm River salt marsh. This normalization overcomes the problem of both horizontal and vertical mobility in the sediment so long as the excess metals are confined to the same depth horizons that contain excess 2'0Pb. (This is possible because the intense human mobilization of metals began about 100 years ago, which is also the range of '"Pb excess.) When such a calculation is made for two cores (Table HI),one analyzed by Thomson et al. (1975) and the other by Benninger et al. (1979), it shows that all the Pb could be explained as of atmospheric origin, and half the zinc and copper. Presumably the additional burden of zinc and copper comes from the coastal high metal-particle sources discussed earlier. 7. SUMMARY

Our observations about the distribution of trace metals in Long Island Sound, as elucidated by our studies of the behaviors of 234Thand "OPb, lead us to the following generalizations: (1) Comparison of trace-metal distribution maps for Cu, Zn, and Pb with a sediment grain-size map shows that the primary control is the association of Cu, Zn, and Pb with the fine-grained fraction. (2) Imprinted on this feature is the increase in concentration of these trace metals near metal-polluting sources such as the Housatonic River and the East River (in New York). Thus, although the association of trace metals with the fine-grained fraction is established, the complete homogenization of this fraction does not proceed fast enough to obliterate intensive local sources of metals. (3) 234Th,produced by the decay of 238Udissolved in seawater, is re-

162

K. K.TUREKIAN e l a/.

moved to the sediments on a very rapid time scale (with a mean residence time in the water column of less than 10 days). The distribution of 234Th in the upper few centimeters of sediments throughout Long Island Sound is controlled by the amount of fine-grained material (identified either by percent loss on ignition or 232Thconcentration), similar to the trace metals. The lack of a positive correlation between the integrated 234Thactivity in the sediments and water depth implies that the fine-grained component tends to be homogenized on a rapid time scale, at least over distances that include a range of depths of water. However, the preservation of trace-metal patterns that reflect local pollution sources suggests that homogenization, although rapid, is not basin wide. (4) In a north-south transect across the Sound, the long-term sediment accumulation rate (based on the sediment thickness to a reflecting horizon inferred to correspond to 8000 years ago) varies by no more than a factor of 2. The "sediment accumulation rates" formally calculated from the exponential decrease of excess '"Pb with depth in cores along this traverse are at least a factor of 2 greater than the long-term rates and vary by an order of magnitude, with the higher values toward the deeper parts of the Sound. We infer that these exponential curves are dominated by particle mixing by infauna which transport 'IOPb deep into the sediment pile. ( 5 ) There is a direct correlation between the formal 210Pb"sediment accumulation rate" and the total standing crop of excess '"Pb in the sediment column. This indicates that although the top centimeter of the sediment core is homogenized on a rapid time scale, the rate and depth of bioturbation determine the amount of storage of "OPb in the sediment column at different locations in the Sound. Examination of macrofauna in the cores studied supports the argument for the biological control on '"Pb inventories. Areas of high *IOPbinventory are characterized by deeper mixing by infauna than are sediments with low 210Pbinventory. (6) The general pattern is seen for copper and by inference for other metals and thus can be used to predict the long-term repositories of trace metals injected since the beginnings of marked human use of metals. (7) Scaling the measured "OPb atmospheric flux in the region to the total atmospheric flux of metals over the past 100 years as determined in a salt marsh, the relative contributions of atmospheric and stream and sewage supply of metals to the Sound sediment can be estimated. In two cores so analyzed, all of the lead is inferred to be of atmospheric origin and about half of the copper and zinc. These proportions can be expected to vary westward along the axis of the Sound as local source terms from municipal and industrial pollution become more important. Thus, in relatively sheltered Long Island Sound, the long-term storage

NUCLIDES IN LONG ISLAND SOUND

163

of anthropogenically mobilized trace metals that became associated with fine-grained material is most affected by the rate and depth of biological particle reworking. In less protected areas such as the New York Bight, the role of physical disruption and redistribution becomes more important. ACKNOWLEDGMENTS This research has been supported mainly by the Department of Energy through grant EY75-5-02-3573. The United Illuminating Company also provided some financial assistance. Acknowledgment is also made to the donors of the Petroleum Research Fund administered by the American Chemical Society.

REFERENCES Aaboe, E., Dion, E. P., and Turekian, K. K. (1980). Be-7 in Sargasso Sea and Long Island Sound waters. J . Geophys. Res. (submitted). Aller, R. C., and Cochran, J. K. (1976). *'4Th?'"U disequilibrium in nearshore sediment: Particle reworking and diagenetic time scales. Earth Planet. Sci. Lett. 29, 37-50. Aller, R. C., Benninger, L. K., and Cochran, J. K. (1980). Tracking particle-associated processes in nearshore environments by use of 234Th?38Udisequilibrium. Earth Planet. Sci. Lett. 47, 161-175. Applequist, M. D., Katz, A., and Turekian, K. K. (1972). Distribution of mercury in the sediments of New Haven (CT) Harbor. Environ. Sci. Technol. 6 , 1123-1124. Benninger, L. K. (1976) The use of uranium-series radionuclides as tracers of geochemical processes in Long Island Sound. Ph.D. Thesis, Yale University, New Haven, Connecticut. Benninger, L. K. (1978) *"Pb balance in Long Island Sound. Geochim. Cosmochim. Acta 42, 1165-1174. Benninger, L. K., Lewis, D. M., and Turekian, K. K. (1975). The use of natural Pb-210 as a heavy metal tracer in the river-estuarine system. In "Marine Chemistry in the Coastal Environment" (T. M. Church, ed.), pp. 202-210. Am. Chem. SOC.Symp. Ser. 18. Benninger, L. K., Aller, R. C., Cochran, J. K., and Turekian, K. K. (1979). Effects of biological sediment mixing on the '"Pb chronology and trace metal distribution in a Long Island Sound sediment core. Earth Planet. Sci. Lett. 43, 241-259. Benoit, G. J., Turekian, K. K., and Benninger, L. K. (1979). Radiocarbon dating of a core from Long Island Sound. Estuar. Coastal Mar. Sci. Bokuniewicz, H. J., Gebert, J., and Gordon, R. B. (1976). Sediment mass balance in a large estuary (Long Island Sound). Estuar. coastal. Mar. Sci. 4, 523-536. Carmody, D. J., Pearce, J. B., and Yasso, W. E. (1973). Trace metals in sediments of the New York Bight. Mar. Pollut. Bull. 4, 132-135. Curran, D., Benninger, L. K., and Turekian, K. K. (1980). Metals, "'Pb and *"Pb in the soft tissue of the blue mussel (Mytilus edulis) as a function of location along the northern shore of Long Island Sound. In preparation. Feng, S.Y., and Ruddy, G. M. (1974). Zn, Cn, Cd, Mn, and Hg in oysters along the Connecticut coast. In "Final Report to office of Sea Grant Programs by the University of Connecticut Marine Sciences Institute," pp. 132-161. Gibbs, R. J. (1967). The geochemistry of the Amazon River System, Part I: The factors that control the salinity and the composition concentrations of suspended solids. Geol. SOC. Am. Bull. 78, 1203- 1232.

164

K.K. TUREKIAN ef d.

Goldhaber, M. B., Aller, R. C., Cochran, J. K., Rosenfeld, J. K., Martens, C. S., and Berner, R. A. (1977). Sulfate reduction, diffusion and bioturbation in Long Island Sound sediments. Report of the FOAM group. Am. J. Sci. 277, 193-237. Greig, R. A., Reid, R. N., and Wenzloff, D. R. (1977). Trace metal concentrations in sediments from Long Island Sound. Mar. Pollut. Bull. 8, 183-188. Gross, M. G. (1976). Sources of urban waste. I n “Middle Atlantic Continental Shelf and the New York Bight” (M. G. Gross, ed.). Am. SOC.Limn. Oceanog. Spec. Symp. Ser. 2, pp. 150-161. Kaufman, A., Li, Y. H., and Turekian, K. K. (1980). The removal rates of Z34Thand 228Th from waters of the New York Bight. Earth Planet. Sci. Lett. (in press). Krishnaswami, S., Benninger, L. K., Aller, R. C., and Von Damm, K. L. (1980). Atmospherically-derived radionuclides as tracers of sediment mixing and accumulation in nearEarth Planet. shore marine and lake sediments: Evidence from ’Be, ”‘Pb, and 239.240Pu. Sci. Lett. 47, 307-318. Lazrus, A. L., Lorange, E., and Lodge, J. P., Jr. (1970). Lead and other metal ions in precipitation. Environ. Sci. Technol. 4, 55-58. Lewis, D. M. (1977). The use of ”‘Pb as a heavy metal tracer in the Susquehanna River system. Geochim. Cosmochim. Acta 41, 1557-1564. McCaffrey, R. J. (1977). A record of the accumulation of sediment and trace metals in a Connecticut, U.S.A., salt marsh. Ph.D. Thesis, Yale University. McCall, P. L. (1977). Community patterns and adaptive strategies of the infaunal benthos of Long Island Sound. J. Mar. Res. 35, 221-266. Nozaki, Y., Thomson, J., and Turekian, K. K. (1976). The distribution of ’“Pb and ”*Po in the surface waters of the Pacific Ocean. Earth Planer. Sci. Let?. 32, 304-312. Sholkovitz, E. R. (1976). Flocculation of dissolved organic and inorganic matter during the mixing of river water and sea water. Geochim. Cosmochim. Acta 40,831-845. Thomson, J., Turekian, K. K., and McCafTrey, R. J. (1975). The accumulation of metals in and release from sediments of Long Island Sound. I n “Estuarine Research” (L. E. Cronin, ed.), Vol. 1, pp. 28-44. Academic Press, New York. Turekian, K. K. (1971). Rivers, tributaries and estuaries. I n “Impingement of Man on the Ocear,” (D. W.Hood, ed.), pp. 9-73. Wiley, New York. Turekian, K. K. (1979). Trace metals. I n “New Haven Harbor Ecological Studies.” Summary Report-United Illuminating Report to Connecticut Dept. of Environmental Protection. Turekian, K. K., Berner, R. A,, and Gordon, R. B. (1972). Marine sediments, New Haven harbor, Connecticut: Results of analyses and proposals for dredge spoil disposal: Addendum 12 of Environmental Report, Coke Works Site, June 1971. (Conducted for the United Illuminating Co. and coordinated by Normandeau Associates, Inc.) Turekian, K. K., Kharkar, D. P., and Thomson, J. (1974). The fates of ”‘Pb and ’“Po in the ocean surface. J. Rech. Atmos. 8, 639-646. Turekian, K. K., Benoit, G. J., and Benninger, L. K. (1980). The mean residence time of plankton-derived carbon in a Long Island Sound sediment core: a correction. Estuar. Coastal Mar. Sci. 11 (in press). Volchok, H. L., and Bogen, D. (1971). Trace metals-fallout in New York City. I n “Health and Safety Laboratory Fallout Program Quarterly Summary Report.” U.S.A.C.E., April I , 1971, New York.

A RECORD OF THE ACCUMULATION OF SEDIMENT AND TRACE METALS IN A CONNECTICUT SALT MARSH RICHARD J. MCCAFFREY" AND JOHNTHOMSON~ Department of Geology and Geophysics Yale University New Haven, Connecticut

1.

2.

3.

4.

Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Experimental Methods and Results . . . . . . . . . . . . . . . . . . . . . . . . 2.1. Site Selection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.2. Peat Properties . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.3. Radionuclides . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.4. Trace-Metals Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.1. *laPb Dating of the Deposit and Evaluation . . . . . . . . . . . . . . . . . 3.2. Physical and Chemical Evidence on Lack of Disturbance and Chemical Immobility . 3.3. Atmospheric Fluxes Recorded in the Salt Marsh . . . . . . . . . . . . . . . 3.4. Salt-Marsh Accretion . . . . . . . . . . . . . . . . . . . . . . . . . . . . Summary and Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

1 65 169 I69 172 181

. . .

183 189 I89 198 210 22 I 221 229

1. INTRODUCTION Salt marshes occur over a wide latitudinal range from above the Arctic circle to the tropics (Chapman, 1960). Some occur inland in association with brackish waters, but most are intertidal grasslands found along seacoasts in relatively protected areas. This work concerns a salt marsh on the Connecticut coast, formed since the last glaciation, during a prolonged period of sea-level rise relative to land. Salt marshes are conveniently classified into zones of so-called high marsh and low marsh. As a first approximation, high marsh may be described as a grass-covered horizontal platform near the mean-high-water level. The characteristic, benchlike nature of the high marsh and its associated, distinctive set of salt-tolerant plant species distinguish this zone from low marsh. High marsh in the study area is covered with dense stands of grasses, typically dominated by Spartina patens (Ait.) Muhl (salt hay). The tall grass, Spartina altern8oru Loisel (salt thatch), almost

* Present address: Graduate School of Oceanography, University of Rhode Island, Kingston, Rhode Island 02881. t Present address: Institute of Oceanographic Sciences, Wormley, Godalming, Surrey, GUS SUB, England. 165 ADVANCES IN GEOPHYSICS,VOLUME

22

Copyright 8 1980 by Academic Press, Inc. All rights of reproduction in any form reserved. ISBN 0-12-018822-8

166

RICHARD J. McCAFFREY AND JOHN THOMSON

exclusively occupies the low areas fringing the high marsh, however, a short form of this species may be found at some higher elevations (Miller and Egler, 1950). The elevational distribution of these and other species is generally considered to reflect tolerance to immersion in saline water (Johnson and York, 1915). Adams (1963), working along coastal North Carolina, found that a given species tends to occupy a limited elevational range above mean sea level for a given tidal regime. Stalter and Batson (1969) showed that survival of high-marsh grass, notably S.patens, was unlikely after transplanting to a lower tidal position. Other environmental factors, such as nutrient availability (Adams, 1963; Valiela and Teal, 1974) and competition for space (Blum, 1968; Niering and Warren, 1974) may also affect distribution. As grasslands, salt marshes are one of the productive facets of the nearshore environment. A portion of the net production appears as thick stands of characteristic vegetation, the remainder as subsurface roots and rhizomes. Silt and clay-sized deposits from outside sources also accumulate within the interstices of the organic matrix. Data from Connecticut salt marshes indicate that peat accumulations of about 1 m are common (Hill and Shearin, 1970), although depths of several meters have been noted elsewhere in southern New England (Johnson, 1967) and more than 3 m in at least one site at nearby Clinton, Connecticut (Bloom and Ellis, 1965). Extensive, qualitative evidence of relative vertical motion of land and sea in this area has now been developed. This evidence includes relicts of former uplands beneath the salt marshes of New England, such as soil profiles, in situ tree stumps, and freshwater peat now overlain by considerable thicknesses of salt-marsh deposits (Cook, 1857; Dawson, 1855; Bloom, 1961). Archeological evidence, numerous anecdotal references to datable historical artifacts, and other evidence supports arguments that relative submergence has been a widespread and generally continuing phenomenon in this area (Davis, 1910,1913; Johnson, 1967; Knight, 1934; Chapman, 1974). The advent of I4C-dating technique provided an opportunity to more closely describe the rise in relative sea level since the last worldwide glaciation. Application of the technique to salt-marsh samples yielded quantitative evidence for the continuity and vertical extent of sea-level rise both worldwide (Shepard, 1963; Scholl and Stuiver, 1967; Milliman and Emory, 1968) and in southern New England (Barghoorn, 1953; Fairbridge, 1961 ;Redfield, 1967). Radiocarbon dating of salt marshes revealed that the vertical development of the deposits has evidently kept pace with

SEDIMENT AND TRACE METALS IN A SALT MARSH

167

relative sea-level rise over the last few thousand years (Redfield and Rubin, 1962; Bloom and Stuiver, 1963). In nearby Clinton, Connecticut about 3 m of salt-marsh deposit has accumulated during the last 3500 years (Bloom, 1967). Dated samples are spaced hundreds to thousands of years apart, however, and the most recent sample is more than 900 years old; so possible variations within the intervals and in historical time cannot be resolved. Accumulation during the colonial and industrial eras is very poorly defined, but the marker bed experiments begun in 1962 by Bloom (1967) and recently updated and extended (Harrison and Bloom, 1974; Harrison, 1975) show that increasing thicknesses of peat are generally found above the marker layers in subsequent years, evidently due to new peat formation. Measurements at monthly intervals in Flax Pond salt marsh on the south shore of Long Island Sound show that accretion in an established S. alterniJ2ora area is practically monotonic and in reasonably good agreement with long-term average peat-accumulation rate (Amantano and Woodwell, 1975; Flessa et al., 1977; Richard, 1978). The manner of accretion, chemical properties, and the considerable physical protection afforded by the peat framework suggest that these deposits may contain useful depositional records. Coastal wetlands are generally considered to be sinks for waterborne particulate matter (Meade, 1972). Sediment storage areas such as mudflats and channels, however, often fail to show stratification, perhaps because these deposits tend to be uniformly fine-grained and/or intensely bioturbated (Ellis, 1962; Rhoads, 1967). In addition, their susceptibility to cycles of erosion and deposition severely limits their scope as historical repositories. Sediment incorporated within the peat matrix, however, is relatively well protected from physical disturbance and, therefore, its depth distribution may reflect historical sediment-generating processes in the environment. The mass of sediment incorporated into the peat is supplied from offshore and in riverine runoff. Ellis (1962) found that the proximate source of sediment input to the Great Marsh, a coastal marsh near Norwalk, Connecticut, was suspended material carried shoreward by tidal currents. Circulation studies of Long Island Sound show that bottom water flows westward into the Sound, then veers northward toward the coast in shoal water (Gordon and Pilbeam, 1975). Thus, sediment entering north-shore coastal salt marshes would tend to originate in the east. The proportioning between offshore and riverine sources, however, will strongly depend on local circumstances, e.g., proximity to the coast and intensity of stratification of adjacent waters. Salt marshes alongside well-stratified rivers some distance from the coast, as is the study marsh, may be expected to receive a much larger share directly from land sources. The present

168

RICHARD J. McCAFFREY AND JOHN THOMSON

work will give evidence of salt-marsh deposition dominated by riverine runoff. Knowledge of the chemical properties of salt-marsh deposits, notably their organic-matter rich, reducing, sulfidic nature, suggests that certain substances may be fixed in place by natural processes, thus possibly forming a record of chemical deposition. H,S is generated in these kinds of deposits, on occasion evidently to excess, by bacterial sulfate reduction (Galliher, 1933). Chemical principles predict that any available metal whose sulfide salt is sufficiently insoluble will spontaneously precipitate as metal sulfide. Berner (1970) has shown in laboratory experiments that iron minerals will react with H2S to rapidly form monosulfides and, ultimately, pyrite (FeS,), a process believed to be ubiquitous in anaerobic marine sediments containing metabolizable organic matter. Formation of metal sulfides besides iron has not yet been demonstrated in salt marshes, but there is good reason to suppose that certain trace-metal sulfides do form in these deposits. For example, Cu, Zn, and Pb sulfides are known to form in the laboratory when sulfide produced by bacterial sulfate reduction reacts with trace metals added in the form of soluble salts (Baas Becking and Moore, 1961 ; Bubela and McDonald, 1969). Furthermore, the addition of sulfide or precursor sulfate is found to effectively control heavy-metal toxicity in anaerobic digesters by precipitation of the metals as their sulfides (Lawrence and McCarty, 1965). Locally, Siccama and Porter (1972) showed that the Pb content was higher in the surface layer of the Farm River marsh than at depth and that the surface concentration decreased in salt marshes further east (away from New Haven), a phenomenon they felt to be due mainly to automotive Pb brought to the marsh surface in freshwater runoff and then chemically fixed. Vertical profiles, by themselves, are generally insufficient to identify sources or to rule out postdepositional rearrangement. Simple deposition, for example, cannot be credibly distinguished from deposition modified by mixing or remobilization. The present work investigates the possibility that a useful, historical record of deposition might be found in the salt marsh. After selecting a site thought to optimize chances of finding a physically intact record (see site selection criteria, Section 2. I), a series of investigations of the vertical distribution of indicator materials in both solid and pore-water phases was begun. The three trace metals, Cu, Zn, and Pb, were selected, a priori, as likely to be chemically immobilized through formation of their insoluble sulfides. Fe and Mn were also included in the study because of their reputed involvement in the conservation of trace elements (Jenne, 1968) and because of their sensitivity to redox conditions. Authigenic iron sulfide

SEDIMENT AND TRACE METALS IN A SALT MARSH

169

(FeS,) is of special interest because it is defucto evidence of the operation of the sulfate reduction process and the persistence of reducing conditions. It is conceivable that processes resulting in pyrite formation, in addition to exemplifying the sulfide-precipitation process, may actually include other trace metals in a chemical sense. A study of the vertical distribution of "'Pb was undertaken because of its utility as a tracer and as a potential means to establish a sediment chronology. This natural radionuclide is produced in the atmosphere by decay of gaseous *"Rn, is rapidly attached to the atmospheric aerosol, and is deposited on all exposed surfaces. After deposition, its chemical behavior is believed to mimic elements with similar chemical properties. Unlike stable elements, however, 2'oPbdecays with a 22-yr half-life (Hohndorf, 1969) and, under suitable conditions, its vertical distribution within the deposit may be used to define a chronology. 2. EXPERIMENTAL METHODSAND RESULTS 2.1. Site Selection

Choice of site for detailed study was based on the following criteria: (1) the surface and subsurface vegetation should be typical high-marsh species, e.g., S. patens; (2) the surface relief should be minimal; (3) the core should penetrate into peat formed before human influence became significant; and (4) the probability of physical disturbance in the past should be low. Salt hay has been cut on many Connecticut salt marshes in historical times for fodder and mulch, and drainage ditches were dug through almost all of them in the early 20th century to control mosquito breeding (Connecticut Agric. Exp. Stn. Rep., 1912). Many ditches are still maintained, so their location is usually obvious and easy to avoid. Relict ditches and areas previously hayed-implying disturbance by men or animals-are harder to avoid. It is possible, however, to minimize the probability of such disturbance by selecting areas that are difficult to approach due to the impediment of the natural drainage pattern. Aerial photographs (Connecticut Department of Environmental Protection, 1971 series, 1 in = 200 ft especially No. 30-1-4) proved valuable for this purpose, and also for identifying relict ditches or other anomalous features, and for finding areas of desirable high-marsh vegetation. Once promising areas within the Farm River marsh were selected, they were examined in the field. Several trial cores were raised using a tool designed for this task (McCaffrey, 1977) and local residents were inter-

170

RICHARD J. McCAFFREY AND JOHN THOMSON I

Farm River e'

X u r -

I

7Z"'W''

7

u FIG.1. Map of the study area.

viewed about the candidate sites. A part-time farmer (Stanley Kaczynski, b. 1915, personal communication), who had personally cut hay on the accessible parts of the salt marsh before 1935 and still lives within view of it, did not recollect anyone cutting hay or otherwise physically disturbing the area eventually chosen for intensive study. The chosen area (Fig. 1) is located in the eastern lobe of the Farm River salt marsh (41"16'00'", 72"51'06W), 7.9 km southwest of the center of

SEDIMENT AND TRACE METALS IN A SALT MARSH

171

the city of New Haven. This marsh fills its valley to a remarkably uniform level and straddles tidal channels carrying freshwater runoff from a 66km2 watershed (Thomas, 1972). The site selected for detailed analysis is located near the center of a rectangular area bounded on two sides by parallel drainage ditches and on the other sides by a natural channel and a small upland “island.” The deep stratigraphy (Fig. 2) was determined with a Davis peat sampler. All sampling was done at this site, except for a few cores taken from a similar salt marsh about 3 km to the east, on the north side of Indian Neck peninsula. Human influence on the watershed was very likely negligible prior to the arrival of European colonists in the 17th century. Since then, the naturally forested watershed has been subjected to extensive clearing and cultivation, and the area is now a rapidly developing suburb of the nearby urban and industrial center of New Haven. Man’s impact on the watershed can be inferred from demographic and land-use trends for this general area. According to Hicock (1970), about 95% of the land was forested before colonization. Reduction of forested areas began after imposition of a colonial economy based on subsistence agriculture. By 1800 about a third of Connecticut’s forests had been cleared to replace nutrient-impoverished farmland or for fuel or timber. The percentage of forested land passed through a minimum around the end of the Civil War (1865) (MacDonald, 1968; Harper, 1918). After this war the rate of farm abandonment exceeded the rate of establishment. This initiated an era of net decline in the area devoted to agriculture and permitted spontaneous reforestation to begin. As farming declined as an occupation, industrialization intensified and urban centers developed. 1

? ISLAND

m

1,0

~

~-

172

RICHARD J. McCAFFREY AND JOHN THOMSON

Census data for New England (cited by Harper, 1918) revealed that the percentage of the population employed in agriculture diminished from 62% in 1840 to 10% in 1910, while industrial employment increased from 28% to over 49%. Demographic trends during this time, shown in Fig. 3, portray the early rapid development of urban centers such as New Haven and the subsequent spillover onto outlying second-growth farmland, typified by the study area in East Haven. 2.2. Peat Properties Cores were obtained at low tide by first excavating a vertical access hole with a post-hole digger, then cutting one or more meter-length cores of about 9 cm x 24 cm cross section from the exposed sides, using the special coring tool. It is a common experience among those who stand in place on the high marsh that the surface slowly depresses and forms a water-filled pool around one’s feet. Based on this observation and the generally watersaturated condition of the peat, we anticipated that water would soon fill

1970 CENSUS NEW W V E N 137701 STATE OF CONN. M32217 EAST HAVEN 15120

1972 1912

1932

1912

1892 1872

1852

1832

1812

1792 17R

YEAR

FIG.3. The population of New Haven burgeoned shortly before the Civil War, in part due to the general movement from rural areas to manufacturing centers, and in part (about a third) due to a surge in immigration (Gilbert and Olmstead, 1910). This rapid increase was sustained until World War I, when census records for the city show an abrupt flattening. This behavior appears to be an artifact produced by growth across fixed political boundaries; suburbs such as East Haven tended to accumulate the spillover. More than 85% of the population of East Haven was added since 1920, especially since World War 11. The pace of change on the watershed has continued to accelerate to the present: by 1970 over 7400 housing units and more than 140 km of roads had been constructed (East Haven, 1970). Despite the population boom, there is still no known discharge from industry into the Farm River, and at least 75% of the domestic sewage is carried out of the watershed through the municipal sewage system.

SEDIMENT AND TRACE METALS IN A SALT MARSH

173

the cavity by seepage, but this proved minimal. The few liters of water that accumulated in the hole during the approximately 15-min coring operation appeared to enter near the top of the hole, rather than by subsurface flow. Vertical zonation, in the form of diffuse colored bands, is apparent in fresh cores, especially within the upper 50 cm. The same general features of the zonation were found to extend laterally over most of the high marsh, except near the creek banks. FRl lB, the core examined in greatest detail, featured: (1) S . patens growing at the top of the core, with apparently live roots and rhizomes extending to about 10-12 cm depth; (2) a thin (? 1cm), pink-colored layer of sediment present at the surface, presumably a result of oxidation; (3) a generally brown-colored peat beneath the surface oxidized layer, but visibly blackened at a depth of roughly 5-10 cm; and (4) a zone of gray-colored fine sediment, conspicuous in the peat between 26 and 42 cm depth, hereafter called the “clay band.” The meter-length core was cut parallel with the marsh surface at 2-cm depth intervals, then the peripheral 1-2 cm was trimmed, leaving a neat, firm, rectangular prism of peat. The prisms were placed in thin-wall plastic bags and the length of each of the 12 edges measured. Peat of this type held its shape well if handled carefully, allowing accurate measurements to be made. The peat samples were weighed in Petri dishes while fresh, after drying to constant weight at about 100°C, and again after ashing at 500 2 25°C. The organic matter of the salt marsh is photosynthesized by grasses growing at the surface. The net primary productivity is apportioned between the conspicuous above-ground growth and the inconspicuous, but appreciable, below-ground growth. The internal structure of the peat is conveniently studied by x radiography. X radiographs of FR11A, a core raised from the same hole as FRllB, are presented as Fig. 4. Another x radiograph, Fig. 5, was made of a very large S . patens tussock from nearby marsh, These x-ray images provide valuable evidence of growth processes and are helpful for identifying zones of apparently homogeneous peat within the cores. A key to the identification of the various grass species forming the peat has been developed by Niering et al. (1977), based mainly on their distinctive rhizome cross sections. Using an x radiograph as a map, subsamples representing the homogeneous zones may be selected for species identification. In this way it is possible to estimate the proportions of species present, and to assign a peat type based on dominance in each zone. Whether or not a significant preservational bias exists is yet to be demonstrated, so inference of the original species composition from old peat remains uncertain.

174

RICHARD J. McCAFFREY AND JOHN THOMSON

FIG.4. X radiographs of 1-cm-thick vertical slices of core FRllA show details of the structure of high-marsh peat, As the percentage of space at any given position occupied by organic matter increases, the penetrability to x rays and consequent lightness of the photographic print also increases. The plant parts visible in the print, mainly consisting of roots and rhizomes of CO.1-1-cm diam., occupy much of the available space and form an intricate, interwoven organic matrix. The light-shaded subround objects appearing throughout most of the core are plant parts viewed in cross section or, in a few cases, holes.

The interstices of the plant matter contain inorganic sediment and are generally saturated with saline water. The mass content of a given volume of material may be conveniently described in terms of bulk densities of its component phases. Mean-bulk densities for a core may be calculated from the weight and volume data averaged over all slices. Examination of the bulk density data for the entire core in Table 1 reveals that water constitutes over four-fifths of its mass, whereas dry

SEDIMENT AND TRACE METALS IN A SALT MARSH

175

FIG.4. (Continued.)

matter accounts for less than one-fifth. Of the dry-matter fraction, about 70% is inorganic and the remaining 30% is organic. Water dominates the mass, to the extent that the mean-bulk density of a fresh core is a mere 1.011 gm/cm3, which demonstrates that this material is essentially neutrally buoyant in water of the estuary. Conversely, the overall mean-bulk density of dry matter is only 0.2 gm/cm3,compared to about 0.65 gm/cm3 for Long Island Sound sediment (Thomson et al., 1975) and about 1.9 g d cm3for sandy upland soil. The surface of the salt marsh is not supported on a column of particles. Also note that the organic fraction mean-bulk density (0.056 gm/cm3) is not much greater than Sound sediment (0.04 gm/cm3).The reputation of the salt-marsh peat as an organic-rich material is hardly justifiable on a mass basis. The impressions of observers are

176

RICHARD J. McCAFFREY AND JOHN THOMSON

FIG. 5. X radiograph of an S. patens tussock. This x radiograph shows details of the internal structure of a relatively large tussock. The darker portions on the periphery and lower central part of the x radiograph are relatively rich in silt. The light-shaded region traversing the base of the tussock is composed of a mass of intertwined, 2-3-mm diam., rhizomes of, presumably, S patens. A number of similar plant parts are oriented vertically within the tussock, and very fine rootlets are evidently well dispersed throughout its entire volume, which extends some 10-12 cm above the adjacent surface of the marsh. Bar is 5 crn long.

SEDIMENT AND TRACE METALS IN A SALT MARSH

177

OF SALT-MARSH TABLEI. BULKDENSITY SEDIMENTS"

Condition

Fraction

Wet Dry Dry Dry Dry Dry Dry Dry Dry

Total Total Total Organic Inorganic Inorganic Inorganic Inorganic Inorganic

Depth Mean bulk interval density (cm) (gm ~ m - ~ ) 0-100 0-100 26-42 0-100 0-100 0-5 10-25 26-42 42-50

1.011 0.194 0.265 0.058 0.135 0.200 0.142 0.201 0.137

" Mean bulk-density values are based on the measured weights and dimensions of all 50 slices of core FRl lB, corrected for saw kerf. Values for the various subsections were estimated graphically using an expanded version of Fig. 6 and are considered accurate to within k0.003 (lu).

evidently biased by the fact that the salt-marsh organic matter is voluminous and still largely intact, whereas in bottom sediment it is comminute and diffuse. Even though the mass of organic matter is in fact small, the x radiographs show that the volume occupied or enclosed by roots and rhizomes is extensive. The bulk density of dry matter present in individual slices varies over the length of the core. When the accumulated mass per unit area is plotted against depth, the curves (Fig. 6) show that dry-matter accumulation has varied in a complex way during the period of accretion. Further insight into the nature of this variation is gained by separating the dry matter into its inorganic and organic phases, also shown in the Fig. 6. It is immediately obvious that essentially all of the variation in dry-matter accumulation is due to a varying inorganic-matter contribution; the organic material, in striking contrast, accumulates in a highly linear manner over depth. This suggests that inorganic and organic matter accumulate independently. Furthermore, the uniformity of the organic-accumulation process suggests that this process is controlled in some manner. The inorganic-matter bulk density is greater in the upper half of the core, reaching highest concentrations within the 0-5 cm and 26-42 cm depth intervals, where it averages 0.2 gm/cm3, or 50% above the mean for the whole core. The mean-bulk density over limited depths was estimated from the slope of the line through the appropriate segment of the

178

RICHARD J . McCAFFREY AND JOHN THOMSON 20

-

16

\o

E

(r

z

12

0

0

D E . P T H (cm)

FIG. 6. Accumulation of high-marsh peat. The variation in accumulated total dry-matter content (gm cm-*) of Farm River salt-marsh core FRllB and Long Island Sound bottomsediment core LIS 72-6 are plotted over each increment of depth. Total dry matter is broken down into inorganic (SWC residue) and organic (weight loss on ignition) fractions for the salt-marsh core. The line corresponding to the organic fraction is very nearly straight over the entire core, except for a region of changing slope within the upper 10 cm or so; whereas the inorganic-matter accumulation varies in a complex manner. The curve representing the accumulated total dry weight of sediment in Long Island Sound is shown for comparison. It exhibits the upward concavity expected for deposits where porosity decreases with depth. In contrast, there is no clear evidence of compaction in the salt marsh, even below the relatively dense layer at 26-42 cm, and the water content remains about 80% throughout.

accumulation curve, as shown in Fig. 6 for the 26-42 cm interval, and summarized in Table I. It is clear that the gray clay band noticed in the fresh core does indeed correspond to a zone of relatively high inorganicmatter content at 26-42 cm. The presence of the 0-5 cm zone of relatively high inorganic-matter content was not apparent in fresh cores, but is also indicated by its bulk density. Direct evidence for the formation of authigenic metal sulfide comes from x-ray microprobe and light-microscope examination of the saltmarsh sediments. In the presence of sulfur, iron monosulfides react to form pyrite, FeS,, which is known to occur as distinctive, characteristic aggregates of octahedral microcrystals of FeS, (framboids; Berner, 1970; Sweeney and Kaplan, 1973). In the Farm River samples, framboidal FeS, was found to be common within at least the upper 14 cm of core examined, either as discrete framboids of -lO-p,m diam. (Fig. 7) or as ordered clusters of framboids. Inspection of polished thin sections reveals a frequent association with the organic matrix, which appears to act as a template for their formation (see Fig. 8).

SEDIMENT AND TRACE METALS IN A SALT MARSH

179

FIG.7. A single pyrite framboid about 20-km diam. Specimen current image (a) shows the subround cross section of the framboid. X-Ray images of sulfur (b) and iron (c) are shown for the same specimen. The scan for copper (d), initiated before the phosphor completely faded (faint grey background), failed to detect a significant concentration within the framboid.

180

RICHARD J. McCAFFREY AND JOHN THOMSON

FIG.8. Ordered distribution of iron sulfide. The photomicrograph (a) shows individual pyrite grains dispersed throughout the matrix and clustered inside an organic (?) envelope, 7 0 x . The graininess typical of pyrite framboids is evident at 320x (b). X-Ray images of sulfur (c) and iron (d) portray the tendency of pyrite to conform to organic templates.

SEDIMENT AND TRACE METALS IN A SALT MARSH

181

These pyrite framboids, considered possible sites for incorporation of other metals, were scanned under the microprobe for iron, sulfur, and trace elements, including Mn, Cu, Zn, and Pb. Iron and sulfur were identified as major components of the framboids, as expected, but no trace metals were detected in them. In fact, all of many attempts to find evidence of trace metals in association with mineral or organic phases proved negative, in agreement with the results of Sweeney and Kaplan (1973). Although the detection limits during element scans were probably not better than 100 or perhaps even 10,OOO ppm, depending on the element and instrument conditions, if trace metals had been present in relatively pure phases in pyrite or other particles over a few microns in size, it is likely that they would have been observed. Since these trace elements are known to be present in bulk sediment, failure to detect them suggests that they are dispersed, perhaps as micron or submicron-size particles, thin surface coatings, or as absorbed species.

2.3. Radionuclides Details of the analytical techniques used to determine the activities of the radionuclides employed are given by Turekian et al. (1973). Briefly, 5 gm samples of peat ash were leached in hot 6 N HCl and the leachate analyzed for 226Raand 210Pb.Radium-226 was determined by recovery of 222Rngas produced by decay of parent 226Rain solution within a closed vessel for at least ten days, followed by alpha scintillation counting of the separated radon. The procedure was repeated at least twice. Blanks, estimated by extraction without sample in the recycle loop, were small relative to sample activity. After completing the 226Raanalysis, '"Pb was separated by ion exchange in the presence of stable Pb carrier, then precipitated as the chromate, mounted, covered with Mylar to absorb 210Poalpha particles, and finally determined by beta-counting in a proportional gas-flow counter at intervals during ingrowth of daughter "OBi. Background corrections were made by back-extrapolating the ingrowing activity to zero time. The total activity of '"Pb and 226Rawas estimated in each slice from core FR11B. '"Pb analyses were performed on contiguous, nominally 2cm-thick slices from the upper two-thirds of the 1-m-longcore. 226Rawas determined at discontinuous intervals throughout the upper half. The total 210Pbactivity comes from two sources: '"Pb in equilibrium with parent nuclides naturally present in sediment ("supported"), and 210Pbcontinually produced in the atmosphere from 222Rnand deposited on all exposed surfaces ("unsupported" or "excess"). Once isolated from

182

RICHARD J. McCAFFREY AND JOHN THOMSON

the atmosphere, the excess activity decays with its characteristic 22-yr half-life, which makes it possible to determine the age of a given layer of sediment. Excess activity is estimated by subtracting the supported activity from the total. The level of supported activity may be estimated from the activity of the parent nuclide if it is in secular equilibrium with its daughter, or it may be allowed to define itself by measuring the activity in sufficiently old sediment of similar composition. The latter course was followed in this work since the 210Pbactivity deep in the core reaches a level that is higher than 226Ra-its long-lived precursor nuclide. The empirical, or self-defined, supported 210Pbactivity of 0.83 dpdgm ash, used to calculate excess 210Pbhigher in the core, was taken as the mean activity over the 34-40-cm level. Below 50 cm the peat was distinctly different in appearance, and the 210Pbvalues probably represent a different depositional regime. The estimated excess activity A and associated uncertainty U, for each slice of core are listed in Table 11. The error, associated with each count is estimated using the propagation-of-errors method (Bevington, 1969). Counting statistics account for most of the imprecision, although uncertainties associated with counter background and efficiency also contribute. Measurements of the integrated standing crop of excess "OPb were made on core FRl 1B and on a nearby upland-soil core. The upland core was taken from the center of a small (60-m diam.) "island" of till in the middle of the salt marsh. The integrated, standing crop in the upland-soil core was taken as the sum of excess 210Pbin each of three subsamples of soil plus the root activity in the surface horizon. The results are itemized in Table IV. All of the excess 210Pbpresent was in the upper 10 cm of soil, about equally divided between the 0-5-cm and the 5-10-cm layers. Over the interval 10-40 cm there was no measurable excess. Although the root fraction contained the greatest specific activity (dpdgm) of 210Pb and 226Ra,it contributed a relatively small fraction (13%) of the standing crop of excess 210Pbwithin that layer. The integrated standing crop in core FRl 1B was calculated from "OPb activity in contiguous 2-cm-thick slices:

where i increments from 0; n is 16, the number of slices containing excess activity; p is the density (gm ash/cm3);A is the total 210Pbactivity ( d p d grn ash); z is the depth (cm); and 0.83 dpm/gm ash is the supported 210Pb activity assumed throughout the core.

SEDIMENT AND TRACE METALS IN A SALT MARSH

183

2.4. Trace-Metals Analysis

Results of trace-metal analyses of cores FR11B and FR5A are presented as Tables I1 and 111, respectively. Samples of ash material (500°C) were treated by exhaustive leach in hot 6 N HCl, with the weight of ash taken for analysis ranging from 1 to 0.1 gm, depending on the anticipated abundance of the elements. Typically, the sample was leached with 50-100 ml of hot, partially refluxing acid in a covered Pyrex beaker, for at least 12 hr, until the residue was bleached to a light color. Afterward, the slurry was evaporated to near dryness and the soluble material was taken up in 0.1 N HCl and passed through filters pre-washed in HCl. The filtrate was added directly to a volumetric flask and diluted with repeated washings of the insoluble residue with 0.1 N HCl. Solutions were analyzed by standard methods (Perkin-Elmer, 1964,1968)using a Perkin-Elmer Model 303 Atomic Absorption Spectrophotometer equipped with a strip-chart recorder. Filters containing insoluble residue were transferred to Teflon beakers and totally dissolved by treatment with concentrated HN03, HF, and HClO, in the proportion 15: 10:2. The mixture was refluxed for at least 24 hr, then allowed to evaporate until dense white fumes evolved. The nearly dry material was finally taken up in hot 6 N HCl, filtered, and diluted to the mark. The small amount of residue sometimes found despite this treatment consisted of two components: a light-colored crystalline material, identified by x-ray techniques as K,NaAlF,, and a dark-colored, graphitic material. Neither residue was processed further. To detect systematic error in the peat analyses, vegetation and rock reference samples of known composition were analyzed (McCaffrey, 1977). National Bureau of Standards Standard Reference Material 1971 (orchard leaves) was dried at 85"C, then wet ashed using HN03-HC104 and also dry ashed at 500°C. Both the orchard leaves and the United States Geological Survey (USGS) silicate standards G-2 and BCR-1 were also treated by the usual 6 N HCl leach, as well as by total dissolution. The results of these analyses show that both wet- and dry-ashing techniques yielded accurate, reproducible results for organic matter; thus metals associated with the organic portion of the peat are probably quantitatively leached by the procedures employed. The results for the standard silicates, if these materials are assumed to represent the inorganic matter deposited in the marsh, indicate that not all of the trace metals are efficiently leached by present methods. Acid leach of granitic silicate (G2) was accurate for Fe, Mn, Cu, and Zn, but underestimated Pb by about 50%. Total dissolution results were satisfactory for all metals except Pb,

TABLE11. TRACE-METAL,RADIOCHEMICAL, AND BULK-DENSITY DATAFOR FARM RIVERSALT-MARSHCORE FRllB (41"16'00"N; 72"51'06"W)"

00

A

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26

0-2.3 2.3-4.7 4.7-6.8 6.8-8.6 8.6-10.5 10.5-12.3 12.3-14.3 14.3-16.2 16.2-18.1 18.1-20.0 20.0-21.9 21.9-23.8 23.8-25.9 25.9-27.9 27.9-29.9 29.9-32.0 32.0-33.9 33.9-36.0 36.0-37.9 37.9-39.9 39.9-41.9 41.9-44.1 44.1-46.4 46.4-48.4 48.4-50.3 50.3-52.4

0.299 0.300 0.249 0.243 0.218 0.183 0. I86 0.21 1 0. I94 0.226 0.191 0.203 0.224 0.239 0.282 0.277 0.269 0.263 0.261 0.267 0.238 0.177 0.175 0.185 0.216 0.232

0.217 0.219 0.166 0.160 0. I50 0.118 0.135 0.161 0.150 0.181 0.137 0.146 0.171 0.184 0.225 0.208 0.208 0.203 0.195 0.205 0.177 0.107 0.136 0.127 0.153 0.161

14.38 f 0.10 13.88 f 0.11 12.40 f 0.11 9.16 f 0.07 10.58 f 0.06 8.85 2 0.07 9.24 k 0.06 7.01 k 0.05 5.00 t 0.06 2.78 2 0.05 4.89 f 0.06 4.04 2 0.09 3.01 f 0.07 2.25 f 0.07 1.79 f 0.10 1.31 f 0.08 0.99 f 0.08 0.82 f 0.08 0.75 f 0.07 0.67 f 0.06 0.79 f 0.07 0.99 f 0.08 0.71 t 0.07 0.85 f 0.06 1.07 k 0.06 0.59 f 0.05

1.o

-

0.23 0.27

-

0.27

-

0.20

-

0.29

-

0.22

0.11 0.19 -

0.24

-

-

45.4 42.1 43.3 44.4 40.9 44.0 39.1 36.6 34.6 35.7 39.4 41.2 38.1 36.4 33.8 32.7

-

-

-

8% 453 374 345 337 345 323 301 303 310 300 315 322 349 324 329 322 312 268 246 237 229 185 261 245 274

124 130 126 147 137 149 128 %

102 73 56

44 33 20 15 15 13 11 9 12 13 15 13 13 10 12

213 222 226 192 233 234 195 193 191 130 165 162

99 68 71 55 52 63 54 46 49 41 33 50 58 53

79 148 101 66 132 146 217 50 189 67 81 62 50 33 28 12 22 18 17 17 20 22 17 17 14

-

L

00 VI

27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50

52.4-54.4 54.4-56.6 56.6-58.4 58.4-60.2 60.2-62.2 62.2-64.2 64.2-66.3 66.3-68.3 68.3-70.2 70.2-72.3 72.3-74.3 74.3-76.4 76.4-78.3 78.3-80.4 80.4-82.5 82.5-84.6 84.6-86.4 86.4-88.7 88.7-90.7 90.7-92.5 92.5-94.3 94.3-%.4 %.4-98.3 98.3-100.0

0.156 0.155 0.141 0.167 0.126 0.101 0.110

0. I14 0.107 0.138 0.157 0.155 0.164 0. I54 0.184 0.159 0.137 0.146

0.148 0.168 0.165 0.183 0.200 0.227

0.119 0.095 0.085 0.104 0.072 0.052 0.062 0.063 0.061 0.080 0.099 0.095 0.102 0.095 0.126 0.105 0.082 0.089 0.098 0.116 0.112 0.129 0.138 0.167

0.62 ? 0.07 0.38 t 0.05 0.75 0.07 0.48 t 0.08 0.16 2 0.06 0.28 f 0.09 0.47 t 0.07

*

-

-

-

-

-

245 200 187 182 158 I20 119 139 130 169 204 209 21 1 260 274 290 224 206 147 224 218 253 303 257

I1 10 6 10 9 10

8

13 9 12 12 15 7 16 17 17 16 17 15 13 14 18 23 13

41 36 38 33 29 19 20 42 31 36 37 42 47 51 61 60 54 48 40 55 49 55 55 55

17 8 i1

26 13 8 23 18 9 17 16 28 5 14 23 23

-

12 15 8 10 27 5

a Depths are corrected for saw kerf. Densities refer to bulk or whole sediment. Concentrations and activities are based on a hot 6 N HCI leach of ash material. All weights refer to 500°C ash material except for dry density (third column), which refers to 110°C dry material. The wet:dry:ash weight ratios for the 100-cm length of core are 7.49: 1.44: 1,respectively. Precision (la)of sample activity, based on repeated counts during ingrowth, is derived from a propagation of errors analysis as discussed in the text.

186

RICHARD J. McCAFFREY AND JOHN THOMSON

TABLE111. LEACHABLE METALCONCENTRATIONS, COREFR5A" Slice 1

2 3 4 5

6 7 8 9 10 I1 12 13 14 15 16 17 18 19 21 22 23 24 30 35 40 45

Sample depth (cm) 0-2.0 2.0-4.0 4.0-6.1 6.1-8.1 8.1-10.2 10.2-12.2 12.2- 14.3 14.3-16.2 16.2- 18.3 18.3-20.2 20.2-22.2 22.2-24.3 24.3-26.2 26.2-28.3 28.3-30.4 30.4-32.1 32.1-34.1 34.1-36.1 36.1-38.0 40.0-41.9 41.9-43.9 43.9-46.0 46.0-48 .O 58.1-60.3 67.9-69.8 77.8-79.7 87.7-89.9

Fe Mn (mg gm-'1 (pg gm-') 52.9 44.9 45.1 40.4 51.0 72.1 63.9 38.6 41.3 39.6 54.2 49.2

I200 497 394 340 34 1 378 355 322 33 1 330 329 329

34.5 35.1 28.7 29.9 29.1 12.9

338 355 285 222 232 180

-

_.

16.8 11.6 17.2 34.5 37.6

-

-

256 168 143 272 193

cu

Zn (pg gm-7

114 125 135 144 230 192 150 126 118 78 84 32 39

I68 188 153 154 166 395 186 I44 182 165 163 153 136 83 83 66 59 54 58 48

(w g m - 7

-

13 10 6 7 6 8 8 6 6 9 8 7

-

44 29 46 60 29 61

-

Pb

(w gm-') 152 156 143 I42 I70 249 145 91 114 81 84 69 133

-

59 40 29 26 24 19 19 14 14 19 16 12 -

Approximately 2-gm samples of 500°C ash were leached for 6 hours in hot 6N HCI with occasional stining and allowed to evaporate to near dryness before taking up in 0.1 N HCI. Concentrations are reported on an ash-weight basis. Sample depths were corrected for saw kerf.

which was overestimated by about 37%. The basalt standard (BCR-1) was generally less efficiently leached. Total dissolution gave more accurate results, but tended toward overestimationof trace metals, again, especially Pb. Leaching and total dissolution experiments carried out on ash samples from the upper (metal-rich)and lower (presumably preindustrial) sections of core FRllB show that all metals in the upper portion (except Mn) are efficiently leached (r87%), but at depth the leaching of Zn, Pb, and perhaps Cu appears to be only 70-80% efficient. To estimate the excess concentration C,, of Cu, Zn, and Pb in the metalrich upper third of the core, relative to the estimated background values

SEDIMENT AND TRACE METALS IN A SALT MARSH

187

at depth, the average metal concentration c b between 30 and 50 cm was subtracted from the total concentration C at any depth in the core; (2.2)

cxs

=

c-cb

Excess metal concentrations calculated in this way are presented in Table V. With the exception of FRSA, cores destined for pore water analysis were squeezed at room temperature with a minimum of delay (normally within 24 hours) and minimum exposure to the atmosphere. Since core tops are normally exposed to the atmosphere, only subsurface peat is likely to be sensitive to oxidation. To minimize this possibility, a core of larger cross section than necessary was transported in a plastic bag to the laboratory for slicing and squeezing. Just before squeezing, an outer rind of a few centimeters was trimmed away, and each slice was placed in a sealed plastic bag and excess air was expelled. Pore water was obtained by squeezing in a press (Manheim, 1966) while still confined in the plastic bag. Initial eMuent was used to flush the apparatus; later portions were collected in a plastic syringe, filtered (Millipore HA 0.45 pm) directly into a plastic vial, then analyzed immediately or acidified and stored cold. Aliquots of the filtered pore water were taken for analysis of pH, chloride, and sulfate. Chloride was determined by titration with AgNO, to a K,CrO, end point; sulfate was determined gravimetrically by precipitation with BaC1, (Kolthoff and Sandell, 1956). Trace metals were determined by flame atomic absorption spectrophotometry.

TABLEIV. STANDING CROP OF EXCESS"OPb

Depth 210Pb (cm) Soil fraction (dpmlgm ash) 0-5

5-10 10-40

Roots, Sieved (<1 mm)

IN

FARMRIVERUPLAND SOIL'

Standing Sample Organic crop excess 226Ra weight matter '"Pb (dpm/gm ash) (gm dry) (% dry wt.) (dpdcm')

53.4 f 0.3 7.9 & 0.1

2.13 f 0.09 0.81 & 0.02

36 611

2.2 f 0.1 0.8 0.1

0.81 f 0.04 1.02 f 0.02

1345 8240

82.6 14.5

1.5 17.0

Subtotal: 4.0 2.0 Total

18.5 7.7 0 26.2

a Soil and root samples were combusted at 475°C then leached in hot 6 N HCI. 226Ra activity is obtained by measurement of the activity of daughter 222Rn,which is assumed to be in secular equilibrium. 'IoPb flux = Total standing cropmean life = 0.82 dpm cmW2 yr-'.

188

RICHARD J. MCCAFFREY AND JOHN THOMSON

TABLEv. HISTORICALINCREASE IN

FLUXOF TRACEMETALS TO A CONNECTICUT SALTMARSH

THE

~

Slice 1

2 3 4 5

6 7 8 9 10 11 12 13 14 I5 16

Depth z (cm) 0-2.3 2.3-4.7 4.7-6.8 6.8-8.8 8.8-10.5 10.5-12.3 12.3-16.2 16.2-18.1 18.1-20.0 20.0-21.9 21.9-23.9 23.9-25.9 25.9-27.9 27.9-29.9 29.9-32.0 32.0-33.9

Excess flux M f a (pg cm-2 yr-')

Excess conc. C,, (pg gm-' ash)

Cu

Zn

Pb

112

163 172 176 142 183 184 145 143 141 80 1 I5 112 49

61 130 83 48 114 128 199 32 171 49 63 44 32

18

15 10

118

114 135 125 137 I16 84 90 61 44 32 21 8 3 3

21 5

-

cu 8.0 6.9 5.6 7.9 5.2 5.7 3.8 2.8 3.5 4.0

f

Zn 2.0

f 1.5 f 1.7 f 3.2 f 1.6 f 2.4 f 1.0 f 0.8 f

1.3

f 0.8

1.1 2 0.3

0.5 0.8 0.2 0.0 0.0

f f f f f

0.2 0.1 0.0 0.0 0.0

11.7 f 2.7 10.0 f 2.1 8.6 f 2.5 8.3 f 3.1 7.7 f 2.3 7.7 f 3.0 4.7 f 1.2 4.8 f 1.3 5.5 f 1.9 5.3 f 1.8 3.0 f 1.0 1.7 f 0.7 1.9 f 0.3 0.4 f 0.0 0.3 f 0.0 0.0 f 0.0

Pb 4 f1 8f2 4 f1 3f1 5 f 1 5 f2

6 I 7 3 2

f2 f0 f3

-c 1 f I

1 f 0.3 1 f 0.2

0 f 0.0 0 f 0.0 0 f 0.0

Excess stable metal concentrations are total concentrations less natural background, which is taken as the mean concentration ( f la) within the 30-50-cm-depth interval and is 12 f 2,50 f 9, and 18 f 3 pg gm-' ash for Cu, Zn, and Pb, respectively. Precision of the excess concentrations is taken to be f 10%.

Analyses of trace metal and sulfate in pore water provide evidence of diagenetic change in salt-marsh sediment. Rapidly processed cores from the Indian Neck and Farm River sites showed normalized SO,/Cl ratios of greater than 1 at certain depths (Table VI). The cores also contained measurable concentrations of dissolved Mn and, at the Farm River site, Fe. The high Mn concentrations seen in Figs. 9 and 10 coincide with the maximum SO.,/Cl ratio. Other metals were not detected, with the possible exception of trace amounts of Zn in one Indian Neck core. Very high sulfate concentrations and low pH values spontaneously developed in cores stored in air for several weeks (Table VII). Sulfate concentration more than double that of seawater and pH values below 3 were measured at certain depths. Trace metals were also found at very high concentrations at the same depths, but only in cores exposed to the atmosphere (Table VIII). This result appears to be analogous to the phenomenon of acid-mine drainage, where acidic, sulfate-rich waters develop in response to oxidation of pyrite (Stumm and Morgan, 1970). The si-

189

SEDIMENT AND TRACE METALS IN A SALT MARSH

multaneous appearance of high-metal and high-sulfate concentrations when these reduced sediments are exposed to oxygen is most simply explained as oxidation of metals present in the sediment in the form of sulfides. 3. DISCUSSION 3.1. "OPb Dating of the Deposit and Evaluation

The flux of '"Pb from the atmosphere to the surface and its subsequent decay during burial can, under suitable conditions, be used to establish an age-depth relationship in sediments. *'OPb has been used to determine accumulation rates of snowfields (Goldberg, 1963; Crozaz et al., 1964), glaciers (Windom, 1969), lake sediments (Krishnaswamiet al., 1971; Bru-

Depth (cm) 0-2 2-4 4-6 6-8 8-10 10-12 12-14 14-16 16-18 18-20 20-22 22-24 24-26 26-28 28-30 32-34 34-36 46-48 66-68 84-86 88-90

FR5Ab

FR13

IN5

IN6

FR14C

FR14D

-

0.795 0.772 0.734 0.728 0.782 0.767 0.752 0.757 0.749

1.48 1.22

2.43 2.16 1.41 1.01 0.89 0.84 0.89 0.86 0.82

-

0.87 0.69

4.14 4.02 4.84 4.70 4.29 3.12 2.75 2.56 2.54

-

2.37 1.99

-

1.52 1.09 0.96 1.20 1.21

-

0.668 0.701 0.688 0.681 0.609

-

-

0.96 0.85

0.81 0.80 0.74 0.70 0.67

-

-

-

-

-

-

-

0.90 0.78 0.76 0.72 0.68 0.56 0.70 0.66 0.68 0.67 0.66

-

1.11 1.34 1.22 0.80 0.71 0.76 0.77 0.77 0.69

-

0.74

-

0.67

-

(SOJCl) designates the molal concentration ratio of sulfate to chloride. In seawater this ratio is 0.0516. Core FR5A stored several weeks in air, all others processed rapidly. 'I

190

RICHARD J. McCAFFREY AND JOHN THOMSON CONCENTRATION (pg / ml)

I

40'

2

6

4

1 A-0

FR14C

.FIG.9. Variation in dissolved manganese and iron with depth in cores FR14C and FR14D. Core FR14D was taken from a S. patens tussock and core FR14C was taken from an adjacent mud surface some 10-cm lower in elevation. DISSOLVED

0

20 I

20

0.5

Mn (pg /ml) I .o

I

-

FIG. 10. Variation in dissolved manganese with depth in cores IN5 and IN6. Core IN6 was taken from a S. patens plateau and IN5 from the lower adjacent surface.

TABLEVII. PORE WATER ANALYSISOF STORED-CORE FRSA Depth (cm)

c1-

(mM)

(M)

PH

-

6.67 3.98 3.24 3.00 2.90

1.8 2.6 4.4 7.7 6.0

2.1 1.4 8.2 16.8 22.1

2.85 2.92 3.55

7.1 6.5 5.5 4.7 3.7

62.6 30.1 16.3 11.3 1.4

0.315 0.319

3.71 3.83 4.01 4.40

3.5 4.0 4.2 1.9 0.5

2.7 6.4 3.1 0.6 0.2

-

4.75

0.5 0.4 0.3 0.2 0.2


0-2 2-4 4-6 6-8 8-10

54.0 53.5 66.0

0.345 0.272 0.253 0.258 0.264

10-12 12-14 14-16 16-18 18-20

64.7 60.4 47.0 42.2 38.9

0.267 0.273 0.292 0.297 0.294

20-22 22-24 24-26 26-28 28-30

40.0 40.2 38.6 32.8 29.2

0.305

30-32 32-34 34-36 36-38 38-40

-

40-42 42-44 44-46 46-48 48-50 50-52 52-54 54-56 56-58 58-60 60-62 62-64 64-66 66-68 68-70 70-12 12-74 74-76 76-78 78-80 80-82 82-84 84-86 86-88 88-90 90-91.5

Trace metals (CLglml)

SO:-

-

-

-

-

26.6

0.340

-

20.2

-

19.4

-

0.359

-

-

-

-

19.3

0.386

-

20.9

-

-

4.00

-

-

5.80 6.53 6.61 6.90

-

6.93

6.80

-

7.19

-

Fe

Mn

0.2 0.3 0.2 0.1 0.2 0.2 0.2 0.3 0.4 a.3

6.90 7.17 6.80

0.4 0.7 0.8 0.3 0.3

7.10 7.06 6.63 6.08 5.59

0.3 0.3 0.5 1.o 1.5

0.407

5.77 6.00

-

5.19

-

-

25.4

0.408

-

1.4 1.4 1.4 0.8 0.7

-

0.441

5.98

1.3

25.2

-

-

-

cu

Zn

-





-

-


-

-


-

Pb

192

RICHARD J. McCAFFREY AND JOHN THOMSON

TABLEVIII. CHANGES IN CONCENTRATION OF TRACEMETALSAND pH IN PORE WATER OF CORESSTORED IN AIRVERSUS THOSE PROCESSED RAPIDLY WITHOUT STORAGE Trace metals (pg/ml) Core

Treatment

Fe

Mn

cu

Zn

Pb

PH

IN4

Stored" Not stored Not stored Not stored

200

8.9 0.3 0.3-0.8' 0.2-0.6'

14.9 nd' nd nd

66

6 nd nd nd

2.5 6.8 6.5-5* 6.5-5'

IN5 IN6

-

-

tP nd nd

" Stored at ca. 20°C for 7 weeks after squeezing.

'Range among subsamples. ' None detected. Trace.

land et al., 1975; Robbins and Edginton, 1975; Brugam, 1976), and coastal marine sediments (Koide et al., 1972, 1973; Krishnaswami et al., 1973; Bruland et al., 1974; Thomson et al., 1975; Amentano and Woodwell, 1975; Goldberg et al., 1977). In many of these deposits, bioturbation. physical disruption, or compaction may occur. Autocompaction of certain deep-peat deposits has been reported (Kaye and Barghoorn, 1964). Bioturbation of sedimentary deposits, due to the activities of living animals or plants, can be a source of disturbance in either upland (Lutz, 1940; Lyford, 1963) or marine (Rhoads, 1967) environments, and appropriate care must be taken in interpreting result s. If the rate of accumulation of sediment and excess *I0Pbhave both been constant, as is often assumed, and the deposit is undisturbed, then the accumulation rate S can be estimated from the best-fit straight-line equation describing radioactive decay; (3.1)

In A, = lnAo - (A/S)z

where A. and A, are activities of excess '"Pb at depths 0 and z , respectively, h is the decay constant, and S is the rate of accumulation of sediment (cdyr). The age t or time elapsed since a given horizon z occupied the surface is then simply z/S. By inspection of Fig. 11 it is evident that the data for the salt marsh do not exhibit such a linear decrease with depth. On the other hand, if the vertical growth of the salt marsh is indeed keeping pace with a long-term, varying rate of sea-level rise, then a simple linear decrease is not expected, rather, the age-depth relationship implicit in the observed distribution of excess 210Pbshould agree with independent historical records of sea-level rise. To test whether or not the salt marsh has accreted at a rate comparable

193

SEDIMENT AND TRACE METALS IN A SALT MARSH

to the rate of rise of relative sea level requires a different approach, one that does not demand accretion rate to have been constant. The observed, complex distribution of excess 'OPb activity may be used to develop a chronology in a manner similar to that used by Sackett (1965) for dating ocean sediments. Although this method allows the accretion rate to vary, it does assume that the flux of excess "'Pb to the surface has been constant and that there, has been no significant redistribution after deposition. In an accumulating deposit, the rate of storage of excess activity beneath a unit surface area, dQldt, is equal to the rate of deposition on the surface minus loss due to decay: dQldt = F - AQ (3.2) where Q is the standing crop (dpm cm-'), F is the flux to the surface (dpm cm-' yr-I), A is the decay constant (yr-I), and t is time (yr). If F is constant and the core contains the entire standing crop of excess activity, then a steady state exists, i.e., dQldt = 0, and the flux to the

- --

+

+

+

+

I

n

a

-0 (u

* k

LI-

0 4:

0.1 I

0

I

10

20

1

30

40

50

60

70

D E P T H (cm)

FIG.11. The depth distribution of total activity of "OPb in contiguous slices of peat from the Farm River salt marsh shows that the semilog plot is convex in the region about 34 cm. Below 34 cm to the top of the distinct D. spicara rhizome layer at 50 cm, the activity is low and approximately constant, averaging 0.83 dpm gm-', which is taken as supported background. Below this layer, the supported "OPb activity is irregularly lower. The dominant species of vegetation varies with depth as indicated, but all species found are characteristic of high marsh.

194

RICHARD J. MCCAFFREY AND JOHN THOMSON

surface may be readily calculated from the equation F = AQ

(3.3)

As long as the steady state persists, the total standing crop beneath the surface at any time is the constant Qo. Continued accumulation will bury the surface to some depth z in time t. If "OPb is conserved except for decay, then the steady-state standing crop of excess activity Qo will have decayed at a rate governed by the familiar decay law: (3.4)

Qi

= Qo

exp( - Ati)

where Qi is the standing crop remaining below depth zi,the bottom of the I'h slice. When data is available in small depth increments, the following

TABLE Ix. 210PbCHRONOLOGY OF THE FARM RIVER SALT MARSH'

Slice Surface 1 2 3 4 5

6 7 8 9 10 11 12 13 14 15

16

Standing crop of excess '"Pb Depth z Q 2 u ( d p d (cm) cm2) 0 2.3 4.7 6.8 8.8 10.5 12.3 14.3 16.2 18.1 20.0 21.9 23.9 25.9 27.9 29.9 32.0

* *

34.5 1.0 27.7 0.9 21.0 & 0.8 16.8 f 0.6 14.4 ? 0.6 11.7 f 0.5 10.0 f 0.4 7.7 5 0.3 5.8 f 0.3 4.6 f 0.2 3.9 f 0.2 2.9 f 0.2 2.0 f 0.1 1.2 f 0.1 0.7 f 0.1 0.3 f 0.1 0.1 f 0.0

Age u (yr)

t f

0 (1972) 7.0 rt 1.4 16.0 rt 1.5 23.1 rt 1.6 28.0 rt 1.6 34.8 k 1.6 39.9 1.6 48.2 rt 1.6 57.3 f 1.8 64.6 ? 2.0 69.8 f 1.9 79.8 2 2.2 97.7 2.5 107 f 3 125 f 4 155 f 7 203 -t 16

*

Depth is corrected for saw kerf and has a l o precision of f0.05 cm. Total standing crop Qo of excess 'l0Pb is the activity accumulated beneath the surface, corrected for self-defined background activity of 0.83 dpm gn-'. The flux of "OPb to the surface ( = h e o ) is yr-'. calculated to be 1.07 0.03 dpm

*

195

SEDIMENT AND TRACE METALS IN A SALT MARSH

0

+.. ' k

1

.-.

t

4.

>

-

Smoothed Tide-Gouge Record

-...-..-

Salt-Marsh PbZf0 Record

-

-

0 -

... -* ;.. -. -. .. " .1.

N

t-

?! 20

-% 0 Q

m

'

0

*

0. L-.*

-

FIG. 12. Comparison of salt-marsh and tide-gauge records. The ages of salt-marsn peat calculated from the distribution of excess *'OPb are plotted as horizontal bars of length 2a centered at depth increments corresponding to the bottom of each slice. The record of annual mean sea level, based on the New York City tide gauge, was smoothed to remove most fluctuations having a period of about 5 yr or less (Hicks, 1973) and is shown as a dotted line. Smoothing caused truncation of the first 3 and last 3 yr of records, so the latest datum (1969) is indefinitely located relative to 1972 elevation. For the purpose of comparison, the curve has been located so that extrapolation from the 1969 datum along a line with the same slope as the relatively linear sea-level rise since 1940 (0.31 c d y r , Hicks and Crosby, 1974) intersects the surface at the end of 1972. In actuality, the surface of the salt marsh is elevated about 1 m above mean sea level.

approximation may be used:

(3.5) where n is the number of increments containing excess activity, p is the bulk density of sediment (gm ~ m - ~and ) , A is the specific activity of excess 'I0Pb (dpm gm-I). Solving the decay equation explicitly for ti yields (3.6)

ti = ln(QdQi)/h

Since a new value of Qi is available at the bottom of each slice, an age

196

RICHARD J. MCCAFFREY AND JOHN THOMSON

may be calculated at each depth interval. The standing crop below each slice and the resulting chronology for core FR11B are presented in Table IX and compared to the tide-gauge record in Fig. 12. Strict adherence to the assumption of constant *"Pb flux is somewhat relaxed in practice, depending on the time scale of interest. Because 'I0Pb deposition in precipitation is a stochastic process, the actual flux will vary widely over short periods. These fluctuations are effectively integrated, however, by taking samples representing the accumulation of several years. No long-term measurements of atmospheric "OPb are available, but the lack of a significant secular trend in annual precipitation recorded in New Haven over the last century or so (Kirk, 1939; U.S. Dept. of Commerce, 1972), and the lack of demonstrable fission-produced "OPb (Beasley, 1969; Feeley and Seitz, 1970), make the assumption of a constant 210Pbflux seem reasonable. Benninger's (1976) measurements of monthly total "OPb deposition in New Haven further indicate that the long-term average flux is rapidly approached. Although monthly deposition varied by a factor of 4 without obvious seasonality, after only 1 yr the integrated annual average deposition rate was found to be indistinguishable from long-term averages based on standing crops of soils (Table X). The responses of a salt marsh and tide gauge to changes in sea level must be presumed to differ. A tide gauge responds reversibly with a time constant measured in minutes. Although the salt marsh appears to keep pace with sea level over the years, the fact that it is regularly flooded by tides demonstrates its relative sluggishness. In addition, accretion of the salt marsh is generally irreversible-erosion, when it does occur, is usually confined to the seaward perimeter of exposed coastal marshes (Bloom, 1967). Slow, irreversible accretion appears to be a normal result of peat formation in New England-type salt marshes. The nature of the response of the salt marsh suggests that the annual sea-level data should be further smoothed to make the comparison. This TABLEX. INTEGRATED FLUXOF EXCESS"OPb Site Branford, Connecticut New Haven, Connecticut Cook Forest State Park, Pennsylvania Maryland

Flux (dpm cm-*yr-')

Method

1.07 0.82 1 .o 1 .o

Standing crop, salt marsh Standing crop, forest soil Total precipitation Standing crop

This work This work Benninger, 1976 Lewis, 1976

1.2

Standing crop, soil

Fisenne, 1968

Ref.

SEDIMENT AND TRACE METALS IN A SALT MARSH

197

has been done using the method of Hicks (1972), which reportedly attenuates oscillations of yr by at least 90%, but allows long-period oscillations, which account for most of the variance in tide-gauge records (Walcott, 1975), to survive. The mean sea-level data, smoothing program, and results are given in McCaffrey (1977). It should be noted that the degree of smoothing is arbitrary and does influence the comparison. If annual or shorter period average values had been used, the increased scatter would envelop the salt-marsh curve, whereas greater smoothing would, of course, further dampen the oscillations. Using the smoothed sea-level curve, as plotted in Fig. 12, the agreement between the rate of salt-marsh accretion and tide-gauge record is reasonably good, except in that part of the record where temporary reversals in the sea-level curve appear. If the intermediate degree of smoothing used is accepted, then the discrepancies between the curves may provide insight into the accretion process. Note that the temporary fluctuations in the New York City tide gauge are not explained by changes in Hudson River discharge (U.S.G.S., 1960) and the 1920-1930 hiatus, at least, is also apparent in other East Coast gauges, thus they appear to be real sea-level changes of some years in duration. There are several ways the salt marsh could conceivably have responded to sea-level reversals of this type. Direct tracking through relative minima by removal of accumulated peat can be discounted with confidence on the basis of observation: General loss of up to 10 cm of surface material, even if somehow possible, could hardly have gone unnoticed, yet no such loss has been reported. An independent line of evidence against loss of peat, discussed in a later section, is that there is no deficiency in the expected standing crop of excess 2’oPb.Other conceivable responses, such as dewatering and compression, or simply temporary cessation of organic growth are, perhaps, somewhat more probable. However, two lines of argument indicate that these responses are also unlikely. First, the magnitude of the reversal, 4 or 5 cm, will decrease the fraction of the time the marsh is flooded by only a few percent, according to Fig, 13. This level of change is probably insignificant, given that the “noise level” of the yearly record is of the same order of magnitude. The second, more powerful argument is that 210Pbdeposition will continue, independent of accretion. If so, cessation of surface accretion will result in accumulation of 2’oPband produce a zone of relatively high activity. The ratio of the amount of activity that would have accumulated ( Q ’ ) compared to that actually found (Q) can be calculated. If deposition of 210Pbcontinued unchanged during a 13-yr pause in surface accretion, as

198

RICHARD J. McCAFFREY AND JOHN THOMSON AVERAGE

1.0

'7I

2.0

-

I

0.8 I

I

F R A C T I O N OF T I M E E M E R S E D

0.6 1

I

0.2

0.4 1

1

I

0.0 I

-

-

Y

-

z

-

0

> w

0.0

-

-1.0

-

-2.0

-

-

-

1

I

I

1

I

I

1

I

FIG. 13. The portion of the time a given intertidal position is exposed to the atmosphere depends mainly upon elevation and local tidal regime. This relationship has been determined for Bridgeport, Connecticut, where emersion for an elevation at MHW is well over 0.9.

the material in slice No. 8 accumulated, then the present ratio, under the condition that the actual interval of 2'oPbaccumulation t ( = 9.1 yr) had been extended by an additional t' ( = 13 yr) years, would be (3.7) Q'/Q = [l + (t'/t)lexp [$(At')]= 3 This would have been readily apparent in the 1972 data, but no such feature is found in Fig. 11, thus it is likely that accretion continued at some finite rate during this interval. In summary, the change in relative sea level measured independently by the New York City tide gauge generally overlies the record of saltmarsh accretion derived from "OPb. This agreement is considered the best evidence that the assumptions made in calculating the 210Pbages are valid; in particular, that "OPb has not been grossly redistributed since burial. This implies that elements with chemical properties similar to Pb, such as Cu and Zn, primarily reflect depositional rather than diagenetic processes.

3.2. Physical and Chemical Evidence on Lack of Disturbance and Chemical Immobility In addition to the evidence based on 'OPb, other physical and chemical evidence exist by which to assess the possible importance of compaction, mixing, and chemical diagenesis at the coring site.

SEDIMENT AND TRACE METALS IN A SALT MARSH

199

If substantial compaction occurred during burial, and loss of organic matter is negligible, then an upward-openingconcavity would be expected in the plot of organic-matter accumulation, Fig. 6. In fact, the curve is highly linear, which demonstrates that bulk density of organic matter is practically uniform over the depth of the core, and indicates that no substantial compaction has occurred. Beneath the core, compaction is judged to be even less likely, because no compacted organic layer was present and muddy estuarine deposits occur within the next meter. The lack of measurable compaction is perhaps understandable, given that (1) fresh peat is neutrally buoyant in seawater, (2) salt-marsh peat is an extensive, intricately interwoven, space-filling, resilient, spongelike organic material, and (3) the general tendency of the high marsh to remain saturated with water throughout the tidal cycle, except for a few centimeters at the surface. This buoyant-peat model is consistent with the observations of Harrison (1975), who measured the variation in elevation of the marsh surface, relative to a bedrock outcrop, during the flood half-cycle of the tide. She detected an increase in elevation of several millimeters as the tide rose, which is the response expected for a buoyant material. The limited range of the buoyant response is thought to be due to the facts that (1) the water table is perched near the surface of the marsh, (2) changes in the free-water level in the peat are restricted by the finite permeability of this material, and (3) most of the floodwater reaches the interior surface of the high marsh by overland flow rather than by flow through the porous medium: Holes fill from above, not below. Another feature of salt-marsh peat restricting its buoyant response is the interconnectedness of the bulk material: Peat near the surface is not isolated, but is enmeshed by its roots into a semiinfinitelayer of peat some tens of centimeters thick, The buoyant-peat model predicts that if periodic recharge of this perched-water table was prevented, then substantial dewatering and compression would occur. As it turns out, this experiment has already taken place: Diking of salt marshes was one of the means used in the past to reclaim them for agricultural purposes, and following that treatment the marshes were reported to have shrunk considerably, in some cases as much as 3 feet (Smith, 1907). The study site was selected on the basis of its a priori low probability of disturbance, but the effects of biological mixing need be neither obvious nor trivial. Physical disturbance could occur as a byproduct of the macrofauna or macroflora. For example, W. A. Niering (Connecticut College, personal communication) suspects that cropping of salt-marsh vegetation by rodents may be significant adjacent to upland refuges, but Nixon and Oviatt (1973) felt that the impact of mammals was slight. Marsh snails (Melumpus bidentatus) and the ribbed mussel (Modiolus demissus)

200

RICHARD J. McCAFFREY AND JOHN THOMSON

either do not dig burrows, or inhabit the perimeter of the marsh, away from the core site. The burrowing crab, Uca pugnax, is evidently present in considerable numbers on the marsh and probably represents the single greatest threat to the sedimentary record. This crab reportedly occurs in mean densities of 205/m2in the S. alternifora marshes of Georgia, but in Rhode Island only 2.7 -+ 3.8/m2were counted in early fall (Teal, 1962;Nixon and Oviatt, 1973). In this study area their burrows were obvious under the relatively open stands of S. alternifora, but the dense stands of S. patens grass often prevented observation of the high-marsh surface. To assess the significance of their burrowing activity under the S.patens covering the sample site, a study of the distribution of crab burrows in three zones along a transect crossing the sample site was undertaken in midsummer, when crabs were present in relative abundance. The results of this census, presented in Table XI, show that the number density of burrows along the S. alternifora margin of the channel is high, approximately equal to densities reported for the Georgia S. alterniflora marsh. In a S. patens zone 2-m away, the densities dropped to -4 of the S. alterniflora value. Near the sample site, 15 m inland, one burrow was found in one 0.25-m2plot and none in the other. Evidently, crab-burrowing activity is highly skewed toward the channel margins and is uncommon at the sample site. The substrate under each type of grass differs markedly and may explain this result. Along the perimeter of the marsh where S. alterniflora thrives, the surface between the emergent shoots is soft mud. In contrast, under S. patens there is a dense, tangled root mat, as seen in the x radiograph (Fig. 4), and it seems likely that this substrate inhibits burrowing by the fiddler crab. Whatever reason for this site preference, it appears that bioturbation by burrowing crabs is practically negligible at the interior sample site, The relative freedom from bioturbation enjoyed by the high marsh distinguishes this location from other aerobic marine sediments where mixing may be intense (Aller, 1977). Perhaps the best available evidence to evaluate the impact of displacement of solids by plant growth comes from the experiments of Bloom (1967). In order to establish the rate of accretion of the salt marsh, he spread colored-aluminum flakes over small areas of the surface at repeated intervals over a period of years. These marker horizons were later recovered by taking small cores through the overgrown peat and measuring their depth. The fact that recovery was feasible during at least the next decade and that in almost all cases the original sequence had been preserved during growth of several centimeters of new peat is good evidence of the lack of massive physical disturbance of such layers within the critical upper levels of peat.

SEDIMENT AND TRACE METALS IN A SALT MARSH

20 1

TABLEXI. DENSITYOF Uca pugnax BURROWS ON THE SURFACE OF THE FARMRIVERSALT MARSH Burrow count" N-cU

Zone S. alterniflora (bordering channel) S . patens (2-m inland of site 1)

S,patens (midmarsh sample area)

(m-*)

254 64

2

40

* 20

2 2 3

Data represent the total number of open (not necessarily occupied) burrows present on 7 August 1973 on two f-m2plots within each zone.

Additional evidence for or against diagenesis may be found in depth profiles of the stable elements and other chemical species of interest. The source strengths of the stable elements, unlike "OPb, cannot be hindcast with confidence, even though it is generally conceded that emission and deposition rates for many of them have increased over the last century. Although a quantitative, historical-deposition model is unavailable, we can examine the existing data for internal evidence of diagenesis. All stable metals in both cores FRl 1B (Table 11) and FRSA (Table 111) show increased concentrations toward the surface. The magnitude and pattern of the increase varies among metals, but three types may be distinguished. Cu, Zn, and Pb are all relatively low and uniform in concentration below 30 cm. At about 30 cm, the concentrations of each of these metals simultaneously increases to a high plateau around 10 cm. The magnitude of the increases in concentration, comparing the upper third of the cores to the lower two-thirds, ranges from roughly 5 times for Zn and Pb to 9 times or more for Cu. The profiles of Fe and Mn are distinct from the trace metals and from each other. Mn exhibits no sustained trends until the surface is approached, where its concentration abruptly increases, almost tripling in core FRSA. Unlike Mn, Fe shows no abrupt surface maximum, although its concentration does increase irregularly by a factor of roughly 2 times over the upper third of the core. Not all of the elements selected for study are expected to behave in the same way, based on general chemical properties, when buried in a reducing, sulfide-rich, marine environment. As a first approximation, the expected concentrations of dissolved metals can be estimated from published solubility-product constants, by assuming a free-sulfide concentration and neglecting complex formation, as has been done in Table XII. The calculated equilibrium concentrations are extremely low, except for Mn and possibly Fe, which suggests that Cu, Zn, and Pb should be fixed as their insoluble sulfides, whereas Mn might be chemically mobile.

202

RICHARD J. McCAFFREY AND JOHN THOMSON

TABLEXII. EXPECTED EQUILIBRIUM CONCENTRATIONS OF SELECTED METALSIN SALT-MARSH PORE WATERS, ASSUMING A TOTAL SULFIDE CONCENTRATION OF lo-) mu

Element

Solubility product' log K sp (25°C)

Expected conc.' log [M2+lf

- 12.6 - 17.2 - 35.2 - 24.1 -26.6

-2.6 -7.2 -25.2 - 14.1 - 16.6

Mn Fe

cu Zn Pb

" Total sulfide concentrationCH2S(=[H2Slf + IHS-I,) of about 1 mM has been measured in a Louisiana salt marsh by Brannon (1973), a value not unusual for reducing marine sediments. Total dissolved sulfide-ion concentration [S2-], may then be calculated for pH (7.5) typical of the salt marsh, using the apparent dissociation constant of hydrogen sulfide (log K; = - 6.9, log K; = - 13.6) measured by Goldhaber and Kaplan

(1975):

[S-I,

=

KiEH2S (H+){l + [(H')/K;l + lK;/(H+)Il = 2.5

x 10-9m

Solubility product constants were selected from Bjerrum et al. (1958) compilation. ' Expected metal-ion concentrations were calculated from the expression:

K'P =

yM2+

[M2+]fys2-[S2-]f

where the activity coefficients for divalent ions are assumed equal to 0.2 and the effects of complexing ignored.

There is no guarantee that equilibrium conditions prevail, of course, especially within the biologically active surface zone of marine sediments (Presley et al., 1972), and a number of complicating factors could act to support trace-metal concentrations at levels above or below those expected on the basis of simple solubility considerations, such as the tendency to absorb on peat (Fraser, 1961a,b; Bertine, 1972) or to form soluble complexes. However, there is reason to expect that immobilization via sulfide formation occurs rapidly. Goldhaber and Kaplan (1975) showed that free sulfide exists mostly as HS-, the chemical form that Pohl (1975) demonstrated to be active in precipitation of at least Fe, Zn, and Cd

SEDIMENT AND TRACE METALS IN A SALT MARSH

203

sulfides. Lawrence and McCarty (1965) found that in anaerobic digesters, where conditions are similar to those in the salt marsh, addition of sulfate to the reducing system rapidly eliminated heavy-metal toxicity, presumably by precipitating dissolved metals as their sulfides. That sulfide is in fact produced in the salt marsh and reacts to form metal sulfides comes from several lines of evidence. The well-defined color change-pink over black-found at 5 1 cm under live S. patens is generally a good indication that iron hydroxyoxides are being converted to iron mono- and disulfides (Galliher, 1933; Berner, 1970). The fact that the normalized S0,ICl ratio under field conditions is usually less than unity (Table VI) strongly suggests that sulfate-reducingbacteria are active. Direct evidence of formation of authigenic metal sulfide comes from xray-microprobe and light-microscope examination of this section of the sediment, where pyrite (Fig. 7) is common within the sediment, often in close association with organic fibers (Fig. 8). Although Fe was the only metal identified in the pyrite grains, there is indirect evidence that trace metals exist in the sediment as sulfides. One piece of evidence comes from the results of oxidation of salt-marsh sediment during prolonged storage in air. During this experiment, highsulfate concentrations developed in the upper part of core FRSA, which coincided with the appearance of high Fe and trace-metal concentrations in the pore water. The concentrations are far greater than concentrations found in rapidly processed cores (Table VIII). Because the SOJCl ratio exceeds unity, the appearance of SO:- must be due to oxidation of a reduced form of sulfur. Thus it is likely that iron sulfide(s)and trace metals present as dispersed sulfides are oxidized during exposure to the atmosphere, yielding soluble trace metals and sulfate. We can evaluate the observed distribution of Cu, as an example of an element tending to form highly insoluble sulfide, and Mn, which is often mobilized in reducing sediments. The depth profiles of Cu and Mn are plotted in Figs. 14 and 15, respectively. Two facts are outstanding: (1) There is excellent agreement between the concentration (p,g/gm ash) profiles of the two cores, with the exception of two Cu points near 10 cm. In the case of Mn, the two profiles are practically congruent. Within a given core the Mn concentrations vary over a factor of about 2, but the detailed agreement between the two cores indicates the variation is real and laterally extensive on a scale of at least meters. (2) The depth profiles of Cu and Mn are distinctly different: Mn concentration exhibits no sustained increase until just beneath the surface, whereas the increases in Cu concentration begins at about 30-cm depth. The Cu concentration appears to reach a maximum around 10 crn, then declines somewhat toward the surface.

204

RICHARD J. MCCAFFREY AND JOHN THOMSON

0

0

250

500

750

I

I

I

$9

10 -

20

1000 0

‘

1250 0-

O0

adg

-

1 0

6

k

30-

03 DO

h

5

Y

40

-

I

0 0 00

O D 0 0 0

50-

w P

so-

B

FRllB o FRSA 0

0 0 0

80 -

8o 8

90-

o o

10

O 8

0 0

8” OO

100

O0

The decline in specific concentration of Cu toward the surface may indicate that a complex explanation is necessary, such as a recent relative decline in Cu-deposition rate or relative increase in clay deposition. Because the inorganic-matter bulk density does vary appreciably (Fig. 17), dilution by clay merits further examination. If the relationship between copper content and ash content in core FRllB (Fig. 16) is examined, however, we find that (1) a proportionality (12-pg Cu/gm ash) exists in the “deep peat,” which is within the range expected for soils, and (2) higher in the core, above the lower clay band, copper content becomes independent of ash content. This suggests, first, that a new Cu source is contributing and, second, that the trend in concentration might be better expressed on a bulk-density basis (pg/cm’), rather than as specific gravity

SEDIMENT AND TRACE METALS IN A SALT MARSH

205

(wg/gm). It is important to note here that if the inorganic matter in the upper part of the core still has the same concentration as it does in the deep peat, then only about 10% of the total Cu is attributable to this source, and dilution by clay cannot account for the decline in the specific concentration of Cu toward the surface in Fig. 15. Indeed, when the Cu data are replotted on a bulk-density basis, as in Fig. 18, the subsurface maximum is eliminated, at least for the dated core. Concentrations expressed in this way increase monotonically toward the surface, within the precision of the measurements. The evidence indicates that there is a new source of Cu, contributing to the salt-marsh sediment, and this source appears to be increasing in strength relative to soil-derived Cu. COPPER ( p g l g r n ash) 0

10

20

30 L5

E

40

0

Y

I

a

50

W P

60

0

D

FRllB FR5A

70

00

90

100

-8 80

-0

0

FIG. 15. Variation in the concentration of leachable Cu with depth in cores F R l l B and FRSA.

206

RICHARD J. McCAFFREY AND JOHN THOMSON

35

UPPER CLAY BAND

30

25 LI

SnALLoW PEAT

m

E

/

' 20 0

4

I

0

I

,''---\ \ * * \\ \

I

'* I

%

Y

K

n IS

I

0,

1

I I

0 0

I

.*

I

I

I

0 :

I

I

I I

I

10

I

\ \ 1

0

'-%',-.,

5

L O W E R C L A Y BAND:

0 0

0.1

/

'I '* -

0.2

A S H (gm/cm3) FIG.16. Scatter diagram showing the relationship of Cu bulk density to inorganic-matter bulk density in various levels of core FRllB.

In contrast to that of Cu, the Mn profile can be adequately explained only by diagenetic remobilization. If eroded-upland soil (typically containing about 800-pg Mdgm ash) is the major source of Mn deposited on the salt-marsh surface, and the Mn profile is a steady-state feature, then it is clear that partial remobilization indeed must occur as the surface layer is buried. Profiles of dissolved Mn are shown in Figs. 9 and 10. In all cases Mn concentration in the pore water is highest in the uppermost sample or reaches a maximum at shallow depth. The samples exhibiting a shallow maximum were those taken from beneath grassy tussocks; the samples having a surface maximum were taken from adjacent areas of relatively low elevation. The evidence for remobilizatio'n of Mn beneath the surface of the salt marsh is qualitatively in accord with reports of

SEDIMENT AND TRACE METALS IN A SALT MARSH

10

-

20

-

BAND

- - : 30

LOWER CLAY BAND

E

0

207

40

I

50-

W

6070

80 90

-

100

FIG. 17. Vertical distribution of inorganic matter in core FRllB. Samples with a density greater than 0.175 (dotted lines) appear as gray-colored“clay bands” due to their relatively high silt-clay content.

0

10

I-

40t =

EXCESS Cu FR l l B

FIG.18. Depth distribution of excess Cu in core FRl lB, where Cu concentration(pg Cu(pg gm-’) x p (gm cm-’) for each slice of the core.

208

RICHARD J. MCCAFFREY AND JOHN THOMSON

postdepositional migration of Mn in coastal marine sediments (Thomson et al., 1975; Graham et al., 1976), lake sediments (Robbins and Callender, 1975), and saturated soils (Ponnamperuma, 1972). A subsurface maximum in pore water concentration has also been reported beneath the land-water interface of Chesapeake Bay (Holdren et al., 1975; Sanders, 1978) and Long Island Sound (Aller, 1977). In an accumulating body of reducing sediment, the occurrence of a maximum in dissolved Mn beneath a surface layer that contains very high Mn concentrations in the solid phase strongly suggests a diagenetic model involving reductive dissolution of Mn. Furthermore, the fact that the dissolved Mn concentration diminishes strongly below the maximum indicates that a removal process, as yet unidentified, is active in the deeper peat. The relatively low leachable Mn concentrations at depth (Fig. 14) are much lower than expected for soils (Wright et al., 1955), which indicates that some of the Mn escapes the sediment and reenters the estuary. Since Mn is fixed below the zone of dissolution, then both the fraction of the deposited Mn that escapes and the rate of escape may be estimated. If the Mn concentration associated with incoming soil particles is taken to be within the range reported for similar gray-brown podzolic soils (800 2 200 pg/gm), then the concentration below 30 cm in Fig. 14 may be used to estimate the fraction lost: 1 - fraction retained = I - conc. at 30 cm/leach. eff. incoming soil conc. = I - (330/0.76) & 10% pg/gm 800 & 200 pglgm

fraction lost (3.8)

=

= 0.46 & 0.14 or about half

The average rate of loss may be calculated from the average deposition rate times the fraction lost, where average deposition rate is estimated from soil concentration and total mass of inorganic matter accumulated within the length of 2'0Pb-datedcore. Figure 6 shows that 5.5 gm cm-* of inorganic matter, estimated to contain 800 f 200 pg Mn/gm accumulated in about 203 2 16 yr, so the input is 5.5 gm cme2 x 800 2 200 pg/gm or 22 2 6 pgcm-*yr-' (3.9) 200 2 16yr and the loss rate is (3.10)

22 t 6

X

0.46

2

0.14

or

10 t 4 pgcm-2yr-'

SEDIMENT AND TRACE METALS IN A SALT MARSH

209

Not all of the soluble Mn escapes, however. A portion evidently reprecipitates, presumably as insoluble oxyhydroxides, as dissolved Mn diffuses to the surface, resulting in transient storage of Mn and high total concentrations in the surface layer. The rate of loss of Mn from the salt marsh is lower by 1-2 orders of magnitude than measured fluxes from the offshore marine sediments of nearby Long Island Sound (Aller, 1977) and Narragansett Bay, Rhode Island (McCaffrey et al., 1980). If the Branford salt marsh, one of the New England types, is generally representative of other marshes in this area, then their contribution of Mn to the estuary-at-large is probably negligible compared to subtidal sediments. This result is more a consequence of the paucity of inorganic-matter deposition rather than the efficiency of remobilization. Some notion of the actual remobilization process may be obtained by comparing the dissolved Mn concentration profile to the SO,/Cl ratios in Table VI. The SOdCl ratio in core FR13 is everywhere less than unity. The nature of the profile indicates that sulfate reduction is very active near the surface, but that sulfate reduction deeper in the core is less intense. In the other cores (FR14C and D, IN5 and 6) sulfate reduction ultimately proceeds to roughly the same degree; however, the fact that the SOdCl ratio is greater than unity at certain depths indicates that oxidation processes are also active. It is possible that transport of molecular 0, via uninterrupted gas spaces in S . alterniflora might act as an oxygen source in its root zone (Teal and Kanwisher, 1966). However, this has not been shown to be true for S . patens, and the coincident appearance of SO:- and dissolved Mn (normally rapidly removed from oxygenated waters) is difficult to reconcile with simple oxidation with molecular oxygen, at least as an equilibrium condition. Morgan (1967) has pointed out a possible alternative, spontaneous oxidation process, namely the reduction of the higher oxides of Mn in the presence of dissolved sulfide yielding dissolved Mn and sulfate: (3.11) 8MnOOH + Sz- + 16H+ = 8Mn2+ + SO:- + 12H20 Some form of manganese oxide is available near the sediment surface and, as it is buried, it is exposed to an environment where bacterial-sulfate reduction is generally active. If manganese oxides do act as oxidants for sulfide, then it is possible to explain, at least qualitatively, the simultaneous appearance of dissolved Mn and SO:- in the pore water. There is sufficient solid Mn present to produce the amount of dissolved Mn found, but this reaction alone cannot begin to account for the amount of SO:generated. Also, the fact that the normalized SOdCl exceeds unity means that some form of previously accumulated sulfide is oxidized. No trace

210

RICHARD J. MCCAFFREY AND JOHN THOMSON

metals other than Mn and Fe were detected in these samples, but the stored-core experiment does show that under sufficiently extreme oxidizing conditions, trace metals can be mobilized. No single, spontaneous oxidation process seems adequate to explain the observations. It is conceivable that the observations refer to transient conditions in the salt marsh rather than equilibrium conditions, and further study is needed to understand them. 3.3. Atmospheric Fluxes Recorded in the Salt Marsh

The special circumstances of "OPb deposition allow the hypothesis of quantitative retention of 'loPb to be tested. This may be done by comparing 210Pbdeposition in the salt marsh to independent estimates of local deposition. Although there is continual deposition of 210Pbto the salt-marsh surface, the quantity of excess *l0Pbthat accumulates beneath a unit surface area (the standing crop) is finite due to radioactive decay. As '"Pb deposition proceeds, the standing crop in a closed system will increase only until its decay rate equals the depositional flux of 210Pb;thereafter a steady state exists. At steady-state, the '"Pb flux F may be calculated from the standing crop Q and the mean life T of "OPb: (3.12)

F = Q/T = AQ

if the flux of 2'oPbfrom the atmosphere is quantitatively retained by the salt marsh, then the flux calculated in this way should agree with independent estimates of the atmospheric 2'oPbflux. Such independent estimates are available from direct measurements of "OPb in total precipitation and from the standing crops present in undisturbed soils. Previous measurements of the atmospheric flux have been made in New Haven by Benninger (1976) using open-bucket collectors sampled at monthly intervals. Soil-standing crops have been measured at various sites in the eastern U.S.,including a forested upland site within the Farm River salt marsh, and are compared in Table X. It is evident that the 210Pb flux derived from salt-marsh core FR11B is indistinguishable from the current rate of deposition of 210Pbfrom the atmosphere in nearby New Haven. The salt-marsh flux is also in substantial agreement with estimates from soil-standing crops, but may be greater than the flux estimated from the Branford forested-soil sample by about 20%. Circumstances suggest that the standing crop measured in nearby soil may tend to underestimate the atmospheric 210Pbflux. The site is on a small island covered with deciduous forest surrounded by an extensive area of salt marsh and, therefore,

SEDIMENT AND TRACE METALS IN A SALT MARSH

21 1

must be presumed to act as a net source of organic matter to its surroundings. Since vegetation intercepts '"Pb deposited by rainfall (Hill, 1960; Francis et al., 1968), leaves and other tree surfaces will contain excess "OPb. Net loss of such material from the island would depress the residual standing crop below that expected on a unit-area basis. Second, the ages of the largest trees in the canopy, based on counts of annual rings, are not more than 60 years (McCaffrey, 1977), suggesting that the forest cover may have been removed within the past century by man or natural causes. An additional loss mechanism may be drainage of soil leachate containing excess 210Pbfrom the upper part of the profile into the groundwater. Overall, the exchange processes between the island and its environment all seem directed toward net loss of 210Pb,so a systematic deficit in the island standing crop is to be expected relative to semiinfinite forest sites or the salt marsh. It is important to understand the relationship between intertidal position and atmospheric exposure. Due to the periodic nature of the tidal cycle, a given elevation within the tidal range may be frequently submerged by tides, but the average fraction of the time a given position is emersed depends strongly upon its elevation relative to mean-sea level. In nearby Guilford, Connecticut, where tide-gauge data are available, the surface of the high marsh was found to lie between mean-high water (MHW) and MHW neaps (R. Gordon, Yale University, personal communication), and the Farm River high marsh is believed to be similarly elevated. The relationship between elevation and exposure has been worked out for Bridgeport, Connecticut (P. Marshall, Yale University, personal communication) and is reproduced as Fig. 13. This curve may be used to estimate the average exposure time for other stations experiencing the semidiurnal tide characteristic of Long Island Sound. It is evident that exposure to direct atmospheric input increases rapidly from slightly more than half-time for a surface positioned at mean-sea level to more than 0.9 of the time for surfaces located at MHW. If this tidal regime is applied to the Farm River salt marsh, then it is clear that the high marsh is exposed to direct atmospheric input most of the time. In summary, deposition from the atmosphere almost always impinges directly upon the high-marsh surface. Moreover, the standing crop of excess 210Pbnow found beneath the salt-marsh surface or nearby soil is in essential agreement with that expected from measured atmospheric fluxes. Therefore, the salt-marsh standing crop of excess 'IoPb seems best explained as quantitative retention of atmospheric deposition. If an atmospheric source of 210Pbaccounts within experimental error for the salt-marsh deposition, then the 210Pbcontribution from all other sources must be small to nonexistent. If we take the simple difference in

212

RICHARD J. McCAFFREY AND JOHN THOMSON

the fluxes calculated for the salt marsh and New Haven sites, 0.07 dpm cm-2 yr-', as a probable upper limit for inorganic-matter sources, this severely constrains them. Since the present rate of accumulation of inorganic matter, judged from the mass found in the surface slice of known age, is (3.13)

0.217 gm/cm3 x 2.32 cm/7.07 yr

=

0.071 gm cm -* yr -'

then the probable maximum specific activity of inorganic material transported to the marsh via the water column is: 0.07 dpm cm-' yr-' or <"1dpm 2'oPbxs/gmash. This estimate is compared to the specific activity of likely sources (upland and marine) in Table XIII. It appears that neither material scoured from the mud-water interface of Long Island Sound nor material eroded from the upland surface is an allowable source, since only material from beneath these surfaces is of sufficiently low activity. It is relatively difficult to effect deep (18-20 cm) erosion of submarine sediment in Long Island Sound, but common for upland soil to erode deeply, especially when land is disturbed by agriculture or construction. This indicates that the source of low-specific-activity inorganic matter is the eroding subsurface material of the watershed. An upland-soil source is also consistent with the metal concentrations found at lower levels in the salt-marsh cores. The iron content is generally proportional to the mass of inorganic sediment and is found in concentrations (almost 4%) considered reasonable for soil (Bowen, 1966). In an analogous plot for Cu,Fig, 16, the concentrations in deep peat and many of the lower clay-band samples are also typical of soil. This explanation cannot account for the increasingly high trace-metal concentrations found nearer the surface. Above 30 cm, the concentrations of Cu and other trace metals simultaneously begin a sustained increase

TABLEXIII. EXCESS"'Pb ACTIVITY IN POTENTIAL DETRITAL SOURCEMATERIAL

Source

Excess 210Pb (dpdgm ash)

Reference

Long Island Sound Sediment Surface 18-20 cm Upland soil, Branford, Connecticut 0-5 cm 5-10 cm 10-40 cm

4.6 1.1

Thomson et al., 1975 Thomson et al., 1975

5.4 1.5

This work This work This work

0.0

213

SEDIMENT AND TRACE METALS IN A SALT MARSH

E X C E S S F L U X , k g c m 2 yr-l 0 1972

4

8

1

4

2

8

12

0

4

t

1912

+-

8

12

-

t+ Pb

1812 1832

1792

FIG.19. Increased rate of deposition of trace metals from the atmosphere as recorded in a Connecticut salt marsh.

(Fig. 15) and become independent of inorganic matter in peat above the lower clay band (Fig. 16). It is possible to calculate the trace-metal flux to the salt-marsh surface using the sediment chronology based on 210Pb,the excess concentrations of metals within each slice C,, (pgdgm ash), and the bulk density p (gm ash/cm3),from the equation (3.14)

M

= pC,,z/I

where z is the slice thickness (cm), t is the age (yr), and M is the metal flux (pgm cm-* yr-I), averaged over each slice. The flux of trace metals over the historical period are listed in Table V. The first indication of an increase in the flux of each of the trace metals begins simultaneously in Fig. 19. This corresponds to the final stage of deposition of the lower clay band, at about the time of the U.S.Civil War (1860-1865), when industrialization in this area greatly intensified. The rate of increase since then has been generally sustained to the present. The increase is nearly linear in the case of Cu and Zn, but deviates considerably from a smooth rise in the case of Pb, and there is no obvious acceleration due to the introduction of leaded gasoline in the 1920s. The trace-metal flux at the time the core was raised (1972) may be calculated by extrapolating the excess flux versus time curves to zero depth. This is a short, linear extrapolation in the case of Cu and Zn and may be carried out with confidence. Because of the large changes in flux with time for Pb, however, the extrapolation is less precise. Deposition from the atmosphere and water column both contribute to

214

RICHARD J. McCAFFREY AND JOHN THOMSON

the total flux of material accumulating at the surface of the salt marsh. Their relative contributions will depend strongly on the material in question. The atmospheric contribution of inorganic mass is likely to be small relative to deposition from the water column. The total inorganic-matter flux, calculated from the known bulk density, age, and depth of the surface sample, is about 0.07 gm cm-2 yr-', whereas annual dust fall measured near the sample site at Tweed-New Haven Airport (DeLouise, 1968) was found to be about 0.004 gm cm-' yr-', Dust fall at an airport is likely to be an overestimate of actual values in the relatively dust-free salt marsh, so at least 94% of the inorganic mass is deposited from the water column. Since the mass of inorganic matter deposited from the atmosphere can account for only a small fraction of the total inorganic-matter accumulation rate, and surface Long Island Sound sediments are excluded on the basis of their high-*'OPbspecific activity, then the principal source is almost certainly the eroding soil of the watershed, carried to the surface of the salt marsh while suspended in runoff. If this is true, then the flux of trace metals associated with the soil may be estimated from the rate of accumulation of inorganic matter and its trace-metal concentration. Although data for soils of the Farm River watershed are not available, the trace-metal content for typical brown-podzolic forest soils may be used as an approximation. Data used in the model calculations are listed in Table XIV. Multiplying the mass flux by the soil concentrations yields the estimated trace-metal fluxes from the water column (Table XV). The results of the model calculations show that the atmospheric contribution to the Fe and Mn flux is negligible compared to deposition from the water column. This is in accord with the estimated atmospheric deposition of Fe to the area of the New York Bight of about 37 pgm cm-2 yr-' (Duce et al., 1976). In the case of the trace elements Cu, Zn, and Pb, contribution from the water column is generally small relative to the atmosphere. The atmospheric contribution to the salt marsh may be estimated in two ways: (1) by attempting to place the study site appropriately within the broad picture of available deposition data at various other sites; and (2) by making model calculations using local atmosphericdata and, lacking local information, deposition-rate parameters determined elsewhere. Using the first approach, the trace-metal fluxes calculated for the salt marsh are compared to measured deposition rates for other sites in Table XVI. These sites were selected to represent circumstances that would be expected to yield fluxes of comparable magnitude, as well as establish reasonable limits to be expected. The station on Hawaii receives deposition from air masses that have presumably been scavenged over the ocean of most of their pollution burden, whereas the large fluxes measured

215

SEDIMENT AND TRACE METALS IN A SALT MARSH

TABLEXIV. SOILCONCENTRATION C, DRY-DEPOSITION VELOCITIES V, , WASHOUT CONCENTRATIONS x USEDIN MODELCALCULATIONS RATIOSW , AND AEROSOL Data

Fe

C (pdgm ashy

V , (cdsec)b Wd

x (wdm3Y

38,000 1.1 250 0.54

Mn 666

cu

Zn

Pb

14

66 0.6 180 0.33

19 0.3 76 1.0

0.5

0.7'

370 0.01 1

140 0.16

~

Soil data from Wright et a/. (1955), except for Fe (Bowen, 1966). Cawse (1974), except where noted otherwise. The deposition velocities used above are somewhat lower than values measured within New York City (Kneip and Eisenbud, 1974) by factors of 2-4, but may be a better approximation to the regional values expected at the study site, where the largest particles have already been partially removed from the aerosol. Estimated from an average MMD (mass median diameter) for Cu of 1.35 pm, based on the work of Lee et al. (1968) and read from the curve of V, versus MMD of Cawse (1974). Dimensionless ratio, (pdkg rain)/(p.glkgair), fromGatz (1975). Probably reliable within a factor of 2. Mean values for Tweed-New Haven Airport (1969-1973; Connecticut DEP 1975). Uncertainty is estimated to be f50%.

in New York City and Philadelphia represent deposition in major urbanindustrial centers on the East Coast. The total fluxes at Wraymires, U.K., a site about 18 km from significant industrial sources, may be representative of semirural deposition in temperate regions. Measurements at Nantucket and Woods Hole, Massachusetts, and the New York Bight, all nonindustrialized East-Coast sites, may indicate depositional fluxes to be expected for the eastern seaboard, away from intense sources. Although the reliability of these estimates varies and is difficult to assess, it is probably not better than &50% and could easily vary by factors of 2 or 3 or more, depending upon experimental technique, length of the observation period, and other factors. Taken as a whole, however, the various TABLExv. RESULTSOF MODELCALCULATlONS OF THE CONTRIBUTION OF TRACE TO THE SAMPLE SITE METALSFROM THE WATER COLUMN AND ATMOSPHERE COMPARED TO TOTALFLUXESTO THE SALT MARSH" Proximate source ~~

Atmosphere Dry deposition Wet deposition Atmospheric subtotal: Water column, suspended sediment Model, total Salt marsh. total ~

Fe

Mn

cu

Zn

Pb

3 2

7 6 13 5 18 17

10 7 17 1 18 -9

~

__

pg cm-z yr-'.

20

0.2

10 30 2700 2730 3600

0.4 0.6 47 48 <78

5 1 6 10

TABLEXVI. CALCULATED EXCESSMETALFLUXTO THE SURFACE OF THE FARM RIVER SALT MARSHCOMPARED TO MEASURED RATESAT SELECTEDSITES' ATMOSPHERIC DEPOSITION Year

Site Branford, Connecticut Wraymires, U.K.

1972 197% 1972

New York Bight New York Cityb

May 1972 1970 1972-73 1%9- 1970 1977- I979

Collection data Dry Total Total Wet Dry Total Wet

Dry Philadelphia, Pennsylvania' Nantucket, Massachusetts' Woods Hole, Massachusettsd Hawaii'

1966-1%7 1966-1%7 1977-1979 1966-1%7

Tod Wet Wet Wet Dry Total! Wet

cu

Zn

8?2 0.07 3.2

12 3 0.61 12 8.1 21 36 32 4

-

5.3 26 9.8

-

4.9 5.6

1.1

"

5

12 23 7.6 4

4 6

0.6

kg cm-* yr-'.

* Environmental Measurements Laboratory,

U.S. Department of Energy. Weather Bureau Airport Station. Filtered rainwater. Redfield Laboratory. Weather Bureau Research Station, Mauna Loa Observatory, Hawaii. Filtered rainwater. Independent measurement by open bucket.

'

Pb 7"4 <0.6 5.5 6.0 12 23 35 7 0.5 8 13 8.5 2 2 3 0.4

Ref. This work Peirson et al., 1973 Peirson et al., 1973 Duce et al., I976 Volchok and Bogen, 1971 Kneip and Eisenbud, 1974 Volchok and Bogen, 1971 Feeley and Larsen, 1979 Feeley and Larsen, 1979 Feeley and Larsen, 1979 Lazrus et al., 1970 Lazrus et al., 1970 Feeley and Larsen, 1979 Feeley and Larsen, 1979 Feeley and Larsen, 1979 Lazrus et al., 1970

SEDIMENT AND TRACE METALS IN A SALT MARSH

217

estimates are generally self-consistent and probably are the best means available to assess the deposition expected at the study site. It is evident from Table XVI that the rates of deposition at the study site generally fall within the range of atmospheric deposition rates represented on the high side by New York and Philadelphia, and on the low side by Hawaii. In fact, the Branford results are in good agreement with intermediate sites taken to represent total deposition along the northeastern U.S. coast. This widespread consistency suggests that the salt marsh is collecting deposition more or less representative of the region, rather than some local source. The good agreement with Wraymires, U.K., presumably reflects the similar composition and behavior of regional aerosols originating in other temperate, industrialized areas. In the second approach, the trace-metal flux from the atmosphere may be divided into two components: wet and dry deposition. The sum of these two components of the atmospheric flux will be calculated, using local atmospheric concentration and precipitation data, estimates of dry deposition velocity as defined by Chamberlain (1960) and reported by Cawse (1974), and washout ratios from Gatz (197% all listed in Table

xv

*

The concentrations x of the trace metals were measured quarterly in the ambient air over Connecticut at as many as 50 stations during 1969- 1973 by personnel of the Department of Environmental Protection (DEP) and the Department of Health (Connecticut DEP., 1975). The values used in the present calculation are arithmetic means of measurements made on samples taken on the roof of the Tweed-New Haven Airport terminal building, which are not significantly different from New Haven (6 stations) or the state as a whole. The wet-deposition contribution, due to the combined effects of rainout and washout, was calculated as Wxhlp, where W is the washout ratio, i.e., the concentration in rainkoncentration in air; p is the density of air (1200 gm/m3);and h is the precipitation rate (mean 1965-1972 = 112 cm3 cm-2 yr-'; Bridgeport, 1972). The flux of trace metals associated with dry deposition is taken as V,x, where V , is the dry deposition velocity. The results of the model calculations of dry and wet deposition to the total atmospheric flux is shown in Table XV. The results clearly show that the atmospheric contribution to the Fe and Mn flux is negligible compared to deposition from the water column. For the trace elements, however, atmospheric deposition dominates. The sum of wet and dry atmospheric deposition expected for Cu, Zn, and Pb falls within a factor of 2 of the surface fluxes found in the salt marsh. The relative contribution of wet and dry deposition is in accord with the measurements of Volchok and Bogen (1971) for New York City, but disagrees

218

RICHARD J. McCAFFREY AND JOHN THOMSON

with the Wraymire, U.K., data (Peirson et al., 1973), where dry deposition accounted for a relatively small fraction of the total. The generally good agreement may be fortuitous, and the local application of V , and W values determined elsewhere is questionable. The lack of consistent results for the same elements at different stations (Peirson et al., 1973; Kneip and Eisenbud, 1974) may be due to the differences in collection techniques as well as in the aerosols. Therefore, the agreement (within a factor of 2) between the model predictions and the calculations based on the salt-marsh results should be considered provisional for the trace metals Cu, Zn, and Pb. Although the atmosphere behaves as a local source of Cu, Zn, and Pb, it is undoubtedly carrying material from more remote sites. In general, evidence that trace metals in polluted air are being transported considerable distances comes from the decline in concentrations away from source areas, their accumulation in remote sediments, and from massbalance calculations. Industrialized urban atmospheres generally contain higher concentrations of many trace metals than suburban areas (Lee et al., 1968; Stern et al., 1973), and lowest concentrations are characteristic of remote land (Rahn, 1971; Zoller et al., 1974; Struempler, 1975) and ocean sites (Hoffman et al., 1972; Duce et al., 1975). Fuel combustion, incineration, various industrial processes, and automobiles have been identified as major contributors to the pollution burden (Winchester and Nifong, 1971; Lee and von Lehmden, 1973; Kneip and Eisenbud, 1974). Such sources have been implicated as major contributors via the atmosphere to Lake Michigan (Winchester and Nifong, 1971; Edgington and Robbins, 1976). In the Los Angeles Basin about a third of the lead emitted by automobiles is carried at least 150-km downwind before being deposited (Huntzicker et al., 1975). Various sedimentary deposits, such as coastal southern California (Bruland et al., 1974), the Baltic Sea (Erlenkeuser et al., 1974), Long Island Sound (Thomson et al., 1975), and Narragansett Bay (Goldberg et d., 1977) also show increased concentrations in recent deposits. Local atmospheric data may be examined for clues to the relative importance of remote and local sources. The long-term average atmospheric concentrations of Cu, Zn, and Pb at certain stations along the north shore of Long Island Sound (Fig. 20) show no apparent trend along the W-E transect from New York City (Morrow and Brief, 1971; Kneip and Eisenbud, 1974) along the shore to eastern Connecticut (Connecticut DEP., 1975). Measured concentrations in the New York Bight area, however, under conditions where aerosols are being mixed and advected out across the Atlantic, show a general decline away from the coast (Duce et al., 1976). Model calculations indicated that roughly 6% of the Pb and 24%

SEDIMENT AND TRACE METALS IN A SALT MARSH

219

r o ‘ DISTANCE (km) 100

2.0 w A

yr)

E , 1.6

P g -

Y

1.2

I-

U

m

5

0.8

W

0

z

2 0.4 0.0 I-

3

2 Bz z

8

i

0 3

z

W W 0

c3

FIG.20. Atmospheric concentrations of Cu,Zn,and Pb along the Connecticut coast.

of the Zn will be deposited within 100 km of the coast, even if there is no rain. It may also be significant that the Nantucket station of Lazrus et af. (1970) was not measurably less polluted than stations in the northeastern U.S. Taken at face value, this suggests either that the sample sites are immersed in a regional atmosphere of generally similar trace-metal concentration, or that each station resides within an “island” of pollution of comparable concentration. It would be interesting to compare the historical record of atmospheric quality found in the salt marsh to independent records of atmospheric burdens of trace metals, but such information is not available. It is possible, however, to use metal-production figures as a crude index of atmospheric emission. As an example, the historical increase in the primary production of Cu in the U.S. as a whole is compared to the increase in Cu deposition recorded in the Farm River salt marsh in Fig. 21. It is evident that the increase in Cu flux recorded in the salt marsh is in general agreement with the production trend over the period of record.

220

RICHARD J. MCCAFFREY AND JOHN THOMSON

EXC888 Cu Flux

U S . Prlmory Prodn.

El m

s: o.3t 0:

LL

o’21 0.1

-

0.01 L I I I r n k 1980 1960 1940 1920 1900 1880 1860 1840

YEAR

FIG.21. Historical increase in the flux of excess Cu to the salt marsh compared to the trend in U.S.primary copper production over the same interval. Both trends are normalized by dividing the rate at any time by the 1970 rate. In the case of the salt marsh, the flux is averaged over several years accumulation, while the primary-production figures represent the rate for a single year, arbitrarily taken at the beginning of every decade from 1850-1970 (U.S.Dept. Interior, 1973).

Contributionsfrom local sources are presumably important, but aerosol residence-time considerations and the general flow of air masses across the continent suggest that there is a very extensive areal contribution to the atmospheric burden of trace metals at the study site. Mean-residence time in the atmosphere depends on particle size and climatic regime. Esmen and Corn (1971) found that the mean-residence time of submicronsize particles in urban air in the absence of wet precipitation was roughly 102-103hr and decreased to 10-100 hr for 1-10 pm particles. Moore et al. (1973) recently reevaluated the mean-residence time of the tropospheric aerosol and concluded that 4-6 days is the most probable value.

SEDIMENT AND TRACE METALS IN A SALT MARSH

22 1

Turekian et al. (1977), using a different worldwide model, arrived at the same value for reasonable mean-transit times of global zonal winds. Rodhe and Grandell (1972) estimated the expected life of aerosol particles in the lower troposphere subjected to precipitation scavenging and found residence times on the order of 100-300 hr in summer and 35-80 hr in winter (Stockholm rainy season). Even when the effects of wet deposition are included, a mean-residence time on the order of days is still likely. Based on a mean-residence time of days, the general eastward sweep of air across the continent (Van Cleef, 1908) and analyses of back trajectories of air masses (Cogbill and Likens, 1974) indicate that virtually any source east of the Mississippi or in southern Canada must be considered a possible contributor to the record of trace metals uncovered in the Farm River salt marsh. 3.4. Salt-Marsh Accretion

There is little direct evidence of the actual mechanism(s) of salt-marsh accretion, but the two processes generally offered in explanation are silt deposition and organic productivity. Silt deposition (more generally, particle deposition from the water column) has received emphasis in the past, depending on the observer and his area of study. Ganong (1903), who observed the marshes along the Bay of Fundy, felt that vegetative growth was not involved in accretion, but that salt marshes were “built up in a subsiding basin out of organic red mud brought in from the sea by the rush of the tides. . . .” His viewpoint represents one extreme, observers elsewhere came to recognize the importance of organic-matter production (i.e., peat formation) in salt-marsh accretion. One of the earliest and most succinct statements of the peat-formation hypothesis of salt-marsh accretion was made by Mudge (1862), after he examined a section of the Romney Marsh, near Lynn, Massachusetts. He said: The surface of the marsh is nearly a dead level, about one foot above ordinary high water mark and only overflowed by the higher Spring tides. The appearance of the marsh soil indicates a gradual formation from the grasses, aided by the fine, rich sediment which the high tides occasionally deposit. The saline grasses grow only above ordinary high water mark, and as the roots in the lowest part of the soil, even eight or more feet below the surface, are in their natural position, showing no distortion, we must conclude that their sirus was above the high water line, and that subsidence has been so gradual that the growth of the plants has never been interrupted.

Mudge’s concept of salt-marsh accretion clearly incorporates the notion that the grasses of the high marsh maintain their position during a gradual change in relative sea level, by upward growth. However, the special

222

RICHARD J. MCCAFFREY AND JOHN THOMSON

mechanism he invoked as a driving force (undercutting of the peat by “a current of water in the diluvium under the clay”) curtailed general application of his ideas. When Mudge made his report, glacial theory had not yet been fully accepted (Flint, 1971), and it was not until somewhat later that the real driving force was recognized to be the general rise in sea level in postglacial time (Davis, 1910). The description enunciated by Mudge contains the tracking hypothesis in embryonic form, but other than mentioning growth of grass, he is vague about the mechanism(s) involved. Shaler’s (1885) conception of salt-marsh accretion, in contrast to Mudge’s, was primarily derived from his observations of the seaward perimeter of salt marshes; he hypothesized that formation of peat occurred during advance of salt-tolerant grass over silted-up mud flats, According to Shaler’s theory, S. alternifora is sufficiently tolerant to thrive at such a low intertidal level, and since other species cannot withstand such prolonged immersion in marine waters, it is the colonizing grass of aggrading mud flats. As S. alternifora colonizes a mud flat, suspended sediment is more efficiently trapped and, once deposited, it is protected from erosion. Colonization thus promotes relatively rapid peat formation, to a level commensurate with local tidal range. The overall phenomenon, Shaler believed, was one aspect of the slow redistribution of sediment on a “steadfast” coast, long after the changes in sea level associated with deglaciation had ended. Accumulations of sediment built up in this way should show a thin surface layer of high-marsh peat overlying muddy S. alterniforu peat, itself overlying mud flat deposits. These two hypotheses of salt-marsh formation were later reconciled by Johnson (1967, originally published in 1925). He recognized that each was based on observations of related but different aspects of salt-marsh formation-and were not necessarily conflicting. Stratigraphic evidence linking the two models has since been discovered in Granite Bay, Branford, Connecticut, where almost a meter of S. patens peat was found overlying S. ultern$ora, itself overlying estuarine mud (Knight, 1934). Redfield (1972) has convincingly documented the overall process at Barnstable, Massachusetts: After S. alternifloru colonizes the mud flats, these stands of grass promote sediment accumulation and rapidly build to a level where high-marsh grasses can become established. Continued rise in relative sea level promotes development of a thick section of highmarsh peat. Eventually, the expanding, thickening wedge’of high-marsh peat overrides the upland. As a result, sections through the landward edge reveal a stratigraphy consistent with the Mudge-Davis model, whereas sections through the seaward edge are consistent with the Shaler model. The tracking hypothesis (Redfield and Rubin, 1962; Bloom and Stuiver,

SEDIMENT AND TRACE METALS IN A SALT MARSH

223

1963), that the vertical growth of the high marsh is keeping pace with the rise in relative sea level, implies that (1) salt-marsh grasses are able to track the long-term change in relative mean sea level, even in the presence of relatively high-amplitude, short-period tidal fluctuations, and (2) that there is some factor related to sea level that is controlling marsh accretion in such a fashion that, ideally, it maintains but does not exceed a given elevational offset. The idea that salt-marsh accretion is a function of the supply of silt is probably correct for Fundy-type salt marshes and during the Shaler stage of the development of the New England type, but this idea seems inadequate to account for accretion in the Mudge-Davis zone of peat formation. The historical emphasis on silt supply is perhaps an outgrowth of the crucial role of sand or silt in early marsh buildup (Shaler, 1885; Ganong, 1903; Chapman, 1960) and of the high inorganic-matter content of many marshes found along the shores of the North Atlantic and coastal plains states (Johnson, 1967). Many deposits referred to as salt marshes do, in fact, consist largely of inorganic-particulate detritus, with only incidental plant remains (Dawson, 1855; Bouma, 1963, 1969; Redfield, 1972). The botannical studies of Richards (1934) and the x radiographs of Bouma (1963) show that inorganic-particle deposition can indeed lead vertical development where supplies are adequate. On the other hand, salt marshes are found in areas lacking an abundant silt supply. Johnson (1967) recognized that the New England-type salt marsh contained a relative lack of inorganic matter and relative excess of organic remains compared to the Fundy type, and used this property as a basis of classification. New England-type salt marshes occur from the mid-Maine coast to northern New Jersey. This work deals with the high-marsh zone of one of the New Englandtype salt marshes and confirms the relative paucity of inorganic matter. Viewed at low tide, the high marsh stands exposed as an elevated volume of apparently firm, but actually highly unconsolidated material, containing over 80% of water by weight and only 0.2 gm cmW3of inorganic matter. Such a low bulk density means that individual grains of inorganic matter are well-dispersed and by themselves are clearly inadequate to support an accreting high marsh. Another mechanism is needed to explain it, presumably related to high-marsh peat formation and the bulk properties of the deposit. The salt marsh is composed largely of plant remains and more or less silt and clay-sized inorganic matter. The detailed structure of the subsurface organic matter is aptly described by Davis (1910) as:

. . . an extensive system of slender, tough, underground stems, from which arise

224

RICHARD J. MCCAFFREY AND JOHN THOMSON

long, fibrous, branching roots; together, these form a turfy structure, that is a mass of intermingled fine fibers, usually whitish or grayish, from intermingled silt . . . it differs from the turf formed by sedges in the persistence with which the underground stems retain their form and individuality, instead of collapsing and flattening . . . (and) in the presence of white or light-coloredfinely branching roots, which penetrate the mass in every direction and make up the great bulk of the material.

It is important to recognize that the material forming the peat is not the emergent portion of the grasses, rather, it is the organic matter grown in place at or just below the surface. Traditional studies of salt-marsh-grass productivity have focused on only the accessible above-ground shoots (Keefe, 1972), but there is reason to believe below-ground productivity is substantial (Valiela and Teal, 1974), perhaps equaling or exceeding above-ground productivity. For example, Broome et al. (1972) showed that more of the net growth of S. alterniflora seedlings is stored belowground than above. Similarly, Amentano and Woodwell (1975) report that 20-40% of the total net production may be stored below ground each year in an established Long Island salt marsh. The plant parts formed below ground, in contrast to aerial parts, are well preserved in the resulting saline peat (Barghoorn, 1949). Even in 4000-5000-yr-old peat, epidermal root structures remained largely intact even though centralcore tissues were often degraded. The bulk properties of the high-marsh peat seem to be largely a consequence of this organic growth. The x radiographs (Figs. 4 and 5) reveal that a major portion of the subsurface space is occupied by or enclosed within the plant structures. This may account for the apparent low permeability of this material, as manifest in the failure of core holes to fill with water from below, despite the generally water-saturated condition of the deposit. Even at low tide the relatively low permeability of the peat in concert with tidal flooding is apparently sufficient to maintain the watersaturated condition within a few centimeters of the high-marsh surface. The uniform distribution of organic matter seen in Fig. 6 indicates that compaction is negligible and that vertical growth is somehow controlled, as if either an upper limit or some sufficient condition had been attained. This combination of circumstances seems to suggest that the primary mechanism of accretion is the formation of neutrally buoyant peat within a perched body of water. If true, this model would account for a number of observations: lack of autocompaction, small, reversible increases in elevation of the surface during tidal flooding, and long-term tracking of the rise in relative sea level. One can speculate on how tracking might be accomplished: Normal changes in relative sea level (at tidal frequency) serve to establish the quasi-steady-state exposure at about the mean high-water level for these

SEDIMENT AND TRACE METALS IN A SALT MARSH

225

moderately salt-tolerant plants. A long-term upward trend in relative sea level, as is occuring in this area, would increasingly stress the plants by increasing immersion; thus new growth would have obvious survival value. Direct evidence of the conditions causing the stress and the growth response of the plants has yet to be developed. The need to respond to a long-term trend in the midst of tidal fluctuations suggests that a highly damped signal, such as the mean height of the perched watertable, might be involved. Whatever the detailed mechanism, the response of the plants appears to be new growth both below and above ground. A rather extreme example of this is shown in the x radiograph of a large tussock in Fig. 5 . The result is to move upward along the emersion curve (Fig. 13) and decrease stress. It also may be more than coincidence that the surface elevation of the high marsh is closely located at that singular point in the emersion curve where further increases in elevation have a rapidly diminishing effect on emersion time, but any lesser elevation would rapidly increase emersion. The overall effect of this growth would be to maintain the plant species within their tolerance limits and to track the rise in relative sea level. If the inorganic matter is incorporated into the peat in proportion to its availability in suspension, then the salt marsh should be a recorder of sediment-generating processes in its vicinity. The inorganic matter occupying the interstices of the peat originates either directly from the land or from offshore sites. Meade (1972) argues that very little of the sediment load of rivers actually escapes to the deepocean basin, but that most of the sediment is temporarily deposited offshore then returned, via landward bottom drift associated with estuarine circulation, to accumulate in estuaries and coastal marshlands. Evidence that offshore marine sediments are supplying silt and clay to certain coastal marshes along the Connecticut shore has been presented (Ellis, 1962; Hill and Sherin, 1970). However, as the standing crop of excess Z'oPbin the Farm River marsh is almost entirely accounted for by atmospheric deposition, the source of silt and clay now depositing on this salt marsh seems not to be the relatively high specific-activity material present at either the marine or upland surface, but a low-specific-activity material, probably from the subsurface of the watershed. The general features of the distribution of clay with depth are shown in Fig. 17. Moving from the oldest peat toward the present, we see that the earliest clay accumulation is beyond the reach of reliable 210Pbdating, but if accumulation is reasonably assumed to have averaged 0.1-0.2 em yr-', then accelerated deposition must have begun during the 18th century. After a long period of increase, clay content remained high for roughly a century, until the time of the U.S.Civil War. Thereafter, clay

226

RICHARD 1. MCCAFFREY AND JOHN THOMSON

content decreased and remained low until after WWII, when renewed high concentrations began and continued at least to the time of sampling in 1972. These general features of the inorganic-matter distribution are in qualitative agreement with the temporal pattern of sediment-generating processes described by Wolman (1967) and appear to be a record of the variation in erosion taking place on a watershed in response to changing land-use activities. Two-thirds of the naturally forested land of Connecticut was cleared for agriculture between the time of settlement in the 17th century and the Civil War (MacDonald, 1968). This generally corresponds to the interval of high-inorganic-matter content forming the lower clay band at 26-42 cm. The subsequent decrease in inorganic-matter content is probably a result of extensive abandonment of farms following the Civil War. The change in the percentage of land area devoted to farming in New Haven County is shown in Table XVII. Farm acreage decreased from a high of 87% at the time of the Civil War, to 21% by 1950, and presumably further decreased in subsequent years. In New England, farmland spontaneously reverts to forest (Lutz, 1928), thus reestablishing cover and diminishing erosion. The increasing inorganic-matter content within the upper few centimeters of peat corresponds to a period of rapidly rising population and urbanization of the watershed. Land-development activities associated with urbanization promote erosion by increasing the rainfallhnoff ratio (Lull and Reinhart, 1972) and by increasing stream velocity in channels adapted to lesser flows (Hewlett and Nutter, 1969). Erosion of soil from the watershed and its transport to the salt marsh was directly observed in the field. Indeed, salt marshes like the Farm River marsh, which lie adjacent to well-stratified tidal rivers, seem biased to receive soil from upland TABLEXVII. PERCENTAGE OF NEW HAVENCOUNTY LANDAREAIN FARMS" Area

(%I

Year

(%I

1850 1860 1870 I880 1890

73.0 87.3 12.9 71.8 63.7 67.5

1910 1920 1925 1930 1950

64.3 48.8 50.6 38.9 31.2

1900 a

Area

Year

Census of Agriculture (1933); U.S. Census (1950).

SEDIMENT AND TRACE METALS IN A SALT MARSH

227

sources rather than marine sources. This occurs because the runoff flood stage of the Farm River characteristically lasts a matter of days after any substantial rainfall, but the tidal period is only 12 hours, thus the entering salt wedge will repeatedly displace the overlying relatively fresh water above the banks of the tidal channel. As this water and its suspended load overspread the marsh, velocity slackens and suspended matter, strongly biased toward upland sources, is deposited on the surface. 4. SUMMARY AND CONCLUSIONS

Peat from the Farm River salt marsh, an estuarine marsh on the submerging coast of Connecticut near New Haven, was sampled by raising essentially undisturbed, meter-long cores, and was found to record tracemetal deposition from the atmosphere and sediment erosion from the land during the past century. The activity of 2foPband concentrations of Fe, Mn, Cu, Zn, and Pb were measured in contiguous, 2-cm-thick slices of peat. The distribution of excess 210Pbwith depth in the salt marsh, by assuming a constant flux to the surface, is used to estimate the age of sediment at the bottom of each slice. The resulting age-depth relationship was compared to the independent measure of sea-level rise recorded by the New York City tide gauge and was found to be in good general agreement. The standing crop of excess 210Pbactivity beneath the salt marsh is equivalent to a constant deposition rate of 1.07 0.03 dpm cm-2 yr-l. This long-term average flux is the same as the total atmosphericdeposition of excess z'OPbmeasured recently in nearby New Haven and is within 20% of the standing crop of excess '"Pb found in local upland soil. These results are considered the best evidence that "OPb is quantitatively retained within the salt marsh and is not significantly redistributed after deposition. They also support the idea that stable elements with analogous chemical properties also represent a depositional record. Fixation of metals as their insoluble sulfides is clearly evident for Fe and is considered the likely method for fixing the other trace metals, except for Mn. The blackened sediment beneath the thin-surface-oxidized layer, the lower S04/C1ratio than that of marine water, and the presence of pyrite framboids (FeS,) show that sulfate is actively reduced to sulfide in the salt marsh. No discrete metal sulfides, besides pyrite framboids, were detected by microprobe examination; however, the simultaneous appearance of SO:- and trace metals in the interstitial water of cores stored in air suggests that other trace metals may be present as finely dispersed insoluble sulfides.

*

228

RICHARD J. McCAFFREY AND JOHN THOMSON

If Cu, Zn, and Pb are fixed after deposition, then the historical flux of trace metals to the surface may be calculated from the 210Pbchronology and sediment chemical composition. The flux of excess metals calculated in this way increased from about the time of the Civil War to the present high values. The fluxes observed at the study site fall within the range expected, based on reported measurements of deposition at selected sites, and also agree generally with model calculations based on concentrations reported for the local aerosol. The atmosphere is thus the principal source of 210Pband trace metals now being deposited in the salt marsh, and this repository appears to contain a useful record of atmospheric quality over the past century or so. The trace metals in the atmospheric aerosol are no doubt partly of local origin, however, their likely residence time of a few days and the sweep of air masses across the continent requires practically any source in the eastern half of the U.S.to be regarded as a probable contributor. In the case of Fe and Mn, the atmospheric supply is practically negligible, so terrestrial and/or marine sources must account for most of the Fe and Mn content. On the other hand, the standing crop of excess 210Pb in the salt marsh is almost entirely accounted for by atmospheric deposition, so the material arriving via the water column apparently contains little or no additional excess activity. Because the excess activity at the surface of both Long Island Sound sediment and the upland soil is relatively high, only material found at much deeper levels (1-2 decimeters) is sufficiently low in activity to act as source of inorganic matter. It is obvious that disturbed soils on the watershed are eroding and that the particulate matter suspended in runoff is deposited on the surface of the marsh during flood tides; indeed, the well-stratified estuary favors terrestrial over marine sources. For these reasons, the origin of the inorganic matter is considered to be upland soil. The temporal pattern of land use on the Farm River watershed provides a qualitative explanation for the distribution of inorganic matter in the peat. A decline in forest cover due to clearing for agriculture proceeded until about 1865, the end of the U.S.Civil War. This was followed by a period of farm abandonment and intensifying industrial development in neighboring New Haven. Aggravated erosion during the conversion to agriculture was followed by gradual amelioration of erosion during spontaneous reforestation of abandoned farmland. This cycle is believed to have produced the feature recognized as the lower clay band. The recent upper clay band appears to be a consequence of intensive construction activities attendant to development of the watershed as a suburb of New Haven. The vertical development of the high-marsh surface is governed by vegetative growth, not particle deposition. The extensive, interwoven,

SEDIMENT AND TRACE METALS IN A SALT MARSH

229

space-filling, resilient roots and rhizomes of the high-marsh grasses deter burrowing by macrofauna and act as a matrix for retention of available, particulate-inorganic matter. The resultant peat has a varying inorganicmatter concentration, but is sufficiently impermeable so that the freewater table of the interior high marsh is usually found only a few centimeters beneath the surface. A rise in relative sea level will tend to force an offset increase in the mean height of the interstitial water table, which is perched near mean high water. Peat beneath the water table in neutrally buoyant and not measurably compressed. Salt-tolerant grasses in the active-growing zone near the surface are forced by this relative rise in water level to occupy new space, by forming tussocks. The overall result is to track the rise in relative sea level by peat formation. ACKNOWLEDGMENTS We wish to thank Karl Turekian for his essential support and patient encouragement of this work. We have benefitted from the advice and willing help of the faculty, students, and staff of Yale University. RJM wishes to acknowledge gratefully the special help of Thomas Siccama, whose pioneering efforts and enthusiasm helped launch this work; William Niering, for making available the facilities of Connecticut College Arboretum; Job Potter for assistance in identification of peat specimens; and Gregg Currier, who was always ready to lend a needed hand, low tide or high. Financial support was provided by the Yale University Fellowship Program, the Yale School of Forestry and Environmental Studies, and an Energy Research and Development Administration Grant E( 1 1-1)3573 to Karl K. Turekian, Principal Investigator.

REFERENCES Adams, D. A. (1%3). Factors influencing vascular plant zonation in North Carolina salt marshes. Ecology 44, 445-456. Aller, R. C. (1977). The influence of macrobenthos on chemical diagenesis of marine sediments. Ph.D. Thesis, Yale University, New Haven, Connecticut. Amentano, T. V., and Woodwell, G. M. (1975). Sedimentation rates in a Long Island marsh determined by *"Pb dating. Limnol. Oceanogr. 20,452-456. Baas Becking, L. G . M., and Moore, D. (1961). Biogenic sulfides. Econ. Geol. 56,259-272. Barghoorn, E. S. (1949). Degradation of plant remains in organic sediments. Harvard Univ. Bot. Mus.Leaflets 14, 1-32. Barghoorn, E. S . (1953). Recent changes in sea level along the New England coast. Science 117, 597-598. Beasley, T. M. (1969). Lead-210 production by nuclear devices, 1946-1958. Nature (London) 224, 573. Benninger, L. K. (1976). The uranium-series radionuclides as tracers of geochemical processes in Long Island Sound. Ph.D. Thesis, Yale University, New Haven, Connecticut. Berner, R. A. (1970). Sedimentary pyrite formation. Am. J . Sci. 268, 1-23. Bertine, K. K. (1972). The deposition of molybdenum in anoxic waters. Mar. Chern. 1, 43-53.

230

RICHARD J . McCAFFREY AND JOHN THOMSON

Bevington, P. R. (1969). “Data Reduction and Error Analysis for the Physical Sciences.” McGraw-Hill, New York. Bjerrum, J., Schwarzenbach, G., and Silltn, L. G. (1958). Stability constants of metal-ion complexes, with solubility products of inorganic substances. Part 11. Inorganic ligands. Chem. SOC.(London) Spec. Pub. No. 7. Bloom, A. L. (1961). Evidence of submergence from Connecticut tidal marshes. Geol. SOC. Am. Spec. Paper 68, 138 (Abstr.). Bloom, A. L. (1967). Coastal geomorphology of Connecticut. Final Report, Office of Naval Research, Task No. 388-065. Bloom, A. L., and Ellis, C., Jr. (1965). Postglacial stratigraphy and morphology of coastal Connecticut. State Geological and Natural Hist. Survey. Guidebook No. 1. Bloom, A. L., and Stuiver, M. (1963). Submergence of the Connecticut coast. Science 139, 332-334. Blum, J. L. (1968). Salt marsh Spartinas and associated algae. Ecol. Monogr. 38, 199-221. Bouma, A. H. (1963). A graphic presentation of the facies model of salt marsh deposits. Sedirnentology 2, 122-127. Bouma, A. H. (1969). “Methods for the Study of Sedimentary Structures.” Wiley (Interscience), New York. Bowen, H. J. M. (1966). “Trace Elements in Biochemistry.” Academic Press, New York. Brannon, J. M. (1973). Seasonal variation of nutrients and physiochemical properties in the salt marsh soils of Barataria Bay, Louisiana. M.S. Thesis, Louisiana State University, Baton Rouge. Bridgeport. (1972). Local climatological data. Bridgeport, Connecticut. U.S. Dept. of Commerce, NOAA National Climate Center, Asheville, North Carolina. Broome, S. W., Woodhouse, W. W.. Jr., and Seneca, E. D. (1972). Propagation of smooth cordgrass, Spartina alterniflora, from seed in North Carolina. Chesapeake Sci. 15, 214-221. Brugam, R. B. (1976). The human disturbance history of Linsley Pond, North Branford, Connecticut. Ph.D. Thesis, Yale University, New Haven, Connecticut. Bruland, K. W., Bertine, K., Koide, M., and Goldberg, E. D. (1974). History of metal pollution in Southern California coastal zone. Environ. Sci. Technol. 8, 425-432. Bruland, K. W., Koide, M., Bowser, C., Maher, L. J., and Goldberg, E. D. (1975). Lead210 and pollen geochronologies in Lake Superior sediments. Quat. Res. 5 , 89-98. Bubela, B., and McDonald, J. A. (1969). Formation of banded sulphides: Metal ion separation and precipitation by inorganic and microbial sulphide sources. Nature (London) 221,465-466. Cawse, P. A. (1974). A survey of atmospheric trace elements in the U.K. (1972-73). AERER7669. Environmental and Medical Sciences Div., U.K. Atomic Energy Authority, Harwell, Oxfordshire. Census of Agriculture. (1933). Farm real-estate values in the New England States, 1850 to 1930. U.S.Govt. Printing Office, Washington, D.C. Chamberlain, A. C. (1960). Aspects of the deposition of radioactive and other gases and particles. Int. J. Air Pollut. 3, 63-88. Chapman, V. J. (1960). “Salt Marshes and Salt Deserts of the World.” Wiley (Interscience), New York. Chapman, V. J. (1974). “Salt Marshes and Salt Deserts of the World” 2nd Suppl. ed. Von J. Cramer, 3301 Lehre, Germany. Coghill, C. V., and Likens, G. E. (1974). Acid precipitation in the northeastern United States. Water Resour. Res. 10, 1133-1137. Connecticut Agric. Exp. Stn. Rep. (1912). Mosquito control work in Connecticut in 1912. pp. 273-274.

SEDIMENT AND TRACE METALS IN A SALT MARSH

23 1

Connecticut Dept. of Environ. Protection (1975). Unpublished data, 1969-1973, Air Compliance Sect. State Office Bldg., 165 Capital Ave., Room 150, Hartford, Connecticut, 06115.

Cook, G. H. (1857). On a subsidence of the land on the sea-coast of New Jersey and Long Island. Am. J. Sci. 24, 341-354. Crozaz, G., Picciotto, E., and DeBreuck, W. (1964). Antarctic snow chronology with Pb210. J. Geophys. Res. 69, 2597-2604. Davis, C. A. (1910). Salt marsh formation near Boston and its geological significance. Econ. Geol. 5, 623-639. Davis, C. A. (1913). Peat deposit of geological interest near New Haven, Connecticut. Bull. Geol. SOC.Am. 24, 700 (Abstr.). Dawson, J. W. (1855). “Arcadian Geology.” Oliver & Boyd, Edinburgh. DeLouise, E. R. (1968). “Air pollution in New Haven, Connecticut. Status Report on the Problem and the Program.” New Haven Health Department. Duce, R. A., Hoffman, G. L., and Zoller, H. (1975). Atmospheric trace metals at remote northern and southern hemisphere sites: Pollution or natural? Science 187, 59-61. Duce, R. A., Wallace, G. T., Jr., and Ray, B. J. (1976). Atmospheric trace metals over the New York Bight. NOAA Tech. Rept. ERL 361-MESA 4, U.S. Dept. of Commerce, Environmental Research Lab., Boulder, Colorado. East Haven (1968, 1970). Community Monographs. (On file with South Central, Regional Planning Agency, 96 Grove Street, New Haven, Connecticut.) Edgington, D. N., and Robbins, J. A. (1976). Records of lead deposition in Lake Michigan sediments since 1800. Environ. Sci. Technol. 10, 266-274. Ellis, C. W. (1962). Marine sedimentary environments in the vicinity of the Norwalk islands, Connecticut. Bull. No. 94, State Geol. and Natural Hist. Survey of Connecticut. Erlenkeuser, H., Suess, E., and Willkomm, H. (1974). Industrialization affects heavy metal and carbon isotope concentrations in recent Baltic Sea sediments. Geochim. Cosmochim. Acta 38, 823-842. Esmen, N. A., and Corn, M. (1971). Residence time of particles in urban air. Atmos. Environ. 5,571-578.

Fairbridge, R. W. (1961). Eustatic changes in sea level. in “Physics and Chemistry of the Earth,” Vol. 4, pp. 99-185. Pergamon, Oxford. Feeley, H. W., and Larsen, R. J. (1979). The chemical composition of atmospheric deposition. Environ. Measure. Lab. Q . Rep. (EML-363, Appendix B) DOE, Oct. Feeley, H. W., and Seitz, H. (1970). Use of lead-210 as a tracer of transport processes in the stratosphere. J. Geophys. Res. 75, 2885-2894. Fisenne, I. M. (1968). Distribution of ”OPb and 226Rain soils. Proc. Ann. Bioassay Anal. Chem. Meet., 13th, Lawrence Radiat. Lab., Oct. 1967. Flessa, K. W., Constantine, K. J., and Cushman, M. K. (1977). Sedimentation rates in a coastal marsh determined from historical records. Chesapeake Sci. 18, 172-176. Flint, R. F. (1971). “Glacial and Quaternary Geology.” Wiley, New York. Francis, C. W., Chesters, G., and Erhardt, W. H. (1968). ”oPolonium entry into plants. Environ. Sci. Technol. 2, 690-695. Fraser, D. C. (1961a). A syngenetic copper deposit of recent age. Econ. Geol. 56,951-962. Fraser, D. C. (1961b). Organic sequestration of copper. Econ. Geol. 56, 1063-1078. Galliher, E. W. (1933). The sulfur cycle in sediments. J. Sed. Petrol. 3, 51-63. Ganong, W. F. (1903). The vegetation of the Bay of Fundy salt and diked marshes. Bot. Gaz. 36, 161-186, 280-302, 349-367,429-455. Gatz, D. F. (1975). Pollutant aerosol deposition into southern Lake Michigan. Water Air Soil Pollut. 5, 239-251.

232

RICHARD J. MCCAFFREY AND JOHN THOMSON

Gilbert, C., and Olmstead, F. L. (1910). Report of the New Haven Civic Improvement Commission. New Haven, Connecticut. Goldberg, E. D. (1963). Geochronology with 210Pb.Proc. Radioactive Dating Symp. Athens, NOV.1962, pp. 121-131. Goldberg, E. D., Gamble, E., Griffin, J. J., and Koide, M. (1977). Pollution history of Narragansett Bay as recorded in its sediments. Estuarine Coastal Mar. Sci. 5,549-561. Goldhaber, M. B., and Kaplan, I. R. (1975). Apparent dissociation constants of hydrogen sulfide in chloride solutions. Mar. Chem. 3, 83-104. Gordon, R. B., and Pilbeam, C. C. (1975). Circulation in central Long Island Sound. J . Geophys. Res. 80, 414-422. Graham, W. F., Bender, M. L., and Klinkhammer, G. P. (1976). Manganese in Narragansett Bay. Limnol. Oceanogr. 21,665-673. Harper, R. M. (1918). Changes in the forest area of New England in three centuries. J. For. 16,442-452.

Harrison, E. Z. (1975). Sedimentation rates, shoreline modification and vegetation changes on tidal marshes along the coast of Connecticut. M.S. Thesis, Cornell University, Ithaca, New York. Harrison, E. Z., and Bloom, A. L. (1974). The responses of Connecticut salt marshes to the recent rise in sea level. Abstr. Geol. SOC. Am. Northeast. Sect. Ann. Meet., 9th, Balfimore pp. 35-36. Hewlett, J. D., and Nutter, W. L. (1969). “An Outline of Forest Hyrdrology.” Univ. of Georgia Press, Athens, Georgia. Hicks, S. D. (1972). Vertical crustal movements from sea level measurements along the East Coast of the United States. J. Geophys. Res. 77, 5930-5934. Hicks, S. D. (1973). Trends and variability of yearly mean sea level 1893-1971. National Ocean Survey, NOAA Tech. Memo. NOS. 12. Hicks, S. D., and Crosby, J. E. (1974). Trends and variability of yearly mean sea level 1893-1972. National Ocean Survey, NOAA Tech. Memor. NOS. 13. Hicock, H.W. (1970). An early history of forests and forestry in Connecticut. Conn. Woodlands 35, 26-40. Hill, C. R. (1960). Lead-210 and Polonium-210 in grass. Nature (London) 187, 211-212. Hill, D. E., and Shearin, A. E. (1970). Tidal marshes of Connecticut and Rhode Island. Bull. 709. Conn. Agric. Exp. Stn., New Haven, Connecticut. Hoffman, G. L., Duce, R. A,, and Hoffman, E. J. (1972). Trace metals in the Hawaiian marine atmosphere. J . Geophys. Res. 77, 5322-5329. Hohndorf, A. (1969). Bestimmung der Halbwertszeit von *l0Pb. Z . Naturforsch. 24A, 612-615.

Holdren, G . R., Jr., Bricker, 0. P., 111, and Matisoff, G. (1975). A model for the control of dissolved manganese in the interstitial waters of Chesapeake Bay. I n “Marine Chemistry in the Coastal Environment” (T. M. Church, ed.), pp. 364-381. Am. Chem. SOC., Washington, D.C. Huntzicker, J. J., Friedlander, S. K., and Davidson, C. 1. (1975). Material balance for autoemitted lead in the Los Angeles Basin. Environ. Sci. Technol. 9, 448-457. Jenne, E. A. (1%8). Controls on Mn, Fe, Co, Ni, Cu and Zn concentrations in soils and water: The significant role of hydrous Mn and Fe oxides. In “Trace Inorganics in Water” (R. F. Gould, ed.), pp. 337-387. Am. Chem. SOC.,Washington, D.C. Johnson, D. S., and York, H. H. (1915). The reaction of plants to tide levels. Publ. No. 206, Carnegie Inst. of Washington. Johnson, D. W. (1967). “The New England-Acadian Shoreline.” Hafner, New York. Kaye, C. A., and Barghoorn, E. S. (1964). Late quarternary sea-level change and crustal

SEDIMENT AND TRACE METALS IN A SALT MARSH

233

rise at Boston, Massachusetts, with notes on the auto-compaction of peat. Geol. SOC. Am. Bull. 75, 63-80. Keefe, C. W. (1972). Marsh production: A summary of the literature. Contrib. Mar. Sci. 16, 163-181. Kirk, J. M. (1939). The weather and climate of Connecticut. Bull. 61, State Geol. and Natural Hist. Survey. Kneip, T. J., and Eisenbud, M. (1974). Trace metals in urban aerosols. Progress Report to the Edison Electric Inst. and American Petroleum Inst., April 30. Knight, J. B. (1934). A salt-marsh study. Am. J. Sci. 28, 161-181. Koide, M., Soutar, A., and Goldberg, E. D. (1972). Marine geochronology with 210-Pb. Earth Planet. Sci. Lett. 14, 442-446. Koide, M., Bruland, K. W., and Goldberg, E. D. (1973). Th-228Rh-232 and Pb-210 geochronologies in marine and lake sediments. Geochim. Cosmochim. Acta 37, 1171-1 187. Kolthoff, I. M., and Sandell, E. B. (1956). “Textbook of Quantitative Inorganic Analysis,” 3rd ed. Macmillan, New York. Krishnaswamy, S., Lal, D., Martin, J. M., and Meybeck, M. (1971). Geochronology oflake sediments. Earth Planet. Sci. Lett. 11, 407-414. Krishnaswamy, S., Lal, D., Amin, B. S., and Soutar, A. (1973). Geochronological studies in Santa Barbara basin: Fe-55 as a unique tracer for particulate settling. Limnol. Oceanogr. 18, 763-770. Lawrence, A. W., and McCarthy, P. L. (1965). The role of sulfide in preventing heavy metal toxicity in anaerobic treatment. J . Water Pollut. Control Fed. 37, 396-406. Lazrus, A. L., Lorange, E., and Lodge, J. P., Jr. (1970). Lead and other metals in United States precipitation. Environ. Sci. Technol. 4, 55-58. Lee, R. E., Jr., and von Lehmden, D. J. (1973). Trace metal pollution in the environment. J. Air Pollut. Control Assoc. 23, 853-857. Lee, R. E., Jr., Patterson, R. K., and Wagman, J. (1968). Particle-size distribution of metal components in urban air. Environ. Sci. Technol. 2, 288-290. Lewis, D. M. (1976). The geochemistry of manganese, iron, uranium, lead-210 and major ions in the Susquehanna River. Ph.D. Thesis, Yale University, New Haven, Connecticut. Lutz, H. J. (1928). Trends and silvicultural significance of upland forest successions in southern New England. Yale School of Forestry, Bull. 22. Yale University, New Haven, Connecticut. Lutz, H. J. (1940). Disturbance of forest soil resulting from the uprooting of trees. Yale School of Forestry, Bull. 45. Yale University, New Haven, Connecticut. Lull, H. W., and Reinhart, K. G.(1972). Forests and floods in the eastern United States. U.S.D.A. Forest Service Res. Paper, NE-226. Northeast Forest Exp. Stn., Upper Darby, Pennsylvania. Lyford, W. H. (1963). Importance of ants to brown podzolic soil genesis in New England. Harvard Forest Paper No. 7. Harvard Univ., Harvard Forest, Petersham, Massachusetts 01366. McCaffrey, R. J. (1977). A record of the accumulation of sediment and trace metals in a Connecticut, U.S.A., salt marsh. Ph.D. Thesis, Yale University, New Haven, Connecticut. McCaffrey, R. J., Myers, A. C., Davey, E., Morrison, G., Bender, M., Luedtke, N., Cullen, D., Froelich, P., and Klinkhammer, G. (1980). The relation between porewater chemistry and benthic fluxes of nutrients and manganese in Narragansett Bay. Limnol. Oceanogr. 25, 31-44. MacDonald, R. D. S. (1968). Multiple use forestry in megalopolis: A case study of the

234

RICHARD J . MCCAFFREY AND JOHN THOMSON

evolution of forest policies and programs in Connecticut. Ph.D. Thesis, Yale University, New Haven, Connecticut. Manheim. F. T. (1966). A hydraulic squeezer for obtaining interstitial water from consolidated and unconsolidated sediments. U.S. Geol. Survey Prof. Paper 550-C, C256-261. Meade, R. H. (1972). Sources and sinks of suspended matter on continental shelves. I n “Shelf Sediment Transport: Process and Pattern” (D. Swift, D. Duane, and 0. Pilkey, eds.), Dowden, Hutchinson & Ross, Stroudsburg, Pennsylvania. Miller, W. R., and Egler, F. E. (1950). Vegetation of the Wequetequock-Pawcatuck tidalmarshes, Connecticut. Ecol. Monogr. 20, 143-172. Milliman, J. D., and Emergy, K. 0. (1968). Sea levels during the past 35,000years. Science 162, 1121-1 123. Moore, H. E., Poet, S. E., and Martell, E. A. (1973). 222Rn,”‘Pb, 21’Bi and 2’”Poprofiles and aerosol residence times versus altitude. J. Geophys. Res. 78, 7065-7075. Morgan, J. J. (1967). Chemical equilibria and kinetic properties of manganese in natural waters. I n “Principles and Applications of Water Chemistry” (S, P. Faust and J. V. Hunter, eds.). Wiley, New York. Morrow, N. L., and Brief, R. S . (1971). Elemental composition of suspended particulate matter in metropolitan New York. Environ. Sci. Technol. 5 , 786-789. Mudge, B. F. (1862). The salt marsh formations of Lynn. Proc. Essex Insr. ZZ (1856-1860) (Paper presented by Mudge, Feb. 13, 1857), pp. 117-119. Niering, W. A., and Warren, R. S . (1974). Tidal marshes of Connecticut: Vegetation, microrelief and environmental impacts. Connecticut College, New London, Connecticut. Niering, W. A., Warren, R. S., and Weymouth, C. G . (1977). Our dynamic salt marshes: Vegetation change as revealed by peat analysis. Connecticut Arboretum Bull. No. 22. Connecticut College, New London. Nixon, S. W., and Oviatt, C. A. (1973). Ecology of a New England salt marsh. Ecol. Monogr. 43, 463-478. Peirson, D. H., Cawse, P. A,, Salmon, L., and Cambray, R. S. (1973). Trace elements in the atmospheric environment. Nature (London) 241, 252-256. Perkin-Elmer Corp. (1964, 1968). “Analytical Methods for Atomic Absorption Spectrophotometry” (with supplements). Norwalk, Connecticut. Pohl, H. A. (1975). The formation and dissolution of metal sulfides. J . Am. Chem. SOC. 76, 2182-2184. Ponnamperuma, F. N. (1972). The chemistry of submerged soils. I n “Advances in Agronomy” (N. C. Brady, ed.), Vol. 24, pp. 29-96. Academic Press, New York. Presley, B. J., Kolodny, Y., Nissenbaum, A., and Kaplan, I. R. (1972). Early diagenesis in a reducing fjord, Saanich Inlet, British Columbia-11. Trace element distribution in interstitial water and sediment. Geochim. Cosrnochim. Acta 36, 1037-1090. Rahn, K. A. (1971). Sources of trace elements in aerosols-an approach to clean air. Ph.D. Thesis, University of Michigan, Ann Arbor. Redfield, A. C. (1967). Postglacial change in sea level in the Western North Atlantic Ocean. Science 157, 687-692. Redfield, A. C. (1972). Development of a New England salt marsh. Ecol. Monogr. 42, 201 -237. Redfield, A. C., and Rubin, M. (1962). The age of salt marsh peat and its relation to recent changes in sea level at Barnstable, Massachusetts. Proc. Nntl. Acad. Sci. U.S.A. 48, 1 728- 1 735. Rhoads, D. C. (1967). Biogenic reworking of intertidal and subtidal sediments in Barnstable Harbor and Buzzards Bay, Massachusetts. J. Geol. 75,461-476.

SEDIMENT AND TRACE METALS IN A SALT MARSH

235

Richard, G. A. (1978). Seasonal and environmental variations in sediment accretion in a Long Island salt marsh. Estuaries 1, 29-35. Richards, F. J. (1934). The salt marshes of the Dovey Estuary. IV. The rates of vertical accretion, horizontal extension and scarp erosion. Ann. Bot. 48, 225-259. Robbins, J. A., and Callender, E. (1975). Diagenesis of manganese in Lake Michigan sediments. Am. J. Sci. 275, 512-533. Robbins, J. A., and Edgington, D. N. (1975). Determination of recent sedimentation rates in Lake Michigan using Pb-210 and Cs-137. Geochim. Cosmochim. Acta 39, 285-304. Rodhe, H . , and Grandell, J. (1972). On the removal time of aerosol particles from the atmosphere by precipitation scavenging. Tellus 24, 442-454. Sackett, W. M. (1965). Deposition rates by the protactinium method. I n “Marine Geochemistry,” (D. R. Schink and J. T. Corless, eds.), pp. 29-39. Univ. of Rhode Island Occas. Publ. No. 3. Sanders, G. (1978). The sources of dissolved Mn to Calico Creek, North Carolina. Estuarine Coastal Mar. Sci. 6 , 231-238. Scholl, D. W., and Stuiver, M. (1967). Recent submergence of southern Florida: A comparison with adjacent coasts and other eustatic data. Geol. SOC.Am. Bull. 78,437-454. Shaler, N. S. (1885). Sea-coast swamps of the Eastern United States. U.S. Geol. Survey Ann. Rep., 6th pp. 359-368. Shepard, F. P. (1963). Thirty-five thousand years of sea level. I n “Essays in Marine Geology in Honor of K. 0. Emery” (T. Clements, ed.), pp. 1-10. Univ. of S. Calif. Press, Los Angeles, California. Siccama, T. G., and Porter, E. (1972). Lead in a Connecticut salt marsh. BioScience 22, 232-234. Smith, J. B. (1907). The New Jersey salt marsh and its improvement. New Jersey Agric. Exp. Stn. Bull. No. 207. Stalter, R., and Batson, W. T. (1969). Transplantation of salt marsh vegetation, Georgetown, South Carolina. Ecology 50, 1087-1094. Stern, A. C., Wohlers, H. C., Boubel, R. W., and Lowry, W. P. (1973). “Fundamentals of Air Pollution.” Academic Press, New York. Struempler, A. W. (1975). Trace element composition in atmospheric particulates during 1973 and the summer of 1974 at Chadron, Neb. Environ. Sci. Technol. 9, 1165-1168. Stumm, W., and Morgan, J. J. (1970). “Aquatic Chemistry: An Introduction Emphasizing Chemical Equilibria in Natural Waters.” Wiley (Interscience), New York. Sweeney, R. E., and Kaplan, I. R. (1973). Pyrite framboid formation: Laboratory synthesis and marine sediments. Econ. Geol. 68, 618-634. Teal, J. M. (1962). Energy flow in the salt marsh ecosystem ofGeorgia. Ecology 43,613-624. Teal, J. M., and Kanwisher, J. W. (1966). Gas transport in the marsh grass, Spartina alterniflora. J . Exp. Bot. 17, 355-361. Thomas, M. P. (1972). Gazetteer of natural drainage areas of streams and water bodies within the state of Connecticut. Bull. No. 1, Connecticut Dept. of Environmental Protection, Hartford. Thomson, J., Turekian, K. K., and McCaffrey, R. J. (1975). The accumulation of metals in and release from sediments of Long Island Sound. I n “Estuarine Research” (L. E. Cronin, ed.), Vol. I , pp. 28-44. Academic Press, New York. Turekian, K. K., Kharkar, D, P., and Thomson, J. (1973). Uranium and thorium decay series nuclide abundances in marine plankton. Final Report to Advanced Project Research Agency, ARPA, Order No. 1793, Contract N00014-67-A-0097-0022.

236

RICHARD J. McCAFFREY AND JOHN THOMSON

Turekian, K. K., Nozaki, Y., and Benninger, L. K. (1977). Geochemistry of atmospheric radon and radon products. Annu. Rev. Earth Planet. Sci. 5 , 227-255. U.S. Dept. of Commerce (1972). Local climatological data: Annual summary with comparative data. Bridgeport, Connecticut. National Oceanic and Atmospheric Administration, Rockville, Maryland. U.S. Dept. of the Interior (1973). United States Mineral Resources. Geol. Survey Prof. Paper No. 820. U.S.G.S. (U.S.Geological Survey) (1960). Water Supply Papers 1302, lB, Compilation of Records of Surface Water of the U.S. through September 1950. Hudson River at Mechanickville, New York, 4500 mi2. Valiela, I., and Teal, J. M. (1974). Nutrient limitation in salt marsh vegetation. I n “Ecology of Coastal Halophytes” (R. J. Reimold and W. H. Queen, eds.), pp. 547-563. Academic Press, New York. Van Cleef, E. (1908). Is there a type of storm path? Mon. Weather Rev. (U.S. Dep. Agric.) 36, 56-58. Volchok, H. L., and Bogen, D. (1971). Trace metals-fallout in New York City. Health Safety Lab. Fallout Program Q.Summary Rep. (U.S.A.E.C.)April, pp. 1-91 to 1-107. Walcott, R. I. (1975). Recent and late quaternary changes in water level. Trans. Am. Geophys. Union 56, 62-72. Winchester, J. W., and Nifong, G. D. (1971). Water Pollution in Lake Michigan by trace elements from pollution aerosol fallout. Water Air Soil Pollut. 1, 50-64. Windom, H. (1%9). Atmospheric dust records in permanent snowfields: Implications to marine sedimentation. Geol. SOC.Am. Bull. 80, 761-782. Wolman, M. G. (1967). A cycle of sedimentation and erosion in urban river channels. Geogr. Ann. 49A, 385-395. Wright, J. R., Levick, R., and Atkinson, H. J. (1955). Trace element distribution in virgin profiles representing four great soil groups. Soil Sci. 19, 340-344. Zoller, W. H., Gladney, E. S., and Duce, R. A. (1974). Atmospheric concentrations and sources of trace metals at the South Pole. Science 183, 198-200.

DIAGENETIC PROCESSES NEAR THE SEDIMENT-WATER INTERFACE OF LONG ISLAND SOUND I . Decomposition and Nutrient Element Geochemistry (S. N. P)

.

ROBERTC . ALLER* Department of Geology and Geophysics Yale Universiry New Haven. Connecticut

.

Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Location of Study and Station Description . . . . . . . . . . . . . . . . . . . . 2.1. Shallow-Water Station FOAM . . . . . . . . . . . . . . . . . . . . . . . . 2.2. Intermediate-Water-Depth Station NWC . . . . . . . . . . . . . . . . . . . 2.3. Deepwater Station DEEP . . . . . . . . . . . . . . . . . . . . . . . . . 3. Sampling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4 . Methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.1. Treatment of Cores . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.2. Analytical Techniques . . . . . . . . . . . . . . . . . . . . . . . . . . . 5 . Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.1. Pore-Water Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.2. Flux Measurements . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.3. Solid-Phase Analyses . . . . . . . . . . . . . . . . . . . . . . . . . . . 6 . Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.1. Decomposition Reactions . . . . . . . . . . . . . . . . . . . . . . . . . . 6.2. Supply and Reactivity of Organic Material . . . . . . . . . . . . . . . . . . . 6.3. Products of Decomposition-Solid Phase . . . . . . . . . . . . . . . . . . . 6.4. Transport Processes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.5. Seasonal and Spatial Variation in Pore-Water Profiles-General Patterns . . . . . . 6.6. One-Dimensional Models of Pore-Water Distributions in the Zone of Bioturbation . . 6.7. Two-Dimensional Models of Pore-Water Distributions in the Zone of Bioturbation . . 6.8. Abiogenic Reaction Controls on Pore-Water Composition . . . . . . . . . . . . 6.9. Flux of NH: and HP0:- between Sediment and Overlying Water . . . . . . . . . 6.10. Stoichiometry of Decomposition . . . . . . . . . . . . . . . . . . . . . . 7 . Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Appendix A . Macrofauna (21 mm) Sieved from Flux-Core Boxes . . . . . . . . . . . Appendix B. Box-Core and Gravity-Core Data from Long Island Sound . . . . . . . . Appendix C. Flux-Core Data . . . . . . . . . . . . . . . . . . . . . . . . . . List of Symbols . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1

2.

238 238 240 244 246 250 252 252 255 257 257 264 269 272 272 274 276 279 282 285 293 303 308 315 317 320 322 340 343 344

* Present address: Department of The Geophysical Sciences. The University of Chicago. Chicago. Illinois 60637. 231 Copyright 0 1930 by Academic Press. Inc . All rights of reproduction in any form reserved . ADVANCES IN GEOPHYSICS. VOLUME 22 ISBN 0.12-018822-8

238

ROBERT C . ALLER

1. INTRODUCTION

Early diagenetic reactions, particularly those directly or indirectly involving the decomposition of organic matter, are most intense and rapid in the upper 1 m and especially the upper 10 cm of marine sediment (Goldhaber et a / . , 1977; Sayles, 1979). It is in this upper zone where most benthic organisms live and interact with sediments and where exchange rates of dissolved and particulate material between sediment and overlying water are largely determined. Knowledge of diagenetic processes taking place in this zone is therefore essential for understanding the chemistry of sediments, the chemistry of water overlying sediments, certain ecological interactions and adaptations of marine organisms, and the longterm recording of historical information in marine deposits; for example, fossil preservation. In this and the following companion article (Part 11), selected early diagenetic reactions associated with the decomposition of organic matter in estuarine deposits of Long Island Sound are examined. Particular emphasis is placed on understanding the role of benthic macroorganisms together with depositional environment in controlling the composition of surface sediments and in determining the flux of solutes between sediment and overlying water. The study has been broken into two parts; the first concentrates on describing diagenetic processes involving organic-matter decomposition and production or consumption of the nutrients SO’,-, NH; , alkalinity, and HPO’,-, The second emphasizes the associated chemical behavior of Fe and Mn (Part 11). Several types of measurements were made: (1) seasonal pore water and solid-phase analyses, (2) direct measurement of solute fluxes out of the sediment, (3) rates of reaction as a function of depth and temperature, and (4) the abundance and composition of the fauna at each station. Taken together, these measurements provide one of the most detailed descriptions of controls on diagenesis near the sediment-water interface that is presently available. 2. LOCATION OF STUDY AND STATION DESCRIPTION

Three stations (FOAM, NWC,and DEEP) representing three different depositional environments in central Long Island Sound were chosen for study (Fig. 1). The relative bathymetry of the stations is shown in the north-south cross section of Fig. 2. As illustrated by the thickness of marine mud accumulated at each site during the last 5000 years, the longterm sedimentation rate does not greatly differ in the three areas (Bokuniewicz et al., 1976). No stations were located in the region of highest

d NWC

A

t

CONN

N

05'

DEEP 0

-

0

I 2 3 4 5 k m

Icm

40°00'

73"

72O

B

FIG. 1. (A) Location of Long Island Sound along northeastern United States. (B) Long Island Sound showing central basin where stations were established. (C) Location of stations FOAM, NWC,and DEEP in the central Sound.

240

ROBERT C. ALLER FOAM

LATITUDE

FIG. 2. Cross section of central basin through approximately 72"53' with thickness of marine mud deposits outlined (after Bokuniewicz et al., 1976). Stations are passed onto this single plane illustrating relative position with respect to bathyrnetry and long-term desposition.

long-term sedimentation rate, but the average rates of -0.02-0.06 cm/yr are so low as to be unimportant in this study. A concise qualitative description of the three study sites is given in Table I. All stations in less than about 20 m of water are subject to constant resuspension of surface sediment creating highly turbid conditions. During major storms the upper few centimeters at these stations may be eroded and redeposited (Aller and Cochran, 1976; McCall, 1978), but normally only the top 1 mm or so is subject to resuspension.(Bokuniewicz, 1976). DEEP station is also subject to resuspension, but only through tidal scour and it is considerably lower in turbidity than either FOAM or NWC. The bottom fauna in the central Sound have distinct bathymetric patterns in distribution. These patterns are delineated by changes in the relative abundance of animals of certain sizes, life habits, life history, and feeding types in different regions of the bottom, and reflect depositional environment and severity frequency of physical disturbances (McCall, 1977). This allows comparison, in an inshore-offshore transect, of diagenetic processes in sediment inhabited by animals of different feeding groups, life habits, and life histories, all of which determine the ways in which animals interact with sediment. Based on direct observation of the bottom on numerous occasions and sieve analysis (-1-mm mesh) of box cores, taken at FOAM, NWC, and DEEP (Rhoads et al., 1977; Yingst and Rhoads, 1978; Appendix l), a composite description of the fauna at each station can be made. These fauna are shown schematically in Figs. 3, 6, and 9. 2.1. Shallow- Water Station FOAM FOAM contains a diverse assemblage of infaunal benthos consisting of mixed trophic groups, but mostly of deposit feeders. Large numbers

TABLEI. LONGISLANDSOUNDSTATIONLOCATION AND DESCRIPTION Location Depth Station

Latitude

Longitude

(m)

FOAM

41O14.5'

72O44.8'

-8

NWC

41Y0.4'

72'56.3'

-15

DEEP

41"03.0

72O53.1'

-34

Relative depositional environment Shallow, partially protected nearshore

Sediment type

Color stratigraphy

-

Comments

Silt-clay, sand present, Top 0 3-cm H2S smell strong below 10 cm fecal-pelleted surface yellow-orange, 3 530% CaCO, shell 10-cm black, olive at debris present in )I0 cm massive layers H2S smell absent or Intermediate depth Silt-clay, sand pockets Top 0 3-cm unprotected offshore at depths 2 1 m fecal- yellow-orange, 3-8faint at all depths peUeted surface cm black, grey at )8 sampled 5 10% CaC03 shell cm. Yellow layer is debris present in seasonal in thickness, discrete, thin layers thin in early summer, thickest in fall. (Rhoads et al., 1977) Deep, unprotected, Silt-clay, thin sand Top 0-0.5-cm H2S smell absent or offshore layers present fecalorange-yellow. faint at all depths pelleted surface 5 5 % Mottled black and sampled CaCO, shell debris orange from 0.5-10 absent or very cm. Grey at 210 cm scattered

-

-

242

ROBERT C. ALLER

FIG.3. Schematic drawing of major fauna at FOAM: A, Ampelisca; C, Clymenella sp.; M . l . , Mulinia lateralis; M.t., Macoma tenta; N.a., Nucula annulata; N.i.,Nephtys incisa; N.t, Nassarius trivitatus ; P., Pagurus :T.a., Tellina agilis; KO., Spiochaetopterus oculatus.

of the surface deposit-feeding amphipod Ampelisca sp. are characteristic of this station. During the summer and fall they form dense mats of dwelling tubes at the interface that bind the sediment surface; these are eroded into irregular clumps during the winter, at which time the population levels are greatly reduced. Other surface deposit feeders present are the tellinid bivalves (Tellina agilis and Macoma tenta). Deep-feeding bivalves are represented by small numbers of Nucula unnuluta and Yoldiu limatula. Other deep feeders or burrowers are the sedentary maldanid polychaete Clymenella sp. and the highly mobile polychaete Nephtys incisa, both of which can burrow to depths > 12 cm. The gastropods, Nassarius trivitatus (deposit feeder) and Retusa sp. (predator), together with the hermit crab Paguris sp. (scavenger) comprise the easily captured mobile epifauna. Suspension feeders include the bivalve Mulinia lateralis and the deepburrowing (>25 cm), sedentary polychaete Spiochaetopterus oculatus. FOAM is characterized by the presence of both mobile and sedentary tube dwellers and dominated by deposit feeders. Most animals are very small in size and, on the basis of direct inspection by diving, vary greatly from year to year in their occurrence. These observations, together with the physical setting, the large numbers of the typical opportunist Am-

DIAGENETIC PROCESSES. I.

243

pelisca, and the high diversity of apparently short-lived (small) benthos, all suggest that the FOAM site is highly physically disturbed and in a state of constant colonization by a wide variety of organisms (see criteria of McCall, 1977, 1978). Representative x radiographs of sediment samples taken at FOAM illustrate sedimentary features indicative of both the physical and biological conditions effective at this station (Figs. 4 and 5 ) . (The sampling and xradiographic techniques used are described under methods.) The small size and transitory nature of the fauna at FOAM and the intensity of its physical setting are demonstrated by the presence of dense and massive shell lag layers and preservation of physically formed laminations at depth. These layers have been subject to only minor biogenic particle reworking. Dwelling burrows and burrows of mobile fauna, the construction of which at low abundance does not homogenize sedimentary particles, are evident, but at low density compared to NWC and DEEP. The shell layers probably obscure many biogenic sedimentary features. Shells are often quite large and are composed of species not common at FOAM,

FIG.4. X radiograph of vertical sediment section at FOAM. The upper 10-12 cm are characterized by abundant shell debris. A laminated layer at -4 cm is disrupted irregularly by shells and biogenic reworking activity. Parts of two vertical maldanid tubes can be seen in the center at a depth of 8-10 cm. Below 10-12 cm, the sediment begins to become laminated. This laminated zone begins at 8 cm in most x radiographs from FOAM (see Fig. 5). (Scale: 3 cm.)

244

ROBERT C. ALLER

suggesting that physical processes tend to accumulate debris in the FOAM area. 2.2. Intermediate- Water-Depth Station NWC

NWC has a relatively stable, low-diversity assemblage of organisms characterized by large numbers of the highly active protobranch bivalves Yoldia lirnatula and Nucula annulata and the errant polychaete Nephtys incisa. These organisms were always present during this study at fairly constant levels (Rhoads et al., 1977; Yingst and Rhoads, 1978). All are deep-feeding deposit feeders. Other less reliably present fauna (Fig. 6) include the suspension-feeding bivalves Pitar morrhuana and Mulinia ; the suspension-feeding polychaete Spiochaetopterus ; and the surfacedeposit feeders Melinna cristata (arnpharetid polychaete) and small numbers of Ampelisca. Permanent tube-dwelling polychaetes other than Spi-

FIG.5 . Parts of two vertically oriented x radiographs taken at FOAM. A massive shell layer of varied thickness (6-8 cm) overlies a well-laminated zone in both cases. There is relatively little evidence of biogenic structures although burrows can be made out in the shell layer. Beneath the laminated zone are additional shell layers down to at least 50 cm. (Scale: 3 cm.)

DIAGENETIC PROCESSES. I.

245

FIG. 6. Schematic drawing of major fauna at NWC: C.a., Ceriantheopsis americanus; P m . , Pitar morrhuana; and S., Squilla; all other abbreviations as in Fig. 3.

ochaetopterus are not common, but the infaunal anemone Ceriantheopsis americanus, seldom recovered by grabs or samples, is plentiful as observed by diving. Evidence of the deep-burrowing shrimp (Squilla)is also found by diving. Easily captured epifauna are the same as FOAM. At NWC shell debris is less plentiful and is often aggregated into distinct layers that probably represent biogenic lag layers resulting from feeding (e.g., Van Straaten, 1952; Schiifer, 1962; Rhoads and Stanley, 1965; Aller and Dodge, 1974). No evidence of physically produced sedimentary structure is found and storm laminations that do form are known to be rapidly destroyed by biogenic particle reworking (Aller and Cochran, 1976). The protobranch bivalves present at NWC are responsible for much of the particle homogenization. Despite extensive homogeneous, particle reworking, a high degree of biogenic and physical heterogeneity is present in sedimentary structures as shown by horizontal x radiographs taken at NWC; an example of which, in addition to a vertical x radiograph, is given in Figs. 7 and 8. Permanent and temporary dwelling burrows create a complex geometry of biogenic shafts through the upper few decimeters of sediment. Construction of these features together with physical processes make it necessary to average measurements, especially solidphase measurements, over as large an area as possible (e.g., Hanor and Marshall, 1971).

246

ROBERT C . ALLER

FIG.7. Vertically oriented x radiograph illustrating sedimentary features typical of NWC. The upper 4-6 cm is intensely burrowed. Yoldia lirnatula can be seen at the far left and right of the radiograph. Abundant vertical burrows, due most probably to Owenia can be made out near the interface. A shell lag layer commonly found at this station occurs -6 cm and results from both feeding and storm activity. Bioturbate structure dominates at depth; burrows coming out of plane of radiograph appear as light holes. (Scale: 3 cm.)

2.3. Deep- Water Station DEEP

The polychaete fauna at DEEP (Fig. 9) are generally larger, more sedentary, and deeper burrowing than at either FOAM or NWC. Many form permanent dwelling tubes. Pherusa affinis, a flabelligerid polychaete, is conspicuous and abundant. Maldanids are also plentiful and are represented by an unidentified species (not recovered live because of burrow depth >25 cm, but tubes present in abundance, probably Asychis elongata) and a Clymenella species that forms Fe-encrusted silt-sand tubes. The terebellid, Pista palmata and the ampharetid Melinna cristata, both

DIAGENETIC PROCESSES.1.

247

sedentary surface deposit feeders, are present. The suspension-feeding polychaete Spiochaetopterus is also present. An occasional protobranch or tellinid are found but these are rare. Very large, infaunal anemones (Ceriuntheopsis ) are plentiful as .is the epifaunal hydroid Corymorpha. Burrowing activity by Squilla is evident (see Myers, 1979).

FIG.8. X radiograph of 5-cm-thick horizontal slab of sediment 10-15 cm deep from NWC. The irregular distribution of shell debris can be seen with most shell material restricted to the right side of the radiograph. Remnants of maldanid tubes can be made out as can a single layer crustacean burrow (left) passing vertically down into the sediment. Maldanids are not presently a common, permanent faunal component at NWC but occur irregularly. (Scale: 3 cm.)

DIAGENETIC PROCESSES. I.

249

FIG.10. Vertical x radiograph from DEEP.Maldanid dwelling tubes are abundant on right side. These commonly extend 20 cm into sediment. Shell debris is generally absent below a few centimeters. Where dwelling burrows are absent the sediment is characterized by a mottled, bioturbate texture (right side). (Scale: 3 cm.)

DEEP also lacks evidence of many sedimentary structures formed by physical processes. Dwelling burrows, often at high abundance, extend great distances into sediment and are characteristic of this station (Figs. 10 and 11). Despite the presence of living shell-bearing animals at DEEP (see also Saunders, 1956), little shell debris is preserved.

FIG.11. Large burrow formed by unknown organism is seen in center of this x radiograph from DEEP.The wall is lined with clods of sediment and small pebbles. Surrounding sediment shows bioturbate structure typical of sediment that is not part of a contemporary dwelling structure. (Scale: 3 cm.)

250

ROBERT C. ALLER 0:Gordon 8 Pilbearn (1974) *:Rhoads,

Aller 8 Goldhaber

(1975)5/74-4/75 .=5/75-2/76

.

0 0

v

. o

0

01 0 0

0

0.

T T T

0 :

.

o w

*09

0

0

00

FIG. 12. Seasonal temperature variation at NWC over 3-yr period. Data of Gordon and Pilbeam (1974) by thermistor, all other by direct insertion of thermometer in surface of divertaken sediment cores.

All stations are subject to a sinusoidal variation in temperature during the year with a low temperature of about 2°C in February-March and a high of about 22" in August-September (Riley, 1956a). A summary of bottom temperatures at NWC, taken by inserting a thermometer into the top 1 cm of diver-taken box cores, for the 2-yr sampling period is given in Fig. 12. 3. SAMPLING

In order to obtain undisturbed samples of consistently high quality, most bottom samples were obtained by scuba divers. Three types of handheld Plexiglas box cores were commonly taken (Fig. 13). These varied in size and dimension according to the type of analyses for which they were used (Table 11). One type, termed chem box, was used for studying interstitial water and sediment chemistry. Thinner cores, called x-ray boxes, were used to study biogenic and other sedimentary structures (Rhoads and Young, 1970; Stanley, 1970). The third type, flux boxes, were used in experiments to estimate sediment-water exchange. All cores are constructed of clear Plexiglas and are essentially boxes having removable tops with vent holes and removable bottoms (Fig. 13, Table 11). The bottom edges of the box are sharpened for easy insertion into sediment. Coring is done by gently pushing the bottomless corer into the sediment to a depth of 10-22 cm in the case of x-ray and chem boxes, or 10-15 cm for flux boxes. The vent holes in the top are then sealed with rubber

25 1

DIAGENETIC PROCESSES. I.

7 1.064cm strips

olt Holes, 48mrn i d

FIG.13. The hand-held, diver-operatedbox corer with details of construction. Dimensions A, B, and C for each corer type are given in Table 11. Vent holes use size 5 stoppers.

stoppers. After removing sediment from the exterior of one side of the core, the bottom plate is slid under the core and locked into place. Large rubber bands are then strapped around the corer to permahently secure the bottom and the core returned to the surface. If done correctly, an undisturbed vertical section of the bottom and its immediately overlying water is obtained. Water overlying the core is clear and surface sediment TABLE11. DIMENSIONS OF BOX CORERS Inside dimensions (cm) A

B

C

Wall thickness (mm)

25.2 25.2 30.5 30.5

29.2 29.2 29.2 28.1

2.5 7.6 10.2 20.5

6.2 6.2 6.2/10.2 10.2

Use X-ray sections Sediment chemistry Flux measurements Large samples, horizontal x rays

252

ROBERT C. ALLER

not resuspended; animals in the cores continue their activity unperturbed. The sealed top and overlying water prevent disturbance of the core during transport by eliminating sloshing action. After recovery, cores were placed in large cooler boxes and returned to the lab for analysis. A specifically designed corer which allowed the taking of horizontal samples in the lab was used to obtain horizontal x radiographs (Fig. 8; Table 11). Gravity cores of -1-m length were taken on one or more occasions at FOAM, NWC, and DEEP. Acetate-butyrate liners 4.7-cm i.d. were used in all cases. The sampling dates and types of cores taken at each station, for which data are presented, are given in Table 111. Often more than one core of each type was taken and only the best (e.g., longest, straightest insertion) was used for analysis. In general, box cores were taken in summer, fall, and winter-spring periods at each station and represent times of warmest, intermediate, and coldest water temperature. Sampling was carried out over 2-yr periods at FOAM and NWC and for 1 yr at DEEP.

4. METHODS

4.1. Treatment of Cores 4.1.1. X-Ray Samples. After storing at 4°C x-ray box cores were x rayed within several hours of collection by placing an 8 in. x 10 in. sheet of Kodak AA Industrial x-ray film on the back and exposing it for -2.5 min. using 60 Kv x rays at a distance of 60-75 cm. A dental x-ray unit was used. Positive prints of the x-ray negatives were then made. Because of gradients in water content near the surface, positives were often printed so as to reduce the resulting gradient in exposure density in the negative. This does not affect interpretation of structures, except by giving better resolution near the interface. 4.1.2. Pore-Water and Sediment Samples. Chem cores were stored in the dark at 4°C for a period of 3-5 hr after collection. Before processing the core, the color stratigraphy, general sediment features, and readily identified animal life were described and recorded. The sides of the box were ruled, using a marking pen, at 1-cm intervals from the sediment-water interface to within 3 cm of the base of the particular core being sampled. Water overlying the core was then rapidly removed using a vacuum line, the core top and fasteners removed, and the entire core placed in a glove bag. This bag was pumped down and purged several times with NZ (0,

253

DIAGENETIC PROCESSES. I.

TABLE111. SAMPLING SCHEDULE Station FOAM-1 FOAM-2 FOAM-3 FOAM-4

Date

FOAM-5 FOAMG

08/5/74 03/13/75 07/8/75 11/5/75 11/17/75 03/18/76 07/8/75

NWC-1 NWC-2 NWC-3 NWC-4 NWC-5 NWC-6 NWCG

07/21/74 11111/74 03118/75 07/16/75 10/29/75 03/23/76 11111/74

DEEP- 1 DEEP-2 DEEP-3 DEEPG

07/23/75 11/12/75 04113/76 11/12/75

Chem box

Flux box

X

X-ray box

Gravity core

X X

X X X X X

X

X

X

X

X

X

X

X X

free) and subsequent sampling of the core done under a nitrogen atmosphere. Sediment was removed at 1-cm intervals using plastic spoons, then packed into 10-cm long sections of butyrate tubing (4.7 cm i.d.). Subsamples of sediment were saved for percent H,O determination. pH and Eh or pS measurements were then made as described later and the tubes sealed with plastic end caps. Care was taken to avoid entrapment of N, in the sediment taken from each interval. Degassing is not a problem, as shown by comparing values of H,S concentrations (methylene blue analysis) obtained by this sampling procedure with other methods (Goldhaber et al., 1977). The sediment in these sections is squeezed within a few hours for removal of interstitial water as later described. Samples were stored at 4°C at times when they were not being processed. Most samples were squeezed within 12 hours, but occasionally a few samples were squeezed as late as 24 hours after collection. The method of Kalil and Goldhaber (1973) was used in all cases. After removal from the storage cooler, the 10-cm-long butyrate sections were quickly uncapped, approximately 1 cm of sediment scooped out from each end, and glass fiber filter paper inserted against the sediment. These sections were then squeezed and pore water passed through in-line (0.45-pm pore size)

254

ROBERT C. ALLER

Millipore filters directly into syringes (Kalil and Goldhaber, 1973). The first -5 ml of expressed water at each end was discarded to avoid air contamination. In the case of summer and fall samples, the sediment was allowed to warm to room temperature before squeezing. The winter 1975 samples were squeezed immediately after removal from the refrigerator; winter 1976 samples were squeezed at 4°C using the cooling units described by Kalil and Goldhaber (1973). Gravity cores were processed in the same way, with two differences. The core was sectioned into 10-cm intervals in a miter box using a hacksaw. When sections were squeezed, water taken from each was either kept separate and assigned to different 5-cm intervals or combined into a single sample. Pore-water samples were treated using the following methods. Ten ml of pore water were passed from a syringe through a short section of tygon tubing into a 10-ml pipet. This pipet was drained into a beaker and the water acidified for SO:- analysis. A second 10 ml, or sometimes 15 ml, sample from the same depth interval was drained into an acid-washed and distilled water-rinsed 2 oz. polyethylene widemouthed bottle. This sample was titrated for alkalinity and acidified within 48 hours (occasionally 72 hours) in the same 2-02 container. After acidification it was used for Fe and Mn analysis as will be described in Part 11, and in some cases Ca. A 2-ml sample for C1- was pipeted immediately into a glass vial for later titration in that same vial (1975-1976 samples only). Separate samples were taken for HPO:-, NH: , and H,S analysis. Samples for HP0:- and NH: were kept in sealed, air-free syringes and placed immediately into a 4°C refrigerator for storage. These were analyzed, or fixed in the case of NH:, within 24 hr. The sediment cake that resulted from squeezing was retained in its butyrate tube, sealed with end caps, and placed in a freezer for later analysis of the solid phase. Ignition loss, percent CaCO,, acid-volatile sulfide, and total nonvolatile sulfur were measured on many samples. Metal analyses were also made (Part 11). 4.1.3. Flux Samples. Two flux boxes were always taken (1975-1976 only); one contained only bottom water and acted as a blank, the other was filled with 10-15 cm of undisturbed bottom sediment and its overlying water. Summer (1975) boxes were kept at 22"C, fall flux boxes at 15"C, and winter at 4°C. The original FOAM fall flux box developed a leak so that a second flux core was taken 12 days later as a replacement. Processing of these boxes consisted of withdrawing water samples from each box at successive times and following the change in overlying water con-

DIAGENETIC PROCESSES. 1.

255

centrations of selected, dissolved ions. The total initial overlying water volume was determined and volumes removed at each sampling were measured. Fluxes were calculated as described subsequently. The summer and fall boxes were periodically aerated. Sampling of overlying water was continued for up to 2 days in the summer and fall cores and for up to 7 days in the case of winter cores. Water samples were obtained by inserting a section of tygon tubing attached to a syringe into a box and withdrawing a sample. This water was immediately filtered through 0.4-pm pore-size Nuclepore filters directly into acid-cleaned sample bottles. The contents of one bottle was acidified with HCI and one bottle was left unacidified and placed immediately into a refrigerator at 4°C. The acidified samples were later analyzed for Fe and Mn (Part 11) and the unacidified samples for NH: and HPOZ-. After water sampling was complete sediment in the flux boxes was sometimes sieved through a 1-mm mesh sieve for the contained animals. This was done at DEEP and FOAM for fall 1975 and winter 1976 samples and at NWC for winter 1976. Regular sampling of fauna at NWC (Rhoads et al., 1977; Yingst and Rhoads, 1978) made sieving of most flux boxes at that station unnecessary. Based on this regular sampling, fall and winter samples should give the maximum ranges found in the abundance and kinds of animals at a given station. 4.2. Analytical Techniques 4.2.1. General. Weight-percent water was determined by drying samples to constant weight at 80°C. No salt con'ection was made. pH and pS or Eh measurements were made in the glove bag by inserting the appropriate electrode directly into one end of the 10-cm squeezing section; electrodes were passed into the bag through a sealed port. The sediment immediately surrounding the electrodes was discarded before squeezing. A rugged, combination glass electrode with Ag-AgC1 reference was standardized at pH 7.414 and 7.0 and used for pH. Platinum (cleaned) and silver-sulfide electrodes were used for Eh and pS measurements relative to the Ag-AgC1 reference of the pH electrode. The sulfide electrode was calibrated as described by Berner (1963). Both pH and pS came to stable readings in the time allowed; Eh did not, but it was not changing rapidly when readings were taken. 4.2.2. Pore Water. The acidified sample for SO:- analysis was diluted with 50-75 ml of distilled water and heated to near boiling. Ten-ml BaSO,

256

ROBERT C. ALLER

(10% w/v) were added, and the resulting precipitate was allowed to cool and digest for at least 24 hr. The precipitate was filtered onto preweighed 25-mm Millipore filters (0.45-~mpore size), dried under a heat lamp, and weighed. The molarity of SO:- in the original sample was then calculated from the appropriate gravimetric factor. Precision of analysis is 2 0.5 mM. Reactive phosphate was determined by a slightly modified molybdenum blue method (Strickland and Parsons, 1968; Presley, 1971). Precision of analysis was better than 3%. Standards were periodically made up and absorbance readings of these standards were the same throughout this study. The phenol-hypochlorite (Indophenol-blue) method was used for ammonia analysis (Solorzano, 1969). Appropriate sample sizes were fixed with phenol within 24 hours of collection (Degobbis, 1973)and stored at all times at 4°C. Analyses were made within 2 weeks of collection. The reagent dilutions given by Presley (1971) were used, but the quantity added was doubled. All reaction and color development was done in polyethylene scintillation vials. Color development was allowed for 6-8 hr to obtain a straight-line standardization curve; blanks (always CO.010 absorbance with respect to distilled water, 1-cmcell) were insignificant compared to sample absorbances and precision better than 3%. Standards were diluted immediately before use. Total alkalinity was determined by titration with standardized 0.1 N HCl (Fisher) using a Gilmont micrometer buret and a microcombination electrode (Gieskes and Rodgers, 1973). The sample was titrated to pH = 5 and allowed to degas until pH drift was minor or absent. At high alkalinities titration to pH -5 was sometimes done more than once. After degassing the sample was titrated to the end point (0.002-ml steps) and the alkalinity determined by curve inflection that was always exceedingly sharp. Precision of analysis was -1% based on overlying water samples. Chloride was done by titration with AgN03 using starch-fluorescein indicator. AgNO, was initially gravimetrically standardized and intercalibration between sample sets checked by titration of standard Long Island Sound sea water that had been sealed in glass ampules (courtesy L. K. Benninger). Ca analyses were made using atomic absorption spectrometry. A K-La buffer was used. Total precision is about 3%. 4.2.3. Solid Phase. Weight loss on ignition of 475°C for 6 hours was measured on sediment samples after drying to constant weight at 80- 100°C. CaCO, was estimated on ashed sediment samples by measuring weight loss after leaching with dilute HCl (e.g., Molnia, 1974).

DIAGENETIC PROCESSES.I.

257

Acid-volatile sulfide was determined on sediment cakes as follows. The frozen sample was thawed slightly and a portion of the center of the cake away from any oxidized exterior taken for analysis. This was weighed, placed in a reaction vessel, and covered with distilled. water. Nitrogen was passed through the chamber to clear out oxygen; gas passed successively through the sample vessel, into a pH = 4 buffer trap, and then into a AgNO, trap. After clearing the line of air, 10 ml of 12 N HCl was slowly added to the sediment sample by use of a syringe mounted on the reaction vessel. .The sample was vigorously agitated throughout by use of a vortex mixer. H2S released from the sample was precipitated as Ag,S in the AgNO, trap. The precipitate was filtered onto preweighed Millipore filters (0.45-km pore size) and weighed. The percent or pmoles/gm acid volatile S in the sample was then calculated. A separate sample of the cake was used to find percent H,O of the sediment and correct the measurements to dry weight. Both Pruden and Bloomfield (1968) and Berner (1974a) have suggested that Fe3+ released from the sample will greatly decrease (up to 50%) the amount of released H2S due to oxidation of H2S to So. Pruden and Bloomfield found no significant effect up to Fe3+concentrations of 50 mg in a sample. They suggest adding 2 gm of SnCl, to samples to prevent this reaction. Unfortunately, before discovering their work, initial experiments with hematite and CdS showed no difference between samples with or without added SnCl,, so it was thought any interference by Fe3+ would be minimal and SnCl, was not used in the natural samples. It is possible that the use of cold reaction conditions and hematite decreases interference because of lower Fe3+ release. In any case SnCl, was not added to the samples and the results may be low. Standard yields were >90% and results precise to within -10%. Total sulfur, expressed as pyrite, was determined on the rinsed and dried samples left from the acid-volatile sulfur runs. A Leco Induction furnace was used. Sulfur-oxide gases released from the burning samples were trapped in 3% H20zsolution, the solution filtered, and SO:- content determined as described previously for pore-water samples. Yields of 84 k 5% were obtained by 12 determinations of natural FeS, and synthetic PbS having 4 different grain sizes; all samples were corrected to 100% yield using this factor. 5 . RESULTS

5.1. Pore- Water Results 5.1.1. Gravity Cores. Sulfate, ammonia, phosphate, and alkalinity profiles over the top meter or more of sediment at each station are shown

258

ROBERT C. ALLER

in Fig. 14 and listed in Appendix B. Note that season of collection may differ between stations (Table 111). The FOAM site has the greatest decrease in sulfate concentration with depth and the expected associated increases in ammonia, phosphate, and alkalinity. (Many additional gravity core data from this station are given in Goldhaber et al., 1977; Rosenfeld, 1977; Martens et al., 1978). Evidence for sulfate reduction and mineralization of organic matter becomes less evident from gravity core porewater data with distance offshore. NWC has a minor decrease in sulfate and regular minor increases in ammonia, phosphate, and alkalinity with depth. DEEP has no evidence of sulfate decrease and no increase in alkalinity, phosphate, or ammonia with depth after an initial increase above seawater values. Both NWC and DEEP have minima in the alkalinity and phosphate profiles that occur at depths of 15-20 cm at NWC and 20-30 cm at DEEP. Ammonia does not show a definite minimum at the same depths, but is nearly constant on either side of the respective depth interval at each station. In addition, at DEEP a slight decrease in concentrations of alkalinity, phosphate, and ammonia below 80 cm occurs. Ca2+ levels are on the average at seawater values except at FOAM ALK. (meq/I)

0.02 Q06 0.1

I

FIG. 14. Pore-water profiles of SO:-, NH: , HPOi-, and alkalinity at FOAM, NWC, and DEEP (gravity cores). Profiles from a given station are arranged horizontally.

259

DIAGENETIC PROCESSES. I. Ca2+(rnM) 4.0

-f Q

60 60 I

6.0

6.0

60

60

60 100

100

FOAM-G

6.0

6.0

60

I20

I20

I

140

140

4.0

NWC-G

DEEP-G

FIG. 15. Pore water Ca2' concentrations as a function of depth at FOAM, NWC, and DEEP (gravity cores).

where there is an unambiguous decrease in dissolved Ca2+ with depth (Fig. 15). This decrease was documented previously by Graber (1975) and Berner et al. (1978) using different analytical techniques. 5.1.2. Box Cores. All stations are characterized by an upper zone of approximately constant sulfate concentrations (Figs. 16- 18; Appendix B). At FOAM this zone varies seasonally, ranging from a few centimeters to greater than 16 cm, and is deepest in the fall. At NWC and DEEP, the zone runs essentially the length of the box cores at all times with a possible exception during summer at DEEP. Summer samples from NWC-1 are irregular owing to analytical problems. A regular pattern of seasonal change in alkalinity, ammonia, and phosphate concentration profiles occurs at all stations (Figs. 19-27). Seasonal changes are most easily delineated at NWC and will be examined first. During the summer the concentration of alkalinity, ammonia, and phosphate at NWC are the highest of the year (Figs. 20, 23, and 26). Concentrations at this time show a tendency to increase below the interface to a maximum and then decrease to a relatively constant or slowly increasing concentration for the remainder of the core length. Fall concentrations are lower at all depths sampled than summer, but remain higher than seawater values even in the top centimeters. The winter profiles show a steady decrease in concentrations as the interface is approached, and in the upper few centimeters become closest to seawater values of any time of the year. There are some minor but important differences between the profiles of the three pore-water constituents at NWC. Structure in the alkalinity profile is smdl because the values are very close to seawater in magnitude, nevertheless the seasonal pattern described is evident. Ammonia most clearly follows the seasonal pattern of change. Concentration max-

260

ROBERT C . ALLER

'

'

O5 10 FOAM-2 Winter- Spring I3 Mar 75

I

FOAM- I Summer 5 Aug 74

1

FOAM-3 Summer

f

10

FOAM- 4 15

FOAM-5

s E 7 5

I

FIG. 16. Seasonal pore-water SO:- profiles from box cores at FOAM; FOAM-1 and FOAM-2 after Goldhaber et al. (1977); overlying water range 21.5-23 mM.

ima are produced in summer cores NWC-1 and NWC-4; these are absent and concentrations uniformly low with depth in fall cores NWC-2 and NWC-5. Convex-upward profiles decreasing smoothly toward the interface occur in winter cores NWC-3 and NWC-6. In all summer and fall cores the ammonia concentrations are about 100 times seawater in the top 0-1 cm interval, giving the impression of a discontinuity between sediment and overlying water concentrations; this discontinuity effect is absent in winter cores. Phosphate profiles follow the general seasonal pattern, but differ from ammonia profiles in that phosphate concentrations decrease regularly near the interface in fall cores and during the winter decrease with a linear or slightly concave-upward profile rather than convex upward, as with ammonia (NWC-2, 3, 5, and 6). FOAM follows the same general time sequence and pattern of change in pore-water concentration profiles as NWC, but with several important differences (Figs. 19, 22, and 25). The high pore-water concentrations at FOAM allow alkalinity to more clearly reflect seasonal patterns. These

DIAGENETIC PROCESSES. I.

01 1

- 5

-5

fa 10 a

1.S . 3 0

I

26 1

2,5

5

I

10 NWC-I Summer

I

NWC- 3 . Winter -Spring 18 Mor 75

I

Y

f

t

a

10

NWC-4

I

-

f

NWC 5

Foll 29 Oct 75

NWC-6 Winter- Spring 23 Mor 76

FIG. 17. Seasonal pore-water SO:- profiles from box cores at NWC; NWC-I scatter results from analytical problems; overlying water range 21 5 2 3 mM.

- 0 10

.’

10

..

25 cL.l

I

I5

DEEP-I Summer 23 July 75

5

15

DEEP- 2 Foll 12 Nov 75

-

DEEP 3 Winter Spring 13 Apr 7 6

-

FIG. 18. Seasonal pore-water SO:- profiles from box cores at DEEP; overlying water range 21.5-23 mM.

262

ROBERT C. ALLER

high values also obscure any summer increases near the interface; no distinct maxima occur in summer FOAM cores, except for phosphate in FOAM-I (ammonia was not measured in FOAM-I). FOAM-3, the second summer core, more closely resembles a winter core than the previous summer core FOAM-1, but it does have the summer characteristic of relatively high values of alkalinity and ammonia in the upper few centimeters compared to later times of the year. The fall core, FOAM-4, has low, constant values in the upper 0-13 cm; a feature typical of fall cores at NWC. Concentrations begin to increase below this zone of relatively constant concentration. Both alkalinity and ammonia have convex-upward concentration profiles during winter (FOAM-2 and FOAM-5) that decrease smoothly to low values near the surface, while phosphate profiles are distinctly concave upward. At DEEP, the highest concentrations of alkalinity, ammonia, and phosphate near the sediment surface occur during the summer (Figs. 21, 24, and 27). These drop to lower values throughout the profile during the fall Alk (mcq/l) 2 4 6 8 101214

-5 f

n

0

2 4 6 8 1 0

15 0

R

10

8

FOAM-2

-

Winter Spring I3 Mar 75

Summer 5 Aug 74

20

2 4 6 8 1012

2 4 6 81012

:r 2 4 6 8 0

10 10

Ill-

0

I5

15

FOAM- 3 Summer

8

JI+

., .

-

FOAM 4

FOAM-5

-

Winter Spring 18 Mar 76

Fall 5 Nov 75

FIG. 19. Seasonal pore-water alkalinity profiles from box cores at FOAM (overlying water -1.8-1.95 meq/Iiter).

263

DIAGENETIC PROCESSES. I. Alk (meq/l) 1.0

3.0

5.0

F0 S

fa 10

NWC-3 Winter Spring 18 Mor 75

-

15 NWC-I Summer 21 July 74 1.0

3.0

NWC-2 Fall II Nov 74 5.0

1.0

3.0

f n NWC-6 Wlnter- Spring 23 Mar 76

15

NWC-4 Summer

NWC-5 Fall 29 Oct 75

16 July 75

FIG.20. Seasonal pore-water alkalinity profiles from box cores at NWC (overlying water -1.8-1.95 meq/liter).

Alk (rnea / I 1

DEEP-I

DEEP-2

DEEP- 3

Summer 23 July 7 5

Foll 12 Nov 75

Winter Spring I3 Apr 76

-

FIG.21. Seasonal pore-water alkalinity profiles from box cores at DEEP (overlying water

- 1.8-1.95 meqlliter).

264

ROBERT C . ALLER

FOAM-2 Winter- Spring 13 Mar 75

t

FOAM-3 Summer 8 July 7 5

L

t

FOAM-5 Winter-Spring 18 Mar 76

FOAM-4 Fall 5 Nov 75

FIG.22. Seasonal pore-water ammonia profiles from box cores at FOAM (overlying water 5 2 WM).

and remain low near the interface in the winter, but increase in the deeper sections of the profile at that time. Like NWC, alkalinity is near seawater values at DEEP. The discontinuity between ammonia concentration in the upper few centimeters at DEEP is not as great as either NWC or FOAM. Phosphate profiles, like the FOAM site, are concave upward during winter and at this station also in fall. At all three stations Ca2+increases during the winter above seawater Ca2+/Cl- ratios near the interface (Fig. 28). This is particularly evident at FOAM where a distinct Ca2+maximum is found. At other times of the year Ca2' is not distinguishable from seawater background levels at any station, except possibly NWC. 5.2. Flux Measurements

No significant change in concentration of NH; and HP0:- in the flux boxes without sediment (blanks) relative to flux boxes with sediment were

265

DIAGENETIC PROCESSES. I. NH: ( 0

p ~ 100 200 300 400

O 5

I

T

5

0 5

f 10

n I5

‘I

NWC- 3 Wtnler, 18 March

15

75

NWC-I NWC-2

Summer, 21 July 74

Fall, I I November 74

o 5

o

r

o

_ / / / / I 0

5 10

NWC-6

15

NWC-5

NWC-4

Winter, 23 March 76

Fall, 2 9 October 75

Summer, 16 July 75

FIG.23. Seasonal pore-water ammonia profiles from box cores at NWC (overlying water 52 pM).

found at any time. Concentration increases were also found to be approximately linear with time. Fluxes were therefore calculated assuming no consumption term in the overlying water as follows. The concentration of either ammonia or phosphate at time of first measurement in the blank box C , was determined. The volume of water overlying the sediment in the sediment flux box at a given time V,, and the concentrations C , of ammonia or phosphate in that water at the same time were also measured.

;r

f a

0.1

0.2

03

10

15

15

DEEP-I Summer 23 July 75

DEEP-2 Fall

12 NOV 75

DEEP-3 Winter - Spring 13 Apr 76

FIG.24. Seasonal pore-water ammonia profiles from box cores at DEEP (overlying water 5 2 FM).

266

ROBERT C. ALLER HPO4 = (mM) 0.1

0.2 0.3

\

FOAM-2 Winter

2o

i I'

13 Mor 75

FOAM -I Summer 5 Aup 7 5

0.1

0.2

0.3

f ," 10 0

FOAM- 3 15

Summer 8 July 75

FOAM-5 15 5NOv 7 5

18 Mor 76

FIG. 25. Seasonal pore-water phosphate profiles from box cores at FOAM (overlying water 5 2 w M ) .

The quantity V,(C, - C,) was then plotted versus time t, a line fit to the data by least-square methods, and the amount of material entering the water per unit time determined as the slope. This number was divided by the surface area sampled by the flux box to calculate the flux of material out of the sediment per unit area per time. The error is taken as the standard deviation of the slope and is about ? 10%. The actual uncertainty must be higher, as will become clear later (Nixon et al., 1980). The last sample points (-30-50 hr) in both the summer and fall sets of flux cores were not used in the calculations. Inclusion of these points often, but not always, resulted in calculation of a higher NH,' (or Mn2+, see Part 11) flux. (In some cases a lower flux is calculated.) This suggested that lower 0, levels were beginning to cause changes in pore-water profiles and perhaps reaction distribution in the sediment. For consistency in calculation none of these later sample points were used. These points were arbitrarily included or discarded in previous calculations (Aller, 1977); the present

267

DIAGENETIC PROCESSES.I.

0.1

(I'

0.2 0.3

'-

1

' A

I/'

NWC-I

NWC- 3 Winter -Spring 18 Mar 75

21 July 74

0.1

i'\,

0.2

0.3 1

NWC-6 Winter Spring 23 Mar 76

NWC-5

.I

16 July 7 5

-

29 OCt 75

FIG.26. Seasonal pore-water' profiles from box cores at NWC (overlying water 5 2 p M ) .

f

%

n

,y 10

10

10

15

15

15

268

ROBERT C. ALLER Ca-CI x 10 Y)

3

18 B 20 21 2 2 2 3

19 20 2t 22 23

I 8 R 20 21 22 23

FIG.28. Pore-water CdCI molar ratios during fall (solid lines) and winter (dashed lines) at FOAM, NWC, and DEEP.

values are believed to be the most acceptable. Usually three points are used in calculation of a line. In general, ammonia followed the straight-line fit rather well. Phosphate did not follow a linear increase in all cases; when iron flux was high or iron precipitation apparently took place in the box (Part 11), phosphate levels showed little or negative change. Tabulations of the flux data are in Appendix C. Plots of the NWC measurements are shown in Fig. 29 to illustrate the trends. The calculated fluxes for each station during the year are listed in Table VII together with the temperature at which they were determined. The highest fluxes were found in the summer boxes and the lowest in the winter. Winter fluxes are an order of magnitude lower than those found in summer or fall. 280

'Summer

i

g 160 0

-5

120

+v I

80

z

->r-I 30 0

g

3

40

I

N*

20 40 60 80 100 120 Time (hr)

BI

20 10

10 30 50 70 90 110 Time (hr)

FIG.29. Plot of NH: and HP0:- released versus time in flux cores from NWC. 0, summer; 0, fall; A, winter. Slopes of lines drawn (wmoles/hr) correspond to flux values given in Table VII.

269

DIAGENETIC PROCESSES. I.

5.3. Solid-Phase Analyses 5.3.1. Zgnition Loss and Organic Matter. Ignition-loss measurements for the three stations are plotted in Fig. 30. FOAM is lowest in percent ignition loss even when corrected on a carbonate-free basis. DEEP is next with values between 5-6%, with most about 6%. Neither FOAM nor DEEP have much depth dependence in ignition-loss profiles. NWC has the highest values near the interface of any of the stations. The profiles at NWC also show depth dependence with decreasing values at deeper levels in the sediment. Ignition loss is proportional to organic matter content, but may not be readily converted to percent organic matter values in fine-grained sediment owing to water loss from hydrated particles. Comparison of ignition-loss values with measurements made using a Leco % Ignition

Loss

'p 5

10

NWC - 2

NWC - I 15

I

I

NWC-4

NWC - 3

15

2 4 6 8

0

f a u

10

a

DEEP

- I 15

-I

15

20

FIG.30. Ignition-loss profiles from selected cores from FOAM (Goldhaber ct al., 19771,

NWC,and DEEP.

270

ROBERT C. ALLER

SUMMER S (a.v.,pmoles/grnl

S (a.v..pmo~es/grn)

S (a.v., prnoles/gm)

a 0 10 15

DEEP-I 20 FOAM-I

FIG.31. Comparison of acid-volatile sulfide (FeS) profiles at FOAM, NWC, and DEEP.

*

*

carbon analyzer indicate that 2.4 0.7% and 2.3 0.4% weight loss on ignition at NWC and DEEP, respectively, are due to sources other than organic matter (AIler and Yingst, 1980). A direct conversion factor for Long Island Sound sediments of absolute ignition loss values to carbon as determined by wet oxidation or Leco methods lies between 3-3.5% (Rosenfeld, 1977; Aller and Yingst, 1980). 5.3.2. Sulfur. Acid-volatile sulfide profiles at each station all show maxima 4-6 cm below the interface then decrease to background values at depth (Fig. 31). DEEP has the largest maximum of any of the stations at about 30 kmoles/gm. Three cores were analyzed at NWC. All three have maximum acid-volatile sulfide at 4 cm, and appear to decrease in a similar way at depth (Fig. 32). The main differences between the three cores is in the upper 3-cm samples. The winter core contained no acid-volatile sulfide in the 0-1-cm sample; the silver-nitrate trap did not show S ( a v ,pnoles/gm)

S fa v , pmoles/gm

I

S ( a v.,pmoles/gm)

9 5

0

N W C - 2 Fall II Nov 74 N W C - 3 Winter 18 Mar 75

N W C - 4 Summer 16 July 75

FIG.32. Seasonal variation in acid-volatile sulfide profiles at NWC. Lowest values near interface occur during winter.

27 1

DIAGENETIC PROCESSES.I. S (pyrite, ,moles/grn)

S (pyrite,pmoies/gm) 0

100

S (pyrite,,moles/grn)

500

300

T

2 5O

O

f

fn

0 10

n

o5

T

n

NWC-3

0

K

10

Wmter 75

201

DEEP-I Summer 7 5

I FOAM-I Summer 74

FIG. 33. Total nonvolatile sulfur (FeS2 + So) profiles at FOAM, NWC, and DEEP.

even a hint of discoloration. The 1-2-cm sample from the winter core is also lower than that found in either of the others. The fall core has the highest near-interface values of acid-volatile sulfide of any time of the year. Nonvolatile (pyrite) sulfur profiles differ at the three stations. FOAM (after Goldhaber et al., 1977) has the highest values of pyrite sulfur, followed by NWC, and the lowest values at DEEP (Figure 33). At all three stations pyrite increases away from the interface and, at all three, measurable levels are found at the interface. Pyrite sulfur shows little evidence of seasonal change as determined at NWC, although there is a possibility of a slight decrease near the interface in fall and winter relative to summer (Fig. 34). Profiles over the upper meter of sediment from the offshore

SO

100

150

200

-i-

loo

Y 50

I

I

8 6

NWC-2

Fall

II Nov 74

NWC-3 Winter 18 Mar 75

200

4

I I'

NWC- 4

Summer 16 July 75

FIG. 34. Seasonal profiles of total nonvolatile sulfur (FeS2 + So) at NWC. A possible minor decrease in interface sulfur during fall and winter is detectable.

272

ROBERT C. ALLER

central Sound show little increase in pyrite sulfur after the top 10 cm (Berner, 1970), so these data can be considered representative of the fixed sulfur concentrations and profiles at each station.

6. DISCUSSION

The pore-water, solid-phase, and flux data indicate that there are many similarities between stations both in the general type of diagenetic reactions taking place and in the sequence of change in sediment properties through time. There are major differences as well, particularly in the way in which products of the anaerobic decomposition of organic matter: HS- , HPOZ-, HCO;, and NH:, build up in the sediment and the magnitude of release of these products into overlying water. These similarities and differences reflect the particular balances between transport and reaction processes which occur in the sediments in each region. By considering the relative contribution of physical, chemical, and biological influences on transport and reaction at different times of the year and in different bottom areas, it is possible to explain the temporal and spatial patterns observed. 6.1. Decomposition Reactions

The major commonly accepted pathways for the biogenic decomposition of organic matter in marine sediments are listed in Table IV (after Richards, 1965; Stumm and Morgan, 1970; Claypool and Kaplan, 1974). The overall stoichiometries of the reactions tend toward mean values determined by the average composition of marine organic matter (Redfield, 1934). In the present case, relative stoichiometries have been left as the variables x, y, and z because preferential breakdown of subcomponents of the average organic material can occur at particular stages during decomposition (Krause, 1959; Grill and Richards, 1964; Sholkovitz, 1973). Of the reactions listed, only methanogenesis (reaction 6) is likely to be inhibited in the upper few decimeters of sediments at the stations studied. Methane production takes place in Long Island Sound deposits, but, as observed in other environments, can be expected to be most rapid in regions or microenvironments within a deposit where active sulfate reduction (reaction 5 ) is not taking place (Martens and Berner, 1974, 1977; Oremland and Taylor, 1978). Decomposition is predominantly anaerobic below the upper few centimeters of sediment at FOAM, NWC, and DEEP. This is shown by the

TABLEIV. IDEALIZED DECOMPOSITION REACTIONS Aerobic respiration: (CHzO),(NH3),,(W'04), + (x + 2~102 + xCOZ + (X + y)HZO + yHNO3 + ZH3PO4 Nitrate reduction:

5(CHzO),(NH,),(H,PO,), + x C O ~+ 3XH20

Manganese reduction:

Iron reduction:

+ 4xNO;

+ 4xHCO; + 2rNz + 5yNH3 + 5zH3PO4 (CHZO),(NH3),(H3P04), + 4xMnOOH + 7xc02 + xH2O + 8xHCO; + 4xMn" + yNH3 + zH3P04

+ )7XCoz ~ (CH20),(NH3),(H,P04 )z + ~ x F ~ ( O H + 8xHCO; + 3xH20 + 4xFe" + yNH3 + zH3P04 (CH,O),(NH, )v(H3P04) z + xS0:+ 2UHCO; + xH2S + 2yNH3

+ 2ZH3P04

Methane production: 2(CHzO),(NH3 ),(H3PO4), + xCOZ + xCH4 + 2yNH3

+ 22H3P04

Sulfate reduction:

Fermentation (generalized): 12(CHzO),(NH3),(H3P04)L + xCH~CH~COOH + xCH3COOH + 2xCH3CH2OH + 3XCO2 + xH2 + 12yNH3 + 12zH3PO4

274

ROBERT C. ALLER

sharp attenuation of O2beneath the sediment-water interface as reflected in low Eh values and the high concentrations of Fez+ and Mn2+ in the same region (Appendix B; Part 11, this volume). Decreases in pore-water NO; concentrations in the upper 1-3 cm also demonstrate the occurrence of denitrification near the interface at all sites (Rosenfeld, 1977). However, because of the low concentration of NO; in overlying water (1- 15 p M ) and the restriction of denitrification predominately to the top of 0-1 cm, this reaction should not greatly influence the buildup of decomposition products in pore water. One of the most important anaerobic pathways of decomposition in marine sediments is sulfate reduction (Berner, 1964; Goldhaber and KapIan, 1974; Jergenson, 1977). Proof that sulfate reduction is taking place in surface sediments at each station of this study comes from the abundance of fixed sulfur in the solid phase and the presence in the pore waters of dissolved sulfide (Figs. 31-34; Appendix B; Goldhaber et al., 1977). Because sulfate reduction presumably dominates the anaerobic decomposition reactions over most of the sampled sediment regions, reaction 5 of Table IV will be assumed as the major model reaction to aid in the interpretation of pore-water and solid-phase property distributions. Additional decomposition reactions involving sulfide oxidation, specific interaction with Fe and Mn oxides, and fermentation (Presley and Kaplan, 1968) occur, but will not be emphasized here.

6.2. Supply and Reactivity of Organic Material Of the three environments, the near-shore station FOAM shows the greatest evidence of sulfate reduction as expressed by depletion of sulfate pore-water concentrations and the production of high concentrations of alkalinity, phosphate, and ammonia (Figs. 14, 19, 22, 25). Evidence for sulfate depletion and mineralization of C, N , and P as expressed in porewater profiles (Fig. 14) decreases with distance from the Connecticut shore and with overlying water depth. This trend could quite reasonably be due to a change in the supply of organic material to bottom sediments, either quantity or quality, as a function of depositional environment. However, ignition-loss measurements (Fig. 30) and the corresponding organic-carbon values (Rosenfeld, 1977; Aller and Yingst, 1980) demonstrate that the quantity or content of organic matter at each station in any given sediment-depth interval does not greatly differ and if anything is lowest at FOAM where evidence of sulfate reduction is greatest. The susceptibility or quality of the organic matter present at each station for use by sulfate reducers can be compared by considering the rates of

275

DIAGENETIC PROCESSES. I.

sulfate reduction obtained in incubation experiments by Goldhaber et al. (1977) and Aller and Yingst (1980) for each station as a relative assay for metabolizable organic matter (Fig. 35). Since the depth distribution of particle-associated metabolizable organic matter is determined by particle-transport processes at each station (as discussed later), the average value for sulfate reduction rate over the top 10 cm rather than the surface value is most appropriately compared. These values are obtained by integrating the functions describing the rate distributions at each site (22°C): FOAM, R , = 137 exp( - 0 . 7 7 4 + 11 (Goldhaber et al., 1977; Rosenfeld, 1977); NWC, R , = 118 exp(-0.34x) + 5; and DEEP, R , = 27.3 + 26.6 exp( - 0 . 7 1 ~ )mMlyr (Aller and Yingst, 1980); giving average rates of R = 29,39, and 29 mM/yr, respectively, over the top 10 cm at each station, These are very similar within error, suggesting that the type of organic matter, in the sense of its ability to support sulfate reduction, is not greatly different in different regions of the central Sound. These considerations lead to the conclusion that to a first approximation the quantity and quality (with respect to sulfate reducers) of organic matter at each of the three stations is nearly the same. The rate of supply of organic material also cannot differ much if it is assumed that a constant proportion of the flux of organics hitting the interface at each station must remain as refractory residue and therefore is reflected in the standing crop. Taken at face value, these data indicate that FOAM, which has the greatest buildup of decomposition products in the pore water, is not necessarily the most productive of the three stations. Additional evidence for relative reaction rates at the three stations comes from solid-phase properties as discussed subsequently. S O$mM/yr

S O f rnM/yr 20

40

60

80

20 40

100

60

0 -

2 ’

,,’

22°C

4 . 6 .

#

t/ ; 22°C

+li I

8 ’ c

a 0)

10-

j

1

FOAM

10

tti

NWC

10 -

it

DEEP

‘7i 14

FIG.35. The rate of SO:- reduction obtained from decomposition experiments as a function of depth in sediment. Data from FOAM after Goldhaber et a/. (1977) and Rosenfeld (1977); NWC and DEEP after Aller and Yingst (1980).

276

ROBERT C. ALLER

6.3. Products of Decomposition-Solid

Phase

The sulfide produced during sulfate reduction reacts rapidly with ironrich particles to form the acid-volatile sulfides mackinawite and greigite (Berner, 1967; Rickard, 1974). These labile sulfides represent transient intermediates which, if conditions are appropriate, commonly convert through several possible mechanisms to pyrite (Berner, 1970; Sweeney and Kaplan, 1973; Rickard, 1969, 1975). Acid-volatile sulfides (FeS) readily oxidize back to Fe2+and SO:- if exposed to oxygen (e.g., Sato, 1960; Bloomfield, 1972), while pyrite oxidizes much more slowly in seawater (Harmsen, 1954), but is also lost at a measurable rate (Part 11). The acid-volatile sulfide profiles that result from these reactions provide an added check on the assertion that rates of sulfate reduction are not greatly different at the three stations and if anything are higher in deeper water. Profiles show that at each station: (1) the shape of FeS depth distributions is similar, (2) the maximum value of FeS attained at a depth of 4-8 cm is of comparable magnitude, and (3) the total quantity of FeS in the upper 10 cm is nearly the same, with the highest inventory at DEEP. Because the quantity of reactive solid-phase Fe (Berner, 1970) is not greatly different at these stations (Part 11), it is unlikely that relative reactivity of the sediment and thus formation rates of FeS in the presence of free HS- differ at the three sites. It is also unlikely that conversion rates of FeS to FeS, differ substantially. Rickard (1975) demonstrated that the initial rate of pyrite formation in aqueous solution at low temperatures has second-order dependence on FeS and first-order dependence on So and H2Sconcentrations. This means that FOAM could have higher rates of conversion than NWC or DEEP below 8-cm depth but that significant differences between stations probably do not exist closer to the interface (see pS data in Appendix B; H,S data of Goldhaber et al., 1977, at FOAM 1). Taken together these observations indicate that the quantity and form of the FeS profiles are similar because the rate of HS- production is similar at each site. In contrast to FeS, there are obvious differences in total or pyrite sulfur distributions in the shallow-to-deepwater transect. Pyrite sulfur is present at the sediment-water interface at all stations as found in previous studies (Harmsen, 1954; Berner, 1970; Sweeney and Kaplan, 1973; Goldhaber and Kaplan, 1974). In addition, the basic form of the FeS, distribution at each station is similar: a zone of relatively low, constant FeS, values overlies a zone of relatively high, constant values. However, the quantity of pyrite present is not the same. DEEP, which has the highest inventory of acid-volatile sulfide, has the lowest pyrite content.

DIAGENETIC PROCESSES. I.

277

The discrepancy between the relative rates of sulfate reduction, FeS inventories, and the resulting quantity of pyrite formed at each station may be due to several processes acting alone or in concert: (1) the rate of conversion of FeS to FeS, may differ, although this was largely discounted previously; (2) FeS may be reoxidized prior to pyrite formation; or (3) FeSz may be oxidized after formation. Evidence that oxidation of solid-phase sulfides takes place continually can be found in the shapes of the profiles and in seasonal changes in profiles (Figs. 33 and 34). The strongest evidence for continual oxidation is the presence of measurable solid-phase concentration gradients in the upper few centimeters. This region is rapidly reworked by protobranch bivalves during feeding (Aller and Cochran, 1976). The maintenance of a gradient in the presence of such mixing requires continuous formation and loss of solid-phase sulfide. The exact form of the gradient at any given time represents a balance between formation, reworking, and oxidation. Lower gradients for FeS, than FeS near the interface are consistent with a much slower oxidation rate for FeS, (Harmsen, 1954; Bloomfield, 1972). Somewhat higher FeS concentrations in the top 0-1 cm at FOAM may result from partial protection from oxidation by enclosure of sediment in the abundant shell debris at that station. The lowest FeS concentrations are found near the interface during the winter cold-water period (Fig. 33). This may also be true for FeS, (Fig. 34). The decrease presumably results from lowered biogenic formation of sulfide during this time, due to lower temperature coupled with continued loss of solid-phase sulfide through oxidation. An estimate of the actual loss of sulfide produced in the upper 10 cm at each station can be calculated as follows. Sulfide production is a function of temperature and has an apparent activation energy of about 19 It 2 kcal/mole (Vosjan, 1974; Goldhaber et al., 1977; Jgrgenson, 1977; Aller and Yingst, 1980). The temperature T of water in the Sound is approximately a sinusoidal function of time t with a maximum and minimum of 22" and 2°C; this variation can be described analytically on the absolute temperature scale as T = T o - T , sin wt with w = 2dyr, T o = 285, and T , = 10. The Arrhenius rate law describing the approximate temperature dependence of the rate of sulfate reduction can then be converted to a time-dependent function A (t), by replacement of temperature with the previous expression for T ( t ) . Since temperature dependence is not a function of depth in sediment this means R , , the rate of sulfate reduction is given by (6.1)

R,(x, t )

=

R , ( x ) A ( t )= R , ( x ) A oexp( - E / k T )

where R ( x ) is the rate at depth x at 22°C (from Fig. 31), A . is the pre-ex-

278

ROBERT C. ALLER

ponential factor of A ( t ) adjusted to set A ( t ) = 1 at 22°C; the remainder of A ( t ) is determined by the apparent activation energy of sulfate reduction E, the constant k = 1.99 callday-mole, and temperature (time of year). The yearly average of A ( t ) is -0.4 by numerical evaluation, so the yearly average R , is 0.4 R , (R, at 22°C). The total sulfur S, fixed at a depth of 10 cm assuming a steady state R , ( x ) distribution is given by

I,

10

(6.2)

ST =

(0.410)

R , ( x )dx

where o is the sedimentation rate and R,(x) is the 22" rate distribution. This assumes that the depth interval considered represents an appropriately long span of time so that an average A ( t ) can be used. Take as an approximation o = 0.5 cm/yr at NWC and DEEP and o = 0.3 at FOAM (these are the *'OPb apparent sedimentation rates found at FOAM, NWC, DEEP, and an additional offshore station by Thomson et al., 1975; Benninger et al., 1979; E. Dion, personal communication;although not a true sedimentation rate because of reworking, this o will be used as the apparent advective transport term for reactive organic matter in the model), then ST = 387, 312, and 232 mM S at FOAM, NWC, and DEEP. The average water content is about 45%, 55%, and 55% at FOAM, NWC, and DEEP giving an expected STof 317,381, and 284 kmoles Slgm-sediment at a depth of 10 cm at each. This corresponds to an actual measured ST of 278, 152, and 137 pmoles/gm at 10 cm at the three stations. Almost all, 88%, of the sulfide produced at FOAM can be accounted for by this model while only 40-48% of the sulfide produced at NWC and DEEP can. Goldhaber et al. (1977) had previously calculated, using the same method but without temperature correction, that only 43% of the sulfide produced at FOAM was retained; the difference arises because of the previous lack of temperature correction and a slightly different form for R,. Although there is considerable uncertainty in the exact values obtained, the important conclusion from this calculation is that a portion of the sulfide produced at each station is lost and that this loss is greater at the offshore stations than at FOAM. The similar HS-production rates at each station as measured experimentally and the similar standing crops of FeS indicate that most loss probably, but not necessarily, takes place after sulfide has reacted with the solid phase rather than before. In contrast, Jgrgenson (1977) argues for -90% loss by oxidation of HS- at the dissolved stage in sediments from Limfjorden, Denmark. To this point, analysis of the data has demonstrated that the average rates of sulfate reduction of sediment in the upper 10 cm are comparable in different areas of the Sound, but that the ultimate fate of at least one of the products of reduction, HS-, differs from one region to the next.

DIAGENETIC PROCESSES.I.

279

Because the reactivity of the sediment toward sulfide and subsequent conversion of solid-phase sulfide is probably not significantly different between stations, these findings suggest that transport rather than reaction processes are predominately responsible for the observed variations in the standing crops of different forms of solid-phase sulfur and other products of decomposition in Long Island Sound (LIS) sediment. Physical and biological transport processes, which determine in part the distribution and abundance of the products or reactants of decomposition within a deposit, will now be considered. 6.4. Transport Processes 6.4.1. Physically Controlled Transport. The similarity in net sedimentation rates and sediment mass properties of FOAM, NWC, and DEEP make it unlikely that physical transport factors such as compaction (Berner, 1971; Lerman, 1976) or resistivity of a deposit to molecular diffusion (Manheim, 1970; Manheim and Waterman, 1974) differ greatly between these stations. There is also no evidence of groundwater movement in the upper meter of sediment at any of the three sites. In contrast, physical mixing of sediment by current and wave action is much more important at shallow than deeper water stations. As shown by x-radiography studies, short-term storm events such as hurricane Belle (1976) are capable of resuspending at least the upper 2-3 cm of sediment in depths of water of 10 m or less. This ability attenuates rapidly offshore and at depths of 15 m only the upper -0.5 cm is influenced (McCall, 1978). During longer Northeastern storms, whose wind fetch is down the axis of the Sound, the upper 2-3 cm of sediment at NWC depths (-15 m) can sometimes be mobilized (Aller and Cochran, 1976), but it is unlikely that sediment at DEEP (34 m) is ever directly influenced by wave activity. These physical mixing events are most important at FOAM and least important at DEEP; they are capable of directly affecting only the upper few centimeters of sediment and are relatively random single events. They cannot account for regular ongoing transport processes or differences between stations other than by indirectly influencing other types of sediment transport such as biogenic reworking as discussed next. 6.4.2. Biogenic Transport Process: Particle Reworking. Unlike most physical transport processes, there are significant variations between stations in the biological transport of sedimentary particles and fluids. There are three independent lines of evidence that biogenic particle reworking is different at FOAM, NWC, and DEEP: sedimentological, radiochemi-

280

ROBERT C. ALLER

cal, and biological. X radiographs (Figs. 4 , 5 , 7 , 8 , 10, and 11) demonstrate the presence of biogenic structures at all stations but preserved, physically formed laminations only at FOAM. Preservation of this kind occurs only in physically disturbed environments where: (1) animals cannot establish themselves long enough to thoroughly rework sediment (Schafer, 1962; Howard, 1975); (2) animals are in low abundance for other reasons, for example, lack of food and oxygen (Calvert, 1964); or (3) sedimentation rates are extremely high (Moore and Scruton, 1957). Sedimentationrates, food source, and water characteristics do not differ significantly between stations, while, as discussed earlier, physical disturbance does. McCall (1977) suggested that in shallow-water regions of LIS, physical disturbance maintained benthic communities in a constant state of recolonization due to increased mortality. This results in a high, transient diversity of small animals having short life spans. Many of these animals are restricted in their activities to near the sediment-water interface. The resulting short residence time of individual populations at FOAM allows the partial preservation of physical structures produced by major storms. For example, laminations at a depth of -12 cm underwent little disturbance over the 2-yr sampling period at FOAM (Figs. 4 and 5). In contrast, x radiographs of sediment structures at NWC and DEEP demonstrate a dominance of bioturbate structure at both with a greater importance of homogeneous reworking relative to dwelling structures at NWC than DEEP (Figs. 7, 8, 10, and 11). The distribution of the particle-associated radionuclides 234Thand *"Pb in surface sediments document differences in the rate and type of particleredistribution processes at each station. 234Th( t = 24. Id) profiles show that the upper few centimeters at all stations are relatively rapidly mixed with fastest reworking at NWC and slowest at DEEP (Aller et al., 1980). A range of particle-diffusion coefficients of 0.1-1.0 x cm2/sec is found. Z1oPb( t = 22 yr) distributions demonstrate that on "'Pb decay time scales, homogeneous particle reworking is greatest at DEEP with excess 210Pbdecreasing smoothly to a depth of about 40 cm (Benninger et al., in preparation). Smoothly decreasing decay curves extend to only 15-20 cm at NWC and FOAM (Benninger et a[., 1979; E. Dion personal communication). In addition, excess 210Pbis found to a depth of -30 cm at a station situated between NWC and DEEP (Thomson et al., 1975). Modeling of these distributions indicates that homogeneous particle reworking on "OPb time scales increases offshore in the inshore-offshore transect. Zones of irregular reworking on 'I0Pb time scales extend to depths of > 1 m and are due to infilling of solitary crustacean burrows rather than homogeneous mixing (Benninger et al., 1979). The sedimentological and radiochemical evidence for trends in the type and extent of particle reworking in the central Sound are consistent with

DIAGENETIC PROCESSES. I.

28 1

the kinds of animals present in each case. Deep-burrowing, sedentary, long-lived, surface deposit, or suspension-feeding benthos increase in abundance offshore in deeper water (McCall, 1977; Figs. 3, 6, and 9). These animals form permanent or temperorary dwelling structures in which they reside for considerable periods of time. One faunal component present at DEEP, the maldanid polychaetes, feed at depths of 15-30 cm and pass particles vertically to the interface (Rhoads, 1974). Most other organisms at DEEP are surface-deposit or suspension feeders and their chief effect on particle transport at depth will be due to burrow construction and the subsequent filling in of burrows when they are vacated rather than particle reworking during feeding. For the case of teribellid or other tube-building polychaetes present at DEEP, for example Pista, burrow formation can result in the construction of biogenic shafts lined with surface sediment rich in organic matter (Aller and Yingst, 1978). These kinds of particle reworking activities are relatively slow on monthly time scales (234Th),but are significant when integrated over many generations (2'oPb time scales). Animals at NWC are generally more mobile and shallow burrowing than at DEEP. The characteristic species are protobranch bivalves that intensively rework the upper 3-4 cm of sediment (Rhoads, 1963; Young, 1971) and are largely responsible for the differences in 234Thprofiles at NWC and DEEP (Aller and Cochran, 1976; Aller, 1977). Maximum burrowing depth of most common benthos such as Nephtys at NWC is 15-20 cm although Spiochaetopterus builds slender burrows to -25 cm. The number of active or open burrows decreases rapidly below the interface. At both NWC and DEEP, large crustacean burrows are found to depths greater than 1 m, but as shown by 2'0Pbprofiles these are not especially important for particle reworking over the period in which data were gathered for this study. As described in Section 2.1, FOAM is characterized by small, shortlived animals representing many species living in the Sound. The most common and predictably present species is the surface-deposit-feeding amphipod Ampelisca sp. that restricts tube building and feeding activity to the upper 0-2 cm. Maldanid polychaetes which were not present during the first year of sampling at FOAM (Goldhaber et al., 1977) appeared in fair abundance in Fall 1975. Despite this, sediment remained laminated at depths >10 cm indicating no extensive, rapid particle reworking by these animals deep in the sediment column. Just as shells protect FeS from oxidation near the interface, enclosure of sediment by shells aids in protecting particles from reworking by animals at FOAM. Physical processes at FOAM apparently have two indirect roles in reducing biogenic particle reworking: disturbance decreases the residence time of given animal populations and therefore

282

ROBERT C. ALLER

efficiency and depth of reworking, and shell accumulations brought about by the current regime hinder homgeneous reworking by the species normally present in Long Island Sound. 6.4.3. Biogenic Fluid Transport. Irrigation of burrows or body cavities for purposes of respiration, excretion, or reproduction results in the biogenic transport of seawater into and out of marine sediment. Compilations of fluid- and particle-transport rates by various kinds of animals show that the weight ratio of fluid-to-solid transport rates by deposit feeders is 100-500 (Aller, 1977, p. 26). Suspension feeders have much higher ratios: -10’-106, but in many cases fluid pumped by suspension feeders does not intimately contact sediment. Irrigation activity is capable of maintaining water within burrows at near seawater concentrations for many ions despite high fluxes of solutes from surrounding sediment (Aller and Yingst, 1978). Extensive exchange of pore water with overlying water can take place across tube walls even in the absence of particle reworking (Aller, 1978). This exchange is brought about by a combination of horizontal diffusion into the burrow core and subsequent vertical advection, by irrigation, out of the sediment. Because of burrow formation and ventilation, the distance a dissolved ion must diffuse to escape sediment is approximately half the distance between burrows. The quantitative effect of this change in diffusion geometry is developed in detail in Section 6.7. From the previous considerations on particle reworking and the types of animals present at each station, it is clear that biogenic fluid transport will vary in intensity and depth of influence in the Sound. Based on density and depth of burrowing the deepest exchange by irrigation will occur at DEEP and the shallowest at FOAM. It is important to note that fluid and particle transport need not be coupled in the same way at all stations because of the dependence of these transport mechanisms on animal type and the geometry of exchange. Because fluid exchange is effectively much more rapid than particle mixing, irrigation is capable of affecting porewater profiles in a short period of time without disturbing physically formed sediment laminations such as occur at FOAM.

-

6.5. Seasonal and Spatial Variation in Pore- Water Profiles-General Patterns

The transport reaction conditions outlined for each station can be used to interpret some of the general seasonal and spatial variations in porewater profiles, The basic seasonal pattern at all stations is as follows:

283

DIAGENETIC PROCESSES.I.

production (NH: , HPO:-, HCO; , HS-) and consumption (SO:-) of interstitial nutrients predominates during initial warming of overlying water during spring and summer. This causes high standing concentrations of NH: , HPO:-, and HCO; in the pore water and increases or replenishes the standing crop of FeS near the sediment-water interface after winter lowering by oxidation. Increases in NH: and HPO:-, which are most easily detected because of the relatively low background concentrations, mirror the depth dependence of production predicted by the decomposition experiments of Goldhaber et al. (1977) and Aller and Yingst (1980) at each station (Figs. 35 and 36). Concentrations at FOAM (FOAM-1) and NWC (NWC-1) increase exponentially toward the interface before decreasing to seawater values; concentrations at DEEP (DEEP-1) increase much less sharply as expected from the lower depth dependence of decomposition at that station (Figs. 22-27). The intensity of this early summer increase was not exactly the same in the two sampling years at NWC and FOAM. This may result from slight offsets in the sampling dates from the actual sequence of change in the sediment or may reflect a longer term cycle in sediment chemistry since a general lowering of concentrations occurred at both NWC and FOAM in the second year of sampling. Additional explanations related to patchiness in biogenic transport are developed quantitatively in Section 6.7. With regard to long-term patterns, Riley (1956b) noted a difference in the intensity of the spring flowering of phytoplankton in Long Island Sound during the years of 1952-1954. This type of change may result in

ti

9 t 4-

FIG.36. Rate of NH: production [(RN/(l + K ) ] at 22” and 4°C obtained from decomposition experiments as a function of depth in sediment. Data uncorrected for adsorption [FOAM after Rosenfeld (1977); NWC and DEEP after Aller and Yingst (1980)l. The curves are given by: FOAM, RN/(1 + K ) = 32.5 exp(-0.73x) + 0.86; NWC, Rd(1 + K) = 28.1 exp(-0.55x) + 0.57; DEEP, RN/(1 + K ) = 6.8 exp(-0.61x) + 1.6 mM/yr.

284

ROBERT C. ALLER

multiyear patterns of supply of decomposable organic material to bottom sediments and influence pore-water compositions near the interface accordingly. These differences in supply would not affect the relative temporal pattern of change in pore-water profiles but might alter the magnitude of excursions in concentration levels from one season to the next. During the late summer and fall (August, September, October) when water temperature peaks and then begins to decrease, biogenic reworking reaches its highest level (Aller and Cochran, 1976). This maximum in the rate of reworking results not only from the increased temperature that increases activity at this time, but from the fact that macrofaunal populations tend to be near their highest abundance and diversity of the year due to earlier recruitment and growth (Rhoads et ul., 1977; Yingst and Rhoads, 1978). The increase in biogenic transport flushes the sediment at a rate faster than would occur by simple one-dimensional molecular diffusion in a homogeneous body. Decompositionrates may also be somewhat depressed at this time due to depletion of fresh organic matter from the earlier plankton blooms (Riley, 1956b; Conover, 1956; Nixon et al., 1980) and to the start of lower water temperatures. The resulting decrease in standing concentration of NH: , HPOi-, and HCO; , or corresponding replenishment of SO:- occurs down to at least the deepest depth intervals sampled by the box cores at NWC and deep (16-18 cm) and to about 11-13 cm at FOAM-4 (Figs. 19-27). If it is assumed that the minima in the gravity core NH: , HPOZ-, and alkalinity profiles at NWC and DEEP correspond to the base of this exchanged zone and that the increased concentrations above these minima represent increased production rates resulting from the depth dependence of R s, then the zone of exchange is about 15-20 cm deep at NWC and 20-40 cm deep at DEEP. This is in remarkable correspondence with the depth to which excess "'Pb is found and the known burrowing depth of animals at these stations. The transport process itself does not involve a great deal of particle reworking, as demonstrated by 234Thand '"Pb profiles and the sediment laminations at FOAM, which put limits on the rate of particle mixing, but is due to porewater exchange across burrows formed by sedentary and mobile organisms. The winter-early spring profiles collected at close to the coldest temperatures of the year represent profiles having minimum biological contribution. NH: and alkalinity profiles appear to be simply diffusion-altered relicts of the fall profiles. HP0:- concentrations decrease with a concave-up form (Figs. 25-27) indicative of consumption, probably as a result of increased adsorption or reaction with Fe oxides as the interface is approached (Li et al., 1973; Syers et al., 1973). This consumption is hidden in most profiles taken at other times of the year by high production

DIAGENETIC PROCESSES. I.

285

and reducing conditions closer to the interface. At DEEP, where production does not increase so radically near the interface as at FOAM and NWC, evidence for HP0:- consumption near the interface also occurs in the fall profile where production is lowered somewhat from summer. Additional reactions not evident from concentration profiles at other times of year also become noticeable at the lowered transport rates that act during the winter. An example of this is Ca. Winter calcium profiles show concentration increases above the seawater CdC1 ratio in a zone just beneath the sediment-water interface (Fig. 28 and in preparation). Profiles of this type have been recreated in the laboratory by simply removing macrofauna and reducing biogenic reworking, which prevents Ca from building up to levels detectable above seawater background (Aller, 1978). This example leads to two very important conclusions: (1) CaCO, is dissolving in near-interface sediment in all regions of the Sound; this will be discussed in a future publication, and (2) when biogenic influence is reduced during cold periods of the year, inorganic reactions become evident and play a relatively greater role in controlling near-interface porewater profiles than during warmer times. The transition between predominantly biologically controlled periods which represent -3 of the year and more physically controlled is smoothly varying and determined by the oscillation of temperature. 6.6. One-Dimensional Models of Pore-Water Distributions in the Zone of Bioturbation

Quantitative transport-reaction models can be used to check for consistency in the qualitative interpretations of controls on diagenesis near the sediment surface. It is commonly assumed that sedimentary deposits are analogous to an accreting, porous slab bounded above by overlying water and below by impermeable basement (Berner, 1971,1976a;Lerman, 1976). The appropriate transport-reaction equation describing the vertical distribution of pore-water constituent in this body is given in one dimension (Cartesian coordinate system) as

where C is the concentration of pore-water constituent being considered; x is the vertical dimension, origin at sediment-water interface, positive

axis into sediment; t is time; w is the sedimentation rate; D is the bulk sediment diffusion coefficient modified for charge coupling and ion pairing; R is the reaction term with functional form to be described; and K is

286

ROBERT C. ALLER

the linear adsorption coefficient. Compaction is ignored and adsorption is expressed as a simple linear isotherm in this formulation (Berner, 1976). Models of this type, suitably modified, have been found to adequately describe pore-water distributions in a wide range of environments (Berner, 1974b; Lerman, 1976). Because pore-water distributions over only the upper few decimeters of sediment will be considered here, advection can be ignored as relatively unimportant compared to diffusion and reaction (Lerman, 1975). Equation (6.3) therefore simplifies, assuming a depth independent D , to (6.4)

Adsorption has been left out at this stage for clarity but can readily be included in subsequent considerations by dividing D and R by the factor 1 + K. In order to apply Eqs. (6.3) or (6.4) to the sediment zone subject to bioturbation or physical disturbance, a number of researchers have suggested that the effects of biogenic or physical reworking on pore-water distributions be lumped into an effective or apparent transport coefficient (Hammond et al., 1975; Vanderborght et al., 1977; Goldhaber et al., 1977). D is arbitrarily adjusted until the observed pore-water profile is fit to the appropriate solution, say, to Eq. (6.4). This approach has been found to work reasonably well in sediments where biogenic or physical disturbance is limited to the upper few centimeters (Vanderborght et al., 1977; Aller, 1978). Reported values of effective pore-water transport coefficients, referred to hereafter as D,, for biogenic reworking range from -1 to -2 x cm2/sec (Goldhaber et al., 1977; Aller, 1978). The one-dimensional effective transport-reaction model can be applied to NH: data available in the present study as follows. NH: distributions are chosen for modeling for a number of reasons: (1) production rates of NH: are known as a function of depth at the stations studied (Fig. 36); (2) NH: is subject to simple linear adsorption reactions, but is unlikely to be irreversibly removed from solution at the concentrations observed (Nbmmick, 1965; Mortland and Wolcott, 1965; Rosenfeld, 1979); and, (3) consumption reactions, if they occur at all, are probably restricted to the sediment-water interface where they can be taken into account in a simple boundary condition. Both the activity of macrofauna and bacteria are temperature sensitive. It was argued earlier that the temperature dependence led to at least a portion of the observed seasonality of pore-water profiles. In the present model it will be assumed that temperature dependence can be described by Arrhenius rate equations using apparent activation energies appropriate for bacterial metabolic activity (the R term) and macrofaunal ac-

DIAGENETIC PROCESSES. I.

287

tivity (the D, term) (see Prosser and Brown, 1961). An appropriate apparent activation energy for NH: production is 18 k 3 kcaYmole (Aller and Yingst, 19801, whereas that for macrofauna is -11 kcal/mole (Ql0 -2; see Mangum, 1963; Young, 1971; Coyer and Mangum, 1973). The model equation (6.4) will first be solved for constant D and R and then altered to take temperature dependence into account. The initial and boundary conditions to Eq. (6.4) are taken as:

(a)

t=O,

OlXsL,

c=cB

(6.5) (b)

t>0,

x = 0,

(4

t>0,

x=L,

c = CT c = CB

This is comparable to starting the model with an NH: profile like that at NWC during the fall (Fig. 23). The lower boundary condition is arbitrary and is based on the observation that concentrations stay approximately constant at depth L 10-15 cm for any given year. It also allows sediment below the modeled interval to act as a source or sink for NH: . In some cases a physical justification for the use of the boundary condition (c) [Eq. (6.5)]at x = L could be that L corresponds to the depth ofpreferential feeding and irrigation by a common macrofaunal species (Grundmanis and Murray, 1977). Pore-water solutes would then be held by localized irrigation at a balance concentration, in this case CB,at the base of the zone. Unfortunately, in the case of NWC at least, animal species which could unequivocally produce such features, for example maldanid polychaetes, are not common and condition (c) of Eq. (6.5) must remain largely an unexplained assertion. A solution to Eq. (6.4) and conditions (a)-(c) of Eq. (6.5) is obtained with R = R , + R o exp(-crx) (R,, Ro are constants; see Fig. 36) by method of the Laplace transform and inversion theorem:

-

(6.6) C = CT + ( C ,

-

x RIx CT)- + -(L L

20

- X)

exp( - n2nZDt/L2) z [( - 1)” - 13 sin(n.rrx/L) nn(n ’IT’DIL~) [( - 1)“ exp(aL) - 11 sin(nvx/L) exp( -nZn’Dt/L’) + 2Ro z n ~ ( D a ’+ n2n2D/L’)

+ 2R,

n=l

n= I

288

ROBERT C. ALLER

To illustrate the behavior of this model, Eq. (6.6) has been graphed for different times at a temperature of 22°C in Fig. 37 using Rl(1 + K ) = 28.1 exp( - 0 . 5 5 ~ ) + 0.57 mM/yr (Aller and Yingst, 1980; Fig. 36), a molecular diffusion coefficient D = 1 x cm’lsec (see Section 6.9), K = 1 (Rosenfeld, 1979), CT = 0.0002 mM, and CB = 0.1 mM. The case of a variable effective transport coefficient D , is considered later. Several important predictions are made by the model: (1) at small times the basic form of the summer NH; profile at NWC can be simulated; (2) very large excursions in NH; concentrations can take place in a short period of time at 22°C; and (3) in the absence of temperature change the system should go to steady state in about 6 months. At the assumed value of D the steady-state profile does not closely resemble the naturally occurring ones. The time dependence of D , and R can be included in the solution by making a number of simplifications. Based on the previous discussion, let

R(x,t) = R(x)g(t)

(6.8)

where bothf(t) and g ( t ) are of the form of the Arrhenius rate function A ( t ) in Eq. (6.1) (see Fig. 38). This gives from Eq. (6.4) (6.9)

:lo

0

15 FIG. 37. Plots illustrating the behavior of transient-state model equation [Eq. (6.611 for the bioturbated zone NH; concentrations at successive times with constant temperature of 22°C. CB = 0.1 mM is the starting value; D = 1 x cm2/sec. The NH; production rate used is that of station NWC in Fig. 36 (T = 22°C).

289

DIAGENETIC PROCESSES. I.

Month

FIG. 38. The temperature- (time) dependent functions assumed to describe the relative seasonal change in effective transport rate D,and NH: production rate R [Eqs. (6.l), (6.7), (6.8)]. The maximum value for each is taken as that occurring at 22°C in the models. Apparent activation energies of 1 I and 18 kcal/mole were used for D. (macrofauna controlled) and R (microbe controlled).

Let (6.10)

t‘ = l f ( t * ) dt*

where t* is a dummy integration variable (Crank, 1956);this reduces Eq. (6.9) to

ac -- D e -a2c+ R ( x ) G ( t ’ ) _ at’ ax *

(6.11)

The solution to (6.11) with the conditions (a)-(c) of Eq. (2.5)] can be obtained from (6.6) by use of Duhamel’s theorem (Bartels and Churchill, 1942). Several simplifications will first be made. Since in A ( t ) , T o S T , it follows from successive use of the binomial and exponential series expansion that (6.12)

-

A ( t ) A . exp( -E/kTo)[l

+ ( T I / T o sin(wt)] )

- A exp( - E/kTo)[1 - ( E T l / k E )sin(wt)

if ET ,/kTZ,is also small. Applying (6.12) to (6.10), then the approximation t’

(6.13)

- F o exp( - Ef/kTo)t

can be made wheref(t) = fo exp( -E,/kt). In the present modelf(t) differs from g ( t ) only in the value of the activation energy in each case, so that (6.14)

G(t)

- G oexp( -EJkTo)[l

- (EGTl/kTo)sin(vt‘)l

with G o = g d f 0 ; E, = E, - E,; k = 1.99 cal/deg/mole; v

=

IT exp(Ef/

290

ROBERT C. ALLER

kTo)fo.Note that iff(t) = g(t), the solution is essentially (6.6) and the system will go to steady state. Using this approximation for G(t) and letting time become long enough (-1 yr) for exponential terms to decay, gives finally a solution to (6.11) in time t as x R IX (6.15) C(x,t) = CT + (CB- CT)- + -(L - x)G(t) L 20,

Ro De a

+--$l -

Rdr -[1 D :L

-

- 2Ro

exp( - a L ) ] G ( t )

' ffi

- 2R'

exp( - ax)]G(t)

[( - 1)" - 11 sin(nnx/L)G*(t)

nn(n2n2De/L2)

n=1

-aL) - 11 sin(nnx/L)G*(t) I: [( - 1)" exp( nn(D,a2 + n2n2De/L2) cs

n=l

where

(6.16) (6.17)

G(t) = Goexp G*(t) =

G~EGTI kG

(vn'T 2DJL ') cos(2at) + v 2 sin(2nt) (n2T2De/L2)*+ 'V

'"(%)[

1

Equation (6.15) is the periodic part only of the solution to (6.11) and describes the behavior of the system after particular initial conditions are no longer important (-1 yr). This is the relevant part of the solution for the present considerations. The solution (6.15) is graphed in Fig. 39 for three values of D,: 1 x lo-', 2 x lo-', and 6 x lo-' cmz/sec (maximum value at 22°C). The values E, = 1 1 and E, = 18 kcal/mole were taken for the temperature dependence of macrofauna and bacteria, respectively. Thef, and goterms were calculated so thatf(t) and g(t) are unity at 22°C. The absorption coefficient K is 1 (Rosenfeld, 1979). The value of R is that from NWC (Fig. 36). Many of the most basic features of time dependence and concentration ranges in the NH; pore-water profiles at NWC can be reproduced by the assumptions of this model. Effective transport coefficients of - 1-2 x lo-' cm2/sec result in the best agreement between modeled and observed con-

29 1

DIAGENETIC PROCESSES. I.

0.2 0.4

0.2 0.4

, ,' I'

I

,

May

I

July

,

lug

FIG. 39. Periodic oscillations in pore-water NH: concentrations predicted for the bioturbated zone by the one-dimensional model Eq. (6.15). The profiles represent of a year separation and corresponding months are indicated. The different temperature dependence of D , and R as illustrated in Fig. 38 produces the oscillations. Three maximum values of D , were assumed: 1 x (dashed line), 2 x (solid), and 6 x cm2/sec (dotted).

centrations (Fig. 39). The model predicts a periodic oscillation in the profiles with minimum values found approximately in February-March. Maximum concentrations migrate up and down the sediment column and become skewed toward the interface during warmer months as is actually observed. The major discrepancy between measured and predicted profiles is found in the fall season where the model badly overestimates the expected concentrations. Similar modest agreement between simulated and observed profiles at the other stations can be obtained by appropriate modification of R , D,, and C , . The major disadvantages of the model are the utilization of the generally unexplained boundary condition (c) of Eq. (6.5) at the base of the bioturbated zone and the arbitrary adjustment of an empirical transport coefficient required to obtain any resemblance to the actual data. In order to determine whether a different class of boundary conditions at x = L might result in better fits of the one-dimensional model to the observed pore-water distribution, Eq. (6.4) was solved with the initial and boundary conditions: (a)

t=0,

OIXCL,

(6.18) (b)

t

> 0,

x = 0,

(4

t

> 0,

x

=

L,

c c

=Ce = CT

aclax = F

292

ROBERT C. ALLER

All definitions are as before and F is a constant. The solution obtained by use of the Laplace transform and inversion theorem is Ro (6.19) C(x,t) = CT + -[l - exp(-m)] Da *

[4RIL2 + 2LFDPn(-1)" 4 " --I: D n-0

+ D(CT - C B ) P ~ ]

x sin(P,,x/2L)exp( - DPit/4L ') p',

{RoPn - [2LROa exp(-aL)](

-4c. n=O

- 1)")

x sin(P,,x/2L)exp( -DP2t14L2)

P',(Da2 + DP2,/4L2)

where P,, = (2n + 1 ) ~ . cm2/sec, This solution is shown in Fig. 40 for the values D = 1 x K = 1, RI(1 f K ) = 28.1 exp( - 0 . 5 5 ~ )+ 0.57 mM/yr, CT = 0.0002 mM, and C B = 0.1 mM as used previously for Eq. (6.6). The value of F is taken as 0.0056 mM/cm measured from the gravity core NWC-G over the depth interval 15-25 cm. The initial patterns of concentration-profile change predicted by this model, like those of Eq. (6.6), result in shapes similar to profiles observed

FIG.40. Plots showing successive NH: concentration profiles predictec At 22°C for the bioturbated zone at NWC by Eq. (6.19). A basal flux rather than concentration constraint at x = L = 15 em is used in this case. The sequence shown would be similar to that which would occur in a box core collected at 10°C, warmed rapidly to 22°C and allowed to sit (except F = 0 in that case). A regular periodic oscillation of temperature instead of a single warming would produce corresponding oscillating profiles similar in shape to the steadystate profile; no maxima would occur.

DIAGENETIC PROCESSES. I.

293

in early summer at NWC. The predicted temporal sequence actually shown is close to that expected for a box core of length L collected at -10°C and subsequently stored at 22°C (F = 0). At long times a stable parabolic-shaped profile is predicted. If periodic values for 0’ and R as a function of temperature are assumed in this case (with F = 0 for simplicity), only parabolic-shaped profiles of rising and falling concentrations at x = L can result. Maxima cannot occur. Such a model might be appropriate as an empirical description of concentration profiles of a relatively unreactive pore-water constituent at stations other than NWC, for example, DEEP, but like the previous one-dimensional models it suffers from the arbitrary definitions of boundary conditions and the ad hoc variable D , required to simulate the observed distributions. If a pore-water constituent more reactive than NH: had been considered, it might be possible to ascribe some of the discrepancies between measured and predicted distributions to precipitation reactions at depth in the sediment and adjust a reaction coefficient accordingly. The adjustment of an additional reaction variable in order to obtain a better fit would simply disguise the lack of agreement between the assumptions of the one-dimensional model and the real world. The problems of the one-dimensional models in describing the distribution of a pore-water constituent not subject to precipitation reactions can be eliminated mostly by considering a more realistic geometry of diffusion within the bioturbated zone as shown in the next section.

6.7. Two-Dimensional Models of Pore- Water Distributions in the Zone of Bioturbation An alternative to the one-dimensional model can be developed by noting that the zone of sediment inhabited by macrofauna is not a homogeneous one-dimensional slab, but instead is a body permeated by cylinders. The water within these cylinders (burrows) is maintained at approximately seawater solute concentrations by the irrigation activity of their animal occupants. Interstitial solutes can therefore diffuse into burrows and be advected out of sediment by irrigation activity as well as diffuse vertically toward the sediment-water interface. Diffusion in this case can be considered as taking place in a system of cylindrical symmetry similar to that occurring in a root-permeated soil (Gardner, 1980; Cowan, 1965; Nye and Tinker, 1977). It is possible to imagine a geometric analog to the average geometry of diffusion in the bioturbated zone as follows. Consider the ideal case where a population of animals builds only vertically oriented tubes of

294

ROBERT C. ALLER

Water

r L

FIG.41. (a) Idealized packing of burrow microenvironmentsenvisionedfor the bioturbated zone. (b) The idealized geometry in vertical cross section. (c) A single model hollow cylinder of length L representing the average diffusion geometry and microenvironment in the bioturbated zone (after Aller, 1978, 1980).

equal length and distributes these uniformly over a given area. An equal volume of material can then be assigned to each tube such that the inhabited zone is envisioned as consisting of a set of closely packed hollow cylinders or annuli (Fig. 41). Looking down on the interface, the deposit would appear like a case of cans filled with sediment. Each can would have a hollow, vertically oriented tube in its center. Only the portion of sediment at the intersection of any three cylinders is unaccounted for in this mental construction and is ignored. The average geometry of diffusion in such a system is simply that of a single hollow cylinder (Fig. 41c). Although this description was derived by considering the ideal situation where only a single kind of relatively immobile animal is present (Aller, 1978, 1980), the fundamental assertion that the average sediment microenvironment is a hollow cylinder can be readily applied to the multispecies case. The same idealized geometry is assumed, only now the dimensions of the individual cylinder are varied to correspond to an average or effective cylinder microenvironment. The inner radius of this cylinder is determined by the statistical size distribution of individuals, whereas the outer radius or distance between individuals is determined by animal abundance and distribution. The transport-reaction equation describing pore-water distributions in the single-cylinder microenvironment is (cylindrical coordinates): ac = -D- a( r z a)c +- D s -+ R (6.20) at r ar ax

295

DIAGENETIC PROCESSES. I.

All symbols are as before and r is the radial dimension measured from the tube axis (Fig. 41c). Advection and compaction are ignored for reasons described previously. D is assumed isotropic and constant. In the present discussion R is assumed to have only vertical (x) dependence although it is known that burrow walls can be sites of high metabolic activity (Aller and Yingst, 1978). Equation (6.20) will be solved only in the steady state since the dimensions of the cylinders to be considered are such that steady-state distributions are rapidly achieved. The boundary conditions are taken as (a) (6.21) (b) (C)

x = 0,

r

=

r2,

x = L,

r = rl, aClar

=

0

actax

=

F

C=CT

Condition (a) specifies a constant concentration of solute along the sediment-water interface and within the burrow core. Condition (b) requires pore-water solute concentrations to go through a maximum or minimum value half-way between any two burrows. The last condition, (c), matches the bioturbated zone to the underlying unburrowed zone by requiring continuity in flux across the lower boundary at depth L. The solution to Eq. (6.20) with conditions (a)-(c) [Eq. (6.20)] and R = R + R o exp( - OUT) is (by separation of variables)

with (6.23)

A, = (n

+ i)(n/L)

(6.24)

Uo(Anr)= K l(Aflr2)Z~(Aflr) +ZI(A~~~)KO(L~)

(6.25)

'"=[

R o exp( -aL)( - 1)" a2+A;

A 3 0

A"

The functions Z J z ) and K,(z) are the modified Bessel functions of the first and second kind respectively of order v (see, e.g., Abramowitz and Stegun, 1964 for values). The solution [Eq. (6.22)] gives the radial and vertical pore-water distribution within a single average microenvironment. Actual measured pore-water distributions are mean vertical distributions. If a core sample

296

ROBERT C. ALLER

integrates over a sufficient number of microenvironments, then the average vertical pore-water concentration is identical to that found in a single model cylinder or microenvironment. Over any vertical depth interval xI-x2,the measured concentration C is, therefore, given by

(6.26)

c=

271

[6:Cr dr dx [6; r dr dx

271

The behavior of the model is illustrated for NH: in Fig. 42. In this case an effective cylinder size has been chosen based approximately on the populations of Nephtys incisa at NWC in July 1974 (Yingst and Rhoads, 1978). The mean radius of the Nephfys present was found by direct measurement of collected individuals as r , = 0.14 cm and the population abundance N gives an estimate of r 2 = (TZV)-”~ = 3.3 cm. In Fig. 42, r2 is taken as 4 cm. [Note that r 2 may also be estimated from the observed length 1 of burrows/crn3sediment in x radiographs as r 2 = IT^)-"^.] The production rate for NH: at NWC is that of Fig. 36; D is taken as 1 x lo-’ cm2/sec(see subsequent text); F = 0.0056 mM/cm from the gradient between 15-25 cm in gravity core NWC-G; L = 16 cm; and CT = 0.0002 mM.

There are a number of important conclusions to be made immediately: (1) the cylinder model produces a vertical profile similar in shape and magnitude to that actually observed by direct, unmanipulated use of faunal data, molecular diffusion coefficients, and production rate data (it will be shown subsequently that minor manipulation of cylinder dimensions can produce almost exact fits); and (2) the radial profiles demonstrate that a range of concentrations can be expected to exist within microenvironments of the bioturbated zone well below the interface. This concept will be resumed later when equilibrium controls on pore water are considered.

NH: profiles at all stations can be successfully fit by the cylinder model. The production rates used in all cases are those given in Fig. 36. These are modified as before to appropriate values at particular temperatures by the Arrhenius rate function [Eq. (6.1)] with an apparent activation energy of 18 kcal/mole (Aller and Yingst, 1980). Estimates of bulk sediment diffusion coefficients, D , at different temperatures were made by multiplying the diffusion coeficients at infinite dilution (Li and Gregory, 1974) by the factor cp2, where cp = porosity (Manhiem, 1970; Lerman, 1978; Krom and Berner, 1980). These give cp2 corrections of 0.44,

297

DIAGENETIC PROCESSES. I .

(a1

?I 0.2 r

_ - -- -

r2= 4cm' r, =O.I4cm

(b)

1

NH;

0.1

I'

2

3

4

Radial Position (cm)

(mM)

0.2

0.3

I

14

16

FIG. 42. (a) A representative average microenvironment predicted for station NWC illustrating radial NHf concentration profiles around the idealized central burrow at three depths. The production rate of NH; is that in Fig. 36, r , = 0.14 and r 2 = 4 cm. (b) The average vertical pore-water gradient in I-cm increments predicted by the microenvironment of (a). See Eqs. (6.22) and (6.26).

0.56, and 0.56 for FOAM, NWC, and DEEP, respectively (Section 6.9). Initial values of r , and r 2 were estimated from the major components of the deep-burrowing (>2 cm) polychaete fauna only. The species included are Nephtys, Pherusa, Clymenella, Pista, and Melinna. r was estimated as an average of direct measurements of preserved collected individuals and initial r 2 magnitudes were estimated from the polychaete abundances (see Yingst and Rhoads, 1978; Appendix A). L was chosen to correspond to an obvious break in pore-water concentration profiles and the known

298

ROBERT C. ALLER

0.1 0.2 0.3 0.4

-5

0.1 0.2

0.3 0.4

5

2 z

:10 0)

n

FIG.43. Fits of two-dimensional hollow-cylinder microenvironment model to pore-water NH: profiles at FOAM (see Table V for model values). Solid lines represent measured profile, dashed represent model.

burrowing depth at each station. In order to fit the profiles, r 2 was varied, but r l was kept fixed at the measured mean value. Table V lists model parameters actually used, and corresponding fits of the model to the profiles are shown in Figs. 43-45. Profiles from the first year of collection at FOAM were not modeled because of the lack of faunal data for that time.

a W

D

15

NWC-4

15

'/

NWC-6

FIG.44. Fits of two-dimensional hollow-cylinder microenvironment model to pore-water

NH; profiles at NWC (see Table V for model values). Solid lines, measured profile; dashed, model.

TABLEV. CYLINDERM~CROENVIRONMENT MODELVALUES

core

("C)

F, basal gradient (mM/cm)

FOAM-4 FOAM-5 NWC-1 NWC-2 NWC-3 NWC-4 NWC-5 NWC-6 DEEP- 1 DEEP- 1 DEEP-2 DEEP-3

14" 2" 20" 13" 3" 19" 16" 3" 19"

0.042 0.036 0.0056 0.0056 0.0056 0.0056 0.0056 0.0056 0

14" 6"

0 0

T

(p2 correction

L (cm)

0.44

13

0.56

0.56

DNW+ (cm'lday)

K

10

0.583 0.401

16 18 16 18 16 16 16

0.723 0.530 0.838 0.780 0.530 0.838

1 1 1 1 1

16 16

0.742 0.588

0.857

R d l + 0" R,/(1 + 0" (mM/day) (mM/day)

1

0.0379 0.00953 0.0624 0.0293 0.00924 0.0562

1 1 1

0.00924 0.0136

0.001 0.00025 0.00126 0.00059 0.00019 0.00114 0.00083 0.00187 0.0032

1 1

0.00794 0.0032

0.00187 0.000754

0.0408

Initial estimatedb

Estimated PI

(cm) 0.075 0.075 0.14 0.14 0.14 0.14 0.14 0.14 0.20 0.20 0.20 0.20

Model r z (cm) 3.5 8

6.5 5.9 14 4.5 4 6.5 4.5 4 4.7 8.8

rz

(cm)

3.2 3.4 3.3 3.9 4.7 7.0 3.8 3.7 2.5-3.7 3.2-4.9

" T correction made assuming 18 kcdmole apparent activation energy on experimental values determined at 22°C (Fig. 36). r 2 estimated from deepburrowing (>2-3 cm) polychaete abundance in box core taken within -20 m of chem core. Difference between estimated and actual model value usually is equivalent to 2-3 more individuals in a faunal sample. DEEP sample higher values use Pherusa abundance only.

300

ROBERT C. ALLER

0I

0.2

0.I

0.3

0.2 0.3 I

~‘i

10 DEEP-2

15

‘5

t

j

FIG.45. Fits of two-dimensional hollow-cylinder microenvironment model to pore-water

NH: profiles at DEEP (see Table V for model values). Solid lines, measured profile; dashed or dotted, model.

In general, the hollow cylinder or average microenvironment model fits the NH: data extraordinarily well with only minor variation in r 2 , the effective density of animals. The required variation in r z presumably comes about because of the uneven distribution of animals within sediment (Jumars et al., 1977) and mobility. It may also be apparent variation due to the arbitrary requirement that r remain fixed. The variation forced into r z by this restriction could actually reflect changing boundary conditions at r l , for example, inhibition to diffusion by burrow linings or a reduction in irrigation activity. A part of the discrepancy between the model and measured profiles could also reflect changes in the production rate of NH: because of factors other than temperature. For example, fall profiles may show a smaller maximum than predicted, due to a lower reaction rate resulting from a depletion of substrate (see Nixon et al., 1980). In order to illustrate how variation of the parameters r (burrow radius) and r 2 (effective population abundance) can influence the steady-state concentration of a pore-water constituent, the predicted average concentration of NH: in the top 16 cm of NWC sediment at 22°C is plotted as a function of either r I or r z in Figs. 46 and 47. Values for D and R are fixed at values appropriate for 22°C at NWC (Fig. 36, K = 1, 0 = 1.03 x lo-’ cmz/sec).Only two members of an infinite family of curves have been plotted for each case of fixed r I or fixed rz. The values of Y , chosen are reasonable end members for macrofauna commonly collected in the upper 50 cm of normal marine sediment. The population abundances equivalent to r z values are also shown for convenience. A typical range of abundance for near-shore sediment would be several hundred to onethousand large macrofauna per square meter. Note that concentrations are most sensitive to animal abundance rather than burrow size and that

301

DIAGENETIC PROCESSES. I.

even at extremely low population densities, the presence of irrigated burrows has a significant influence on the buildup of pore-water constituents. This can be seen by comparing the predicted mean concentrations of the two-dimensional model with that of the one-dimensional, vertical diffusion (molecular) model indicated as the upper dashed line. It should be noted that the cylinder model also relatively accurately predicts the SO:- profiles although these have not been plotted. If, for example, the consupption rate depicted for NWC in Fig. 35 is used in cm2/secassumed the model and a SO:- diffusion coefficient of -5 x (Krom and Berner, 1980), then a nearly vertically constant SO:- profile is predicted for the upper 15 cm. An average cylinder size of r , = 0.14 cm; r 2 = 3.3 cm was used. A general offset of -0.5 mM SO:- from overlying seawater concentrations and a slightly greater offset (- 1.5 mM) at 3-4-cm depth are predicted by the model. Such a pattern would be essentially undetectable with the analytical method used for SO:-. This 1.0

F

-

_No_ Burrows _ _ - -o n-e 4- -Vertical - - -Case -7,

= OIcm

,, 0.5cm I.OCrn 4 =

:

co E

T Twical Cmwntration Range

-I 0

0-16em

0.1

I

0 C

.c

E!

+ C 8 0

001

0

0 c U Q)

I

+*0001 I Z

_ _ _ _ _Overlying _ _ - _Water __--0.OoOl~ 0

"

2

'

1

4

"

6

"

8

"

10

"

12

Half-Distance between Burrows (crn) I

3183 795 354 199 127 BB

65

50

39

32 26

22

Population Abundance (/m2)

FIG.46. Average pore-water NH; concentration in the top 0-16 cm at NWC predicted at 22°C by the hollow-cylinder model as a function of rz (half-distancebetween individuals) or effective population abundance. Burrow radius r , is fixed at 0.1 (A), 0.5 (A),or 1.0 crn (0) for each curve. The lower bound is the seawater value 0.0002 mM and the upper bound is the one-dimensional model [Eq. (6.4)] no burrow case. Typical ranges indicated.

302

ROBERT C.ALLER 1.0

No Burrows

one-D

E------------

Vertical Case - --

r2 = 6cm N = m/m2

r2 = 4 c m N = 199/m2

I

Typical Average Concentrollon Range 0-16cm N W C

r2 = 3 c m N = 354/m2

+ L C 0)

r2 = 2 cm

-

0 C

u0 C

0

0001

N = 795/m2

Typical S i z e Range at N W C

to

I

z

_ _ _ - Overlying _ - _ _ _Water __OOOOll

"

"

"

"

I

'

"

0 I 0 2 03 0.405 0.60.7 0.8 0.9 I 0 I I 1.2

Burrow Radius (cm)

FIG.47. Average pore-water NH: concentration in the top 0-16 cm at N W C predicted at 22" by the hollow-cylinder model as a function of r , (burrow radius) at fixed values of r 2 (half-distance between burrow axes) or N (equivalent population abundance). Upper and lower boundaries as in Fig. 46. Typical ranges of values at NWC are indicated.

suggests that more sensitive SO:- methods should be employed within the zone of bioturbation to rigorously test the model (e.g., Sayles and Mangelsdorf, 1976). The successful application of the cylinder model does not contradict the assertion of relatively lower biological influence on sediment chemistry during winter periods. This lower influence is expressed in the macrofaunal component of the cylinder model by a required larger r 2 (lower effective abundance of macrofauna) and in bacterial activity by the diminished production rate R during the same period. Up to this point, the cylinder model has been applied only to the upper few decimeters of sediment inhabited by the easily captured macrofauna. Deeper regions of sediment, particularly at NWC and DEEP, are known to be burrowed to depths > 1 m by the mantis shrimp Squillu (see Meyers, 1979). The resulting large-diameter burrows are clearly responsible for

DIAGENETIC PROCESSES. I.

303

the deep exchange of pore water observed at these offshore stations. In order to model the entire burrowed zone to depths of a meter or more, it may be necessary to utilize either a stacked cylinder model in which stacked cylinders of variable r l and r 2 are imagined as the effective microenvironment at multiple depths, or perhaps a single cylinder of smoothly varying dimensions with depth (Fig. 48). Such a model, which will not be pursued here, would reflect the fact that deeper burrowers are less abundant, but generally of larger individual size than the more shallow dwellers. The two-dimensional, single-cylinder model can be considered the far superior of the two end-member models for the bioturbated zone presented in this article. Although idealized, it is basically realistic and allows direct input of measurable physical and biological parameters with little mysticism. It will now be shown that this simple but relatively successful description of transport conditions within the bioturbated zone can provide insight into thermodynamic equilibrium controls on pore-water distributions. 6.8. Abiogenic Reaction Controls on Pore- Water Composition In addition to transport and biologically mediated reaction controls on pore-water composition, interstitial solutes are subject to abiogenic reactions with specific solid phases in the sediment (Suess, 1976, 1979; Emerson, 1976; Hartman et al., 1976; Sayles and Manheim, 1975; Martens et al., 1978). The interaction of particular solid phases with pore-water compositions can be inferred in part by examining the disequilibriumstate of pore water with respect to selected compounds likely to be present.

1 (0)

(b )

FIG.48. (a) Geometrical analog for diffusion in the zone of biogenic reworking with burrow size and abundance variable with depth. (b) Simplified geometry of single, stacked hollow cylinders corresponding to (a).

304

ROBERT C. ALLER

Calculations should be done in conjunction with direct information on the composition of the solid phase, for example, measurement of sulfides or carbonates. Unfortunately, this is not always analytically possible and in the present study no attempt to do this was made except for the sulfides. In order to calculate the saturation state of pore water with respect to a given phase, it is necessary to determine the individual activities of ions in solution. This requires calculation of specific (ion pairing) and nonspecific interactions influencing the solution activity of a given ion. Techniques used in the calculation of total activities in natural waters are discussed in Garrels and Christ (1965) and Berner (1971). In the present study, a computer program developed and written by G. R. Holdren (Holdren, 1977) was employed. Specific interactions involving ion pairs of: Na+, K+,Ca", MgZ+, Fe", Mn2+, H+,OH-, Cl-, HCO;, CO:-, SO:-, H,POi, HPO:-, and HS- are taken into account, using stability constants (25°C) tabulated in the literature (Sillen and Martel, 1964; Morgan, 1967). After iteration to obtain the free-ion concentration and a constant ionic strength, activity coefficients are determined by use of the extended Debye-Huckel equation. At the ionic strengths (0.40-0.55) present in Long Island Sound pore waters, the use of this equation introduces an error of -10-15% in divalent ion activities, but as will become apparent, this is not an important influence on the conclusions to be drawn. A detailed presentation of this model is given in Holdren (1977) and will not be repeated here. The Fe2+and MnZ+data used are discussed in Part I1 and listed in Appendix B, Part I. Total alkalinity was assumed equivalent to carbonate alkalinity, a reasonable approximation based on previous studies (e.g., Berner et al., 1970). Na+, K+,and Mg2+were calculated from C1- by use of seawater concentration ratios. The model gives reasonable total activity coefficients at FOAM, NWC, and DEEP, for example: Y T (Fez+)= 0.14-0.19, Y T (Mn2+) = 0.11-0.14, y T (Ca") = 0.21, and also agrees well in its predictions when compared to calculations using apparent solubility products in the case of CaCO,. Charge balances were found to be 20.1% of neutrality. No correction for temperature variation has been made. Only phosphate mineral formation will be considered here. Sulfide equilibria are discussed in detail in Part I1 and calcium-carbonate saturation states will be considered in another paper (or see Aller, 1977). NH; is unlikely to be involved in the formation of specific insoluble phases such as struvite (MgNH,P04 6H20), given the measured concentrations of the ions involved (Martens et al., 1978). The major interactions of NH: with sedimentary particles include reversible and irreversible adsorption reactions (Ndmmik, 1965; Mortland and Wolcott, 1965; Muller, 1977; Rosenfeld, 1979).

-

305

DIAGENETIC PROCESSES. I.

Phosphate is also subject to both reversible and irreversible adsorption reactions (Li et al., 1972; Syers el al. 1973) and is known to form a wide range of relatively insoluble Ca, Fe, and Mn compounds in marine and lake sediments (Bray, 1973;Troup, 1974;Bricker and Troup, 1975;Nriagu and Dell, 1974;Tessenow, 1974). Specific adsorption on Fe oxides (Parfitt et al., 1975) is thought to greatly influence HP0:- mobility and concentrations in surface sediment (Hayes, 1964; Mortimer, 1971). Expressed as an adsorption isotherm, phosphate can be shown to be much more strongly absorbed under oxygenated ( K 25-50) than anoxic (K 0.5-8) conditions, probably as a result of a reduction of Fe oxides and a lowering of strong, specific Fe3'-PO, interaction in the anoxic relative to the oxygenated case (Olsen, 1964; Li et at., 1972; Krom and Berner, 1980). Distinct, end-member phosphate phases that are likely to form in anoxic marine sediments include: Ca,(PO,),OH (Hydroxyapatite), Ca,(PO,), (Whitlockite), Fe,(PO,), - 8H,O (Vivianite), and Mn,(PO,), 3H,O (Reddingite). Struvite, mentioned earlier, may also occur in inshore LIS sediments, but is highly undersaturated at the stations examined in this study (Martens et al., 1978). The presence of a pure phase is the exception rather than the rule in low-temperature disgenesis (Suess, 1979), but in the absence of information on the solid-phase composition, comparison of pore-water saturation states with respect to only the end member will be treated here. The solubility products and reactions used here as a guideline to saturation states are given in Table VI. The results of the calculations for phosphate compounds are plotted as -log IAP (IAP is the ion activity product) as a function of depth at each station (Figs. 49 and 50). Only data from box cores collected during 1975-1976 and some selected horizons from the gravity cores are shown. Hydroxyapatite was supersaturated by a factor of lO3-lOioat all stations and is not plotted; precipitation of this phase is known to be kinetically hindered in seawater (Martens and Harriss, 1970). Bray (1973) and Norvell(l974) inferred likely equilibrium of pore waters with whitlockite [Ca,(PO,),] in Chesapeake Bay and anoxic lake sediments, respectively. Long Island Sound pore waters also tend to have activity products close to those predicted for saturation with respect to whitlockite, although distinct undersaturation is found for most sediment intervals at NWC. A zone of possible saturation with vivianite occurs first beneath the interface at all stations during the summer and at FOAM and DEEP during winter (Fig. 49). Otherwise pore water is undersaturated with respect to this phase. A similar depth distribution of saturation states is found for reddingite at FOAM and NWC. At DEEP, saturation can occur well below 4 cm in summer and winter box cores. The near-interface bands of sat-

-

-

FOAM

NWC

- log(lAP),

DEEP

Grovity Cores

COJ[PO~)~ 26 0 5

2

~A

6

I

Gq

&

f n

410 dm

'A 0

0

n

10

0 1

A 'On

90

,

@

0ID

- log(1AP).

120

I

O I 15

80

I5

'(lp

I5

Whitlockite

Kw

A 0 10

I

Whit.

I

Ksp

Kw

Whit.

Whil.

FeJ(P04)i 8H.p

0

@ 8

0 0

f

80 A

0 0 15

'

KSp V i r

0

A

120 I

I Ksp

Vir

: A

0

ao o

00 I5

0

A 0

00

0

0 0

Ksp

Viv.

K , ~ Vlvianiie

FIG.49. Plots of -log IAP versus depth at each station for selected solid phases: whitlockite [Ca3(PO4),] and vivianite [Fe,(pO,), * 8H,O]. Box cores: 0, summer (1975); 0, fall (1975); A.

a W W 0

0 Y)

2

H : : : : : : b Y)

-

m

308

ROBERT C. ALLER

TABLEVI. SOLUBILITY PRODUCTS AT 25°C FOR SELECTED PHOSPHATES Solubility product (KSJ Reference 10-28.5 Duff (1971)

Reaction Whitlockite Ca3(po4l2~t 3Ca2+ + 2P0:Vivianite Fe3(P04)2* 8H20 3Fe2+ Reddingite Mn3(P04)2* 3H20

* 3Mn2'

10 - 36.0

+ 2PO:- + 8H20 + 2PO:- + 3&0

Nriagu (1972)

10-34.6 Nriagu and Dell (1974)

uration of both vivianite and reddingite are caused by the large increase in Fez+ and Mn2+concentrations from the dissolution of oxides in this zone. There are several reasons why undersaturation with respect to pure Fe and Mn phosphates could occur: (1) solid-solution formation (Tessenow, 1974); (2) inaccurate values for KSp;(3) the pore waters are not in equilibrium with an Fe and Mn phosphate; and (4) the average pore-water concentrations cannot be used to calculate saturation states. This last reason applies only to the bioturbated zone and comes about because a wide range of concentrations actually exists in any sediment interval of this region (see Fig. 42, for example). Production and consumption rates of HP0:- would be required to check this possibility in general. In the present case, the continued undersaturation of pore waters below the bioturbated zone argues for a reason, or reasons, other than Eq. (6.4) to explain the discrepancy. In summary, the equilibrium calculations suggest that Ca, Fe, and Mn phosphates could be forming in sediments of Long Island Sound. The exact compositions of any such compounds being formed remain equivocal and could quite easily be changing from season to season in a given sediment interval as transport-reaction conditions change. These considerations show that the complexity of transport conditions in the bioturbated zone requires that production-consumption rates, transport geometries, and detailed information on the solid phase all be known in order to explain both the average magnitude and relative abundances of reactive pore-water solutes such as HPOi-. 6.9. Flux of N H ; and H P 0 : - between Sediment and Overlying Water

The sea floor is an important site for nutrient regeneration and acts as a source of continual supply of remineralized organic constituents to

DIAGENETIC PROCESSES. I.

309

overlying water (Rittenberg et al., 1955). The magnitude and relative importance of this nutrient source in any given area can be determined in three independent ways: (1) direct measurement by short-term incubation of bottom water and underlying sediment (Tessenow, 1972; Fanning and Pilson, 1974; KampNielsen, 1974; Hale, 1975; Rowe et al., 1975; Hartwig, 1976; Nixon et al., 1976, 1980); (2) calculation of fluxes by measurement of the appropriate pore-water concentration gradient and employing a Fick’s first law relation with an assumed diffusion coefficient (Bray et al., 1973; Billen, 1975; Norton and Sasseville, 1975); and (3) determination of the production rate term of the solute within the sediment body and calculation of the resulting flux at steady state (Elderfield, 1976; Aller, 1977; Billen, 1978; Ullman and Aller, 1980).

In this study, all three of these types of measurements are available to estimate the sediment-water flux of NH; and the first two techniques are available for estimation of HP0:- exchange. In addition to providing information about the influence of sediment-water fluxes on water-column processes, a comparison of the different estimates can be used to further elucidate transport-reaction processes near the sediment-water interface. For application of Fick’s first law it will be assumed that advection due to compaction or otherwise is not an important component of the flux (see Imboden, 1975; Berner, 1976b). This gives the flux of a constituent C in one dimension, with reference frame as before, as (6.27)

J,

=

cpD(dC/dx)x

where J, is the flux of dissolved material across plane x , and cp is the porosity at x. This equation will generally be evaluated at-the sediment-water interface x = 0, and the cp value will be taken as -1. It was shown previously that use of a one-dimensional transport-reaction model may in turn require use of an effective transport coefficient D,rather than D in order to account for the influence of biogenic reworking on solute mobility. The effective transport coefficient can be estimated from Eq. (6.27) simply by matching the directly measured flux with that required to balance the measured concentration gradient. The concentration gradient at x = 0 in each case was estimated by first assuming that over the top few centimeters C = a , + a l x + a 2 x 2 + * * * (a , a I , a 2 , . . ., are positive or negative constants). Then for any ver-

310

ROBERT C. ALLER

tical finite sampling interval x1-x2, the average concentration C is r xz

(6.28) Because of the shape of the profiles, particularly at NWC, a is determined almost entirely by the seawater value and the first 0-1 cm sampling interval. The addition of polynomial terms beyond x 2 does little to change a , if only the top few data points in a profile are fit to the equation. If concentration values below the top few centimeters are included in the polynomial fit, however, a l is significantly lowered. In this study the simplest possible approach was taken, C was assumed to be given by C = a. + a l + a2x2,a. was fixed at the seawater value (-0), and the top two 1-cm sampling intervals were used to find a I and a by use of Eq. (6.28). For DEEP-3 the intervals 0-2 and 2-3 cm were used. In general, this procedure maximizes the calculated flux. It is important to realize that other approaches to calculating the concentration gradient-for example, inclusion of deeper profile points, or not accounting for the finite sampling interval-significantly lower the calculated flux. The use of only near-surface samples is consistent with the fact that most decomposition is thought to take place at or just below the interface. In the absence of sediment resistivity measurements, the bulk sediment diffusion coefficients D for NH: and HP0:- at different temperatures were estimated by multiplying the appropriate infinite dilution values of Li and Gregory (1974) by Q" with n 2 (Manheim, 1970; Manheim and Waterman, 1974; Lerman, 1978; Krom and Berner, 1980). The Q' correction factors are 0.44, 0.56, and 0.56 for FOAM, NWC, and DEEP, assuming a sediment density of 2.5 gm/cm3 and average water contents in of 45%, 55%, and 55%, respectively (Table V). These values of the reciprocal formation factor agree well with those determined by direct measurement of diffusioncoefficients in Long Island Sound muds (Krom and Berner, 1980; Goldhaber et al., 1977). The infinite dilution values of D used at 22", 15", and 4" for NH: are 18.5 x 15.7 x and 11.4 x cm2/sec.For HP0;- they are 6.86 x 5.74 x and 3.99 x (Li and Gregory, 1974). Ion-pairing and charge-coupling influence on D are ignored as relatively unimportant (Lasaga, 1979; Sayles, 1979). If the production rate R of an ion in the sediment has the form R o exp( - ax) + R I , then at steady state from Eq. (6.4) or (6.27), with Q = latx = 0

-

(6.29)

Jo

- - ( R ~ / ( Y-)RIL

This assumes that the overlying water has a concentration of the respec-

311

DIAGENETIC PROCESSES. I.

tive solute of -0, only a sediment interval of thickness L is considered as a significant source of the ion involved, and the term R o e x p ( - d ) 0. This equation simply states that at steady state the quantity of an ion being produced must equal that leaving a volume of sediment of thickness L (flux from below L is ignored). The sediment-water fluxes of NH: and HPOZ- estimated by direct measurement, from pore-water concentration gradients [Eq. (6.271, and from production rates [Eq. (6.29); Fig. 361 when available are tabulated for each station and season in Table VII. Both the production rate and gradient estimates may be as much as -10% high because the surface porosity was neglected. The production rate estimates of NHZ and flux are adjusted for temperature using the experimentally determined apparent activation energy for NH: production of 18 2 3 kcalldeg-mole (Aller and Yingst, 1980). An adsorption coefficient K value of -1 was assumed in correcting the experimental rates to absolute production rates (Rosenfeld, 1979) and the equation evaluated over a depth of 10 cm in each case.

-

6.9.1. NH,+ Flux. The three independent estimates of sediment-water NH: flux are generally in excellent agreement. Given the uncertainties in measurement, spatial heterogeneity of bottom fauna and sediments, and the assumptions used in the calculations, these estimates can be considered virtually identical. In some respects this is surprising because the direct measurements include a component of NH: from excretion by bottom fauna. For example, based on NH: excretion rates for Yoldiu and Nephtys of 0.059 and 0.049 Fmoles/mg/dry wt./day (Harris, 1959) and average dry weights for these species at NWC of 0.048 and 0.036 gm, TABLEVII. FLUXESTIMATES~ OF NH; Type of

Summer (22")

AND

HP0:-

Fall (15')

Winter (4")

flux

Station

FOAM NWC

DEEP

estimateb NH:

DM DP PP DM DP PP DM DP PP

2.2 2.8 2.9 3.2 4.0 3.1 1.3 1.4 1.5

HP0:-

N/P

NH; HP0:-

N/P

NH:

0.10 0.07

22

0.84

15

0.32 0.17

10

0.26 0.43 0.39 0.28 0.25 0.42 0.19 0.24 0.20

1.1

-

0.20 0.059

-

6.5

1.4 2.2 1.4 1.5 0.33 0.54 0.71

0.057 0.085

-

0.40 0.025 0.020 0.032

-

5.5

17

H P O - N/P

Values in mmoles/m2/day. DM, direct measure; DP, diffusion predicted; PP, production predicted.

-

-

0.050

-

0.021 0.029

13

-

-

0.011 0.019

17

-

312

ROBERT C. ALLER

respectively (D. C. Rhoads, personal communication), contributions of about 2.8 pmoles/individual/day and 1.8 pmoles/individual/dayfor Yoldia and Nephtys can be expected. This means that at summer temperatures excretion could supply 1 mmole/m2/daygiven normal population densities at NWC (Yingst and Rhoads, 1978). The direct flux measurements may, therefore, be underestimating the true flux somewhat, producing agreement with the other estimates that is more apparent than real. Additional factors that may cause the direct measurements to be minima include lack of current or wave stirring (Hale, 1975), a decrease in aerobic activity due to slightly lowered 0, concentrations in the flux cores (Mangum and Van Winkle, 1973), or loss of NH: by biological consumption or nitrification. Production-rate estimates may also be somewhat low. It is known that macrofaunal reworking of sediments can stimulate microbial activity (Hargrave, 1970,1976; Fenchel, 1970; Aller, 1978; Aller and Yingst, 1978; Yingst and Rhoads, 1980). The lack of macrofaunal grazing and transport effects in the decomposition experiments can be expected to make the rates obtained minima. Overall, however, the similarity between the values obtained by the various methods of flux estimation is evidence that the measurement techniques are reasonably accurate. The agreement between the flux estimates from production rates [Eq. (6.29), Fig. 361 and other measurements shows that the flux is limited predominately, but not entirely, by the rate of reaction or production rather than transport effects (Aller, 1980). Essentially all the NH; being produced is escaping the sediments at an equivalent rate independent of biogenic reworking or burrow formation. This would not be true of an element such as Si whose production rate depends on the Si pore-water concentration and is therefore highly sensitive to transport. The fact that one-dimensional estimates of Fick’s first law agree well with other estimates without the necessity of using an apparent transport coefficient reflects the exponential attenuation of production rate with depth. If production rates were high at depth in the sediment, the horizontal flux into burrows would be correspondingly high. Irrigation of this NH; out of the sediments would then result in a higher net flux than could be accounted for by the vertical concentration gradient at the interface. It would be necessary to assign this excess flux to an apparent transport rate when in fact it simply reflects two-dimensional diffusion. In the present case, production rate is sharply attenuated with depth so that most of the NH: is formed near the sediment surface where it can diffuse predominately vertically out of the sediment and is accounted for in the one-dimensional gradient calculation. Equation (6.22), the so-

-

DIAGENETIC PROCESSES. I.

313

lution to the two-dimensional model for pore-water distributions, can be used to show that at FOAM and NWC as much as -10-40% of the total flux could come from radial diffusion into burrows, given the observed depth dependence in production rates. On the average, -70% of the flux can be accounted for by vertical diffusion alone at these stations. At DEEP, a somewhat higher percentage would be attributed to irrigation of burrows (radial diffusion) ranging from -32% during summer to -60% during fall and winter. Production-rate control of NH: fluxes has the important consequence that fluxes are highly seasonal and increase with increases in temperature (Fig. 51). This seasonality has been found in other environments as well (Hale, 1975; Rowe et al., 1975; Hartwig, 1976; Nixon et al., 1976). If it is assumed that the temperature dependence of the fluxes can be described by the expression Jo = J' exp( - E/kT), where Jo is the flux; J', the preexponential factor normalized to 22°C; TI the absolute temperature; E, the apparent activation energy; k, the gas constant; then it is found that E = 19.0, 22.7, and 16.3 kcaYmole at FOAM, NWC, and DEEP, respectively. Corresponding values of J' are exp(33.21, exp(40.3), and exp(27.9). The apparent activation energies are almost identical to that of 18 k 3 kcal/mole found for NH: production by bacteria (Aller and Yingst, 1980). This follows directly from Eq. (6.29) and is further evidence for production rather than transport control, of NH: fluxes. The expressions for temperature dependence are useful in calculating the yearly average fluxes of NH: from the bottom. Based on the direct flux measurements these are 0.87, 1.5, and 0.46 mmoles/m2/yrat FOAM, NWC, and DEEP assuming T = 285 - 10 sin wt, w = 2dyr as before. This shows that bottom areas differ in importance as NH: sources to overlying water. Protobranch-inhabited bottom areas such as NWC may be the most important source regions.

10

2 0 30

T ("C)

FIG. 51. The flux of NH: out of bottom sediment as a function of temperature (season) at FOAM, NWC, and DEEP.

314

ROBERT C. ALLER

Harris (1959) estimated the average daily N requirement of plankton in the central Sound for 1952-1954 to be 6.1 (1952-1953) to 5.1 (1953-1954) mmoles/m2/dayof which he could account for roughly 4.0-2.2 mmoles/m2/day,respectively, from zooplankton excretion. The remainder was believed to come from bacterial remineralization of organic matter in the water column or sediments. Depending on how average areas of bottom are assigned to corresponding flux rates like that of FOAM, NWC, or DEEP, it is apparent that bottom regeneration can account for much of the remainder, assuming continuity in N utilization between the periods of 1952-1954 and 1975-1976. This would represent about 10-30% of the daily requirement. Nixon et al. (1976) have argued that about half of the N regeneration on the bottom is in the form of organic N not NH: so that these estimates of the role of bottom regeneration of N in the water column N cycle are minima. Overall, the very similar range of values obtained for the required bottom regeneration rates (Harris, 1959), and the actual measured values of this study is good evidence that the N budgets for the central Sound were reasonably well described by the calculations of Harris (1959) and Riley (1956b).

6.9.2. Phosphate Fluxes. The direct measurements of phosphate fluxes and those predicted by use of Fick’s first law are also in good agreement. Only the fall estimates at NWC and summer values at DEEP are substantially different. This could be due to a dominance of the flux by macrofaunal excretion or to underestimates of HPOZ- pore-water concentrations due to oxidation and scavenging of P by Fe oxides during sampling (Bray et al., 1973). The contribution of HP0:- from excretion can be estimated from the data of Harris (1959) in the same way as for NH: to give values of -0.29 pmoles/individual/dayand -0.19 pmoles/individuaV day for Yuldia and Nephtys, respectively, at, for example, NWC. This means that a contribution from macrofaunal excretion as high as -0.1 mmoles/m2/daycould be expected at summer temperatures. This would be a substantial proportion of the observed flux. If the small differences between measured and calculated HP0:- fluxes are assumed real, then direct measurements tend to be higher than predicted during summer and somewhat lower than predicted from porewater gradients during fall and winter. The fall NWC sample is the obvious exception. The summer differences are consistent with an excretion contribution, whereas the fall and winter differences are consistent with a loss of HP0:- due to scavenging by Fe oxides at the sediment-water interface (Mortimer, 1941; Hayes, 1964). Such scavenging is expected from the concavities in HPOZ- profiles at these times. In general, how-

DIAGENETIC PROCESSES. I.

3 15

ever, agreement is good suggesting that like NH: , HP0:- fluxes are basically production-rate not transport controlled and that these production rates must attenuate rapidly with depth below the interface. This latter requirement follows from the lack of need for an effective transport coefficient to account for fluxes and the arguments made previously for the case of NH:. The temperature dependence of the measured fluxes give apparent activation energies E of 14, 26, and 24 kcal/mole for FOAM, NWC, and DEEP with corresponding preexponential factors J’ of 23.2, 45.0, and 41.5. Mean values for E and J’ of 22.6 and 38.5 kcal/mole are obtained by pooling all the flux data. Using these values, the mean yearly fluxes of HP0:- are 0.038, 0.12, and 0.076 mmoles/m2/yrat FOAM, NWC, and DEEP, respectively. Riley (1956b) estimated phosphate regeneration rates from bottom sediments by use of oceanographic data (also used in Harris, 1959, for N budget calculation). Average seasonal required bottom fluxes for 1952-1954 were calculated as 0.25,0.24, and - 0.02 mmoles/m2/dayfor summer, fall, and winter. The summer measurements made in the present study agree well with Riley’s calculations but are lower in fall and higher in winter than those he estimated. The average yearly plankton requirement of HP0:- was determined by Riley as 0.36 mmoles/m*/dayfrom 1952-1954. Based on the yearly mean HP0:- flux estimated previously, the bottom supplies from 10-30% of the requirement on a yearly average. Seasonal utilization of HP0:- by plankton varied from 0.13-1.1 mmoles/m2/day (Riley, 1956b) so that at certain times bottom fluxes can account for as much as 50-100% of the plankton requirement. This range of contributions of the bottom to overlying water-plankton HPO:.- utilization are similar to those reported for other shallow marine environments (Hartwig, 1976; Nixon et al., 1980). 6.10. Stoichiometry of Decomposition

The measured fluxes and pore-water concentrations can be used to obtain information on the relative composition of decomposing organic matter at different stages of breakdown and in different regions of Long Island Sound. It was demonstrated that the flux both of NH: and HP0:- are predominately production-rate controlled. This means that the ratio of the two fluxes can be used directly without correction for diffusion or adsorption to obtain limits on the N/P ratio in the organic material being metabolized near the sediment-water interface [from Eq. (6.29)]. These are listed in Table VII. Such ratios are only rough estimates because

316

ROBERT C. ALLER

the dissolved organic N and P fluxes, which can be substantial (Hartwig, 1976; Nixon et al., 1976, 1980), are not taken into account. Adsorption of HP0:- by iron oxides at the interface can also bias the ratios to higher N/P values. The mean NIP ratio obtained by pooling all stations is 13 2 6 with a range of 5.5-22. There are no obvious seasonal or spatial trends. Harris and Riley (1956) determined that the average N/P ratio of Long Island Sound plankton was 16.7 (phytoplankton) to 24 (zooplankton). Harris (1959) later determined that excretion of NH,+and HP02,- by zooplankton and bottom fauna in the Sound varied in ratio between 2.6 and 21.4, but was mostly between 7 and 15. The measured limits on N/P ratios determined from the fluxes are therefore generally lower than the source organic material and are in the range of the expected contribution from excretion (Harris, 1959; Johannes, 1968). In the lower range, the N/P ratios are in agreement with the expected preferential release of P relative to N during early stages of decomposition (Grill and Richards, 1964; Johannes, 1968). This release leaves organic material depleted in P relative to N as substrate for later diagenesis in the sediment body (Sholkovitz, 1973). For example, Martens et a2. (1978) determined that below the bioturbated zone at FOAM, the mean N/P ratio of decomposing organic matter is 33. This is substantially higher than the 15-22 found in this study for surface sediments from FOAM. Unlike the N/P ratios, there is definite evidence that the average C/N ratio of decomposing organic matter in the upper 10 cm of sediment increases with the bathymetric depth of the station. Evidence for this comes directly from the decomposition-rate experiments at FOAM, NWC, and DEEP (Figs. 35 and 36). The ratio of the average SO:--reduction rate R , to the average NHZ-production rate &/(I + K ) in the upper 10 cm at each station can be taken as identical to the expected C/N ratio of decomposing organic matter. This assumes K 1 (Rosenfeld, 1979) and the stoichiometry of SO:- reduced to carbon oxidized is CIS = 2 (Goldhaber and Kaplan, 1974). The experimental ratios are 5.5, 6.7, and 10.7 at FOAM, NWC, and DEEP, respectively (after Goldhaber et al., 1977; Rosenfeld, 1977; Aller and Yingst, 1980). This means that C/N ratios of decomposing organic matter change from values less than the Redfield ratio of 6.6 at FOAM to values much greater at DEEP, indicating a progressive depletion of N relative to C offshore. Martens et al. (1978) also found evidence of increasingly lower C/N ratios of metabolized organic matter at stations inshore from FOAM thereby extending the general trend shoreward. The RJR, ratio is known to increase with depth in the sediment at individual stations as well (Aller and Yingst, 1980). Many studies have demonstrated preferential release of N relative to C during early decomposition of organic matter in marine waters or sed-

-

DIAGENETIC PROCESSES. I.

317

iments (for example, Waksman, 1933; Anderson, 1939; Kaplan et al., 1963; Krause, 1959; Sholkovitz, 1973; Hartman et al., 1973). It seems likely that the observed increase of C/N in both station water depth and depth in the sediment reflects increasing time after the start of decomposition in each case. The increasing water depth could allow a substantial decrease in N content of planktonic debris hitting the interface due to the known rapid initial loss of N constituents from decomposing plankton (Krause, 1959). The increase of the ratio with depth in a deposit represents the slower long-term preferential release commonly observed in studies on sediment diagenesis (Sholkovitz, 1973; Hartman et al., 1973; Martens et al., 1978). Solute transport, adsorption, and precipitation reactions can make it difficult to reconstruct decomposition stoichiometries from pore-water profiles alone, particularly within the bioturbated zone. Because the general trends and limits on N/P and N/C stoichiometry are evident from the previous considerations no further modeling will be done here. An adequate explanation for the higher NH:/HPO:- -concentration ratios in the deeper pore water at inshore relative to offshore stations (Fig. 14) requires decomposition experiments below 10 cm, together with direct examination of the solid phase for phosphate compounds and solute transport modeling. 7. SUMMARY (1) Pore-water profiles of SO:-, alkalinity, HPOi-, and NH: in the top 1 m of sediment differ radically in different depositional environments of Long Island Sound. The greatest evidence for extensive sulfate reduction and organic matter decomposition as measured by standing concentrations of pore-water metabolites is found in shallow water along the north coast of the Sound; in deeper water the buildup of decomposition products or depletion of SO:- in sediment pore water is greatly lowered. (2) In contrast to pore-water data, analysis of the solid phase shows extensive and comparable production of sulfide fixed as FeS (acid-volatile sulfide) in all areas. Deep-water stations show extensive loss of solidphase sulfide either prior to or after conversion of FeS to FeS2 (pyrite). All stations have comparable standing crops of organic matter and all stations are capable of supporting approximately the same rate of sulfate reduction in the upper 10 cm of sediment where most decomposition takes place. (3) All stations show similar seasonal variation in pore-water profiles near the sediment-water interface. These are periodic and repeatable from year to year. During the late spring and summer, microbial activity

318

ROBERT C. ALLER

results in high production of NH: , HPOi-. and alkalinity throughout the sediment column but particularly near the interface. Because decomposition rates often show exponential decrease with depth below the interface, seasonally high production often results in maxima in pore-water concentrations a few centimeters below the sediment surface. In late summer and in the fall, biogenic reworking, particularly fluid transport associated with solute diffusion into burrows, results in extensive exchange of sediment pore water with overlying water and a general net lowering of NH: , HPOZ,-, and alkalinity concentrations or raising of SO:- from the earlier summer levels. During the winter, pore-water profiles reflect greatly decreased biological influence. The year can therefore be divided into a relatively warm period (spring, summer, fall) of different types of apparent biological dominance of diagenesis followed by a shorter cold period (winter) when processes are more physically controlled. (4) The oscillatory behavior of pore-water profiles near the interface can be accounted for by the different temperature dependence of microbial and macrofaunal activity that control production (consumption) and biogenic transport processes respectively. A non-steady-state model is used to show this is the case. (5) A two-dimensional transport-reaction model incorporating both radial transport into burrows and vertical diffusion is presented. This model is capable of predicting both the form and magnitude of pore-water profiles extraordinarily well at all stations. A one-dimensional model in which an effective transport coefficient is used to account for the influence of reworking and burrow construction on solute movement is far less satisfactory in predicting the observed profiles. (6) The buildup of pore-water constituents in the bioturbated zone is extremely sensitive to the number and, to a lesser extent, size of burrows present. This demonstrates that at any given time sediment chemistry may vary in accordance with the abundance, individual size, and relative mobilities of the animals present. The often-observed decrease in population abundance during winter, for example, may in part be responsible for the reduction in biological influence on pore-water profiles at that time. The spatial variation in pore-water profiles in the Sound are shown to be consistent with the bathymetric variation in the animal communities present. (7) Equilibrium calculations indicate that HP0:- concentrations may in some cases be saturated with respect to various Ca, Mn, and Fe phosphates. Saturation states vary seasonally. Because of microenvironments created by burrows, the use of average pore-water concentrations to examine saturation states in the bioturbated zone can give misleading

DIAGENETIC PROCESSES. I.

3 19

results. Direct examination of the solid phase together with transportreaction modeling are necessary to correctly evaluate equilibrium controls on pore-water distributions within this sediment zone. (8) The flux of NH: and HPOZ- out of the bottom varies seasonally and is highest in summer, lowest in winter. Three independent methods were used to estimate fluxes: direct measurement, pore-water concentration gradients, together with Fick’s first law, and in the case of NH:, reaction rates within the sediment. These estimates are in good agreement. Because most decomposition takes place near the sediment-water interface, the fluxes of these constituents are not greatly influenced by burrow construction and can be adequately predicted by molecular diffusion and vertical concentration gradients in the upper few centimeters. The flux of NH: is shown to be limited by production rate and relatively insensitive to transport regime. During fall and winter, HPOZ- fluxes may be lowered slightly due to adsorption on Fe-oxides at the sediment water interface. On a yearly average basis, bottom fluxes are capable of supplying 10-30% of the daily N and P requirement of plankton in the overlying water. A higher percentage may be supplied by the bottom during specific seasons. (9) The flux measurements show that P is released preferentially to N during early diagenesis. Decomposition experiments demonstrate that N is released preferentially to C near the sediment surface. These observations are in agreement with previous studies on decomposition. (10) This study demonstrates that the chemical composition of the bioturbated zone of marine sediments, like that of any natural body, represents a balance between transport and reaction processes taking place within and around it. The influence of macrofauna on solute transport in this zone derives largely from the change in the geometry of diffusion brought about by burrow construction and irrigation. Depending on the reactions governing the behavior of a given element, this change may or may not influence the netflux of its dissolved form between sediment and overlying water. In contrast, the buildup of pore-water solute concentrations is highly sensitive to the transport regime and the presence of infaunal communities.

320

ROBERT C. ALLER

APPENDIXA. MACROFAUNA(>1 mm) TABLEA l . FAUNAL ~~

Sample Flux-

FOAM

17 Nov. 1975 Species Bivalvia Mulinia lateralis Pitar Morrhuana Tellina agilis Macoma tenta Nucula annulata

Nohample

No./mZ

Nohample

Noh2

8

269

I

34

I

3 1

4

Yoldia limatula Gastropoda Nassarius rrivitatus Retusa sp. Polychaeta Spiochaetoptzrus oculatus Pherusa afinis Pista palmata Melinna cristata Clymenella sp. Nephiys incisa Sigambra tentaculata Phyllodoce sp. Crustacea Ampelisca sp. Pagurus sp. Cumacean A. Ostracoda A. Hydrozoa, Corymorpha pendula

19 March 1976

-

-

101 34 134

34 34

I

34 34

1

-

34 34

-

34

-

-

269 34

134 134

-

34 72

2417

1 1

34 34

-

-

-

34 34

321

DIAGENETIC PROCESSES. I.

SIEVED FROM

FLUX-COREBOXES

SAMPLES(21 mm) Core Box

NWC 23 March 1976

DEEP 12 Nov. 1975

14 April 1976

Nohample

No./mZ

Nohample

NohZ

No./sample

No./mZ

-

-

672

II

369

-

-

20 (<2 mm) 8 2 1

67 34

34 134

-

(-I mm) 8

269

235

-

-

269

-

-

-

-

1

7 (large)

235

4 (large) 1

-

-

2 3 5

101

1

67 I68 34

322

ROBERT C. ALLER

APPENDIXB. BOX-COREAND GRAVITY-CORE DATA FROM LONGISLAND SOUND Symbols meq/liter M mM CLM pmoles g (gm)

0

milliequivalents/liter moles/liter millimoles/liter micromoles/liter micromoles gram; in the case of solid-phase metal analyses, this is grams of ash where ash refers to residue from ignited sample (organic free) refers to questionable analysis ~

~~

Addirional analysts''

so: -

NH: (Kjeldahl) CaCO,

M. B. Goldhaber, A. Ruggiero (FOAM-], NWC1, NWC-2); J. K. Rosenfeld (FOAM-2, FOAMG) M. B. Goldhaber (FOAM-1, NWC-I) M. B. Goldhaber (FOAM-I) J. K. Rosenfeld (FOAM-2) M. B. Goldhaber (FOAM-1); B. Brockett (DEEP-

FeS Cu, FeSL

M. B. Goldhaber, A. Ruggiero (FOAM-1) M. B. Goldhaber (FOAM-1)

Alkalinity

HS

1)

a For the sake of completeness, data for FOAM-1 and FOAM-2 published in Goldhaber et al. (1977) are also listed here.

TABLEB1. FOAM-1 Box CORE, 5 AUGUST1974 (T = -20°C)"

0- 1 1-2 2-3 3-4 4-5 5-6 6-7 7-8 8-9 9-10 10-1 1 11-12 12-13 13-14 14-15 15-16 16-17 17-18 18-24

-

-

22 63 91 87 75 73 131 143 134 287

20.9 19.6 20.6 20.8 20.2 20.9 20.9 20.9 (13.9) 19.7 17.7 18.7 18.6 18.7 20.0 17.4 15.8 17.0 15.2

5.8 6.6 6.7 6.6

6.3 8.5 6.9 7.9 8.2 8.9 9.5 9.6 10.2 10.3 10.0 11.0 11.7 11.7 12.8

In part after Goldhaber et a/., 1977.

161 243 207 204 193 160 157

230 168 181 235 243 217 223

237 174 136 98.8 87.7 40.2 76.2 37.7 22.0 18.6 15 16 12 14 13 8.2 11 11 11

35.3 40.1 15 2.3 6.8 1.8 4.8 2.3 2.3 1.8 2.3 1.6 2.7 2.1 1.9 2.9 4.1 2.7 3.2

4.17 3.87 3.46 3.44 3.22 3.24 3.39 2.79 2.62 2.57 3.05 3.28 3.07 3.10 3.25 4.78 4.89 3.66 3.53

10.3 11.1 17.8 12.5 10.8 31.6 (30) 22.2 29.0 31.5 27.4 29. I 25.2 26.3 23.4 19.6 14.4 12.4 9.6

7.5 11 8.1 12 23 20 18 13 11 12 11 7.2 8.4 6.6 7.2 10 10 10 7.5

90 110 1180 200 190 I70 I80 170 180 180 I70 210 260 310 343 381 334 340

28.0 25.8 26.9 28.5 28.5 22.9 24.7 20.0 20.1 20.6 19.7 24.0 25.2 27.2 28.7 35.0 33.7 30.2 30.5

599 573 58 1 567 588 520 518 455 440 465 477 569 614 67 1 747 906

885 757 71 1

TABLEB2. FOAM-2 Box CORE( T = 3.0-3.5"C)o Pore water

cI1

0- 1 1-2 2-3 3-4 4-5 5-6 6-7 7-8 8-9 9-10 10-11

56.4 50.2 49.8 50.9 54.7 53.9 39.6 38.5 37.5 38.7 36.9

7.36 7.17 7.09 7.00 6.92 6.96 7.14 7.33 7.45 7.55 7.53

22.8 22.2 22.4 22.1 21.7 21.1

-

19.5 19.7 19.1 18.6

71.2 124 181 245 322 361 442 453 542 590 63 I

84

135 240 264 283 328 407 514 530 652 628

12.6 52 65 91 151

258 309 23 1 308

294 288

59.0 204 193 I85 262 286 249 22 1 155 142 I22

3.85 278 385 743 1022 363 170 24.9 5.4 11

16

3.47 4.25 4.96 5.16 6.31 7.39 7.43 8.24 9.00 9.51 9.95

* NH; analysis on 5-ml samples by J. Rosenfeld using Kjeldahl distillation, good correspondence with phenol-hypochlorite method (0.1-0.2 ml) is found. ' H2S detectable by smell below 9 cm. In part after Goldhaber et al., 1977.

TABLEB3. FOAM3 Box CORE,8 JULY1975 (T = 18°C) Pore water

~

N VI

0- 1 1-2 2-3 3-4 4-5 5-6 6-7 7-8 8-9 9-10 10-1 I 11-12 12-14 14-16 16-18

~~

48.2 42.4 40.8 39.9 39.9 41.9 43.7 45.9 45.9 49.2 55.7 60.0 62.4 55.3 48.7

~~

7.25 7.47 7.43 7.34 7.31 7.27 7.30 7.34 7.39 7.41 7.43 7.41 7.49 7.49 7.55

- 160

0.434 0.434 0.427 0.432 0.429 0.434 0.436 0.436

- I85 - 200 - 180 - 210 - I80 - 180 -210

0.436 0.437 0.436 0.438 0.440 0.437

380 50 30 60 20 40 - I45

-

22.2 21.8 21.9 21.8 21.6 21.7 20.6 21.2 21.0 21.1 20.8 21 .o 20.7 20.3 20.0

3.42 3.72 4.06 4.48 4.56 5.56 6.00 6.07 6.10

-

7.24 6.62 7.73 7.39 8.31 Bottom water:

145 205 269 318 331 378 468 398 439 520 509 504 547 565 614 2.0

13 20 32 82 80 I24 142 I37 I09 127 159 100

159 149 I63 2.8

~~

348 322 303 I59 138 I17 94. I 85.2 76.4 60.4 50.1 50. I 41.7 33.7 24.4

~

2.5 52.6 (6.5) 42.8 21.7 7.3 3.8 2.3 0.77 0.91 2.7 6.8 2.3 1.3 2.0

TABLEB4. FOAM-^ Box CORE, 5 NOVEMBER 1975 (T = 14-14.5"C) Pore water

0- 1 1-2 2-3 3-4 4-5 5-6 6-8 8-10 10-13 13-16

51.6 46.6 44.8 42.1 41.3 40.8 40.3 40.6 49.9 54.9

7.37 7.37 7.46 7.48 7.49 7.47 1.43 7.42 7.35 7.31

-

13.1 11.6 10.5 10.2 10.1 9.8

0.427 0.430 0.430 0.431 0.430 0.429 0.43 1 0.432 0.436 0.439

7.90 7.93 7.96 7.% 7.87 8.11 8. I4 8.08 8.14 8.23

22.7 22.2 22.4 22.3 22.3 22.7 22.3 22.8 22.5 22. I

2.64 2.80 2.94 2.95 3.05 3.03 3.15 3.11 3.25 4.17

Bottom water:

71.4 I30 133 139 143 143 161 173 170 2%

17 33 45 46 55 55.1 55.6 52 48 79

0.3

1.2

157 135 48.4 35.7 54.6 47. I 45.5 38.2 30.4 30.4

I .3 4.3 6.4 2.9 3.4 2.0 1.6 2.0 2.0 2.7

TABLEBS. FOAM-5Box CORE, 18 MARCH1976 (T = 2.2"C)

Pore water Depth w 4 N

0- 1 1-2 2-4 4-6 6-8 8-10

c1-

Ca2

so:-

Alk

PS

(MI

(mM)

(mM)

(mesfliter)

NH: (wW

HP0:-

PH 7.34 7.37 7.25 7.25 7.25 7.34

-

0.410 0.415 0.422 0.428 0.431 0.432

7.90 8.17 9.63 9.15 9.08 8.17

21.o 21.6 21.7 21.8 21.4 21.1

3.26 3.71 4.56 4.82 5.48 5.78

39.6 79.1 160 227 302 393

16.6

H20

(cm) 54.8

46.2 41.1 41.2 45.6 47.3

16.3 9.7 9.5

+

Bottom water:

0.1

Mn

(14 (CLM)

Fe (CLM)

89.3 181 164 137 109 69.2

0.50 1.3 58.8 38.5 4.04 2.81

(6.6) 29.8 77.7 127 180

0.68

TABLEB6. Box CORE,NWC-I 21 JULY 1974 (T = 20°C) Pore water

: m

Depth (cm)

HzO (%)

so:-

0- 1

72.8 64.9 65.4 66.3 67.7 61.8 59.2 61.5 65.1

(16.0) 23.2 21.5 22.3 21.9 (17.2) (17.1) 21.9 (18.2) (15.6) (17.9) (17.6) (17.7) 21.5 (19.9) 21 .o

1-2 2-3 3-4 4-5 5-6 6-7 7-8 8-9 9-10 10-1 1 11-12 12-13 13-14 14-15 15-16

64.0 55.6 55.7 54. I 54.9 54.9 55.4

(mM)

Alk (medliter) 3.7 4.5 4.4 4.7 4.4 4.3 3.9 4.0 3.7 3.7 3.8 3.7 3.6 4.0 . 3.8 3.9

NH;

Solid phase

HPOZ-

(t.W

(PMl

190

153 294 235

349 349 392 370 326 270 290 270 278 269 287 277 299 304 289

190

144 135 87 119 89 87 74 69 67 87 87 81

Mn OIM) 26 I 189 172 76.8 56.4 43.9 47.0 44.2 31.9 30.2 23.7 29.3 25.7 34.2 33.9 36.2

Fe

Organic matter

(FM)

(W

Fe (mp/gm)

13.3 33.7 12.1 8.30 4.89 3.46 3.04 2.94 2.13 0.82 0.66 1.88 0.82 1.83

6.32 6.43 7.17 7.60 7.72 6.03 5.92 5.91 6.13 4.75 4.62 3.72 3.76 3.82 4.01 4.52

39.5 38.5 39.1 39.8 37.3 26. I 32.7 38.4 38.5 33.6 35.6 35.5 35.8 36.8 35.3 34.3

1.15

1.3

Mn (pdgm)

929 718 702 639 588 566 627 680

686 566 570 543 568 560 5% 568

TABLEB7. NWC-2 Box CORE,11 NOVEMBER 1974 (T

' W

0-1 1-2 2-3 3-4 4-5 5-6

6-7 7-8 8-9 9-10 10-11 11-12 12-13 13-14 14-15 15-16 16-17 17-18

72.7 66.5

62.4 56.1 54.6 55.2 52.8 49.2 47.1 49.9 50.4 49.7 51.3 53.1 50.9 54.2 52.5 53.6

7.42 7.47 7.35 7.40 7.58 7.51 7.51 7.47 7.62 7.55 7.58 7.58 7.57 7.49 7.44 7.47 7.47 1.47

22.9 23.5 22.7 23.2 22.9 23.3 23.2 23.4 23.6 23.8 23.5 23.3 22.8 23.3 23.8 24.3 23.8 23.8

3.16 3.62 3.24 3.36 3.37 3.32 3.38 3.46 3.32 3.28 3.16 3.19 3.02 3.00 3.04 3.00 2.92 2.98

122 141 115

142 145 158 158 164

175 185 195 181 162 184 159 150 146

I51

13 28.8 43.6 55.3 54.3 54.8 55.8 53.0 52.7 50.5

52.1 48.7 42.6 46.8 41.5 39.7 36.7 38.3

193 193 116 57.7 40.8 31.1 27.1 30.8 27.7 25.5 14.9 24.0 2.4 15.3 20.8 20.6 19.5 7.5

0.77 1.2 9.76 5.98 1.1 2.61 1.3 1.9 1.3 1.6 0.73 1.1

0.52 0.84 0.84 0.84 0.97 0.52

6.81 6.79 5.97 4.83 4.61 6.24 4.94 4.75 3.87 4.12 4.70 4.80 4.50 4.43 4.16 3.93 4.09 4.42

=

13.2")

3.82 5.78 5.33 7.05 7.76 7.51 6.69 6.66 9.61 7.34 8.49 4.69 5.89 4.91 6.40 5.71 4.47 3.93

0.6 9.04 13.7 16.5 14.0 9.36 10.6

80 93 %

104 125 146 168

36.6 36.7 34.3 30.3 31.5 31.9 30.7 33.1 28.8 32.2 30.8 33.4 32.4 30.4 33.7 29.4 32.0 34.4

1326 72 1 583 518 516 526 509

538 464 494 504 529 489 526 499 489 508 561

TABLEB8. NWCJ Box CORE,18 MARCH1975 (T = 3°C) Pore water

so:-

Depth (cm)

HzO (%)

PH

(mM)

Alk (meqfliter)

0-1 1-2 2-3 3-4 4-5 5-6 6-7 7-8 8-9 9-10

72.2 65.8 63.4 61.3 51.6 47.1 46.6 48.6 48.4 42.9

7.34 7.51 7.55 7.49 1.44 1.46 7.41

22.4 23.0 22.3 22.0 22.5 22.6 22.3 22.4 22.8 22.5

2.61 2.93 3.09 3.23 3.29 3.31 3.39 3.41 3.44 3.55

1.50

7.43 7.43

NH;

Solid phase HP0:-

Mn

Fe

Organic matter

S-FeS

(FW

(PW

(PW

(PW

(%I

(did

34 78 94.2 I29 136

2.5 20 42 63 58 69 65 78 52 52

35.7 I65 135 113 86.8 63.7 50.4 47.5 52.2 26.9

0.48 16.6 19.3 5.9 1.8 0.91

6.36 6.48 6.48

0.00 2.1 19.0 23.3 15.2 8.45 5.71 4.99 3.18 1.7

160

167 165 I76 188

1.1

1.1 1.2 0.57

5.55

4.37 3.85 3.79 3.66 3.99 3.61

S-FeS, Fe Mn (CL/gm) (mg/pm) ( d g m ) 83 79 103 119 119 153 168 172 156 150

38.9 39.0 39.6 36.4 30.5 31.3 31.3 31.5 30.8 25.5

1325 860 763 689 579 569 517 553 537 414

TABLEB9. NWC-4 Box CORE,16 JULY1975 ( T = 19T) Solid phase

Pore water

w W

0-1 1-2 2-3 3-4 4-5 5-6 6-7 7-8 8-9 9-10 10-1 1 11-12 12-14 14-16 16-18

71.6 64.2 62.6

-

55.9 54.5

-

52.1 50.3 48.0

-

51.2 54.2 53.8 54.1

7.32 7.25 1.29 7.33 7.21 7.27 1.27 1.25 1.21 1.23 7.23 1.23 1.23 7.21 7.19

190 90 60 65 80 60 0 -95 35 -90 - loo

*

-90 -90 -90 - 120 - 150

0.443 0.442 0.438 0.438 0.438 0.438 0.431 0.437

23.2 22.3 21.9 22.1 22.6 21.8 22.3 22.0

2.89 3.27 3.32 3.35 3.65 3.32 3.52 3.30

182 186 203 195 199 191 190

26 72.8 96.8 102 97.3 82.4 81.9 57.9

251 217 117 89.7 98.3 61.9 53.3 45.0

1.4 33.5 27.8 13.8 9.20 3.6 2.1 3.9

0.431 0.436 0.437 0.436 0.431 0.438 0.439

21.9 22.3 21.6 (19.5) (19.5) 22.0 22.5

3.25 3.39 3.40 3.36 3.16 3.04 3.01

160 162 173 155 156 138 135

52.0 51 52 49 41 38 34

40.0 34.6 30.6 28.0 25.5 24.8

1.4 1.2 1.4 1.3

2.4

1.6

Bottom water:

160

0.68

1.1 0.82

6.71 6.31 6.15 5.72 5.63 5.38 4.88 4.46

0.097 7.36 16.9 18.6 12.0 10.1

-

6.86

98 98 I03 I13 I65 163 174 154

39.1 38.0 36.2 33.9 36.8 35.6 34.5 32.3

1277 823 645 597 601 580 548 524

TABLEB10. NWC-5 Box CORE, 29 OCTOBER1975 (T = 16.4"C) Pore water

0- 1 1-2 2-3 3-4 4-5 5-6 6-8 8-10 10-13 13-16

67.4 62.4 58.0 55.3 54.1 54.9 54.4 54.8 57.9 58.6

7.37 7.44 7.44 7.46 7.43 7.43 7.46 7.49 7.43 7.47

-

-

14.3 14.1 13.6 13.0 12.8 12.5 12.6

0.438 0.440 0.440 0.440 0.440 0.436 0.43 0.443 0.443 0.446

8.26 8.41 8.47 8.32 8.23 8.35 8.35 8.17 8.08 8.11

23.2 23.3 22.5 22.8 22.4 23.1 22.7 22.8 22.8 22.9

2.43 2.55 2.60

2.63 2.16 2.48 2.14 2.58 2.74 2.79

Bottom water:

64.9 92.2 103 104 115 112 111 81.6 104

111 0.2

6.8 14 35 38 38 26 29 20 21 24

2.5

175 141 74.6 51.0 41.3 31.5 30.9 29.3 25.3 19.7

0.86 1.4 0.86 1.2 0.91 1.1 1.o 0.86 1.1 1.o

TABLEBII. NWC-6 Box CORE,23 MARCH 1976 (T = 3.2”C) Pore water

Mn

Depth (cm) 0- I W W

1-2 2-3 3-4 4-6 6-8 8-10 10- I2

(PM)

68.2 61.4 60.0 58.7 55.3 56.7 57.9 54.1

7.11 7.24 7.22 7.31 7.29 7.30 7.31 7.29

13.2 11.1 10.5

0.416 0.416 0.418 0.419 0.417 0.423 0.425 0.426

8.10 7.90 8.10 7.76 7.62 7.83 7.62 8.10

21.8 21.7 22.1 22.2 21.8 22.2 22.3 22.0

2.39 2.36 2.49 2.56 2.61 2.60 2.66 2.64

Bottom water:

18.6 33.4 42.2 53.4 68.9 83.4 %.5 107 0.9

7.0 12.3 23.7 31.2 29.6 31.9 33.6 35.1 1.05

35.7 64.4 65.3 61 .O 37.3 27.1 25.1 24.2

0.59 9.63 6.99 5.28 0.92 0.68 0.55 0.59

TABLEB12. DEEP-1 Box CORE,23 JULY 1975 (T = 185°C) Pore water Depth H20 (cm) (%)

pH

69.5

6.88

0-1

Eh (mv) 30

Cl(M)

Alk (meql (mM) liter)

SO:-

Solid phase

NH;

HP0:-

Mn

Fe (I*M) (CLM)

(CCM)

(+M)

62.9

12

135

S-FeS S-FeS2 Organic matter CaCO, ( p d (pd (%) (%) gm) gm)

Fe (mpl gm)

(pd

Mn gm)

5.37

5.1

0.62

50

32.6

753

54.6 30.4

4.98 5.40

4.3 4.6

4.43 9.92

50 53

33.6 32.2

567 532

46.2

5.47

3.5

11.0

62

35.1

579

6.19

-

22.7

56

34.2

61 I

48.3

6.25

-

30.7

72

36.6

684

231

13.1

5.89

0.0

30.8

115

36.5

663

121

228

27.9

6.02

-

23.3

100

37.1

6%

120

39

110

5.44

5.87

-

24.6

112

36.8

677

3.09

128

77

135

2.5

6.26

2.1

16.0

115

36.7

679

3.69

123

73

165

2.2

5.89

-

13.9

115

37.9

727

0.446

23.3

2.49

0.448 0.448

23.1 22.8

.2.47 2.65

95 115

34 51

128 109

0.449

22.7

2.77

165

26

125

0.446

22.7

2.88

144

%

117

0.446

22.9

3.40

160

117

202

0.446

22.4

3.37

150

(19)

0.448

22.7

3.22

160

0.448

23.0

2.95

0.447

22.3

0.447

22.5

0.77

f 10

1-2 2-3

53.7 56.3

6.97 7.06

3-4

57.5

7.12

- 45 -75

f40 W

W p

- 88 rt 38

4-5

56.8

7.37

5-6

55.0

7.33

- I28 rt 27 - 145

4.55

240 6-7

55.2

7.31

7-8

53.7

7.38

8-10

54.4

7.43

10-12

52.2

7.39

12-14

49.3

7.35

- 153 2 23 - 140 2 55 - 125 f40 - 135 f 30 -115 It 30

TABLEB13. DEEP-2 BOX CORE,12 NOVEMBER 1975 ( T = 14°C)

Pore water

0- 1

W

v,

1-2 2-3 3-4 4-5 5-6 6-8 8-10 10-13 13-16

70.6 59.0 58.6 57.8 59.1 57.6 57.3 54.4 54.3 49.7

7.05 6.96 7.00 7.00 7.16 7.33 7.34 7.43 7.41 7.43

13.9 13.5 12.7 13.2

0.441 0.441 0.441 0.443 0.442 0.441 0.442 0.442 0.442 0.443

8.11 8.29 8.26 8.26 8.08 8.14 8.26 8.17 8.29 8.17

23.1 23.3 23.0 22.9 22.5 22.8 22.7 22.9 22.7 22.6

2.29 2.32 2.56 2.45 2.57 2.98 3.01 3.24 3.20 2.99

Bottom water:

27.2 48.6 62.7 65.O 77.7 108 111 117 120 127 0.1

6.9 5.7 9.6 17 25.3 34.9 47.1 47.5 41.7 44.5 1.15

109 120 76.1 64.6 59.3 61.0 70.1 69.7 56.4 46.2

1.o 25.2 66.4 19.5 10.7 7.86 7.54 2.7 1.5 1.5

TABLEB14. DEEP-3 Box CORE,13 APRIL1976 (7' = 5.8"C) Pore water

m

Depth (cm)

HzO (%)

PH

0- 1 1-2 2-3 3-4 4-5 5-6 6-8 8-10 10-12 12-14

69.1 63.8 62.6 60.5 59.3 57.7 54.3 54.5 53.6 52.5

7.04 6.91 6.91 7.17 7.11 7.21 7.44 7.28 7.42 7.43

PS

15.6 12.9 12.7 12.6 12.2

CI-

Caz+

so: -

(W

(mM)

(mM)

Alk (meqhiter)

0.415 0.418 0.422 0.422 0.424 0.431 0.423 0.423 0.430 0.43 1

8.03 8.14 8.32 8.20 8.29 8.55 8.17 8.58 8.11 8.14

21.1 22.3 22.0 22.4 22.0 22.0 21.9 22.0 22.0 22.2

2.36 2.21 2.28 2.54 2.76 3.36 3.58 3.88 4.06 4.12

Bottom water:

HP0:(PW

31.3 33.2 52.1 72.0 86.1 116 132 144 176 175

5.16 8.75 18.5 34.6 49.4 I03 I43 147 190 158

0.7

1.21

Mn (kW 27.1 71.4 131 157 I77 249 187 190

198 202

Fe (PW

0.87 52.3 106 85 95 73.5 16.2 9.14 2.03 1.90

TABLEB15. FOAM-G GRAVITYCORES Pore water

July 8, 1975 Depth (cm) 0-5 5-10 10-15 15-20 20-25 25-30 30-35 35-40 40-45 45-50 w

W . l

PH

Ca2 (mM)

7.36

8.26

7.47

7.%

+

105-1 10 110-115

115-120 120- I25 125- I30 130-135 135-140

(PM)

Hm:(PM)

Mn (CLM)

(PW

Depth (cm)

PS

3.88

0.253

42

18.8

13.3

0-5

11.5

9.04

0.618

156

25.9

2.5

10-15

10.2

13.42

0.981

219

15.5

3.4

30-35

10.2

19.44

1.15

224

3.5

6.24

60-65

9.7

20.86

1.45

215

2.5

3.4

90-95

9.4

25.66

I .56

250

3.3

8.2

120-125

9.4

26.12

1.58

234

2.2

2.2

27.58

1.76

219

2.0

2.2

29.28

1.82

242

I .5

4.7

30.58

I .90

239

I .5

6.3

32.74

2. I5

245

2.0

2.7

34.0

2.12

245

I .8

1.7

34.80

2.23

252

1.5

2.0

35.00

2.41

26 I

I .3

2.9

Alk (meqfiiter)

NH:

Fe

19.4 16.9 7.59 13.0 6.92

t

50-55

55-60 60-65 65-70 10-15 75-80 80-85 85-90 90-95 95-100 100-105

so:(mM)

October 14, 1975

10.9 6.64 8.8 8.85 7.5

1.4

6.30 4.9 6.15 3.6 6.00 2.1

J.

5.88 1.6 5.85 0.8 5.88 0.4 5.87 0. I

5.37 0.4

TABLEB16. NWC-G GRAVITY CORES

Pore water November 11, 1974 Depth (cm)

W

g

0-5 5-10 10-15 15-20 20-25 25-30 30-35 35-40 40-45 45-50 50-55 55-60 60-65 65-70 70-75 75-80 80-90 90-100 100-110 130-140

Ca2'

so: -

(mM)

(mM)

8.31

22.4 22.8 22.1 (23.2) 21.9 21.9 22.0 22.0 22.0 21.6 21.9 21.2 21.5 21.5

-

7.39 8.21 7.78 7.78 8.10 8.06 7.76 7.98 8.13 8.11 7.66 7.98 8.21

-

21.1 20.1 19.8 19.6 17.7

Alk (meqhiter) 3.40 3.21 3.19 3.44 3.75 4.21 4.46 5.00 5.08 5.26 5.70 5.98 6.01 6.15 6.42 6.68 6.60 7.16 7.62 9.66

NH:

September 16, 1975

HPOZ-

(PW

(FW

183 193 192 220 26 1 299 317 354 381 400

58.3 35.5 34.9 37.6 54.0 71.5 74.2 81.6 79.5 92.7

428

97.8

458 464

98.6 103

489 515 574

109 109 115

Mn ((LICI) 122 32.6 31.7 27.7 25.3 29.7 26.2 25.5 22.2 27.3 25.1 24.2 25.0 24.4 23.0 23.7 21.7 21.9 21.5 23.7

Fe

(PW 1.3 0.8 0.8 1.5

1.0 1.4 1.5 1.3 1.4 2.6 1.3 1.6 1.4 2.5 2.0 1.9 2.0 1.9 2.9 3.0

Depth (cm)

PH

3-8 18-23 43-48 68-73 78-83 88-93 93-98 98-103

7.30 7.13 7.14 7.18 7.15 7.12 7.21 7.20

October 29, 1975 Depth (cm)

PH

PS

30-35 50-55 80-85 110-115 150-155

7.38 7.37 7.36 7.33 7.30

12.5 12.5 12.4 12.6 12.9

TABLEB17. DEEP-G GRAVITY CORE,12 NOVEMBER 1975 Pore water Depth (cm) 0-10 10-20 20-30 30-40

PH 7.22 7.23

PS

-

c1-

Ca2

so:-

(M 0.444

(mM)

(mM)

7.78 8.03 7.% 8.03

22.3 22.7 22.8 22.4 22.1 22.1 22.5 22.5 22.3 22.4

0.447 0.447 0.441

7.%

40-50

50-60 60-70 70-80 90-100 100-1 10

+

0,444 0.447 0.446

8.08 7.93 8.93 8.08 8.16

Mn

Fe

(CLW

(PW

(CLM

53.1 39.4 27.4 36.9 53.3 59.6 57.4 66.6 54.0 61.5

66.3 46.8 21.7 28.6 46.1 47.6 49.5 63.2 46.6 46.8

4.34 1.6 1.7 1.1 1.o 1.3 1.1 1.1 1.1 0.8

HP0:3.% 3.72 3.44 3.58 3.95 4.02 3.95 4.08 3.59 3.44

186 159 159 156 163 151 152 104 95.8

340

ROBERT C . ALLER

APPENDIXC. FLUX-CORE DATA TABLE Cl. FLUX-CORE DATA:FOAM

Summer (8 July 1975) 0 5.5

16 29 54"

5.31 4.99 4.62 4.30

1.98 2.66 6.23 16.6

-

1.20 1.49 1.88

2.18 3.36

0.2 2.00 2.37 6.37 23.5

0.018 0.039 0.027 0.081 0.064

Fall (17 November 1975) 0 5.6 23.3 35.9 50.8"

5.15 4.78 4.41 4.06

0.32 1.4 3.9 8.93 13.4

2.81 2.87 3.05 3.37 3.59

0.2 1.62 1.8

3.3 3.8

0.09' 1.6' 0.02 0.04 0.01

Winter-Spring (18 March 1976) 0 3 12.75 25.25 48.42 14.75 99.75

5.97 5.49 5.01 4.54

4.12 3.64

0.10 0.19

0.68 1.21b

1.58

1.15'

3.43 4.75 7.38 10.0

1.23' 1.23' 1.21' 1.32'

0.2 2.2' 2.1b 2.0b 2.2'*d 2.9' 4.1'

-

-

-

Sampling time not used in calculations. Resuspension, points not used. ' Points used in calculation for Fe or winter core. Concentration assigned as t = 0 for flux calculation.

a

34 1

DIAGENETIC PROCESSES. I.

TABLEC2. FLUX-CORE DATA:NWC

Summer (16 July 1975) 0 3 16 27.5 50.5"

5.92 5.58 5.24 4.90

2.4 4.58 13.8

-

64.0

1.63 1.88 3.22 3.75 8.02

0.07 2.3 11.3 24.2 62.7

0.047 0.066 0.063

1.3 2.7 8.0 21.3 43.7

0.032 0.039 0.068 0.287 0.077

0.091

0.18

Fall (29 October 1975) 0 6.75 12.75 23 35.75"

4.58 4.21 3.84 3.48

0.2 2.6 7.3 14.5 23.2

2.46 3.27 4.33 5.56 6.78

Winter-Spring (23 March 1976) 0 3.25 12.25 33.25 49.50 100.75 172.75

5.34 4.98 4.61 4.24 3.87 3.50

0.9 1.3 1.4 3.0 4.4 9.9

-

1.05 1.09 1.15

1.28 1.36

-

0.07 0.34 0.43 0.62 0.84 1.46

2.22

' Sampling time not used in flux calculation.

342

ROBERT C. ALLER

TABLE C3. FLUX-COREDATA:DEEP

Summer (23 July 1975) 0 6 22 35 47.5”

0.36“ 5.4 6.1 16.0 18.4

5.62 5.28 4.95 4.61

1.15’ 2.61 3.34 4.27 4.92

0.2’ 7.1 14.0 18.0 24.8

0.0Ib

0.029 0.086 0.13 0.37

Fall (12 November 1975) 0 6.33 25.33 31.33 48.83”

4.44 4.08 3.69 3.33

0.14 2.6 4.9 5.8 10.1

2.55 2.88 3.04 3.11 3.29

0.24 4.0 5 5.5

6.1

0.018 1.lY 0.14 0.14 0.063

Winter-Spring (13 April 1976) 0 29 48 73 99 ~~~

3.07 2.64 2.21 1.78 ~~~~~

a

1.21 2.25 3.79 7.81 11.8

0.694 0.896 0.996 1.24

0.16 0.86 0.89 1.16 2.00

~

Sampling time not used in flux calculation.

’ Collected at 23 m.

‘Points used in Fe fall flux calculation.

-

-

-

DIAGENETIC PROCESSES. I.

343

ACKNOWLEDGMENTS This contribution is based on a portion of a Ph.D. dissertation done in the Department of Geology and Geophysics, Yale University. Modifications and, I hope, improvements were made at the University of Chicago while I was supported by an Alfred P. Sloan Foundation Fellowship. M. B. Goldhaber and J. Y. Yingst deserve special thanks as valued co-workers during critical stages of this project. I also wish to thank L. K. Benninger, J. K. Cochran, G. R. Holdren, and J. K. Rosenfeld for much help and discussion. My major diving partners over the years were J. Y. Yingst, W.J. Ullman, and M. Pimer. R. Wells was an indispensable aid in the field and lab. M.Pimer and M. Reed captained the boats used in sampling. K. K. Turekian, D. C. Rhoads, R. A. Bemer, and V. Barcilon provided critical comments, guidance, and advice at various stages. Thanks to M. D. Krom for critical review of the manuscript. Research support was predominately by ERDA grant EY-764-02-3573 (K. E. Turekian, principal investigator) and by NSF grant GA-42-838 (D. C. Rhoads, principal investigator). Personal support was additionally provided by a NSF Fellowship, a Yale Graduate Fellowship, ERDA grant EY-764-02-3573 and EPA grant R804-090-010 (D. C. Rhoads, principal investigator).

LISTO F SYMBOLS Attenuation constant Arrhenius temperature-dependent rate function Preexponential factor in A(t) Pore-water solute concentration at x and t Molecular diffusion coefficientmodified for porosity, tortuosity, charge coupling, and ion pairing Effective transport coefficient Activation energy Concentration gradient at depth x = L Solute flux at depth x Preexponential factor for temperature dependence of J, Linear, adsorption coefficient Gas constant, 1.99 caYdeghole Thickness of bioturbated zone Integer summation variable Porosity General reaction rate at depth x and time t Sulfate reduction rate at depth x and time t Ammonium production rate at depth x and time t Radial distance from burrow axis Total solid-phase sulfur Temperature Time Frequency as defined Frequency as defined Sedimentation rate. thicknesshime x Vertical depth measured positive from sediment-water interface

344

ROBERT C. ALLER

REFERENCES Abramowitz, M., and Stegun, I. A., eds. (1964). “Handbook of Mathematical Functions with Formulas, Graphs, and Mathematical Tables.” Dover, New York. Aller, R. C. (1977). The Influence of Macrobenthos on Chemical Diagenesis of Marine Sediments. Ph.D. dissertation, Yale University, New Haven, Connecticut. Aller, R. C. (1978). Experimental studies of changes produced by deposit feeders on pore water, sediment, and overlying water chemistry. Am. J . Sci. 278, 1185-1234. Aller, R. C. (1980). Relationships of tube-dwelling benthos with sediment and overlying water chemistry. In “Marine Benthic Dynamics” (K. R. Tenore and B. C. Coull, eds.), pp. 285-308. Univ. South Carolina Press, Columbia. Aller, R. C., and Cochran, J. K. (1976). z’4Th/Z3*Udisequilibrium in nearshore sediment: Particle reworking and diagenetic time scales. Earth Planet Sci. Lett. 20, 37-50. Aller, R. C., and Dodge, R. E. (1974). Animal-sediment relations in a tropical lagoon: Discovery Bay, Jamaica. J . Mar. Res. 32,209-232. Aller, R. C., and Yingst, J. Y. (1978). Biogeochemistry of tube-dwellings: A study of the sedentary polychaete Amphitrite ornafa (Leidy). J . Mar. Res. 36, 201-254. Aller, R. C., and Yingst, J. Y. (1980). Relationships between microbial distributions and the anaerobic decomposition of organic matter in surface sediments of Long Island Sound, U.S.A. Mar. Biol. 56, 29-42. Aller, R. C., Benninger, L. K., and Cochran, J. K. (1980). Tracking particle associated disequilibrium. Earth Planet. processes in nearshore environments by use of 234Th/23*U Sci. Lett. 47, 161-175. Anderson, D. Q. (1939). Distribution of organic matter in marine sediments and its availability to further decomposition. J . Mar. Res. 2, 225-235. Bartels, R. C. F., and Churchill, R. V. (1942). Resolution of boundary problems by the use of a generalized convolution. Bull. Am. Math. SOC.48, 276-282. Benninger, L. K., AHer, R. C., Cochran, J. K., and Turekian, K. K. (1979). Effects of biological sediment mixing on the ’”Pb chronology and trace metal distribution in a Long Island Sound sediment core. Earth Planet. Sci. Lett. 43, 241-259. Berner, R. A. (1963). Electrode studies of hydrogen sulfide in marine sediments. Geochim. Cosmochim. Acta 27, 563-575. Berner, R. A. (1964). Distribution and diagenesis of sulfur in some sediments from the Gulf of California. Mar. Geol. 1, 117-140. Berner, R. A. (1967). Thermodynamic stability of sedimentary sulfides. Am. J . Sci. 265, 773-785. Berner, R. A. (1970). Sedimentary pyrite formation. Am. J . Sci. 268, 1-23. Berner, R. A. (1971). “Principles of Chemical Sedimentology.” McGraw-Hill, New York. Berner, R. A. (1974a). Iron sulfides in Pleistocene deep Black Sea sediments and their Paleooceanographic significance. In “The Black Sea-Geology, Chemistry and Biology.” AAPG Memoir No. 20, pp. 524-531. Berner, R. A. (1974b). Kinetic models for the early diagenesis of nitrogen, sulfur, phosphorus and silicon in anoxic marine sediments. In “The Sea” (E. D. Goldberg, ed.), Vol. 5, pp. 427-450. Wiley, New York. Berner, R. A. (1976a). Inclusion of adsorption in the modeling of early diagenesis. Earth Planet Sci. Let?. 29, 333-340. Bemer, R. A. (1976b). The benthic boundary layer from the viewpoint of a geochemist. In “The Benthic Boundary Layer” (I. N. McCave, ed.), pp. 33-35. Plenum, New York. Berner, R. A., Scott, M. R., and Thomlinson, C. (1970). Carbonate alkalinity in the pore waters of anoxic marine sediments. Limnol. Oceanogr. 15, 544-549. Berner, R. A., Westrich, J. T., Graber, R., Smith, J., and Martens, C. S . (1978). Inhibition of aragonite precipitation from supersaturated seawater: A laboratory and field study. Am. J . Sci. 278, 816-837.

DIAGENETIC PROCESSES. I.

345

Billen, G. (1975). Nitrification in the Scheldt Estuary (Belgium and the Netherlands). Estuarine Coastal Mar. Sci. 3, 79-90. Billen, G. (1978). A budget of nitrogen recycling in North Sea sediments off the Belgian coast. Estuarine Coastal Mar. Sci. 7, 127-146. Bloomfield, C. (1972). The oxidation of iron sulphides in soils in relation to the formation of acid sulphate soils and of ochre deposits in field drains. J. Soil Sci. 23, 1-16. Bokuniewicz, H. J. (1976). Estuarine sediment flux evaluated in Long Island Sound. Ph.D. dissertation, Yale University, New Haven, Connecticut. Bokuniewicz, H. J., Gebert, J., and Gordon, R. B. (1976). Sediment mass balance of a large estuary, Long lsland Sound. Estuarine Coastal Mar. Sci. 4,523-536. Bray, J. T. (1973). The behavior of phosphate in the interstitial waters of Chesapeake Bay sediments. Ph.D. dissertation, The Johns Hopkins University, Baltimore, Maryland. Bray, J. T., Bricker, 0. P., and Troup, B. N. (1973). Phosphate in interstitial waters of anoxic sediments: Oxidation effects dunng sampling procedure. Science 180, 1362-1364. Bricker, 0. P., and Troup, B. N. (1975). Sediment-water exchange in Chesapeake Bay. In “Estuarine Research, 1. Chemistry, Biology and the Estuarine System” (L. E. Cronin, ed.), pp. 3-27. Academic Press, New York, Calvert, S. E. (1964). Factors affecting distribution of laminated diatomaceous sediments in the Gulf of California. Am. Assoc. Petroleum Geol. Mem. 3, 311-330. Claypool, G. E., and Kaplan, I. R. (1974). The origin and distribution of methane in marine sediments. I n “Natural Gases in Marine Sediments” (I. R. Kaplan, ed.), pp. 99-139. Plenum, New York. Conover, S. A. M. (1956). Oceanography of Long Island Sound, 1952-1954. IV. Phytoplankton. Bull. Bingham Oceanogr. Collect. 15, 62-1 12. Cowan, I. R. (1965). Transport of water in the soil-plant-atmosphere system. J . Appl. Ecol. 2, 221-239.

Coyer, P. E., and Mangum, C. P. (1973). Effect of temperature on active and resting metabolism in polychaetes. I n “Effects of Temperature on Ectothermic Organisms” (W. Wieser, ed.), pp. 173-180. Springer-Verlag, Berlin and New York. Crank, J. (1956). “The Mathematics of Diffusion.” Oxford Univ. Press (Clarendon), London and New York. Degobbis, D, (1973). On the storage of sea water samples for ammonia determination. Limnol. Oceanogr. 18, 146-150. Duff, E. J. (1971). Orthophosphates. Part IV. J. Chem. SOC.A , 921-926. Elderfield, H. (1976). Manganese fluxes to the oceans. Mar. Chem. 4, 103-132. Emerson, S. (1976). Early diagenesis in anaerobic lake sediments: Chemical equilibria in interstitial waters. Geochim. Cosmochim. Acta 40,925-934. Fanning, K. A., and Pilson, M. E. Q. (1974). Diffusion of dissolved silica out of deep-sea sediments. J. Geophys. Res. 79, 1293-1297. Fenchel, T. (1970). Studies on the decomposition of organic detritus derived from the turtle grass Thalassia testudinum. Limnol. Oceanogr. 15, 14-20. Gardner, W. R. (1960). Dynamic aspects of water availability to plants. Soil Sci. 89,63-73. Garrels, R. M., and Christ, C. L. (1965). “Solutions, Minerals and Equilibria.” Harper, New York. Gieskes, J. M., and Rogers, W. C. (1973). Alkalinity determination in interstitial waters of marine sediments. J. Sediment. Petrol. 43, 272-277. Goldhaber, M. B., and Kaplan, I. R. (1974). The sulfur cycle. I n “The Sea” (E. D. Goldberg, ed.), Vol. 5, pp. 569-655. Wiley, New York. Goldhaber, M. B., Aller, R. G., Cochran, J. K., Rosenfeld, J. K., Martens, C. S., and Berner, R. A. (1977). Sulfate reduction, diffusion and bioturbation in Long Island Sound sediments: Report of the FOAM group. Am. J. Sci. 277, 193-237. Gordon, R. B., and Pilbeam, C. C. (1974). Environmental consequences of dredge spoil

346

ROBERT C.ALLER

disposal in Long Island Sound: Geophysical studies, 1 October 1972-30 September 1973, Report to U.S. Army Corps of Engineers, New England Division, Waltham, Massachusetts. Graber, R. (1975). CaCO, saturation in anoxic sediments. Senior Thesis, Yale University, New Haven, Connecticut. Grill, E. V., and Richards, F. A. (1964). Nutrient regeneration from phytoplankton decomposing in sea water. J. Mar. Res. 22, 51-69. Grundmanis, V., and Murray, J. W. (1977). Nitrification and denitrification in marine sediments from Puget Sound. Limnol. Oceanogr. 22, 354-368. Hale, S. S. (1975). The role of benthic communities in the nitrogen and phosphorus cycles of an estuary. Proc. ERDA Symp. Augusta, Georgia, May, 1974 pp. 291-308 (ERDA ser.). Hammond, D. E., Simpson, H. J., and Mathieu, G . (1975). Methane and Radon-222 as tracers for mechanisms of exchange across the sediment-water interface in the Hudson River Estuary, pp. 119-132. ACS Symp. Ser. 18. Hanor, J. S., and Marshal, N. T. (1971). Mixing of sediment by organisms. I n “Trace Fossils” (B. F. Perkins, ed.), pp. 127-136. Louisiana State Univ. Misc. Publ. 71-1. Hargrave, B. T. (1970). The effect of a deposit-feeding amphipod on the metabolism of benthic microflora. Limnol. Oceanogr. 15, 21-30. Hargrave, B. T. (1976). The central role of invertebrate faeces in sediment decomposition. p. 285-299. Proc. Symp. Brit. Ecol. SOC., 17th pp. 285-299. Harmsen, G. W. (1954). Observations on the formation and oxidation of pyrite in the soil. Plant Soil 5, 324-348. Harris, E. (1959). The nitrogen cycle in Long Island Sound. Bull. Bingham Oceanogr. Collect. 17, 31-65. Harris, E., and Riley, G. A. (1956). Oceanography of Long Island Sound, 1952-1954. VIII. Chemical composition of the plankton. Bull. Bingham Oceanogr. Collect. 15,315-323. Hartmann, M., Muller, P., Suess, E., and Van der Weijden, C. H. (1973). Oxidation of organic matter in recent marine sediments. Meteor Forschungsergeb. Reihe C 12,74-86. Hartmann, M., Muller, P. J., Suess, E., and Van der Wejdan, C. H. (1976). Chemistry of late Quaternary sediments and their interstitial waters from the N.W. African continental margin. Meteor Forschungsergeb. Reihe C 24, 1-67. Hartwig, E. 0. (1976). The impact of nitrogen and phosphorus release from a siliceous sediment on the overlying water. I n “Estuarine Processes” (M. L. Wiley, ed.), Vol. 1, pp. 103-117. Academic Press, New York. Hayes, F. R. (1964). The mud-water interface. Oceanogr. Mar. Biol. Annu. Rev. 2,121-145. Holdren, G . R., Jr. (1977). Distribution and behavior of manganese in the interstitial waters of Chesapeake Bay sediments during early diagenesis. Ph.D. dissertation, The Johns Hopkins University, Baltimore, Maryland. Howard, J. D. (1975). The sedimentological significance of trace fossils. I n “The Study of Trace Fossils” (R. W. Frey, ed.), pp. 131-146. Springer-Verlag, Berlin and New York. Imboden, D. M. (1975). Interstitial transport of solutes in non-steady state accumulating and compacting sediments. Earth Planet. Sci. Lett. 27, 221-228. Johannes, R. E. (1968). Nutrient regeneration in lakes and oceans. Adv. Microbiol. Sea 1, 203-2 12.

J@rgensen,B. B. (1977). The sulfur cycle of a coastal marine sediment (Linfjorden, Denmark). Limnol. Oceanogr. 22,814-832. Jumars, P. A., Thistle, D., and Jones, M. L. (1977). Detecting two dimensional spatial structure in biological data. Oecologia 28, 109-123. Kalil, E. K., and Goldhaber, M. (1973). A sediment squeezer for removal of pore waters without air contact. J. Sediment. Petrol. 43, 553-557. Kamp-Nielsen, L. (1974). Mud-water exchange of phosphate and other ions in undisturbed

DIAGENETIC PROCESSES. I.

347

sediment cores and factors affecting the exchange rates. Arch. Hydrobiol. 73,218-237. Kaplan, I. R., Emery, K. 0..and Rittenberg, S. C. (1963). The distribution and isotopic abundance of sulphur in recent marine sediments off southern California. Geochim. Cosmochim . Acta 27, 297-33 1. Krause, H. R. (1959). Biochemische Untersuchungen iiber den postmortalen Abbau von totem Plankton unter aeroben und anaeroben Bedingungen. Arch. Hydrobiol. 24, 297-337.

Krom, M., and Berner, R. A. (1980). The experimental determination of the diffusion coefficients of sulfate, ammonium, and phosphate in anoxic marine sediments. Limnol. Oceanogr. 25, 327-337. Lasaga, A. C. (1979). The treatment of multi-component diffusion and ion pairs in diagenetic fluxes. Am. J . Sci. 279, 324-346. Lerman, A. (1975). Maintenance of steady state in oceanic sediments. Am. J . Sci. 275, 609-635.

Lerman, A. (1976). Migrational processes and chemical reactions in interstitial waters. pp. 695-738. In “The Sea” (E. D. Goldberg, I. N. McCave, J. J. O’Brien, and J. H. Steele, eds.), Vol. 6. Wiley, New York. Lerman, A. (1978). Chemical exchange across sediment-water interface. Ann. Rev. Earth Planet. Sci. 6, 281-303. Li, W. C., Armstrong, D. E., Williams, J. D. H.,Harris,R. F., and Syers, J. K. (1972). Rate and extent of inorganic phosphate exchange in lake sediments. Soil Sci. SOC.Am. Proc. 36, 279-285. Li, Y.-H., and Gregory, S . (1974). Diffusion of ions in sea water and in deep-sea sediments. Geochim. Cosmochim. Acta 38,703-714. McCall, P. L. (1977). Community patterns and adaptive strategies of the infaunal benthos of Long Island Sound. J . Mar. Res. 35,221-266. McCall, P. L. (1978). Spatial-temporal distributions of Long Island Sound infauna: The role of bottom disturbance in a nearshore marine habitat. In “Estuarine Interactions” (M. L. Wiley, ed.), pp. 191-219. Academic Press, New York. Mangum, C. P. (1963). Studies of speciation in maldanid polychaetes of the North American Atlantic coast-111. Intraspecific and interspecific divergence in oxygen consumption. Comp. Biochem. Physiol. 10, 335-349. Mangum, C. P., and Vanwinkle, W. (1973). The response of aquatic invertebrates to declining oxygen conditions. Am. Zool. 13, 529-541. Manheim, F. T. (1970). The diffusion of ions in unconsolidated sediments. Earth Planet. Sci. Lett. 9, 307-309. Manheim, F. T., and Waterman, L. S. (1974). Dsusimetry (diffusion constant estimation) on sediment cores by resistivity probe. Initial Rep. Deep Sea Drilling Project 22, 663-670.

Martens, C. S., and Berner, R. A. (1974). Methane production in the interstitial waters of sulfate-depleted marine sediments. Science 185, 1167-1 169. Martens, C. S., and Berner, R. A. (1977). Interstitial water chemistry of anoxic Long Island Sound sediments: Dissolved gases. Limnol. Oceanogr. 22, 10-25. Martens, C. S., and Harriss, R. C. (1970). Inhibition of apatite precipitation in the marine environment by magnesium ions. Geochim. Cosmochim. Acta 34,621-625. Martens, C. S., Berner, R. A., and Rosenfeld, J. K. (1978). Interstitial water chemistry of anoxic Long Island Sound sediments. 2. Nutrient regeneration and phosphate removal. Limnol. Oceanogr. 23,605-617. Molnia, B. F. (1974). A rapid and accurate method for the analysis of calcium carbonate in small samples. J . Sediment. Petrol. 44, 589-590. Moore, D. G., and Scruton, P. C. (1957). Minor internal structures of some recent unconsolidated sediments. Am. Assoc. Petrol. Geol. Bull. 41, 2723-2751.

348

ROBERT C. ALLER

Morgan, J. J. (1967). Chemical equilibria and kinetic properties of manganese in natural waters. I n “Principles and Applications of Water Chemistry” (S. P. Faust and J. V. Hunter, eds.), pp. 561-622. Wiley, New York. Mortimer, C. H. (1941). The exchange of dissolved substances between mud and water. 1 and 11. J . Ecol. 29, 280-329. Mortimer, C. H. (1971). Chemical exchanges between sediments and water in the great lakes-speculations on probable regulatory mechanisms. Limnol. Oceanogr. 16, 387-404. Mortland, M. M., and Wolcott, A. R. (1965). Sorption of inorganic nitrogen compounds by soil materials. Agronomy 10, 150-197. Miiller, P. J. (1977). C/N ratios in Pacific deep-sea sediments: Effect of inorganic ammonium and organic nitrogen compounds sorbed by clays. Geochim. Cosmochim. Acta 41, 765-776. Myers, A. C. (1979). Summer and winter burrows of a mantis shrimp, Squilla empusa, in Narragansett Bay, Rhode Island (U.S.A.) Estuarine Coastal Mar. Sci. 8, 87-98. Nixon, S. W., Oviatt, C. A., and Hale, S. S. (1976). Nitrogen regeneration and the metabolism of coastal marine bottom communities. Proc. Symp. Br. Ecol. Sac., /7rh, April 1975, pp. 269-283. Nixon, W. S., Kelly, J. R., Furnas, B. N., Oviatt, C. A., and Hale, S. S. (1980). Phosphorus regeneration and the metabolism of coastal marine bottom communities. I n “Marine Benthic Dynamics” (K. R. Tenore and B. C. Coull, eds.), pp. 219-239. Univ. South Carolina Press, Columbia. NBmmik, H. (1965). Ammonium fixation and other reactions involving a nonenzymatic immobilization of mineral nitrogen in soil. Agronomy 10, 198-258. Norton, S.A., and Sasseville, D. R. (1975). Flux of nutrients by diffusion through the lake sediment-hypolimnion interface for 9 lakes in Maine, U.S.A. Znt. Ver. Theor. Angew. Limnol. Verh. 19, 372-381. Norvell, W. A. (1974). Insolubilization of inorganic phosphate by anoxic lake sediment. Soil Sci. SOC.Am. Proc. 38, 441-445. Nriagu, J. 0. (1972). Stability of vivianite and ion pair formation in the system Fe3(P04) - H3P04- H20. Geochim. Cosmochim. Acta 36,459-470. Nriagu, J. O., and Dell, C. I. (1974). Diagenetic formation of iron phosphates in recent lake sediments. Am. Mineral. 59, 934-946. Nye, P. H., and Tinker, P. B. (1977). “Solute Movement in the Soil-Root System. Studies in Ecology,” Vol. 4. Univ. California Press, Berkeley and Los Angeles. Olsen, S. (1964). Phosphate equilibrium between reduced sediments and water. Verh. Znt. Verein. Limnol. 15, 333-341. Oremland, R. S., and Taylor, B. F. (1978). Sulfate reduction and methanogenesis in marine sediments. Geochim. Cosmochim. Acta 42, 209-214. Parfitt, R. L., Atkinson, R. J., and Smart, R. St. C. (1975). The mechanism of phosphate fixation by iron oxides. Soil Sci. SOC.Am. Proc. 39, 837-841. Presley, B. J. (1971). Techniques for analyzing interstitial water samples. Part I: Determination of selected minor and major inorganic constituents. Initial Rep. Deep Sea Drilling Project I , Part 2, 1749-1755. Presley, B. J., and Kaplan, I. R. (1968). Changes in dissolved sulfate, calcium, and carbonate from interstitial water of near-shore sediments. Geochim. Cosmochim. Acta 32, 1037-1048. Prosser, C. L., and Brown, F. A. (1961). “Comparative Animal Physiology.” Saunders, Philadelphia, Pennsylvania. Pruden, G., and Bloomfield, C. (1968). The determination of ferrous sulphide in soil, in the presence of ferric oxide. Analyst 93, 532-534.

DIAGENETIC PROCESSES. I.

349

Redfield, A. C. (1934). On the proportion of organic derivatives in sea water and their relation to the composition of plankton. pp. 176-192. I n “James Johnson Memorial Volume.” Univ. Press, Liverpool. Rhoads, D. C. (1963). Rates of sediment reworking by Yoldia limatula in Buzzards Bay, Massachusetts and Long Island Sound. J. Sediment. Petrol. 33, 723-727. Rhoads, D. C. (1974). Organism-sediment relations on the muddy sea floor. Oceanogr. Mar. Biol. Annu. Rev. 12, 263-300. Rhoads, D. C., and Stanley, D. J. (1965). Biogenic graded bedding, .I Sediment. . Petrol. 35, 956-963. Rhoads, D. C., and Young, D. K. (1970). The influence of deposit-feeding organisms on sediment stability and community trophic structure. J. Mar. Res. 28, 150-178. Rhoads, D. C., Aller. R. C., and Goldhaber, M. B. (1977). The influence of colonizing macrobenthos on physical properties and chemical diagenesis of the estuarine seafloor. pp. 113-138. I n “Ecology of Marine Benthos” (B. C. Coull, ed.). Belle W. Baruch Library in Marine Science, No. 6. Univ. South Carolina Press, Columbia. Richards, F. A. (1965). Anoxic basins and fiords. I n “Chemical Oceanography” (J. P. Riley and G. Skirrow, eds.), Vol. 1, pp. 611-645. Academic Press, New York. Rickard, D. T. (1969). The chemistry of iron sulfide formation at low temperatures. Stockholm Contrib. Geol. 20, 67-95. Rickard, D. T. (1974). Kinetics and mechanism of the sulfidation of goethite. Am. J. Sci. 274, 941-952. Rickard, D. T. (1975). Kinetics and mechanism of pyrite formation at low temperatures. Am. J. Sci. 275,636-652. Riley, G . A. (1956a). Oceanography of Long Island Sound, 1952-1954. 11. Physical oceanography. Bull. Bingham Oceanogr. Collect. 15, 15-46. Riley, G. A. (1956b). Oceanography of Long Island Sound, 1952-1954. IX. Production and utilization of organic matter. Bull. Bingham Oceanogr. Collect. 15, 324-344. Rittenberg, S . C., Emery, K. O., and Orr, W. L. (1955). Regeneration of nutrients in sediments of marine basins. Deep Sea Res. 3, 23-45. Rosenfeld, J. K. (1977). Nitrogen diagenesis in nearshore anoxic sediments. Ph.D. dissertation, Yale University, New Haven, Connecticut. Rosenfeld, J. K. (1979). Ammonium adsorption in nearshore anoxic sediments. Limnol. Oceanogr. 24, 356-364. Rowe, G . T., Clifford, C. H., Smith, K, L., Jr., and Hamilton, P. L. (1975). Benthic nutrient regeneration and its coupling to primary productivity in coastal waters. Nature (London) 255, 215-217. Sanders, H. L. (1956). Oceanography of Long Island Sound, 1952-1954. X. The biology of marine bottom communities. Bull. Bingham Oceanogr. Collect. 15, 345-414. Sato, M. (1960). Oxidation of sulfide ore bodies. 11. Oxidation mechanisms of sulfide minerals at 25°C. Econ. Geol. 55, 1202-1231. Sayles, F. L. (1979). The composition and diagenesis of interstitial solutions. I. Fluxes across the seawater-sediment interface in the Atlantic Ocean. Geochim. Cosmochim. Acta 43, 527-546. Sayles, F. L., and Mangelsdorf, P. C., Jr. (1976). The analysis of SO; in seawater by difference chromatography. Limnol. Oceanogr. 21, 899-905. Sayles, F. L., and Manheim, F. T. (1975). Interstitial solution and diagenesis in deeply buried marine sediments: Results from the Deep Sea Drilling Project. Geochim. Cosmochim. Acta 39, 103-127. Schtifer, W. (1962). “Aktuo-Palaontologie.” Kramer, Frankfurt am Main [“Ecology and Paleoecology of Marine Environments” (I. Oertel, trans.; G. Y. Craig, ed.). Univ. Chicago Press, 1972.1

350

ROBERT C. ALLER

Sholkovitz, E. (1973). Interstitial water chemistry of the Santa Barbara Basin sediments. Geochim. Cosmochim. Acra 37,2043-2073. Sillbn, L. G.,and Martell, A. E. (1964). “Stability Constants of Metal-Ion Complexes.” The Chemical Society, Burlington House, London. Solorzano, L. (1969). Determination of ammonia in natural waters by the phenolhyochlorite method. Limnol. Oceanogr. 14, 799-801. Stanley, S. M. (1970). Relation of shell form to life habits in the Bivalvia. Geol. SOC.Am. Mem. 125. Strickland, J. D. H., and Parsons, T. R. (1968). A practical handbook of sea water analysis. Can. Fish. Res. Bd. Bull. 167. Stumm, W., and Morgan, J. J. (1970). “Aquatic Chemistry,” 583 pp. Wiley, New York. Suess, E. (1976). Nutrients near the depositional interface. I n “The Benthic Boundary Layer” (I. N. McCave, ed.), pp. 57-79. Plenum, New York. Suess, E. (1979). Mineral phases formed in anoxic sediments by microbial decomposition of organic matter. Geochim. Cosmochim. Acra 43, 339-352. Sweeney, R. E., and Kaplan, I. R. (1973). Pyrite framboid formation: laboratory synthesis and marine sediments. Econ. Geol. 68,618-634. Syers, J. K., Hams, R. F., and Armstrong, D. E. (1973). Phosphate chemistry in lake sediments. J . Environ. Qual. 2, 1-14. Tessenow, U.(1972). Losungs-, Diffusions- und Sorptionsprozesse in der Oberschicht von Seesedimenten. I. Arch. Hydrobiol. Suppl. 38, 353-398. Tessenow, U. (1974). Losungs-, Diffusions-, and Sorptionsprozesse in der Oberschicht von Seesedimenten. IV. Arch. Hydrobiol. Suppl. 47, 1-79. Thomson, J., Turekian, K. K., and McCaffrey, R. J. (1975). The accumulation of metals in and release from sediments of Long Island Sound. I n “Estuarine Research, 1. Chemistry, Biology and the Estuarine System” (L. E. Cronin, ed.), pp. 28-44. Academic Press, New York. Troup, B. N. (1974). The interaction of iron with phosphate, carbonate and sulfide in Chesapeake Bay interstitial waters: A thermodynamic interpretation. Ph.D. dissertation, The Johns Hopkins University, Baltimore, Maryland. Ullman, W. J., and Aller, R. C. (1980). Dissolved iodine flux from estuarine sediments and implications for the enrichment of iodine at the sediment-water interface. Geochim. Cosmochim. Acta 44 (in press). Vanderborght, J. P., Woliast, R., and Billen, G. (1977). Kinetic models of diagenesis in disturbed sediments. Part I. Mass transfer properties and silica diagenesis. Limnol. Oceanogr. 22, 787-793. Van Straaten, L. M. J. U. (1952). Biogene textures and the formation of shell beds in the Dutch Wadden Sea. 1, 11. K. Ned. Akad. Wer. Proc. Ser. B 55, 500-516. Vosjan, J. H. (1974). Sulphate in water and sediment of the Durch Wadden Sea. Neth. J . Sea Res. 8,208-213. Waksman, S . A. (1933). On the distribution of organic matter in the sea bottom and the chemical nature and origin of marine humus. Soil Sci. 36, 125-147. Yingst, J. Y.,and Rhoads, D. C. (1978). Seafloor stability in central Long Island Sound. Part 11. Biological interactions and their potential importance for seafloor erodibility. I n “Estuarine Interactions” (M. N. Wiley, ed.), pp. 245-260. Academic Press, New York. Yingst, J. Y., and Rhoads, D. C. (1980). The role of bioturbation in the enhancement of microbial turnover rates in marine sediments. In “Marine Benthic Dynamics” (K. R. Tenore and B. C. Coull, eds.), pp. 407-421. Univ. of South Carolina Press, Columbia. Young, D. K. (1971). Effects of infauna on the sediment and sestonofasubtidal environment. Vie Milieu Suppl. 22,557-571.

DIAGENETIC PROCESSES NEAR THE SEDIMENT-WATER INTERFACE OF LONG ISLAND SOUND. 11. Fe and Mn ROBERTC.ALLER* Department of Geology and Geophysics Yale University New Haven, Connecticut

1. 2. 3. 4.

5.

Introduction . . . . . . . . . . . . . . . . . . . . . . . . , . . Location and Description of Study Area . . . . . . . . . . . . . . . Sampling . . . . . . . .... ..... ...... ... Methods . . , . . . . . . . . . . . . . . . . . . , . . .. . 4.1. Treatment of Cores . . . .... .. .... ....... 4.2. Analytical Techniques . . . . . . .. ... ..... ... Results . . .. . ... ... .... . .. ..... .. . 5.1. Pore-Water Profiles, Gravity Cores . . . . . . . . . . . . . . . 5.2. Pore-Water Profiles, Box Cores . . . . . . . . . . . .. ... 5.3. Solid Phase . . . . . . . . . . . . . . . . . . . . . . . . . 5.4. Flux Measurements . . . . . . . . . . . . . . . . . . . . . . Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.1. Production of Fe2+ and Mn2+ . . . . . . . . . . . . . . . . . 6.2. Precipitation Reactions and Saturation States . . . . . . . . . . 6.3. Seasonality of Pore-Water Fe and Mn Profiles Near the Sediment-Water 6.4. Production and Precipitation Rates . ... ,,........ 6.5. Flux of MnZ+and Fez+ into Overlying Waters . . . . . . . . . . . Summary . . , . . . . . . . . . . . . . . . . . . . . . . . List of Symbols . . . . . . . . . . . . . . . . . . . .. . References . . . . . . . . . . . . . .. . .. .. . ..

.

.

.

.

.

.

.

.

.

. .

.

.

6.

.

.

.

..

7.

. .

..

.

.

.. ..

. . . .

... . .... ..,... .... .. .... .. . . . , . . .... .. ...... ...... .. .... .... .. ...... .... .. ...... Interface . . .. .. .. ...... ....... ..... . ......

35 1 352 353 353 353 354 355 355 357 359 364 367 368 375 382 384 401

406 409 410

1. INTRODUCTION

The concentration and distribution of many metals in seawater and sediments are influenced by adsorption or coprecipitation with Fe and Mn oxides (Goldberg, 1954; Krauskopf, 1956; Jenne, 1968; Murray and Brewer, 1977). Because of this and because of the direct or indirect involvement of both Fe and Mn in the metabolic or biogeochemical activities of marine organisms, knowledge of the behavior of these two elements is particularly essential to understanding marine sediment and water * Present address: Department of The Geophysical Sciences, The University of Chicago, Chicago, Illinois 60637. 351 ADVANCES IN GEOPHYSICS, VOLUME 22

Copyright 0 1980 by Academic Press, Inc. All rights of reproduction in any form reserved.

ISBN cm-oiaazz-8

352

ROBERT C. ALLER

chemistry (Turekian, 1977). The general geochemistries of Fe and Mn are well known: both metals are relatively mobile under reducing, aqueous conditions, but are rapidly precipitated when exposed to oxygenated, neutral-to-basic solutions (Krauskopf, 1957). On the other hand, specific rates and mechanisms of mobilization, precipitation, and transport of Fe and Mn between natural reservoirs as well as identification of particular solid phases controlling their solubilities in different environments are still being worked out. Coastal marine environments are of special interest in this regard because of their role in determining the influence of continental runoff on ocean composition. Near-shore sediments are commonly reducing a few centimeters below the sediment surface and are overlain by oxygenated waters. Coupled with the oxidation-reduction properties of Fe and Mn, these conditions promote the mobilization-reprecipitation of both metals in a sedimentary zone where interaction with living organisms or other elements in both sediment and overlying waters is greatest. With these facts in mind, in this article I examine the chemical diagenesis of Fe and Mn in the near-shore sediments of Long Island Sound. Particular attention is given to quantifying the physical and biological transportreaction processes controlling both the distribution of these metals within the upper few decimeters of a deposit and the exchange of their soluble forms with overlying water. Several kinds of chemical measurements were made in each of three distinct depositional environments from the Sound. These are: (1) seasonal variation of pore-water solute profiles over l- or 2-year periods, (2) solid-phase analyses of total Fe and Mn, and (3) direct flux measurements of Fe and Mn released from bottom sediments. In addition, several laboratory experiments were performed to help substantiate or disprove interpretations of field data. The availability of 234Th-or *"Pb-particle reworking rates in surface sediments (Benninger et al., 1979; Aller et al., 1980), together with extensive faunal descriptions (Aller, this volume, Part I), allow an especially complete consideration of the influence of macrobenthic activity on Fe and Mn distributions at each site.

2. LOCATION AND DESCRIPTION OF STUDY AREA The three stations chosen for study-FOAM, NWC, and DEEP-are located in predominantly silt-clay regions of central Long Island Sound and lie in 9, 14, and 34 m of water, respectively. The location, description, faunal analysis, and examples of x radiographs of sediment at these stations are given in detail in this volume, Part I, p. 238, and will not be repeated here. The three stations represent a general inshore-offshore transect running from FOAM to DEEP.

DIAGENETIC PROCESSES. 11.

353

3. SAMPLING

Most bottom samples were obtained by Scuba divers using specially constructed Plexiglas box cores (Fig. 13 of Part I). The method of obtaining cores is described in Part I. Gravity cores were also taken at each station on one or more occasions as previously described. Box cores were taken in summer, fall, and winter-spring periods from 1974 to 1976; sampling was carried out over 2-year periods at FOAM and NWC and for 1 year at DEEP (Table 111 of Part I, this volume, p. 253).

4. METHODS

4.1. Treatment of Cores 4.1 .I. Pore- Water and Sediment Samples. The cores used for analysis of Fe and Mn were the same used for the pore-water and solid-phase analyses given in Part I. The treatment of cores is therefore identical. In summary: box cores were stored at 4°C for 3-5 hours after collection. Sediment was removed at 1-cm intervals under a N, atmosphere in a glove bag, placed in precut 10-cm-long acetate-butyrate tube sections (4.7-cm o.d.), and squeezed using the method of Kalil and Goldhaber (1973) within 12-24 hours after collection. Pore water was expressed first through a glass fiber prefilter and then through 0.45-pm-pore Millepore filters directly into plastic syringes without air contact. Gravity cores were processed in the same way except that the core was sectioned in 10-cm intervals by use of a hacksaw and a miter box. Water taken from each end of a section was either kept separate and assigned to different 5-cm intervals or combined. In all cases, the first -5 ml of pore water expressed was discarded and care was taken to avoid air contact during packing and squeezing operations to minimize oxidation reactions (Troup et al., 1974). Pore-water samples for Fe and Mn analysis were treated in the following way: 10 or 15 ml of pore water were passed rapidly from a syringe through a short section of Tygon tubing into a 10- or 15-ml pipet, depending on sample size. This sample was immediately drained into an acid-washed and distilled-water rinsed 2-oz polyethylene wide-mouth bottle. The sample was titrated for alkalinity and acidified to pH 1-2 within 48 (occasionally 72) hours in the same 2-oz container (Part I). The resulting acidified solutions were later analyzed for Fe and Mn. Experimental checks of this procedure demonstrated that if acidification (titration) was performed within 1-2 weeks of sample storage that 100 k 1% of the Fe and 100 f 2% of the Mn was recovered (Aller, 1977,p. 426). Distilled water blanks

354

ROBERT C. ALLER

run through the same alkalinity titration procedure gave Fe and Mn blanks below detection limit for 10-ml samples. The sediment cake resulting from squeezing was retained in its butyrate tube, sealed with end caps and frozen. Ignition-loss (475"C), % CaCO,, acid-volatile sulfide, and total nonvolatile sulfur were measured on many samples (Part I). Fe and Mn determinations were made as described later. Solid-phase analyses for Cu were reported in Aller (1977). 4.1.2. Flux Samples. Core boxes designed for measuring sedimentwater solute exchange were taken at each station during the summer, fall, and winter of 1975-1976. Two boxes were collected during a single sampling; one contained only bottom water and acted as a blank, the other was filled with 10-15 cm of undisturbed bottom sediment and its overlying water (Part I). Both summer boxes (1975) were kept at 22"C, fall (1975) flux boxes at 15"C, and winter (1976) at 4°C. Processing of these boxes consisted of withdrawing water samples from each box at successive times and following the change in overlying water concentrations of selected, dissolved ions. The total initial water volume was determined and volumes removed at each sampling were measured. By correcting the concentration changes for volume differences between one sampling and the next, a flux of a given ion out of the sediment could be calculated as described subsequently. The summer and fall boxes were periodically aerated. Water samples were obtained by inserting a section of Tygon tubing attached to a syringe into a box and withdrawing a sample. This water was immediately filtered through 0.4-pm-pore size Nuclepore filters directly into acid-cleaned sample bottles. The contents of one bottle was acidified with HCI and later analyzed for Fe and Mn. pH and Eh of the water were occasionally checked by insertion of the appropriate electrode into a water sample. After water sampling was complete, sediment in the flux boxes was sometimes sieved through a 1-mm mesh sieve for the contained animals (Part I). 4.2. Analytical Techniques 4.2.1. General. pH and pS or Eh measurements were made in the glove bkg by inserting the appropriate electrode directly into sediment samples; electrodes were passed into the bag through a sealed port. Calibration and specifics are described in Part I. Both pH and pS came to stable readings in the time allowed; Eh did not, but it was not changing rapidly when readings were taken. 4.2.2. Water Analyses. Fe analyses were made using the colorometric reagent Ferrozine and the reductant NH20H HCI (Stookey, 1970; Lewin

-

DIAGENETIC PROCESSES. 11.

355

and Chen, 1973). Samples were not boiled but were allowed to react for -1 hr prior to color development. Aliquot size ranged from 1-10 ml for pore water to 45 ml for flux samples. Precision of analysis was -1% in most cases. In the 0.02-p.M range, precision was -20-30%, but in this range the analysis is quite close to background Fe in the reagents used and the results questionable in any case. Testing of the Ferrozine reagent on standard rock samples (GSP-1, BCR-1, G-2) indicated that this reagent reproduced the accepted analytical Fe content of those standards, whereas several other methods of Fe analysis [e.g., atomic absorption (AA), 2,2'-dipyridine] following recommended procedures, did not. Mn analyses were made using flame AA by direct aspiration of samples. Precision was -2-3%. Additional analyses using formaldoxime (Goto et al., 1962)were made on some pore-water samples and found to agree with AA analyses to within -5%. Interferences were occasionally found with the formaldoxime method for samples taken from depth in a deposit. These were readily recognized and samples could be accurately reanalyzed by the simple technique of diluting the aliquot so that interfering substances, presumably HPOZ- and Ca2+(Henriksen, 1966), were at low enough concentrations so as to not affect the analysis (Aller, 1977). Winter Mn flux samples at all stations and fall fluxes at FOAM were measured by use of formaldoxime. Precision was -2%. 4.2.3. Solid-Phase Analyses. Solid-phase analyses for Fe and Mn were done as follows: weighed aliquots of ashed (475°C) sediment were leached for 24 hours in hot 12 N HCl(30 ml), the solutions taken to near dryness, the residue picked up in hot 0.1-0.2 N HCl, and filtered through Whatman 42 paper. The resulting leachate was diluted to known volume and appropriate further dilutions analyzed for Fe using Ferrozine reagents and Mn using flame AA. Fe and Mn standards (in 0.1 NHCl) and the analytical techniques used were checked against analyses of totally dissolved standard rock samples (GSP-1, BCR-1, G-2) and found to have good agreement. Precision of all analyses was -3%. Analyses of unashed sediment aliquots leached with hot aqua regia gave good agreement with the ashedsediment-HCl leach, indicating that pyrite was destroyed with the normal analytical procedure.

5. RESULTS 5.1. Pore- Water Profiles, Gravity Cores Pore-water Fe and Mn from gravity cores at each site are given in Fig. 1. Because the cores were collected in different seasons, the concentra-

74 1

FOAM- G (8 July 7 5 ) Fe ( p M ) 2 4 6 8 10 12 14 _ _ - -

DEEP-G (12 Nov 7 5 ) Fe(pM) 1 2 3 4 5

a, i

I-

z

40.

40

80-

80

120.

I20

Mn ( p M ) 1 0

E w

20

30

f-y-?fy 40/1.40

;

8ol

120

40

*

80: 120. L

L.-.,.

$ - - - - - - - -

1 I"

.

;)I

4,

40

/

I20

I

80

1

FIG. 1 . Pore-water Fez+ and MnZ' profiles from gravity cores at FOAM, NWC,and DEEP. The region near the sediment-water interface should not be compared because of different collection times of cores.

DIAGENETIC PROCESSES. 11.

357

tions found in the upper sampling intervals cannot be readily compared, as will be shown later when results from high-resolution and seasonal sampling are presented. Below the upper 10-30 cm, dissolved Fe generally ranges between 1-4 p M with highest values found at FOAM and lowest at DEEP. There is a tendency for dissolved Fe to increase with depth in sediment at NWC. The irregular nature of the profile at FOAM may result from sampling problems during squeezing or may be real; values average -4 p M . Dissolved Mn below 30 cm shows an opposite station trend to that of Fe. Pore-water Mn is lowest at depth in the sediment at FOAM and highest at DEEP. Sulfide and pH measures were not always made on the same gravity cores used for Fe and Mn pore-water analyses, but the general values of these parameters are relatively predictable deep in the sediment at each station based on other cores. These are listed in Appendix B of Part I along with pore-water analyses. Sulfide levels are highest at FOAM (Goldhaber et al., 1977)and lowest at DEEP where they are below the detection limit of a sulfide electrode, pS 2 16, in the gravity-core samples (but not in the box cores). 5.2. Pore- Water Profiles, Box Cores

Pore-water profiles of dissolved Fe and Mn obtained from box cores are illustrated in Figs. 2-7. There are several important seasonal and spatial trends in the data. At all stations, dissolved Fe reaches a maximum concentration in a 1-cm sampling interval 1-3 cm below the interface and decreases rapidly on either side of the maximum. Concentration gradients are generally greater above the maximum than below. The highest peak concentrations are reached at FOAM and DEEP and the lowest at NWC. There are distinct seasonal trends in profile shape and the maximum Fe value attained at each site. High dissolved Fe is found during both winter and summer. Concentrations as high as 1 mM are measured, well over lo5 times that of overlying water collected in flux blank boxes. The maximum yearly concentrations are obtained during winter at FOAM and DEEP and during summer at NWC. During the fall, dissolved Fe peaks are diminished in either maximum value or width compared with other times of year at any given station. As shown by data from NWC and FOAM, the general patterns are repeatable from year to year. Dissolved Mn has a somewhat different seasonal and spatial pattern than does Fe (Figs. 5-7). At FOAM and NWC, Mn decreases exponentially below the interface during the summer and fall as determined by 1-cm sampling intervals (Figs. 5 and 6). This means that a maximum in

-

I'

-

- - - - + - - -

FOAM- 2 Winter- Spring I3 Mar 75

0,IL L

- 5 -

5 ? 5a

10 )

y

& -

.-

' -I

.

- -

- r

.

. . . , . , . . . #

- - - - - - _ - -I - - -41-I -*

359

DIAGENETIC PROCESSES. 11.

Fe ( p M )

-

0

01

20

I0 1

30

.

I .......................

5.

=

__I. '.r.I..

...I

'

:lo:/' 3

;.I

NWC-I

NWC-2

NWC-3

Summer 74

Foll 74 ( I I November)

Winter 75

NWC-5

NWC-6

Foll 7 5 (29October)

Winter 76 (23 March)

(18 March)

0 ....................................... ............... - - { - - --J

5 10

15

NWC-4 Summer 75 (16 July)

+0OJ 5. 10

15:

FIG.3. Seasonal change in pore-water Fez+ profiles in box cores from NWC.

Mn concentration is reached in the top 0-1 cm at these times. Like Fe, Mn concentrations in the fall cores decrease from those measured in summer. Unlike Fe, Mn concentrations continue to decrease into the winter at which time the maximum value shifts from the top 0-1 cm to a lower level in the sediment column. At DEEP, the form of the Mn concentration profiles differs from that at FOAM and NWC, there is either less or no tendency for exponentially decreasing values beneath the interface (Fig. 7). During the summer and fall, high Mn is found in the top few centimeters; these decrease to much lower levels in these same sampling intervals in winter. There is a tendency for a maximum to occur over an interval from 5-8 cm corresponding to an increase in organic matter at that depth (Part I). High concentrations of Mn occur deep in the sediment during winter and the profile decreases relatively smoothly toward the interface at that time. The lowest concentrations below 3-cm depth are measured in the fall.

5.3. Solid Phase Total Fe and Mn concentrations from several box cores are plotted as a function of depth in Figs. 8 and 9. As determined at NWC (Fig. 8), there is no easily detectable seasonality in solid-phase distributions of either Fe or Mn. At NWC, total Fe shows a tendency to decrease slightly with

360

ROBERT C . ALLER

20

15L

60

15

DEEP - I

DEEP-2

Summer 23 July 75

Fall 12 Nov 75

510 :- 1I’

40

I-

*

I-

- - ,I

-

- I ’

I-

3

I

I5

-

depth. This is true even after correction for dilution by CaCO, (see Part I, Appendix B, Table B8). Total Mn decreases exponentially below the sediment-water interface at NWC reaching a background concentration of -550 kg/gm ash by 4-5 cm. DEEP also has surface enrichment of Mn, but less so than NWC (Fig. 9). The Mn content passes through a minimum and then begins to increase again at depth. Fe has no surface enrichment but instead increases regularly with depth. Correction for dilution by CaCO, does not eliminate this pattern, although the gradients between surface concentrations and deeper sediment are lowered somewhat (Part I, Appendix B, Table B12). Solid-phase Fe and Mn profiles at FOAM are relatively irregular in comparison to other stations (Fig. 9). Mn has no special surface enrich-

3

t I

\

n

x

0 D

.-e

.-C

362

ROBERT C. ALLER

0

; 10:

NWC-2

I

,

15;

)I

Summer 74 (21 July)

1

Winter 75 ( 18 March)

(I1 November)

15

o/F 0

5 11'

NWC-5

NWC-4 Summer 7 5 (16 July)

151

I'

NWC-6 Winter 76 (23 March)

Fall 75 (29 October)

5p

FIG.6. Seasonal change in pore-water Mn2+ profiles in box cores from NWC.

40

120

200 280

0

fa 10

15

15

15t DEEP- I Summer 23 July 75

DEEP-2 Fall 12 Nov 75

DEEP-3 Winter Spring 13 Apr 76

-

FIG.7. Seasonal change in pore-water Mn" profiles in box cores from DEEP.

NWC- 2

NWC- I Fe (mg/qm-ash)

NWC

Fe 25

-3

NWC

Fe

=-?

-4

Fe

25

40

4

I

5 02r+---

I

0 5

lo 1 0

15

2

I

200

900

1100

5

II '

I

'

I

I

I It I

I

I I

I

Mn

Mn (pg/gm-ash) 500

r

500

700

900

Mn IKX)

1300

500

TOO

900

Mn 1100

I300

600

800

1000

I200

0

FIG.8. Distribution of Mn and Fe in solid phase at NWC. Gross features of the distributions are repeated in separate samples. There is no detectable regular seasonal change for either metal.

364

ROBERT C.ALLER

DEEP-I

(23 July 7 6 )

Fe +g/gm-ash)

FIG.9. Fe and Mn solid-phase distribution at FOAM and DEEP. Dashed lines represent concentration on a CaC03-free basis at FOAM.

ment at this station and below 10 cm begins to increase in concentration. Fe shows essentially the same kind of distribution. CaCO, is abundant at FOAM (-30% by weight, Goldhaber et al., 1977), but correction of Fe and Mn distributions to a CaC0,-free basis does not radically alter profile shapes (Fig. 9). 5.4. Flux Measurements

Fluxes of Fe and Mn were estimated as follows. The concentration Co of either Mn or Fe at the time t of first measurement in the blank box

365

DIAGENETIC PROCESSES. 11.

I

>

Summer

2401

200

c -

m

O

280

' 0.6

c

a3 -

o 160

.+5

N t

2

120

80

40 10 30 50 70 90 110 130 150 170

T i m e (hr)

-0.1

10

20 30 Time (hr)

40

FIG.10. Amount of Mn" and Fe2' released from flux-core sediment versus time after collection. Mn" increases are approximately linear with time and are plotted relative to the starting concentration in the blank flux boxes (Co, t = 0). Only data from the first 30 hr are used to calculate summer and fall flux rates. Fez' increases are plotted relative to the concentration of the first sample taken from the incubated flux core (C,,, r = 0). An approximate initial linear rate of increase is assumed and correction for Fe3+ precipitation made in calculating the flux (Section 5.4).

was determined. (C,can also be taken as the first measurement from the flux box and the values o f t offset accordingly.) The volume of water V overlying the sediment in the flux box at a given time and the concentrations C of Mn or Fe in that water at the same time were also measured. The quantity M = V(C - C,) for a given metal was then plotted versus time t, a line fit to the data, and the amount of material entering the water per unit time determined as the slope (Fig. 10). The assumption is that (5.1)

dMldt

=

AJ

where J is the constant flux out of sediment. This assumes that the porewater concentration gradient and overlying water concentration does not change sufficiently during the time of measurement to influence the flux. A is the surface area of core. Neither dissolved Fe or Mn is conservative under oxygenated conditions. Precipitation kinetics of these metals from solutions at fixed pH and 0, are known to be first order or pseudo-first order (Stumm and Lee, 1961; Kester et al., 1975; Morgan, 1967; Lewis, 1976). The concentration change is therefore subject to a loss term of the form (5.2)

dCldt = -kC

366

ROBERT C. ALLER

where k is the first-order rate constant. From (5.2) and the definition of M,

dM/dt = A J - k(M

(5.3)

-

+ VCo)

For simplicity, the conditions: VCo constant, t assumed. These give a solution to (5.3) as (5.4)

M

=

=

0, and M = 0 are

[VCO- (AJlk)] e-kt + (AJlk) - VCo

so that

dM/dt

(5.5)

=

(AJ - kVCo)e-kt

An estimate of the value of k for both Fe and Mn was made by simply transfering water from one of the flux cores into a core box without sediment and measuring the decrease of both dissolved Fe and Mn in that water following isolation (Aller, 1977, p. 429, 440). At 22°C kFe 7.5 x 10-3/min and k,, 8 x lO-'/min with pH 7.2-7.4. The kFe value is in fair agreement with the controlled studies of Kester et al. (1975) and Murray and Gill (1978) who report kFe values in seawater of 3-5 x min at pH 7 and 3-5 x lO-'/min at pH 8 (converted from their reported pH and 0,concentrations and rounded off). These estimates of k allow an evaluation of the use of Eq. (5.1) in flux calculations. Because the factor exp(kt) for Mn in Eq. (5.5) is -1.1 at 22°C after 24 hr, the approximation of Eq. (5.1) is valid. In contrast, the factor exp(kt) for Fe is -10 after only 5 hours. This means dissolved Fe flux estimates using Eq. (5.1) could be low by at least a factor of 10. In order to estimate an order of magnitude of Fe flux, the following approximations were made: (a) the value of Cofor Fe is taken as the first sample from the flux box not the blank box; (b) t = 0 for purposes of calculation is set as the time of the first Fe flux sample, not the time of collection or blank value as listed in Appendix C of Part I; (c) volume in the term VCo is the mean overlying water volume over the sampling period; and (d) dM/df is roughly constant for an initial period of one-two samples points beyond the starting value. This last condition requires from Eqs. (5.4) and (5.5) that

-

(5.6)

-

dMldt

-

- A J - kVCo

The values for kFe in flux cores are assumed to be -7.5 X 10-3/min at 22"C, and based on a temperature dependence for Fe precipitation rates of 10 timedl5"C found by Stumm and Lee (19611, kFe -1.6 X 10-3/min at 15°C. Data of Lewin and Chen (1973) that illustrate Fe precipitation at 10°C in sample bottles of similar size to the flux cores give a calculated

DIAGENETIC PROCESSES.11.

367

kFeof -1.7 x 10-3/min(Aller, 1977, p. 441) indicating that the 15°C value used here is a reasonable one. The last sample points (-30-50 hr after collection) were not used in the summer and fall flux calculations. Mn fluxes were calculated using Eq. (5.1) and one-three sample points in addition to the first water value from the blank box. Fe fluxes were estimated as described earlier. The exact points used are indicated with the flux data in Appendix C of Part I. These points differ somewhat from those used by Aller (1977), and the resulting Mn fluxes are generally lower than those reported previously. In the earlier work, points taken a long time (24-48 hr) after collection often resulted in calculation of higher fluxes than if only initial points were used. In some cases decreases in flux rate were also observed (e.g., DEEP). It is assumed here that the initial rates are the best approximations to in situ rates, particularly because cores were not continuously aerated. Fe fluxes are higher than reported in Aller (1977) due to correction for precipitation loss by use of Eq. (5.6). Curves illustrating types of time-concentration changes in Mn and Fe observed in the flux cores are shown in Fig. 12. The calculated fluxes are listed in Table 11. Mn fluxes are highest in summer and lowest in winter. The observed range is -0.01-4 mmoles/ m*/day over the entire year. No Fe flux was detectable during winter and at other times of year the flux estimate is dominated by the precipitation correction in Eq. (5.6). Order of magnitude estimates for summer and fall are listed in Table I. These range over 0.04-0.5 mmoles/m2/day.

6. DISCUSSION The forms of the pore-water and solid-phase profiles of Fe and Mn found in Long Island Sound (LIS) are similar in many basic features to those reported from other sedimentary basins, particularly the southern Chesapeake Bay (Bricker and Troup, 1975; Robbins and Callender, 1975; Holdren et al., 1975; Elderfield, 1976; Froelich et al., 1979). The exact profile shape and depth distribution of metal concentrations from a given depositional environment reflect the depth-dependent transport-reaction processes occurring there. It was shown in Part I that in LIS, biogenic transport of particles and solutes differs from one area to another and that these differences cause significant variation in the way products of decomposition build up in the upper meter of sediment. These same transport processes can be expected to directly or indirectly influence the sedimentary distribution of Fe and Mn, depending on the type and rate

368

ROBERT C. ALLER

TABLEI. Mn2+ AND Fe2' SEDIMENT-WATER FLUXESTIMATES~ ~

Station FOAM NWC DEEP

Summer (22")

Type of flux estimateb DM DP DM DP DM DP

Fall (15")

Winter (4")

Mn2'

Fe2+

Mn"

Fe2+

Mn2'

0.67 1.7

0.08

0.5

0.08

0.006

3.7

0.1

0.16 0.62 3.5

0.25 0.03

1.6

0.1 0.07

1.4 0.8

0.1

0.2

0.04 0.9 0.001 0.097 0.2 0.56 0.06

0.1

Fe2'

0.01 -

0.001

0.013 0.09 0.1

" Values in mmoles/m2/day. All direct measure estimates of Fe2" flux are corrected for precipitation loss in the flux cores (see Section 5.4): DM, direct measure; DP, diffusion predicted.

of reactions to which these metals respond. Within this context, transport and reaction controls on Fe and Mn at the three LIS stations will now beconsidered indetail and compared with thoseproposedforotherdeposits. Reactions that can result in the elevation of dissolved Fe and Mn concentrations in sediments compared to those found in overlying waters will be considered first. This is followed by a qualitative explanation for the observed seasonality in pore-water profiles and quantitative consideration of reaction rates and sediment-water solute exchange. 6.1. Production of Fez+ and Mn2+ 6.1.1. Oxide Reduction. Fe and Mn oxides can be solubilized as Fe2+ and Mn2+under reducing conditions by both biogenic and abiogenic reactions (Krauskopf, 1957). Because Fe, Mn oxides may act as terminal electron acceptors for some microbial metabolic pathways or as oxidants of reduced products of microbial metabolism, the reduction of each of these oxides in surfcial sediments is often depicted as a generalized metabolic reaction (Bostrom, 1967; Stumm and Morgan, 1970; Froelich et al., 1979; Berner, 1980):

(6.1)

CH20 + 2Mn02 + 3co2

+ H20

or [MnOOH] or [Mn304]+ 2Mn2+ + 4HCOa

(6.2)

CH20 + 4FeOOH

+ 7C02 + H20

or [Fe(OH),I or [FeZ03]+ 4Fe2+ t 8HC0;

DIAGENETIC PROCESSES. 11.

369

The exact oxidation state of the solid in the case of Mn (Bricker, 1965; Crerar and Barnes, 1974) or extent of hydration in the case of both metal oxides, are probably variable and some optional representations are indicated. Of the two metals, Mn oxides are reduced at a higher Eh than are Fe oxides and are therefore more energetically favorable for direct metabolic use as an oxidant (e.g., Garrels and Christ, 1965; Froelich et al., 1979). The rates of these reactions can be viewed as limited by either the quantity of metabolizable organic matter or the abundance of available oxides. The exact influence of the kind of organic matter on reaction rates is, to my knowledge, unknown. However, the quality of organic matter is an important control of microbial activity generally and can be expected to affect the rates of reactions (6.1) and (6.2) directly or indirectly regardless of the quantity of oxide present (Halvorson and Starkey, 1927; Mann and Quastel, 1946; BCtrCmieux, 1951). 6.1.2. Oxidation of Sulfides. In addition to the production in sediments of dissolved Fe2+ and Mn2+during oxide reduction, Fe2+at least could also be released into solution under oxidizing conditions by the oxidation of the reduced phases FeS or FeS, (e.g., Hart, 1963; Stumm and Morgan, 1970; Bloomfield, 1972; van Breemen and Harmsen, 1975; Harmsen and van Breemen, 1975a). Both FeS and FeS, are present near the sediment-water interface at all stations studied from LIS and are subject to continual oxidation during biogenic reworking of surface sediments (Goldhaber et al., 1977; Aller, this volume). The oxidation of pyrite (FeS,) can be represented by the reactions (Stumm and Morgan, 1970): FeS2, + 7/2 O2 + H 2 0 = Fe2+ + 2SO:- + 2H+ (6.3a) Fe2+ + 114 O2 + H’ = Fe3+ + 112 HzO (6.3b) Fe3+ + 3Hz0 = 3 H+ + Fe(OH), (6.3~) FeS2 + 14Fe” + 8 H 2 0 = 15Fe2+ + 2SO:- + 16H’ (6.3d)

At low pH, oxidation of Fe2+to Fe3+is the rate determining step (Singer and Stumm, 1970a), but it is unlikely that this would be true at the high pH levels of marine sediments where ferrous oxidation is rapid (e.g., Kester et al., 1975; Murray and Gill, 1978). Thiobacillus, sulfide-oxidizing bacteria which can help catalyze the previous reactions, are readily cultured from LIS surface muds (Vishniac and Santer, 1957). It seems reasonable to conclude that at least some Fez+could be released into solution near the sediment-water interface as a reactive intermediate during the oxidation of pyrite. The oxidation of FeS is much more rapid than that of FeS, and readily takes place abiogenically at high pH (e.g., Harmsen, 1954; Sato, 1960;

370

ROBERT C. ALLER

Hart, 1963). FeS oxidation releases SO:- into solution (Bloomfield, 1972) so that the overall reactions (6.4a) 4FeS + 8 0 2 = 4Fe2+ + 4SO:(6.4b)

4Fe2+ + O2 + 10H20= 8H+

+ 4Fe(OH),

can result in formation of a ferrous intermediate and acid production. The total production of acid during FeS oxidation is much less than that formed during pyrite oxidation. Because the oxidation of reduced phases is not commonly considered as a source for dissolved Fe or Mn near the sediment-water interface, several simple experiments were performed to determine whether reactions such as Eqs. (6.3) and (6.4) are significant. In one experiment, done with the aid of M. B. Goldhaber, 281 gm of sediment from the FOAM site (mixed from depth intervals 10-11, 13-14, and 16-17 cm of FOAMl), and 2.7 gm of freshly ground (<62 pm) pyrite were each placed in narrow-mouth jars and covered with 2 liters of LIS seawater. The quantity of pure pyrite used was approximately equivalent to the amount in the sediment sample (Goldhaber et al., 1977). A third jar filled only with seawater was used as a blank. A large, glass-covered stirring bar was added, glass tubes passed through rubber stoppers were placed into each jar, the jars sealed, and the solutions vigorously stirred and aerated by use of magnetic stirrers and by passing filtered (0.4-pm), water-saturated air through the glass tubes. The water in each jar was periodically sampled, rapidly filtered through 0.4-pm-pore size Nuclepore filters, and acidified for later analysis of Fe and SO:- using the techniques described earlier or in Part I. A second set of experiments was similar to the first, but in this case sediment was taken from 0-1, 1-2, 2-3, and 3-4-cm-depth intervals of the NWC-4 box core. In each case, a portion of sediment, still frozen following squeezing, was weighed and then placed in a narrow-mouthed, glass jar. Care was taken to remove visually oxidized portions of sediment. Sample 3-4 cm had become crumbly and was oxidized from storage and previous sampling so it was used only for pyrite-loss determination. A portion of the frozen sediment was dried for water analysis in order to determine the dry weight of sediment added to each jar. After covering with 1 liter of seawater, addition of stirring bars and aeration tubes, the samples were vigorously stirred and aerated. Periodic sampling of water and sediment in these and a blank jar was done as described earlier. The pH and alkalinity were measured on unacidified subsamples of jar water in addition to Fe and Mn analyses on acidified samples. The analytical results of the experiments are listed in Tables I1 and 111. These show that in some but not all cases, measurable Fe appears in

37 1

DIAGENETIC PROCESSES. 11.

TABLE11. OXIDATION EXPERIMENT: PYRITE-FOAM SEDIMENT Blank

Pyrite Fe

Time (hr) 0

5 43 116 192 341

SO, (mM) 20.2 21.1 20.6 21.8 21.7 21.2

FOAM Sediment Fe

(PM) 0.066 0.081 0.098 0.075 0.075 0.075

S04(mM) 20.2 20.1 20.6 21.2 21.2 21.7

(~m) 0.066 0.21 0.13 0.084 0.084 0.084

Fe

S04(mM) 20.2 19.9 20.5 21.8 22.1 22.8

(PM) 0.084 1.60 0.22 0.150 0.129 0.142

“solution” following exposure and aeration of sediment (Figs. 11 and 12). When an increase in Fe does occur, concentrations reach a maximum and then decrease to a background level (Fig. 12). An initial increase in Mn also takes place. This increase is apparently more immediate than that of Fe and concentrations again decrease with time (Table 111). There is a tendency for a slight increase in SO:- over week-long time periods in the FOAM sediment jar (Table 11). More analytically dramatic and rapid changes occur in the alkalinities of the jar waters (Fig. 13). Alkalinity decreases regularly with time and the magnitude of the decrease increases with the depth interval from which the sediment sample was obtained. The pH in the NWC jars was initially low due to storage of the seawater used in the experiment, but quickly rose to a relatively constant value during aeration. These results are consistent with the occurrence of reactions (6.3) and (6.4) during sediment aeration. Additional evidence of sulfide oxidation 2 .o I .6

z -3

1.2

N i

2

0.0

0.4

50

100 TIME

I50

200

350

(hr)

FIG.1 1 . Fez’ concentration in overlying water versus time of resuspension in pyrite-FOAM sediment oxidation experiment: 0, FOAM sediment; A,ground pyrite; 0, seawater blank.

372

ROBERT C. ALLER TABLE111. OXIDATION EXPERIMENT: NWC SEDIMENT

Time (hr) 0

Sample (cm) Blank 0- 1

3.5

1-2 2-3 Blank 0-

6.5

10

24

31

180.5

1

1-2 2-3 Blank 0- I 1-2 2-3 Blank 0- 1 1-2 2-3 Blank 0- 1 1-2 2-3 Blank 0- 1 1-2 2-3 Blank 0- I 1-2 2-3

pH 7.77 7.77 7.75 7.77 7.82 7.77 7.79 7.81 7.91 7.81 7.82 7.82 7.92 7.93 7.91 7.93 7.96 7.88 7.89 7.91 8.01 7.93 7.94 7.88 8.01 7.92 7.93 7.89

Alkalinity (meq/liter)

Fe Mn ( p M ) ( p M )

2.28 2.28 2.28 2.28

0.1

0.01 0.01 0.01

2.16 2.20 2.22

3.3

0.34 0.089 0.095

2.14 2.16 2.18

0.97

2.23 0.045 0.051

2.12 2.10 2.10

I .6

3.97 0.068 0.051

2.10 2.08 2.06

0.61

1.17 0.063 0.058

2.04 I .94 1.82 2.30 2.04 1.84 1.46

0.51

1.02 0.054 0.10

0.3

0.37 0.092 0.053

Sample (cm)

Wt dry sediment /liter (gm)

0- I 1-2 2-3

6.32 6.88 7.25

is provided by measurement of total sulfur remaining in NWC samples from 0-1- and 3-4-cm depth after 4.5 months of exposure. An average loss rate of 50 ? 12 pmoles S/gm/yr was found (sulfur analytical method described in Part I). No Fe release was detectable in the NWC samples from 1-2 and 2-3 cm, presumably because rapid Fe precipitation onto suspended particles prevented a measurable buildup in each case. The abundance of fresh organic matter in the 0-1-cm interval or slight differences between sam-

373

DIAGENETIC PROCESSES. 11.

4 h

3

3

+

N

9

2 I

t (hr) FIG.12. Fe2’ concentration in overlying water versus time of resuspension and oxidation of NWC sediment from 0-1 cm.

ples in the nature of that material may have aided in maintaining Fe in solution for a long enough time to be observed during vigorous aeration in both the 0-I-cm NWC and FOAMjars (e.g., Moore and Maynard, 1929; Rashid and Leonard, 1973; Picard and Felbeck, 1976). At the lower in situ 0, concentrations present in surface sediments, the precipitation of Fez+produced during sulfide oxidation (which occurs at a lower Eh than Fe” oxidation) would be slower than in the present experiments and a resulting measurable buildup of the Fe2’ intermediate would be more likely.

a

1.4

10

20

30

40

t (hr)

.,

50

60

70

80

FIG. 13. Alkalinity of overlying water versus time of resuspension and oxidation of N W C sediment from different depth intervals (0,0-1 cm; 1-2 cm; A,2-3 cm).

374

ROBERT C. ALLER

The cause of the increase in dissolved Mn during suspension is unknown. The immediate appearance of Mn in solution as compared with the more gradual apparent release pattern of Fe (Table 111, Fig. 12), suggests that desorbtion of Mn2+from sediment particles upon exposure to low Mn” waters could be one explanation (e.g., Murata, 1939). Oxidation or dissolution of an unidentified reduced phase is also possible. The decrease in alkalinity and the increasing magnitude of the decrease associated with deeper sediment samples can be explained in part by reactions (6.3) and (6.4) and the known increase in FeS below the sediment-water interface at NWC (Part I). It can be shown that the small effect of sulfide oxidation on pH is in agreement with the magnitude of the alkalinity change and the experimental conditions. Because of vigorous aeration, the experimental jars can be considered to be held at a constant COz pressure (by the end of the experiment) so that if the carbonic acid system is controlling pH, the relation (6.5)

pH1 - PHZ = log Alki - log Alkz

describes to a good approximation the expected change in H’ activity related to change in carbonate alkalinity (asymptotically equal to titration alkalinity). This assumes a H %- KS where u H is the H’ activity and K ; is the second, apparent dissociation constant of H,CO, in seawater. With the maximum observed change in alkalinity in the 2-3-cm jar, a pH difference relative to the final blank jar value of only -0.19 units is expected, the measured value is 0.12. Considering the measurement errors and the approximations involved in this calculation, agreement seems good so that the small change in pH is not inconsistent with both the production of acid and the observed lowering of alkalinity. There are several quantitative problems with interpreting all the changes observed in the oxidation experiments as resulting from the dissolution of FeS or FeS,. The first is that -4 times too much Fe is observed in solution (0-1 cm, NWC sample) as can be accounted for by the total amount of FeS (-0.6 pmoles) added or the estimated average pyrite oxidation rate of -50 pmoles/gm/yr. The observed decrease in alkalinity in the samples from 1-2 and 2-3 cm also requires -3-4 times the quantity of FeS actually added if the oxidation reaction (6.4) as written as stoichiometrically correct and the average long-term pyrite oxidation rate is again assumed. The release of additional, perhaps organic, acids, oxidation of other reduced species, or a more rapid initial pyrite oxidation rate seem likely. In summary, theoretical predictions indicate that Fe2+can be released by oxidation of FeS and FeS, near the sediment-water interface. A destruction of alkalinity is associated with these reactions. The preliminary

375

DIAGENETIC PROCESSES. 11.

experimental results are consistent with but do not unequivocally prove this expectation. As will be shown later, field measurements are also suggestive of solid-phase sulfide oxidation as a source of Fe2+ and alkalinity consumption in surface sediments. 6.2. Precipitation Reactions and Saturation States After they are generated by dissolution reactions such as those outlined previously, both Fe2+and Mn2+can reprecipitate as oxides upon exposure to O2or, under anoxic conditions, form insoluble phases with anions commonly produced during anaerobic metabolism. Some geologically likely reduced-metal carbonate and sulfide phases, together with their solubility products are listed in Table IV (see reviews of Nriagu and Dell, 1974; Rickard, 1969; Berner, 1970; Troup, 1974; Holdren, 1977); representative Fe-, Mn-phosphate compounds are listed in Table VI of Part I. (p. 308). The removal of Fe2+or Mn2+from anoxic pore waters by such reactions can be readily demonstrated by incubating sediment and following the Fe" or Mn2+ concentrations as the products of anaerobic metabolism buildup (e.g., Norvell, 1974). This is illustrated for Mn in Fig. 14, which shows pore-water Mn2+and SO',- concentrations as a function of time during the anoxic incubation of salt-marsh sediment from LIS [The incubation experiments and SO$- analyses are those reported in detail by Martens and Berner (1974) who kindly made pore-water subsamples available for Mn determination.] Mn concentrations rapidily respond to the presence of anaerobic metabolities formed during SO',- consumption, and once SO',- is used up, come to a relatively constant level suggestive of saturation with respect to a reduced phase. TABLEIv.

SOLUBlLlTY PRODUCTS AT 25°C AND

1 Atm FOR SELECTED SULFIDE AND CARBONATE COMPOUNDS

Solubility product (log KqJ

Reaction FeS (amorphous) FeS (mackinawite) Fe,S, (greigite) FeCO, (siderite) MnS (alabandite) MnS2 (hauerite) MnCO, (rhodochrosite)

e Fe2+ + S2-

Fez+ + SzSo F?3Fe2+ + 3s'# F e z + + CogF? Mn2+ + Sz# Mn2+ + SzSo F? MnZr + C0:#

+

+

- 16.9 - 17.5 - 18.2

- 10.2 - 17.8

-20.6 - 10.4

Ref. Bemer (1967) Berner (1967) Berner (1967) Singer and Sturnm (1970b) Mills (1974) Mills (1974) Morgan (1967)

376

ROBERT C. ALLER

TIme (days)

FIG.14. Concentration of MnZ' (open circles) as function of time during anoxic incubation of salt-marsh sediment. SO:- decrease during same period (Martens and Berner, 1974) is plotted as dashed line.

The likelihood of a specific solid phase forming in sediment and influencing the concentrations of dissolved Fe and Mn can be inferred from direct analysis of solid components, together with solubility calculations based on pore-water composition (Bricker and Troup, 1975; Holdren et al., 1975; Suess, 1979). Unfortunately, direct analysis by x-ray or wetchemical techniques is impossible for any but the most abundant phases, and in the case of wet-chemical analyses, often involves operational definitions of a component. Equilibrium-solubility calculations using porewater analyses are limited by analytical uncertainties (e.g., speciation, abundance) that become magnified in importance during calculation; by kinetic effects; by the fact that sediments are open systems and near the interface are often nonsteady state; and by lack of thermodynamic information on any but a few pure, well-crystallized end-member phases. Additional problems can arise because of multidimensional diffusion in the bioturbated zone as described in Part I and discussed subsequently. Despite these drawbacks, the use of both direct measurement and equilibrium calculations can provide insight and establish limits on the possible reactions and transport processes controlling Fe and Mn pore-water distributions (e.g., Bricker and Troup, 1975; Holdrkn, 1977; Martens et al., 1978; Suess, 1979). Ion activity products (IAP) for FeS, FeCO,, MnS, and MnC03 were calculated from pore-water samples taken at FOAM, NWC, and DEEP

DIAGENETIC PROCESSES. 11.

377

during 1975-1976 (see Appendix B of Part I for data). These calculations, together with the solubility products for the compounds listed in Table IV, are plotted for each station in Figs. 15 and 16. Calculations for phosphate compounds are given in Part I. Saturation of pore water with respect to elemental sulfur is assumed in the case of greigite and hauerite activity product comparisons. Ion activities were estimated in each case by use of an ion-pairing model written and developed by Holdren (1977) as described in Part I. The range of the total activity coefficient predicted by this model for Fe2+ in LIS pore waters is 0.14-0.19 and for Mn2+,is 0.11-0.14. These are in general agreement with those used in other studies (e.g., Li et al., 1969; Troop, 1974; Murray and Gill, 1978). No correction for temperature variation was made. The carbonate solubility calculations assume that titration alkalinity is equivalent to carbonate alkalinity (Berner et al., 1970), thereby slightly overestimating carbonate activities. pS values for summer box cores (1975) were estimated from Eh measurements by use of the conversion equation of Berner (1963), which requires that the sulfide-elemental sulfur couple control Eh. These estimates are approximate, but reasonable, based on subsequent direct pS measurements at each station. pS values for FOAM and NWC gravity cores were obtained from different cores than the actual Fe2+analyses (Appendix B, Part I). 6.2.1. Fe Equilibria. In the case of Fe, pore waters at all stations are generally distinctly undersaturated with respect to siderite (FeCO,); although near saturation may occur in a thin band in the upper -5 cm corresponding to the zone where Fe2+concentrations are high or, at the FOAM station, below 40 cm (Fig. 15). Because total alkalinity was used, these undersaturations are minima. Pore waters in the upper 20 cm tend to be saturated or supersaturated with respect to fine-grained FeS (amorphous) at FOAM, variably saturated or undersaturated with respect to mackinawite or greigite at DEEP, and generally undersaturated with respect to these phases at NWC. Below 20 cm at NWC and DEEP, pore waters are undersaturated with respect to the monosulfides. If sulfide equilibria control Fe concentrations below 20 cm, then the stability of the corresponding solid-phase sulfide must increase at deeper water stations. Calculated undersaturations occur despite the abundant presence of Fe sulfides in the solid phase (Part I). Departures from equilibrium can be due to any of a number of factors, for example: (1) analytical measurements of Fe2+or pS are inaccurate; (2) Fe2+ is complexed by organics (results in supersaturation) (e.g., Nissenbaum et al., 1972; Krom and Sholkovitz, 1978); (3) the solubility products are incorrect; (4) an uni-

FOAM

- log(lAP1,

Gravity Cores

FeS

18

16

DEEP

NWC

20

20

18

16

16

20

18

22

I8

16

20

0

I

I

0 0

0

5

0'

DEEP

10

I 10

I I

I

1

I

;

I 1 Ksp greigite

I K~:

K sp

14

fna

o n o

o no OO#

80

l

o

1

o

0

1

8

,Cbn ?P

3 I

0

B

0

I Ksp

'

o n 0

I d o

14

12

I

"0

I l

I

I

I

1 Ksp

I

I I

I K , ~ greigite

I Ksp

greigite

omorph.

amorphous

FeC03

'BO0O

I

greigite

160

I 1 KSp

b o n

b

@

@

0

mackinawite

12

I

15

amorphous

- log(lAP), 10

120

I no

0

I I

Ksp

I

-

80

siderite

I 0 I K s p siderite

:I

I0

I

40

m

' d B '

15

I

I Ksp

0

0 siderite

Ksp

siderite

FIG. 15. Plots of -log IAP for solid phases: FeS and FeCO, as a function of depth in sediment at FOAM, NWC, and DEEP. Box cores:

0.summer (1975): 0, fall (1975); A,winter (1976). Gravity cores: 0 , FOAM; W, NWC; A,DEEP.

0

0

9

!

-

O D -

;i 0

t

0

*

t

.

a

Y)

u)

0

(

0

D

0

0

0

*-'

E

n f

-

0

0

0

E

v) -

'D

0

m 0

p

0

7 '1

0

0

*

.

u)

(

0

D

41

0

E

E

0

!!?

.

d

8

W

2m 3

6 z

m

*

0

'r

..

L

2

380

ROBERT C. ALLER

dentified metastable sulfide phase controls solubility (undersaturation or supersaturation); ( 5 ) the presence of microenvironments created by burrows prevents accurate calculation of saturation states from average porewater concentrations (Part I); or (6) additional reactions other than sulfide formation are controlling Fez+.Of these possibilities, I think it is likely that pS values are slightly high in the case of undersaturation (NWC, DEEP) and that organic complexes are a likely explanation for supersaturation at FOAM (Berner et al., 1978). In the absence of abundant sulfide it is possible that reduced Fe phosphates such as vivianite might precipitate (Nriagu and Dell, 1974; Bricker and Troup, 1975; Martens et al., 1978). This possibility can be checked in part as follows. The interconversion between FeS (mackinawite) and Fe,(P0,)2 - 8H20(vivianite) is represented by (6.6) 3FeS + 2PO:- + 8 H 2 0 = Fe2(P04)2.8 H 2 0 + 3.3’which from the relative free energies (Nriagu, 1972; Berner, 1967) gives at equilibrium (6.7) (s2-)3/(po:-)2 = 1 0 - 2 2 . 2 or pP04 =2pS - 1 i . l Data from the upper 8 cm at all stations where both pS and pP0, were detectable are plotted together in Fig. 17 with the equilibrium line given by Eq. (6.7). Most points lie within the mackinawite rather than the vivianite fields, demonstrating that sulfides should form in preference to vivianite in this zone of LIS sediments. It was shown in Part I that vivianite could form in the upper few centimeters of sediment during at l3

t

Mackinawite /Vivionite

12

-

II

-

0 a 10

o

-onoo

9 10

12

14

16

18

20

PS

FIG. 17. pP0, versus pS for upper 8 cm at FOAM, NWC, and DEEP where pS was detectable. The line drawn is the equilibrium relation between vivianite (Fe3(P0& . 8Hz0) and mackinawite (FeS). 0, FOAM; 0. NWC; A,DEEP. All seasons plotted. The formation of sulfides is generally favored, but regions exist near the sediment-water interface where coexistence of phases or phosphate dominance are possible.

DIAGENETIC PROCESSES. 11.

38 1

least some times of the year at all stations, but at that depth, pore waters are undersaturated. Such undersaturation is consistent with the present independent calculations. In summary, equilibrium calculations based on pore-water composition indicate that Fe carbonate or phosphate are unlikely to form below the top few centimeters of sediment and that siderite formation is unlikely generally. Undersaturations by a factor of K,,/IAP -100 are found. The presence of solid-phase sulfide is evidence for the formation of Fe sulfides. However, pore waters are not always saturated with respect to the common Fe-sulfide minerals and under-or supersaturations by a factor of -10 are calculated. These deviations may be due to such problems as organic complexing or cumulative analytical and sampling errors, but the possibility that other phases are influencing Fe2 concentrations cannot be excluded on the basis of these data. +

6.2.2. Mn Equilibria. The range of values reported for the K s pof MnS (see Holdren, 1977, for summary) make it impossible to eliminate the likelihood of MnS formation in LIS on the basis of solubility calculations alone. MnS reported from sediments is generally of the hexagonal polymorph (Suess, 1979). The K , , listed in Table IV and plotted in Fig. 16 is, however, based on the free energy of formation of the cubic polymorph (alabandite) because of better documentation of its value (Mills, 1974). Ion-activity products in pore waters from the upper 20 cm at NWC and DEEP can be quite close to this K,,, but FOAM waters are substantially supersaturated. All pore waters, except in the gravity core from DEEP, are supersaturated with respect to MnS2 (hauerite). Just as in the case of MnS, because of the large uncertainty in the accepted K , , for compounds such as Mn3(P0,)2* 3 H 2 0 (reddingite) and the tendency for solid solutions with Fe (Tessenow, 1974), it cannot be said with any certainty that Mn phosphates are not forming in LIS. It was demonstrated in Part I that IAPs in pore waters at DEEP come closest to the K , , (reddingite) selected for use. MnC03 (rhodochrosite) or mixed Mn, Ca, Mg carbonates are believed to be forming in many marine environments and to be acting as important controls on pore-water Mn” concentrations (Li et al., 1969; Calvert and Price, 1970, 1972, 1977; Holdren et al., 1975; Suess, 1979). IAP calculation indicates that at FOAM the pore-water samples taken from box cores are very near saturation with MnC03 below -4 cm during the summer and winter but are undersaturated during the fall (Fig. 16). The same pattern is found at DEEP (Fig. 16). NWC box cores are undersaturated (-5-10 x ) with respect to rhodochrosite. Gravity cores indicate that deeper pore waters are slightly (-5 X ) undersaturated at NWC and DEEP and are

382

ROBERT C. ALLER

undersaturated to a greater extent at FOAM. Part of this discrepancy could be due to not having pH values on the same gravity cores at NWC and FOAM for which alkalinity and Mn were obtained, but because of the slight variation in pH, this should not be a major problem. In summary, it seems likely that below a few centimeters depth in LIS sediments, Mn comes into fairly rapid equilibrium with solid-phase components such as MnCO,, although at present the exact components remain unknown. 6.3. Seasonality of Pore- Water Fe and Mn Profiles Near the Sediment- Water Interface

The shape and seasonality of the dissolved Fe and Mn pore-water profiles in the upper 20 cm can be qualitatively explained by the relative seasonal importance of transport processes and the production-consumption reactions just described. During the summer, the warming of overlying water and the deposition of organic matter from the winter-spring plankton bloom cause an increase of microbial metabolic activity within the sediments (Part I). Oxygen is relatively rapidly consumed near the interface at this time as reflected in the seasonal migration upward of the average boundary between oxidized (yellow-brown) and sulfide-rich (black) sediment (Rhoads et al., 1977). As a result of this increase in metabolic activity, Mn and Fe oxides are reduced near the sediment-water interface causing a buildup of both Mn2+and Fe2+in the upper few centimeters (Figs. 2-7). Mn2+builds up to higher levels closer to the interface than does Fe2+because Mn oxide is reduced at a higher Eh than Fe oxide and Mn2+is less rapidly reoxidized than is Fe2+(e.g., Krauskopf, 1957; Garrels and Christ, 1965; Morgan, 1967; Kester et al., 1975). Once solubilized, Fe2+ and MnZ+diffuse either away from the sediment-water interface where precipitation as a reduced phase can take place or toward the interface where they are oxidized or released into overlying water. Lateral diffusion of Fe2+and Mn2' into burrows can also lower concentrations at depth in the sediment, as will be discussed later (see also Part I). The net results of these processes are maxima in average Fez+and Mn2' concentration close to the sediment-water interface. The exact magnitude and depth distributions of these maxima depend on the particular balance between transport and reaction occurring at each station. For example, in addition to having high-Mn2+concentrations near the interface during the summer, DEEP has a maximum between 5-8 cm and high Mn2+ to at least 14 cm. This may result from the much smaller attenuation of microbial activity with depth below the interface at this station (Aller and

DIAGENETIC PROCESSES. 11.

383

Yingst, 1980), which allows production of Mn2+ at depth, together with the abundant presence of tube-dwelling benthos that prevent the buildup of metabolites such as HCO; that might lower MnZ+by anoxic precipitation (Part I). In the fall the maximum concentrations of Fe” and Mn”, or the vertical extent of concentration peaks, decrease relative to those found at the same station during summer. Peak concentrations are sometimes deeper in the sediment and concentrations can also be generally lower throughout the sampled region as occurs for Mn2+at DEEP (Fig. 7). These changes apparently result from a combination of lower production rates of Fe” and Mn2+,together with the relatively increased importance of biogenic transport at this time compared to summer. Production rates decrease because of decreased microbial activity associated with lower water temperatures and depleted food sources [e.g., Eqs. (6.1) and (6.2)]. This allows the transport effects of biogenic reworking and irrigation to dominate the form of the fall solute profiles relative to those of the summer period. Similar changes are also observed for other porewater constituents (Part I). Such differences in profile shapes resulting from temporal changes in the relative importance of biogenic transport and reaction rates can be reproduced in the laboratory. NWC sediments reworked in laboratory tanks by Yoldia limatula, a deposit-feeding bivalve common at that station, contained a depleted Fe” profile compared with that found in a tank without macrofauna, but otherwise equivalent (Aller, 1978). Relative differences in the importance of biogenic reworking between the two tanks therefore produced differences in pore-water profiles similar to those found in LIS sediments between summer (or winter) and fall. During the winter, Mn” concentrations continue to decrease in the top few centimeters compared to fall profiles, Maximum concentrations occur deeper in the sediment column and the vertical extent of concentration peaks can broaden or Mn2+ increase well below the interface (Figs. 5-7). These changes are consistent with a continued seasonal decrease in the production rates of Mn2+,as well as a decrease in consumption rates by precipitation with anaerobic metabolites or biogenic transport at depth in the sediment. Both changes are caused by the low winter temperatures (2-4°C) and the associated decrease in biological activity. Except for the added complications of precipitation reactions, the overall seasonal patterns of change in Mn2+pore-water distributions in the upper 20 cm are therefore similar to those found for metabolites such as NH; at each station (Part I). In contrast to Mn”, Fe2+ concentrations increase again in winter, sometimes to the highest concentrations of the year (Figs. 2-4). Peaks

384

ROBERT C. ALLER

in Fe2+ concentration can also occur closer to the sediment-water interface than Mn2+ maxima at this time (Figs. 2, 4, 5, 7). At DEEP, the winter maximum in Fez+ is associated with a slight decrease in the alkalinity profile (see Fig. 21 or Alk-Cl ratios for Appendix B, Part I), although no such concavities were detected at FOAM or NWC. Similar decreases in alkalinity that mimic Fe2+maxima were found during winter at the New Haven dump site (1975, unpublished) and can also be produced in the laboratory when macrofauna are removed from sediment (Aller, 1978). These observations, together with the known winter decrease in FeS content of surface sediments (Part I) and the findings of the oxidation experiment described previously, indicate that oxidation of Fe sulfide is a major source of Fez+at this time. Because these reactions can be entirely abiogenic, they would not only explain small decreases in alkalinity and the fact that Fe2+maxima are closer to an oxygenated interface than Mn2+ maxima, but would also explain the large increase in released Fez+during times of the year when biological activity is lowest. The decrease of Fez+ with depth indicates that sulfide formation is still taking place and that Fe-oxide reduction, as in the case of Mn2+,contributes to Fe" production during the winter as well. It is concluded that Fez+ production, unlike MnZ+production, is due to two sources that vary in a relative importance seasonally: (1) the biogenic reduction of Fe oxide that dominates during summer, and (2) the abiogenic (or biogenic) oxidation of Fe sulfides (an indirect metabolic product) that dominates during winter. 6.4. Production and Precipitation Rates 6.4.1, Mnz+ and Fe2+ Production-Rate Estimates from Solid-Phase Distributions. The redistribution of Fe and Mn within sediments during early diagenesis has been recognized and modeled for many years (Murray and Irvine, 1895; Lynn and Bonatti, 1965; Manheim, 1965; Bender, 1971; Bonatti et ai., 1971; Robbins and Callender, 1975; Harmsen and van Bremeen, 1975b; Thomson et al., 1975; Elderfield, 1976; among many). The common enrichment of, for example, Mn in surface relative to deeper sediment is a necessary consequence of pore-water profiles such as depicted in Figs. 5-7, Fick's first law of diffusion, and net sedimentation. The resulting concentration gradients in solid-phase distributions of Mn or, equivalently, Fe allow estimates at steady state of the Mn2+or Fe2+ production or precipitation rates required to sustain those gradients (e.g., Michard, 1971; Robbins and Callender, 1975). The average vertical distributions of solid-phase Fe or Mn are determined by biogenic and physical reworking, net sedimentation, compaction, and dissolution-precipitation reactions. These distributions can

385

DIAGENETIC PROCESSES. 11.

often be adequately described by the general one-dimensional transportreaction equation (after Berner, 1980) (6.8)

aps(1 - d C S=

at

2 De apA1 - W, - aops(l ax

ax

ax

+

where metal concentratiodmass of dry sediment space coordinate, origin at sediment-water interface, positive axis into sediment time DB particle-mixingcoefficient o sedimentation rate in lengthltime p. particle density cp porosity R generalized reaction term C, x t

Only solid-phase distributions in the top 5 cm of LIS sediments will be considered here. It is known from previous studies of excess 234Thand ’“Pb activity profiles that in this zone of LIS deposits DB >> o (Aller and Cochran, 1976; Benninger et al., 1979; Aller et al., 1980). DB can also be considered approximately constant in the top 5 cm (Aller and Cochran, 1976), allowing simplification of Eq. (6.8) in this case to

ac,/at

(6.9)

cs

=

DB(d’c^Jdx’)

+R

where is the metal concentratiodvolume total bulk sediment. Because of physical disturbance, the solid-phase Mn and Fe profiles at FOAM are irregular and cannot be effectively treated quantitatively. The distributions at NWC and DEEP are sufficiently regular to be modeled and those at NWC are generally repeatable enough that an assumption of steady state for Eq. (6.9) seems reasonable (Figs. 8 and 9). Steadystate solid-phase distributions are assumed for DEEP as well. It is not possible to determine an absolute reaction rate from any of these profiles because the percentage of the total metal that is diagenetically mobile at any given depth is unknown. A relative net reaction rate for the upper region compared to deeper sediment can be calculated by subtracting an assumed background concentration value from the concentration in the zone to be modeled. This means that only a profile of “excess” Mn or Fe is utilized to estimate reaction. The net reaction rate will therefore represent an average dissolution rate (R negative) of excess Mn or Fe in the modeled zone. This will be a good estimate of the actual dissolution rate if most reprecipitation takes place at or outside the boundaries of the modeled regions. No estimate of Fe dissolution can be made at DEEP because total Fe decreases toward the interface. Relative Mn loss rates for the upper 4 cm at NWC are determined as follows. The Mn concentration (pg/gm) between 3 and 4 cm was sub-

386

ROBERT C. ALLER

tracted from each of the overlying sample intervals. These “excess” concentrations were converted into units of mass/total volume by assuming a sediment-particle density of 2.5 gm/cm3and utilizing the measured water contents (Appendix B, Part I). Excess concentrations were plotted and found to be adequately fit by simple exponential functions; A t . , = ACso exp( -ax) (Fig. 18; Table V). These functions were then substituted into Eq. (6.9) and an estimate of R was obtained using D B = 0.43 x cm2/ sec (Aller et al., 1980). This D e is the average, not instantaneous, value for the upper 5 cm at NWC during summer; winter-spring values at NWC or summer values at nearby stations do not differ significantly from this D B(Aller and Cochran, 1976; Aller et al., 1980). Resulting estimates of dissolution rate functions of the form R = R, exp( -ax) in the upper 4 cm at NWC are listed in Table V. The lowest rate is found for NWC-1, whereas the estimates from the three other cores are in fairly good agreement. It is possible that a small amount of erosion (e.g., 0.5 cm) could have occurred at NWC prior to collection of NWC-1 and thereby caused a lower estimate (see, for example, Benninger et al., 1979). The Mn profile at DEEP was treated in the same way, except that only the top 3 cm was considered. DB from 234Thexcess activity profiles (3 cm) is -0.15 x cm2/secat this station (Aller et al., 1980). Resulting exponential functions for excess Mn (At.,) and reaction rates from Eq. (6.9) are listed in Table V. It is possible to estimate a turnover time T for excess Mn in the top few centimeters L at each station from the excess Mn profiles Ats and the estimated reaction rates R. T is given by (6.10)

T

=

( L j k A t s dx)/(Jo“R dx)

The calculated values of T are listed in Table V. Except for the high estimate from NWC-1 these show that the excess Mn present in the top few centimeters of sediment is completely dissolved on the order of every 60-100 days. This is a more rapid estimated turnover than distributions

FIG. 18. Excess Mn and Fe profiles for cores NWC-2 and NWC-4. The curves plotted are those used to calculate solid-phase dissolution rates. Note that the vertical axis for excess Fe is offset to a depth of 1 cm.

TABLEv. Mn AND Fe EXCESSDISTRIBUTIONS AND REACTION RATES

Core NWC-I NWC-2 NWC-3 NWC-4

DEEP-I

Core NWC-2 NWC-3 NWC-4

Model interval (cm) 0-4 0-4 0-4 0-4 0-3

Excess Mn (~dcm’) 11I exp( - .077x) 477 exp( - 1 . 2 ~ ) 366 exp( - 1 . 0 8 ~ ) 501 exp( - I .35x) 162 exp( - 1 . 5 ~ )

Model interval (cm) 1-4 2-5 1-4

Mn dissolution rate (~glcm’lday) 2.4 exp( - 0 . 7 7 ~ ) 26 exp( - 1 . 2 ~ ) 16 exp( - I .08x) 34 exp( - I .35x) 4.7 exp(- 1 . 5 ~ )

Excess Fe (mg/crn3) 2.6 4.3 2.2

+ 0 . 4 7 ( ~- 1) - 0.61(~+ 0.40(~- 2) - 0 . 8 5 ( ~- 2)2 + 0.41(~- I) - 0.18(~- 1)’

Turnover time T (days) 185

13 92 59 103

Fe dissolution rate (pglcrn3/day1 45 63 13

First-order dissolution rate const. k (/day)

Production-supported flux estimate JMc. (mmoles/mz/day)

0.022 0.054 0.043 0.067 0.029

0.57 3.9 2.7 4.6 0.57

Turnover time (days) 104 I16 232

T

Production-supported flux estimate J F e (mmoles/m2/day) 24 34 7

388

ROBERT C. ALLER

from other environments predict (e.g., Robbins and Callender, 1975), due in part to the availability of biogenic mixing coefficients (D.) from 234Th activity profiles in this study. Previous calculations have assumed average long-term sedimentation rates as the major transport agent of particles within the sediment column. The present estimates are in agreement with laboratory experiments, which showed that Mn profiles similar to those found in the field could form in -45 days (Aller, 1978). In some studies it has been assumed a priori that Mn2+production rates or, equivalently, oxide dissolution rates, can be represented as a firstorder reaction where the rate of solid-phase dissolution is proportional through a constant k to the quantity of excess or sometimes total oxide present, that is, R = - k ( A e , ) (Holdren et al., 1975; Elderfield, 1976). Substitution of this function into Eq. (6.9) shows that this is the same as assuming an exponentially decreasing dissolution rate having an attenuation coefficient = (k/D,)”’. The attenuation coefficients a found in this study by direct substitution of solid-phase distribution data into Eq. (6.9) can be equated to (k/D.)”’to obtain values of k for comparison with other work. The values of k = a2DBare listed in Table V and range from 2.6-7.8 x lO-’/sec or 8.2-25/yr. These are about 103-104times greater than model values calculated for other near-shore sediments (Holdren et al., 1975). One internal check on whether the Mn2+production rates are reasonable can be made by calculating the maximum sediment-water flux of Mn2+ that could be supported by these rates. At steady state, this flux JMnis simply equal to the average production rate over the modeled interval L, which from Table V and Eq. (6.9) is (6.11)

JMn= (Ro/a)[l- exp(-aL)]

This requires that at steady state, what Mn2+leaves the sediment can be no greater than that produced. Because of precipitation of oxides at the interface, the flux out of the sediment into overlying water can, of course, be less than predicted by (6.11). The maximum steady-state sediment-water fluxes predicted by (6.11) and the estimated production rates are listed in Table V under JMn.Estimates range from 0.6 mmoles/m2/dayat DEEP to 5 mmoles/m’/day at NWC and are in excellent agreement with the magnitudes of actual measured fluxes (Table I). This suggests that: (1) the production rate estimates made from the solid phase are reasonable, and (2) that a large fraction of the Mn’’ produced in these sediments near the interface escapes into the overlying water and is reprecipitated there rather than in the sediment deposit. The agreement may be more apparent than real because only minimum production rates can be estimated from the excess metal profiles. Additional checks on the calculated rates come from the pore-water distributions as developed later.

DIAGENETIC PROCESSES. 11.

389

Estimates of Fe2+ production at NWC can also be made in the same way as done for Mn”. Only cores NWC-2, 3, and 4 have solid-phase Fe profiles that lend themselves to modeling of this type (Fig. 8). Slightly different depth intervals were used for calculation of excess Fe for each core because of small differences in the forms of the profiles. In all cases the top 0-1 cm was considered an obvious region of net precipitation and not used. The modeled depth intervals are 1-4 cm (NWC-2, NWC-4) and 2-5 cm (NWC-3). The Fe content of the bottom-most sample in each selected interval was subtracted from overlying samples to obtain measThe bottom sample selected as a base concenures of excess Fe, tration is either a minimum value in the respective Fe profile or representative of a relatively constant Fe concentration below that depth. Excess concentrations were plotted and found to be fit relatively well as a function of depth by a second-order polynomial (Fig. 18, Table V). Insertion of these functions into Eq. (6.9) results in an estimate of a constant dissolution rate for Fe in each modeled interval. These range over 0.013-0.063 mg/cm3/dayand are best considered order of magnitude rates because of the subtraction of the large background value from the profile in each case. Values of the turnover time 7 for excess Fe in each interval were calculated using Eq. (6.11) and vary from -100-200 days, somewhat longer but similar to those found for excess Mn. The maximum expected steady-state flux of Fe2+from NWC sediments into overlying water is approximately J F e = RL, where L is the thickness of the modeled interval and R is the constant Fe dissolution rate in that interval. Calculated fluxes range from 7-34 mmoles/m2/dayand are considerably higher (10-100 x ) than directly estimated fluxes from incubated box cores (Tables I and V). This is consistent with the rapid oxidation kinetics of Fe” in seawater (Aston and Chester, 1973; Kester et al., 1975) and the expected precipitation of most mobilized Fez+ from solution as the oxygenated interface is approached. One major conclusion from these calculations is that there is at least as much Fe mobilized as Mn and probably as much as 10 times more. This redistributed Fe is much harder to analytically detect above its lithologic background.

AeS.

6.4.2. Mnz+ and Fez+ Reaction Rates from Pore- Water Distributions. The dissolved Mn2+ and Fez+ profiles can also be modeled by transport-reaction equations equivalent to Eq. (6.9) or (6.10) (Anikouchine, 1967; Spencer and Brewer, 1971; Michard, 1971; Robbins and Callender, 1975; Holdren et al., 1975). This allows independent checks on the form and magnitude of MnZ+and Fe2+ production rates estimated from the solid-phase distributions. In addition, modeling allows estimation

390

ROBERT C. ALLER

of precipitation rates and evaluation of the effects of biogenic irritation on pore-water profiles. It was shown in Part I that it is possible to describe solute profiles in the bioturbated zone by defining an average sediment microenvironment and determining solute distributions within it. Because the sediment is composed of many such microenvironments, an average vertical solute profile found in the sediment must correspond to that in a single microenvironment. The average microenvironment is assumed to be a finite hollow cylinder or annulus of sediment oriented vertically in the deposit. This portion of sediment essentially corresponds to that associated with a single ideal infaunal animal. The inner radius of the cylinder r , is determined by the average size of macrofauna inhabiting the sediment and irrigating burrows, and the outer radius of the annulus r2 is determined by the effective abundance of individuals. In the simplest case where animals are immobile tube dwellers and are evenly distributed, r , is the average tube radius and r z is simply half the distance between individuals as measured from the tube axis (Aller, 1977, 1978, 1980, Part I). The vertical length of the cylinder microenvironment L corresponds to the effective thickness of the bioturbated zone. The transport-reaction equation describing pore-water distributions in the imaginary cylinder is (cylindrical coordinates) (6.12)

'a = D-a(r$) at

r ar

ac

+ D y + R

ax

where all symbols are as before and r is the radial dimension measured from the cylinder axis. Porosity is assumed to be approximately constant with respect to pore-water volume and is not explicitly written into the equation. Advection is also ignored because pore-water distributions in the top -20 cm over short time periods only are considered here (e.g., Lerman, 1975). Adsorption is not included because only the steady-state case will be considered (Berner, 1976). D is the molecular diffusion coefficient in the bulk sediment and is assumed both isotropic and constant [equivalent to D , of Berner (19801. If there are no or very few irrigated burrows present in the sediment, lateral diffusion is not significant and the r dependence of Eq. (6.12) can be ignored. In that case, the equation becomes the more traditional onedimensional transport-reaction equation used to model pore-water solute profiles where advection is relatively unimportant (Berner, 1971; 1980; Lerman, 1979). Both the cylindrical microenvironment model and the onedimensional Cartesian coordinate model will be used here to quantify the Mn2' distributions at NWC and DEEP.

DIAGENETIC PROCESSES. 11.

39 1

The reaction term R in Eq. (6.12) is determined as follows. Mn2+ is produced by the dissolution of solid-phase Mn oxide and is subject to reprecipitation as either an oxide or reduced phase. Because oxide reduction begins very close to the sediment-water interface, I assume that little reprecipitation as an oxide actually takes place within the deposit or that reprecipitation takes place so near to the interface that it cannot be differentiated from a boundary condition. Therefore, the Mn2+distribution can be considered as influenced dominantely by production and anoxic precipitation reactions over most of the sampled interval. The production term was shown in the previous section to be of the form R = R o exp( -ax) where R o and a are constants and x is the depth in the deposit. Precipitation reactions are commonly assumed to follow firstorder or pseudo-first-order kinetics such that R = k l ( C - Ceq)where k , is a first-order rate constant and C,, represents a depth-dependent equilibrium concentration (Holdren et al., 1975;Robbins and Callender, 1975). In LIS sediments the concentrations of many anions such as HCO; ,which might precipitate with Mn”, are roughly constant over the top -20 cm of sediment. This is true in particular at NWC and DEEP (Part I). It will therefore be assumed that C , , is constant over the depth interval of interest and that its value is the concentration to which a profile asymptotes at depth. Taken together these considerations suggest that an appropriate reaction term for Mn2+in the present case is R

=

Roexp(-aLu)

- kl(C -

Ceq)

Equation (6.12) will be solved only for the steady-state case (dClat = 0) because the dimensions of the effective cylinder microenvironment are such that steady-state solute distributions are quickly achieved (Part 1). The boundary conditions on the cylinder are taken as (6.13a)

x=O,

r=rl,

(6.13b)

r =

(6.13~)

x = L,

r2,

C=CT

aClar = 0 aclax

=

o

Condition (6.13a) specifies that the solute concentrations along the sediment-water interface and within the ideal burrow are equal and constant. The second condition, (6.13b), requires that concentrations go through a maximum or minimum halfway between individual microenvironments. Condition (6.13~)matches the bioturbated zone with the underlying unburrowed zone by requiring a continuity of flux between the two regions. In this case, the gradient is taken as -0 because of the requirement that

392

ROBERT C. ALLER

C asymptote to a constant C,, and because the gradients observed in cores are approximately zero at the base of the zone of interest. The solution to Eq. (6.12) with conditions (6.13a)-(6.13c) and R = R o exp( - a)- k , (C - C,,) is (by separation of variables)

with

The functions ZJz) and K,(z) are the modified Bessel functions of the first and second kind, respectively, of order v (see e.g., Abramowitz and Stegun, 1964, for values). The vertical concentration gradient measured in the sediment pore waters corresponds to the average concentration in the effective microenvironment over the finite depth interval x , - x 2 and is given by (6.15)

=

( Z . ~ i : . . . . ) / ( l . ~ ~ ~ r ~ r ~ ~ )

Values for r l , r2, and L in the present case are taken as those found by modeling NH; distributions in the same cores analyzed for Mn2+(Part I). It is assumed that NH; is not subject to precipitation reactions and can be used to obtain the effective geometry of diffusion. Once obtained, by using a relatively conservative element this geometry can be used to model Mn” distributions and estimates of reaction rates can be made. To illustrate the behavior of the cylinder model and also to demonstrate how irrigated burrows can be expected to influence Mn2+profiles, a representative vertical profile predicted by Eqs. (6.14) and (6.15) for Mn” has been plotted for the case k , = 0 (Fig. 19). The production rate for Mn”, r l , r 2 , and L are those for core NWC-4 based on the solid-phase dissolution rate of Section 6.4.1 (Table V; divided by average porosity ~0.750)and the cylinder-model values of Table V in Part I. The value of D is estimated from the molecular diffusion coefficient at infinite dilution T = 19°C (Li and Gregory, 1974), multiplied by a correction factor for sediment structure of 0.56. This factor was approximated by cp2 with an average porosity of cp = 0.75 (Manheim, 1970; Manheim and Waterman, 1974; Krom and Berner, 1980). For comparison, the one-dimensional

DIAGENETIC PROCESSES. 11.

393

0

2 4

c

,k'

10

t

Q

l2

n

14

a,

16

no p r e c i p i t o t l o n

18

FIG.19. Comparison of the one- and two-dimensional models for Mn2' distribution in the top 0-18 cm of sediment at NWC. The production rate in both cases is that found for core NWC-4. The anoxic precipitation rate is assumed to be zero. The effective cylinder geometry used in the two-dimensional model is that determined for NH; in Part I; r l = 0.14 cm, r 2 = 4.5 cm. The basal gradient is constrained to be zero. The diffusion geometry created by irrigated burrows results in a vertical pore-water solute profile exhibiting apparent precipitation.

case (rl + 0; r2 --* m) is plotted for the same upper and lower boundary conditions. The cylinder model demonstrates that the presence of irrigated burrows can produce a decrease of Mn2+concentration with depth that would commonly be interpreted as evidence for precipitation when, in fact, no such precipitation is occurring. In order to estimate precipitation rate constants k l for each core, the production rates determined in Section 6.4.1 from solid-phase distributions and the model cylinder geometries determined from NH: distributions were first used to fix values of R0,ct, r l , rz, and L. Estimates of D in each case were made by modifying the infinite dilution value at the appropriate core collection temperature (Li and Gregory, 1974) by the cpz correction factor 0.56 as outlined previously. The average porosity at both NWC and DEEP is -0.75 in the modeled zone, assuming a sediment particle density of 2.5 gm/cm3. CTis -0.2 pit4 based on flux box cores (Appendix C, Part I). C,, was estimated from the asymptotic concentration in each core and is -25-27 pA4 at NWC and -125 at DEEP. These values and the corresponding values of k , determined by profile fitting for each core are listed in Table VI. Plots of model profiles are given in Figs. 20 and 21. Summer and fall profiles at NWC are fit quite well by the model. The winter profile is not; this may reflect an increase in Mn" oxidation rate

TABLEVI. PORE-WATER Mn2+ MODELVALUES'

Cylinder Model

Sample

TCC)

Inner radius r , (cm)

NWC-2 NWC-3 (dash line) (dot line) NWC-4 DEEP- 1

13.2" 3"

0.14 0.14

19" 18.5"

0.14 0.2

Outer radius r 2 (cm) 5.9 14 4.5 4

D RO a L (cm2/day) (pmoles/cm3/day) (/cm) (cm) 0.244 0.169 0.287 0.283

0.631 0.389 0.825 0.114

1.2 1.08 1.35 1.5

18 16 18 16

Anoxic precipitation Estimated flux CT C,, const. k , acrossx = 0 (pM) (pM) (/day) (mmoles/mz/day) 0.2 0.2 0.2 0.2

One dimension

Sample N WC-2 NWC-3 NWC-4 DEEP-I a

See Section 6.4.2.

Anoxic precipitation Estimated flux across x = 0 const. k l (/day) (mmoles/m*/day) 0.95 0.85 0.67 1

2.1 1.3 3.O 1

25

0.57

1.7

27 50

0.6 0.7 0.45

0.99 0.82 2.4 1.2

25 125

0.5

395

DIAGENETIC PROCESSES. 11.

MnZ+ b M 1

MnZ' (uM) 100

18

200

100

300

200

300

h!

FIG.20. Pore-water Mn" profiles predicted by the two-dimensional cylinder model (dash or dot bars) and one-dimensional model (continuous profile line) compared with the actual measured profile in cores NWC-2 and NWC-3 (solid bars). Model values are given in Table VI.

near the interface or is due to an overestimate of Mn2+ production rate because 234Thmixing rates can integrate over a several-month period. The profile at DEEP is relatively insensitive to most model values and largely reflects the choice of Ceq.The subsurface maximum in Mn2+not predicted by the production-rate function in this case demonstrates that production of Mn2+ is underestimated at depth. All profiles predict values of k , O.5-0.7/day or 3.5-49 x 10-4/min. Except for the winter core at NWC (NWC-3), the sediment-water Mn" fluxes predicted by the model pro-

-

Mnz+(uM) 100

200

300

Mn"(uM1 400 I

6

12 14 16

FIG.21, Pore-water Mn2' profiles predicted by the two-dimensional cylinder model (dash or dot bars) and one-dimensional model (continuous profile line) compared with the measured profile in cores NWC-4 and DEEP-1 (solid bars). Model values are given in Table VI.

396

ROBERT C. ALLER

files are in the range, -1-2 mmoles/m2/day, observed by direct measurement (Table 11). The magnitudes of k , determined from these model fits are lo2-lo3 times the highest values yet reported (Robbins and Callender, 1975; Elderfield, 1976). This is due predominantly to the higher estimates of R o used here than in previous work. As stated previously, the general agreement of the direct flux measurements with the estimated range of production rates indicates that the present model values are at least internally consistent with all available data. In order to determine how estimates of k l would change if the effect of irrigated burrows on diffusion geometry were ignored, Eq. (6.12) was also solved at steady state for the one-dimensional case, that is

-=

(6.16)

at

o = Da2c T+ ROe--M- kl(c- ceq) ax

The solution to Eq. (6.16) with the boundary conditions (6.13a)-(6.13~) is (6.17) C(x) = A cosh[A(x - L ) ] + Ceq - [Rd(azD- kdle-" with A =

(kl

lD)1'2j A = [(C,- Ceq

+ Ro)/(aZD- k~)]/cosh(-AL)

The resulting model profiles are plotted as the continuous curves in Figs. 20 and 21. All appropriate model parameters (CT, Ceq,Ro, a,L, and D ) are identical to those used for the cylinder calculations. Values of k I were varied to obtain the plotted curves. Estimates of k , ranged from 0.7-11 day and are generally higher than, but within a factor of, 1.5-2 of those determined by the cylinder model. The estimate of DEEP is relatively insensitive and k , ' s in the range 0.5-l/day give reasonable fits. Sediment-water fluxes estimated by the one-dimensional model are also in the range observed (Table VI). The use of the one-dimensional model allows estimates of reaction rates to within a factor of 1.5 of those predicted by the more complex twodimensional model. This close agreement is due largely to the rapid attenuation of MnZ+production rates with depth and suggests that in such cases the distribution of a nonconservative pore-water constituent can be modeled reasonably accurately by use of the transport-reaction equation in one dimension. With this in mind, a one-dimensional transport-reaction model will be used to obtain corresponding reaction rate constants for Fez+.These are very approximate because of the large uncertainty in Fez+ production rates, oxidation rates, and variation in sulfide production over the zone

-

DIAGENETIC PROCESSES. 11.

397

of interest (Goldhaber et al., 1977; Aller and Yingst, 1980). It is imagined in this case that the sediment column can be divided into three vertical zones of fixed thickness. In the upper zone near the sediment-water interface, Fe2+distribution is controlled by diffusive loss and precipitation as an oxide. The rate of oxidation depends on the 0,concentration and pH, which change rapidly as the sediment-water interface is approached (Stumm and Lee, 1961; Kester et a/., 1975). It will be assumed that an average rate constant can be applied in this zone to characterize precipitation. In zone 2, Fe2+is produced at an approximately constant rate as calculated in Section 6.4.1 and is also subject to precipitation as a sulfide (Part I). Fe-sulfide precipitation kinetics are complex (Rickard, 1974), but it will be assumed here that at constant pH and approximately constant pS the kinetics can be described as first order in Fe2+with a decreasing rate as an equilibrium value C,, is approached. Zone 3 is taken as dominated by the same net precipitation of Fe2+ as in Zone 2, but without further production. The equations describing these distributions are at steady state: a2C zonel, O = D ? - k z C , 05xSL1 ax a2C (6.18b) zone 2, 0 = D 7 + R o - k3(C - Ceq), L I 5 x 5 L Z ax a2C L2 5 x 5 w (6.18~) zone 3, 0 = D 7 - ka(C - Ceq), ax (6.18a)

where k 2 and k 3 are the oxic and anoxic reaction constants, respectively, R o is a constant FeZ+production rate, C,, is an equilibrium Fe2+ concentration presumably determined by sulfide equilibria, and all other symbols are as before. The boundary conditions are (a)

x = 0,

(b)

x = LIP

(el

x = L2,

(0

x--,

MI

c

= CT

Czone 1 = Czone 2

(ac1ax)zone 2

=

(ac/dx)zone 3

C + Ceq

These require continuity of flux and concentration across zone boundaries as well as a constant boundary value at the surface and an asymptotic

398

ROBERT C. ALLER

concentration at depth. The solutions are (6.20)

zone 1, zone 2, zone 3,

C C C

+

= A 1 sinh(ulx) CTcosh(ulx) = A 2 sinh(u2x) B2 cosh(u2x) = A 3 expf - w ) C,,

+ +

+ Rdk, + C,,

where (TI

= (kz/D)'",

0 2

= (k3/D)1'2

F; = uIcosh(a2L cosh(ulLl) - u2sinh(o2L sinh(o1L A2 =

(Ce,

+ R/k3)u1cosh(all, I) - [R0/[k3exp(u2L2)1- UICT exp(cr2LI)[ulcosh(ulLl) + uz sinh(uILI)l

B 2 = -A2 - Ro/[k3exp(u2L2)l

A1 =

- A I exp(u2LI ) + [ 1 - cosh(u& I)lexp(u2L2)lRdk3 sinh(utLI) CT cosh(alli) + Ceq -sinh(ulL I)

A 3 = -A2

+ [exp(uzL2)- cosh(u~~~)lRdk3

The constants used in the model are as follows. The D terms were estimated in the same way as done for Mn by correcting the infinite dilution values of Li and Gregory (1974) to the core collection temperature and multiplying the corrected D in each case by the factor 0.56 corresponding to 'p2 (Lerman, 1978). Ro, L , , and L z were determined from solid-phase distributions as discussed in Section 6.4.1. A porosity of 0.75 was assumed in correcting R o to a rate relative to pore-water volume. k 2 does not significantly alter the shape of the model profiles because zone 1 is relatively thin. A value of Uday was used in all cases. This is 1/10 the magnitude of the water column precipitation constant kFe, estimated in the flux experiments (Section 5.4) and found to result in sediment-water fluxes in the range of those given in Table 1. CT was fixed at 0.02 p M [based on flux-core blank samples (Appendix C, Part I)] and C,, at 1.1 F M ,corresponding to the observed asymptotic concentration at depth in cores NWC-2, NWC-3, and NWCd. With all other variables fixed, k , was then varied to produce as close a fit as possible to the observed porewater profiles. Fez' profiles predicted by the model are plotted in Fig. 22. Corresponding model values are listed in Table VII. Estimates of the anoxic precipitation rate constant k , range from about 10-120/day and are the

399

DIAGENETIC PROCESSES. 11.

Fe2'bM) 10

20

Fez' (JIM)

Fe2+(NM) 10

30

20

2

10

30

20

30

40

2 -

4

-E

-

6

8

u 10 c c

12

0"

14

a

16

8 1 10

NWC-3 10

--

16

18

FIG.22. Pore-water Fez+ vertical'concentration profiles (continuous line) predicted from the three-zone, one-dimensional model compared with the measured profile (solid bars). Model values are listed in Table VII.

same magnitude as those reported for Fe" oxidation in well-aerated seawater (Kester et al., 1975; Murray and Gill, 1978). These k , values are 10-100 times the anoxic precipitation rate constants determined for Mn, a fact consistent with the higher production rates calculated for Fez+. The use of the square wave production rate distribution is an obvious approximation and a Gaussian distribution for the production rate, such as used for Mn" by Robbins and Callender (1975), would be more realistic. A smoothly increasing, then decreasing, Fe" production-rate distribution would result in better fits to the actual data at the boundaries of zone 2 and lower estimates of k,. k , would also be slightly smaller if the cylindrical coordinate model were used as previously illustrated for Mn2+.However, because of the very approximate nature of the production rates and the simplifications made concerning the kinetics of oxic and anoxic precipitation, more sophisticated modeling is not warranted. In summary, the pore-water Fez+ and Mn2+profiles can be used together with solid-phase distributions, measurements of sediment-water exchange, and knowledge of the biogenic diffusion geometry as obtained from modeling relatively conservative solutes to place constraints on reaction rates. If anoxic precipitation kinetics are assumed to be first order, then rate constants are calculated for both Mn2+and Fez+losses that are larger than any others reported in the literature of which I am aware. Both the shapes of the pore-water profiles, particularly at NWC, as well as the apparent rapid reaction rates indicate that most anoxic precipitation of Fez+and MnZ+takes place in the zone of solid-phase excess (e.g., Figs. 2-7). Precipitation rates and production rates could therefore easily be minima.

TABLEVII. PORE-WATER Fe" MODELVALUES"

Sample

T ("C)

D (cm2/day)

LI (cm)

L2 (cm)

NWC-2 NWC-3 NWC-4

13.2" 3" 19"

0.256 0.184

1 2 1

4 5 4

a

See Section 6.4.2.

0.297

R (mmoles/cm3/day) 1.1 1.5

0.31

CT

C., (pM)

0.02 0.02 0.02

1.1 1.1 1.1

Oxic precipitation k2 (/day) 1 1

1

Anoxic precipitation k , (/day)

Estimated flux acrossx = 0 (mmoles/m2/day)

1 20

0.01

80

0.001 0.06

9.5

40 1

DIAGENETIC PROCESSES. 11.

6.5. Flux of Mn2+and Fez+ into Overlying Water

The flux of Mn” into the overlying water varies seasonally in a generally regular fashion and is a strong function of temperature (Fig. 23, Table I). An exponential function of the Arrhenius type used in Part I to describe the temperature dependence of NH: fluxes can also be used in this case to describe the temperature dependence of Mn2+ fluxes. If J = J’ exp( -E/gT), where J is the flux, J’ the constant preexponential factor, T the absolute temperature, E the apparent activation energy, and g the gas constant, then the apparent activation energies and preexponential factors are FOAM:

E = 18.2 kcallmole;

J’

NWC:

E = 45.9 kcaYmole;

J’ = exp(80.6)

DEEP

E = 40.7 kcal/more;

J’

= exp(30.4)

= exp(69.7)

The average yearly Mn” fluxes calculated from these activation energies and assuming T = 285 - 10 cos(2vt) [t is time (yr)], are 0.23, 2.6, and 0.36 mmoles/m2/day at FOAM, NWC, and DEEP, respectively. Values of E calculated from NH$ fluxes were shown to agree well with those found for the temperature dependence of microbial metabolic activity of -19 kcal/mole (Part I). In contrast, the apparent activation energies of Mn2+ fluxes can be much higher even though Mn2+production should also be controlled directly or indirectly by metabolic rates. One of the likely reasons for this difference is that Mn2’ release from the sediment is influenced by abiogenic precipitation reactions in addition to production. The changing relative importance of oxic or anoxic consumption reactions and Mn” production rates from season to season can be expected to produce a temperature dependence of sediment-water

T (“C)

FIG.23. Directly measured Mn2+flux out of bottom sediment at each station as a function of temperature (season).

402

ROBERT C. ALLER

Mn2+ flux different from that controlled by production alone. For example, during winter the surface sediment is more highly oxygenated than at other times of year because of lowered microbial activity. As a result, Mn2+oxidation rates near the interface should be relatively high at that time compared to production rates. This change in the relative rates of consumption-production reactions would result in a greater decrease of the flux from the sediment than could be explained by the temperature dependence of production. Consumption of Mn" near the sediment-water interface can be demonstrated in part by discrepancies between flux predictions made using Fick's first law and actual measured fluxes. Minimum flux estimates expected on the basis of concentration gradients alone can be made by assigning the MnZ+concentration in the top 0-1 cm of each core to a depth of 0.5 cm and calculating a linear concentration gradient between that depth and overlying water (x = 0, Mn2+ 0.2 pM).This gradient is a minimum, assuming no precipitation reactions, because in some cores the maximum Mn2+concentration must occur closer to the interface than 0.5 cm (e.g., all summer cores), whereas in others, not accounting for the finite sampling interval minimizes the gradient. Diffusion coefficients were estimated as before by correcting the infinite dilution values of Li and Gregory (1974) to the flux-core temperatures giving 6.38, 5.32, and 3.65 x cm2/secfor 22", 15", and 4"C, respectively. These were then multiplied by the factor cp2 to approximately correct for sediment structure (Lerman, 1978; Krom and Berner, 1980); cp2 = 0.44 at FOAM and 0.56 at NWC and DEEP as approximated from average water contents (see Part I, Section 6.9). Ficks first law gives the diffusive flux in sediments as (Berner, 1980)

-

(6.21)

J,=o = -cpI)(dC/dx),=o

The minimum fluxes calculated from (6.21), assuming cp = 1 at x = 0, and values of I) and (dcldx) determined as described above are listed in Table I along with the directly measured fluxes. At FOAM the minimum flux predicted on the basis of concentration gradient is always larger than the measured flux. Similar overestimates occur for winter cores at both NWC and DEEP and the fall core at DEEP. These discrepancies are consistent with a partial loss of Mn2+from solution in the top centimeter or at the sediment-water interface in each case. It should be noted that although the calculated flux estimates are approximate and either overestimate or underestimate the flux for the reasonsjust discussed, they do predict the measured flux to within a factor of 2-6 in all cases.

DlAGENETlC PROCESSES. 11.

403

Because of physical disturbance the solid-phase Mn profile at FOAM is irregular and a production rate could not be estimated from it (Section 6.4.1). An order of magnitude production rate can be inferred from the flux data as follows. There is overall agreement in the magnitude of estimated production rate, measured fluxes, and estimated fluxes at NWC and DEEP. Mn” flux at FOAM is lower, but similar in range to NWC and DEEP. In addition, measured fluxes at FOAM are within a factor of 3 of those predicted from pore-water gradients indicating that precipitation losses are not unusually high compared to the other stations. Taken together these facts indicate that Mn2+ production rates at FOAM are in the same range as at NWC and DEEP. In addition to seasonal differences in flux due to temperature-controlled reactions and transport, the magnitude of the flux changes with depositional environment. Fluxes in this case are highest at NWC and lowest at FOAM. Subsequent measurement of Mn2+fluxes made during summer 1977 at a variety of stations in LIS has shown that the magnitude is directly correlated with the abundance of organic matter or fine-grained sediment in the top few centimeters (Aller, 1979). The Mn2+flux can also be increased by biogenic reworking as demonstrated in laboratory experiments (Aller, 1978). This increase presumably comes about because of sediment irrigation as well as the continual transport of oxide-rich particles into otherwise reduced regions of a deposit. The presence or absence of deposit-feeding organisms is also correlated with grain size and depositional environment (e.g., Sanders, 1956). Mobile deposit feeders such as protobranch bivalves are most abundant at NWC and may be partly responsible for relatively high Mn” release rates there compared with other stations. Because of the rapid attenuation of production rates away from the interface, the construction of burrows per se does not greatly influence the Mn2+flux in a steady-state system. At DEEP, the cylindrical model predicts that as much as -40% of the flux could come from burrows, whereas at NWC only 3-7% has a radial source. The use of average concentrations within sampling intervals to calculate gradients will lower the effective radial component still further. This relatively small effect, given measurement uncertainties, illustrates why the flux estimates made previously using the molecular diffusion coefficients are reasonably close to that actually observed. The contribution of radial diffusion to the flux could change if reaction rates are different around burrows than in surrounding sediment (Aller and Yingst, 1978) or if the rate is not rapidly attenuated with depth. The magnitude of the Mn2+fluxes from different areas of LIS are in general agreement, although usually somewhat higher than those reported

404

ROBERT C. ALLER

from other near-shore regions. Graham et al. (1976) report Mn2+fluxes from Narragansett Bay, Rhode Island also taken in June-July, 1975. They obtained an average for this time of about 0.4 mmoles/m2/day, which is lower than that measured at any LIS station. No information was given on the depositional environment so direct comparison with stations in the present study is not possible; nevertheless, the values obtained are not dissimilar to those found at FOAM. Benthic Mn2' fluxes reported from the Chesapeake Bay range from 0.3-7 mmoles/m2/dayin summer (July, 1977) to 0.01-0.02 mmoles/m2/day in spring (April, 1978) (Eaton, 1979). The high value of 7 mmoles/m/day was taken in a region of anoxic bottom water and is not directly comparable to measures made in LIS. The lower values are similar in range to those found in this study, but again no description of environment was given. Elderfield (1976) estimated a range of fluxes from near-shore sediment of 0.003-0.3 mmoles/m2/daybased on pore-water profile data of Calvert and Price (1972). He implies that this flux is to surface sediment and that it does not represent Mn*+,which actually escapes into overlying water. If it is assumed that the Mn2+does escape, these estimates are generally lower than the yearly average fluxes calculated in this study. Because of the large uncertainty in Fe2+flux measurements, no spatial patterns in magnitude are discernable. Any regular seasonal pattern is also obscured, except that Fe2+fluxes are definitely lowest in winter. The measured fluxes listed in Table I have been corrected for precipitation loss in the core boxes and in some cases are 10-100 times higher than that predicted by the actual uncorrected measured concentration increase in the flux core water (Section 5.4). Oxidation is therefore so rapid that the release might best be conceptualized as a colloid flux. The expected flux of Fe" from the sediment if no precipitation at the interface takes place can be estimated from the pore-water concentration gradients in a similar way as done for Mn2+.A linear concentration gradient was estimated for each core by assigning the measured Fe2' concentrations at 0- 1 and 1-2 cm to the midsample depth and calculating the corresponding gradient. The overlying water Fe2+concentration was assumed to be zero. Values of DFez+were determined in the same way as described for Mn2+using the same sediment correction factors (cp'). The infinite dilution values are: 6.59, 5.56, and 3.94 x cm2/sec at 22", 15", and 4"C, respectively (Li and Gregory, 1974). The calculated fluxes are listed in Table I. Fluxes calculated from the linear gradient are within a factor of 2 of corrected measured fluxes at 22°C. Gradient calculations predict lower fluxes at 15°C and higher at 4°C than the corrected measured fluxes. Approximate agreement during warm periods and overestimates during cold

DIAGENETIC PROCESSES. 11.

405

periods are consistent with relatively low and high oxygenation of surface sediments at those times and correspondingly low and high precipitation rates near the interface. The discrepancies of 3-100 times between the two estimates during the fall ( W C ) period are not readily explained except by measurement errors and assumptions, or by requiring production of Fe2+along the sediment-water interface and burrow walls as a result of biogenic reworking and possibly oxidation of sulfides. Uncertainties in the measurement of Fe2+fluxes can be reduced in future studies by determining the flux of both particulate and dissolved Fe from the bottom sediments and by more stringent monitoring of precipitation rates within the flux-core boxes than done in this study. The release of Mn" and Fe2+into the water column, their subsequent oxidation, and the constant resuspension of particles coated with freshly precipitated Fe, Mn oxides as a result of diagenetic remobilization should play a major role in controlling the water chemistry of the estuary (Aston and Chester, 1973; Turekian, 1977). Continual scavenging of Fe- and Mnoxide-seeking elements, such as Zn, Cu, Co, and Ra, by coprecipitation or adsorption on colloids and larger sedimentary particles must take place. This sweeping-out process will maintain low standing concentrations of many metals even in the presence of significant natural or anthropogenic inputs into the basin (Broecker et al., 1973; Aller et al., 1980). Benthic fluxes of Mn2+and reprecipitation are known to influence the spatial distribution of solid-phase Mn within near-shore basins (Evans et al., 1977; Grill, 1978). These fluxes may also influence the distribution of Mn within the ocean basins as a whole. For example, Turekian (1977) calculated that if the annual Mn2+flux found in LIS (-1 mmole/m2/yr) were assumed to apply to -1% of the sea floor and if a few percent of this annual flux were lost to the deep sea, then the excess Mn accumulation rate of -1 pg/cm2/yr(Bender et al., 1970; Elderfield, 1976) in deepsea sediments could be accounted for. The salient conclusion is not necessarily that near-shore sediment-water Mn2+ fluxes cause excess Mn accumulation in the deep sea but that these fluxes can be high enough to affect large-scale distributions. Because of the lithologic abundance of Fe, it is not possible to readily demonstrate the effect of diagenetically released dissolved and colloidal Fe on sea-floor distributions. However, remobilization can be reflected in Fe contents of suspended sediment (Murray and Gill, 1978). Dissolved and colloidal Fe release from sediments may also be important in explaining seaward migrations or isotopic exchange of Fe from one basin to another as suggested by the distribution of "Fe inventories (Labeyrie et al., 1976). Fallout 55Fein turn may be a useful tool for quantifying such large-scale Fe mobilization.

406

ROBERT C. ALLER

7. SUMMARY (1) Pore-water profiles of Fez+ and MnZ+from three stations in Long Island Sound have general depth-dependent concentration distributions similar to those reported from other sedimentary basins: concentrations rise above seawater values to a maximum below the interface and then decrease again or remain constant deeper in the deposit. Beyond these general features, specific features of the profiles reflect the internal transport-reaction regime effective at each station. (2) The production of Mn2+ in pore waters is directly related to the rate of reduction of Mn oxides during the decomposition of organic matter, both as a function of depth in the sediment as well as seasonally. Fez+, on the other hand, is produced both by the reduction of Fe oxides and by abiogenic or biogenic oxidation of Fe sulfides. The result of different sources for the two dissolved metals is a different seasonality of the interstitial water profiles near the sediment-water interface. The temporal changes in both Mn2+and Fez+profiles are repeatable from year to year. During the summer, pore-water MnZ+in the top few centimeters reaches the highest concentration of the year. In the fall, MnZ+concentrations are lowered in magnitude throughout the sediment column as a result of both decreased production and a relative increase in the effect of biogenic transport processes that exchange sediment solutes with overlying waters. During the winter, Mn2+ profiles reflect lowered rates of production. Concentrations near the sediment-water interface decrease because of the increased dominance of precipitation reactions and diffusive loss to overlying water compared with production rates. Overall MnZ+production mimics the seasonal and, to a lesser extent, the depth-dependent production patterns of metabolites such as NH: (Part I). Fez+ is also formed in abundance near the sediment-water interface during the early summer when the overlying water first begins to warm. Because of the associated increase in sulfide production, Fez+ does not always rise to its maximum yearly concentration at this time. In the fall, like many other ions whose source is in the sediment, Fez+ concentrations below the top few centimeters drop to their lowest value of the year. Unlike Mn", during the winter Fez+ may reach its maximum standing concentration of the year at some stations. This presumably results from a net oxidation and loss of FeS and FeSzfrom surface sediment and an associated release of Fez+ that does not immediately depend on strongly temperature-controlled microbial metabolism. (3) Both Mn2+ and Fe2+ react rapidly with anions formed during decomposition. Equilibrium calculations and direct measurement of solidphase sulfides indicate that away from oxidized portions of sediment, Fez+

407

DIAGENETIC PROCESSES. 11.

concentration is determined in part by the formation of FeS-FeS,, whereas Mn2+ concentrations appear to be related in some cases to the formation of MnC03. Other reduced phases may also influence Fe+ and Mn+ solubilities. (4) Although Mn” and Fez+ may maintain a set mass action relation with a given solid phase, the overall magnitudes of their concentrations reflect transport processes within the sediment. Biogenic transport is particularly important in this regard because the buildup at depth of anions that can precipitate Mn2+or Fez+ is determined to a large extent by burrow construction and irrigation (Part I). If, for example, it is assumed that Mn2+ concentrations are controlled by MnCO, solubility, then the increased Mn” below 10 cm at deep water relative to inshore stations could result from the decreased alkalinity buildup associated with increased biogenic transport offshore (Part I). ( 5 ) The solid-phase Mn and Fe profiles, together with estimates of biogenic reworking rates obtained from 234Thdistributions, allow calculation of solid-phase dissolution rates or equivalent Mn2+and Fez+production rates in the top 5 cm at two stations. The rates obtained indicate that the turnover time of excess Mn above the average sediment background is -60- 100 days. Excess Fe turns over on a similar time scale of 100-200 days. Excess Mn dissolution rates decrease exponentially beneath the sediment-water interface and can be described by functions of the form R = R o exp( - ax)where R o 5-30 pg/cm3/day,a 1-1 S/cm in magnitude, and x is depth in the sediment. Fe dissolution was estimated only as an average rate over selected depth intervals from 1-4 or 2-5 cm beneath the interface. A range of Fe dissolution of -10-60 pg/cm3/day was calculated. (6) Pore-water profiles were used together with solid-phase dissolution rates in diagenetic models to determine first-order anoxic precipitation rate constants for both Mn and Fe. A two-dimensional cylindrical coordinate model was employed to account for the effects of biogenic irrigation of burrows on pore-water MnZ+distributions. Two-dimensional diffusion can result in a decrease in Mn” with depth that would be interpreted as evidence for precipitation and cause overestimation of precipitation rates in a one-dimensional model. Anoxic precipitation rate constants of -0.5-0.7/day are estimated from the cylindrical coordinate model. A traditional one-dimensional model predicts rates 1.5-2 times higher. Agreement is relatively close between the two models in this case because Mnz+ production rates attenuate rapidly with depth, Fez’ precipitation rates of about lO-l20/day are estimated from a three-layer one-dimensional model. (7) The rapid oxic and anoxic precipitation of both Mn” and Fez+

-

-

-

-

408

ROBERT C. ALLER

apparently results in substantial reprecipitation within the zone of excess solid-phase Mn and Fe. This means that the estimated production rates are probably minima although agreement between production rates, the estimated sediment-water fluxes of MnZ+and Fez+that they could support, and the actual measured benthic fluxes are evidence that the production rates are approximately correct. (8) Both MnZ+and Fez+escape into overlying water. Mn" flux is highest in the summer and lowest in the winter. The seasonal range encompassing values at all stations is 0.01-4 mmoles/m2/day and the yearly average flux range is 0.23-2.6 mmoleslmz/dayin the central Sound. MnZ+ fluxes can be predicted to within a factor of 2-6 by use of Fick's first law of diffusion and assuming a linear concentration gradient from the top 0-1 cm of sediment to overlying water. Diffusion coefficients were estimated from the infinite dilution value multiplied by the factor Q' (Manheim, 1970; Lerman, 1978). This general agreement indicates that precipitation loss lowers the flux of Mn2+from sediment, but not radically. The high reactivity of Fez' under oxygenated conditions caused major problems in measurement of its flux out of the bottom. Because of this it is not possible to say with certainty what the magnitude or seasonality of the fluxes are, although the lowest fluxes definitely occur in winter. Measured fluxes corrected for precipitation loss range from about 0.001 to 0.5 mmole/m2/day. (9) The flux of Mn2+ from near-shore sediments is sufficiently high to influence both small- and large-scale distribution of Mn in the ocean basins, The lithologic Fe background prevents easy recognition of Fe deposition patterns resulting from diagenetic remobilization, but it is likely that an absolute quantity of Fe similar to or greater than that found for Mn is mobile. (10) In both Part I and I1 it has been shown that consistent explanation and quantitative modeling of diagenesis near the sediment-water interface can be made if a wide range of biological, chemical, and geological data are available for a given area. The importance of interrelationships between benthic communities, physical depositional environment, and chemical properties of the elements under consideration in controlling diagenesis and sediment-water exchange has been documented and emphasized. ACKNOWLEDGMENTS This article is based on a portion of a Ph.D. dissertation done in the Department of Geology and Geophysics, Yale University. Modifications and, I hope, improvements were made at the University of Chicago while I was supported by an Alfred P. Sloan Foundation Fellowship. M. B. Goldhaber and J . Y. Yingst deserve special thanks as valued co-workers during

DIAGENETIC PROCESSES. 11.

409

critical stages of this project. I also wish to thank L. K. Benninger, J. K. Cochran, G. R. Holdren, and J. K.Rosenfeld for much help and discussion. My major diving partners over the years were J. Y. Yingst, W. J. Ullman, and M.Pimer. R. Wells was an indispensable aid in the field and laboratory. M. Pimer and M. Reed captained the boats used in sampling. K. K. Turekian, D. C. Rhoads, R. A. Berner (Yale), and V. Barcilon (Chicago) provided critical comments, guidance, and advice at various stages. Thanks to G. R. Holdren and an unidentified reader for critical review of the manuscript. J. Pasdeloup deciphered and typed the final manuscript copy. Research support was predominantly by ERDA grant EY-764-02-3573 (K. K.Turekian, principal investigator) and by NSF grant GA-42-838 (D. C. Rhoads, principal investigator). Personal support was additionally provided by a NSF fellowship, a Yale Graduate Fellowship, ERDA grant EY-764-02-3573 and EPA grant R804-909-010 (D. C. Rhoads, principal investigator).

LISTOF SYMBOLS Constant Attenuation constant Constant Concentration Pore-water solute concentration at x, r, and t Concentration in solid-phase (masshnit volume total sediment) Excess Mn or Fe concentration in solid-phase above background (masslunit volume total sediment) Average pore-water solute concentration over defined interval Asymptotic concentration of pore-water solute at depth in the sediment Molecular diffusion coefficient modified for porosity and turtuosity (charge coupling and ion pairing assumed neglible) Biogenic particle-mixing coefficient Activation energy Constant Gas constant; 1.99 cal/deg/mole Solute flux at depth x Preexponential factor for temperature dependence of J, First-order solid-phase dissolution rate constant First-order anoxic precipitation rate constant for solute Mn2+ in sediment First-order oxic precipitation rate constant for solute Fez+ in sediment First-order anoxic precipitation rate constant for solute Fe2+ in sediment First-order oxic precipitation rate constant for solute Fez+in water overlying sediment First-order oxic precipitation rate constant for solute Mn2+ in water overlying sediment Thickness of modeled sediment interval, usually the intensively bioturbated zone Integer summation variable Porosity General reaction term Reaction constant Radial distance from central axis of hollow cylinder (burrow) Temperature Time

410 7

w X

ROBERT C. ALLER

Excess Mn or Fe characteristic turnover times in sediment Sedimentation rate, thicknessltime Vertical depth in sediment measured positively from sediment-water interface.

REFERENCES Abramowitz, M., and Stegun, I. A. (1964). “Handbook of Mathematical Functions with Formulas, Graphs, and Mathematical Tables.” Dover, New York. Aller, R. C. (1977). The influence of macrobenthos on chemical diagenesis of marine sediments. Ph.D. dissertation, Yale University, New Haven, Connecticut. Aller, R. C. (1978). Experimental studies of changes produced by deposit feeders on pore water, sediment, and overlying water chemistry. Am. J . Sci. 278, 1185-1234. Aller, R. C. (1979). Spatial and temporal patterns of Mn++ and NH: fluxes from bottom sediments of Long Island Sound. Am. SOC.Limnol. Oceanogr., Absrr. Winter Meet., lst, Corpus Christi, Texas, Jan. p. 1. Aller, R. C. (1980). Relationships of tube-dwelling benthos with sediment and overlying water chemistry. I n “Marine Benthic Dynamics” (K. R. Tenore and B. C. Coull, eds.), pp. 285-308. Univ. South Carolina Press, Columbia. disequilibrium in nearshore sediment: Aller, R. C., and Cochran, J. K. (1976). 234Th/23*U Particle reworking and diagenetic time scales. Earth Plane?. Sci. Lett. 20, 37-50. Aller, R. C., and Yingst, J. Y. (1978). Biogeochemistry of tube-dwellings: A study of the sedentary polychaete Amphitrite ornata (Leidy). J . Mar. Res. 36, 201-254. Aller, R. C., and Yingst, J. Y. (1980). Relationships between microbial distributions and the anaerobic decomposition of organic matter in surface sediments of Long Island Sound, U.S.A. Mar. Biol. 56, 29-42. Aller, R. C., Benninger, L. K., and Cochran, J. K. (1980). Tracking particle associated processes in nearshore environments by use of 234ThIL38U disequilibrium. Earth Planet. Sci. Lett. 47, 161-175. Anikouchine, W. A. (1967). Dissolved chemical substances in compacting marine sediments. J . Geophys. Res. 72, 505-509. Aston, S. R., and Chester, R. (1973). The influence of suspended particles on the precipitation of iron in natural waters. Esfuarine Coastal Mar. Sci. 1, 225-231. Bender, M. L. (1971). Does upward diffusion supply the excess manganese in pelagic sediments? J . Geophys. Rev. 76,4212-4215. Bender, M . L., Ku,T.-L., and Broecker, W.S. (1970). Accumulation rates of manganese in pelagic sediments and nodules. Earth Planet. Sci. Lett. 8, 143-148. Benninger, L. K., Aller, R. C.,Cochran, J. K., and Turekian, K. K. (1979). Effects of biological sediment mixing on the ’“Pb chronology and trace metal distribution in a Long Island Sound sediment core. Earth Planet. Sci. Left. 43, 241-259. Berner, R. A. (1963). Electrode studies of hydrogen sulfide in marine sediments. Geochim. Cosmochim. Acta 27, 563-575. Berner, R. A. (1967). Thermodynamic stability of sedimentary iron sulfides. Am. J . Sci. 265, 773-785.

Berner, R. A. (1970). Sedimentary pyrite formation. Am. J. Sci. 268, 1-23. Berner, R. A. (1971). “Principles of Chemical Sedimentology.” McGraw-Hill, New York. Berner, R. A. (1976). Inclusion of adsorption in the modeling of early diagenesis. Earth Planet. Sci. Lett. 29, 333-340. Berner, R.A. (1980). “Early Diagenesis: A Theoretical Approach,” Princeton Univ. Press, Princeton, New Jersey, 241 pp.

DIAGENETIC PROCESSES. 11.

41 1

Berner, R. A., Scott, M. R., and Thomlinson, C. (1970). Carbonate alkalinity in the pore waters of anoxic marine sediments. Limnol. Oceanogr. 15, 544-549. Berner, R. A,, Westrich, J. T., Graver, R., Smith, J., and Martens, C. S.(1978). Inhibition of aragonite precipitation from supersaturated sea water: A laboratory and field study. Am. J . Sci. 218, 816-837. B&trCmieux,R. (1951). Etude expbrimentale de I’bvolution du fer et du manganese dans les sols. Ann. Agron. (Paris) 2, 193-295. Bloomfield, C. (1972). The oxidation of iron sulphides in soils in relation to the formation of acid sulphate soils and of ochre deposits in field drains. J . Soil Sci. 23, 1-16. Bonatti, E., Fisher, D. E., Joensuu, O., and Rydell, H. S. (1971). Postdepositional mobility of some transition elements, phosphorous, uranium and thorium in deep sea sediments. Geochim. Cosmochim. Acta 35, 189-201. Bostrom, K. (1967). Some pH-controlling redox reactions in natural waters. Am. Chem. SOC. Adv. Chem. Ser. 67,286-311. Bricker, 0 . P. (1965). Some stability relations in the system Mn-02-H20at 25” and one atmosphere total pressure. Am. Miner. 50, 12%-1354. Bricker, 0. P., and Troup, B. N. (1975). Sediment-water exchange in Chesapeake Bay. I n “Estuarine Research, 1, Chemistry, Biology and the Estuarine System” (L. E. Cronin, ed.), pp. 3-27. Academic Press, New York. Broecker, W. S., Kaufman, A., and Trier, R. M. (1973). The residence time of thorium in surface sea water and its implications regarding the rate of reactive pollutants. Earth Planet. Sci. Lett. 20, 35-44. Calvert, S. E., and Price, N. B. (1970). Composition of manganese nodules and manganese carbonates from Loch Fyne, Scotland. Contrib. Mineral. Petrol. 29, 215-233. Calvert, S. E., and Price, N. B. (1972). Diffusion and reaction profiles of dissolved manganese in the pore waters of marine sediments. Earth Planet. Sci. Lett. 16, 245-249. Calvert, S. E., and Price, N. B. (1977). Shallow water, continental margin and lacustrine nodules: Distribution and geochemistry. I n “Marine Manganese Deposits” (G. P. Glasby, ed.), pp. 45-86. Elsevier, Amsterdam. Crerar, D. A., and Barnes, H. L. (1974). Deposition of deep-sea manganese nodules. Geochim. Cosmochim. Acta 38, 279-300. Eaton, A. (1979). The impact of anoxia on Mn fluxes in the Chesapeake Bay. Geochim. Cosmochim. Acta 43,429-432. Elderfield, H. (1976). Manganese fluxes to the oceans. Mar. Chem. 4, 103-132. Evans, D. W., Cutshall, N. H., Cross, F. A., and Wolfe, D. A. (1977). Manganese cycline in the Newport River estuary, North Carolina. Estuarine Coastal Mar. Sci. 5, 71-80. Froelich, P. N., Klinkhammer, G. P., Bender, M. L., Luedtke, N. A., Heath, G. R., Cullen, D., Dauphin, P., Hammond, D., Hartman, B., and Maynard, V. (1979). Early oxidation of organic matter in pelagic sediments of the eastern equatorial Atlantic: Suboxic diagenesis. Geochim. Cosmochim. Acta 43, 1075-1091. Garrels, R. M., and Christ, C. L. (1965). “Solutions, Minerals and Equilibria.” Harper, New York. Goldberg, E. D. (1954). Marine geochemistry. I. Chemical scavengers of the sea. J . Geol. 62, 249-265.

Goldhaber, M. B., Aller, R. C., Cochran, J. K., Rosenfeld, J. K., Martens, C. S., and Berner, R. A. (1977). Sulfate reduction, diffusion and bioturbation in Long Island Sound sediments: report of the FOAM group. Am. J. Sci. 211, 193-237. Goto, K., Komatsu, T., and Furukawa, T. (1962). Rapid colorimetric determination of manganese in waters containing iron. Anal. Chim. Acta 21, 331-334.

412

ROBERT C. ALLER

Graham, W. F., Bender, M. L., and Klinkhammer, G. P. (1976). Manganese in Narragansett Bay. Limnol. Oceanogr. 21,555-673. Grill, E. V. (1978). The effect of sediment-water exchange on manganese deposition and nodule growth in Jervis Inlet, British Columbia. Geochim. Cosmochim. Acta 42, 485-494.

Halvorson, H. O., and Starkey, R. L. (1972). Studies on the transformations of iron in nature. Part 11. Concerning the importance of micro-organisms in the solution and precipitation of iron. Soil Sci. 24, ‘381-402. Harmsen, G. W. (1954). Observation on the formation and oxidation of pyrite in the soil. Plant Soil 5, 324-348. Harmsen, K., and Van Breemen, N. (1975a). Translocation of iron in acid sulfate soils: 11. Production and diffusion of dissolved ferrous iron. Soil Sci. SOC. Am. Proc. 39, 1148-1153.

Harmsen, K., and Van Breemen, N. (1975b). A model for the simultaneous production and diffusion of ferrous iron in submerged soils. Soil Sci. SOC.Am. Proc. 39,1063-1068. Hart, M. G. R. (1963). Observations on the source of acid in empoldered mangrove soils: 11. Oxidation of soil polysulfides. Plant Soil 19, 106-114. Henriksen, A. (1966). An automated modified formaldoxime method for determining low concentrations of manganese in water containing iron. Analyst 91, 647-651. Holdren, G. R., Jr. (1977). Distribution and behavior of manganese in the interstitial waters of Chesapeake Bay sediments during early diagenesis. Ph.D dissertation, The Johns Hopkins University, Baltimore, Maryland. Holdren, G. R., Jr., Bricker, 0. P., and Matisoff, G. (1975). A model for the control of dissolved manganese in the interstitial waters of Chesapeake Bay. In “Marine Chemistry in the Coastal Environment (ACS Symposium Series 18)” (T. M. Church, ed.), pp. 364-381. American Chemical Society, Washington, D.C. Jenne, E. A. (1968). Controls on Mn. Fe, Co, Ni, Cu, and Zn concentration in soils and water: Significant role of hydrous Mn and Fe oxides. Adv. Chem. Ser. 73, 337-387. Kalil, E. K., and Goldhaber, M. (1973). A sediment squeezer for removal of pore waters without air contact. J. Sediment. Petrol. 43, 553-557. Kester, D. R., Byme, R. H., Jr., and Liang, Y.-J. (1975). Redox reactions and solution complexes of iron in marine systems. ACS Symp. Ser. 18, 56-79. Krauskopf, K. B. (1956). Factors controlling the concentration of thirteen rare metals in sea water. Geochim. Cosmochim. Acta 9, 1-32. Krauskopf, K. B. (1957). Separation of manganese from iron in sedimentary processes. Geochim. Cosmochim. Acta 12, 61-84. Krom, M. D., and Bemer, R. A. (1980). The experimental determination of the diffusion coefficients of sulfate, ammonium, and phosphate in anoxic marine sediments. Limnol. Oceanogr. 25,327-337. Krom, M. D., and Sholkovitz, E. R. (1978). On the association of iron and manganese with organic matter in anoxic marine pore waters. Geochim. Cosmochim. Acta 42,607-612. Labeyrie, L. D., Livingston, H. D., and Bowen, V. T. (1976). Comparison of the distributions in marine sediments of the fallout derived nuclides ”Fe and 239. ’%I:A new approach to the chemistry of environmental radionuclides. ERDNIAEA Znt. Symp. Transuranium Nuclides Environ., San Francisco, Nov. 1975 . Lerman, A. (1975). Maintenance of steady state in oceanic sediments. Am. J. Sci. 275, 609-635.

Lerman, A. (1978). Chemical exchange across sediment-water interface. Ann. Rev. Earth Planet. Sci. 6 , 281-303.

DIAGENETIC PROCESSES. 11.

413

Lerman, A. (1979). “Geochemical Processes: Water and Sediment Environments.” Wiley, New York. Lewin, J., and Chen, C. (1973). Changes in the concentration of soluble and particulate iron in sea water enclosed in containers. Limnol. Oceanogr. 18, 5904%. Lewis, D. M. (1976). The geochemistry of manganese, iron, uranium, lead-210, and major ions in the Susquehanna River. Ph.D. dissertation, Yale University, New Haven, Connecticut. Li, Y.-H., and Gregory, S. (1974). Diffusion of ions in sea water and deep-sea sediments. Geochim. Cosmochim. Acta 38, 703-714. Li, Y.-H., Bischoff, J. L., and Mathieu, G. (1%9). The migration of manganese in the Arctic Basin sediments. Earth Planet. Sci. Lett. 7 , 265-270. Lynn, D. C., and Bonatti, E. (1965). Mobility of manganese in diagenesis of deep-sea sediments. Mar. Geol. 3, 457-474. Manheim, F. T. (1%5). Manganese-iron accumulations in the shallow marine environment. Narragansett Mar. Lab. Univ. R.I. Occas. Publ. 3, 217-276. Manheim, F. T. (1970). The diffusion of ions in unconsolidated sediments. Earth Planet. Sci. Lett. 9, 307-309. Manheim, F. T., and Waterman, L. S. (1974). Diffusimetry (diffusion constant estimation) on sediment cores by resistivity probe. Initial Rep. Deep Sea Drilling Proj. 22,663-670. Mann, P. J. G., and Quastel, J. H. (1946). Manganese metabolism in soils. Nature (London) 158, 154-156.

Martens, C. S., and Berner, R. A. (1974). Methane production in the interstitial waters of sulfate-depleted marine sediments. Science 185, 1167-1 169. Martens, C. S . , Berner, R. A., and Rosenfeld, J. K. (1978). Interstitial water chemistry of anoxic Long Island Sound sediments. 2. Nutrient regeneration and phosphate removal. Limnol. Oceanogr. 23,605-617. Michard, G. (1971). Theoretical model for manganese distribution in calcareous sediment cores. J . Geophys. Res. 76, 2179-2186. Mills, K. C. (1974). “Thermodynamic Data for Inorganic Sulfides, Selenides, and Tellurides.” Butterworths, London. Moore, E. S., and Maynard, J. E. (1929a,b,c). Solution, transportation, and precipitation of iron and silica. Econ. Geol. 24, Part I: 272-303; Part 11: 365-402; Part 111: 506-527. Morgan, J. J. (1967). Chemical equilibria and kinetic properties of manganese in natural waters. In “Principles and Applications of Water Chemistry” (S. P. Faust and J. V. Hunter, eds.), pp. 561-622. Wiley, New York. Murata, K. J. (1939). Manganese in river and ocean muds. Am. J. Sci. 237, 725-735. Murray, J., and Irvine, R. (1895). On the manganese oxides and manganese nodules in marine deposits. Trans. R. SOC.Edinburgh 37, 721-742. Murray, J. W., and Brewer, P. G. (1977). Mechanisms of removal of manganese, iron and other trace metals from sea water. In “Marine Manganese Deposits.” (G. P. Glasby, ed.), pp. 291-326. Elsevier, Amsterdam. Murray, J. W., and Gill, G . (1978). The geochemistry of iron in Puget Sound. Geochim. Cosmochim. Acta 42, 9-19. Nissenbaum, A., Baedecker, M. J., and Kaplan, I. R. (1972). Studies on dissolved organic matter from interstitial water of a reducing marine fjord. Adv. Organic Geochem. 1971, 427-440.

Norvell, W. A. (1974). Insolubilization of inorganic phosphate by anoxic lake sediment. Soil Sci. SOC.Am. Proc. 38,441-445.

414

ROBERT C. ALLER

Nriagu, J. 0. (1972). Stability of vivianite and ion pair formation in the system Fe2(P04)H3POI-H20.Geochim. Cosmochim. Acta 36,459-470. Nriagu, J. O., and Dell, C. I. (1974). Diagenetic formation of iron phosphates in recent lake sediments. Am. Mineral. 59,934-946. Picard, G. L., and Felbeck, G. T. (1976). The complexation of iron by marine humic acid. Geochim. Cosmochim. Acta 40, 1347-1350. Rashid, M. A., and Leonard, J. D. (1973). Modifications in the solubility and precipitation behavior of various metals as a result of their interaction with sedimentary humic acid. Chem. Geol. 11, 89-97. Rhoads, D. C., Aller, R. C., and Goldhaber, M. B. (1977). The influence of colonizing macrobenthos on physical properties and chemical diagenesis of the estuarine seafloor. I n “Ecology of Marine Benthos” (B. C. Coull, ed.), pp. 113-138. Belle W. Baruch Library in Marine Science, No. 6. Univ. of South Carolina Press, Columbia. Rickard, D. T. (1%9). The chemistry of iron sulfide formation at low temperatures. Stockholm Contrib. Geol. 20, 67-95. Rickard, D. T. (1974). Kinetics and mechanism of the sulfidation of goethite. Am. J. Sci. 274,941-952.

Robbins, J. A., and Callender, E. (1975). Diagenesis of Manganese in Lake Michigan sediments. Am. J. Sci. 215, 512-533. Sanders, H. L. (1956). Oceanography of Long island Sound, 1952-1954. X.The biology of marine bottom communities. Bull. Bingham Oceanogr. Collact. 15, 345-414. Sato, M. (1960). Oxidation of sulfide ore bodies. 11. Oxidation mechanisms of sulfide minerals at 25°C. Econ. Geol. 55, 1202-1231. Singer, P. C., and Stumm, W. (1970a). Acidic mine drainage-the rate determining step. Science 167, 1121-1123. Singer, P. C., and Stumm, W. (1970b).Solubility of ferrous iron in carbonate bearing waters. J. Am. Water Works Assoc. 62, 198-202. Spencer, D. W., and Brewer, P. G. (1971). Vertical advection diffusion and redox potentials as controls on the distribution of manganese and other trace metals dissolved in waters of the Black Sea. J. Geophys. Res. 76, 5877-5892. Stookey, L. L. (1970). Ferrozine-a new spectrophotometric reagent for iron. Anal. Chem. 42, 779-781.

Stumm, W., and Lee, G. F. (1961). Oxygenation of ferrous iron. Indian Eng. Chem. 53, 143- 146.

Stumm, W., and Morgan, J. J. (1970). “Aquatic Chemistry.” Wiley, New York. Suess, E. (1979). Mineral phases formed in anoxic sediments by microbial decomposition of organic matter. Geochim. Cosmochim. Acta 43, 339-352. Tessenow, U. (1974). Losungs-, Diffusions-, und Sorptionsprozesse in der Oberschicht von Seesedimenten. IV. Arch. Hydrobiol. Suppl. 47, 1-79. Thomson, J., Turekian, K. K., and McCaffrey, R. J. (1975). The accumulation of metals in and release from sediments of Long Island Sound. I n “Estuarine Research, I. Chemistry, Biology and the Estuarine System” (L. E. Cronin, ed.), pp. 28-44. Academic Press, New York. Troup, B. N. (1974). The interaction of iron with phosphate, carbonate and sulfide in Chesapeake Bay interstitial waters: A thermodynamic interpretation. Ph.D. dissertation, The Johns Hopkins University, Baltimore, Maryland. Troup, B. N., Bricker, 0. P., and Bray, J. T. (1974). Oxidation effect of the analysis of iron in the interstitial water of recent anoxic sediments. Nature (London) 249, 237-239.

DIAGENETIC PROCESSES. 11.

415

Turekian, K. K. (1977). The fate of metals in the oceans. Geochim. Cosmochim. Acta 41, 1 139- 1 144. Van Breemen, N . , and Harmsen, K. (1975). Translocation of iron in acid sulfate soils: I. Soil morphology, and the chemistry and mineralogy of iron in a chronosequence of acid sulfate. Soil Sci. SOC.Am. Proc. 39, 1140-1148. Vishniac, W., and Santer, M. (1957). The thiobacilli. Bacteriol. Rev. 21, 195-213.

This Page Intentionally Left Blank

INDEX

A

Acoustic-reflection profiling, 26-29, 33-34 Aerobic respiration, in decomposition reaction, 273 Alabandite, 375-376, 379, 381 Alkalinity, near sediment-water interface,

256-260, 262-264, 284, 317-318, 322-339 oxidation of sulfides, 371-375 Ammonia, near sediment-water interface, NH; sampling, 254-260, 262,

264-266, 268, 283-293, 2%-298, 300-302, 308-319, 322, 324-342 Amphipod, 242 Anaerobic decomposition of organic matter,

272, 274 Anemometer, 42,72 Anoxic precipitation of iron and manganese,

375-376, 397-399, 407 Antarctic ice sheet, 7 Appalachians, erosion, 4 Atmosphere, trace metal deposition from,

130, 136, 159, 161-162,210-220, 228

Microenvironment model of bioturbated zone Bivalvia, 242, 244-245, 281, 320-321 Block Island, 9 Block Island Sound, 42, 50,76 Bloom and Stuiver sea level curve, 9,13 Bottom currents, 115 Bottom disturbance, by wind and storm, 54,

74-75 Bottom stability, and sediment transport,

87-95 Bottom water, 24-25, 72-73, 114 nontidal displacement, 82-83 wind stress, 76-82, 84 Box-core sampling, 250-252, 259-267, 353,

357-359 tabulation of data from, 322-336 Bridgeport, Connecticut tide gauge, 42 trace metals in oysters, 143 water level deviations, 50 Bristol Channel, specific dissipation, 30 Burrows and burrowing animals, 281-282,

293-297, 300-303, 390, 392-393 burrowing crab, in salt marsh, 200-201 Bursting phenomena, 109

B Bay of Fundy, tidal dissipation, 22 Beach sand, in shore materials, 85 Bedrock, in shore materials, 85 Belle (hurricane), 76-78, 89,279 Benthic fauna, 240, 242-249, 320-321 HPO:- excretion rates, 314 NH: excretion rates, 311-312 and pore-water distributions, 286-303 sediment and fluid transport, 280-282 sediment processing, 31, 91, 93, 98,

101-102, 114 Benthic manganese fluxes, 404-405 Beryllium, 149-150, 153-154 Biogenic sediment and fluid transport,

279-282, 284,407 Bioturbation, 192, 285-303, 390; see also

C

Cable and Anchor Reef, 71, 74-75 Cadmium, in oysters and mussels, 143, 146 Calcium, near sediment-water interface,

254,256,259,264,268,285,326-327, 332-333, 335-339 Calcium carbonate, near sediment-water interface, 254, 256, 285, 322-323, 329,

334, 364 Captain Islands, 8, 10 Carbonates, solubility products, 375,377-381 Carbon-14, in sediments of Long Island Sound, 131, 155-158 Carbon-14-carbon-12 ratio, in sediment of Long Island Sound, 131

417

418

INDEX

Carbon-14 dating technique, salt-marsh sam238-241, 246-250, 252-253, 255, ples, 166-167 258-316, 352-353, 356-403 Carbon-nitrogen ratios, of decomposing orbox-core data, 334-336 ganic matter, 316-317 flux-core data, 342 Cathles’ sea level curve, 13-14 gravity-core data, 339 Cenozoic glaciations, 32 macrofauna, 321 Cesium, 131 Deglaciation, Long Island and southern New Chandeleur Sound, tidal dissipation, 22 England, 7, 9-12 Chao Phya estuary, 102 Deposit-feeding fauna, 240, 242, 244, 246, Charlestown moraine, 8 281-282 Chem box core, for bottom samples, 250, Diagnetic processes near sediment-water 252-254 interface Chesapeake Bay, 101, 109 decomposition and nutrient element geomanganese fluxes, 404 chemistry, 237-342 Chloride, near sediment-water interface, box-core and gravity-core data, 322-339 254, 256, 268, 325-327, 331-336, 339 decomposition reactions, 272-274 Clay band, of salt marsh sample, 173, 178, flux-core data, 340-341 207, 225-226, 228 location, methods, and results of study, Cobalt, in streams, 132 238-272 Connecticut macrofauna, 320-321 glacial erosion, 7 nutrient flux between sediment and moraines, 10 water, 308-315 outwash deposits, 11-12 organic material, supply and reactivity, sea level changes, 15-17 274-215 till layers, 5 , 33 pore-water composition, abiogenic reConnecticut highlands, 5 action controls, 303-308 Connecticut River, 10,22,24-26,43,86-87, pore-water distributions, models of, 130 285-303 metal transport, 132-133 pore-water profiles, variation in, 282-285 trace metals in mussels, 145 products of decomposition, 276-279 Continental shelf, as sediment source in stoichiometry of decomposition,315-3 17 Long Island Sound, 85-86, 99 transport processes, 279-282 Copper iron and manganese concentrations, in Long Island Sound, 137, 139, 141, 351-415 159- 162 flux into overlying water, 401-405 in New Haven Harbor, 134 location, methods, and results of study, in oysters and mussels, 143-145 352-367 in salt marsh, 136, 138, 168, 183-186, 188, precipitation reactions and saturation 191-192, 201-207, 212-219, 221, states, 375-382 227-228 production of, 368-375, 384-400 in streams, 134 seasonality, 382-384 Crustacea, 320-321 Dredge spoils, in sediment, 131 Current meter, 42, 44,46, 48, 51, 75-79 Dredging, effect on sediment storage capacCurrents, in Long Island Sound, 75-82, 115 ity, 103 unidirectional particle transport, 108 Current velocity, 115-1 16

E D Dammina of rivers. 132-133 DEEP station, central Long Island Sound,

Easterly winds, 50 East River, 22-23 tidal locks, effect on sedimentary regime, 31

419

INDEX

Eddy, in estuary, 116 Eddy-diffusion coefficient, in sand transport, 118-121, 124-125 Effective fetch, 72 Elmhurst moraine, 8 Embayment, energy balance formulation, 61-65 End moraine, 7-8, 10 acoustic reflection profile, 27 in shore materials, 84-85 Energy, in estuarine sedimentary processes, 99-101 Energy balance, in embayment, formulation Of, 61-65 Erosion glacial, of New England, 7, 12 rate, on land surface supplying sediment to east coast of North America, 4, 12 river banks, 87 shoreside, as sediment source, 84-85, 101 Estuarine circulation, I , 23-25, 43, 75 salinity gradient, 82 and sand flux, 115 Estuary, 130 sediment storage capacity, 30 sediment systems, 1-2, 99-103 Eustatic sea level curve, 13-14

F Falkner Island, 5 , 71-72 Fall Line, of Long Island Sound, 2, 4, 6 Fall Zone surface, of Long Island Sound, 2-5, 10-11 Farm River, 11 Farm River salt marsh, Connecticut dating of deposits, 193-198 history, 170-172 Lack of disturbance and chemical immobility, 199-210 peat properties, 172-181 sediment and trace metals, 161, 168-169, 181-189 atmospheric sources, 210-220 silt and clay sources, 225-227 Fauna, benthic, see Benthic fauna Fecal pellets, in sediments, 31-32, 91-93, 102, 114, 241 Fermentation, in organic decomposition, 273

Ferrous oxide ion activity products, 376-378 reduction, 368-369, 384, 406 Ferrous sulfide, at sediment-water interface, 257, 270-271, 276-278, 317, 322-323, 329-33 1, 334, 377 oxidation, 369-370, 374, 384, 406 Fishers Island Sound, 31 Floods, and sediment transport, 109 Flow velocity, see Current velocity Fluid mud, 102 Flux-box core, for bottom samples, 250, 253-255, 354, 364-367 data, 340-341 macrofauna, 318-319 FOAM station, central Long Island Sound, 238-244, 252-253, 255, 258-316, 352-353, 356-403 box-core data, 323-327 flux-core data, 340 gravity-core data, 337 macrofauna, 320

G Gastropoda, 242, 320-321 Glacial drift, 7-8, 14 Glacial sand, 26 Glaciations, eastern North America, 5,7,32 Grass, of salt marsh, 165-167, 173,221-224, 229 Gravity-core sampling, 252, 254, 257-259, 306-307, 353, 355-357 tabulation of data from, 322, 337-339 Great Marsh, Connecticut, 167 Greigite, 276, 375, 377

H Hammock River Marsh, 13 Harbor Hill moraine, 8, 10 Hauerite, 381 Heavy metals, in sediments, 32 Hermit crab, 242 High marsh, 165-166 Hitchcock, Lake, 10-11, 32 Housatonic River, 22 trace metals, 132, 143, 145 Hurricane, effect on bottom water and sediment, 76-78, 89, 279

420 Hydrogen sulfide in salt marsh, 168 near sediment-water interface, 322-324 Hydroxyapatite, 305 Hydrozoa, 320-321

INDEX

atmospheric fluxes, 210-220 dating of deposition, 189-198 in streams, 132, 134 Ledyard moraine, 8, 10 Long Island Sound acoustic reflection profiles, 26-29 bedrock geology, 6 I contours of water depth, 71 Ice-contact drift, 7, 10 currents, 75-82, 115 Industrial wastes, 101, 130, 132 dated organic matter, location of, 9 Interglacial period, denudation rate of land diagenetic processes near sediment-water surface, 32 interface, 237-415 Irish Sea, tidal dissipation, 22 decomposition and nutrient element Iron, see also Ferrous oxide; Ferrous sulfide geochemistry, 237-342 reduction, in organic decomposition, 273 iron and manganese concentrations, in salt marsh, 168-169, 178-181,183-186, 35 1-4 15 190-192,201-203,210,214-215,217, estuarine circulation, 43 227-228 geological history, 2-12, 122, 130 near sediment-water interface, 254-255, geometric and tidal characteristics, 104 257, 268, 274, 276, 323-415 glacial drift, depth of, 8 flux into overlying water, 401-405,408 islands, composition of, 5 location, methods, and results of study, nuclides, sources and sinks of, 129-164 352-367 physical oceanography, 20-25 precipitation reactions and saturation power characteristics, 84 states, 375-382 river flow, 82 production, 368-375,384-400,406-408 salinity, 24-25 seasonality, 382-384, 406 sand content variation, 120-121 in streams, 132 sand transport at floor, 107-128 sea level rise, 12-20 sedimentary system, 1-39 J sedimentation parameters, 98 sediment sources, 84-87 Jameco formation, 7 sediment transport and deposition, 69-106 K shore materials, 85 storm energy, 48-55 Knight Inlet, 100 temperature, 24-25 tidal dissipation estimate, 65-66 L tidal energy, 43-47, 60, 70, 84 water level deviations, 55-60 Lacustrine deposits, 27, 95 waves, 70, 72-75 Lag effect, in sediment transport, 108 Lordship outwash, 10 Land elevation curve, 13-14 Low marsh, 165-166 Lead in Long Island Sound, 131, 137, 139, 141, Lunar semidiurnal tide, 20-21 149- 150,153- 159,161-162,280-28 1, 284 in mussels, 146, 152 M in New Haven Harbor, 134 in salt marsh, 136-138, 168-169, 181-188, Mackinawite, 276, 375, 377, 380 191-392, 201-202, 227-228 Madison moraine, 8, 10

42 1

INDEX

Manganese reduction, in organic decomposition, 273 in salt marsh, 168, 183-186, 190-192,

Near-bottom flow velocity, 116 New England, geological history, 4-5, 7,

201-204,206,208-210,214-215,217, 227-228 near sediment-water interface, 254-255, 274, 323-415 flux into overlying water, 401-405, 408

New Haven, Connecticut clay mineralogy of sediments, 32 fluid mud, 102 population growth, 172 sand-mud ratio, 119, 121 sewer outfalls, 133-135 tide gauge, 42, 51 trace metals in oysters and mussels, 143,

location, methods, and results of study, 352-367

precipitation reactions and saturation states, 375-382 production, 368-375,384-399,406-408 seasonality, 382-384,406 Manganese oxide reduction, at sedimentwater interface, 368-369, 406 Mannetto formation, 5 Marsh, Long Island Sound shoreline, 85 Martha’s Vineyard, glaciations, 5 Mattituck sill, 9, 14-15, 34, 71, 110-111, 122-123, 125

glacial outwash, 95-96 sand flux, 90 sand waves, 114-1 15 Meltwater curve, 13 Mercury, in New Haven Harbor, 133-135 Mersey River, 103 Metals, see Trace metals Methane production, in organic decomposition, 272-273 Microenvironmentmodel of bioturbated zone, 294-303, 390-396 Mill River, 11

Mineral material in sediment, 91 Montauk, New York, 42, 50 Moraine, 7-8, 10, 84-85 acoustic reflection profile, 27 Morner’s sea level curve, 13-14 Mud, 26-27, 87, 89, 101-102, 110, 114; see also Sand-mud transition zone deposition, 95-98 transport, 91-95 Mussels, trace metals in, 142, 144-147, 152

N Narragansett Bay, 22, 404 Nauerite, 375 Naugatuck River, 132

10-12

145

water level deviations, 48-50 New London, Connecticut mean sea level, 15-16 tide gauge, 42, 55 trace metals in oysters, 143 water level deviations, 48-50, 55-60 Newport, Rhode Island, water level deviations, 48-49 New Rochelle, New York, 42 Nickel, 148 in mussels, I46 Nitrate reduction, in organic decomposition, 273 Nitrogen-carbon ratios, of decomposing organic matter, 316-317 Nitrogen concentration, near sedment-water interface. see Ammonia Noank, Connecticut, trace metals in oysters, 143

Norwalk, Connecticut, trace metals in oysters, 143 Norwalk Islands, 8, 10 Nuclides, see Trace metals Nucula annulata, 93, 242, 244, 320-321 NWC station, central Long Island Sound, 238-241,244-247,250,252-253,255, 258-316, 352-353, 356-403

box core data, 328-333 flux-core data, 341 gravity-core data, 338 macrofauna, 321

0

Ocean basin volume adjustment, 13 Old Saybrook moraine, 8, 10 One-dimensional model of pore-water dis-

422

INDEX

tributions, 285-293, 396-399 Organic matter dated remains, location of, 7, 9 at sediment-water interface, 269-270, 323, 328-331, 334 decomposition products, 276-277 decomposition reactions, 272-274 stoichiometry of decomposition, 315-317 supply and reactivity, 274-275 Outwash deposits, 7, 10-12, 26-27, 85, 95-96 Oysters, trace metals in, 142-144

P Peat, in salt marsh, 13, 166-167, 172-181, 199, 220-227, 229 trace-metal analysis, 185-189 Pelletized material, see Fecal pellets Phosphate mineral formation in sediments, 304-308 near sediment-water interface, HP0:sampling, 254-260, 262, 261-268, 283-284, 305, 308-31 I , 314-319, 323-342 Plankton, 24, 283-284, 314-316 Plutonium, 131, 154-156 Polluted streams, and trace metals in Long Island Sound, 132-135 Polonium, in mussels, 152 Polychaeta, 242,244-245,247,281,297,299, 320-321 Pore water, near sediment-water interface of Long Island Sound, 252-268, 282-308, 317-318 box-core data, 323-336 gravity-core data, 337-339 iron and manganese concentrations, 353-362, 394-395 Port Jefferson, New York, 42, 50 Potomac estuary, 76, 100-101 Pyrite in salt marsh, 168-169, 178-181,203,227 near sediment-water interface, 257, 271-272,216-277,322-323,329-331, 334 oxidation, 369-371, 314

Q Quinnipiac River, 5 , 11, 13 metal concentrations, 132-134

R Race, the, Connecticut, 71 Radionuclides, see Trace metals Radium, in salt marsh, 181-184, 187 Reddingite, 305, 307-308, 381 Resonant basin, 20 Resuspended sediment, 31-32, 34-35, 5 5 , 87-88, 93-95, 114, 125, 240, 279 Rhodochrosite, 375-376, 379, 381-382, 407 Rivers flow and discharge into Long Island Sound, 82, 86-81 metal concentrations, 131-1 34 power, in estuarine sedimentary processes, 100 sediment supply rate, 98 Rogers Lake, 7, 9 Ronkonkoma moraine, 7-8 Rotary tides, 20, 22 S

Sachem Head, Connecticut, sand-mud ratio, 119, 121 Salinity, Long Island Sound, 24-25, 82, 114 Salt hay, see Spartina patens Salt marsh accretion, theories of, 220-227 dating of samples, 189-198 elevation, 16-20, 34 sediment and trace metals, 136-138, 161, 165-236 Salt thatch, see Spartina alterniflora Sand, 26, 85, 87, 89, 241 distribution, in Long Island Sound, 110, 112-113 grain size, 113 trace-metal content, 137 transport, 89-91,96-97, 107-128 Sand-mud transition zone, 27, 113 formation, 116-122, 125 sand distribution, 123-124 Sand wave, 89-91, 96, 114-115

423

INDEX

Sandy Hook, New York, 42, 50 Savannah River, 103 Sea-level changes, 9, 11-20, 34 rise, and sediment storage capacity, 30 and salt-marsh accretion, 192-198,222-225 Sea level curve, 13-14 Sea-surface elevation, 4, 15 Seawater transport, biogenic, 282 Sediment accumulation rate, 30, 130 bulk density, 175, 177-178, 183-184 composition, in Long Island Sound, 110-113 iron and manganese concentrations, 353-355 Long Island Sound’s sedimentary system, 1-39 resuspended, 31-32, 34-35, 55, 87-88, 93-95, 114, 125, 240, 279 in salt marsh, 165-236 sources, 84-87, 101 storage capacity, 30-31, 102-103 thickness, in sand-mud transition zone, 117 trace metals, 137-142, 147-163 transport and deposition, 25-33, 69-106, 108-110, 113-116, 279-282 and bottom stability, 87-95 power sources, 70-84 in stormy period, 5 5 , 61 Sediment trapping efficiency, of Long Island Sound, 30, 32-33, 35, 98-99 Sediment-water interface, diagenetic processes near, see Diagenetic processes near sediment-water interface Sediment yield, 4 of rivers entering Long Island Sound, 26, 30, 32 Severn estuary, 102 Sewage wastes, 23-24, 101, 130, 133-135 Shell debris, in sediments, 241, 243-247, 249, 281-282 Shoals, and sediment transport, 109 Shoreside erosion, 84-85, 101 Siderite, 375-378, 381 Silt, deposition of, and salt-marsh accretion, 220, 223 Silt-clay sediment, 25-26,30-32,91,93,102, 24 1

deposition rate, 97 of salt marsh, 225 Silver, in streams, 132-134 Spartina alterniflora, 165,167,200,209,222, 224 Spartinapatens, 165-166,169,173,176,190, 193, 200, 209, 222 Specific dissipation, 30-31, 99-101, 103 Stagnation-zone retreat, in deglaciation of New England, 7 Stony Creek, Connecticut, salt marsh leveling measurements, 16-17 Storm energy, 48-55 Storms and flow velocity, 114 resuspension of sediment, 31, 279 and sand distribution, 123-124 and sediment transport, 109 Stratified drift, 7 Struvite, 304-305 Sulfate near sediment-water interface, SOf sampling, 254-261, 283-284, 317, 322-339 SO:- concentrations during anoxic incubation, 375-376 Sulfate-chloride ratio, in salt marsh, 188-189, 191, 203, 209, 227 Sulfate reduction in organic decomposition, 273-278, 316-317 in salt marsh, 168-169 Sulfide oxidation, at sediment-water interface, 369-375 in salt marshes, 168-169, 202-203, 227 solubility products, 375-381 Surficial sediment, in Long Island Sound, 110 Suspension-feeding fauna, 242, 244, 247, 281-282 Susquehanna River, 132

T Tay, Firth of, 100 Temperature, and flux of iron and manganese into overlying water, 368, 401-403

424

INDEX

Thames estuary, 102 Thames River, 22 Thiobacillus , 369 Thorium, 131, 149-151, 153-154, 161-162, 280-281, 284 Tidal characteristics of Long Island Sound, 20-22 Tidal currents, 1 I3 and sand flux, 115 Tidal dissipation estimate, for Long Island Sound, 65-66 Tidal energy, 41, 43-47, 60,70, 84 Tidal locks, effect on sedimentary regime, 31 Tidal stream, 75, 79-80 Tidal velocity, 79-80, 108-109, 116, 122 Tidal wave, and sediment transport, 109 Tide, 113 salt marsh submersion, 21 1 and sedimentary processes, 101, 107-109 Tide gauge data, 15-16, 42, 51, 55 Till, 5, 33 Tilt rate, east coast of North America, 4, 10 Totoket bog, 9 Trace metals in Long Island Sound, 129-164 distribution in sediments, 137-142 in mussels and oysters, 142-147 processes affecting deposition and accumulation in sediments, 147-161 sources of, 131-136 in salt marsh, 165-236 atmospheric supply, 136-138 Two-dimensional model of pore-water distributions, 294-303, 390-3% U

Uca pugnax, in salt marsh, 200-201 Unbound sediment, 98

V Valley train, 7, 10-1 1 Vivianite, 305-306, 308, 380

W Water column scavenging of trace metals, 149- I50 Water level, in Long Island Sound, 15-16, 48-53, 55-60 Water velocity at bottom, 72-73 in stormy period, 54 Wave gauge, 44 Wave recorder, 74 Waves, 70, 72-75, 87, 89, 100 West River, 11 Whitlockite, 305-306,308 Winds, 70, 72-75 and currents, 76-82, 84 and flow velocity, 114 and sand distribution, 123 and sedimentary processes, 100-101, 109 and tidal stream velocity, 54 and water level deviations, 48 in winter storm, 50-51 Wisconsinan glaciations, 5, 7, 32

X

X-ray box core, for bottom samples, 250, 252-253

Y Yoldia limatuia, 93, 242, 244, 246, 311-312, 314, 320-321, 383

2 Zinc in Long Island Sound, 137, 139, 141, 161-162 in New Haven Harbor, 134-135 in oysters and mussels, 142-143, 146 in salt marsh, 136, 138, 168, 183-186, 188, 191-192, 201-202, 213-219, 227-228