OCR error (C:\Conversion\JobRoot\00000AOP\tiff\2000WB12.tif): Unspecified error
-------
Regional Center for Cm ironmentdl I
US EPA Region III
1650AichSt
Philadelphia, PA 19103
-------
903R88005
,.!. rental Protection
1 u,rornidlion Resource
„ ...... ; uPi.'^>
U41 Gi'.ssinut Street
hia, PA 1910? ,-j
Understanding the Estuary:
Advances in
Chesapeake Bay Research
Proceedings of a Conference
29-31 March 1988
Baltimore, Maryland
Maurice P. Lynch and Elizabeth C. Krome
Editors
August 1988
CRC Publication No. 129
CBP/TRS 24/88
Chesapeake Research Consortium • P.O. Box 1280 • Solomons, Maryland 20688
-------
DISCLAIMER
Mention of trade names or commercial products does not constitute endorsement or
recommendation for use. Statements made or ideas expressed in these Proceedings are those of the
identified authors and are not to be construed as positions or policies of the agencies or institutions
which may employ the authors or of the sponsors who provided support for publication of these
Proceedings.
-------
ACKNOWLEDGEMENTS
Plenary session moderators and session presenters were conscientious in
furnishing materials to aid in summaries. Kevin Kasowski provided helpful notes
from the conference. Leon Clancy, Robert J. Lukens, Ruth Hershner, Janet
Walker, and Pamela M. Owens assisted in various phases of electronic conversion
and word-processing. G. Glynn Rountree gave valuable help in proofreading.
Joseph A. Mihursky's comments on galley proofs provided useful feedback.
Pamela M. Owens deserves special mention for her dedicated and diligent work on
all the papers from concurrent sessions. Karen L. McDonald provided invaluable
guidance, support, and encouragement.
SPONSORS
This conference was held under the auspices of
the Chesapeake Bay Program and was funded or supported by:
Chesapeake Bay Liaison Office
U.S. Environmental Protection Agency
Scientfic and Technical Advisory Committee
Chesapeake Bay Program
Virginia Graduate Marine Science Consortium
Virginia Sea Grant College Program
University of Maryland
Maryland Sea Grant College Program
The Academy of Natural Sciences of Philadelphia
Chesapeake Research Consortium, Inc.
Metropolitan Washington Council of Governments
Chesapeake Bay Foundation
Chesapeake Bay Trust
The Procter and Gamble Company
Joseph Meyerhoff, II
-------
CONFERENCE STEERING COMMITTEE
Dr. Maurice P. Lynch and Dr. Joseph Mihursky, Co-chairs
Dr. William Rickards Ms. Gail MacKieman
Dr. Richard Jachowski Dr. Jonathan Garber
Dr. Thomas Osborn Dr. Kent Mountford
Ms. Karen L. McDonald: Conference Coordinator
Ms. Pam Owens: Conference and Proceedings
Production Assistant
CONFERENCE ASSISTANTS
Elizabeth Krome
Kevin Kasowski
Karen Kelly
Tammy Rowe
Claudia Walthall
Julia Wilcox
Susan Pultz
Janet Norman
Chao-Hsi Chang
Staff of National Aquarium of Baltimore
JV
-------
FOREWORD
A conference entitled "Understanding the Estuary: Advances in Chesapeake Bay Research"
was held 29-31 March 1988 in Baltimore, MD. The conference was primarily oriented towards
scientists engaged in research on fundamental estuarine processes in Chesapeake Bay and
secondarily oriented to managers with scientific and technical backgrounds. Scientifically and
technically conversant citizens were also encouraged to attend.
The conference examined recent research findings in several areas that are relevant to a
wide range of estuarine processes. Conveners attempted to provide a context for assessing the
relevance of these scientific findings to the long-term efforts to protect and restore the Chesapeake
watershed.
Estuarine research is coming of age in terms of scientific credibility. For years, many
scientists engaged in oceanographic or limnological studies considered estuarine research as a
spinoff or splinter effort of blue-water or freshwater work. Now, the realization is growing that
estuaries are unique areas, worthy of focused attention. This recognition comes none too soon.
Coastal demographic pressures and readily visible degradation of habitats, living resources, and
water quality, have created a strong public outcry to "do something" about coastal and estuarine
pollution.
The Chesapeake Bay area is fortunate in that estuarine research has been nurtured and
encouraged by the regional states, which established laboratories dedicated to estuarine studies of
the Bay as far back as the 1920's. In the past two decades, interaction of Bay scientists, resource
managers, and the public has both nurtured and been nurtured by major federally funded
Chesapeake Bay studies. The most recent of these studies was funded through the U.S.
Environmental Protection Agency and resulted in a coordinated federal, Maryland, Virginia,
Pennsylvania, and District of Columbia Chesapeake Bay Restoration and Protection Program.
This program is considered a model for estuarine management both nationally and internationally.
The emergence of the Chesapeake Bay Program as a national and international model of
estuarine management is based to no small extent on accumulated information developed by the
Bay research community. Maintaining Chesapeake Bay Program's momentum will depend on two
things. The Bay scientific and technical community must improve understanding of the vital
estuarine processes. Equally important, this understanding must be communicated to the managers
and the public. It is hoped mat this conference contributed to this improvement in understanding
and communication.
We wish to thank the individuals who contributed to these Proceedings and the conference
by attending the sessions, presenting the papers, and participating in the discussions. We hope to
continue these exchanges through similar conferences and activities in the future.
Maurice P. Lynch
Joseph A. Mihursky
Conference Co-chairs
-------
TABLE OF CONTENTS
KEYNOTE ADDRESS
Uncertainty in Estuarine Research
JohnCostlow 3
PLENARY SESSIONS
PHYSICAL PROCESSES
Eric C. Itsweire, Moderator
Changes in Understanding of the Circulation of the Chesapeake Bay
William C. Boicourt 5
Opportunities to Improve Understanding of the Circulation of the Chesapeake Bay
Thomas R. Osborn 6
Stratification Control and Bay-Shelf Exchange: Two Physical Processes and Their Implications for
Ecosystem Dynamics
David M. Goodrich 7
TOXICS
James G. Sanders, Moderator
Chemical and Physical Processes Influencing Bioavailability of Toxics in Estuaries: An Overview
James G. Sanders 13
Factors Affecting the Bioavailability of Toxic Trace Elements to Estuarine Organisms
Gerhardt F. Riedel and James G. Sanders 14
Bioavailability of Organic Pollutants to Aquatic Organisms
Robert C. Hale and Robert J. Huggett 26
GENETICS AND SPECIES CONSERVATION
Robert W. Chapman, Moderator
Introduction to Plenary Session
Robert W. Chapman 33
New Methodology for Immunologic Discrimination of Stocks or Populations in Fish Species
Raymond Simon 34
Mitochondrial DNA Analysis of Chesapeake and Delaware Bay Populations ofFundulus heteroditus
Michael W. Smith, Robert W. Chapman, and Dennis A. Powers 36
VII
-------
Genetic Analysis of Oysters that Grow at Different Rates
KenPaynter 38
BENTHIC-PELAGIC COUPLING
Jonathan Garber, Moderator
Introduction to Plenary Session
Jonathan Garber 39
Production of Paniculate Organic Matter in the Mesohaline Reach of the Chesapeake Bay
Thomas Malone 40
Deposition of Organic Matter to the Sediment Surface
W. Michael Kemp 44
Uptake and Release of Nutrients from Sediments
Walter R.Boynton 46
PELAGIC TROPHIC STRUCTURE
Peter G. Verity, Moderator
The Trophic Structure of Pelagic Communities: Hypotheses and Problems
Peter G. Verity 49
Trophic Structure of the Chesapeake Mesohaline Ecosystem
Robert Ulanowicz 55
The Importance of the Microbial Loop in the Chesapeake Bay and its Tributaries
Kevin G. Sellner 60
"WHERE DO WE GO FROM HERE?"
Joseph A. Mihursky, Moderator
Living Resources
Edward Houde 66
Dissolved Oxygen
Eugene Cronin 67
Submerged Aquatic Vegetation
Robert J. Orth 67
Toxics
Nicholas Fendinger 68
-------
Physical Processes
EricC. Itsweire and Owen Phillips 69
Nutrients
CarlCerco 69
"HOW DO WE GET THERE; WHO PAYS THE FARE?"
Joseph A. Mihursky, Moderator
Remarks by Monica Healy, Chris D'Elia, John W. Daniel, Frank Perkins, Ian Morris,
David Challinor, and Sheldon Samuels 70
NOTE: Because of the informal nature of the setting, coverage of the excellent luncheon addresses by the
Honorable Joseph Gartlan and Dr. Alvin Morris is not included in these Proceedings.
CONCURRENT SESSIONS AND POSTER SESSION
LIVING RESOURCES
Chairs: William Hargis and Linda Schaffner
Effect of Changing pH and Salinity on Chesapeake Bay Striped Bass Larvae (abstract only)
AndrewS. Kane, Richard O. Bennett, and Eric B. Mary 77
Grazing and Egg Production by Chesapeake Bay Zooplankton in Spring and
Summer (abstract only)
Jacques R. White and Michael R. Roman 78
The Effects of Suspended Sediments on Microzooplankton Grazing in the Patuxent River,
A Subestuary of the Chesapeake Bay
David C. Brownlee, Kevin G. Sellner, and Kevin R. Braun 79
Temporal and Spatial Variations in Zooplankton of the Mesohaline Portion of
Chesapeake Bay (abstract only)
Michael R. Roman, Jacques R. White, and Anne L. Gauzens 91
Simulating the Vertical Motion of Nekton in the Estuarine Environment - Scientific
Considerations and Speculations
Charles Bostater and Robert Biggs 92
Effects of Natural Environmental Fluctuations on Defense-related Oyster Hemocyte
Activities (abstract only)
William S. Fisher and Marnita M. Chintala 109
Evidence for Loss of Suitable Benthic Habitats for Oysters in Tributaries of the Chesapeake Bay
H. H. Seliger and J. A. Boggs Ill
Stabilized Coal Ash as Substratum for Larval Oyster Settlement: A Pilot Field Study
Kent S. Price, Joel Rosenfeld, KarolynM. Mueller, and Thomas Warren 128
The Sedimentary History of Submerged Macrophytes in the Upper Chesapeake Bay (abstract only)
Grace S. Brush and WilliamB. Hilgartner 137
IX
-------
Alternative Sampling Strategies for a Survey of Submerged Aquatic Vegetation in the
Chesapeake Bay
Paul H. Geissler 138
The Status of Yellow Perch Populations in Five Chesapeake Bay Tributary Streams (abstract only)
Holly Greening, A. Janicki, andP. Saunders 147
Bay Anchovy Ecology in Mid-Chesapeake Bay (abstract only)
T. A. Newberger, E. D. Houde, andE. J. Chesney, Jr 149
Hatchability and Viability of Striped Bass Eggs: Effects of Tributary Source and
Female Sire (abstract only)
Edward D. Houde, C.E.Zastrow, andE. H. Saunders 150
Current Research to Improve the Understanding of Habitat Use, Distribution, and Population Status
of Canvasbacks on the Chesapeake Bay (abstract only)
G. Michael Haramis and Dennis G. Jorde 151
Patterns of Post Larval Availability and Settlement in the Blue Crab: Effects of Time, Space, and
Habitat (abstract only)
Eugene J. Olmi, in, Jacques van Montfrans,
Romuald N. Lipcius, and Robert J. Orth 152
Regulatory Mechanisms of Postlarval Blue Crab Recruitment: Settlement, Metamorphosis,
and Developmental State (abstract only)
RomualdN. Lipcius, Eugene J. Olmi, III, and Jacques van Montfrans 153
Variation in Postlarval Blue Crab Settlement on Artificial Substrates in the York River,
Virginia (abstract only)
Jacques van Montfrans and Robert J. Orth 154
NUTRIENTS
Chairs: George Simmons and Carl Cerco
Radionuclide Concentrations in Susquehanna River and Chesapeake Bay Sediments — Implications for
Transport and Distribution of Particle-Reactive Pollutants
R. I. McLean, S. L. Domotor, J. K. Summers,
V. Dickens, and C. R. Olsen 157
Quantifying Pollutant Sources to Rock Creek Estuary: The Patapsco River, Sediment
Remineralization, and Non-point Source Runoff
RogerS. Copp and Robert M. Summers 170
Nutrient Regeneration Rates in Chesapeake Bay Bottom Sediments Based on Bulk Properties
James M. Hill and Jeffrey Halka 184
Delineation of Regional Sediment Resuspension Potential in Chesapeake Bay, with Implications
for Bottom Sediment Management
Richard W. Faas 201
Comparison of Sediment Landscapes in Chesapeake Bay as Seen by Surface and Profile Imaging
R. J. Diaz and L. C. Schaffner 222
Use of In Situ Nutrient Addition and Dilution Bioassays to Detect Nutrient
Limitation in the Tidal Freshwater Potomac
R. Christian Jones 241
-------
Numerical Tagging of Phosphorus in the James Estuary (abstract only)
Wu-SengLung 253
Some Response Patterns of Phytoplankton to Nullification Based on Enrichment Experiments
and Apparent Influence of Silicon (abstract only)
T. J. Smayda and T. A. Villareal 254
The Importance of Submarine Groundwater Discharge to Nutrient Flux in Coastal
Marine Environments
George M. Simmons, Jr 255
Nitrogen Cycling in Chesapeake Bay Sediments: Balance Between Regeneration and
Denitrification (abstract only)
W. Michael Kemp, P. A. Sampou, J. M. Caffrey, M. S. Mayer,
K. Hendriksen, and W. R. Boynton 270
Variability of Groundwater Nitrate Concentrations in Non-Agricultural Ecosystems
Timothy B. Parkin, E. E. Codling, J. J. Meisinger, and J. L. Starr 272
Chesapeake Bay Sediment Monitoring for Water Quality Model Development
Lewis C. Linker 283
PHYSICAL PROCESSES
Chairs: William Boicourt and Evon P. Ruzecki
Hampton Roads Circulation: The Combined Effects of General and Meso-Scale
Features (abstract only)
Evon P. Ruzecki and David A. Evans 294
A Theory of Tidal Intrusion Front and its Practical Application (abstract only)
Albert Y. Kuo, Robert J. Byrne, Evon P. Ruzecki,
J. M. Brubaker and Paul V. Hyer 295
Changes in Circulation and Salinity from Increased Channel Depth in the Baltimore Harbor
Peter Olson and Vincent Grano 296
A Numerical Investigation of Circulation and Salt Distribution in the Patuxent River Estuary
Chris Kincaid, Peter Olson, and Harry Wang 323
Lagrangian Drift Model of Suspended Sediment Transport in Chesapeake Bay
Kurt W. Hess 352
How to Estimate the Thickness of Benthic Boundary Layers in Estuaries With and
Without Tides
Sergei A. Kitaigorodskii 369
High-Resolution Thermistor Chain Observations in the Upper Chesapeake Bay (abstract only)
Charles C. Sarabun, Jr 373
Satellite Observations of Turbidity Variations in Chesapeake Bay, Spring-Summer 1987
Richard P. Stumpf 374
Evaluation of Conowingo Reservoir Release for Controlling Salinity in the
Upper Chesapeake Bay
Bernard B. Hsieh 385
xt
-------
Controlled Energy Dissipation from River Inflow as a Factor in
Managing Estuarine Water Quality
T. A. Wastler 399
Application of a Multiple Linear System to Identify Tidal Signals in the Upper Chesapeake Bay
Bernard B. Hsieh 410
Time Variations of Bottom-Water Inflow at the Mouth of an Estuary
Glenn A. Cannon 424
TOXICS
Chairs: Harriette L. Phelps and Robert C. Hale
Dynamics of Organic Pollutants in Blue Crabs, Callinectes sapidus, Collected from the Lower
Chesapeake Bay Region (abstract only)
Robert C. Hale 430
Bioassay for Phytotoxicity of Toxicants to Sago Pondweed
W. James Fleming, Jeffrey!. Momot, and M. Stephen Ailstock 431
Inducible Adaptations as Bioassays
Brian P. Bradley, Roxana Hakimzadeh, and Roger M. Davis 441
The Role of Photochemistry in the Bioavailability of Toxic Substances in Estuaries;
Implications for Monitoring (abstract only)
George R. Helz and Robert J. Kieber 449
Dechlorination: Is it the Answer? (abstract only)
George R. Helz, Anthony C. Nweke, and Philip J. Kijak 450
Implications of Toxic Materials Accumulating in the Surface Microlayer in Chesapeake Bay
H. Gucinski, J. T. Hardy, and H. R. Preston 451
Assessing the Impact of DOD's Installations on the Water Quality of the Chesapeake
Bay Region
L. Peter Boice and Steven D. Garbarino 468
Some Histologic Gill Lesions of Several Estuarine Finfishes Related to Exposure to
Contaminated Sediments: A Preliminary Report
W. J. Hargis, Jr. and D. E. Zwerner 474
Chemical and Biological Analysis of the Effluents from Oil and Water Separators in
Commercial Shipyards (abstract only)
Charles Banks 488
Acidic Conditions in Maryland Coastal Plain Streams: Potential Impacts on Anadromous
Fish (abstract only)
A. Janicki, C. Knapp, H. Greening, andR. Klauda 489
Sediment Toxicity Testing in the Chesapeake Bay
Harriette L. Phelps 491
Development of an Estuarine Solid-Phase Sediment Bioassay Using Molluscan
Larvae (abstract only)
Harriette L. Phelps, Kelton Clark, and Kimberly Warner 505
xn
-------
DISSOLVED OXYGEN
Chairs: William Richards and Jack Greer
Organic Carbon, Oxygen Consumption, and Bacterial Metabolism in Chesapeake Bay (abstract only)
R. B. Jonas, J.H. Tuttle, J.T. Bell, and D.G. Cargo 509
Fluxes of Carbon, Nitrogen, and Oxygen Through Estuarine Bacterioplankton
Hugh W. Ducklow, Emily A. Peele, Suzanne M. Hill, and Helen L. Quinby 511
Planktonic Respiration in Mesohaline Waters of Chesapeake Bay: Seasonal Patterns of
Size-Fractionated Oxygen Consumption with Reference to Depletion of Bottom Water
Oxygen (abstract only)
Peter Sampou, W. T. Randall, and Michael Kemp 524
The Relative Significance of Macrophyte Decomposition and Phytoplankton Respiration in the
Consumption of Oxygen in the Lower Chesapeake Bay
Luis M. Lagera, Jr. and Joseph C. Zieman, Jr 525
Ecological Changes in Chesapeake Bay: Are They The Result of Overharvesting the American Oyster,
Crassostrea virginical
Roger Newell 536
Oxygen Fluctuations and Fish Population Dynamics in a Chesapeake Bay Oyster Bed
Denise Breitburg 547
Role of the Organic and Inorganic Carbon Systems in the Dissolved Oxygen Regime of the
Chesapeake Bay
T.A. Wastler 558
Long Term Pattern of Anoxia in the Chesapeake Bay
H. H. Seliger and J. A. Boggs 570
Insights to the Chesapeake Bay's Eutrophication Process through Modeling
Charles App and James J. Fitzpatrick 584
Potential Biological Effects of Modeled Water Quality Improvements
Resulting From Two Pollutant Reduction Scenarios
Kent Mountford and Robert C. Reynolds 593
Variability of Dissolved Oxygen in the Mesohaline Chesapeake Bay (abstract only)
L. P. Sanford, K. G. Sellner, and M. H. Bundy 607
Metabolic and Respiratory Compensation During Long Term Hypoxia in Blue Crabs,
Callinectes sapidus
Peter L. deFur and Anita L. Pease 608
SUBMERGED AQUATIC VEGETATION
Co-chairs: Robert J. Orth and Ken Moore
Submerged Aquatic Vegetation in the Chesapeake Bay: A Barometer of Bay Health
Robert J. Orth and Ken Moore 619
NOTE: There has been some rearrangement of papers in the Proceedings, primarily in the distribution of
poster session papers to related topical sessions.
Xlll
-------
PLENARY SESSIONS
MODERATORS
Physical Processes
Eric C. Itsweire
The Johns Hopkins University
Baltimore, Maryland 21211
Bioavailability of Toxics
James G. Sanders
The Academy of Natural Sciences
Benedict Estuarine Research Laboratory
Benedict, Maryland 20612
Genetics and Species Conservation
Robert W. Chapman
The Johns Hopkins University
Chesapeake Bay Institute
Shady Side, Maryland 20764-0037
Benthic-Pelagic Coupling
Jonathan Garber
University of Maryland
Chesapeake Biological Laboratory
Solomons, Maryland 20688
Pelagic Trophic Structure
Peter G. Verity
Skidaway Institute of Oceanography
Skidaway Island
Savannah, Georgia 31416
Summary and Panel Discussion
Joseph A. Mihursky
University of Maryland
Chesapeake Biological Laboratory
Solomons, Maryland 20688
-------
KEYNOTE ADDRESS
Uncertainty in Estuarine Research
John Costlow, Director
Duke University Marine Laboratory
Beaufort, North Carolina 28516-9721
My question today is why there is any uncertainty
left in estuarine research. After all, AERS began
looking at estuarine research in 1947. A monumental
research symposium was held in 1964, and the Estuar-
ine Research Federation was formed shortly thereafter.
Within this time we have also seen the establishment of
the U. S. Office of Naval Research, the National
Science Foundation, the Environmental Protection
Agency, the National Oceanographic and Atmospheric
Administration, Sea Grant, and the Department of
Energy. With all this activity, why are there any
uncertainties? We deal in facts, unlike the soft sci-
ences. For example, in the physical sciences, every-
thing has been reduced to equations. Tide tables are
available for dates far in the future—there is no
uncertainty about them. The National Estuarine
Inventory Data Analysis contains detailed information
on the James and other rivers. The chemical facts of
water are undisputed—water contains two atoms of
hydrogen and one of oxygen. The biological sciences
do still have a few areas that are not as determined as
the chemical and physical sciences. Nevertheless,
biologists have made major strides in recent years —
for example, in DNA, RNA, and genetic manipulation,
and the buzzword now in Washington is marine bio-
technology.
Questions do remain. In North Carolina, for
example, there is the question of the striped bass. There
are not many left, and people ask why. There is that
gift from Florida, the red tide. Blame for these prob-
lems has been placed, in some quarters at least, on the
Russians flushing their bilges or on a vicious plot by
developers.
Outside the research laboratories, in the real world,
are we dealing with certainty or uncertainty? Some
things are certain. For example, we can look at an
aerial view of an entrepreneur's effort to make water-
front property available to more of mankind—a canal
system. It is certain that this enterprise will lead to
more people, which will lead to more sewerage.
Inevitably this enterprise leads to sewage pollution.
Must we deliberate about whether an algal bloom will
follow, and then a fish kill? Lawyers and administra-
tors and multitudes of scientists are not necessary to
verify that this sequence of events is a certainty.
Another example is the building of ports. Where
there are ports, there are ships, and where there are
ships, there is oil, which sooner or later spills. The
consequence is the death of wildlife and the deteriora-
tion of marshes.
In the watershed we can look at the "improvement"
of streams, which amounts to straightening, with the
cutting down of streamside vegetation. The runoff
carrries the consequence downstream, where the result
will be the posting of a sign announcing "Closed to
shellfishing."
So if all these things are certain, what are the
uncertainties of estuarine research?
What do we really know? What do we need to know
as we approach the end of the 20th century? The situa-
tion is really one of ignorance in a sea of knowledge.
An example of determining what we really know is a
report drafted by Kirby Smith and me reviewing 4,693
reports on the Newport River system in North Carolina.
These reports represent over 190 collective years of
research at a number of labs, and what they tell us is
summarized in 80 pages of lay language. This docu-
ment should furnish some guidance to decision-makers
about what we know and where we should go from
here.
There has been plenty of deliberation in the past
about where we should be going. Past publications
have outlined what needs to be done next. One thing
we must do next is adopt an interdisciplinary approach.
Interdisciplinary research has arrived for the blue-water
oceanographers, and perhaps one day this will happen
to the shallow-water researchers as well. Are we as
scientists practicing what we are preaching? A useful
document along this line is "A Good Bay Agreement
and Ways to Make it Better." Another is a draft report
for the first year of the Albemarle Pamlico project,
entitled "Citizens Guide to Coastal Resource Manage-
ment."
Another area of uncertainty is this: To what degree
-------
PLENARY SESSION
Keynote Address
are we successful and dedicated in getting citizens
involved in the effort? If we don't know what needs to
be done, it will be hard to persuade citizens. We should
include legislators and local governments in our
thinking. Albemarle-Pamlico, for instance, is sur-
rounded by 20 coastal counties, each of which has a
board of supervisors, usually five men. These are lay
people who earn their living at something besides
marine science, and they may have little understanding
of what these scientific documents mean. (The EPA
Journal may be a little more understandable.)
Implementation of sound management practices is
the largest uncertainty. We can see this problem in a
hypothetical situation, perhaps described as a faraway
place, where, unlike our respective home states, a
myriad of agencies are presumed to be responsible for a
myriad of carefully delineated problems. The scenario
might concern stormwater runoff. The rain goes
downhill, into an estuary. Imagine a large estuarine
system. Over the course of 400 years it has evolved a
tradition of fishing, first commercial and then recrea-
tional. The state has established a mechanism for
identifying water quality status: a rating of SA indicates
fewer than 14 fecal coliforms per 100 cc of water; SB
and SC follow as higher fecal coliform counts indicate
water of progressively poorer quality. Shellfish are
taken only from SA waters. There is an absolute: if the
level goes over 14/100 cc, one agency calls another,
and die water is closed to shellfishing until the count
goes down again. Over 320,000 acres have been closed
over the last 40 years. Motels, shopping malls, and
condominiums have arrived, bringing with them
dollars, profits, and accelerated runoff including fecal
coliforms. Will laws and regulatory agencies take care
of this degradation? The Coastal Improvement Agency
issues permits only for development above high water.
The Environmental Quality Agency is responsible for
water quality at the water molecule level but has no
responsibility for the watershed or its resources.
Another group (the Resource Agency) is responsible for
living resources. Citizens trying to unravel a problem
get very frustrated trying to find someone who has
authority to deal with it. These different agencies rarely
meet, and they report to different cabinet officers. The
citizens must wonder whether the bureaucrats really
care about the environment. Is there a chance, they
wonder, that partisan politics and economic gains may
be more important?
A final uncertainty is long-term, sustained funding
for specific objectives. A new bill, the Chaffee-
Mitchell bill on Marine Research Funding, would
provide $30 million for research in 10 geographic
districts, in environmental issues in coastal waters. We
need to look at what this bill would achieve, and
whether this bill is the best way to do it.
We need a better understanding of what we know
and what we need to know. We need to bring citizens,
legislators, and local government into closer coopera-
tion. We need to be sure that good sound management
practices evolve from this information. And we need to
be sure adequate funds are available for long-term
research; we do not need pork-barrel efforts that will
come back to haunt us.
-------
PHYSICAL PROCESSES
Eric Itsweire, moderator
Changes in Understanding of the Circulation of the Chesapeake Bay
William Boicourt
The Johns Hopkins University
Chesapeake Bay Institute
The Rotunda, Suite 340
711 West 40th Street
Baltimore, Maryland 21211
We have been aware of the primary circulation of
the Chesapeake Bay for about 40 years. The fresh
water is introduced by the rivers and is amplified, so
that the circulation in the Bay is considerably larger
than the flux of the rivers coming into it. The paradigm
is tidal action over a rough bottom, mixing salt and
fresh water and driving the gravitational circulation.
During the last 40 years we have learned about meteor-
ologically forced circulation and topographically
controlled circulation. We are challenged by questions
such as where nutrients, dissolved oxygen, suspended
sediments, and toxics go. There is much we do not
understand about transport and mixing. We thought the
primary mechanism was primary circulation, but now
we realize that something else is also at work, and at a
relatively small scale. Our current knowledge is
inadequate to describe transport processes, as it
principally pertains to large-scale processes.
A number of discomfitting possibilities must be
considered:
• Mixing may be non-uniform in space and time.
• Transport may be dominated by transient, episodic
events.
• The vertical exchange may be localized.
• Topographic controls may dominate.
• The boundary conditions for mathematical model-
ing may defy straightforward formulation.
The variability we observe includes tidal variability
(time span of a tidal period); quarter-wave seiche (1.7
day period); local forcing (4-10 day period); non-local
forcing (4-10 day period), and gravitational circulation.
The meteorological response can be determined if the
effects of local forcing are removed.
Two new tools are available for research. The first
is the bottom-mounted acoustic Doppler profiler, which
can measure the horizontal flow of water very pre-
cisely. For instance, in the Patuxent River, intrusions of
high-salt water across the entrance sill are being studied
with this instrumentation. Results are showing that the
middle portions of the water column are more quies-
cent, and that the role of wind coupled with the quarter-
wave seiche can be quite important, as important as the
diurnal tide. Wind can excite a near-resonance process,
with large-amplitude seiches and strong surges and
anti-surges.
The other new tool now available both to reveal the
physics and to integrate the catalog of processes is
numerical modeling. Process modeling is less expen-
sive than simulation modeling, but it is more difficult to
compare directly to observations. Both kinds of
modeling are needed.
Recommendations for research directions include:
• Long-term (3-5 years) measurements
• More three-dimensional studies of circulation
processes
• Boundary and lateral meeting processes
• Wind stress field (data are woefully inadequate)
• Bottom stress field
• Process modeling
• Simulation modeling
• Model-observation comparisons
• Exchange between the estuary and the shelf
QUESTIONS
Q: Are we missing a component in not studying long-
term flushing? What is its importance compared to that
of the quarter-wave seiche?
A: We have not done quantitative checking on this, but
the outflow of a quarter-wave seiche is lost to the
system, and the inflow is totally new water. So
seiching provides an indication of integrated inflow
from the continental shelf.
Q: Internal waves were not mentioned. What is the
period of the lateral seiche?
A: The process catalog includes longitudinal, lateral,
and mixing processes. The lateral internal seiche time
is 8-20 hours. A higher-frequency lateral seiche may be
important as well. The surface gravity wave seiche is
on the order of minutes.
5
-------
PLENARY SESSION
Opportunities to Improve Understanding of the Circulation of the Chesapeake Bay
Thomas Osborn
The Johns Hopkins University
Chesapeake Bay Institute
The Rotunda, Suite 315
711 West 40th Street
Baltimore, Maryland 21211
We have a need for future research in estuaries, and
we need to consider how to use modern technology.
Research is being driven by a strong need for sensible
management decisions. The Bay must be dealt with as
the large ecosystem that it is. It interacts with itself, its
watershed, and its population. Physical circulation is of
primary importance, and we need to understand its role.
We understand the circulation qualitatively for the
mean flow and for some of the low-frequency fluctua-
tions. Our understanding of short-duration and small-
scale variations is still rudimentary.
The processes responsible for the horizontal and
vertical transport are understood: advection by the
mean flow, and advection by the fluctuating flow terms
such as turbulent diffusion.
Is sufficient fundamental research being done to
support the decision-making process for Chesapeake
Bay restoration and preservation? No, because most
Bay questions, while appearing simple on the surface,
require a detailed understanding for a satisfactory
answer. An example is the "simple" question of how
much nutrient loadings should be reduced. This simple
question generates others: How much and how fast will
it improve the quality of water in the Bay? Does it
matter where the reductions are made? These are very
difficult questions to answer. Thus it is obvious that we
need to do fundamental research on the Bay's physical
processes so that we know what makes the system run.
New opportunities for information are available:
• Remote and large-scale sampling of surface
currents is possible.
• High-resolution Doppler profiles of the current
field and the velocity field can be procured, either
from moored instruments or as transects from
moving vessels.
• Data from airborne and satellite remote sensing are
available (although it must be remembered that use
of satellite data is difficult and time-consuming,
and resolution is low).
• Measurements can be made of small-scale turbu-
lent mixing processes.
For example, in the Delaware Bay NOAA used data
from a radar backscatter system to show a two-dimen-
sional circulation. This kind of monitoring (CODAR)
can and should be done on the Chesapeake, as it
provides answers to long-term questions.
As another example, Doppler profiles show that
circulation in the Chesapeake has three layers: down-
stream at the top and bottom, and upstream in the
middle.
The Bay Program needs a substantial commitment to
sampling as well as modeling the circulation. If
sampling is inadequate we do not know whether all the
processes that are taking place are being reflected in the
model.
QUESTIONS
Q: Is the three-layer circulation ephemeral? What are
the implications for sediment transport?
A: We don't know yet; longer-term studies, over
several days or many tidal cycles, are needed.
Q: What were the velocities on the cross-Bay transect?
A: In November, 10 m/sec.
Q: What about the labor force necessary to perform the
research you outline?
A: The number of people required is not that large for
CODAR and the bottom-mounted acoustic doppler pro-
filer. The money goes into equipment rather than staff.
Q: How about staff for interpreting the results?
A: This does require staffing. Although support has
been available for interpretation in open-ocean re-
search, funding for this activity in estuaries is falling
through a crack between the National Science Founda-
tion and the Environmental Protection Agency. It looks
"too applied" to NSF and "not applied enough" to EPA.
I think we need a serious, peer-reviewed scientific
program running on the order of $1 million a year, and
we need a long-term commitment.
-------
Physical Processes
Stratification Control and Bay-Shelf Exchange: Two Physical Processes and
Their Implications for Ecosystem Dynamics
David M. Goodrich
NOAA Office of Climatic and Atmospheric Research
6010 Executive Blvd. Room 825
Rockville, Maryland 20852
Much of the interest in estuarine research revolves
around ecosystem effects, and it is important to link the
results of physical process research to the ecosystem. It
is axiomatic that the major advances in estuarine
research are interdisciplinary, and it is these interdisci-
plinary results that are of most interest to management.
Two recent examples are the studies of Hurricane
Agnes [Chesapeake Research Consortium, 1976] and
the investigation of anoxia in the Bay in the early
1980's [Officer et al. 1984]. I will present here two
examples of physical processes that have implications
for the living resources, with some suggestions for
future research directions.
SEASONAL STRATIFICATION
The vertical density distribution has a strong influence
on the distribution of nutrients in the water column, and
the degree of vertical mixing has been shown to have a
strong influence on the seasonal succession of phyto-
plankton. The natural variability in stratification also
affects the seasonal development of the anoxic water
mass in the Bay. Seliger and Boggs [1987] have
suggested that streamflow-induced change in stratifica-
tion is the major source of interannual variability in
summer volume of anoxic water. This possibility has
major implications for nutrient control strategies.
Processes controlling stratification
Four major processes influence stratification in the
Chesapeake: runoff, tidal mixing, wind mixing, and the
dynamic balance of heating and cooling effects. In an
annual cycle (Figure 1), salinity generally drops in early
spring, responding to the spring freshet, and reaches a
minimum in June. Stable stratification sets up in May
and persists through the summer. The first major wind
event in the fall brings together salinities in the top and
bottom water, and they are mixed intermittently there-
after [Goodrich et al. 1987]. A large wind event can
leave the water column completely mixed (Figure 2).
The salinity of the lower Bay shows much more
variability than does the upper Bay, as the exchange
with the ocean is a major factor in the salt balance. The
tide is a background source of mixing energy, which is
constant in time but spatially variable; the amplitude in
the upper Bay is half that in the lower Bay. The
condition of stratification is thus a dynamic balance
between the buoyancy inputs (runoff and surface
heating) and mixing energy inputs (tide, wind, and
surface cooling).
Prospects for prediction and hindcasting
The processes controlling stratification are well
understood on a large scale, but more work is needed to
better define the small-scale processes. A large
historical data base exists on the internal behavior of
the estuary and on its boundary forcing (including
current, salinity, and temperature for the former, and
runoff, sea level, and wind for the latter). Large-scale
hydrodynamic modeling is in a relatively advanced
state. For example, the model developed by Blumberg
and Goodrich has simulated the overturn that was
observed in September 1983. Prospects are good for
determining the natural variability in stratification using
a combination of hydrodynamic modeling and historical
observations. This natural variability could then be
removed as a factor in the interannual variations in the
anoxic water mass. Unlike the two previous speakers, I
am not convinced that a large-scale field program is
necessary to solve this particular problem.
BAY-SHELF EXCHANGE AND BLUE CRAB LARVAL
RETENTION
Another example of the impact of physical processes
on the ecosystem is found in the exchange between the
Bay and the shelf, and its effect on the larval population
of blue crabs.
Crab populations are seemingly unaffected by
pollution or fishing pressure, but the natural variability
-------
PLENARY SESSION
MID-BAY
Figure 1. Continuous surface and bottom
salinity from two current meter moorings
in the mid-Bay region (off Patuxent River
entrance) and in the lower Bay (between
the Rappahannock and York rivers). Data
have been lowpass-filtered to remove
variance at tidal frequencies and above.
Data from NOAA Chesapeake Bay
Circulation Survey.
Nov.l Dec.31 Feb.28 Apr.29 Jun.28 Aug.27 Oct.26
1982 1983
in stocks is so great that any anthropogenic signal
would be difficult to discern. Much of this variability
stems from the blue crab's early life history. The larvae
develop offshore, and the return of the postlarvae
(megalopae) to the Bay depends on physical transport.
Current data (Figure 3) indicate that the Eulerian mean
flow field at the Bay entrance was relatively stable in
the two years of observation. Superimposed on this
flow is a strongly variable wind-driven exchange, the
magnitude of which can be accurately estimated from
sea level records at virtually any time (Figure 4).
Three-year daily records of megalopae collected at
the VIMS pier in the York River show an episodic dis-
tribution. Thirteen of the 18 peaks observed over the
three years correspond to positive anomalies in subtidal
volume, including the largest 1985 peak, which
occurred during the massive storm surge associated
with Hurricane Juan (Figure 5). This correspondence
suggests that an important mechanism for the return of
megalopae into the Bay is transport via wind-driven
exchange events. Analysis of the frequency of these
exchange events over a 28-year period indicates that
they reach a fall peak in the last two weeks of Septem-
ber, during which time an average of roughly two
10
E
20
S-2041 S'1962 S'I95I
to)
CHESAPEAKE BAY
SMITH POINT TRANSECT
SAi.INITY.pp1
SEPT 22,1981
2022-2200
® STATION 2 INSTRUMENT
LOCATIONS
S»2Q34 5»B.96 S=B 51 5-1929
10
E.
20
301-
(b)
SEPT 24,1981
1558-1716
DISTANCE.*™
10
DISTANCE, km
20
Figure 2. Lateral salinity sections at Smilh Point (south of Potomac River entrance) before (a) and after (b) the storm of
September 23-24, 1981.
-------
Physical Processes
20 h
• OBSERVED EUIERIAN MEAN CURRENT
X EXTRAPOLATED
U! 23 DAYS BEGINNING 12 MAY 1982
6 10
Km
Figure 3. Lateral current structure at the Chesapeake Bay entrance. Velocities are in cm/sec and are normal to the transect.
Positive velocities are out of the Bay. The 1982 data are from NOAA, the 1980 data from the Chesapeake Bay Institute.
inflow events greater than a tidal prism can be expected
(Figure 6). The 28-year record indicates that an
average of ten events of at least this magnitude can be
expected during the July-November period when
megalopae are present. The number of inflow events
increases in the late summer, reaching a maximum
around the equinox before leveling off. This is signifi-
cant because water temperatures decline rapidly after
the equinox until the megalopae cannot sustain activity.
The main point here is that these transport events are
not fortuitous but rather must be considered a stable
feature of the flow climatology of the Bay entrance. It
seems plausible that the blue crab has evolved a
strategy to take advantage of this feature.
The larvae of a number of other commercially
important species, such as menhaden and croaker, must
also pass through the mouth of the Bay in relatively
non-motile forms. If recruitment studies for these
organisms are to resolve the dominant transport
processes, they should include daily sampling. Weekly
or biweekly sampling cruises are likely to miss the
wind-associated recruitment events that may dominate
the recruitment process.
A climatological approach to physical studies of the
Bay can be very productive. As the above example
suggests, small changes in the long-term physical
behavior of the system can have significant effects on
particular species. A climatological analysis of the
many available long-term records is needed both for
ecological applications and as a benchmark for numeri-
cal model simulations.
-25000-
-50000-
Aug. 18
Sept. 18 Oct. 17
1983
Nov. 16
Figure 4. Volume flux and velocity at the Chesapeake Bay entrance. Solid line is the sublidal volume flux as calculated from
sea level records at Baltimore, Solomons, and Kiptopeke. Dashed line is the filtered principal axis velocity at the Chesapeake
Bay entrance (station 40, surface; sec Figure 3 for location).
-------
PLENARY SESSION
Aug.1
CO
Aug.31
Sept.30
Oct.30
Nov.29
Figure 5. Mean number of megalopae per trap at VIMS pier (solid) and Chesapeake Bay subtidal volume (dashed), 1985-87.
Zero on subtidal volume axis represents series mean. Arrow indicates passage of Hurricane Juan in November 1985.
s.
o
2.0 -i
1.0 -
0.5 -
0.0
7/1 8/1 9/1 10/1 11/1
Two Week Period Beginning
r30
£
-20 -3
I
•10
Figure 6. Frequency of inflow events
greater than one tidal prism (solid), 1955-
86, and monthly mean water temperature
at the Chesapeake Bay Bridge Tunnel
(dashed) [Dowgiallo et al. 1984].
-------
Physical Processes
QUESTIONS
Q: It should be noted that there is a contradiction
between your assessment of the suitability of the
longer-term data bases and the assessment that Boicourt
has made.
REFERENCES
Blumberg, A., D. Goodrich. Modeling of wind-induced
destratification in Chesapeake Bay. Submitted to J.
Geophys. Res.
Chesapeake Research Consortium, 1976. The effects of trop-
ical storm Agnes on the Chesapeake Bay estuarine system.
Johns Hopkins University Press, Baltimore, MD. 639 pp.
Dowgiallo, M., M. Predoehl, S. Green, R. Dennis, 1984.
Marine environmental assessment, Chesapeake Bay, 1983
annual summary. NOAA National Environmental
Satellite, Data and Information Service, Washington, D. C.
81pp.
Goodrich, D., W. Boicourt, P. Hamilton, D. Pritchard, 1987.
Wind-induced destratification in Chesapeake Bay. J.
Phys. Oceanogr. 17:2232-2240.
Officer, C., R. Biggs, J. Taft, L. Cronin, M. Tyler, W.
Boynton, 1984. Chesapeake Bay anoxia: origin, develop-
ment and significance. Science 223:22-27.
Seliger, H., J. Boggs, 1987. Anoxia in the Chesapeake Bay
(abstract). EOS Trans. Am. Geophys. Union 68(30),
1692 pp.
11
-------
BIOAVAILABILJTY OF TOXICS
James G. Sanders, moderator
Chemical and Physical Processes Influencing Bioavailability of Toxics in Estuaries:
An Overview
Within the Chesapeake Bay region, pressure from
increased population continues to mount. Population
increases bring increases in water use, shipping,
recreation, the location of industries on the water—and
it is inevitable that the Chesapeake Bay will be sub-
jected to continued loading of toxic substances.
We as a regional society must address the potential
impact of contaminants in the Chesapeake Bay. To
predict and assess impact, however, we must first
understand the processes that control contaminant
toxicity and availability. This session is an attempt to
present the current view of our knowledge in this area
—how contaminants move through the estuary and the
coastal zone, and what abiotic and biotic processes con-
trol contaminant movement, fate, and bioavailability.
This session will be composed of two speakers; one
will generally address inorganic contaminants; the other
will address organic substances. The focus will be not
only on what we know, but also on what we must still
discover. In that regard, I offer the following recom-
mendations for further research. I ask that you keep
them in mind as you listen to the following speakers.
• Determine how partitioning affects the toxicity of
model compounds. Is complexation a detoxifica-
tion mechanism? Are paniculate-associated
contaminants still toxic? How important is
concentration in the surface microlayer?
> Understand the role of biota in the transfer of
pollutants. Is transport across the sediment/water
interface an important process? What about fecal
pellet production? What is the fate of contami-
nants associated with biological tissues?
1 Examine the importance of anoxia on transport and
availability. Is availability increased or decreased?
Is the reservoir capacity of the sediments increased
or decreased?
• Evaluate the importance of sediments as a control-
ling mechanism. Are they a sink or source, and for
which contaminants? What controls the flux of
carbon and contaminants across the sediment/water
interface?
• Understand the role of communities in the transport
and transformation of contaminants. Can we move
away from laboratory experiments with single
species? Can we move toward natural community
manipulations and microcosm approaches?
—J.G.S.
13
-------
PLENARY SESSION
Factors Affecting the Bioavailability of Toxic Trace Elements to Estuarine Organisms
Gerhardt F. Riedel and James G. Sanders
The Academy of Natural Sciences
Benedict Estuarine Research Laboratory
Benedict, Maryland 20612
INTRODUCTION
The many factors altering the bioavailability of toxic
trace elements may be considered in two major group-
ings. One group consists of processes that somehow
change the availability of the contaminant present, but
do not alter its total amount or basic chemical form.
This group of factors includes organic and inorganic
complexation, ionic strength, pH, redo* reactions, and
competition with similar compounds. Tnese factors are
rather general in nature, applying to large groups of
organisms and large groups of elements in a wide
variety of environments.
The other major group of factors controlling
bioavailability comprises processes that control the
concentration, distribution, or chemical form of the
contaminant within the system. Factors in this group
include sources and sinks, such as adsorption, floccula-
tion, sedimentation, and remobilization, redox reac-
tions, and the formation of organometallic compounds.
While these factors do operate in most environments,
their relative importance in each environment with
respect to a particular element is highly variable, and to
adequately quantify their effects requires study of a
specific site and element.
It is also important to note that organisms are not
merely affected by the trace elements in question; their
activities in the environment also help determine in
large measure the factors that control bioavailability of
trace elements.
GENERAL ROUTES OF TRACE ELEMENT UPTAKE
In discussing the bioavailability of trace elements to
organisms, it is useful to consider groups of compounds
that have similar patterns of availability. A focus of
recent research has been the effect of chemical form, or
speciation, of trace elements on their availability to
aquatic biota. There are three groups of contaminants
for which generalizations concerning bioavailability are
available: cationic trace elements, anionic trace
elements, and organic compounds (Figure 1).
For cationic elements, free ion activity appears to be
the major factor determining bioavailability [Sunda and
Guillard, 1976; Anderson and Morel 1978; Sunda et al.
1978; Engel and Sunda 1979; Rueter and Morel 1981;
Anderson and Morel 1982; Harrison and Morel 1983;
Morel and Morel-Laurens 1983; Zamuda 1984; Wright
and Zamuda 1987]. For example, the uptake of copper
by oysters is largely controlled by the copper free ion
concentration of the medium, and to a lesser extent by
the copper content of the food [Zamuda 1984]. How-
ever, other chemical forms may be available. For
example, an uncharged complex may provide a second
route of access to the cell. For the grass shrimp
(Palaemonetes pugio) [Engel et al. 1981] and several
species of phytoplankton [Sanders and Abbe 1987a,b],
it appears that the calculated activity of the uncharged
species AgCl better explains the availability and
toxicity of silver than does the calculated free silver ion
activity.
For anionic trace elements (e.g., chromate, mo-
lybdate, selenate, arsenate, germanate), bioavailability
is most often controlled by competition for uptake with
a similar nutrient ion of greater abundance (i.e., sulfate,
phosphate, or silicate). For those elements whose
availability varies inversely with sulfate (chromate
[Riedel 1984, 1985a], selenate [Schrift 1974; Wheeler
et al. 1982], molybdate [Howarth and Cole 1985; Cole
et al. 1986]), salinity plays a dominant role in bioavaila-
bility. This is shown in Figure 2 for the effect of salinity
and sulfate on chromate uptake by phytoplankton
[Riedel 1984]. Arsenate behaves as a phosphate analog
[Blum 1966; Planas andHealy 1978; Sanders 1979],
and germanate behaves as an analog of silicate [Azam
and Volcani 1981], so the availability of these metals
varies according to the concentration of their analog,
which in turn is highly dependent on biological activity.
A third group of trace elements are those which form
lipophilic organo-metallic compounds, such as methyl
14
-------
Bioavailability of Toxics
ANIONS: COMPETITION
FOR TRANSPORT SITES
CrO4>- S04»-
Figure 1. Dominant routes of uptake for three
important trace element groups.
CuOH*
CuCO,—* Cuji
^
CATIONS:
TRANSPORT
OF FREE IONS
=0=
CrO4J-
INQESTION
OF PARTICLES
CONTAINING
CONTAMINANTS
(CH5),Hg
LIPOPHILIC METALLO-ORGANICS:
SOLUBLE IN LIPIO DILAYERS
and dimethyl mercury, methyl and ethyl lead species,
and methyl and butyl tin. There are few direct compari-
sons of the bioavailability of alkyl compounds and their
inorganic precursors, but in general it appears that their
uptake and toxicity is enhanced by their greater lipid
solubility. For example, in a study of the uptake of
inorganic and methylmercury by the diatom Skele-
tonema costatum and the copepodAcarfw clausi,
methylmercury resulted in much greater uptake of total
mercury than inorganic mercury [Fujiki 1980]. For
inorganic and organo-tin compounds, there is a general
correlation between the octanol/water partition coeffi-
cient (PyJ, an index of the lipophilicity of the com-
pound, and the toxicity to phytoplankton [Wong et al.
1982] and to animals [Zuckerman et al. 1978].
UPTAKE AND INCORPORATION
The ability of marine organisms to accumulate trace
elements has been well documented; the literature is
replete with studies of pollutant levels in organisms
from the Chesapeake Bay [Young et al. 1980; Abbe
1982; Bieri et al. 1982; Di Guilio and Scanlon 1982;
Hung 1982; Phelps et al. 1985; Abbe and Sanders 1986;
Shigenaka and Calder 1987; Wright and Foster 1987].
Less well known, and currently under study, is the
relative role of various uptake processes. For those
organisms (primarily animals) that can live in a medium
containing a dissolved toxic element, contact sediment
with high concentrations of the element, and ingest
other organisms that have already incorporated the
element, the logical query is which source contributes
most to the uptake of the toxic. The answer, of course,
varies for different organisms and pollutants; however,
in general, water appears to be the most important
source of most toxics to the greatest number of organ-
o
o
o
QJ
o
OJ
o
0
O>
o
-15.0n
-16.0-
-17.0-
-18.0-
-8.0 -7.0 -6.0 -5.0 -4.0
-15.0. Log Cr(VI) (M)
-16.0-
-17.0-
-18.0-
B
-4.0 -3.0 -2.0 -1.0 0.0
Log Cr(Vl)/SO 2" (M)
Figure 2. Uptake of chromate by phytoplankton at several
combinations of salinity and sulfate concentration. (A) Uptake
as a function of total chromate in medium. (B) Uptake as a
function of the ration of sulfate to chromate in the medium
[fromRiedell985a].
isms. Much of the metal present in food is not passed
across the gut wall. Generally speaking, uptake of
contaminants from food is more important for larger
animals, for which surface absorption is less important.
Animals have been shown to take up large amounts of
75
-------
PLENARY SESSION
metals from contaminated sediments, but it is usually
not clear whether the materials have been ingested or
absorbed directly from interstitial water.
The uptake of silver by oysters from Chesapeake
Bay is solely from silver dissolved in water. Uptake
from either algal food or sediment is insignificant in
comparison [Sanders and Abbe 1986; Sanders and
Abbe 1987c]. However, copper uptake by the same
organism is more complex. Uptake of dissolved copper
occurs readily, largely controlled by the availability of
the free copper ion. Copper is also taken up from food,
although to a lesser extent. It appears that colloidal
organo-cupric complexes are also available [Zamuda
and Sunda 1982; Zamuda 1984; Zamuda et al. 1985].
Through the predator/prey relationship, the potential
exists for toxic trace elements to be passed up the food
chain from autotrophs to herbivores to carnivores. In
past years, such food chain transfer was thought to lead
to increased body burdens at each step in the chain,
resulting in top carnivores with extremely high concen-
trations of toxic metals, toxic responses within the
population, and even potential toxicity to human
consumers. Recent work, however, has shown that
most trace elements do not get magnified as they are
passed up the food chain [Young 1984]. Biomagnifica-
OISSOLVED
METAL ION
RELEASED
ORQANOMETAL
(A*. Hfl>
RELEASED
REDUCED METAL
(As)
PHOTOREDUCTION
VIA EXCRETED ORGANICS
Dpii (Cr. Cu, Mn, F«)
COMPLEXATION
. BY EXCRETED ORGANICS
(Fe, Cu)
Figure 3. Possible interactions of dissolved metal ions with
phytoplankton, including changes in partitioning (A), direct
transformation of chemical speciation (B), and indirect
facilitation of complexation and photoreduclion (C).
INCORPORATION
AND SEDIMENTATION
(Cr)
C.
LIGHT
DISSOLVED
METAL ION
tion through food appears to be a significant factor for
very few trace elements; methylmercury [Fowler 1982]
appears to be a notable exception.
Aside from the concentration and sequestering of
trace elements, organisms have the capability to alter
the chemical form or partitioning of many trace
elements. Such shifts alter biological reactivity and
toxicity of these elements and can alter their rate of
transport through the estuary as well as their eventual
fate. Potential metal/phytoplankton transformations are
illustrated in Figure 3.
FACTORS THAT ALTER UNAVAILABILITY OF
TRACE ELEMENTS AT FIXED CONCENTRATIONS
Ionic strength
In concentrated salt solutions, the activity of an ion is
reduced because of the interaction of its charge with
other nearby ions. In general, the individual ion activity
coefficient ranges from 1.1 for uncharged species to
approximately 0.1 for triply charged species in full-
strength seawater [Stumm and Morgan 1970]. In
practice, changes in bioavailability due to ionic strength
alone must be considered in conjunction with changes
due to inorganic and organic complexation.
Inorganic complexation
A number of ions are present in seawater in almost
constant ratio to one another. Although these ions are
not particularly strong complex formers, their concen-
trations in seawater are high enough that the most
predominant chemical species of most trace elements
are ion pairs or other complexes with these ions. Other
ligands (e.g., sulfide) are of biological origin and have
wide variation independent of salinity. Under some
circumstances, however, they may be important in
determining the chemical speciation of inorganic
pollutants.
Inorganic complexation of the various trace elements
differs greatly. For example, the inorganic speciation
of copper is dominated by complexation by hydroxide
and carbonate species; it is only slightly complexed by
the very abundant chloride ion [Sunda 1975]. Silver
and cadmium, however, have comparatively high
binding constants for chloride, and their inorganic
speciation is dominated by chloride complexes [Engel
et al. 1981; Jenne et al. 1978; Sanders and Abbe
1987b]. The binding of various ions is a highly
interdependent process since they compete for common
ligands; however, the majority of the important
equilibrium constants are well enough known that
computer models can readily estimate the binding by
inorganic ligands.
16
-------
Bioavailability of Toxics
Organic complexation
In addition to the inorganic ligands present in aquatic
systems, organisms also produce, either through
excretion, leakage or decay, a variety of organic
compounds that may also complex and reduce the free
ion activity of trace elements. Dissolved organic
compounds (DOC) range from 0.1 mg 1"' in unpolluted
freshwater to 1-2 mg 1-1 in seawater to 10 mg 1"' and
higher in highly productive or polluted water. Concen-
trations in the Chesapeake Bay range from 1.9 to 13 mg
1 -1 [Sanders 1982; Sigleo and Helz 1982; Newell 1983;
Zamuda 1984; Newell and Sanders 1986]. The
composition of the organic matter varies depending on
the source, but it contains many functional groups with
metal-complexing capabilities. Unlike the situation with
inorganic complexing ions, natural organics contain a
large number of compounds of varying complexing
strengths [Dzombak et al. 1986; Fish et al. 1986].
Therefore, except for a few studies on specific chelators
found to be excreted by algae or released into the
environment by human activity, the studies of com-
plexation by natural DOC are largely empirical.
Only a few metals (iron, copper, zinc, and mercury,
which have the greatest affinity for ligands) are
complexed significantly by natural DOC. Copper is
organically complexed to a large extent (50-98%)
[Hanson and Quinn 1983; Zuelke and Kester 1983;
Mills and Quinn 1984; Sunda and Hanson 1987], by a
variety of dissolved organic compounds, some of which
are relatively labile, and some of which are kinetically
inert. In fresh water, where the sources of DOC are
highly variable, there is little correlation between DOC
and copper complexation capacity between different
sites; however, in marine and estuarine waters, includ-
ing the Chesapeake Bay, where the DOC is largely of
autochthonous origin and is more homogeneous on
spatial and temporal scales, there is a significant
correlation between DOC concentration and copper
complexation capacity [Newell and Sanders 1986; van
den Berg etal. 1987].
pH
Many of the inorganic compounds important in the
complexation of some toxic metals (e.g. carbonate,
phosphate, and sulfate), as well as many of the pollut-
ants, have equilibria dependent on pH within the pH
range of aquatic systems [Zirino and Yamamoto 1972].
Control of the relative abundance of such species
through these equilibria is perhaps the most direct way
in which pH determines the bioavailability of pollut-
ants. In addition, the active sites of most organic
complexing agents are pH-sensitive groups (e.g.,
carboxylic acids, amines, and sulfhydryl residues), so
that pH has a strong influence on the extent to which
organics complex toxic trace elements. Finally, the
active sites of enzymes and uptake sites of trace
elements contain the same variety of pH-sensitive
groups, and the activity of these systems varies with pH
as well. Thus, the effect of pH on the availability of a
given element to a particular organism in a specific site
depends on a variety of effects. Some of these effects
(e.g., the effect of pH on inorganic complexation) are
rather straightforward to predict, whereas others (the
effect of pH on organic complexation and on uptake of
the metal) are more difficult.
Algal photosynthesis results in the removal of
inorganic carbon from the water column and a resulting
increase in pH; during bloom events, such increases can
be substantial. In the freshwater section of the Potomac
River during 1984, a persistent bloom of Microcystis sp.
caused substantial variability in pH, with maximum pH
of >10 [Seitzinger 1986]. Such a large shift in pH can
greatly affect the bioavailability of trace elements.
PROCESSES AFFECTING TRACE ELEMENT CON-
CENTRATION, DISTRIBUTION, AND FORM
In the previous section we have discussed a number of
factors that can affect the bioavailability of a trace
element to estuarine organisms at a fixed total concen-
tration. There are also factors that control the total
concentration, distribution and form of the element to
which organisms are exposed in the Bay. These are
discussed below, and are diagrammed in Figure 4.
Sources
There are several sources of toxic trace elements to the
Chesapeake Bay (Table 1). One of the most important
is runoff of fresh water from the land. Industry,
agriculture, and municipal effluents each contributes a
significant load of compounds. Many toxic trace
elements are present in the air as suspended particles
(e.g., copper and zinc) or as vapors (e.g., mercury, lead,
selenium, and arsenic) and can be deposited by rain or
through dry fall. Aeolian flux of some metals has been
estimated for Chesapeake Bay, and although the
loadings are not large compared to runoff, they are
significant. Aeolian inputs may be a particularly
significant source to the surface microlayer where a
variety of toxics are concentrated, and to areas remote
from industry or other concentrated sources.
Distribution
Estuaries are extremely dynamic systems, moving and
changing constantly in response to winds, tides, and
runoff, so it is difficult to discuss the distribution of a
given trace element in the water and sediment except on
a statistical basis. Nevertheless, some general trends in
17
-------
PLENARY SESSION
SURFACE MICROLAYER
RIVER_
INPUT"
SEAWARD
ADVECTION
-»• ADSORPTION »- DESORPTION
FLOCCULATION
REMOBILIZATION
-1
SEDIMENTATION
-—.—.
Figure 4. Processes affecting transport and availability of trace elements in estuaries.
LANDWARD
ADVECTION
trace element distributions can be discussed. Elements
whose sources are in fresh water are inversely corre-
lated to salinity; that is, mean concentrations decrease
going from the head of the Bay to the ocean, and from
tributaries to the main Bay. This pattern has been
observed for metals such as cadmium, copper, and lead
[Kingston et al. 1982]. Conversely, for those few trace
elements whose concentrations are normally higher in
seawater than in fresh water, a reverse trend is ob-
served, with bottom water and more saline waters
having higher concentrations [Kingston et al. 1982;
Sanders 1985]. Elements that are reduced and solubil-
ized in sediments (e.g., iron, manganese, arsenic, and
zinc) also tend to be enriched in deep waters of the Bay
[Carpenter et al. 1975; Kingston et al. 1982; Troup and
Bricker 1975; Sanders 1985].
The phenomenon of seasonal anoxia in the Chesap-
eake Bay is undoubtedly a dominant factor in the
spatial and temporal distributions of redox-reactive
elements such as arsenic, chromium, iron, manganese,
and selenium. Unfortunately, virtually no published
studies on this aspect of the anoxia are available yet.
This should be an active area of research in the near
future.
The incorporation of a trace element by biota may
alter its transport within the estuary. In general,
because organisms try to maintain position within a
specific environment, incorporation within tissues will
lead to the contaminant remaining within the estuary.
For example, silver has a high affinity for particle
surfaces and is rapidly taken up and incorporated by
phytoplankton as well as by suspended sediments. As
these particles move down the estuary, the silver is
desorbed from the suspended particles and remains in a
dissolved state, complexed with chloride. Silver
associated with phytoplankton, however, does not
desorb, but remains in the Chesapeake Bay and
probably recycles [Sanders and Abbe 1987a,b].
Surface films
Trace elements can become concentrated in two areas
within water bodies: the sediment and the surface
microlayer. The surface microlayer is the upper 50 \iun
to 1 mm surface film that lies between the water and the
atmosphere. Many organic compounds and contami-
nants concentrate there, due to their hydrophobic nature
[Baier et al. 1974; Hardy 1982; Hardy et al. 1986]. The
film also contains a variety of particles, including a
unique fauna and flora [Hardy et al. 1986]. In addition,
the presence of natural organics ensures the presence of
trace contaminants that normally bind to organic com-
plexing agents [Garrett and Duce 1980]. Recent studies
Table 1. Estimated sources of toxic trace elements, in metric tons per year, to the Chesapeake Bay. Data taken
from Bieri et al. [1982].
Source
Cd
Cr
Cu
Pb
Zn
Fe
Mn
Major tributaries
Industry
Municipal wastewater
Urban runoff
Shore erosion
Wet fall
75
178
6
7
1
3
551
200
200
10
83
ND
517
190
99
9
29
28
402
ND
ND
20
ND
25
307
155
68
111
28
34
1,444
167
284
63
3
825
199,683
2,006
625
977
57,200
87
19,000
ND
ND
22
ND
5
18
-------
Bioavailability of Toxics
in the Chesapeake Bay have found high concentrations
of many pollutants, including metals, in the surface
microlayer [Bellama andZoller 1983; Gucinski 1986;
Hall et al. 1986; Hardy et al. 1987]. Because of the
high concentrations and the potential for direct uptake
by biota, this layer may represent an efficient and
important transfer mechanism for pollutants. If so, this
mechanism may be particularly important for periodi-
cally emergent organisms, such as sedentary animals in
the tidal zone, or emergent vegetation or neuston,
including commercially important fish eggs.
Adsorption
Adsorption reactions may be the important controlling
factors for the distribution of many toxic trace metals in
estuaries, including copper, lead, and zinc [Harris et al.
1975]. These reactions are partly reversible, so that the
material can be desorbed from solids by lowering the
concentration of the pollutant in solution, or by chang-
ing the chemical system to favor the release (e.g., a
change in pH, or an increase in competing ions).
Flocculation
Suspended solids and colloids carrying adsorbed trace
elements can aggregate to form larger particles, which
become more susceptible to the processes of sedimenta-
tion and filtration. An important component of the
suspended solids in most river systems is clay particles.
Clay particles aggregate quite slowly in fresh water but
readily in seawater [Stumm and Morgan 1970]. Floc-
culation can lead to segregation and concentration of
trace metals. In studies of metal interactions with Che-
sapeake Bay floes, most metals were concentrated in
the smaller size fractions [Gibbs 1982, 1986].
Sedimentation
The sediments are the ultimate repository of most
particles, including large amounts of toxic trace
elements. In particular, the sediments of the turbidity
maximum zone downstream of a major source are often
a major repository. Within the Chesapeake Bay,
Officer et al. [1984] andHelz et al. [1985] determined
that most of the paniculate material and associated
toxics entering from the Susquehanna River is depos-
ited at the head of the Bay.
The combined processes of adsorption, flocculation,
and sedimentation are no doubt largely responsible for
the high concentration of toxics found in the sediments
near sources of pollution [Huggett et al. 1971; Pheiffer
1972; Owens et al. 1974; Villa and Johnson 1974;
Goldberg et al. 1978; Sinex and Helz 1981; Wong and
Moy 1984; Di Guilio and Scanlon 1985; Wright and
Foster 1987]. For example, the concentrations of
arsenic, copper, manganese, nickel, lead, tin, and zinc
in Chesapeake Bay sediments decline seaward from
maxima at Baltimore Harbor, Susquehanna River, and
Elizabeth River [Helz and Huggett 1987].
Organisms within the water column and on the
bottom produce aggregated fecal pellets containing
unassimilated organic matter and inadvertently ingested
sediment particles. Studies of zinc have indicated that
rapid recycling between the sediment and water column
takes place in the northern Chesapeake Bay. This cycle
has been attributed to uptake by phytoplankton,
ingestion by zooplankton, and rapid return to the
sediments in fecal pellets [Carpenter et al. 1975].
Remobilization
Incorporation into sediments and subsequent burial
does not forever remove trace elements from possible
uptake by organisms. Many organisms reside in the
sediments, and are thus exposed to the toxic material.
Toxic substances may be present in the pore waters of
sediment in higher concentration than the water
column. For example, arsenic present in water as
arsenate can be adsorbed onto surface sediments.
When buried in reducing sediments the arsenic is
reduced to arsenite, which is present in much higher
concentration in the interstitial waters than in the water
column. In the Patuxent River, the arsenite concentra-
tion in the pore waters is about 50 times higher than
arsenate in the surface waters [Riedel et al. 1987].
Other metals coprecipitated with iron and manganese
oxyhydroxides, such as copper, zinc, and lead, can be
remobilized in pore waters, particularly in the northern
Chesapeake Bay, where low sulfate concentrations
result in less sulfide formation under reducing condi-
tions [Carpenter et al. 1975; Troup and Bricker 1975].
Cadmium also may be remobilized from Chesapeake
Bay sediments [Helz et al. 1975].
Trace elements in sediment pore waters can also be
returned to the water column by diffusion, resuspension
of the sediments, or the activities of benthic organisms.
Benthic infauna alter the transport and availability of
pollutants in sediments in several ways. Organisms
may change the distributions of contaminants in sedi-
ments in a variety of ways, including mechanical dis-
turbance (mixing, pelletization, and sorting), chemical
changes (increased oxygen penetration or organic en-
richment), increase of microbial activity, ingestion of
sediment constituents, increase of surface area, and
direct uptake of metals [Rice and Whitlow 1985].
Burrowing activities of Nereis succinea in contami-
nated sediments, for example, increased the flux of
arsenic by a factor of 5 [Riedel et al. 1987]. This
enhancement corresponded to an approximately equal
increase in the surface area of the sediment due to
worm burrows (Figure 5).
19
-------
PLENARY SESSION
Redox reactions and equilibria
Several of the toxic trace elements participate in redox
reactions that alter their biological availability. Most of
the important redox reactions of trace elements are
relatively slow compared with inorganic and organic
complexation reactions.
In the Chesapeake Bay, organic carbon from the
highly productive surface waters fuels a strong oxygen
demand in the sediments and bottom waters. This
demand produces anoxic, sulfide-rich interstitial water
and seasonally anoxic bottom waters in the deep
channel of the Bay. In such waters a variety of metals
(arsenic, chromium, iron, manganese, and selenium)
can be reduced to forms different from the oxidized
form commonly present. Arsenic, for example, is most
stable in oxidized waters as the arsenate ion. However,
in anoxic bottom and interstitial waters, arsenate is
largely reduced to arsenite [Peterson and Carpenter
1983; Sanders 1985; Riedel et al. 1987]. Arsenite in the
sediments and bottom waters can diffuse or be mixed
up into the surface waters, where it slowly oxidizes
back to arsenate. Arsenate is primarily available and
toxic to phytoplankton, which have a requirement for
phosphate, whereas arsenite is more toxic to fauna
[Peoples 1975].
Indirect evidence suggests that sulfide is present in
oxidized seawater and in the Chesapeake Bay at
concentrations of 10 ~10 to 10 "n M [G. Cutter, personal
communication; Elliot et al. 1985], enough to have
significant effects on the speciation of trace metals such
as silver and mercury. However, standard analytical
techniques are not sensitive enough to measure sulfide
at those concentrations. Moreover, in the Chesapeake
Bay, where there is occasional anoxia and a constant
source of sulfide diffusing from the sediments, concen-
trations are likely to be more variable.
37
LEACHABLE BOUND As
38,000
Another source of reduced metal ions is photoreduc-
tion. The interaction of light, or ultraviolet radiation,
on DOC or on metal organic complexes can either
directly reduce certain metal ions or produce redox
active compounds such as peroxide or superoxide that
can reduce metals. It has been shown that peroxide
concentrations in the Patuxent River undergo a diurnal
cycle, with concentrations increasing in the daylight
hours and decreasing at night [Kieber and Helz 1986].
Peroxide, in turn, has been shown to reduce chlorine
from industrial sources to chloride [Helz and Kieber
1985], and reduce some copper(II) to copper(I) [Moffet
and Zika 1983]. Chromate is reduced photochemically
in the presence of high DOC levels [Riedel 1985b; G.
Helz and R. Kieber, personal communication]. Since
photochemical reactions are favored by the presence of
high DOC and abundant light, it is likely to be most
important in shallow, organic-rich estuarine systems.
Organo-metallic compounds
Several metals (antimony, arsenic, lead, mercury,
selenium, and tin) are present in the Chesapeake Bay
partly as covalently bound organo-metallic compounds
(e.g., methylarsonic and dimethylarsinic acids [Sanders
1985, 1986; Riedel et al. 1987], alkyl lead, methylmer-
cury [Zoller et al. 1983], organic selenium [Takayanagi
and Wong 1984], and methyl and butyltin compounds
[Hallas and Cooney 1981; Hallas et al. 1982; Brinck-
man et al. 1983; Gilmour et al. 1985; Hall et al. 1986;
Westbrook et al. 1986]). In some cases (methylarsenic
compounds, methylmercury, organic selenium, and
methyltins), these compounds are formed in the
environment from inorganic metal by the local biota.
Metals can also be methylated by exchange reactions
with other methylated compounds [Manders et al. 1984;
Bellama et al. 1985; Brinckman et al. 1985].
LEACHABLE BOUND As
38.000
>
7
Nereis
12
Figure 5. Cycling of arsenic in undisturbed and bioturbated estuarine sediment contaminated with arsenic [from Riedel et al.
1987].
20
-------
Bioavailability of Toxics
For example, biological uptake of arsenic leads to
the production of reduced and methylated arsenic
compounds, some of which are more toxic to higher
trophic levels than was the original arsenic compound
[Peoples 1975; Sanders 1980; Sanders 1985]. Within
the Chesapeake Bay, large fractions (up to 80%) of the
arsenic may be present in these reduced or methylated
forms (Figure 6).
SUMMARY AND RESEARCH DIRECTIONS
The overall conclusion must be that the complexity of
the interactions between physical, chemical, and
biological factors is extreme. This complexity hampers
our ability to fully understand (and more important,
predict) pollutant transport, availability, uptake, and
impact.
Past efforts have focused largely upon determining
the concentration of contaminants within the various
compartments of the system. Although this research
has value, we must move more toward careful study of
the processes themselves.
ACKNOWLEDGEMENTS
The preparation of this article was supported by the
Academy of Natural Sciences.
QUESTIONS
Q: Did you measure oxic vs. anoxic events in your
bioturbation studies?
A: No, these were strictly laboratory experiments; but
the sediments were mostly anoxic. The worms made
themselves an oxic area, and it may be that this increase
of surface area exposed to oxygen accounts for the
increase in flux.
T.O -i
o
x
o>
o
S.O -
1.0 -
500 -1 B
400 -
300 -
200 -
100 -
n -
1
1
\
\ "-
\ S '»
\ .!(
V'l
i
\
\
\
\
\
\
1
\
\
\
\
\
1
1
» \
1 'I
• ;'V
v \
\
\
N
10
IS 20
30
SALINITY. Voo
Figure 6. Correlation between species composition of
phytoplankton in Chesapeake Bay with methylated arsenic
[from Sanders 1985].
RELEVANT LITERATURE
Abbe, G.R. 1982. Growth, mortality and copper-nickel
accumulation by oysters, Crassostrea virginica, at the
Morgantown steam electric station on the Potomac River
Maryland. J. Shellfish Res. 2: 3-13.
Abbe, G.R. and J.G. Sanders. 1986. Condenser replacement in
a coastal power plant: copper uptake and incorporation in
the American oyster, Crassostrea virginica. Mar.
Environ. Res. 19: 93-113.
Anderson, D.M. and F.M.M. Morel. 1978. Copper sensitivity
of Gonyaulax tamarensis. Limnol. Oceanogr. 23:283-295.
Anderson, M.A. and F.M.M. Morel. 1982. The influence of
aqueous iron chemistry on the uptake of iron by the coastal
diatom, Thalassiosira \veissflogii. Limnol. Oceanogr
27:789-813.
Azam, F. and B.E. Volcani. 1981. Germanium-silicon
interactions in biological systems, p. 43-67. In: T.L.
Simpson and B. E. Volcani (eds.), Silicon and siliceous
structures in biological systems. Springer-Verlag.
Baier, R.E., DAY. Goupil, S. Perlmutter and R. King. 1974.
Dominant chemical composition in sea-surface films,
natural slicks, and foams. J. Rech. Atmos. 8:571-600.
Bellama, J.M., and W.H. Zoller. 1983. An investigation of
the character of and constituents of slicks on water
surfaces near power plants in the Chesapeake Bay. Report
to MD Power Plant Siting Program, PPRP -83.
Bellama, J.M., K.L. Jewett and J.D. Nies. 1985. Rates of
methylation of mercury(II) species in water by organotin
and organosilicon compounds, p.239-247. In: A. E.
21
-------
PLENARY SESSION
Martell and K. J. Irgolic (eds.), Environmental inorganic
chemistry. Verlag Chemie.
Bieri, R., O. Bricker, R. Byrne, R. Diaz, G. Helz, J. HiU, R.
Huggett, R. Kerhin, M. Nichols, E. Reinharz, L. Schaffner,
D. Wilding and C. Strobel. 1982. Toxic substances, p.
263-375. In: E.G. Macalaster, D.A. Barker and M.
Kasper (eds.), Chesapeake Bay Program technical studies:
a synthesis. EPA, Washington, D. C.
Blum, J.J. 1966. Phosphate uptake by phosphate starved
Euglena. J. Gen. Physiol. 49:1125-1136.
Blust, R., E. Verheyen, C. Doumen, and W. Decleir. 1986.
Effects of complexation by organic ligands on the bi-
oavailability of copper to the brine shrimp, Anemia sp. Aq.
Toxicol. 8:211-221.
Brinckman, F.E., G.J. Olson and J.S. Thayer. 1985. Biological
mediation of marine metal cycles, p.239-258. In: A. Sigleo
and A. Hattori (eds.), Marine and estuarine geochemistry,
Lewis Publ.
Brinckman, F.E., J.A. Jackson, W.R. Blair, G.J. Olson and W.
P. Iverson. 1983. Ultratrace speciation and biogenesis of
methyltin transport in estuarine waters, p. 39-72. In: C.S.
Wong, E. Boyle, K.W. Bruland, J.D. Burton, and E.D.
Goldberg (eds.), Trace metals in sea water. Plenum.
Carpenter, J.H., W.L. Bradford and V. Grant. 1975. Processes
affecting the composition of estuarine waters (HCO3, Fe,
Mn, Zn, Cu, Ni, Cr, Co, and Cd), p. 188-214. In: L.E.
Cronin (ed.), Estuarine Research, Vol. 1. Academic Press.
Chau, Y.K. and P.T.S. Wong. 1978. Occurrence of biological
methylation of elements in the environment, p. 39-53. In:
F.E. Brinckman and J.M. Bellama (eds.), Organometals
and organometalloids ocurrence and fate in the environ-
ment. ACS Symposium series 82.
Church, T.M., J.M. Tramontane, J.R. Scudlark, T.D. Jickells,
J.J. Tokos, Jr., A.M. Knap, and J.N. Galloway. 1984. The
wet deposition of trace metals to the western Atlantic
Ocean at the mid-Atlantic coast and on Bermuda. Atmos.
Environ. 18:2657-2664.
Cole, J.J., R.W. Howarth, S.S. Nolan and R. Marino. 1986.
Sulfate inhibition of molybdate assimilation by planktonic
algae and bacteria: some implications for the aquatic
nitrogen cycle. Biogeochemistry 2:179-196.
Di Guilio, R.T. and P.P. Scanlon. 1985. Heavy metals in
aquatic plants, clams, and sediments from the Chesapeake
Bay, U. S. A. Implications for the waterfowl. Sci. Tot.
Environ. 41:259-274.
Dzombak, D.A., W. Fish and F.M.M. Morel. 1986. Metal-
humate interactions. 1. Discrete ligand and continuous
distribution models. Environ. Sci. Technol. 20:669-675.
Elliot, S., E. Lu, and F.S. Rowland. 1985. Marine tropo-
spheric HjS and OCS imply significant levels of sulfide
anions in seawater. Trans. Amer. Geophys. Union 66:
1309.
Engel, D.W. and W.G. Sunda. 1979. Toxicity of cupric ion to
the eggs of the spot Leiostomus xanthurus and the Atlantic
silverside Menidia menidia. Mar. Biol. 50:121-126.
Engel, D.W., W.G. Sunda and B.A. Fowler. 1981. Factors
affecting trace metal uptake and toxicity to estuarine
organisms. I. Environmental parameters, p. 127-144. In:
W.B. Vernberg, P.P. Thurberg, A. Calabrese, and F.J.
Vernberg (eds.), Biological monitoring of marine
pollutants. Academic Press.
Fish, W., D.A. Dzombak and F.M.M. Morel. 1986. Metal-
humate interactions. 2. Application and comparison of
models. Environ. Sci. Technol. 20:676-683.
Fowler, S.W. 1982. Biological transfer and transport
processes, p.1-66. In: G. Kullenberg (ed.), Pollutant
transfer and transport in the sea, Vol. II. CRC Press.
Fujiki, M. 1980. Methylmercury accumulation in plankton
and fish, p. 485-492. In: R.A. Baker (ed.), Contaminants
and sediments, Vol. 2. Ann Arbor Science.
Garrett, W.D. and R.A. Duce. 1980. Surface microlayer
samplers, p. 471-490. In: F. Dobson, L. Hasse, and R.
Davis (eds.), Air-sea interaction instruments and methods.
Plenum Press.
Gibbs, R.J. 1982. The differentiation of toxic metals in
estuaries. Thalassia Jugosl. 18:181-190.
Gibbs, R.J. 1986. Segregation of metals by coagulation in
estuaries. Mar. Chem. 18:149-159.
Gilmour, C.C., J.H. Tuttle, and J.C. Means. 1985. Tin
methylation in sulfide bearing sediment, p. 239-258. In:
A. Sigleo and A. Hattori (eds.), Marine and estuarine
geochemistry. Lewis Publishers.
Goldberg, E.D., V. Hodge, M. Koide, J.J. Griffen, E. Gamble,
O.P. Bricker, G. Mattisoff, G.R. Holdren, and R. Braun.
1978. A pollution history of the Chesapeake Bay.
Geochim. Cosmochim. Acta 42:1413-1425.
Gucinski, H. 1986. The effect of sea surface microlayer
enrichment on TBT transport, p. 1266-1274. In: Oceans
86 Conference proceedings, Vol. 4. Washington D.C.
Hall, L.W., M.J. Lenkevich, W.S. Hall, A.E. Pinkney, and S J.
Bushang. 1986. Monitoring organotin concentrations in
Maryland waters of Chesapeake Bay, p. 1275-1279. In:
Oceans 86 Conference proceedings, Vol. 4. Washington,
D.C.
Hallas, L.E. and J.J. Cooney. 1981. Tin and tin resistant
microorganisms in Chesapeake Bay. Appl. Environ.
Microbiol. 41:446-471.
Hallas, L.E., J.C. Means, and J.J. Cooney. 1982. Methylation
of tin by estuarine microorganisms. Science 215:1505-
1507.
Hanson, A.K., Jr. and J.G. Quinn. 1983. The distribution of
dissolved and organically complexed copper and nickel in
the Middle Atlantic Bight. Can. J. Fish. Aquat. Sci. 40
(Suppl. 2):151-161.
Hardy, J.T. 1982. The sea-surface microlayer: biology,
chemistry and anthropogenic enrichment. Prog. Oceanogr.
11:307-328.
Hardy, J.T., E.A. Crecelius, and R. Kocarn. 1986. Concentra-
22
-------
Bioavailability of Toxics
lion and toxicity of sea-surface contaminants in Puget
Sound. Final report to NOAA, OAD.
Hardy, J.T., E.A. Crecelius, L.D. Antrim, V.L. Broadhurst,
P.D. Boehm, and W.G. Steinhauer. 1987. Aquatic surface
contamination in Chesapeake Bay. Draft final report to
MD DNR, Power Plant Research Program.
Harris, R.L., G.R. Helz, and R.L. Cory. 1975. Processes
affect the vertical distribution of trace components in the
Chesapeake Bay, p. 176-185. In: T.M. Church (ed.),
Marine chemistry in the coastal environment. ACS
Symposium series 18.
Harrison, G.I. and F.M.M. Morel. 1983. Antagonism between
cadmium and iron in the marine diatom Thalassiosira
weissflogii. J. Phycol. 19:495-507.
Helz, G.R., and R.J. Huggett. 1987. Contaminants in
Chesapeake Bay; the regional perspective. In: S.K.
Majumdar, L.W. Hall, and H. Austin (eds.), Contaminant
problems and management of living Chesapeake Bay
resources. Penn. Academy Sci., in press.
Helz, G.R. and R.J. Kieber. 1985. Hydrogen peroxide: a
minor but significant contribution to chlorine demand, p.
1033-1040. In: R.L. Jolley, R.J. Bull, W.P Davis, S. Kate,
M.H. Roberts, and V. Jacobs (eds.), Water chlorination:
environmental impact and health effects, Vol. 5. Lewis
Publ.
Helz, G.R., R.J. Huggett, and J.M. Hill. 1975. Behavior of
Mn, Fe, Cu, Zn, Cd, and Pb discharged from a wastewater
treatment plant into an estuarine environment. Water Res.
9:631-636.
Helz, G.R., S.A. Sinex, K.L. Ferri, and M. Nichols. 1985.
Processes controlling Fe, Mn, and Zn in sediments of
northern Chesapeake Bay. Est Coast. Shelf Sci. 21:1-16.
Howarth, R.W. and J.J. Cole. 1985. Molybdenum availabil-
ity, nitrogen limitation and phytoplankton growth in
natural waters. Science 229:653-655.
Huggett, R.J., M.E. Bender, and H.D. Slone. 1971. Mercury
in sediments from three Virginia estuaries. Ches. Sci.
12:280-282.
Huggett, R.J., F.A. Cross, and M.E. Bender. 1975. Distribu-
tion of copper and zinc in oysters and sediments from
three coastal plain estuaries, p. 224-238. In: F.G. Howell,
J.B. Gentry, and M.H. Smith (eds.), Mineral cycling in
southeastern ecosystems. NTIS CONF-740513.
Hung, Y-W. 1982. Effects of temperature and chelating
agents on cadmium in the American oyster. Bull. Environ.
Contam. Toxicol. 28:546-551.
Jenne, E.A., D.C. Girvin, J.W. Ball, and J.M. Burchard. 1978.
Inorganic speciation of silver in natural waters—fresh to
marine, p. 41-61. In: D.A. Klein (ed.), Environmental
impacts of artificial ice nucleating agents. Dowden,
Hutchinson, & Ross.
Kieber, R.J. and G.R. Helz. 1986. Two method verification
of hydrogen peroxide determinations in natural waters.
Anal. Chem. 58:2312-2315.
Kingston, H.M., R.R. Greenberg, E.S. Beary, B.R. Hardas,
J.R. Moody, T.C.Rains, and W.S. Liggett. 1982. The
characterization of the Chesapeake Bay: A systematic
analysis of toxic trace elements. EPA-79-D-X-0717.
Macek, K.J., S.R. Petrocelli and B.H. Sleight III. 1979.
Consideration in assessing the potential for, and signifi-
cance of biomagnification of chemical residues in aquatic
food chains. In: L.L. Marking and R.A. Kimrle (eds.)
Aquatic Toxicology, ASTM special publication 667,
American Society of Testing and Materials, Philadelphia,
pp.251-268.
Manders, W.F., G.J. Olson, F.E. Brinckman, and J.M.
Bellama. 1984. A novel synthesis of methyl tin tri-iodide
with environmental implications. J. Chem. Soc. Commun.
1984:538-540.
Mills, G.L. and J.G. Quinn. 1984. Dissolved copper and
copper-organic complexes in the Narragansett Bay
estuary. Mar. Chem. 15:151-172.
Moffet, J.W. and R. G. Zika. 1983. Oxidation kinetics of
Cu(I) in seawater: implications for its existence in the
marine environment. Mar. Chem. 13:239-251.
Morel, F.M.M. and N.M.L. Morel-Laurens. 1983. Trace
metals and plankton in the ocean, p. 841-870. In: C.S.
Wong, E. Boyle, K.W. Bruland, J.D. Burton, and E.D.
Goldberg (eds.), Trace metals in sea water. Plenum.
Newell, A.D. 1983. An investigation of copper-organic
complexes in the Patuxent River, Maryland. Thesis,
University of North Carolina, Chapel Hill.
Newell, A.D. and J.G. Sanders. 1986. The relative copper
binding capacities of dissolved organic compounds in a
coastal-plain estuary. Environ. Sci. Technol. 20:817-821.
Nichols, M., R. Harris, and G. Thompson. 1982. Fate,
transport, and transformation of toxic substances:
significance of suspended sediment and fluid mud. EPA-
600/8-82-089.
Officer, C.B., D.R. Lynch, G.H. Setlock, and G. R. Helz.
1984. Recent sedimentation in Chesapeake Bay, p. 131-
158. In: V.S. Kennedy (ed.), The estuary as a filter.
Academic Press.
Owens, J.P., K. Stefansson, and L.A. Sirkin. 1974. Chemical,
mineralogic, and palynologic character of the Upper
Wisconsinan-Lov> ur Holocene fill in parts of Hudson,
Delaware, and Chesapeake estuaries. J. Sed. Pet. 44:390-
408.
Peoples, S.A. 1975. Review of arsenical pesticides, p. 1-12.
In: E.A. Woolson (ed.), Arsenical pesticides. ACS
Symposium series 7.
Peterson, M.L. and R. Carpenter. 1983. Biogeochemical
processes affecting total arsenic and arsenic species
distributions in an intermittently anoxic fjord. Mar. Chem.
12:295-321.
Pheiffer, T.H. 1972. Heavy metal analyses of bottom
sediment in the Potomac River estuary. EPA Region III
Technical Report 49.
23
-------
PLENARY SESSION
Phelps, H.L., D.A. Wright, and J.A. Mihursky. 1985. Factors
affecting trace metal accumulation by estuarine oysters,
Crassosirea virginica. Mar. Ecol. Prog. Ser. 22:187-197.
Planas, D. and P.P. Healey. 1978. Effects of arsenate on
growth and phosphorus metabolism of phytoplankton. J.
Phycol. 14:337-341.
Rice, D.L. and S. I. Whitlow. 1985. Early diagenesis of
transition metals: a study of metal partitioning between
macrofaunal populations and shallow sediments, p. 21-30.
In: The fate and effects of pollutants. MD Sea Grant publ.
UM-SG-TS-85-02.
Riedel, G.F. 1984. Influence of salinity and sulfate on the
toxicity of chromium (VI) to the estuarine diatom,
Thalassiosira pseudonana. J. Phycol. 20:496-500.
Riedel, G.F. 1985a. The relationship between chromium (VI)
uptake, sulfate uptake, and chromium (VI) toxicity in the
estuarine diatom Thalassiosira pseudonana. Aq. Toxicol.
7: 191-204.
Riedel, G.F. 1985b. Photoreduction of Cr(VI) in natural
waters.Trans. Amer. Geo. Union 66:1257.
Riedel, G.F., J.G. Sanders, and R.W. Osman. 1987. The
effect of biological and physical disturbances on the
transport of arsenic from contaminated estuarine sedi-
ments. Est. Coast. Shelf. Sci. 25:693-706.
Rueter, J.G. and F.M.M. Morel. 1981. The interaction
between zinc deficiency and copper toxicity as it affects
silicic acid uptake mechanisms in the marine diatom
Thalassiosira pseudonana. Limnol. Oceanogr. 26:67-73.
Sanders, J.G. 1979. Effects of arsenic speciation and
phosphate concentration on arsenic inhibition of Skele-
tonema costatum (Bacillariophyceae). J. Phycol. 15:424-
428.
Sanders, J.G. 1980. Arsenic cycling in marine systems. Mar.
Environ. Res. 3:257-266.
Sanders, J.G. 1982. Chlorination of estuarine water: the
occurrence and magnitude of carbon oxidation and its
impact on trace metal transport. Environ. Sci. Technol.
16:791-796.
Sanders, J.G. 1985. Arsenic geochemistry in Chesapeake
Bay: dependence upon anthropogenic inputs and
phytoplankton species composition. Mar. Chem. 17:329-
340.
Sanders, J.G. 1986. Alteration of arsenic transport and
reactivity in coastal marine systems after biological
transformation. Rapp. P.-v. Reun. Cons. int. Explor. Mer
186:185-192.
Sanders, J.G. and G.R. Abbe. 1986. The processes and
pathways determining silver uptake and accumulation in
American oysters in Chesapeake Bay. Progress report for
1985 prepared for Baltimore Gas & Electric Co.
Sanders, J.G. and G.R. Abbe. 1987a. The role of suspended
sediments and natural phytoplankton in the partitioning
and transport of silver in estuaries. Cont. Shelf Res.
7:1357-1361.
Sanders, J.G. and G.R. Abbe. 1987b. Silver transport and
impact in estuarine and marine systems. In: G. Suter and
M. Lewis (eds.), 11 Symposium on aquatic toxicology
and hazard assessment, in press.
Sanders, J.G. and G.R. Abbe. 1987c. The processes and
pathways determining silver uptake and accumulation in
American oysters in Chesapeake Bay. Progress report for
1986 prepared for Baltimore Gas & Electric Co.
Schrift, A. 1954. Sulfur-selenium antagonism. I. An-
timetabolite action of selenate on the growth of Chlorella
vulgaris. Am. J. Bot. 41:223-230.
Seitzinger, S. 1986. The effect of pH on the release of
phosphorus from Potomac River sediments. Final report
to EPA Chesapeake Bay Program. ANSP report 86-8F.
Seliger, H.H., J.A. Boggs, and W.H. Biggley. 1985. Cata-
strophic anoxia in the Chesapeake Bay in 1984. Science
228: 70-73.
Shigenaka, G. and J.A. Calder. 1987. Organic compounds and
metals in selected fish, bivalve molluscs, and sediments of
the Chesapeake Bay: preliminary results from the
National Status and Trends Program, p. 69-80. In: M.P.
Lynch and E.G. Krome (eds.), State of the Chesapeake
Bay—Second annual monitoring report—Compendium.
Chesapeake Research Consortium publ. 125.
Sigleo, A.C. and G.R. Helz. 1982. Composition of estuarine
colloidal material: major and trace elements. Geochim.
Cosmochim. Acta 45:2501-2509.
Sinex, S.A. and G.R. Helz. 1981. Regional geochemistry of
trace elements in Chesapeake Bay sediments. Environ.
Geol. 3:315-323.
Stumm, W. and JJ. Morgan. 1970. Aquatic chemistry.
Wiley Interscience.
Sunda, W.G. 1975. Relationship between cupric ion activity
and the toxicity of copper to phytoplankton. Dissertation,
Massachusetts Institute of Technology.
Sunda, W.G. and R.R.L. Guillard. 1976. Relationship
between cupric ion activity and toxicity of copper to
phytoplankton. J. Mar. Res. 34:511-529.
Sunda, W.G. and A.K. Hanson. 1987. Measurement of free
cupric ion concentration in seawater by a ligand competi-
tion technique involving copper sorption onto CIS SEP-
PAK cartridges. Limnol. Oceanogr. 32:537-551.
Sunda, W.G., D.W. Engel, and R.M. Thuotte. 1978. Effect of
chemical speciation on toxicity of cadmium to grass
shrimp Palemonetespugio: importance of free cadmium
ion. Environ. Sci. Technol. 12:409-413.
Takayanagi, K. and G.T. Wong. 1984. Organic and colloidal
selenium in Chesapeake Bay. Mar. Chem. 14:141-148.
Troup, B.N. and O.P. Bricker. 1975. Processes affecting the
transport of materials from continents to oceans, p. 133-
151. In: T.M. Church (ed.), Marine chemistry in the
coastal environment. ACS Symposium series 18.
van den Berg, C.M.G., A.G. Merks and E.K. Duursma. 1987.
Organic complexation and its control of the dissolved
24
-------
Bioavailability of Toxics
concentrations of copper and zinc in the Scheldt estuary.
Est. Coast. Shelf Sci. 24:785-798.
Villa, D., Jr. and P.G. Johnson. 1974. Distribution of metals
in Baltimore Harbor sediments. EPA Technical Report 59.
Westbrook, D.J., E.J. Travelstead, F.A. Espourteille, C.D.
Rice, and R.J. Huggett. 1986. Tributyltin in whole water
and sediment collected from marinas and the Hampton
Roads area in southern Chesapeake Bay. Final Report to
Johns Hopkins University, Applied Physics Laboratory.
Wheeler, A.E., R.A. Zingaro, K. Irgolic, and N.R. Bottino.
1982. The effect of selenate, selenite, and sulphate on the
growth of six unicellular algae. J. Exp. Mar. Biol. Ecol.
57:181-194.
Wood, J.M. 1974. Biological cycles for toxic elements in the
environment. Science 183:1049-1052.
Wong, G.T. and C.S. Moy. 1984. Cesium-137, metals and
organic carbon in the sediments of the James River
estuary, Virginia. Est. Coast. Shelf Sci. 18:37-49.
Wong, P.T.S., Y.K. Chau, O. Kramer, and G.A. Bengert.
1982. Structure-toxicity relationship of tin compounds on
algae. Can. J. Fish. Aquat. Sci. 39:483-488.
Wright, D.A. and G.D. Foster. 1987. Toxicant monitoring in
the upper Chesapeake Bay, p. 55-62. In: M.P. Lynch and
E.G. Krome (eds.), State of the Chesapeake Bay—second
annual monitoring repor—Compendium. Chesapeake
Research Consortium publ. 125.
Wright, D.A. and C.D. Zamuda. 1987. Copper accumulation
by two bivalve molluscs: salinity effect is independent of
cupric ion activity. Mar. Environ. Res. 23:1-14.
Young, D.R. 1984. Methods of evaluating pollutant biomag-
nification in marine ecosystems, p. 261-278. In: H.H.
White (ed.), Concepts in marine pollution measurements.
MD Sea Grant publ. UM-SG-TS-84-03.
Young, G.R., J.A. Swader, and S.W. Bingham. 1980. Kepone
uptake by algae and effects on growth, photosynthesis, and
respiration. Va. J. Sci. 31: 47-54.
Zamuda, C.D. 1984. The bioavailability of copper to the
American oyster Crassostrea virginica. Dissertation,
University of Maryland.
Zamuda, C.D. and W.G. Sunda. 1982. Bioavailability of
dissolved copper to the American oyster Crassostrea
virginica. I. Importance of chemical speciation. Mar.
Biol. 66:77-82.
Zamuda, C.D., D.A. Wright, and R.A. Smucker. 1985. The
importance of dissolved organic compounds in the
accumulation of copper by the American oyster, Crassos-
trea virginica. Mar. Environ. Res. 16:1-12.
Zirino, A. and S. Yamamoto. 1972. A pH-dependent model
for chemical speciation of copper, zinc, cadmium, and lead
in seawater. Limnol. Oceanogr. 17:661-671.
Zoller, W.H., G.K. Barley, M.S. Germani, and D.S. Bal-
lantine. 1983. Fate of mercury release from power plants.
Final report to MD Power Plant Siting Program PPRP-79.
Zuckerman, J.J., R.P. Reisdorf, H.V. Ellis, and R.R. Wilker-
son.1978. Organotins in biology and the environment, p.
388-422. In: F.E. Brinckman and J.M. Bellama (eds.),
Organometals and organometalloids: occurrence and fate
in the environment. ACS Symposium series 82.
Zuelke, R.W. and D.R. Kester. 1983. Copper speciation in
marine waters, p. 773-788. In: C.S. Wong, E. Boyle,
K.W. Bruland, J.D. Burton, and E.D. Goldberg (eds.),
Trace metals in sea water. Plenum.
25
-------
PLENARY SESSION
Unavailability of Organic Pollutants to Aquatic Organisms
Robert C. Hale and Robert J. Huggett
Division of Chemistry and Toxicology
Virginia Institute of Marine Science
College of William and Mary
Gloucester Point, Virginia 23062
INTRODUCTION
Settlement of the Chesapeake Bay region began in
earnest in the early 17th century. The native American
population and early colonists were impressed by the
abundance of fish and shellfish and located their
population centers to take advantage of these and other
natural resources. Introduction of wastes into the bay
was coincident with this settlement. As the human
population increased, so did the pressure on the
ecological system. In the 20th century significant
quantities of synthetic chemicals began to be intro-
duced, many of which were toxic and nonbiodegradable
[Faust and Hunter 1971]. Today the areas surrounding
the bay are experiencing unprecedented development;
introduction of toxic organic pollutants has correspond-
ingly increased. Water quality and the abundance of
many aquatic organisms have suffered accordingly
[O'Connor and Huggett 1988].
The bioavailability of organic pollutants is of recent
concern and has not been as thoroughly studied as that
of toxic metals. The initial belief was that "insoluble"
organic xenobiotics were not available to aquatic
organisms. They were assumed to be eliminated by
irreversible binding to bottom sediments, which
supposedly removed any significant threat to the
ecosystem. More recently, we have come to realize that
the term insoluble is a misnomer. All organic com-
pounds possess some water solubility. In fact, it has
been established that many low-solubility compounds
are bioaccumulated or biomagnified to high concentra-
tions in the tissues of organisms. Within several
homologous series of organic compounds, toxicity has
been negatively correlated with water solubility [Veith
et al. 1983; Konemann 1981]. However, since chemi-
cally dissimilar compounds may exert their toxic effects
via different mechanisms and influence different
physiological functions, the solubility/toxicity relation-
ship may not be directly applicable across series; the
case of alcohols [Veith et al. 1983] versus organo-
phosphorus pesticides [DeBruin 1976] is a good
example. Sediments in many cases have been deter-
mined to be a source of toxic compounds, rather than
merely a sink [Willford et al. 1987]. As a consequence,
we have been forced to re-evaluate our understanding
of the bioavailability of organic pollutants and their
significance to the health of the Chesapeake Bay
ecosystem.
PHYSICAL, CHEMICAL, AND BIOLOGICAL BASIS
OF BIOAVAILABILITY
Water
Water is an obvious and important route for the
exposure of aquatic organisms to organic pollutants.
The water solubility of a compound has a profound
influence on its environmental fate and bioavailability.
Biota may come into contact with high concentrations
of compounds that exhibit significant solubility, e.g.,
alcohols, phenols, and benzenes. Accidental spills and
untreated effluents may result in toxic concentrations of
these compounds in the water, causing obvious acute
effects such as fish kills.
Some of the most toxic organic compounds possess
low water solubilities. Classes of these compounds
detected in the Chesapeake Bay include polynuclear
aromatic hydrocarbons (PAHs), heterocyclic aromatic
compounds (HACs), and halogenated pesticides. Table
1 gives several specific examples. These compounds
may be taken up by biota directly from water, although
dissolved concentrations will generally be very low.
The mechanism of accumulation is believed to be
simple partitioning from water into lipid-rich biological
tissues [Esser 1986]. Active biologically mediated
transport mechanisms are thought to be less prevalent
for organic pollutants than for trace metals; many of the
metals have critical functions in enzyme systems. The
hydrophobic nature of lipophilic xenobiotics provides
26
-------
Bioavailability of Toxics
the driving force for the partitioning process. The
presence of nonpolar organic solutes is not compatible
within the polar water phase.
Since laboratory experiments designed to determine
bioavailability directly using fish and invertebrates are
expensive and time-consuming, surrogate tests have
been proposed. For example, the bioaccumulation
tendency of a lipophilic organic generally correlates
with the n-octanol/water partition coefficient or Kw.
Octanol has been suggested as a substitute for biologi-
cal lipids in these experiments, although some research-
ers have suggested that critical differences exist in the
thermodynamics of partitioning between water, fish
lipids, and octanol [Opperhuizen et al. 1988]. The
classical approach for the determination of an octanol/
water partition coefficient involves the addition of the
test compound to a vessel containing mutually saturated
octanol and water phases. The contents of the vessel
are then thoroughly mixed and the system is allowed to
come to equilibrium. The concentrations of the test
compound in each of the two phases are subsequently
determined and the coefficient calculated [Karickhoff
and Brown 1979]. Difficulties (e.g., emulsions,
detection limits, and contamination) are inherent in de-
terminations for compounds possessing Kow values
greater than 100,000. Indirect measurement of the Kow
has also been suggested. For example, high-perform-
ance liquid chromatography (HPLC) has been used
[Brooke et al. 1986]. Basically, a series of chemically
similar compounds, for which the partition coefficients
are known, are co-injected with the compound of
interest onto a reverse-phase HPLC column. A
correlation of retention time with K^is then deter-
mined and the partition coefficient for the compound of
interest is calculated from this relationship. Kow values
are often expressed as logarithms, due to the magnitude
of the values and the relationship of this parameter to
bioconcentration factors (BCFs). Table 2 lists the log
BCF and log Kowof several organic compounds which
have been detected in the environment. Note that the
BCFs are less than the Kow values.
Equations relating Kow directly to bioconcentration
factors (BCF) in various types of organisms and to
water solubility have been reported [Isnard and Lam-
bert 1988; Esser 1986]. It has been reported that
concentrations of lipophilic pollutants will be similar in
aquatic organisms in general, provided exposure has
been equal, equilibrium has been established, and the
relative lipid contents of the biota have been normalized
[Adams 1988]. Obviously, differences in the biotrans-
formation capabilities of the organisms may alter this
relationship.
Kowdeterminations provide no information concern-
ing biotransformation or biological effects. These
Table 1. Water solubility [May et al. 1978] and LC50
[Trucco et al. 1983] for some common environmental
contaminants. Values for LC50 encompass a variety of
organisms and conditions and are used for illustrative
purposes only.
Compound
(mg/1)
Solubility
LC50 (mg/1)
Naphthalene 31.7
Ronnel 1.08*
Phenanthrene 1.00
Benz(a)anthracene 0.009
Arochlor 1254 0.012+
p,p'-DDT 0.003*
1.00
0.49T
0.10
0.01
0.003+
0.0002t
* Data from Chiou et al. (1977).
t Data from Johnson et al. (1980).
+ Data from National Research Council (1979).
phenomena are often significant. For example, English
sole exposed to benzo(a)pyrene and Aroclor 1254
exhibited progressive accumulation of PCBs, but little
accumulation of the PAH. This difference was attrib-
uted to extensive metabolism of benzo(a)pyrene by the
fish [Malins et al. 1987]. Information on metabolism is
very important in assessing the fate and effects of
chemicals, especially since biotransformation may
result in the production of more toxic or mutagenic
products [Buhler and Williams 1988]. Bruggeman et
al. [1984] observed that guppies bioaccumulated
hexachlorobiphenyl, but not hexabromobenzene during
aqueous exposures. They attributed this observation to
the existence of an upper molecular size threshold,
limiting transport across membranes. The importance
of molecular volume has also been suggested by other
workers [Doucette and Andren 1987]. Thus the use of
octanol/water partition coefficients alone may result in
Table 2. The log BCF and log Kow of several common
environmental contaminants are given. Values are from
Veith et al. [1979], except as noted.
Compound
log BCF
log K0
Naphthalene
Pentachlorophenol
Phenanthrene
p,p'-DDT
Chlordane
Aroclor 1254
2.63
2.89
3.42
4.47
4.58
5.00
3.37*
5.01
4.46
5.75
6.00
6.47
*Chiouetal. (1977).
27
-------
PLENARY SESSION
an overestimation of the BCF in the case of extremely
large molecules.
The major avenue of entry into biota for lipophilic
compounds varies depending on the organism's size,
morphology, and ecology [Knezovich and Harrison
1987]. For example, sorption to the general cell surface
may predominate for single-cell biota, such as algae or
heterotrophic bacteria. In higher organisms the body
surface may contribute, but specialized respiratory
structures (e.g., gills in fish) often are more important.
Because of the large amounts of water processed by
gills and the high lipid/water partition values (often
> 1,000,000), significant amounts of xenobiotics may
rapidly bioaccumulate. Physiological changes may
alter the bioavailability and thus the toxicity of organics
to organisms. Conklin and Rao [1978] exposed molting
adult grass shrimp to pentachlorophenol during various
phases of the ecdysial cycle. They observed greater
sensitivity and enhanced uptake immediately after
ecdysis and attributed these effects to increased
permeability of the cuticle. Newly molted blue crabs
were observed to contain higher burdens of radiola-
beled benzo(a)pyrene-derived material, compared with
nonmolting crabs [Hale 1988] after laboratory exposure
of these organisms.
Sediments
As previously mentioned, many organic compounds are
rapidly sorbed to paniculate matter. These compounds
may adsorb to the mineral surface directly or to organic
constituents of particles. The latter site is thought to be
the most important [Hodson and Williams 1988].
Means et al. [1980], in determinations of batch equilib-
rium sorption isotherms, reported that adsorption of
PAHs and HACs onto suspended soil/sediment was
independent of substrate pH, cation exchange capacity,
textural composition, or clay mineralogy. This state-
ment is probably an oversimplification for direct appli-
cation to waters of the Chesapeake Bay, because of the
effects of factors such as pH on dissolved organics that
may sorb lipophilic pollutants. These parameters are
quite important in the behavior of polar organics and
heavy metals. Grain size has been mentioned as a
significant factor by some researchers [Marcus et al.
1988]. However, sediment grain size and organic con-
tent are generally correlated in the aquatic environment.
The relationship of pollutant adsorption to organic
content of sediments is well documented [Karickhoff
and Brown 1979]. Equations relating sediment
adsorption have been reported and generally employ
coefficients such as Kow [Dzombak and Luthy 1984].
The exact physical/chemical mechanism of sorption
of chemicals to particulates is still uncertain. It has
been postulated that the reduction of the water-organic
interfacial area achieved by sorption is critical [Mackay
and Powers 1987]. The mechanisms of uptake into
biological lipids and the organic constituents of
particulates appear similar. In comparison, parameters
such as hydrogen bonding and gross electrostatic
attractions contribute little to the sorption of nonpolar
organics to sediments [Voice and Weber 1983].
Sorption of organic pollutants reduces but does not
eliminate availability to biota. Malins et al. [1987]
reported significant correlations of the occurrence of
liver disease in bottom-dwelling fish with the level of
sediment PAH contamination. Huggett et al. [1987]
found a variety of abnormalities in fish of the Elizabeth
River, Virginia, compared with specimens from less
polluted rivers. The Elizabeth is heavily contaminated
with PAHs. Hargis et al. [1984] reported acute toxicity
and lesion formation in fish held over Elizabeth River
sediments in the laboratory. Fish exposed to water
alone that had passed over these contaminated sedi-
ments also showed signs of chemical stress. This result
indicated that the xenobiotics were quite bioavailable.
Duration of contact between lipophilic compounds
and sediments has been observed to affect their rate of
desorption and bioavailability [Varanasi et al. 1985;
Voice and Weber 1983]. Similar observations in regard
to the elimination of lipophilic residues from aquatic
biota, i.e., biphasic depuration patterns, have been
reported [Spacie and Hamelink 1982]. Haddock et al.
[1983] reported that extraction efficiency of PAHs,
using organic solvents, was a function of PAH/sediment
contact time. Coal is an extreme illustration of the
bioavailability question. It contains a large variety and
quantity of PAHs. However, these compounds were
found to be biologically unavailable to oysters during
laboratory exposures to suspended coal dust [Bender et
al. 1987]. Indeed, coal and activated charcoal are used
in water treatment for the removal of dissolved organ-
ics. An analogous example may be the low availability
of metals incorporated in the crystalline lattice structure
of mineral grains. The transfer of lipophilic xenobiotics
from organic reservoirs of sediments to those in biota
does not present the favorable thermodynamics
observed for partitioning from water to sediments or
biota. Indeed, transfer from sediments to biota may
entail movement into the polar water phase as an
intermediate step. The octanol/water partition coeffi-
cient will describe the tendency for this movement to
occur, desorption obviously being less favorable than
sorption. Wood et al. [1987] observed that lower
chlorinated congeners were preferentially desorbed
from PCB-contaminated sediments. They also ob-
served that dipteran larvae present in these sediments
bioaccumulated PCB congeners containing two to four
chlorines in preference to higher chlorinated congeners.
28
-------
Bioavailability of Toxics
More typically, the bioconcentration of PCBs has been
highest for congeners possessing five to seven chlo-
rines, in accordance with their Kw.
Evidence that accumulation and toxicity of organic
xenobiotics are related to the positions of the substitu-
ents is also available [Kuehl et al. 1987]. Sediments
collected from sites in the Hampton Roads area of
Virginia were analyzed by capillary gas chromatogra-
phy and a halogen-selective Hall detector [Hale, in
preparation]. Significant concentrations of both PCBs
and highly chlorinated polychlorinated terphenyls
(PCTs) were identified at one location. Clams collected
at this same site exhibited only PCBs, but the accumu-
lation of lower-chlorinated PCT congeners observed in
bivalves at another site indicated that PCTs were
bioavailable.
Exposure of biota to paniculate-associated xenobiot-
ics may occur via direct epidermal contact with
sediment particles. Contact with interstitial water,
which contains higher concentrations of pollutants than
the overlying waters, may be even more important
[Knezovich and Harrison 1987]. Suspension feeders
such as oysters may filter contaminated particles from
the water, and other organisms (e.g., benthic worms)
may ingest bulk sediment [Reynoldson 1987]. Again,
bioavailability of this adsorbed material is lower than
that of xenobiotics dissolved in water.
Resuspension of contaminated sediments via
dredging operations has been observed to increase the
bioavailability of Kepone in the James River [Lunsford
et al. 1987]. Severe hydrographic events, e.g. floods
and hurricanes, may resuspend sediments [Wood et al.
1987]. Bioturbation has also been reported to increase
bioavailability of organic toxics. Reynoldson [1987]
reviewed the results of several studies of bioturbation.
He reported that tubificid worms transported 90% of
hexachlorobenzene, pentachlorobenzene, and trifluralin
to the sediment surface from a uniformly mixed
sediment in laboratory experiments over a 50-day
period. This resulted in a four- to six-fold increase in
contaminants in the overlying water, compared with
unperturbed systems. This work was done by Karick-
hoff and Morris.
Estuarine circulation and chemistry may result in the
formation of turbidity maxima. This phenomenon is
observed when suspended particulates flocculate upon
contact with saline water and settle. The particles are
then transported up-estuary by tidal currents to less
saline water, where they may dissociate and again be
carried downstream. This cycling leads to the develop-
ment of a turbid zone with a high sedimentation rate.
Lipophilic pollutants sorb to particulates and may
concentrate in these zones, as was observed for Kepone
in the James River [Huggett et al. 1980].
Surface Microlayer
Organisms on the surface of the water may be exposed
to elevated concentrations of xenobiotics via the surface
microlayer. Lipophilic organics accumulate there
because of solubility, surface tension, and specific
gravity considerations. This accumulation may also be
critical to eggs and larvae of species that frequent this
layer [Hardy et al. 1987]. The surface microlayer will
be discussed further during this conference.
Suspended and Dissolved Organic Matter
Naturally occurring organic matter, e.g., humic acids,
may sorb lipophilic xenobiotics. McCarthy [1983]
reported that dissolved and colloidal organic matter
reduced the bioavailability of PAHs to daphnia in
laboratory experiments. The effect may be physical or
chemical. The molecular size of the complex may limit
bio-uptake, as mentioned previously. The chemical
nature, e.g., the polarity, may also affect the bioaccu-
mulation potential of the bound pollutant.
Carter and Suffet [1982] reported that DDT may
sorb to dissolved humic materials and thus remain in
the water column for extended periods of time. Para-
meters such as pH may affect the residence time of
naturally occurring organic matter and, in turn, affect
that of associated pollutants. Sorption of lipophilic
xenobiotics to organics and particulates has been
reported to have varying effects on persistence [Leslie
et al. 1987]. Important routes of degradation include
photochemical and microbial pathways.
Food
Organic pollutants may be available to biota via
consumption of contaminated prey items or plant
material. Ingestion of contaminated particulates was
discussed previously. Biomagnification has been
identified as a significant factor for the transmission of
some persistent xenobiotics. For example, food has
been observed to serve as the major vehicle for uptake
of Kepone in blue crabs. Exposure of these crustaceans
to Kepone via water, during laboratory experiments,
resulted in minimal body burdens. However, when
contaminated food (in the form of either James River or
laboratory-exposed oysters) was provided, significant
concentrations of Kepone were accumulated [Schimmel
et al. 1979]. Obviously, differing feeding strategies
will put certain organisms at greater risk, both because
of the items selected, and because of the efficiency of
uptake. The efficiency of transfer of lipophilic xenobi-
otics from the tissues of prey to consumers will be, as
from the organic reservoirs of particulates, much less
than that from water. Transfer of pollutants to humans
via consumption of contaminated seafood is also of
considerable concern.
29
-------
PLENARY SESSION
SUMMARY OF BIO AVAILABILITY AND
ASSESSMENT TECHNIQUES
A variety of techniques have been utilized to assess
bioavailability. Laboratory procedures are appealing
since variables may be more easily controlled. Bioac-
cumulation studies, using tissue burdens of parent
compounds, suffer since biota are often able to metabo-
lize xenobiotics which normally go undetected by
conventional procedures. Radiotracers have the
advantage that total xenobiotic material present in the
organism may be determined; however, they are
expensive, have limited availability, and require
specialized handling. Bioaccumulation and toxicity
studies often use environmentally unrealistic exposure
scenarios, e.g., solvent carriers and filtered seawater, to
obtain "reasonable results." Application of organic
pollutants to sediments and immediate use in experi-
ments ignores the effects of contact time on pollutant/
sediment binding. The shortcomings of octanol/water
determinations have been discussed. Adsorption
isotherm determinations provide valuable information,
although the observation that organic components of
particulates are the major constituent into which
lipophilic pollutants partition must be considered.
Nonetheless, these techniques provide information
critical to our understanding of the general processes
controlling bioavailability.
Organic pollutants are readily accumulated by
aquatic organisms. Many compounds exhibit appre-
ciable toxicity. Of particular concern are compounds of
low solubility. These compounds may be bioaccumu-
lated or biomagnified to high concentrations. They
sorb to particulates, which may be ingested or settle to
the bottom. Sediment-associated contaminants exhibit
reduced bioavailability, but detrimental effects have
been documented. Little information regarding the
properties of the sediment/water interface is available,
considering its importance with respect to pollutant fate
and bioavailability. Differing feeding strategies and
habitat selection may result in toxics being more avail-
able to some organisms than to others. Because the
Bay is shallow, bioturbation, weather events, construc-
tion, boat traffic, and dredging may render sediment-
adsorbed organics more available via re-working or
resuspension.
The Chesapeake Bay represents a particularly
complex theater for the investigation of the bioavaila-
bility question. Salinity regimes range from fresh to
saline. These conditions may have measurable effects
on the residence time of organic species and the
physiology of aquatic organisms. Content of suspended
particulates and content of colloidal and dissolved or-
ganics also differ drastically within the Bay. Tidal and
riverine flows affect the disposition of toxics in both the
water column and sediments. Turbidity, dissolved
oxygen content, and temperature all influence the de-
gradation rates of xenobiotics. These complexities will
manifest themselves when laboratory data concerning
bioavailability are applied to actual field situations.
REFERENCES
Adams, W. J. Bioavailability of neutral lipophilic organic
chemicals contained on sediments: a review. In: Dickson,
K. L.; Maki, A. W.; Brungs, W., eds. Fate and effects of
sediment bound chemicals in aquatic systems. New York:
Pergamon Press, 1988: p. 219-244.
Bender, M. E.; Roberts, M. H., Jr.; deFur, P. O. Unavailabil-
ity of polynuclear aromatic hydrocarbons from coal
particles to the Eastern oyster. Environ. Poll. 44:243-260;
1987.
Brooke, D. N.; Dobbs, A. J.; Williams, N. Octanol:water
partition coefficients (P): measurement, estimation, and
interpretation, particularly for chemicals with P > 10s.
Ecotox. Environ. Safety 11:251-260; 1986.
Bruggeman, W. A; Opperhuizen, A.; Wijbenga, A.; Hutzin-
ger, O. Bioaccumulation of super-lipophilic chemicals in
fish. Toxicol. Environ. Chem. 7:173-189; 1984.
Buhler, D. R.; Williams, D. E. The role of biotransformation
in the toxicity of chemicals. Aquat. Toxicol. 11:19-28;
1988.
Carter, C. C.; Suffet, I. H. Binding of DDT to dissolved
humic materials. Environ. Sci. Technol. 16:735-740; 1982.
Chiou, C. T.; Freed, V. H.; Schmedding, D. W.; Kohnert, R.
L. Partition coefficient and bioaccumulation of selected
organic compounds. Environ. Sci. Technol. 11:475-478;
1977.
Conklin, P. J.; Rao, K. R. Toxicity of sodium pentachloro-
phenate to the grass shrimp, Palaemonetes pugio, in
relation to the molt cycle. In: Rao, K. R., ed. Pentachloro-
phenol: chemistry, pharmacology, and environmental
toxicology. New York: Plenum Press, 1978:181-192.
DeBruin, A. Biochemical toxicology of environmental
agents. Elsevier, Amsterdam; 1976.
Doucette, W. J.; Andren, A. W. Correlation of octanol/water
partition coefficients and total molecular surface area for
highly hydrophobia aromatic compounds. Environ. Sci.
Technol. 21:821-824; 1987.
Dzombak, D. A.; Luthy, R. G. Estimating adsorption of
polycyclic aromatic hydrocarbons on soils. Soil Sci.
137:292-307; 1984.
30
-------
Bioavailability of Toxics
Esser, H. O. A review of the correlation between physico-
chemical properties and bioaccumulation. Pestic. Sci.
17:265-276; 1986.
Faust, S. J.; Hunter, J. V., editors. Organic compounds in the
aquatic environment. New York: Marcel Dekker, Inc.;
1971.
Haddock, J. D.; Landrum, P. F.; Glesy, J. P. Extraction
efficiency of anthracene from sediments. Anal. Chem.
55:1197-1200; 1983.
Hale, R. C. Dynamics of organic pollutants in blue crabs,
Callinectes sapidus, collected from the lower Chesapeake
Bay region. In Lynch, M.P., and Krome, E.G., eds.
Understanding the Estuary: Advances in Chesapeake Bay
Research. Proceedings of a Conference, 29-31 March
1988, Baltimore, MD. CRC publication no. 129; Chesa-
peake Bay Program publication no. 24-8S/
Hardy, J. T; Crecelius, B. A.; Antrim, L. D.; Kiesser, S. L.;
Broadhurst, V. L.; Boehm, P. D.; Steinhauer, W. G.
Aquatic surface microlayer contamination in Chesapeake
Bay. 1987; 43 p. Maryland Power Plant Research
Program; PPRP-100.
Hargis, W. J.; Roberts, M. H., Jr.; Zweraer, D. E. Effects of
contaminated sediments and sediment-exposed effluent
water on an estuarine fish: acute toxicity. Mar. Environ.
Res. 14:337-354; 1984.
Hodson, J.; Williams, N. A. The estimation of the adsorption
coefficient (K^.) for soils by high performance liquid
chromatography. Chemosphere 17:67-77; 1988.
Huggett, R. J.; Bender, M. E.; Unger, M. A. Polynuclear
aromatic hydrocarbons in the Elizabeth River, Virginia. In:
Dickson, K. L.; Maki, A. W.; Brungs, W., eds. Fate and
effects of sediment bound chemicals in aquatic systems.
New York: Pergamon Press, 1987: p. 327-341.
Huggett, R. J.; Nichols, M. M.; Bender, M. E. Kepone
contamination of the James River Estuary. In: Baker, R.
A., ed. Contaminants and sediments, Volume 1: Fate and
transport, case studies, modeling, toxicity. Ann Arbor:
Ann Arbor Science; 1980: p. 33-52.
Isnard, P.; Lambert, S. Estimating bioconcentration factors
from octanol-water partition coefficient and aqueous
solubility. Chemosphere 17:21-34; 1988.
Johnson, W. W.; Finley, M. T. Handbook of acute toxicity of
chemicals to fish and aquatic invertebrates. 1980; 98 p.
Resource Publication 137-U.S. Fish and Wildlife Service.
Available from U.S. Government Printing Office.
Karickhoff, S. W.; Brown, D. S. Determination of octanol/
water distribution coefficients, water solubilities, and
sediment/water partition coefficients for hydrophobic
organic pollutants. Environmental Research Laboratory-
Athens, Ga. 1979; 25 p. Available from NTIS; PB80-
103591.
Knezovich, J. P; Harrison, F. L. The bioavailability of
sediment-sorbed organic chemicals: a review. Water Air
Soil Poll. 32:233-245; 1987.
Konemann, H. Quantitative structure-activity relationships in
fish toxicity studies. Part 1: Relationship for 50 industrial
pollutants. Toxicol. 19:209-221; 1981.
Kuehl, D. W.; Cook, P. M.; Batterman, A. R.; Butterworth, B.
C. Isomer dependent bioavailability of polychlorinated
dibenzo-p-dioxins and dibenzofurans from municipal
incinerator fly ash to carp. Chemosphere 16:657-666;
1987.
Leslie, T. J.; Dickson, K. L.; Jordan, J. A.; Hopkins, D. W.
Effects of suspended solids on the water column biotrans-
formation of anthracene. Arch. Environ. Contam. Toxicol.
16:637-642; 1987.
Lunsford, C. A.; Weinstein, M. P.; Scott, L. Uptake of
Kepone by the estuarine bivalve Rangia cuneata, during
dredging of contaminated sediments in the James River.
Water Res. 21:411-416; 1987.
Mackay, D.; Powers, B. Sorption of hydrophobic chemicals
from water: a hypothesis for the mechanism of the particle
concentration effect. Chemosphere 16:745-758; 1987.
Malins, D. C.; McCain, B. B.; Brown, D. W.; Varanasi, U.;
Krahn, M. M.; Myers, M. S.; Chan, S. Sediment-
associated contaminants and liver diseases in bottom-
dwelling fish. Hydrobiol. 149:67-74; 1987.
Marcus, J. M.; Swearingen, G. R.; Williams, A. D.; Heizer, D.
D. Polynuclear aromatic hydrocarbon and heavy metal
concentrations in sediments at coastal South Carolina
marinas. Environ. Contam. Toxicol. 17:103-113; 1988.
May, W. E.; Wasik, S. P.; Freeman, D. H. Determination of
the solubility behavior of some polynuclear aromatic
hydrocarbons in water. Anal. Chem. 50:997-1000; 1978.
McCarthy, J. F. Role of participate organic matter in
decreasing accumulation of polynuclear aromatic hydro-
carbons by Daphnia magna. Arch. Environ. Contam.
Toxicol. 12:559-568; 1983.
Means, J. C.; Wood, S. G.; Hassett, J. J.; Banwart, W. L.
Sorption of polynuclear aromatic hydrocarbons by
sediments and soils. Environ. Sci. Technol. 14:1524-1528;
1980.
National Research Council. Polychlorinated Biphenyls. 1979;
182 p. Available from the National Academy of Sciences.
O'Connor, J. M.; Huggett, R. J. Aquatic pollution problems,
North Atlantic coast, including Chesapeake Bay. Aquat.
Toxicol. 11:163-190; 1988.
Opperhuizen, A.; Seme, P.; Van der Steen, J. M. D. Thermo-
dynamics of fish/water and octanol/water partitioning of
some chlorinated benzenes. Environ. Sci. Technol. 22:286-
291; 1988.
Reynoldson, T. B. Interactions between sediment contami-
nants and benthic organisms. Hydrobiol. 149:53-66; 1987.
Schimmel, S. C.; Patrick, J. M., Jr.; Faas, L. F.; Oglesby, J.
L.; Wilson, A. J., Jr. Kepone: toxicity and bioaccumula-
tion in blue crabs. Estuaries 2:9-15; 1979.
Spacie, A.; Hamclink, J. L. Alternative models for describing
the bioconcentration of organics in fish. Environ. Toxicol.
31
-------
PLENARY SESSION
Chem. 1:309-320; 1982.
Trucco, R. G.; Engelhardt, F. R.; Stacey, B. Toxicity,
accumulation and clearance of aromatic hydrocarbons in
Daphniapulex. Environ. Poll. Ser. A, 31:191-203; 1983.
Varanasi, U.; Reichert, W. L.; Stein, J. E.; Brown, D. W.;
Sanborn, H. R. Bioavailability and biotransformation of
aromatic hydrocarbons in benthic organisms exposed to
sediment from an urban estuary. Environ. Sci. Technol.
19:836-841; 1985.
Veith, G. D.; Call, D. J.; Brooke, L. T. Structure-toxicity
relationships for the fathead minnow, Pimephales
promelas: Narcotic industrial chemicals. Can. J. Fish.
Aquat. Sci. 40:743-748; 1983.
Veith, G. D.; DeFoe, D. L.; Bergstedt, B. V. Measuring and
estimating the bioconcentration factor of chemicals in fish.
J. Fish. Res. Board. Can. 36:1040-1048; 1979.
Voice, T. C.; Weber, W. J., Jr. Sorption of hydrophobia
compounds by sediments, soils and suspended solids.
Water Res. 17:1433-1441; 1983.
Willford, W. A.; Mac, M. J.; Hesselberg, R. J. Assessing the
bioaccumulation of contaminants from sediments by fish
and other aquatic organisms. Hydrobiol 149:107-111;
1987.
Wood, L. W.; Rhee, G.; Bush, B.; Bernard, E. Sediment
desorption of PCB congeners and their bio-uptake by
dipteran larvae. Water Res. 21:875-884; 1987.
32
-------
GENETICS AND SPECIES CONSERVATION
Robert Chapman, moderator
Introduction to Plenary Session
In writing the genetics chapter for Perspectives, we homing extensively in the Bay, whereas the males
were not able to cover all the areas of interest, and the appeared to wander. This apparent divergence of
speakers for this session were chosen in the hope that behavior awaited confirmation after analysis of the
they might fill some of the gaps left by the chapter. 1987 data. The 1987 data have been analyzed and they
One particular point that was mentioned in do support the apparent divergence.
Perspectives was that female striped bass seemed to be —R.C.
33
-------
PLENARY SESSION
New Methodology for Immunologic Discrimination of Stocks or
Populations in Fish Species
Raymond Simon
U.S. Fish and Wildlife Service
Kearneysvllle, West Virginia 25430
The immunogenetic section of the National Fish
Health Research Laboratory has worked with striped
bass immunology for three years. Preliminary work
shows significant differences between populations, but
to be useful the work must reach the point where an in-
dividual fish of unknown origin can be assigned to the
population of its origin with reasonable certainty.
A variety of approaches can be categorized as
immunological, including the use of some natural
agglutinins found in the serum of pigs. These aggluti-
nate red blood cells without the addition of any anti-
body, to a different extent in different individuals and
different populations. In some Pacific salmon species,
this approach has indicated a divergence between Asian
and North American stocks. Approaches using
antiserum to purified red blood cells of Pacific salmon
species have shown variable agglutination strengths in
large assemblages from widely separated geographic
areas. Yet another approach has been the induction of
antibody production in rabbits and other animals using
the serum of various salmon species, with double
diffusion in agarose and an analysis of precipitin bands
present and missing.
These three techniques do not lend themselves to
quantitation and are extremely insensitive in compari-
son to more recently developed techniques, primarily
radioimmune assay in combination with a competitive
binding assay. This approach, in use since 1980,
focuses not on the division of breeding groups within a
species, but rather on a taxonomic assessment of the
identity of the species. Serum proteins are used, as
these secreted proteins display many more readily-
acceptable amino acid substitutions than do more
functional proteins such as enzymes.
The competitive binding assay (Figure 1) has two
steps. In the first step, the antibody that has been
prepared against a purified protein is incubated with an
unknown, unpurified serum that contains that antigenic
protein. If the intact serum is identical to the source
from which the antibodies were prepared, then all the
antibody will be bound; whereas if there is lesser
identity, some of the antibody against which the anti-
serum was prepared will be unbound and available for
subsequent reactions.
In the second step a competition mixture of antigen,
antibody, and antigen-antibody complex is poured into
a receptacle coated with the antigen used to prepare the
antibody. If free antibody remains that has not been
consumed by the antigen excess, it is available to react
with the antigen that has been immobilized on the
receptacle. That amount can be quantitated by using a
second antibody, labeled with either a radioactive
CLOSELY RELATED
ANTIGEN-ANTIBODY
COMPIZX IS
DISTANTLY RELATED
Figure 1. Diagram of amour:' of antibody bound with
different degrees of similarity to reference antigen, where
much antibody remains unbound when the relationship is
remote and more is consumed with relationships of increasing
similarity.
34
-------
Genetics and Species Conservation
isotope or an enzyme for which specific substrate and
cofactors can be introduced, and the resulting color
development (corresponding to the amount of protein
bound) can be measured spectrophotometrically.
Although this approach works well there are
complications. The first level of complication is that
two populations may produce individuals with identical
absorbance readings. The situation is analogous to
examining a polyacrylamide gel from the end rather
than from the side — three bands at entirely different
locations may still produce the same absorbance. The
advantage of this system, however, is that the antigen-
antibody complex can be removed with the use of
affinity columns (provided that the antigen is not
identical to the antigen against which the antibody was
prepared), so that the effluent contains only unique
antigen recognition sites. That portion of the protein
now is unique to the population with regard to another
population. In examining larger numbers of popula-
tions, a battery of immune reagents is necessary. It is
also necessary to be careful about defining reference
standards, since a sample of any geographic population
is likely also to contain individuals that originated else-
where. The finding that males are much less likely than
females to return to their place of origin should caution
striped bass researchers to segregate their data on the
basis of gender.
RELEVANT LITERATURE
Chapman, R.W. 1987. Changes in the population structure of
male striped bass, Morone saxatilis, spawning in the three
areas of Chesapeake Bay from 1984 to 1986. U. S. Fish
and Wildlife Service, Fish. Bull. 85(1): 167-170.
Gushing, J.E. 1956. Observations on serology of tuna. U.S.
Fish and Wildlife Services, Special Scientific Report,
Fisheries (183). 14pp.
Fabrizio, M.; Saila, S.B. 1986. Identification of striped bass
stocks in coastal waters from New Jersey to Massachu-
setts: 1983 to 1985. Rhode Island Division of Fish and
Wildlife, Annual Progress Report, Project AFC 4-4,
Wakefield, RI.
Fujino, K. 1970. Immunological and biochemical genetics of
tunas. Trans. Am. Fish. Soc. 99:152-178.
Lowenstein, J.M.; Molleson, T.; Washburn, S.I. 1982.
Piltdown jaw confirmed as orang. Nature 299:294.
Lowenstein, J.M.; Ryder, O.A. 1985. Immunological
systematics of the extinct quagga (Equidae). Experientia
41:1192-1193.
Rago, P.J.; Richards, R.A. 1987. Emergency striped bass re-
search study report for 1986. U.S. Dept. of the Interior &
U.S. Dept. of Commerce (unnumbered document). 79 pp.
Ridgway, G.J.; Klontz, G.W. 1961. Blood types in Pacific
salmon. Int. North Pacific Fish. Comm., Bull. (5):49-55.
Ridgway, G.J., Gushing, J.E., Durall, G.L. 1961. Serological
differentiation of populations of sockeye salmon, On-
corhynchus nerka. Int. North Pacific Fish. Comm., Bull.
(3):5-10.
Ridgway, G.J., Klontz, G.W., Matsumoto, C. 1962. Intra-
specific differences in serum antigens of red salmon
demonstrated by immunological methods. Int. North
Pacific Fish. Comm., Bull (8): 1-13.
35
-------
PLENARY SESSION
Mitochondrial DNA Analysis of Chesapeake and Delaware Bay Populations
of Fundulus heteroclitus
Michael W. Smith, Robert W. Chapman, and Dennis A. Powers
Department of Biology
The Johns Hopkins University
Charles and 34th Street
Baltimore, Maryland 21218
Chesapeake Bay Institute
The Johns Hopkins University
Shady Side, Maryland 20764
The molecular approach to genetics has opened
doors to questions in population and evolutionary
biology that were previously unapproachable by
conventional methods. The analysis of proteins has
revealed extensive genetic variation in natural popula-
tions. The molecular approach is allowing us to address
questions concerning population structure, gene flow,
and evolution. Dr. Powers' laboratory at Johns
Hopkins makes use of the mummichog (Fundulus
heteroclitus) as a model marine organism to address
these issues. F. heteroclitus is found all along the
Atlantic coast from Nova Scotia to Georgia, and of
course in the Chesapeake Bay. In sampling along the
coast and determining the gene frequencies for lactate
dehydrogenase B (Ldh-B), it has been shown that one
allele is fixed in Florida and another in Maine. Similar
clines have been observed for many other loci.
Two models have been proposed to explain these
clines. One hypothesis is that the populations have
always been in contact with each other, and the clines
are the product of chance and adaptive forces, selection,
and nonrandom migration. Alternatively, the clines
could result from historical isolation of the two popula-
tions with subsequent removal of the isolating barrier
and resumption of contact. (This is known as secon-
dary intergradation.) Recent work with mitochondrial
DNA by Gonzalez and Powers has allowed us to
eliminate the first possibility. This research has shown
that the Maine type fish and the Georgia type fish are
very different, having been separated sometime during
the last three to six million years.
PROPERTIES OF MITOCHONDRIAL DNA
Mitochondrial DNA (mtDNA) is a small circular DNA
found in the mitochondria of all eukaryotes. In F.
heteroclitus it is about 17 kB in size. It is maternally
inherited, allowing the tracing of maternal lineages.
The molecule evolves rapidly and variation is sensitive
to small population sizes.
The mechanics of the technique involve isolating the
mitochondria by cell fractionation and subsequent
extraction of the nucleic acids. Then a restriction
enzyme that recognizes specific nucleotide sequences is
used; it cleaves the mtDNA where those sequences
appear, generating fragments that can be sized on an
agarose gel. The Fundulus mtDNA shows two major
patterns. For analysis a battery of restriction enzymes
is used to generate a "fingerprint" of the mtDNA. The
five restriction enzymes usually used to distinguish
between northern and southern types of Fundulus in the
Bay detect a minimum difference of six nucleotide
changes. Gonzalez and Powers have shown that there
are two main F. heteroclitus mtDNA patterns along the
Atlantic coast. The current zone of secondary intergra-
dation is found near northern New Jersey, with some
mtDNA variation within the groups and substantial
differences between the two groups.
REGIONAL HISTORY
Fifteen thousand years ago the Chesapeake Bay had not
yet been formed. The coastline was near the continen-
tal shelf, and the glaciers extended to northern New
Jersey. A relevant question is whether the zone of
secondary intergradation ever existed further south-
ward. If so, it would be reasonable to expect some zone
of secondary integradation in the Delaware and/or
Chesapeake Bays. To answer this question we exam-
ined 540 fish from 20 locations, typing them for
36
-------
Genetics and Species Conservation
mtDNA and nuclear-encoded enzymes. Isolated
refugia of F. heteroclitus with northern type mtDNA
were found in five areas of the Bay. Eleven of the
populations had only southern or northern mtDNA
types, and the remaining nine populations contained
both mtDNA types. Populations within a river are very
different. In fact, populations in the upper James and
upper Potomac resemble each other more closely than
they do the populations in the lower portions of their
respective rivers. The presence of the northern type
mtDNA fish in these rivers and in the upper Bay imply
that these fish once occupied the entire Bay and were
probably present at the mouth of the Bay when it was
initially forming.
Similar clines are also observed in the gene frequen-
cies of northern and southern alleles of allozyme loci.
For example, Ldh-B allele frequencies have similar
clines but none of the populations are fixed for either
allele.
The population in the upper Potomac, near Mount
Vernon, is fixed for the northern mtDNA type but
contains the southern Ldh-B alleles. This pattern could
indicate that "southern" males are dispersing and
breeding with more sedentary females. Conversely,
populations near the mouth of the Bay are fixed for
southern mtDNA and contain the northern Ldh-B
alleles, presumably contributed by northern males
dispersing and breeding. These patterns suggest that
the males are heavily involved in gene flow in F.
heteroclitus. Alternatively, selection or genetic drift
could explain these patterns.
CONCLUSIONS
Both northern and southern races of Fundulus inhabit
the Chesapeake and Delaware Bays. The northern race
was found in refugia in the upper bays and in the upper
James, Potomac, Big Choptank and Patuxent rivers; the
southern race was found elsewhere in the bays. If more
rivers had been sampled, it is likely that more refugia
would have been found, as every river sampled in the
upper tidal region had a refugia. These isolated refugia
result in multiple zones of secondary intergradation.
We conclude that the northern race fish on the Atlantic
coast once lived as far south as the present mouth of the
Chesapeake Bay.
These studies emphasize two points. One, the his-
tory of these populations is connected to previous glaci-
ations. The northern race either was present or came
into the Bay as the glacier retreated, and the southern
race then followed. Thus the history of glaciation has
had a strong influence on the biota present in the Bay
and their genetics. Two, this work with F. heteroclitus
and Chapman's data on striped bass have shown th.
mtDNA research can provide valuable insights into the
living resources in the Bay and elsewhere.
QUESTIONS
Q: Could mtDNA techniques be used to determine
how much interchange there is between, for instance,
the populations in the lower James and the population
in the lower Potomac?
A: Yes, with direct sequencing or enzymes that
recognize more sites.
Q: Have you considered that the difference in mtDNA
frequencies could be related to salinity differences?
A: It could be that the northern-type mtDNA popula-
tions are adapted to lower salinity, as their refugia are
generally of lower salinity; but the northern type also
lives in high-salinity water in Maine.
Q: Fundulus is usually thought of as a territorial
species. Do you agree that these data suggest that the
females are more territorial?
A: I think the data suggest that males could be dispers-
ing more, but to say that females are more territorial is a
further extrapolation from the data.
Q: Could a freshwater species of Fundulus be respon-
sible for the northern type of mtDNA?
A: The mtDNA of Fundulus diaphanus, a sympatric
species in upper tidal areas, has a very different pattern
from that of F. heteroclitus. Granted, there are many
species of Fundulus, and the mtDNA could have come
from one of them, but I think that's unlikely since we
find the same northern mtDNA pattern in Maine.
Q: For your scenario you must postulate that there is
no larval transport downriver.
A: These fish are generally in very still water, in
marshy areas, and for them to be transported downriver
they would have to venture into water that for them
would be very rough.
Q (followup): Has any data been collected on the
salinity tolerances of eggs and larval stages?
A: There was some work done on salinity tolerances.
Unfortunately the authors were not aware that they
were dealing with two different races, so comparisons
are difficult.
A: One of the things we don't know is how far south
one can find the northern type of F. heteroclitus in a
river. It is possible, for example, that the Cape Hatteras
area, where there are a lot of currents, could be where
the original isolating barrier was located.
37
-------
PLENARY SESSION
Genetic Analysis of Oysters that Grow at Different Rates
Ken Paynter
Chesapeake Bay Institute
4800 Atwell Road
Shady Side, Maryland 20764
Two years ago I met a man who claimed he could
grow oysters in less than a year. He grows them
cultchless in a floating raft culture. They look different
from native oysters, with a pointed umbo and a very
deep valve on one side and a very flat valve on the
other. These inbred oysters also have a much higher
percentage meat than native oysters, because native
animals have a much thicker shell. As you would
expect, the inbred animals show little variation in allele
frequencies. Differences in allele frequencies are
evident in native vs. inbred oysters.
Because the animals are so different, we wanted to
study them. The inbred oysters are grown in a cultch-
less manner that the grower maintains as a secret. We
were able to use epinephrine to induce the cultchless
metamorphosis of eyed larvae. We tested grow-out of
these oysters in floating rafts. Tens of thousands of
oysters can be grown in this way in a small space. We
generated larvae from native animals and from some
broodstock of the inbred strain and raised them in
floating rafts in a tidal creek and compared the growth
rates.
At 3.5 months the animals were all just short of 2
inches long. The inbred animals grew faster than the
native ones. The growth rate for the inbred strain wa^
at least 50% higher during the juvenile period and abc...
20% higher at the end of a four-month period. Both
sets of oysters grew very well in the cultchless floating
raft system; most of the animals would have reached
market size in less than a year. It appears this approach
would be well-suited to aquaculture.
Why are these animals different in their growth
rates? We approached the question genetically and
compared large and small representatives (1-2 SD
below and above mean size) of both strains.
Several alleles present in the native population were
absent in the inbred population, presumably due to loss
during inbreeding. Within groups, one large difference
is found at Lap-2C. None of the very large animals
contained a C allele. Otherwise the size groups within
the strains were not very different.
Analysis of variance with respect to weight and
length vs. strain and genotype showed that two genes
appear to affect growth in the inbred strain, and in the
native animals, four or five loci have an effect. So far
only Lap-2 can be associated with a specific effect:
animals with the C allele do not do as \vell as animals
without it. We cannot say that growth rate is inhibited
by the enzyme.
Genetic analysis of oysters with respect to growth
rate and other measures of fitness may be a productive
strategy in learning more about the basic biology of
oysters; this kind of analysis may suggest useful
management strategies for shellfish and perhaps other
species in the Bay as well.
QUESTIONS
Q: Work has been done at VIMS on developing
cultchless oysters by removing them from cultch once
they have set.
A: Yes, this work is different in that the epinephrine
induces what is thought to be a hormonal response, a
willingness to metamorphose without substrate.
Q: Your work is very similar to work on Long Island
with Mytilis edulis, which has shown that Mytilis edulis
with a certain genotype is less efficient at using or
conserving protein.
A: By chance or not, the same locus is involved there
as here.
Q: Osmoregulation may also be associated with this
allele; what effects have been seen wiih alteration of the
salinity?
A: We have not yet tried altering the environment, but
one of the beauties of this sytem is that the environment
can be manipulated to a great degree.
38
-------
BENTH1C-PEIAGIC COUPLING
Jonathan Garber, moderator
Introduction to Plenary Session
The idea that waters and sediments of aquatic sys-
tems are linked by exchanges of energy and nutrients is
not new. References to the concept can be found in the
limnological literature in the late 1930s, and the founda-
tions for the ideas were set down even earlier in the
century. What has developed in the last 10 years is an
appreciation of the importance of these exchanges in
shallow marine ecosystems. Nixon, for example, has
hypothesized that nitrogen availability and ultimately
the productivity of these systems is regulated by the
processing of planktonic organic matter in benthic
communities. Kemp and Boynton have emphasized the
importance of nutrient cycling at estuarine interfaces,
and one of the important interfaces is between the
sediments and the waters. Sanders in his presentation
yesterday said that in addition to the exchanges of
organic matter and nutrients, exchanges across the
sediment-water interface can influence the bioavailabil-
ity of toxics and other xenobiotic compounds.
The impetus, however, for the Perspectives article
and the presentation today, came from those charged
with modeling water quality in the Bay. They discov-
ered that without feedback loops between the waters
and the sediments, the Bay just didn't work right. Our
purpose today is to continue trying to understand how
these processes fit together and how the waters and
sediment are coupled.
The talks that follow will be organized around a
general conceptual model (Figure 1); they will focus on
the mid-mesohaline portion of the Bay; and they will
aim toward assembling quantitative budgets. The
model is divided into three parts: the formation of or-
ganic matter, the processing of that matter in the water
column and its sinking; and its fate on the bottom.
— J.G.
\ Sources of "New" Nutrients
Rivers, Runoff, Ra'nfal, Sewage, Marshes
Sediment
Oxygen
Demand
Water
Sediment
Figure 1. Conceptual model of nutrient element flows in plankton-based systems.
39
-------
PLENARY SESSION
Production of Particulate Organic Matter in the Mesohaline Reach of the Chesapeake Bay
Thomas Malone
Horn Point Environmental Laboratory
University of Maryland
P.O. Box 775
Cambridge, Maryland 21613
One of the most important conclusions that emerged
from early work on this problem (a conclusion that is
now debatable) was that most of the input of organic
material into the Bay downstream of the turbidity
maximum is accounted for by phytoplankton. More
recently, work by Ducklow, Peele, Tuttle, and others
has shown that a large fraction of this production
appears to be cycled through bacterioplankton. Thus
we can view the production and fate of paniculate
organic material in the water column as a function of
two interacting networks of flow: (1) a phytoplankton-
based network, which accounts for most biomass yield
and paniculate organic matter flux to metazoan
consumers; and (2) a bacterioplankton-based network
(the "microbial loop"), which accounts for most internal
flows (to the water column) and most of the flux of
paniculate organic matter to microbial consumers
(within the water column). In the context of benthic-
pelagic coupling, it is likely that the phytoplankton
network dominates the flux of organic matter to the
benthos while bacterioplankton are more directly
involved in the rapid turnover of carbon and associated
nutrients in the water column.
This presentation addresses the problem of the
response of phytoplankton to nutrient input. Hugh
Ducklow will speak later today in a concurrent session
on the bacterial component of the system. Four points
important to the topic are:
• Distributions of phytoplankton productivity and
biomass in both time and space;
• Environmental factors that are responsible for this
variation;
• The fate of phytoplankton production as deduced
from patterns in the distribution of phytoplankton;
• Nitrogen vs. phosphorus limitation of the system as
a whole.
The area we have concentrated on is the mesohaline
reach of the Bay between the Bay Bridge and the
Patuxent River. Many properties vary longitudinally
down the axis of the Bay from the freshwater areas
down to the sea, and there are good reasons for focus-
ing on the mesohaline reach. This is where most of the
nutrient input delivered by the Susquehanna River is
assimilated and where accumulations of phytoplankton
biomass are highest.
Harding et al. [1986] recently found that integral
phytoplankton production over the water column could
be described as a function of chlorophyll concentration
and the amount of light absorbed by phytoplankton in
the euphotic zone. Thus the variation in integral pro-
duction down the axis of the Bay can be parameterized
in terms of biomass and light as a first approximation.
Lateral variability has been studied by a number of
people, with the earliest probably being Flemer in the
late 1960s. The findings then were similar to the 1984
findings: biomass (as indexed by chlorophyll a) was
higher on the western shore. More recent studies
[Malone et al. 1986] also show that the production per
unit of chlorophyll is generally higher on the eastern
shore. It is possible that on the western shore biomass
is accumulating while on the eastern shore it is turning
over rapidly, perhaps because of removal by benthic
and pelagic consumers.
SEASONAL VARIABILITY
Annual cycles of phytoplankton biomass and
productivity
Annual cycles of phytoplankton biomass and produc-
tivity in the mesohaline reach of the Bay are seasonally
out of phase, with biomass peaking during spring and
productivity peaking during summer (Figure 1). Large
interannual variations in phytoplankton productivity
also occur (Figure 1). Although there is evidence that
phytoplankton productivity is phosphorus-limited
during spring and nitrogen-limited during summer
[D'Elia et al. 1986], annual phytoplankton production
appears to be more sensitive to N than to P loading
(Figure 2), and the Bay as a system removes the inor-
ganic N input more efficiently than the P input [Fisher
et al. in press] (Tables 1 and 2). The major external
source of N to the mesohaline reach of the Bay is the
40
-------
Benthic-Pelagic Coupling
Figure 1. Upper panel: annual cycles of
Susquehanna River discharge (D),
chlorophyll a content of the water column
( + ), and phytoplankton productivity (<0>)
of the euphotic zone (monthly averages
over the mesohaline reach of the Bay for
1986-1987 as a fraction of the maximum
monthly means). Lower panel: interan-
nual time series of phytoplankton produc-
tivity in the mesohaline reach of the Bay.
ANNUAL CYCLES OF FLOW. CHL, PP
1f
1.1 •
1 •
OJ-
U-
•*-
•.1 •
PHYTOPLANKTON PRODUCTIVITY. 1972-1987
UHJMBCT tr M. * tMUNE B M.
Susquehanna River, which supplies >90% of the inor-
ganic N input and 70-80% of the total N input [Schubel
and Pritchard 1986; Fisher et al. in press]. As is typical
of mid-latitude rivers, the annual cycle of freshwater
discharge exhibits a spring maximum and a summer
minimum. Consequently, 50-60% of the annual N in-
put occurs during the spring freshet which precedes the
spring peak in biomass by about one month (Figure 1).
Comparison of variations in nitrate input from the
Susquehanna and total phytoplankton biomass in the
mesohaline reach of the Bay suggests that seasonal and
interannual variations in phytoplankton biomass are
related to the riverine input of N (Figure 2). Water
column and euphotic zone levels of biomass (as nitro-
gen, g N) are significantly correlated with the supply of
nitrate-N (Q-N) by the least square regressions:
(1) water column phyto-N = 4.6 + 8 (Q-N) (r = 0.93)
(2) euphotic zone phyto-N = 3.5 + 4 (Q-N) (r = 0.88)
As indicated by the slopes of these regressions, a unit
increase in nitrate-N input results, on average, in a four-
fold increase in phytoplankton biomass in the euphotic
zone and an eightfold increase in the water column as a
whole. While these calculations must be considered
rough approximations, they are indicative of the effects
of two-layered estuarine circulation and N recycling on
the accumulation of phytoplankton biomass within the
mesohaline reach of the Bay. The development of the
spring biomass maximum appears to be due to sedimen-
tation and accumulation of biomass and to the advec-
tion of biomass with bottom water from downstream.
41
-------
PLENARY SESSION
The annual cycle of phytoplankton productivity is
characterized by a winter minimum and a summer
maximum (Figure 1). Consequently, nitrogen assimila-
tion by phytoplankton is inversely related to N supply
(Figure 2). This suggests the importance of light and
temperature as parameters of phytoplankton productiv-
ity, as Harding et al. [1986] have shown. However, as
reported by Boynton et al. [1982], the magnitude of the
summer productivity maximum shows large interannual
variations, which are related to variation in phytoplank-
ton growth rate. These variations were correlated (r =
0.96) with vertical salinity stratification. This correla-
tion and the results of nutrient enrichment studies
indicating that phytoplankton growth is likely to be N-
limited during summer [D'Elia et al. 1986] support the
hypothesis that the magnitude of the summer productiv-
ity maximum is a function of physical processes that
regulate the recycling of ammonium from below the
pycnocline into the euphotic zone [Malone et al. 1986].
Thus the annual cycle of phytoplankton productivity
in the mesohaline reach of the Bay is governed by light,
temperatures, and nitrogen recycling. Seasonal and
interannual variations in phytoplankton biomass occur
in response to variations in freshwater flow and
associated variations in nutrient flux. The occurrence
of a summer productivity maximum and interannual
variations in the magnitude of this maximum reflect the
effects of vertical stratification on the return flux of
ammonium from the benthos to the euphotic zone.
FATE OF PHYTOPLANKTON PRODUCTION
The observation that biomass accumulates during
spring while growth rates are relatively low and that
most of the accumulation occurs below the euphotic
zone implies that the removal rate of phytoplankton is
also low. Measurements of chlorophyll degradation
products of primary consumers are consistent with this
conclusion and suggest that the rapid decline of the
spring bloom during late May and early June is a
consequence of increased grazing pressure. Vertical
distributions of chlorophyll a, low grazing pressure, and
the dominance of diatoms also suggest that the vertical
flux of phytoplankton biomass to the benthos should
exhibit a seasonal maximum in the spring. The rapid
turnover of phytoplankton biomass during summer
probably reflects the combination of high phytoplank-
ton growth rate and grazing pressure [White and Roman
1988], which suggests that vertical flux during this
period is dominated by fecal material of phytoplankton.
Several things can be learned from looking at the
variations in chlorophyll and the percentage of nitrogen
in the dissolved organic nitrogen (DON) pool vs. the
nitrate pool, down the axis of the Bay along the salinity
PRODUCTION vi N-tOADMO
600
400
o
o
200
1»1072
1070
2' 1.
1806
1*1874
1*1075
1060.
1077* »6
13 , »,
tl4 12 ,
.10
10
20 30 40
0 N m-' yr-i
SO
4X)
30
a 20
*>
o
10
0
10
•D 6
Z
d
*o 4
o
O .00
01234
RIVER INPUT, 108g N d-1
Figure 2. Relationships between (upper panel) annual
phytoplankton production and nitrogen loading and (lower
panels) monthly mean phytoplankton biomass (10* g N,
euphotic zone = 0, water column = Q ), nitrogen assimila-
tion (10'g N d •'), and riverine nitrate input (from Malone et
al., in press).
gradient [MCarthy et al. 1977, Fisher et al., in press]. It
appears that most of the nitrogen in the water column
ultimately ends up in the DON pool. Total nitrogen
decreases, but in the water column there is a shift from
mainly nitrate to mainly DON. This would imply that
nitrogen is being exported from the. Bay in the form of
DON or is lost via denitrification and burial. The ratios
of total nitrogen to phosphorus decrease along the axis
of the Bay, from very high ratios (>60) to ratios
approaching Redfield. This change implies a net loss of
42
-------
Benthic-Pelagic Coupling
Table 1. Nutrients in Chesapeake Bay and its coastal plume: changes in distribution and stoichiometry of N and P
from freshwater to the sea [Fisher et al., in press].
Fresh
Sea
TN(jjM)
129
18
TPQuM)
1.4
1.4
NO3
67
3
TON (%)
28
83
P04
27
31
TOP (%)
73
69
Inorganic Organic
235 35
6 17
approaching Redfield. This change implies a net loss of
nitrogen relative to phosphorus, and phosphorus is
actually exported from the Bay. The concentration of
phosphate in the freshwater endmember and that in the
seawater endmember are similar as are the percentages
of inorganic and organic P. Phosphorus recycles
rapidly as it goes down the axis, but its distribution
between pools does not change significantly. This
suggests that the community metabolism of the Bay as
a whole is nitrogen-limited.
QUESTIONS
Q: We have little data on seaweed and microbenthic
algae, but some of these species might also be a major
source of organic material in the system, especially the
detritus feeders. The microbenthic bloom in the spring
does not occur at the same time as the phytoplankton
bloom, and there are other discrepancies. There could
be another whole system here that we know very little
about.
A: I agree that the conclusion of Flemer and Biggs
(that phytoplankton are the major source of organic
material downstream of the turbidity maximum) needs
to be reevaluated, especially in the light of your
comment. But in terms of the accumulation of phyto-
plankton biomass, there is so little light reaching the
bottom at the time of the phytoplankton bloom that you
would not expect microbenthic production to be a major
contributor during that time.
Table 2. Nutrient levels (|iM) and ratios in the coastal plume of Chesapeake Bay [Garside, personal comm.]
Month
REFERENCES
DIN
PCX
Si (OH)4
N/P
N/Si
February
April
June
August
0.22
1.35
0.22
0.29
0.29
0.23
0.63
0.78
0.14
0.73
2.98
1.82
0.8
5.9
0.3
2.7
1.6
1.8
0.1
0.2
Boynton, W.R.; Kemp, W.M.; Keefe, C.W. A comparative
analysis of nutrients and other factors influencing
estuarine phytoplankton production. In: Kennedy, V.S.
(ed.) Estuarine comparisons. Academic Press, New York,
p. 69-90; 1982.
D'Elia, C.F.; Sanders, J.G.; Boynton, W.R. Nutrient
enrichment studies in a coastal plain estuary: phytoplank-
ton growth in large-scale, continuous cultures. Can. J.
Fish. Aquat. Sci. 43:397-406; 1986.
Fisher, T.R., Harding, L.W., Stanley, D.W., Ward, L.G.
Phytoplankton, nutrients, and turbidity in the Chesapeake,
Delaware, and Hudson estuaries. Estuar. Coast. Shelf Sci.
(in press).
Harding, L.W.; Meeson, B.W.; Fisher, T.R. Phytoplankton
production in two east coast estuaries: photosynthesis-
light functions and patterns of carbon assimilation in
Chesapeake and Delaware Bays. Estuar. Coast. Shelf Sci.
23:773-806; 1986.
McCarthy, J J.; Taylor, W.R.; Taft, J.L. 1977. Nitrogenous
nutrition of the plankton in the Chesapeake Bay. Limnol.
Oceanogr. 22:996-1011.
Malone, T.C.; Kemp, W.M.; Ducklow, H.W.; Boynton, W.R.;
Tuttle, J.H.; Jonas, R.B.; Lateral variation in the pro-
duction and fate of phytoplankton in a partially stratified
estuary, Mar. Ecol. Prog. Ser., 32:149-160; 1986.
Schubel, J.R., Pritchard, D.W. Responses of upper Chesa-
peake Bay to variations in discharge of the Susquehanna
River. Estuaries 9:236-249; 1986.
White, J.R., Roman, M.R. Grazing and egg production by
Chesapeake Bay zooplankton in spring and summer. In
Lynch, M.P., and Krome, E.G., eds. Understanding the
Estuary: Advances in Chesapeake Bay Research. Proceed-
ings of a conference, 29-31 March 1988, Baltimore, Md.
Chesapeake Research Consortium, Gloucester Point, VA.
CRC publication no. 129.
43
-------
PLENARY SESSION
Deposition of Organic Matter to the Sediment Surface
W. Michael Kemp
University of Maryland
Horn Point Environmental Laboratories
P. O. Box 775
Cambridge, Maryland 21613
My topic is the connection between participate
organic matter in the overlying water and its deposition
to the sediments. In addition, I want to discuss carbon
balancing.
The technique we are using to measure deposition is
fixed moorings of sediment trap arrays collecting at
three depths with multiple cups. The upper cup is in the
upper mixed layer, or the euphotic zone, the middle cup
is designed to be at the pycnocline, and the lower cup
(which largely measures resuspension) is in the lower
layer.
TEMPORAL DISTRIBUTION
For 1985 and 1986 we measured deposition rates,
chlorophyll standing stocks in the water column, and
the carbon production rates in the overlying waters. We
have over three years of data on deposition rates, and a
distinct annual pattern emerges: a spring pulse of
deposition of chlorophyll, a hiatus, then relatively high
deposition rates in the summer, and often high deposi-
tion in the fall (which we know little about). The
question that arises is what is the source of the chloro-
phyll that is deposited. We are measuring only
fluorometric chlorophyll and lack data on pigment
degradation. What we can address is the relationship of
the chlorophyll standing stocks, the production rates,
and the deposition patterns, particularly the pulses.
The carbonxhlorophyll ratios are relatively low in
the spring and high in the summer; they give us some
idea of the nature of the material that is being depos-
ited. When the ratio is high, we suspect that the
material being deposited has already undergone some
degradation, since chlorophyll degrades more rapidly
than carbon. The nitrogen:phosphorus ratios relative to
carbon can give some idea about nutrient limitation,
although interpretation based on these ratios should be
cautious. Throughout the spring and summer of 1984
and 1985, both the carbomphosphorus and
nitrogen:phosphorus ratios suggested a deficiency of
phosphorus. This may suggest that paniculate produc-
tion, the phytoplankton themselves, may be phos-
phorus-limited at this season.
VERTICAL DISTRIBUTION
In both 1985 and 1986, samples from the surface waters
indicated phosphorus deficiency. The relative propor-
tion of phosphorus also increases as particles fall from
the surface toward the bottom. It may be that these
particles strip the available phosphorus from the water
as they settle, probably through chemical sorption
processes. It is also worth noting (although time
precludes a discussion here today) that there are definite
differences between seston and the material in the
sediment traps.
Year to year variations are apparent in the
phosphorusxarbon and phosphorus:nitrogen ratios. A
comparison of 1984, a relatively wet year, with 1986, a
relatively dry year, shows definite separation of these
ratios, with river flow appearing as a dominant forcing
function. Associated with river flow are the changes in
stratification and in delivery of nutrients.
Euphotic zone chlorophyll and deposition of
chlorophyll are correlated, but differently in the spring
than in the summer. The relationship is still consistent
with the hypothesis to be presented by Malone and
Roman and White later in this meeting, as far as the
mechanism for delivery of phytoplanl; ton to the bottom.
In the summer the chlorophyll stocks in the water turn
over much more rapidly than in the spring, perhaps
largely because of grazing. In genera) the relationship
between primary production and deposition is weak.
Long-term well-integrated data sets (covering years)
might reveal these relationships, but at the scale we are
studying (month to month) they are not readily found.
Such a relationship was found, however, in 1984, which
was an exceptional year. A plot of primary production
on the x-axis vs. sedimentation of carbon on the y-axis,
with data partitioned for the flanks and channel, shows
44
-------
Benthic-Pelagic Coupling
a relationship in the flanks but not in the water directly
above the traps. If we interpret this mechanistically, we
must invoke deposition, resuspension, and ultimately
some lateral transport.
A plotting of production vs. water column respira-
tion for the spring-summer period shows the positive
relationship that one would expect to see, but with a
great deal of scatter.
If net daytime primary production is calculated, and
then the respiration of the upper layers and the lower
layers is .ibtracted, the residual should be the material
that is deposited on the sediment surface. That residual
correlates very closely with actual trap collections.
Either there are lots of compensating errors, or these
data are beginning to be believable. We can begin to
try to develop carbon budgets for the water column, to
understand relationships between production and
deposition, and to examine the community metabolism
in the benthos.
45
-------
PLENARY SESSION
Uptake and Release of Nutrients from Sediments
Walter R. Boynton
University of Maryland
Horn Point Environmental Laboratories
P. O. Box 775
Cambridge, Maryland 21613
The Chesapeake Bay is a very appropriate place to
study sediments. Tom Morton in his recent book Bay
Country put it as well as it can be put: "When you
think Chesapeake Bay, think skinny." We have lots of
watershed and lots of sediments, but almost no water,
even though the Chesapeake is one of the largest
estuaries, in terms of area, in the U.S. They Bay is
shallow, and hence we could expect a substantial
relationship between the waters and the sediments.
Under these circumstances it makes sense to study the
benthos, which we can think of as the memory, or the
modulator of many processes in the Bay.
The uptake and release of nutrients from sediments
has been measured at a number of stations, most of
them in Maryland. An opportunity has arisen recently
to collaborate with Virginia workers in the collection of
these data for the lower portion of the Bay. Data from
these additional stations should give us a better perspec-
tive on these processes along a substantial estuarine
gradient.
Our instrumentation is simply a box core that is
lowered to the surface of Bay sediments. The micro-
cosm that is obtained clearly shows the small but
important structures of the bottom. Water in the
microcosm is sampled every half hour for five to six
hours, and the change in concentration over time of
such variables as oxygen and nutrients is noted. Then it
is possible to calculate the flux from the water to the
sediments or vice versa.
SPATIAL PATTERNS
In some of the tributaries, the sediment oxygen demand
(SOD) is quite high — in the range of 1.5-1.7 g O2m2
d'1. In the mainstem the rates are lower, for reasons we
are not sure of. Extremely high ammonium fluxes are
seen in two places: the lower Potomac River and mid-
Bay regions. Both of these locations experience
frequent summer anoxia. The rates of ammonium flux
and SOD at these locations are comparable with rates
for very productive estuaries. Large fluxes of phospho-
rus are also seen at two stations that become anoxic.
The spatial patterns on the Bay bottom are emerging,
but slowly. Variability is high so that patterns can be
difficult to detect, but we believe that we are now
beginning to see some real patterns,such as those
mentioned above.
TEMPORAL PATTERNS
A large peak in SOD is seen in the spring, probably
largely in response to the deposition of large amounts
of material from the water column to the sediments.
SOD is lower but still substantial through the summer
and fall. This pattern is found at many of the stations.
Another common pattern, found in the upper tributaries,
is peak ammonium fluxes not in the spring but in the
summer. So there is a kind of disjointedness about
SOD being high in the spring but ammonium release
from the sediments tending to be higher in the summer,
particularly at the deep hypoxic and anoxic stations.
REGULATING FACTORS
A number of factors may be regulating sediment fluxes:
bottom water concentrations of oxygen or some other
nutrient; the rate at which organic matter is deposited to
the sediment surface; the characteristics of the sediment
itself; and temperature.
A plot of bottom water oxygen concentration vs.
SOD is so scattered that it is hard to draw any conclu-
sions at all. One point is clear, however; when oxygen
concentration in the bottom water is below about 2 mg/
1, high SOD is never found. Perhaps at this range
oxygen concentration is limiting the demand. More
important, this kind of plot tells us that we must stop
thinking only of simple, single-factor analysis.
Another paradigm is that as deposition increases, so
does SOD. Our data indicate that this could be the
case. The hint of relationship between deposition and
46
-------
Benthic-Pelagic Coupling
SOD response is heartening. Malone has talked about
forcing from the land in plankton response, and Kemp
has seen a response in deposition from plankton
characteristics. We are seeing parts of this model
coming together: some response of SOD to the magni-
tude of deposition.
There is a range of about a factor of three in the
amount of paniculate nitrogen in the surficial sediments
(approximately the top 1 cm). Does that paniculate
nitrogen in the very top sediments — recently deposited
material — send us any message regarding what enters
or leaves the sediments? The answer seems to be yes.
The flux of nitrogen from the sediments to the water
increases as the amount of paniculate nitrogen in the
sediments increases. This may imply that the sediment
memory is short, on the order of months to a year,
rather than years to decades.
We have also found that temperature and ammonium
flux are closely related, in an exponential function.
Comparison of these data in the Chesapeake with other
data from Narragansett Bay reveals the same shape, but
a different curve. The differences may stem from
differences in depositional rates. Phosphorus flux is
also exponentially related to temperature, and the
curves in the Chesapeake and Narragansett bear the
same relationship with each other.
FUTURE STUDIES
We know a little about total sediment metabolism; we
measure oxygen consumption; but there is a lot of
carbon and nitrogen and phosphorus processing in areas
of the Bay where there is no oxygen. We need to know
more about the metabolism in anoxic areas as well.
Another area for future study is nitrification and
denitrification. Denitrification is important because it is
a terminal sink for nitrogen. Since nutrient control is a
management issue, understanding the magnitude and
factors regulating these processes is of prime impor-
tance.
Finally, appropriate time scales for measurements
are all-important. Not only is it important to think
about what we are measuring, but it is also essential to
think about how often variables are measured. Appro-
priate time scale can mean the difference between
meaningless data and data that answer some questions.
47
-------
TROP"'C STRUCTURE
Peter u. Verity, moderator
The Trophic Structure of Pelagic Communities: Hypotheses and Problems
Peter G. Verity
Skidaway Institute of Oceanography
P. O. Box 13687
Savannah, Georgia 31416
INTRODUCTION
A number of factors influence the trophic structure
of pelagic communities, and their interrelationships can
be very complex. The forcing functions range from
natural environmental phenomena to anthropogenic
inputs to inter- and intraspecies interactions. The
pertinent time scales can vary from hours to decades.
The effects may be direct or indirect, they may operate
on an organism's physiology, tolerances, and resource
supply, or they may influence its predators. There are
also random fluctuations to consider, as well as
interannual variability and long-term periodicities.
Traditionally, natural populations were thought to be
limited primarily by the supply of food, habitat, or other
resources. Beginning in the 1960's, theoretical ecolo-
gists proposed that predation was as important as
resource limitation in structuring plant and animal com-
munities. The field of aquatic ecology was not immune
to these developments. For example, enhanced primary
production was long considered to be due almost exclu-
sively to eutrophication. However, limnologists quick-
ly tested the predation theory, and their results demon-
strated that the presence or absence of primary and
secondary carnivory was a dominant force in determin-
ing the magnitude of primary production. According to
this theory, changes at the top of the food web cascade
down to alter the structure at lower trophic levels.
Both of these scenarios (top-down and bottom-up
regulation) emphasize changes in abundance. Perhaps
more important to the structure of food webs are
changes in community composition. This point is often
overlooked in the current emphasis on rate measure-
ments and flux calculations, but a growing body of
evidence, both anecdotal and quantitative, suggests its
importance. To illustrate this, I have selected three
trophic groups that appear to be important in the
Chesapeake Bay (dinoflagellates, copepods, and
gelatinous predators). I have listed some of the potential
effects of their dominance, effects that may cascade
both up and down pelagic food webs. Here I should
emphasize that the proposed relationships and mech-
anisms are those thought to be operating in other
estuaries, and therefore are best viewed as food for
thought in the context of Chesapeake Bay. After
illustrating these, I will underscore some of the difficul-
ties in ascribing cause and effect in trophic responses,
which include interannual variability, long-term
climatic changes, and natural fluctuations.
DINOFLAGELLATES
The dinoflagellate class of phytoplankton is abundant in
the Bay, particularly in the summer. They have certain
physiological characteristics with significant ramifica-
tions for both pelagic and benthic food webs:
• Attenuated irradiance and subsurface nutricline
favor motile taxa [Eppley and Harrison 1975].
• Layered, reverse-flow structure favors retention of
migrating cells [Tyler and Seliger 1978,1981] and
accumulation of benthic cysts which can seed
subsequent blooms [Tyler etal. 1982].
• Some dinoflagellates excrete substances that inhibit
growth rates of diatoms and their synthesis of anti-
bacterial compounds [Aubert andPesando 1971;
Uchida 1977].
• Several species are poor food for copepods,
altering their behavioral feeding patterns and
inducing reverse peristalsis and regurgitation
[Huntley et al. 1986; Gill and Harris 1987].
• The summer-dominant copepod (Acartia tonsa)
does not feed actively on co-occurring Chesapeake
Bay dinoflagellates, and avoids vertically migrat-
ing populations [Sellner and Olson 1985].
• Blooms of inedible dinoflagellates have been im-
plicated in regulating copepod dynamics elsewhere
[Lindahl andHernroth 1983].
• Dinoflagellates are poor food for oyster larvae
49
-------
PLENARY SESSION
[Davis and Chanley 1956], perhaps due to interfer-
ence by ejection of trichocysts [Ukeles and
Sweeney 1969].
• At least one species of dinoflagellate passes undi-
gested through guts of adult oysters [Galtsoff
1964].
• Bloom concentrations of non-toxic species have
been implicated in mortality of larval fish [Potts
and Edwards 1987], perhaps due to exudates that
clog their gills [Jenkinson 1986].
• Ciliate grazing can regulate the timing and magni-
tude of dinoflagellate blooms [Watras et al. 1985].
• Dinoflagellates are good food for larval and
juvenile menhaden [June and Carlson 1971;
Friedland et al. 1984; Stoecker and Govoni 1984].
From such data, one might conclude that copepods
and oysters would have difficulty in a dinoflagellate-
dominated system. In contrast, the simultaneous
occurrence of large populations of dinoflagellates,
ciliates, and menhaden in Chesapeake Bay may not be
entirely coincidental.
SPRING BLOOM HERBIVORY
The spring diatom bloom in the Bay seems to occur at a
time and place in which an abundant crustacean
zooplankton community is seldom present. Observa-
tions relevant to this trophic mismatch include:
• Phytoplankton biomass peaks in the mesohaline
mainstem in late spring [Malone 1987].
• Total copepod abundance is low and declining
[Allan et al. 1976; Burton et al. 1986; Brownlee
and Jacobs 1987].
• The winter-dominant copepod Eurytemora affinis
is apparently caught in a temperature-salinity
squeeze [Roddie et al. 1984].
• Salinities are generally too low for Acartia hudson-
ica [Jeffries 1962] and, where salinities are accept-
able, temperatures are too warm [Sullivan and
McManus 1986].
• Spring temperatures are too low for the copepod
Acartia tonsa.
• Rotifers and protozoans are seasonally abundant
but spatially heterogeneous [Allan et al. 1976;
Brownlee and Jacobs 1987]; their distribution with
respect to the spring bloom has not been described.
The interesting point is that in estuaries to the south
of Chesapeake Bay, temperatures are warm enough for
Acartia tonsa to dominate year-round. Thus, one might
speculate that the minimal grazing of the spring bloom
by copepods may be partially determined by the
geographic location of the Bay. If it were located
further north, either Eurytemora or Acartia hudsonica
would do fine, depending on salinity; and if it were
further south, the euryhaline Acartia tonsa would
dominate. This apparent absence of significant pelagic
grazing by crustaceans is a point I will return to later.
GELATINOUS PREDATORS
Evidence that ctenophores and medusae can drastically
reduce crustacean zooplankton stocks is extensive
[Huntley and Hobson 1978; Moller 1979; Deason and
Smayda 1982]. They appear to be abundant in Chesa-
peake Bay, especially in summer, where their impact
may be considerable.
• Associated cascading effects on other trophic
levels may include diminished survival of fish
larvae and reduced recruitment [Greve and Reiners
1980; Moller 1980; Parsons and Kessler 1987], and
elevated phytoplankton production [Deason and
Smayda 1982].
• Interannual variability in gelatinous predators in
Narragansett Bay was strongly correlated with
crustacean zooplankton biomass immediately prior
to their development [Deason and Smayda 1982].
• Survival and growth of juvenile ctenophores and
medusae is enhanced by an abundant protozoan
community [Stoecker et al. 1987].
• Most oyster larvae in Chesapeake Bay could
potentially be removed by ctenophore predation.
In Barnegat Bay [Nelson 1925] ctenophores con-
tained an average of 14 larvae = daily ingestion of
168-336 larvae/ctenophore, using appropriate gut
residence times [Reeve 1980; Sullivan and Reeve
1982]. Oyster larvae concentrations in Choptank
River were 12 x 103/m3 [Seliger et al. 1982].
Ctenophore concentrations are 10-100/m3 (Miller
1974; Kremer and Nixon 1976]. Combining these
data suggests a potential daily ingestion of 14-28%
to 140-280%.
Two important conclusions can be derived from
these considerations. First, the magnitude of gelatinous
predator stocks is a function of the food available at the
time of their appearance. This correlation has potential
implications for the relationship between phytoplankton
species composition and zooplankton success. Second,
young ctenophores and medusae show better survival
and growth on diets of microzooplankton than on diets
of crustacean zooplankton. Since dinoflagellates are a
good food for microzooplankton, a direct trophic link
may exist between dinoflagellate blooms and large
stocks of gelatinous predators.
VARIABILITY
I emphasize that these proposed trophic relationships
are hypothetical, at least in the context of the Chesa-
-------
Trophic Structure
High winter/spring
stream/low
High nutrient input
Relevant studies: (1, 2) Malone 1987; (3, 4) Tuttleetal 1986
(5) Malone 1987; (6) Smayda 1984; (7) Davis 1984, Anderson and
Sorensen 1986; (8) Verity 1987; (9) Turner and Anderson 1983
Stoecker and Sanders 1985; (10) Stoecker et al. 1987, Verity and
Smayda in preparation, (11-13) Deason and Smayda 1982.
* Asterisks indicate correlations that have not been demonstrated
in the field but that have been suggested by these studies.
Large spring
phytaplankton bloom
(3)
->
Large bacterial
community
Large summer
anoxia
Large early summer
zooplankton bio mass
f
(9)
Large proto/microbial
food web
Jr
Large summer
gelatinous predator stocks
(5)
Jr
Small summer crustacean
zooplankton stocks
Large summer
phytoplankton bloom
Continued anoxia?
More menhaden ?
Figure 1. Hypothetical trophic relationships among major plankton components in the Chesapeake Bay during spring and
summer in years with high freshwater runoff. & r &
peake Bay. The difficulty in ascribing cause and effect
reflects the difficulty in identifying the number of
dependent and independent variables operating in
natural systems at any given moment, and in regulating
them during experimental studies. As an example,
consider the effects of interannual variability on these
processes (Figure 1).
Nutrient input to the Bay is thought to be a function
of streamflow: years with high flow introduce more
nutrients than low-flow years. The magnitude of the
spring bloom is related to this input. Bacterial produc-
tion is coupled to phytoplankton production, and their
degradation of the settled bloom is thought to be a pri-
mary driver of summer anoxia. A by-product of that
metabolic activity is regenerated nutrients, such that the
summer bloom may be determined by the magnitude of
the spring bloom, which is ultimately a function of
streamflow.
What is especially provocative about the Bay (if it is
true) is the lack of a direct trophic connection between
spring pelagic production and summer dynamics. As
discussed earlier, this may partially reflect the geogra-
phy of the Bay. Elsewhere, there appears to be a more
direct trophic connection, although the phenomenon is
51
-------
PLENARY SESSION
poorly documented. Alternatives may exist in the
Chesapeake either via grazing of bacteria by protozoans
or their direct ingestion of phytoplankton. In either
case, an active microbial/protozoan food web may sup-
port both crustacean omnivores and larval and juvenile
gelatinous predators. In any event, the factors regulat-
ing the magnitude of the early summer zooplankton
biomass may be critical, as elsewhere the magnitude of
this biomass determines the success of gelatinous or-
ganisms. In those systems, years with large stocks of
gelatinous predators are years with extensive summer
phytoplankton blooms. If this relationship is applicable
to the Chesapeake Bay, there are several possible
effects of such interannual variability on food web
structure.
If interannual variability were not enough of an ob-
stacle to interpreting changes in trophic structure, there
are also long-term climatic effects and periodicities to
resolve. Documented environmental phenomena
include:
• Declines in incident irradiance over the northeast
Atlantic, 1948 to 1965 [Gushing and Dickson
1976]
• The dark decade in New England during the 1970's
[Smayda 1984]
• Eleven-year and 180-year sunspot cycles [South-
ward et al. 1975]
• Temperature cycles [Taylor et al. 1957; Jeffries
and Terceiro 1985; Brady 1976; Gushing and
Dickson 1976; Sutcliffe et al. 1977]
• Long-term changes in wind strength [Taylor and
Stephens 1980] and circulation patterns of the Gulf
Stream, the northeast Atlantic, and the Mediterra-
nean [Taylor and Stephens 1980; Vucetic 1983]
• Periodicity in precipitation and runoff [Biggs and
Smullen 1987]
Proposed effects of these phenomena on pelagic
food webs include changes in:
• Timing, magnitude, duration, and community
composition of phytoplankton blooms and of
zooplankton responses [Gushing and Dickson
1976; Southward 1983; Smayda 1984]
• Overwintering survival of ctenophores [Frank
1986]
• Fecundity of planktivorous fish [Tanasichuk and
Ware 1987]
• Recruitment and year-class strength of planktivo-
rous and piscivorous fish [Gushing and Dickson
1976]
• Yield of commercial fisheries [Sutcliffe et al 1977]
• Shifts in dominance from fish to gelatinous
predators [Lindahl andHemroth 1983]
Finally, natural fluctuations can also alter the
structure of pelagic food webs. Perhaps the best
example is from the English Channel, where phyto-
plankton, crustacean and gelatinous zooplankton, and
demersal and pelagic fish showed dramatic changes in
community structure associated with climatic variabil-
ity (Figure 2). As changes in current patterns caused
warmer water to invade the English Channel for 30-40
years, the cold-water flora and fauna of the 1920's were
replaced by warm-water species in the 1950's. Nutrient
concentrations declined, large calanoid copepods
became rare, the chaetognath Sagitta elegans was
replaced by S. setosa, and herring disappeared and were
replaced by pilchards. In the 1970's, cold water
returned and the plankton reverted to that characteristic
of the 1920's. What is provocative is that the herring
did not return, but instead the pilchards were replaced
4,000r
3.000
2,000
1,000
Sagitia elegans
mean for May-August
1920
1930
1940
1950
I960
1970
12.0001
10,000
8.000
6.000
4,000
2,000
Pilchard eggs
mean for April-July
0 --.
1920
ll 1
1930
940
ill
1950
,11
I960
ill,,
1970
SOO-i
400-
2 300-
200-
100-
Young fish and
phosphate
1930
T940
06 -_
c.
if
04 1
on
ji
a.
'2*
1950
I960
1970
Figure 2. Long-term (1920-1979) changes in plankton com-
munity structure in the English Channel [Southward 1980].
52
-------
Trophic Structure
by mackerel. This change is thought to be due to
coincidence between the timing of the return of favor-
able conditions and natural fluctuations in herring and
mackerel recruitment Such a hysteresis underscores an
assumption that is often inherent in ecosystem analysis
and in management strategies, that if the addition of a
species or a change in environmental forcing has a
certain effect, then its removal will reverse that effect.
These kinds of data emphasize, however, that different
constraints may be operating at that time, and new
species interactions may be more persistent than
previous ones. Thus, the structure of pelagic food webs
is perhaps best viewed as the product of processes with
strong historical components, such as eutrophication,
and those processes requiring more continual expres-
sion, such as predation. Our challenge for the future is
to quantify and predict their relative roles in regulating
the trophic structure of pelagic communities.
RELEVANT LITERATURE
Allan, J.D.; Kinsey, T.G.; James, M.C. Abundances and
production of copepods in the Rhode River subestuary of
Chesapeake Bay. Ches. Sci. 17:86-92; 1976.
Andersen, P.; Sorensen, H.M. Population dynamics and
trophic coupling in pelagic microorganisms in eutrophic
coastal waters. Mar. Ecol. Prog. Ser. 33:99-109; 1986.
Aubert, M.; Pesando, D. Telemediateurs chemiques et
equilibre biologique oceanique. Deuxieme partie. Nature
chemique de 1'inhibiteur de la synthese d'un antibiotique
produit par un diatomee. Revue int. Oceanogr. med.
21:17-22; 1971.
Biggs, R.B.; Smullen, J.T. Freshwater inflow to the upper
Chesapeake Bay and Potomac River: seasonal to century-
long variations. Eos 68:1733; 1987.
Brady, D.K. Are the Chesapeake Bay waters warming up?
Chesapeake Sci. 17:225-227; 1976.
Brownlee, D.C.; Jacobs, F. Mesozooplankton and microzo-
oplankton in the Chesapeake Bay. In: Majumdar, S.K.;
Hall, L.V. Jr; Austin, H.M., eds. Contaminant problems
and management of living Chesapeake Bay resources.
Philadelphia: Pennsylvania Acad. Sci., in press; 1987.
Burton, W.; Moss. I.; Jacobs, F. Chesapeake Bay water
quality monitoring progam mesozooplankton component:
August 1984 - December 1985. Maryland Office of
Environmental Programs, Baltimore, Maryland; 1986.
Gushing, D.H.; Dickson, R.R. The biological response in the
sea to climatic changes. In: Russell, F.S.; Yonge, M.,
eds. Adv. Mar. Biol. London: Academic Press, Inc.: 14:2-
122; 1976.
Davis, H.C.; Chanley, P.E. Effects of some dissolved
substances on bivalve larvae. Proc. Natl. Shellfisheries
Assoc. 96:59-74; 1956.
Davis, P.G. Geographic, seasonal and diel distribution of
phototrophic and heterotrophic components of the
picoplankton and nanoplankton in the North Atlantic.
PhD Thesis Univ. Rhode Island; 1984.
Deason, E.E.; Smayda, TJ. Ctenophore-zooplankton-phyto-
plankton interactions in Narragansett Bay, Rhode Island,
USA, during 1972-1977. J. Plank. Res. 4:203-217; 1982.
Eppley, R.W.; Harrison, W.G. Physiological ecology of
Gonyaulaxpolyhedra, a red water dinoflagellate of
southern California. In: LoCicero, V.R., ed. Proc. First
Int. Conf. on Toxic Dinoflagellate Blooms. Wakefield,
MA: Mass. Sci. Technol. Foundation; 11-22; 1975.
Frank, K.T. Ecological significance of the ctenophore
Pleurobrachia pileus off southwestern Nova Scotia. Can.
J. Fish. Aquat. Sci. 43:211-222; 1986.
Friedland, K.D.; Haas, L.W.; Merriner, J.V. Filtering rates of
the juvenile Atlantic menhaden Brevoortia tyrannus
(Pisces: Clupeidae), with consideration of the effects of
detritus and swimming speed. Mar. Biol. 84:109-117;
1984.
Galtsoff, P.S. The American oyster Crassostrea virginica
Gmelin. Fish. Bull. 64:219-238; 1964.
Gill, C.W.; Harris, R.P. Behavioral responses of the
copepods Calanus helgolandicus and Temora longicornis
to dinoflagellate diets. J. Mar. Biol. Assoc. U.K. 67:785-
801; 1987.
Greve, W.; Reiners, F. The impact of predator-prey waves
from estuaries on the planktonic marine ecosystem. In:
Kennedy, V.S., ed. Estuarine Perspectives. New York:
Academic Press, Inc.: 405-435; 1980.
Huntley, M.E.; Hobson, L. Medusa predation and plankton
dynamics in a temperate fjord. J. Fish. Res. Bd. Can.
35:257-261; 1978.
Huntley, M.; Sykes, P.; Rohan, S.; Martin, V. Chemically-
mediated rejection of dinoflagellate prey by the copepods
Calanus pacificus and Paracalanus parvus: mechanism,
occurrence and significance. Mar. Ecol. Prog. Ser. 28:105-
120; 1986.
Jeffries, H.P. Succession of two Acartia species in estuaries.
Limnol. Oceanogr. 7:354-364; 1962.
Jeffries, H.P.; Terceiro, M. Cycle of changing abundances in
the fishes of the Narragansett Bay area. Mar. Ecol. Prog.
Ser. 25:239-244; 1985.
Jenkinson, I.R. Oceanographic implications of non-newto-
nian properties found in phytoplankton cultures. Nature
323:435-437; 1986.
June, F.C.; Carlson, FT. Food of young Atlantic menhaden,
Brevoortia tyrannus, in relation to metamorphosis. Fish.
Bull. 68:493-512; 1971.
Kremer, P.; Nixon, S.W. Distribution and abundance of the
ctenophore, Mnetniopsis leidyi, in Narragansett Bay. Est.
Coast. Shelf Sci. 4:627-639; 1976.
Lindahl, O.; Hemroth, L. Phyto-zooplankton community in
coastal waters of western Sweden - an ecosystem off
balance? Mar. Ecol. Prog. Ser. 10:119-126; 1983.
Malone, T.C. Nutrient limited phytoplankton production in
the mesohaline reach of Chesapeake Bay. EOS 68:1688;
1987.
Miller, R.J. Distribution and biomass of an estuarine cteno-
phore population, Mnemiopsis leidyi (A. Agassiz).
53
-------
PLENARY SESSION
Chesapeake Sci. 15:1-8; 1974.
Moller, H. Significance of coelenterates in relation to other
plankton organisms. Meeresforsch. 27:1-18; 1979.
Moller, H. Scyphomedusae as predators and food competitors
of larval fish. Meeresforsch. 28:90-100; 1980.
Nelson, T.C.. On the occurrence and food habits of cteno-
phores in New Jersey inland coastal waters. Biol. Bull.
48:92-111; 1925.
Parsons, T.R.; Kessler, T.A. An ecosystem model for the
assessment of plankton production in relation to the
survival of young fish. J. Plank. Res. 9:125-137; 1987.
Potts, G.W.; Edwards, J.M. The impact of a Gyrodinium
aureolum bloom on inshore young fish populations. J.
Mar. Biol. Assoc. U.K. 67:293-297; 1987.
Reeve, M.R. Comparative experimental studies on the
feeding of chaetognaths and ctenophores. J. Plank. Res.
2:381-393; 1980.
Roddie, B.D.; Leakey, R.J.G.; Berry, AJ. Salinity-tempera-
ture tolerance and osmoregulation in Eurytemora affinis
(Poppe) (Copepoda: Calanoida) in relation to its distribu-
tion in the zooplankton of the upper reaches of the Forth
estuary. J. Exp. Mar. Biol. Ecol. 79:191-211; 1984.
Seliger, H.H.; Boggs, J.A.; Rivkin, R.B.; Biggley, W.H.;
Aspden, K.R.H. The transport of oyster larvae in an
estuary. Mar. Biol. 71:57-72; 1982.
Sellner, K.G.; Olson, M.M. Copepod grazing in red tides of
Chesapeake Bay. In: Anderson, D.M.; White, A.W.;
Baden, D.G., eds. Toxic dinoflagellates. New York:
Elsevier, Inc., 245-250; 1985.
Smayda, T.J. Variations and long-term changes in Narra-
gansett Bay, a phytoplankton-based coastal marine
ecosystem: relevance to field monitoring for pollution
assessment. In: White, H.H.;ed. Concepts in marine
pollution measurements. College Park, MD: Maryland
Sea Grant College, Univ. Maryland: 663-679; 1984.
Southward, A.J. The Western English Channel - an incon-
stant ecosystem? Nature 285:361-366; 1980.
Southward, A.J. Fluctuations in the ecosystem of the Western
Channel: a summary of studies in progress. Proceedings
17th European Marine Biology Symposium, Brest,
France, 27 September -1 October. Oceanol. Acta No. SP:
187-189; 1983.
Southward, A.J. Butler, E.I.; Pennycuick, L. Recent cyclic
changes in climate and in abundance of marine life.
Nature 253:714-717; 1975.
Stoecker, D.K.; Govoni, J.J. Food selection by young larval
gulf menhaden (Brevoortiapatronus). Mar. Biol. 80:299-
306; 1984.
Stoecker, D.K.; Sanders, N.K. Differential grazing by Acartia
tonsa on a dinoflagellate and a tintinnid. J. Plank. Res.
7:85-100;1985.
Stoecker, D.K.; Verity, P.G.; Michaels, A.E.; Davis, L.H.
Feeding by larval and post-larval ctenophores on microzo-
oplankton. J. Plank. Res. 9:667-683; 1987.
Sullivan, B.K.; McManus, L.T. Factors controlling seasonal
succession of the copepods Acarlia hu.dson.ica and Acartia
tonsa in Narragansett Bay, Rhode Island: temperature and
resting egg production. Mar. Ecol. Prog. Ser. 28:121-128;
1986.
Sullivan, B.K., Reeve, M.R. Comparison of estimates of the
predatory impact of ctenophores by two independent
techniques. Mr. Biol. 68:61-65; 1982.
Sutcliffe, W.H. Jr.; Drinkwater, K.; Huir, B.S. Correlations of
fish catch and environmental factors in the Gulf of Maine.
J. Fish. Res. Bd. Can. 34:19-30; 1977.
Tanasichuk, R.W.; Ware, D.M. Influence of interannual
variations in winter sea temperature of fecundity and egg
size in Pacific herring (Clupea harenguspallasi). Can. J.
Fish. Aquat. Sci. 44:1485-1495; 1987.
Taylor, A.H.; Stephens, J.A. Latitudinal displacements of the
the Gulf Stream (1966 to 1977) and their relation to
changes in temperature and zooplankton abundance in the
NE Atlantic. Oceanol. Acta 3:145-149; 1980.
Taylor, C.C.; Bigelow, H.B.; Graham, H.W. Climatic trends
and the distribution of marine animals in New England.
Fish. Bull. 57:293-345; 1957.
Turner, J.T.; Anderson, D.M. Zooplankton grazing during
dinoflagellate blooms in a Cape Cod embayment, with
observations of predation upon tintinnids by copecods.
Publicazioni della Stazione Zoologica di Napoli I: Marine
Ecology 4:359-374; 1983.
Tuttle, J.H.; Malone, T.C.; Jonas, R.B.; Ducklow, H.W.;
Cargo, D.G. Nutrient-dissolved oxygen dynamics in
Chesapeake Bay: the roles of phytoplankton and micro-
heterotrophs under summer conditions, 1985. Final
Report to EPA, Chesapeake Bay Program: UMCEES Ref.
No. 86-125a-CBL; 1986.
Tyler, M.A.; Coats, D.W.; Anderson, D.M. Encystment in a
dynamic environment: deposition of dinoflagellate cysts
by a frontal convergence. Mar. Ecol. Prog. Ser. 7:163-
178; 1982.
Tyler, M.A.; Seliger, H.H. Annual subsurface transport of a
red tide dinoflagellate to its bloom area: water circulation
patterns and organism distributions in the Chesapeake
Bay. Limnol. Oceanogr. 23:227-246; 1978.
Tyler, M.A.; Seliger, H.H. Selection for a red tide organism:
physiological responses to the physical environmental.
Limnol. Oceanogr. 26:310-324; 1981.
Ukeles, R.; Sweeney, B.M. Influence of dinoflagellate
trichocysts and other factors on the feeding of Crassostrea
virginica larvae on Monochrysis lutheri. Limnol.
Oceanogr. 14:403-410; 1969.
Uchida, T. Excretion of a diatom-inhibitory substance by
Prorocentrum micans Ehrenberg. Jap. J. Ecol. 27:1-4;
1977.
Verity, P.G. Abundance, community composition, size
distribution, and production rates of tintinnids in Narra-
gansett Bay, Rhode Island. Est. Coast. Shelf. Sci. 24:671-
690; 1987.
Vucetic, T. Fluctuation in the distribution of the scypho-
medusa Pelagia noctiluca (Forskal) in the Adriatic. Pro-
ceedings 17th European Marine Biology Symposium,
Brest, France, 27 Sept. -1 Oct., Oceanol. Acta. No.
SP:207-211;1983.
Watras, C.J.; Garcon, V.C.; Olson, R.J.; Chisholm, S.W.;
Anderson, D.M. The effect of zooplankton grazing on
estuarine blooms of the toxic dinoflagellate Gonyaulax
tamarensis. J. Plank. Res. 7:891-908; 1985.
54
-------
Trophic Structure
Trophic Structure of the Chesapeake Mesohaline Ecosystem
Robert Ulanowicz
Chesapeake Biological Laboratory
University of Maryland
Box 38
Solomons, Maryland 20688
My subject today will be some details of the carbon
flow network of the Chesapeake mesohaline ecosystem.
Some of you may have seen some of these results, but
they bear repeating, partly because they may seem less
ridiculous on second hearing, and partly because they
fit well into the context set by the two talks preceding
mine.
The topic is represented, oddly enough, by a tiny
item appearing in the Baltimore Sun on Monday,
without attribution of any kind, saying: "The most
productive fisheries are found in colder areas of the
world's oceans, and researchers now believe the reason
is that here fish do not have to compete with bacteria
for food." Many readers of the Sun no doubt scratched
their heads in puzzlement; most of us, however, with
backgrounds in ecology, understand that the question is
one of indirect trophic effects.
The hypothesis that these two elements of the
ecosystem do compete in some way would have been
approached 10-15 years ago by the creation of a large
mathematical model of the Chesapeake Bay system.
This approach has fallen out of favor, because these
models' predictive ability is problematical and many
biologists feel that mathematics do a poor job of
representing living, growing, evolving systems. So
what approach will work? We must again become
ecologists in the true sense of the word, studying the
interactions between organisms and their environment.
One of the most palpable ways of doing this is to
measure the transfers of material and energy through
the ecosystem.
This kind of study is just what we proposed to the
Maryland Department of Natural Resources several
years ago. Dr. Dan Baird collected data from the
literature and developed an estimated food web, and I
did some of the analysis I'll present now. The area
covered was the mesohaline area, about 6-18 ppt. We
represented the Bay in 36 compartments (Figure 1).
Subdividing it, you will find a planktonic realm,
consisting of primary producers, mesozooplankton, and
gelatinous predators, as well as the microbial loop or
web. In the benthic area are the diatoms, filter feeders,
and deposit feeders, and the blue crab. The third large
domain is the nekton, consisting primarily of the filter
feeders, benthic carnivores, and pelagic piscivores.
This kind of network can be shown for each of the
seasons, and it is interesting to see that the topology of
the system does not change greatly from season to
season. Some species migrate out and back in, and the
tempo undergoes wide swings, but the only qualitative
major change that affects the whole food web is the
summertime cropping of the mesozooplankton by the
huge stocks of ctenophores.
One of the things that can be done with a network
like this, with over 160 exchanges among 36 compart-
ments, is to analyze the exchanges in a matrix. It is
possible through mathematical manipulation to estimate
the amount transferred between two species and to
examine all the indirect effects between them. The ma-
trix of total dependency coefficients that can be calcu-
lated yields information about indirect diets, that is, the
percentage of a species' total consumption that at some
previous time was part of another species (Table 1). In
the case of bluefish (species 30), for instance, 7.2% of
its intake spent time at some point in compartment 5,
bacteria. There are some interesting effects here. For
example, 65.8% of striped bass intake was at one time
mesozooplankton; the percentage for bluefish was
considerably less at 28.7%. Polychaetes show another
striking difference: 48.0% of the bluefish diet was at
one time polychaetes, whereas the figure for striped
bass is only 1.8%. Bacteria eventually make a 7.6%
contribution to striped bass and a 7.2% contribution to
bluefish.
Figure 1 (overleaf). Estimated food web of the Chesapeake
Bay during summer, represented in 36 compartments in the
planktonic, benthic, and nektonic realms. Flows are depicted
as mg C m"2 per 90 days.
55
-------
PLENARY SESSION
-------
Trophic Structure
CN*
,22
I
c
£
^H
H
\.f\
*LJ
ro
s
„
CO
CO
CO
O
ON
CN
rS
c~
CN
fNen— 'OoovifS'— '
»-ienTrovoiooooo
en 0(S-H
"-ienQOvO>/">oO>/")
i—i
oo c*J c*J fO ON vo i/~i oo
TTenOONvo«noo-H
rn ON ^H
en en 0 ^ ON vo en cs
cs)cy~ir^CNiO|/"ioot^
c^ o\
TTC^^O — r^^X VO i— ' VOOTf'
tN r- TT
t--HOOOOOOOOOOOOOOOOOOOOOOOt~-voo
F*H ON f*«»
oooooooooooooooooooooooooooo
0-vOCNOOOt-^^OONOt-voONOOOOOOOOOOvoencN
^ ^"^^^JoS '"iri1^
O—1 OOOONeNCNtNV^OOO'-i'^'OOOOOOOONOOvOTj-in
*^ en i/^i «—< vo (*~ vo en
vo en i— < V)
ON^OOOOOOOOOOOOOOOOOOOOOOOenTto
8 ^K
»*H
fM^^CSfMOONVOTfONenOOvoooOOOooOOOOOOeNvooo
^J r«4 «« H r^H ^H ^D ^D ^D
Tr^ONenTrrrovOON^^OOenOOOO-HOOOOOOOOenO
*"H r^*' 1^1 if\ ^O C5 ^P ^^ ^5
r^ <-H CN r^ o
co ^H oo co co w*i i—* ^? 0s) ON T ^5 C5 ^D ^3 ^5 C* ^5 ^H ^5 C5 ^5 ^? C^ ^3 ON V"i ol
^"^ (^ fS) I^J ^O ^^> \^J QQ ^^
^O "f ^H MD ON
i-H en ^ vo vo 1—1 en vo ON f^
r— rN •— < vo ON
en-HOOr~-enr~vDO(SenTj-OOenOOOOi-iOOOOOOt~-envo
I-H ON oo oo */"> vo 10 vo r^ o
VO 1— i— VO ON
^f ^^ ON CO ^f f"J f^ V^ ON ^H **f ^^ ^^ f^ ^^ ^^ ^^ ^^ ^^ ^^ ^^ ^^ ^^ J^ ^^ QQ f<^ ^^
^H ^H V^ l/^J >^ ^ \^ ^^ f"~)
oo CN r- 0
CNCNfNcNeNCMCNfNeNfN enenenenenen
57
-------
PLENARY SESSION
1337830
18873
1512
1281.7
616.6
4.5
888723 1
I+D
84.6%
1131637
-N
II
35.7%
403684
III
9.8%
39733
IV
11.0%
4370.3 t
V
8.6%
370.7
VI
3.4%
12.5 ,
VII
0.8%
0.104
1 —
VIII
253469 >^ 606578
218479
449107
133972
17688.2
2284747
16392.67
6.7
Figure 2. Lindemann-type diagram of the trophic chain. Flows are depicted as mg C m'2 per 90 days.
The same matrix algebra can be used to assess the
trophic distance of each species from the sources of
primary production, and the "average trophic position"
at which each species feeds can be calculated (Table 2).
An interesting feature of this ranking is that the sea
nettle (position 26) has an average trophic level of 3.44,
higher than many of the commercially important
species in the Bay. The highest trophic position is held
by bluefish, but the trophic level is only 4.59, lower
than the 5.0 that was deemed significant by Hutchinson
[1948] and Pimm [1982].
If we know how much of the diet is coming over
pathways of various lengths, we can construct a
Lindemann-type diagram of the trophic chain (Figure
2). Each of the boxes now represents parts of popula-
tions. One thing to notice is that not much gets past
trophic level V, although some pathways reach length
VIII. The ratio of detritivory to herbivory in the Bay is
10.5:1, which is really more like a terrestrial system
than the open ocean. If we merge the detritus with the
plants, we see the classical trophic pyramid that
uniformly decreases going up the Lindemann spine.
One exception is at trophic level IV, where the ciliates
are very efficient at passing on what they have taken in.
We can take the whole system and subtract from it
the cycling pathways. Several conclusions can be
drawn from the resulting picture. First, it is a bipartite
structure, with two non-overlapping domains of carbon
recycling in the system, one planktonic and one
benthic-nektonic. Among the planktonic, most of the
members of the microbial food web are missing. At
least for carbon the microbiota do not form a loop, but
Table 2. Trophic rankings and average annual trophic levels of the major components of the Chesapeake Bay
mesohaline ecosystem.
Rank
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
Component
Phytoplankton
Dissolved organic carbon
Suspended POC
Sediment POC
Benthic diatoms
Suspended POC bacteria
Sediment POC bacteria
Free bacteria
Oysters
Mya arenaria
Misc. suspension feeders
Zooplankton
Meiofauna
Ciliates
Menhaden
Bay anchovy
Heterotrophic microflagellates
Misc. polychaetes
Trophic
level
1.00
1.00
1.00
1.00
1.00
2.00
2.00
2.00
2.08
2.09
2.09
2.16
2.67
2.75
2.77
2.84
3.00
3.00
Rank
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
Component
Nereis
Macoma spp.
Crustacean deposit feeders
Ctenophores
Fish larvae
Alewife & blue herring
Shad
Sea nettle
Blue crab
Weakfish
Striped bass
Hogchoker
White perch
Flounder
Spot
Croaker
Catfish
Bluefish
Trophic
level
3.00
3.00
3.00
3.03
3.16
3.16
3.16
3.44
3.51
3.84
3.87
3.91
3.98
3.99
4.00
4.00
4.00
4.59
58
-------
Trophic Structure
rather serve more as a sink. The filter feeders are not
engaged in this recycle; they serve as a shunt taking
carbon and energy from the planktonic water column
into the benthos and the nekton. Critical arc analysis of
the planktonic systems shows that in the summer the
sea nettles are probably controlling this system.
The overall picture shows high summer productiv-
ity, which along with predator control on the zooplank-
ton fuels "excess" productivity. There are two apparent
routes for this excess productivity. One is dissipation in
the microbial loop, which now looks like a shunt out of
the system; the other is a heavy subsidy to the deposit
feeders on the bottom.
In our analysis we were surprised by how much
more influential the deposit feeders were than the filter
feeders. The question we must now pose is, was this
always so? Or was there a change sometime in the last
few decades? If so, what was responsible? The
ctenophores' grazing control? The quality of the phyto-
plankton?
REFERENCES
Pimm, S. L. Food webs. Chapman and Hall, London, 1982.
219pp.
Hutchinson, G. E. The circular causal systems of ecology.
Ann. N.Y. Acad. Sci. 50:221-246,1948.
59
-------
PLENARY SESSION
The Importance of the Microbial Loop in the Chesapeake Bay and its Tributaries
Kevin G. Sellner
The Academy of Natural Sciences
Benedict Estuarine Research Laboratory
Benedict, Maryland 20612
In recent years many conceptual models of the
microbial food web have been developed in terms of
carbon transformations, passage to higher trophic
levels, and remineralization [Pomeroy 1974; Parsons
1979; Williams 1981]. One such model is shown here
(Figure 1) [Azam et al. 1983], and in it you should
focus on the second suite of arrows representing the
cyanobacteria. The idea is that the microbial food web
could be a source of food and energy for higher trophic
levels, or it could be a sink.
It should be remembered that although the microbial
food web is drawing a lot of attention currently, it has
only been in the last 10 to 15 years, with the develop-
ment of higher-resolution techniques, that we have been
able to establish the importance of the very small
heterotrophs and autotrophs within all aquatic systems.
The microbial food web is very well established in
the Chesapeake Bay, and one question of great interest
is whether it can be related to water quality. The
forcing functions have been reviewed for almost every
level in the trophic system by Verity [1987]. Without
repeating all of what he said, we should briefly summa-
rize a few points. First, when Smayda [1983] reviewed
phytoplankton in estuaries, he documented that high
nutrient loading rates in a stratified estuarine system
resulted in the predominance of picoplanktonic and
flagellated forms (autotrophs) in contrast to the normal
dominance of large flagellates and diatoms. Second,
Oviatt et al. [1986] working in mesocosms in Rhode
Island showed that high nutrient loading rates induced
the most dramatic responses in the bacterial commu-
nity. The highest nutrient loading rates resulted in a
very large increase in bacterial numbers and activity,
and water column respiration rather than benthic
respiration dominated oxygen demand. Sediment input
from agricultural and development practices in the
Chesapeake Bay could result in similar plankton
responses [Fogg 1986; Glover et al. 1986; Parsons et
al.; 1986]. For example, several investigators have
suggested that picoplanktonic forms such as the
cyanobacteria may be better adapted to low-light
conditions than other planktonic forms. Thus high
turbidities of the upper Chesapeake and the light
limitation documented there by Flemer [1970] and
Harding et al. [1986] may be selecting for picoplank-
tonic forms. Similarly, as Sanders and others men-
tioned yesterday in summarizing the input of toxics,
much work suggests that the input of metals, hydrocar-
bons, and other materials at sublethal concentrations
results in a shift of phytoplankton species composition
and sizes. The shift in size is important; sublethal
concentrations of metals and toxics may cause the
predominance of small cells [Greve and Parsons 1977].
Thus three factors that we know encourage the develop-
ment of a microbial food web exist in the Chesapeake
Bay: high nutrient loadings, high turbidity, and
numerous toxic point sources.
What evidence supports the actual existence of a
microbial food web in the Bay? Surveying the litera-
ture for such evidence confirms the recent advent of
interest in the topic, as all of the work has been done in
the last few years (Table 1). Ducklow and Tuttle have
estimated bacterial densities in the Bay as 109-1010
MICRO - ORGANISMS
11 N 0 T M lot mm
Figure 1. Semi-quantitative model of planktonic food chains.
Solid arrows represent flow of energy and materials; open
arrows, flow of materials alone [Azam et al. 1983],
60
-------
Trophic Structure
Table 1. Evidence for a well developed microbial loop in Chesapeake Bay.
Trophic
level
Density
(no71)
Area
Reference
Bacteria
Heterotrophic
nano flagellates
Picoplankton
Microzooplankton
(whole-water)
Microzooplankton
(>44nm)
10wo
1.5-6.5x10*
0.8-6.1x10*
-------
PLENARY SESSION
Table 2. Metabolism of the microbial loop.
Trophic level
Metabolism
Area
Reference
Bacteria
Microheterotrophs
Microzooplankton
grazing
30->50%
(bact prod/prim prod)
OX cons.=0.0045(Bact)
+0.007; r=0.95
P/R<1
Euphotic zone BOD via
oxidation of participates
micro/macro > 1
Mesohaline region
Mesohaline region
Mesohaline region
Mesohaline region
Mesohaline region
Malone et al. 1986
Turtle etal. 1987a,b
Kemp & Boynton 1980
Sellner 1987
Turtle et al. 1987b
Brownlee et al. 1986
Sellner et al. in prep.
[Sellner and Brownlee 1985-88]. Whole-water counts
show total densities of 15,000 to 24,000/liter.
[Brownlee and Jacobs, 1987]. The larger microzo-
oplankton (rotifers, tintinnids, etc.) are found at
densities from <100 to 1500/liter (Figure 3).
Other indications that we may have a well-devel-
oped microbial food web come from metabolism
studies (Table 2). Ducklow, working with Malone and •
Tuttle on tilting [Malone et al., 1986] estimated that
between 30% and >50% of the primary productivity
passes through the bacteria. Ducklow et al. have
summarized these data for the last four years and will
present more long-term data in a session this afternoon.
Tuttle and co-workers [1987b] have also estimated
summer oxygen consumption in the mesohaline Bay,
and found that consumption in the water column is
directly related to bacterial densities. This relationship
suggests that measuring bacterial abundance alone
could serve as an indicator of oxygen consumption
rates; however, this relationship needs to be confirmed
in other salinity regimes and in all seasons before this
measure can be applied everywhere.
An active microbial web in the Bay is also suggested
by several other data sets. For example, water column
respiration off Calvert Cliffs in mesohaline Chesapeake
Bay exceeded production in 18 of 23 summer months,
1975-1980 [Sellner 1987]. In addition, euphotic zone
oxygen demand in mesohaline summer conditions is
primarily due to the oxidation of paniculate matter
[Tuttle et al., 1987b] reflecting total microbial commu-
nity metabolism, rather than solely the bacteria.
A comparison of microzooplankton grazing with
macrozooplankton grazing indicates that the former is
generally much greater than the latter [Brownlee et al.,
1986]. Furthermore, recently collected data on the very
small ciliates suggest that the grazing pressure during
May, June, and July is primarily due to small ciliates [J.
Dolan, unpublished data]. In other words, herbivory
takes place primarily through these small cells.
These data indicate the presence of a strongly
established microbial food web in Chesapeake Bay.
One of the questions we should address is: if we can
reduce loading, will we see reduced eutrophication,
improved water quality, a decrease in anoxia and an
increase in fish yields? A theoretical diagram from
Parsons [1979] (Figure 4) suggests that we can con-
ceive of current conditions in the Chesapeake Bay as a
microautotroph community (the picoplankton and
flagellates) being grazed on primarily by a protozoan
food web, eventually reaching the coelenterates
(discussed by Verity [1987]) as gelatinous predators. If
we could reduce the loading rates, could we expect to
see selection of large diatoms, large flagellates, larger
copepod grazers, and ultimately pelagic fish? Can we
eventually, through regulation of inputs of stresses —
nutrients, suspended sediments, and toxics — shift the
community in this way?
RECOMMENDATIONS
Based on presentations by Verity, me, and later,
Ducklow et al., future Bay conditions might be pre-
dicted with additional support in several areas. We
should continue to estimate the importance of the
microbial food web. We need more information on the
distribution and the activities of these very small
plankton.
We need to do more experiments along the line that
Sanders et al. [1987] have pursued, i.e., looking at the
effects of nutrient loading and light limitation on
62
-------
Trophic Structure
Trophic l«vtl
Prtdatort
Secondary
producer*
Primary
productrt
Whaltt
Palaglc Fish
Co«l«nter«l«i
,.->rCop«podt
Cruttacaan* ',"
0*lr*cod»
Protozoa
Cyanophycaaa
QU/- Dinoflagtllatts
Rhodophycaa*
Chlorophycaa*
500
400
300 200
Year* (x 10')
100
Figure 4. Theoretical diagram of evolutionary relationships between primary producers, secondary producers, and predators.
inducing changes in the microbial webs. Specifically,
we should find out whether nutrient increases and light
limitation cause an increase in the smaller components
in a stratified system. Once the stress is removed, does
the system return to diatoms and larger flagellates and
copepods? Or does it shift, as Verity suggested earlier,
to selection of another metazoan food web, slightly
different but still passing carbon to a higher but
different trophic level?
Much more work needs to be done toward establish-
ing the relationship between bacterial densities and
activities and the oxygen demand in the water column.
We have very little historical data on the importance of
the microbial food web. It may have been well-
established all along, only becoming apparent with the
development of high-resolution techniques. A possible
approach to this problem would be to examine degrada-
tion products of the pigments associated with the
picoplanktonic fraction under anoxic conditions. Then
using cores and substrata analysis in the same manner
as Brush (University of Maryland), coupled with HPLC
analysis, we might be able to determine whether the
dominant members of the picoplankton have been
present over the last several centuries.
RELEVANT LITERATURE
Azam. F..T. Fenchel, J.G. Field, J.S. Gray, L.A. Meycr-Rcil
and F. Thingstad. 1983. The ecological role of water-
column microbes in the sea. Mar. Ecol. Prog. Ser. 10: 257-
263.
Brownlee, D.C. and F. Jacobs. 1987. Mesozooplankton and
microzooplankton in the Chesapeake Bay. In: S.K.
Majumdar, L.W. Hall, Jr. and H.M. Austin (eds.).
Contaminant problems and management of living
Chesapeake Bay resources. PA Acad. Sci., Easton, PA: p.
217-269.
Brownlee, D.C.; F. Jacobs; S.G. Brownlee, and K. G. Sellner.
1986. Relationship between nutrients and plankton in
Chesapeake Bay. III. Potential roles of micro- and
macrozooplankton. Abstract and presentation, ASLO,
URI, 23-26 June 1986.
Dolan J.R. and D.W. Coats. 1987. Heterotrophic microflage 1-
63
-------
PLENARY SESSION
lates in the Chesapeake Bay seasonal abundances in
different parts of the water column. EOS 68: 1728.
Ducklow, H.W. 1982. Chesapeake Bay nutrient and plankton
dynamics. I. Bacterial biomass and production during
spring tidal destratification in the York River, Virginia,
estuary. Limnol. Oceanogr. 27: 651-659.
Flemer, D.A. 1970. Primary production in Chesapeake Bay.
Ches. Sci. vol II. pp. 117-129.
Fogg, G.E. 1986. Light and ultraphytoplankton. Nature 319:
96-99.
Glover, H.E., M.D. Keller and R.R.L. Guillard. 1986. Light
quality and oceanic ultraphytoplankters. Nature 319: 142-
143.
Greve, W. and T.R. Parsons. 1977. Photosynthesis and fish
production. Hypothetical effects of climate change and
pollution. Helgol. wiss. Meeresunters. 30: 666-672.
Harding, L.W., B.W. Meeson and T.R. Fisher Jr. Phytoplank-
ton production in two east coast estuaries: photosynthesis-
light functions and patterns of carbon assimilation in
Chesapeake and Delaware Bays. Est. Coast Shelf. Sci.
23:773-806; 1986.
Kemp, W.M. and W. R. Boynton. 1980. Influence of biologi-
cal and physical processes on dissolved oxygen dynamics in
an estuarine system: implications for measurement of
community metabolism. Estuarine Coastal Marine Science,
vol. II, p. 401-431.
Malone, T.C., W.M. Kemp, H.W. Ducklow, W.R. Boynton,
J.H. Tuttle and R.B. Jonas. 1986. Lateral variation in the
production and fate of phytoplankton in a partially stratified
estuary. Mar. Ecol. Prog. Ser. 32: 149-160.
Marshall, H.G. and R.V. Lacouture. 1986. Seasonal patterns of
growth and composition of phytoplankton in the lower
Chesapeake Bay and vicinity. Est. Coastal Shelf Sci. 23:
115-130.
Oviatt, C.A., A.A. Keller, P.A. Sampou and L.L. Beatty, 1986.
Patterns of productivity during eutrophication: A meso-
cosm experiment. Mar. Ecol. Prog. Ser. 28: 69-80.
Parsons, T.R. 1979. Some ecological, experimental and
evolutionary aspects of the upwelling system. S. Afr. J. Sci.
75: 536-540.
Parsons, T.R., P. Thompson, W. Yong, C.M. Lalli, H. Shumin
and X. Huaishu. 1986. The effect of mine tailings on the
production of plankton. Acta Oceanolog. Sinica 5: 417-423.
Pomeroy, L.R. 1974. The ocean's food web, a changing
paradigm. Bioscience 24: 499-504.
Ray, R.T. 1986. The role of picoplankton in phytoplankton
dynamics of a temperate coastal plain estuary. M.S.
Thesis, VIMS, Gloucester Pt., VA. 85 pp.
Sanders, J. G., S. J. Cobik, C. F. D'Elia, and W. R. Boynton.
1987. Nutrient enrichment studies in a coastal plain
estuary: changes in phytoplankton species composition.
Can. J. Fish. Aquat. Sci. 44:83-90.
Sellner, K.G. 1987. Phytoplankton in the Chesapeake Bay:
Role in carbon, oxygen and nutrient dynamics. In: S.K.
Majumdar, L.W. Hall, Jr. and H.M. Austin (eds.).
Contaminant problems and management of living
Chesapeake Bay resources. PA Acad. Sci., Easton, PA:
p. 134-157.
Sellner, K.G. and D.C. Brownlee. 1985-1988. Maryland
Department of the Environment. Chesapeake Bay water
quality monitoring program. Phytoplankton and micro-
zooplankton component, data summary. The Acad. Nail.
Sci., Benedict Est. Res. Lab., Benedict, MD.
Smayda,TJ. 1983. The phytoplankton of estuaries. In: B.H.
Ketchum (ed.). Estuaries and enclosed seas. Elsevier,
Amsterdam: p. 65-102.
Tuttle, J.H., R.B. Jonas and T.C. Malone. 1987a. Origin,
development and significance of Chesapeake Bay anoxia.
In: S.K. Majumdar, L.W. Hall, Jr. and H.M. Austin (eds.).
Contaminant problems and management of living
Chesapeake Bay resources. PA Acad. Sci., Easton, PA:
p. 442-472.
Tuttle, J.H., T.C. Malone, R.B. Jonas, H.W. Ducklow and
D.C. Cargo. 1987b. Nutrient-dissolved oxygen dynamics
in Chesapeake Bay: The roles of phytoplantcton and
microheterotrophs under summer conditions, 1985. U.S.
EPA, CBP/TRS 3/87, 158 pp.
Verity, P.G. 1987. Factors driving changes in the pelagic
trophic structure of estuaries, with implications for the
Chesapeake Bay. In: Perspectives on the Chesapeake
Bay: Recent advances in estuarine sciences. U.S. EPA,
CBP/TRS 16/87, CRC publication no. 127: p. 35-56.
Williams, PJ. leB. 1981. Incorporation of microheterotrophic
processes into the classical paradigm of the planktonic
food web. Kieler Meeresforsch. Sondh. 5: 1-28.
64
-------
CONCLUDING SUMMARIES AND PANEL DISCUSSION
"Where do we go from here?"
— Summaries of concurrent sessions, identifying
major research questions and associated costs
Participants
Living Resources
Ed Houde
Chesapeake Biological Laboratory
University of Maryland
Dissolved Oxygen
Eugene Cronin
Coastal Consultant
Submerged Aquatic Vegetation
Robert J. Orth
Virginia Institute of Marine Science
College of William and Mary
Physical Processes
Eric Itsweire
Chesapeake Bay Institute
The Johns Hopkins University
Owen Phillips
The Johns Hopkins University
Toxics
Nicholas Fendinger
Agricultural Research Service
U. S. Department of Agriculture
Nutrients
Carl Cerco
U. S. Army Corps of Engineers
"How do we get there? Who pays the fare?"
— Panel discussion with members from federal and
state funding agencies, scientific institutions, and
private organizations
Panel Members:
Monica Healy
Office of Governor
State of Maryland
Chris D'Elia
National Science Foundation
John W.Daniel III
Secretary of Natural Resources
Commonwealth of Virginia
Frank 0. Perkins
Virginia Institute of Marine Science
College of William and Mary
Ian Morris
Center for Environmental and Estuarine Sciences
University of Maryland
David Challinor
The Smithsonian Institution
Sheldon Samuels
Health, Safety, and Environment
AFL-CIO
65
-------
PLENARY SESSION
"Where do we go from here?"
Joseph A. Mihursky, moderator
LIVING RESOURCES: Ed Houde
The perspective of this conference on living re-
sources has been a broad one, including all of the living
organisms in the Bay, as well as factors that influence
their well-being, abundance, and changes in abundance.
The Chesapeake Bay is a life support system, and we
have looked at living resources from an ecosystem
perspective. We may not have given enough attention
to non-harvestable living resources in the past. We are
making a good start at understanding the Bay ecosys-
tem, and are beginning strategic planning to continue
expanding our knowledge. This conference has been a
little disappointing in its lack of coverage of the top
levels, that is, the fishes; but why the people doing this
research did not join us here is hard to say.
In speaking about benthic-pelagic coupling, Walter
Boynton mentioned how thin a layer of water overlies
the sediments of the Chesapeake and suggested that this
interface of sediments with the water column is an
important area of biological activity. A lot of the
dynamics of the Bay is associated with this interface,
and progress has been made in understanding its
relation to nutrient dynamics and associated production
processes. We have tracked the seasonally of many of
the factors associated with this interface that influence
nutrient flux and primary production. This research is
not specifically organism-oriented; there is more
interest in the total system and its ability to produce.
Clearly the anoxia problem, with its association with
sediment oxygen demand, has an important impact on
the living resources.
The presentations on pelagic trophic structure, in
contrast to the data-oriented presentations on benthic-
pelagic structure, were related to ideas and hypotheses.
Our understanding of the Bay in this area is still
limited. Roman, White, and Brownlee have reported
some results on grazing potential and abundance of
organisms in the water column, but these results are
only a start The question of top-down vs bottom-up
control is important, since we know that nutrients have
increased at the bottom, and many of the top predators
have been removed as well. It was suggested that the
removal of a very large biomass of oysters, relatively in
the middle of the trophic structure, has had a significant
effect on that structure and consequent flow pathways
of energy and material.
The basic production mechanisms in the Bay may
have changed. The bacterial food web may have
become more important than the traditional metazooan
food webs, and may now dominate; in any case it
certainly appears to have gained in importance. A
question unanswered thus far is how important it was in
the past; it may have been important but unstudied.
It is possible that the food pathways may have
shifted toward jellyfish, away from the metazooan web,
and that this may have contributed to the decline in
harvestable b'ving resources.
The recruitment process has not been much dis-
cussed here, although these episodic events do influ-
ence populations. The blue crab work at VIMS, which
is highly resolved in time and space, is a good example
of the kind of work we need to do to understand the
recruitment process and how it affects the abundance of
harvestable resources. We need to take an ecosystem
approach as well as investigating these specific ques-
tions of recruitment. The recruitment process is highly
variable, and is influenced by two major factors: adult
stocks and environmental conditions in a specific year.
It seems now that the environment may be a more
important factor than adult stocks. We also need
traditional kinds of fisheries research, as well as single-
species and multi-species modeling efforts. It is worth
repeating here a recommendation that was made at a
Sea Grant conference on fisheries in February: that
fisheries research should concentrate on the first 100
days of life. It is here that the physics of the system,
the available food, the productivity of the Bay, prey
relationships, the potential for disease and contami-
nants, and the role of habitat all are most important.
As far a research needs, there is a critical need for
strategic planning. This process is beginning, through
STAC, the Monitoring Committee, and the Living
Resources Committee. We need time-series analyses,
with monitoring; the appropriate time and space scales
for monitoring are still open to question. We need to
integrate Chesapeake Bay system research better with
fishery research.
Restoration strategies are critical, not only for SAVs,
but for other resources such as striped bass and other
anadromous fishes. We need to learn how to bring
these resources back to their former levels.
66
-------
"Where do \ve go from here?'
It is hard to estimate the costs of the research that is
needed, except to say that they will be large. Restoring
or stabilizing the resources of the Chesapeake Bay is
worth doing, but it won't come cheap. One way to
DISSOLVED OXYGEN: Eugene Cronin
This conference has shown much evidence of the
progress being made in this field. Interest is broad, and
some efforts have been interdisciplinary. Sea Grant
programs have selected dissolved oxygen as an area of
exceptional importance and have supported it over a
number of years. The ten papers presented here do not
adequately represent what we know or what is being
worked on, so it is difficult to summarize on the basis
of these papers alone. Some of the needs in anoxia-
related research were suggested by papers given in
other concurrent sessions: on the causes and degree of
vertical stratification in the Bay; the effect of anoxia on
the availability of toxics in sediments; the chemical
state of inorganic toxicants; the absence or presence of
oxygen related to sediment flux. Where and with what
intensity is deoxygenation happening? What processes
are causing it? What are the ultimate chemical and
biological effects? What management is possible?
What can and what cannot be controlled?
As sources of oxygen consumption, the microbiota
appear to be growing in importance, accounting for
perhaps as much as 30-40% of the oxygen demand. In
responses to nutrients some obvious possible linkages
are seen, but much is yet unclear. Modeling is still
useful, as it breaks the world down into understandable
pieces. The possibility was raised here that the 99%
removal of oysters has contributed significantly to
anoxia, as the filtration capacity of this species has been
carry out research cost-effectively, an intermediate
option between a small tank in the laboratory and the
whole Bay, is the mesocosm approach. This approach
should be explored and used more fully.
lost without this control on phytoplankton crops, they
have increased in response to added nutrients, and the
decay of the large phytoplankton biomass may be
contributing to the oxygen demand processes.
The biological effects of anoxia have been studied in
the goby; this species is not very mobile and can give
some idea of the biological effects in a spatially limited
area.
The significance of anoxia has been debated. One
presenter contends that anoxia has not increased over
time, but that the changes observed have been largely a
result of spring flows — the most important element,
especially in the upper part of the system. One problem
may be a mismatch between the predators and the
location of the food supply.
Modeling continues to be a vital tool, serving several
functions. The importance of verification must not be
underestimated.
This conference has been important for its interdisci-
plinary discussions. We must become predictive, we
must develop hypotheses, we must go to the field and
test them, then discuss our results and fight it out.
People must be aware that we are in a hydrodynami-
cally fluctuating system.
In discerning what research needs to be done, the
most effective approach may be to ask experts what
research should be funded outside of their field.
SUBMERGED AQUATIC VEGETATION: Robert Orth
Submerged aquatic vegetation gives us a Bay-wide
perspective, a unique blend of the scientist, citizen, and
manager in action. In 1978 little was known about
SAVs. Since then there has been a tremendous change
in the attitudes of scientists and managers, and citizens
have joined the effort by monitoring. The importance of
this citizen involvement should not be underestimated.
For instance, the citizens' perceptions of hydrilla have
changed, and with the change has come a different
approach to control. The chemical approach that led to
application of 2,4D to milfoil 30 years ago has given
way to mechanical control.
Distribution and abundance: a decade of change.
The first task ten years ago was to find out the current
distribution. Mapping surveys of SAVs have been con-
ducted annually since 1978. No major shifts have been
found; most of the SAV is still in the lower Bay. We
still do not know what the distribution was in the
1960s. It is possible that areas now denuded were
choked with vegeation then. For the last 10 years snap-
shot pictures have been used to study local changes in
abundance. Now it is possible to overlay the distribu-
tion data with water quality data. A resurgence is evi-
dent in the York, the Rappahannock, and the mid-Bay
area, probably due to climatic changes. The declines
came from the upper Bay to the lower, and from
upriver to downriver. The increases now being
observed are primarily in the vicinity of existing beds.
Water quality criteria. Major monitoring efforts in
the York and Choptank Rivers will contribute to the
67
-------
PLENARY SESSION
development of water quality criteria for SAV. As the
SAV are immobile, relatively small shifts in water
quality can make big differences in SAV growth. This
sensitivity has implications for water quality manage-
ment efforts.
Natural resource value. To clarify the natural
resource value of SAVs, physical oceanographers and
blue crab specialists need to pool their efforts. SAVs
have a role in the early life stages of the blue crab, but
the relationship has not been fully described.
Restoration. Restoration of SAVs is important
because many appropriate areas are too far from
existing areas for the natural spread to revegetate them.
Most of the test programs are using whole-plant
material; these programs need information on plant
spacing, use of fertilizer, and ideal plant size. The
latest effort is to use seed processes, as these are less
labor-intensive. Transplanting is an important tool for
understanding relationships between plants and the
environment. Generally transplanting is not feasible in
the upriver historical sites far removed from existing
beds, because the cost of moving the plant material is
too great. The difficulty in guaranteeing the success of
SAV makes mitigation a risky proposition. Emphasis
should be on protection, not mitigation. Studies should
continue with propagules such as seeds and with early
successional species and mixed plantings.
Monitoring. The monitoring program should con-
tinue, with annual photography so we continue to
acquire data even if it is not all mapped. This informa-
tion has been valuable to managers in assessing the
impact of development. The lack of historical data has
been a big obstacle to us; knowing this we should not
fail to acquire data now for the future. We need a better
link between investigations in the upper Bay and those
in the lower Bay. Microcosm experiments are needed to
help clarify the effect of suspended sediments. We must
take a broader view to understand the relationships
between blue crabs and their habitat. We need to
continue testing restoration methods, considering what
effect population pressure may have on the future of
SAV in the 21st century.
TOXICS: Nicholas Fendinger
The area of toxics is extremely complex, involving
the physical properties of the Bay, physicochemical
properties of toxic compounds, and complex interac-
tions with the biota of the Bay. Toxicants entering the
water from agricultural and industrial uses, from spills,
and from fugitive emissions will be distributed through-
out the Bay, and their distribution will be determined in
part by the physicochemical characteristics of the
compounds. Once distributed, toxic materials will both
affect and be affected by the biota. Three areas have
been addressed here: source assessment, dynamics of
pollutants in the environment, and the interaction of
pollutants with biota.
Source assessment. Papers here reported that about
a quarter of the 65 Department of Defense installations
in the Bay watershed had significant effects on water
quality, but most of the impacts were fairly local. Data
were also presented on levels of toxic materials in
effluent from commercial shipyards, and on their levels
of toxicity.
Dynamics. Toxicants accumulate in the microlayer,
possibly affecting blue crab larvae and other living re-
sources. Microlayer concentrations can reach 10-1000
times the concentration in the bulk water underneath.
Photochemical oxidants, such as hydrogen peroxide
and ozone, in Bay water can also influence the behavior
and toxicity of chlorine and chromium in the Bay. The
question was raised whether dechlorination does any
good, or, conversely, any harm. Evidence on its
benefits is mixed. As far as harm, the conversion of
sulfite to sulfate has no toxic intermediates, but some
copper is released in the process.
Relation to biota. Body burdens of PAHs and PCBs
in blue crabs appear to be unrelated to the areas where
the crabs are sampled. This lack of relationship may
refect the blue crab's mobility through its life history;
or it may be due to the shedding of the exoskeleton.
Nutrients are widely considered to be the principal
culprit in the decline of the SAV. However, toxics may
also be important contributors to their decline. Data
were presented here on a bioassay for the phototoxicity
of contaminants on SAV. This toxicity test could be
used to establish water quality criteria for SAV, and
effluents could then be tested for SAV toxicity.
Induceable adaptations have been used in bioassays,
specifically in a stress protein immunodetection assay.
This assay gives a uniform response regardless of
whether stress is caused by temperature, oxidants, or
chemicals, and it is inexpensive and easy to perform. It
may have broad applicability for testing in the Bay.
Toxic effects on fish include four types of deforma-
tions of gill tissue: blood clots, enlargement of cells,
abnormal cell numbers, and broken filaments with
altered growth patterns. Although the deformities
obviously impair the health of the organisms, it is hard
to judge their significance at population levels.
A rapid toxicity test for sediments presented here
may prove to be a reliable test for sediment toxicity;
another useful test is behaviorally based and utilizes the
burial rates of clams.
68
-------
"Where do we go from here?'
Important areas of toxics not covered here include
the important research on the physical dynamics of
pollutant transport, understanding of the basic chemical
properties of compounds, the importance of sediments
and determination of whether they serve as a source or
as a sink, and impacts of toxic materials on growth,
reproduction, immunology, and genetics of Bay orga-
nisms. In addition, the ability to address toxic effects
on populations and communities is in a primitive state.
PHYSICAL PROCESSES: Eric Itsweire and Owen Phillips
This summary has two parts: what we know about
circulation, and what we need to know.
Models. Local models of circulation seem to be
successful. For instance, the circulation model for
Hampton Roads developed by the Virginia Institute of
Marine Science is a good blend of field observations,
simulations, and simple theoretical and numerical
analysis. Models of Baltimore Harbor and the Patuxent
River are also useful for prediction of general features,
but only for average characteristics with an arbitrary
choice of parameters. The model of an estuarine plume
developed by Horn Point Environmental Laboratory
can reproduce basic behavior with simple geometry.
The global model developed by NOAA has not been
very successful, and it is unclear whether it can be
tuned for greater usefulness.
Remote sensing. NOAA research shows Bay-wide
patterns of near-surface sediments in response to high
runoff, but includes no quantitative estimates of sedi-
ment transport. Acoustic and thermistor chain measure-
ments made by Johns Hopkins scientists show transient
mixing events and internal waves at the pycnocline, but
a possible bias in the sampling must be considered.
Measurements of dissolved oxygen and salinity by
HPEL and the Chesapeake Biological Laboratory show
a negative correlation between DO and salinity at the
Choptank-Patuxent transect. Efforts are being made to
use physical data to formulate managerial decisions.
Future prospects. The work of Pritchard has
provided a base to build on, but we now realize that the
impact of transient or sporadic events is still poorly
understood. There is a need to integrate remote sensing
from acoustic, airborne, and satellite sources with high-
resolution in-situ instrumentation in prioritized, well-
planned, well-directed studies. In particular, studies
should address frontogenesis and frontal evolution,
turbulent mixing, and the spatial-temporal response to
transient weather events. The corollary to this effort
should be an intensive investment in modern equipment
for simultaneous observations.
Models need to be developed for description of
transient and local events and the reestablishment of
water characteristics thereafter; these models should be
closely tied to the observational program.
The multidisciplinary approach should be combined
with a careful scrutiny of what long-term studies would
provide useful scientific and managerial information.
Critical but rapid peer evaluation of the results of
research is necessary, with dissemination of the results
in clear, nontechnical language to both the biological
community and managers and legislators. Reviewed,
technical literature is an inefficient means of communi-
cation beyond the physical oceanography community.
It is not enough for communication to the larger
audience to be reliable and complete; it must first of all
be comprehensible.
NUTRIENTS: Carl Cerco
Rather than summarizing all the work that has been
presented here, I will mention several papers that struck
me with the greatest impact. First, the groundwater flow
of nutrients into Chincoteague Bay is being measured,
and the flow is very substantial. Yet very little is being
done in this field. Considering how important nutrients
are and how little is being done, groundwater flow is a
research area that we need to expand in the Chesapeake
system.
Second, the importance of microecology is growing.
Nutrient cycling and respiration need more study in
bacteria, picoplankton, and microzooplankton. We
need more than applied ecology and in situ measure-
ments; we need fundamental research. The picoplank-
ton, and microzooplankton have not been well enough
studied generally.
Third, sediments are a hot topic. The whole question
of nitrification/denitrification is of great importance, as
denitrification is the only exit from the sediments for
nitrate nitrogen. We need far greater coverage of the
Bay with sediment traps; the trap about to be placed is
only the second one in the entire Bay. There is no
substitute for actual measurements.
Finally, there is a great need for more spatial
resolution in data from the lower Bay and the lower
tributaries.
69
-------
PLENARY PANEL
"How do we get there? Who pays the fare?"
Joseph A. Mihursky, moderator
REMARKS BY MONICA HEALY
The great strides EPA, the other federal agencies
involved in the Bay clean-up program, and the other
three Bay states have made in recent years cannot be
exaggerated. Last December they took a giant step
forward by signing a new Chesapeake Bay Agreement
to guide us into the next century. I won't go into the
details of the document because I know other speakers
have already done a thorough job of that. However, I
will say that the agreement may prove to be the most
significant step in the history of the Bay clean-up
program. It sets tough new pollution control standards
for the entire Bay.
This new document gives us reason for celebration,
but it must be tempered with a measure of sobriety.
Government officials will have to bite the bullet and
infuse billions of dollars into the clean-up effort. They
will be required to make the difficult choices between
competing interests to guarantee the Bay's vitality.
Assuring the flow of federal funds will not be easy,
since the administration and Congress are crying to
grapple with the stark realities of our fiscal crisis.
There have been major cutbacks in federal programs in
an effort to reduce the deficit. Because many sources
of federal funds for environmental programs are simply
drying up, states and localities have to carry more of the
burden. The situation will get worse before it gets
better.
The shrinking federal pie will surely reduce the
amount of overall federal funding flowing into estuaries
over the next several years. However, we have a great
deal going for us in our fight to get a pretty good slice
of that smaller pie.
We have, first of all, a highly popular and highly
visible campaign to clean up the Bay. The Bay's close
proximity to Washington makes members of Congress
much more sensitive to its problems, and much more
aware of what the states are doing to address them.
We also have a good case on the merits. There is a
demonstrated need — a $27 million EPA study has
documented that our nation's largest and most produc-
tive estuary is dying.
Moreover, we have tremendous political clout in
Congress. The House members and Senators who
represent the Bay states have formed the Chesapeake
Bay Congressional Caucus. They are a potent political
force in Congress and have been very successful in the
past in capturing federal funds and passing legislation
to further our efforts. Furthermore, it certainly doesn' t
hurt that many members of Congress have direct,
personal connections to the Bay. When President
Reagan singled out the Bay during his 1984 State of the
Union Message to a joint session of Congress, a
reporter asked a friend of mine why there was such a
big round of applause for such a parochial concern.
"Simple," my friend told the reporter, "they all have
boats on the Bay."
Finally, the lessons of the Bay clean-up program will
serve as a useful model for similar campaigns for other
estuaries around the country. Thus any investment in
the Chesapeake will reap benefits far larger than a
healthier Bay.
In sum, it's a good news/bad news scenario for the
Bay in terms of prospects for future federal funding.
The bad news is there is limited funding available
for programs to aid the clean-up effort The good news
is that we have all the right ingredients to give us a leg
up on other states who will be competing for the same
money. If we use those ingredients to our advantage, if
we work even harder at our lobbying efforts — and that
means grassroots support from all of you — and if we
continue to work as a team, I'm optimistic that we can
be successful in obtaining the federal monies and
programs needed to return the Bay to a productive,
healthy state.
You're probably wondering how can you help
further our cause on Capitol Hill. You can and should
be an integral part of our grassroots lobbying efforts.
You have done a tremendous job in producing the
scientific evidence that has far-reaching ramifications
for the future of the Chesapeake Bay, and that research
can make a real difference during the decision- making
processes on Capitol Hill.
But the research is of little use in Congress unless
policymakers can understand its value in practical
terms. To be blunt, you have to spoon- feed them.
First, politicians and their staffs often have little or no
scientific background, so it's important for you to
translate what you're doing into the simplest possible
terms. For instance, many don't even understand basic
terms like "non-point source pollution." It helps to
70
-------
"Who pays the fare?'
explain that in language they can understand. Second,
you have to show them how your research translates
into tangible action for the Bay. Often, scientists come
to the Hill and talk about the wonderful research they're
doing, but what's going on in politicians' minds when
tney listen to that is, what good is it for the Bay? For
example, you have to explain how scientific research
has led to specific standards for discharge permits. Or
you could tell them how research led to the ban on
TBT.
If you explain science in plain language and show
politicians how what you do translates into tangible
action, I can guarantee that federal research funds will
continue to flow — as they should.
REMARKS BY CHRIS D'ELIA
I am the Program Director of the Biological Ocean-
ography Program at the National Science Foundation,
and I have been very appreciative of the efforts Ed
Houde made while in the same position several years
ago. His work on strategic planning has helped make
NSF more active in the integration of future research.
Within the Ocean Sciences Division there are four
research programs: biological oceanography, chemical
oceanography, marine geology and geophysics, and
physical oceanography. Out of a total budget of
roughly $140 million this fiscal year, the strict research
programs garner about half the funds. The other half
goes to the Ocean Centers and Facilities Support, which
pays for the ships and other infrastructure necessary for
ocean research.
The budget in Biological Oceanography this year is
about $15 million, of which about $3 million goes into
research in the estuarine and coastal areas. Our budget
this year did not do as well as we had hoped. Before
the stock market aberration in October we had about a
19% increase, but what we actually received in the
omnibus appropriation act was about a 3% increase,
which means we are not quite keeping up with inflation.
To make the best use of these limited funds we are
trying to plan more actively than in the past. The NSF
works on a peer-reviewed system in which an individ-
ual investigator with an idea puts together a proposal
that then goes out for peer review. This system is good
at turning the best ideas into meaningful long-term
output. The problem is that it is very bad for strategic
planning and for directing research in a particular area.
Thus we are trying to develop budget initiatives,
specific areas we feel need to be stressed more, and we
are asking the scientific community to tell us what
needs to be done in these areas.
Our primary efforts currently are involved with the
Global Ocean Flux Study (GOFS), which deals with the
large issues of global climate, the CO2 effect and the
associated problems. Specifically there is interest in
transport of carbon to the deepsea sediment reserves of
carbon, the overall impact of this process, and the
various trophic pathways involved. Another initiative
is a study of recruitment: where do organisms come
from and how do they interact? The biotechnology
initiative in Biological Oceanography has focused par-
ticularly on molecular biology and genetic approaches
to understanding oceanic processes and ecology.
Ridge-crest processes, which involves the study of
deep-sea vents and rifts, has a small share for us.
The most important initiative to us here is the land-
sea initiative. We need to have the scientific commu-
nity be very active in developing a land-sea initiative
that will be a specific budget line item in the NSF
budget. A small amount is available in the next request,
and we have a small program jointly with another NSF
division, the Ecosystems Dynamics Program — the
land margin ecosystem research initiative. This ini-
tiative has about $700,000 available this year, which
obviously will not go far in the long run. We are trying
to develop a significant enhancement to the NSF budget
that will involve a lot of community effort in putting
together planning documents and stating what the needs
are. I hope to see substantial results from this effort
within five years, because we understand from top
management in NSF that we should not expect en-
hancements to our core program; we need to talk about
more integrated programs. Congress needs to see spe-
cific items when we talk about budget enhancements.
REMARKS BY JOHN W. DANIEL
Earlier this month we funded $55 million for
Chesapeake Bay programs. A large percentage of this
money is direct contribution to the scientific commu-
nity. An observer from the political arena such as
myself who may not understand exactly what you
scientists do can either ignore the intricacies of your
research, or develop a healthy respect for it. I hope I
have developed a healthy respect for it.
We have come to a point in our Chesapeake Bay
programs that the level of scientific contribution to
policy and programs has become more important, with
a prospect of increasing importance in the future. Some
71
-------
PLENARY PANEL
years ago, for instance, the relationships between open
privies near streams and tributaries, degraded water
quality, and the negative impact on shellfish were
evident — to scientists and to non-scientists as well.
But contrast this to the absolute essential scientific base
necessary to draft and approve and implement a
prohibition on the use of tributyltin paints. There is a
substantial difference. Without that scientific basis and
the capacity to perform the research, that particular
action by states bordering on the iay to improve water
quality simply could not have occurred.
Thus in 1988 there is a much greater acceptance in
the non-scientific community than there would have
been in the past for a scientific conference with very
technical topics, and there is a willingness to address
the economics necessary to accomplish that research.
Policy decisions and regulations must be well founded
on scientific fact. To be successful, regulatory pro-
grams must be legally defensible, and that can only be
achieved with a sound scientific basis.
All of this is expensive, and who will pay the fare?
It is incumbent on those of us responsible for policy
decisions, especially at the state levels, to coordinate
scientific research with our regulatory programs. It
seems to me that, as with TBT, the link between the
science and the utilitarian value of that research to pro-
grammatic functions is a tremendous catalyst to helping
you pay the cost. By virtue of the Bay Agreement, for
instance, the states are obliged to move forward with
toxic control programs in our respective water control
agencies. The solutions to those concerns will be
individualized to the jurisdictions, and they will not be
identical. They will be alike, however, in requiring a
sound scientific basis to be legally defensible. If we
don't do this, we move backward rather than forward.
So who pays? The easy answer is that we all pay.
Federal, state, and local governments pay, the regulated
community pays through capital improvements and
technological investment, and individuals pay with
changed habits and investments of time and energy.
The private sector and foundations pay with grant
funding and charitable contributions.
We need, however, to maximize what is produced
with these payments. This necessitates cooperation
rather than competition, and the beneficiary of these
efforts must be not individuals but the Chesapeake Bay.
If we can show the regional and Bay-wide value of a
dollar spent in any individual jurisdiction and demon-
strate to those who pay how that science relates to
useful programs, we benefit the understanding of the
estuary and we demonstrate the necessity of continued
funding for advances in Chesapeake Bay research.
REMARKS BY FRANK PERKINS
We are working in an exciting field, one that is ap-
pealing to the general public. There are individuals
who have the resources and ability to fund our work at
a very high level. We are able to make a presentation
to a potential donor that is fascinating, exciting, fun,
and shows the practical value of our work. We have a
lot going for us; and we can look to the Cousteau model
for effective presentation of exciting material. The
creatures of the Bay are just as exciting as those in the
open ocean highlighted by Cousteau.
One of the things that corporations respond to
readily is the need for superior instrumentation. An
effective selling point is that an institution can compete
much better for federal funding when the expense of
instrumentation does not have to be added to proposal
budgets. As an aside, I might note that a number of
federal agencies have excellent shops and fabrication
plants (NASA comes to mind right away), and we as
marine scientists need to establish relationships with
those agencies because a lot of the instrumentation we
need can be constructed de novo. When we approach
private donors, especially individuals, we must make
them aware of what we are doing and try to involve
them to some extent in day-by-day operations. Obvi-
ously this kind of thing if overdone could be a real
burden, but done properly, the hands-on experience in
'he living world can be a powerful aid in persuading a
potential donor to support our work.
Private foundations seem to be orienting themselves
increasingly toward public policy and resource manage-
ment, leaving it to the federal agencies to take care of
the basic research in biological and physical sciences;
but the potential of private foundations should not be
dismissed.
User fees, which have already been mentioned, may
be an overlooked source of funds.
As we develop programs for approaching the private
sector we must accept the necessity of diverting some
funds into the use of video to capture the imagination of
the public. This is hard when the money is diverted
from research, but video is a tremendous tool and we
should be making better use of it.
More specifically, we must be concerned not only
about the health of the Bay overall, but also about the
health of organisms per se. We need a strong branch
analogous to medicine or veterinary science, to deal
with this. And this kind of work should be very easy to
find support for among private donors.
72
-------
"Who pays the fare?"
State funding is being covered by other speakers, but
one obvious point that can use emphasis is that when
you go to the legislature or the executive branch, you
should be prepared to propose something that has a
short-term payoff. We know enough about the system
that in some areas we can offer a proposal with a high
probability of success — oyster culture comes to mind.
This kind of proposal is appealing to those who have to
justify their appropriation of funds to constituents
expecting results.
REMARKS BY IAN MORRIS
The remarks to follow reflect a personal bias and
prejudice. They have not been subjected to the usual
rigor of literature comparisons for any originality or to
peer review for any intellectual validity. They should
be read with these apologia clearly in mind.
At the end of this conference, I want to make three
brief comments. Firstly, I would like to congratulate
the organizers on creating a highly successful gathering,
and to highlight one of the most prominent features of
the past two and a half days. I do this, not because this
prominent feature is hidden to all others here (it has
been a feature of much of the coffee discussion), but
more because it will lead me into the two other points I
wish to discuss. Rarely, if ever, does one see at a
technical conference such as this a mix of scientists,
government officials, and interested citizens such as we
have seen here. It is one of the most intriguing features
of recent and current activities in the Chesapeake Bay
region. Scientists want their work to matter to those
charged with making decisions on the future of the Bay.
These decision-makers want to hear the scientists and
the citizens. Citizens want to learn of the latest
technical information and thinking, and to influence
further studies and future decisions. All of this is a
major feature of the Chesapeake Bay today. There is
no better example of it than in the attendance at this
conference. It might be argued, too, that any continued
success in addressing the "problems of the Bay" will
depend on this closeness of interaction between
scientists, citizens, and government. Yet the nature of
that interaction must be right, with each segment
playing its appropriate role.
If the above comments are true, I would like to
suggest that they have some profound implications for
all of us. Here, I want to address only two. The first
concerns what can and should be expected from the
scientific and technical community. The second
concerns the need for truly transdisciplinary input of
information into any decision-making process about an
ecosystem and its natural resources. I shall try to make
my remarks relate to this conference and so be slightly
relevant to the title of this final session.
Much (but not all!) of the research presented at this
conference is supported by regulatory agencies; that is,
by agencies whose primary responsibility is to recom-
mend the passage of certain laws and to enforce them,
so as to lead to improved management of an important
natural resource such as the Chesapeake Bay. Inevita-
bly, such agencies need research which guides their
decision-making process and which supports them once
decisions are made. Under conditions of limited
resources and of rigorous accountability in the expendi-
tures of public funds this need becomes even greater.
Thus, such agencies are severely constrained so that
support for any specific piece of research must be
justified by pointing to a direct link between it and a
specific decision or action. Similarly, a scientist who
lives in the real worM knows that support for a piece of
research will depend on arguing for such a direct link.
Yet there seem to me to be real dangers in arguing such
immediacy and directness between specific research
and specific management actions.
On other occasions, I have emphasized the dangers
to the scientist, particularly the younger one, where
such constraints can severely affect the creativity of a
particular study. Here, however, I wish to make a
different point. It may be that we in the environmental
scientific community cannot deliver this directness and
immediacy of connectedness between specific research
and specific management needs. It may be, too, that the
sponsoring management agencies are misled in asking
for such direct linkages. Rather, we might argue, the
purpose of research on a particular ecosystem or
environmental problem is to weave a tapestry of
knowledge which would provide a framework of
understanding, within which specific management
actions are made. Perhaps, too, we should recognize
any particular decision or action improved because of
such a framework and not because of any specific
research. It would be revealing to analyze the historical
record of management decisions on the Bay within this
context. It may be that past decisions have benefited
more from our increased understanding gained over
years of study and not from a specific piece of research.
If this is true, it seems to me that it has profound impli-
cations for the way in which regulatory agencies can
benefit from, and therefore should sponsor, research.
My second point concerns the crucial interdiscipli-
nary nature of research needed to provide the frame-
work of understanding mentioned above. We often
73
-------
PLENARY PANEL
stress the need for interdisciplinary studies by pointing
to the desirability of physicists, chemists and biologists
working together. But I think this is missing the point.
The system of the Chesapeake Bay is much bigger than
the water, the sediments, and the aquatic living re-
sources. In the Chesapeake Bay the ratio of watershed
area to water volume is very high. Thus it is crucial
that we expand our horizons beyond the edge of the
water and go into the vast watershed. When we do, we
shall realize that the Chesapeake Bay system includes
the land use, the demography of population, the culture,
the history, the economics, and the politics. Decision-
making must be done against a background of aware-
ness of this system "as a whole." We should be
speaking of an "information-management" interface
and not the narrow "science-management" interface.
And the information needs to cut across disciplines.
REMARKS BY DAVID CHALLINOR
As manager of a large research budget one of the
things I have to do is find funds to support it, and this
can require a lot of imagination as well as hustle. For
example, the Smithsonian was given a part of a group
of islands on the Eastern Shore that about 200 years ago
formed one large island of probably 1000 acres. For
the last 200 years this island has been disintegrating
rapidly, and now there are half a dozen small islands
that erode at least 10 feet a year. The Smithsonian was
under a great deal of pressure to "do something" about
this erosion, particularly because these islands harbored
one of the largest great blue heron rookeries on the
Chesapeake Bay. We spent more than $100,000 trying
to put in bulkheads, which lasted about two years.
Finally in a very wise move the Smithsonian sold those
islands. In 20 years they may well be gone.
While we had the islands, however, it was important
to try to figure out what we should do. One suggestion
was to build them up, reversing nature's process. A
large company proposed to use the islands as a reposi-
tory for plastic-baled rubble from the redevelopment of
Baltimore and Philadelphia. Biologists were opposed,
fearing among other things the release of toxic metals.
The developer painted a rosy picture of how the rubble
would be brought in by barge and used to create a
lovely hilly island covered with blue heron rookeries.
We were able to get a private foundation to give us
some money to find out just how practical these ideas
were. It was found that the likelihood of punctured
bales leaking toxic materials was fairly high, and it
became clear that although the project might solve
someone's waste disposal problem, it was unsound
from an environmental point of view. In any case, we
were able to get funding from the company that wanted
to dispose of the waste and also from a private founda-
tion that was simply intrigued by a rather interesting
idea. We at the Smithsonian put in a little of our own
money as well; companies and foundations are more
likely to feel comfortable spending their money on your
project if you show them that you are willing to spend
your own on it.
So I see two things as useful: first, put some of your
own money into the project to make it more "believ-
able;" and second, somehow lure the key characters
into the lab or onto the vessel where their excitement
and interest gives you a high likelihood of success.
REMARKS BY SHELDON SAMUELS
Science requires public support, but to gain it does
not require that we respond to the immediate or even
the long-range needs of society. The practical demands
for understanding our urgent needs in the environment
or other issues are driving forces in every field. Science
is also driven by technology. We are, as Bergson per-
ceived, Homofaber, man the tool-maker. But we also
remain Homo sapiens, man the seeker of knowledge.
Sadly, Homo sapiens is more often forgotten than
remembered. The changes in federal disbursements
traditionally earmarked for science are an example.
Funds are being diverted from science to the regulatory
agenda of our federal government at an unconscionable
rate. It is not only unwise, it is dangerous to place the
future of science in the hands of the regulator.
An example from personal experience began in
1979, when the federal government decided to regulate
benzene as a carcinogen in the workplace. The
petroleum industry challenged this and brought it to the
Supreme Court. My organization challenged the
petroleum industry. The Supreme Court's decision was
one I did not quite understand. I understood risk
assessment, and in fact have written several peer-
reviewed papers on quantitative risk assessment, but I
did not understand what the Supreme Court meant
when it said that priorities for regulation ought to be
based on risk assessment. Kyler Hammond, senior
vice-president of the American Cancer Society and the
74
-------
"Who pays the fare? "
leader at the time in the field of risk assessment,
advised me, when I asked about this requirement, that I
should not worry, that the back of an envelope and a
pencil stub would be adequate. By 1984, however, they
were not enough. Mr. Ruckelshaus gave his famous
speech at the National Academy of Sciences on how
use of risk assessment would speed up the regulatory
process. In reviewing a draft of his speech I told him
that this would not speed up regulation.
Four years later the federal government is now
spending $50 million a year on quantitative risk
assessment research, and regulation is not speeded up.
And $5 million of this money came from the National
Science Foundation, diverted directly from basic
research. Risk assessment cannot and will not speed up
regulation, and it is a perfect example of how funds are
unwisely diverted into the hands of people who have
some immediate political needs, who reshape the
science itself. Homo sapiens has not made his case.
In 1944 a famous BBC broadcast involved an
interview between the nuclear physicist Polanyi and the
philosopher Bertrand Russell, whose principles of
mathematics formed part of the basis for quantum
physics. The discussion was on nuclear energy
research: was the work begun with an application in
mind? Would there be a practical result? Both Polanyi
and Russell said no. Thirty days later the bomb
dropped on Hiroshima.
There are two lessons here. One, there was no way
to forecast that there would be an application. Two,
there was long-term support from the public for that
kind of research. Remember that Einstein's paper on
special relativity was given at a conference sponsored
by a mining company, which expected no practical
application of the theory but sponsored the conference
regardless — because of fascination and curiosity, and
because of the realization that the good and the true is
often indistinguishable from the beautiful. The public
is us. We are not basically different from the banker or
the baker.
We need to do a better job of making our case for
science qua science, and also to use our political powers
to protect our programs. At the same time we need to
do a better job of understanding what other disciplines
are contributing to an understanding of the problems for
which we can supply only one facet of the solution. We
need only remember a principle of science that we
constantly assume — the continuity of nature. It makes
sense to look at cancer in fish as evidence of human
carcinogenesis. It makes sense to look at the work
environment for evidence of what may occur in other
species. It makes sense to model the Bay, using
techniques borrowed from the physical sciences. It
makes sense to use techniques of population study
ecologists have developed to analyze the dynamics of
human populations. It makes sense first because it is
heuristic, from a scientific point of view, and second
because the public perception is of the whole, and we
must respect that perception if we expect public support
to be forthcoming.
QUESTIONS FROM CONFERENCE PARTICIPANTS
Q: There has been a lack of concern here for the edges
of the Bay, the wetlands. In the Chesapeake Bay there
is an important gap that would tie the biological and
energy contributions of our marshes, both fresh and
salt, to the productivity of the Chesapeake. This is
important right now, because at least in Maryland
extensive mosquito control operations, replacing pesti-
cides with upper marsh management, are destroying
many of the natural marshes around the Chesapeake;
and I am not aware of anyone anywhere monitoring this
kind of activity. It is justified by those doing it as being
better than the pesticide alternatives, but someone
should be taking some interest in the fate of our salt
marshes as a result of these destructive management
practices.
A: As Ed Houde pointed out, this particular conference
did not cover the whole Chesapeake world. Our
embryonic plan is, if a conference is held in 1990, to
cover a range of different topics.
A: NSF is funding studies of marshes, and we have
made progress in understanding them, but it is true that
this work has not been done in the Chesapeake.
Q: Most of the work NSF has funded has been in the
coastal regions of Georgia, South Carolina, and New
England - areas not suffering declines - and these
studies have shown the nutrient dynamics and docu-
mented the value of marshes to coastal food chains.
We need similar studies in the Chesapeake Bay,
especially in the mid-Bay region where rapid changes
are occurring. We need to know what these changes
mean for resources in the Bay.
Q: Another topic omitted from this conference is
dredging and its many impacts.
A: A call for papers went out around the Bay, and if
topics have not been addressed here it is partly because
the people doing the work are not in attendance here. If
everyone were aware that a Chesapeake-based confer-
ence would be scheduled every two years, people
would have a better chance in the future to contribute to
the conference and get their interests discussed.
75
-------
PLENARY PANEL
A: Under the Bay Agreement, a comprehensive re-
search plan is to be developed, and scientists around the
Bay are integrally involved in contributing ideas for it.
Q: How can we institutionalize the acquisition of long-
term datasets?
A: Give money to the lab directors and trust them.
A: The question of institutionalizing data sets is one
that concerns a lot of us. This country is not using its
computers and networks to make data really available.
We need a central, independent, quasi-governmental
data organization. The kind of data that people here
collect is directly relevant to human ecology and should
not be hidden away in specialized databanks.
A: One thing that would help support the long-term
monitoring effort would be an agreement at least
between the governors of Virginia and Maryland that
would commit those states to monitor in the long term.
76
-------
CONCURRENT SESSIONS
AND
POSTER SESSION:
LIVING RESOURCES
Chairs:
William Hargis and Linda Schaffner
Virginia Institute of Marine Science
College of William and Mary
-------
Understanding the Estuary: Advances in Chesapeake Abstract only
Bay Research. Proceedings of a Conference. 29-31
March 1988. Baltimore, Maryland. Chesapeake Research
Consortium Publication 129. CBPlTRS 24/88.
Effect of Changing pH and Salinity on
Chesapeake Bay Striped Bass Larvae
Andrew S. Kane, Richard O. Bennett,
and Eric B, May
University of Maryland
School of Medicine
Department of Pathology
10 South Pine Street
Baltimore, Maryland 21201
A laboratory study was conducted to examine the effects of rainfall
phenomena on striped bass (Morone saxatilis) larvae under conditions known to
occur in Chesapeake Bay striped bass spawning grounds. The two variables
under observation were salinity dilution and pH depression as they affected
hatchery-reared 10, 13, 16, 19, 22 and 30 day old (post-hatch) larvae from
Choptank River, a tributary of the Chesapeake Bay, broodstock. A continuous
flow dosing system was developed to expose larvae to three pH depression
regimes by metering three pH-controlled test media (pH 7.5, 6.5 and 5.5) into
the exposure chambers. Initial pH for all exposures was 7.5. Final pH in the
exposure chambers after 24 hours was 7.5, 6.8 and 6.2 for each pH depression
regime respectively. The system also controlled salinity maintenance at 1.1
ppt as well as dilution from 1.1 ppt to 0.6 ppt over a 24-hour exposure
period. Tests were conducted at 18°C. Test media consisted of "soft
reconstituted" bioassay water with a hardness of 42-45 mg/L (as CaCO-j) .
Statistical analysis revealed that age, pH and salinity were all
significant variables affecting larval mortality. Salinity reduction
significantly increased mortality in fish exposed at pH 7.5, 6.8 and 6.2 at 10
days of age, pH 6.8 and 6.2 at 13 days, and pH 6.2 at 16 days. Although not
statistically significant, elevated mortality in salinity reduced exposure
groups was empirically observed in all age groups (tapering off with age)
throughout all pH regimes. Organisms in all age exposure groups maintained at
1.1 ppt salinity were not significantly affected by changes in pH. Further,
pH depression did not significantly affect mortality in larvae exposed to
reduced salinity at 19, 22 and 30 days of age.
The data indicates that survival of 10 day old striped bass larvae is
dependent on maintenance of 1.1 ppt salinity independent of pH. This is
supported by elevated mortalities obtained at all three pH regimes under
salinity reduction. At reduced salinity conditions pH appears to be important
to larval survival at 13 and 16 days of age- At 1.1 ppt the effects of pH on
mortality is negated. Mortality due to salinity reduction was significantly
decreased at pH 7.5, 6.8 and 6.2 with 13, 16, and 19 day old larvae,
respectively. Heavy rainfall events temporally occurring during critical
periods of larval development may significantly affect young-of-the-year
recruitment.
77
-------
Understanding the Estuary: Advances in Chesapeake Abstract only
Bay Research. Proceedings of a Conference. 29-31
March 1988. Baltimore, Maryland. Chesapeake Research
Consortium Publication 129. CBPlTRS 24/88.
Grazing and Egg Production by Chesapeake Bay
Zooplankton in Spring and Summer
Jacques R. White and Michael R. Roman
University of Maryland
Horn Point Environmental Laboratories
P. O. Box 775
Cambridge, Maryland 21613
Grazing of primary production by zooplankton (> 200 urn) was measured in
the surface water of the mesohaline portion of Chesapeake Bay in May and
August of 1986 and by zooplankton (> 64 um) in March, May and August 1986.
Egg production rates were also measured for the dominant copepod species
Acartia tonsa in May and August of 1987. Weight specific filtering rates of
zooplankton populations, averaged over 24h, were higher in August (333 and 841
ml/mgC/h for 1986 and 1987 respectively) than in May (205 and 57 ml/mgC/h).
Average water column grazing pressure was up to an order or magnitude higher
in August (3.6 and 14.2 L/nr/h for 1986 and 1987 respectively) than in May
(0.8 and 0.4 L/nr/h) , due to increased zooplankton biomass during August for
both years. Copepod nauplii had the highest grazing impact of zooplankton >
64 um in March (0.5 L/nr/h) and August (3.0 L/m^/h) of 1987 while polychaete
larvae had the highest grazing rates in May (0.06 L/m /h) of that year. We
estimate that zooplankton > 64 um can remove 23% of the primary production in
March, 7% in May and 79% in August. Water column egg production rates of
Acartia females averaged 0.37 mgC/m /d and were not significantly different
between May and August, even though nauplii biomass and numbers were twice as
high in August. Therefore, the discrepancy between production and standing
stock of nauplii in the mesohaline portion of the Bay during May 1987
represents a loss to predation or advection.
78
-------
Understanding the Estuary: Advances in Chesapeake
Bay Research. Proceedings of aConference. 29-31
March 1988. Baltimore, Maryland. Chesapeake Research
Consortium Publication 129. CBPlTRS 24/88.
The Effects of Suspended Sediments on Microzooplankton Grazing in the
Patuxent River, A Subestuary of the Chesapeake Bay
David C. Brownlee, Kevin G. Sellner,
and Kevin R. Braun
The Academy of Natural Sciences
Benedict Estuarine Research Laboratory
Benedict, Maryland 20612
INTRODUCTION
As a eutrophic temperate zone estuary, Chesapeake Bay is typified by
high phytoplankton densities and chlorophyll concentrations
throughout the system. Nano- and microphytoplankton densities exceed
10 cells/1 and chlorophyll concentrations range from several to
several hundred /ig/1. The abundant primary producers should support
an active planktonic suspension feeding community, ultimately passing
the abundant primary production to higher trophic levels in the
nekton and benthos. In Chesapeake Bay, however, fish and shellfish
stocks have dramatically declined over the last several decades
suggesting some limitation in the Bay from the top (e.g.
over-fishing) or from below through reductions in food levels and
quality, habitat, etc. (Verity 1987).
Because phytoplankton standing stocks are high and cell size and
quality of the Bay's eucaryote phytoplankton is similar to
phytoplankton diets supporting high zooplankton growth in other
estuaries and cultures, we hypothesized that some factor might limit
the quantity of phytoplankton carbon ingested by the planktonic
secondary producers in the watershed, i.e. that feeding in the
plankton might be reduced resulting in lower secondary production and
ultimately biomass of higher trophic levels.
Over the last fifty years, water clarity has declined, as documented
by a decrease in secchi disc depth (Heinle et al. 1980). The
79
-------
decline is attributed to an increase in total suspended particulate
material as phytoplankton and inorganic suspended sediments. The
latter particulate pool, suspended sediments, has increased due to
population growth in the watershed. For example, Schubel (1981) has
estimated that there has been a 100-fold increase in sediment loading
in the Bay since colonial times due to urbanization and suburbani-
zation. Land use also dramatically alters sediment loading rates to
the Bay with minimal input from forested lands (0.5-49 Ibs/acre/yr)
and highest contributions from croplands and residential areas
(11-2461 Ibs/acre/yr; EPA 1982).
Ambient concentrations of suspended sediment range from 1-600 mg/1
in the Bay and its tributaries (Schubel 1968; Roberts and Pierce
1976; Bennett 1983; OEP 1984). This mixture of non-nutritious
inorganic material and phytoplankton provides a suite of potential
food items for the planktonic suspension feeders and, potentially,
inadequate energy intake should sediments interfere with ingestion or
digestion of the planktonic autotrophs.
Deleterious effects of high and moderate concentrations of suspended
sediments on mesozooplankton have been reported (Arruda et al. 1983;
McCabe and O'Brien 1983) though little research has been conducted on
the effects of suspended sediments on microzooplankton. The size of
suspended sediments in the Chesapeake Bay (>90% less than 3um,
Schubel 1969) overlaps the optimal food size for many microzoo-
plankters (Spittler 1973; Heinbokel 1978a,b; Fenchel 1980;
Rassoulzadegan and Etienne 1981; Rassoulzadegan 1982). The presence
of these predominantly non-nutritious fine particles could interfere
with the feeding and, subsequently, growth of these organisms. If
microzooplankton feed only on the basis of size and not "taste"
(suggested by Fenchel 1980) and these inert particulates represent a
significant portion of ingestion, then reduced growth should occur.
Ciliates are known to ingest inert particles, e.g., plastic beads
(Fenchel 1980; Borsheim 1984; Albright et al. 1987) and even iron
particles (Rifkin and Ballentine 1976). Microzooplankton could also
respond to non-nutritious particulate matter by reducing their
feeding rates and therefore total ingestion also resulting in reduced
growth. If some microzooplankton also exhibit chemoreception, as
noted by Stoecker et al. (1981) for a large tintinnine ciliate,
ingestion of significant quantities of the silt particles might be
avoided, but time spent in handling and rejecting inert particles
would decrease ingestion rates of nutritious particles.
Positive effects of suspended particulates on ciliate feeding and
growth can also be postulated. Because sediment particles
accumulate dissolved organic material and microbial (e.g. bacterial)
populations, the organic-inorganic aggregates might supply needed C,N
and P resulting in an increase in ciliate growth rates. The
association of bacteria and suspended particles has been repeatedly
described (e.g. Goulder 1977; Pedros-Alio and Brock 1983). For some
species of ciliates grown in axenic and organically rich culture
conditions, inert beads must be added to the media to promote growth;
beads concentrate nutrients on their surfaces and provide
organic-rich particulate material for ingestion. Albright et al.
(1987) found that ciliates isolated from marsh floe feed more
80
-------
efficiently on bacteria associated with particulates (beads) while
ciliates from a tidal creek showed preferential feeding on unattached
bacteria. With these alternative hypotheses in mind, studies were
designed to examine the effects of suspended sediments on feeding in
microzooplankton.
METHODS
The effects of increasing concentrations of suspended sediments on
microzooplankton grazing rates were studied at the Benedict
Estuarine Research Laboratory, Benedict, Maryland from winter, 1985
through summer, 1986. All species examined and water samples used in
the experiments were collected from the laboratory pier. The
sediment concentrations tested ranged between ambient (19-62) to 246
mg/1. Radioisotope labeling techniques were used in short term
experiments (4-10 h) initially using a single label (l^C-bicarbonate,
and subsequently using dual labels (^ C-bicarbonate and ^H-methyl
thymidine) so that clearance rates for both phytoplankton ( C) and
bacteria (^H) could be measured.
Test Organisms
Test organisms included representatives from the three dominant
microzooplankton taxa in the Chesapeake Bay, i.e. rotifers,
tintinnine ciliates, and oligotrichine ciliates. The rotifer
employed, Synchaeta sp., had a mean length (205 /im) just slightly
greater than its width in the preserved state. Morphologically this
species was very similar to S. baltica though approximately half its
size. The tintinnine ciliate studied was Eutintinnus angustatus
which has a hyaline lorica with length and oral diameter of 205 and
60 fan, respectively. The oligotrichine ciliate Strombidium sp. was
also studied. This ciliate was rather rounded with a mean length and
width of 46 and 43 /urn, respectively, and the length of its preoral
groove was 60% of its total length.
Sediment Preparation
Sediments were collected from farmland north of Harwood, Maryland.
Sediments were ground with mortar and pestle until fine, mixed with
Patuxent River water, and allowed to settle for 9 minutes; the upper
two thirds was subsequently decanted through a series of screens
(505, 202 and 153 ^m mesh nytex nylon). The resulting slurry was
mixed with unfiltered river water to obtain the desired final
sediment concentrations.
Experimental Protocol
The dual labeling method of Lessard and Swift (1985) was used with
minor modifications. Initially, only ^C was used but subsequent
studies used dual isotopes as discussed above. Lessard and Swift's
method involves: (1) gentle concentration of the microzooplankton
with a fine mesh net which should concentrate the microzooplankton
without affecting the available food concentration; (2) addition of
label to the concentrate; (3) incubation (4 h); and (4) isolation of
the organisms from the radioactive milieu by rinsing then
micropipetting individuals into scintillation vials for subsequent
estimation of activities.
81
-------
Adapting this method, originally used in oligotrophic waters, to the
relatively eutrophic Chesapeake Bay required some modifications. In
preliminary experiments, it was found that the concentration step (1)
not only concentrated microzooplankton but the larger algae as well.
The carbon fixation of the concentrate was two fold higher than the
rate in unconcentrated water. To eliminate this problem net
concentrated microzooplankton were micropipetted through at least one
wash in unfiltered river water and then into the experimental vessels
(55 ml test tubes) containing unfiltered river water. The tubes were
spiked with isotope, mixed and placed in a water cooled incubator
equipped with flourescent lighting. Following incubation, individual
animals were micropipetted through a series of filtered seawater
washes. An analysis of the radioactivity remaining in water from
each of six rinses of the organisms following an experiment suggested
that 6 rinses were necessary to remove residual non-fixed label. In
addition, a time-zero control was introduced to control for surface
attachment of the label and feeding during the rinse period. The
controls were set up in an identical manner to the experimental
tubes; however, after the addition of label the tubes were
immediately sacrificed.
RESULTS
The relationship between sediment concentration and clearance rate
for Synchaeta sp. was determined in a set of experiments conducted on
February 5, 1986 (4.4 °C, 12.7 °/oo). Though the clearance rates
were relatively low, a significant increase in grazing rate was found
at moderate sediment concentrations (38 mg/1) relative to ambient
(Figure 1). Clearance rates at 120 and 136 mg/1 were not
significantly different from ambient while clearance rate at; the
highest sediment concentration was significantly lower than at any
other sediment concentration.
Concurrent with this experiment, an incorporation experiment was
conducted in which the rotifers, after having fed on labeled food
for eight hours, were placed in unlabeled food and the loss of label
monitored approximately every 2 hours for 10 hours. The results
indicated a consistent 5% loss rate per hour.
In a later experiment (July 23, 1986, 29.0 °C, 12.8 °/°°)» the dual
label method was used to compare the response of clearance rate on
phytoplankton and bacteria to increasing suspended sediment
concentrations. A tintinnine ciliate Eutintinnus angustatus was the
test organism in these experiments. Clearance rates on both
phytoplankton and bacteria were stimulated at moderate sediment
concentrations and decreased at the highest levels (Figure 2).
However, the peak in grazing rate on phytoplankton occurred at
74 mg/1 suspended sediment and the highest sediment concentration
resulted in a clearance rate significantly less than that at the
ambient sediment concentration; in contrast, the clearance rates on
bacteria peaked at 99 mg/1 and the rate at the highest sediment
level was significantly greater than the rate at the ambient
suspended sediment level. This may be due to the ingestion of
bacteria adhered to sediment particles.
82
-------
CLEARANCE RATE (ul/hr/ind)
O O
£) C/f
N —
50 £
C" r*
^ O
ffi
' >—.
. (70
ro
C/J
co
ro o
o 01
cr1
^ 0)
co <
O5 0)
CO
CD
CO
M
s:
H
O
O
2!
O
2:
50
O
o
en
p
b
O
CO
CO
CO
O
CO
o
en
o
Figure 1. Relationship between suspended sediment concentration and
clearance rates of phytoplankton by the rotifer Synchaeta
sp. In the Duncan's test, a break in the underline
denotes a significant difference.
83
-------
•a
e
•HH
W
w
o
2
OS
<
Cd
U
<
ct:
ui
O
0.36
0.32 -
0.28 -
0.24 -
0.2 -
0.16 -
0.12 -
0.08
60 100 140 180 220
SEDIMENT CONCENTRATION (MG/L)
Duncan's Lowest to Highest (0.05 Level)
14C Grazing Rate: 246 32 93 99 74
260
20 60 100 140 180 220
SEDIMENT CONCENTRATION (MG/L)
Duncan's Lowest to Highest (0.05) Level)
3H Grazing Rate: j}2 246 93 74 99
260
Figure 2. Relationship between suspended sediment concentration and
clearance rate on phytoplankton (above) and bacteria
(below) by the tintinnine ciliate Eutintinnus angustatus.
In the Duncan's test, a break in the underline denotes a
significant difference.
84
-------
A similar study was conducted for Strombidium sp., an oligotrichine
ciliate (August 6, 1986, 29.6 °C, 11.8 o/oo). Clearance rates on
bacteria followed a similar pattern to that demonstrated by E.
angustatus feeding on bacteria with highest clearance rates at about
100 mg/1 suspended sediment (Figure 3). However, clearance rates on
phytoplankton showed a divergent pattern with rates increasing
sharply between 62 (ambient) and 81 mg/1, similar rates between 81
and 127 mg/1, and a dramatic increase at 230 mg/1. In addition,
behavioral changes were noted with increasing sediment concentra-
tion. At the highest level of suspended sediment, this species
showed errant swimming behavior: continuous rapid jumping and
twirling was obvious at the highest turbidity levels.
DISCUSSION
The stimulation of microzooplankton clearance rates at moderate
levels of suspended sediment and the depression of these rates at
higher levels has also been observed for other zooplankters including
the copepod Acartia tonsa (Sellner et al. 1987 a,b). Suspended
sediment levels observed in Chesapeake Bay and its tributaries,
therefore, might prove beneficial to particle (and energy) intake in
small protozoa, perhaps favoring enhanced growth and reproduction.
However, previously reported data in studies of sediments and
planktonic crustaceans do not support one consistent pattern with all
taxa. In some studies, even moderate levels of suspended sediment
resulted in decreased reproduction and growth of mesozooplankton
(Arruda et al. 1983; McCabe and O'Brien 1983); however, this pattern
is not always observed as Sellner et al. (1987a,b) saw no decrease in
growth or reproduction until very high suspended sediment
concentrations were employed.
The clearance rates observed in the present study are similar to
rates previously reported. For rotifers feeding on phytoplankton,
clearance rate can vary over two orders of magnitude (Starkweather
and Gilbert 1977) depending on food concentration, high levels of
food resulting in a low clearance rate. The rates we obtained for
phytoplankton (0.05-0.13 ^1/h/ind) are characteristic of high food
concentrations and low temperatures (Bogdan and Gilbert 1982;
Gilbert and Bogdan 1984). During the grazing experiments with
Synchaeta. chlorophyll-a concentrations were relatively high (24.4
Mg/1) and temperature low (4.4 °C), consistent with the relatively
low grazing rates observed.
When feeding on phytoplankton or bacteria clearance rates for the
tintinnine ciliate E. angustatus (0.84-1.9 and 0.09-0.34 pl/h/ind,
respectively) compared well with published values (phytoplankton -
Spittler 1973, 0.5-8.5; Heinbokel 1978a,b, 0.5- 9.0; Capriulo and
Carpenter 1980, 1-85; Capriulo 1982, 2-65; Lessard and Swift 1985,
0-32; bacteria - Hollibaugh et al. 1980, 0.042; Lessard and Swift
1985, 0-11 pl/h/ind. Heinbokel (1978a) found that high food
densities resulted in lowered clearance rates (e.g. from 5 to
1 /il/h/ind when offered low and high food densities, respectively).
Chlorophyll-a concentrations during the E. angustatus study were high
(25 Mg/1) and thus clearance rates on the lower side of these ranges
would be expected.
85
-------
T)
c
H
W
U
2
a
u
0.6 -
0.5 -
0.4 -
0.3 -
0.2
80 100 120 140 160 180 200
SEDIMENT CONCENTRATION (MG/L)
Duncan's Lowest to Highest (0.05 Level)
14-C Grazing Rate: _62 81 100 127 230
220
60 80 100 120 140 160 180
SEDIMENT CONCENTRATION (MG/L)
Duncan's Lowest to Highest (0.05 Level)
3H Grazing Rate: .62 230 127 .81 100
200 220
Figure 3. Relationship between suspended sediment concentration and
clearance rate on phytoplankton (above) and bacteria
(below) by the oligotrichine ciliate Strombidium sp. In
the Duncan's test, a break in the underline denotes a
significant difference.
86
-------
Strombidium sp. exhibited clearance rates on phytoplankton (1.6- 2.2
/^1/h/ind) that are within the range of reported values for other
oligotrichine ciliates (Rassoulzadegan 1982, 2.3-8.9; Scott 1985,
0.083-1.5; Lessard and Swift 1985, 0-213 /il/h/ind). The rates on
bacteria (0.24-0.63 ^1/h/ind) are generally lower than those
reported (Lessard and Swift 1985, 0-7.6; Rivier et al. 1985, 0.6-73
pl/h/ind) but are within the values reported for tintinnine ciliates
(see above).
The decreased clearance rates at high sediment concentrations for
Synchaeta and E. angustatus could be the result of impaired feeding
or due to a behavioral response (i.e. decreased feeding) in the
presence of elevated levels of suspended sediment. The later
possibility may be an adaptation to survive short periods of high
turbidity. The reason for the opposite response found at high
sediment concentrations for clearance rates on phytoplankton and
bacteria by Strombidium is unknown. One possible explanation might
be that the erratic behavior of this species observed at the highest
sediment concentration might have been associated with rejection of
sediment particles and capture of phytoplankton cells. Though the
movements appeared energetically expensive, this behavior may be an
alternative mechanism for surviving in high suspended sediment
concentrations for short periods of time.
One of the assumptions made in estimating clearance rates from these
types of experiments is that significant elimination of label
(excretion, defecation, respiration) does not occur during the time
course of the experiment. This assumption may be violated when using
the 4 h incubation time. Little information exists to determine how
long all of the label might remain within the various taxa of
microzooplankton. However, the results of the incorporation study
with Svnchaeta suggest that the label has already obtained a constant
rate of elimination by two hours. As the duration of all experiments
was greater than or equal to 4 hours, it is suggested that measured
clearance rates may be better interpreted as assimilation or
incorporation rates. However, the incubation period does not alter
the effects of suspended sediments on relative clearing rates on
bacteria and phytoplankton.
Future studies are anticipated to determine the effects of suspended
sediments on growth and reproduction of microzooplankton. The
question remains as to whether or not stimulation in clearance rate
at moderate suspended sediment concentrations will result in
increased growth (representing increased ingestion of nutritious
particles along with sediments), decreased growth (the harmful
effects of ingesting large quantities of inorganic particles
outweighing any beneficial effect of ingesting more prey), or
equivalent growth (clearance rates are increased to counterbalance
the decreased ingestion of nutritious particles due to interference
in feeding caused by the suspended sediments). If deleterious long-
term effects are observed, then the increases in sediment loads to
the Chesapeake Bay and its tributaries could interfere with the
transfer of carbon to higher trophic levels directly through
reducing assimilation and therefore reproduction in zooplankton or
87
-------
indirectly by altering the species composition of the plankton. If
less phytoplankton are grazed, more would remain to settle to bottom
waters for subsequent decomposition and oxygen demand in bottom
waters.
ACKNOWLEDGMENTS
We thank Stella Brownlee for excellent technical assistance in the
laboratory, and Evelyn Lessard and Peter Verity for valuable
discussions concerning the results. Supported by Maryland Department
of the Environment and the Maryland Governor's Council.
88
-------
LITERATURE CITED
Albright, L.J., E.B. Sherr, B.F. Sherr and R.D. Fallen. 1987.
Grazing of ciliated protozoa on free and particle-attached
bacteria. Mar. Ecol. Prog. Ser. 38: 125-129.
Arruda, J.A., G.R. Marzolf and R.T. Faulk. 1983. The role of
suspended sediments in the nutrition of zooplankton in turbid
reservoirs. Ecology 64: 1225-1235.
Bennett, J.P. 1983. Nutrient and sediment budgets for the tidal
Potomac River and estuary. In: Dissolved loads of rivers and
surface waters: Quantity/quality relationships. Proc. Hamburg
Sympos., August, 1983. IAMS Publ. No. 141, 217-227.
Bogdan, K.G. and J.J. Gilbert. 1982. Seasonal patterns of feeding by
natural populations of Keratella. Polvarthra. and Bosmina:
Clearance rates, selectivities, and contributions to community
grazing. Limnol. Oceanogr. 27: 918-934.
Borsheim, K.Y. 1984. Clearance rates of bacteria-sized particles by
freshwater ciliates, measured with monodispersed fluorescent
latex beads. Oecologia 63: 286-288.
Capriulo, G.M. 1982. Feeding of field collected tintinnid
micro-zooplankton on natural food. Mar. Biol. 71: 73-86.
Capriulo, G.M. and E.J. Carpenter. 1980. Grazing by 35-202 urn
micro-zooplankton in Long Island Sound. Mar. Biol. 56: 319-326.
EPA (U.S. Environmental Protection Agency). 1982. Chesapeake Bay
technical studies: A synthesis. Washington, DC. 634 pp.
Fenchel, T. 1980. Suspension feeding in ciliated protozoa: Feeding
rates and their ecological significance. Microb. Ecol. 6: 13-25.
Gilbert, J.J. and K.G. Bogdan. 1984. Rotifer grazing: In situ
studies on selectivity and rates. Pages 97-133 in: D.G. Meyers
and J.R. Strickler (eds.), Trophic interactions within aquatic
ecosystems. AAAS, Washington, D.C.
Goulder, R. 1977. Attached and free bacteria in an estuary with
abundant suspended solids. J. Appl. Bacteriol. 43: 399-405.
Heinbokel, J.F. 1978a. Studies on the functional role of tintinnids
in the Southern California Bight. I. Grazing and growth rate in
laboratory cultures. Mar. Biol. 47: 177-189.
Heinbokel, J.F. 1978b. Studies on the functional role of tintinnids
in the Southern California Bight. II. Grazing rate of the field
populations. Mar. Biol. 47: 191-197.
Heinle, D.R., C.F. D'Elia, J.L. Taft, J.S. Wilson, M. Cole-Jones,
R.B. Caplins and L.E. Cronin. 1980. Historical review of water
quality and climatic data from Chesapeake Bay with emphasis on
89
-------
effects of enrichment. U.S. EPA Chesapeake Bay Prog. Final
Rept., CRC Public. No. 84, Annapolis, MD. 128 pp.
Hollibaugh, J.T., J. Fuhrman and F. Azam. 1980. Radioactivel)'
labeling in natural assemblages of bacterioplankton for use in
trophic studies. Limnol. Oceanogr. 25: 172-181.
Lessard, E.J. and E. Swift. 1985. Species-specific grazing rates of
heterotrophic dinoflagellates in oceanic waters, measured with a
dual-label radioisotope technique. Mar. Biol. 87: 289-296.
McCabe, G.J. and W.J. O'Brien. 1983. The effects of suspended silt on
feeding and reproduction of Daphnia pulex. Amer. Midland
Natural. 110: 324-337.
OEP (MD Office of Environmental Programs). 1984. Patuxent estuary
water quality survey. 1983 data summary. MD Dept. Health and
Mental Hygiene, Technical Rept. No. 3, Baltimore, MD. 57 pp.
Pedros-Alio, C. and T.D. Brock. 1983. The importance of attachment to
particles for planktonic bacteria. Arch. Hydrobiol. 98: 354-379.
Rassoulzadegan, F. 1982. Dependence of grazing rate, gross growth
efficiency and food size range on temperature in a pelagic
oligotrichous ciliate Lohmanniella spiralis Leeg., fed on
naturally occurring particulate matter. Inst. Oceanogr. 58:
177- 184.
Rassoulzadegan, F. and M. Etienne. 1981. Grazing rates of the
tintinnid Stenosemella ventricosa (Clap, and Lachm.) Jorg. on
the spectrum of the naturally occurring particulate matter from
a Mediterranean neritic sea. Limnol. Oceanogr. 26: 258-270.
Rifkin, J.L. and R. Ballentine. 1976. Magnetic filtering of the
ciliated protozoan Tetrahymena pyriformis. Trans. Amer. Micros.
Soc. 95: 187-197.
Rivier, A., D.C. Brownlee, R.W. Sheldon and F. Rassoulzadegan. 1985.
Growth of microzooplankton: A comparative study of bactivorous
zooflagellates and ciliates. Mar. Microb. Food Webs 1: 51-60.
Roberts, W.P. and I.W. Pierce. 1976. Deposition in upper Patuxent
estuary, Maryland 1968-1969. Est. Coastal Mar. Sci. 4: 267-280.
Schubel, J.R. 1968. Turbidity maximum of the northern Chesapeake
Bay. Science 161: 1013-1015.
Schubel, J.R. 1969. Size distributions of the suspended particles of
the Chesapeake Bay turbidity maximum. Neth. J. Sea Res. 4:
283-309.
Schubel, J.R. 1981. The Living Chesapeake. Johns Hopkins, Baltimore,
MD. 113 pp.
Scott, J.M. 1985. The feeding rates and efficiencies of a marine
90
-------
ciliate, Strombidium sp., grown under chemostat steady-state
conditions. J. Exp. Mar. Biol. Ecol. 90: 81-95.
Sellner, K.G., M.H. Bundy and J.W. Deming. 1987a. Influences of
suspended sediments on feeding and production of the copepod
Acartia tonsa. EOS 68: 1732, Abst. (31G-10).
Sellner, K.G., J.W. Deming and M.H. Bundy. 1987b. The effects of
suspended sediment on copepods and carbon transfer in Chesapeake
Bay. Abstract and presentation at the 50th Annual Meeting of
the American Society of Limnology and Oceanography, June 14-18,
1987.
Spittler, P. 1973. Feeding experiments with tintinnids. Oikos
15(Suppl.): 128-132.
Starkweather, P.L. and J.J. Gilbert. 1977. Feeding in the rotifer
Brachionus calyciflorus. 2. Effect of food density on feeding
rates using Euelena gracilis and Rhodotorula glutinis. Oecologia
28: 133-139.
Stoecker; D., R.R.L. Guillard and R.M. Kavee. 1981. Selective
predation by Favella ehrenbergii (Tintinnina) on and among
d-inoflagellates. Biol. Bull. 160: 136-145.
Verity, P.G. 1987. Factors driving changes in pelagic trophic
structure of estuaries, with implications for the Chesapeake
Bay. pp 35-56. In: M.P. Lynch and E.G. Krome (eds.),
Perspectives on the Chesapeake Bay: Advances in estuarine
sciences. CRC Publication No. 127. Chesapeake Research
Consortium. Gloucester Point, Virginia.
90a
-------
Understanding the Estuary: Advances in Chesapeake Abstract only
Bay Research. Proceedings of a Conference. 29-31
March 1988. Baltimore, Maryland. Chesapeake Research
Consortium Publication 129. CBPlTRS 24188.
Temporal and Spatial Variations in Zooplankton
of the Mesohaline Portion of Chesapeake Bay
Michael R. Roman, Jacques R. White,
and Anne L. Gauzens
University of Maryland
Horn Point Environmental Laboratories
P. 0. Box 775
Cambridge, Maryland 21613
As part of the NOAA-Sea Grant Program on Anoxia in Chesapeake Bay, we
conducted measurements of the short-term (tidal, diel, < 10-day) variability
in the distribution of zooplankton (> 64 um) in the mesohaline portion of
Chesapeake Bay during May and August 1986 and 1987. Hydrographic and
biological parameters were measured along with the vertical distribution of
zooplankton at both 30 h anchor stations in the mid-Bay channel and along a 5-
station transect across the Bay. During both years, there was a 2-5 fold
increase in zooplankton biomass between May and August. There was an order of
magnitude increase in the abundance of nauplii in the surface waters between
May (x = 10/L) and August (x = 100/L). Considerable short-term variability
occurred during both months. Zooplankton samples collected in mid-Bay every
1.5 h for 30 h varied by a factor of 5 in May and 10 in August. In general,
we found that significant diel vertical migration occurred in August but not
in May. Transect stations across the Bay showed that in general: there were
more zooplankton present in the middle of the Bay than on the flanks, and that
there were greater concentrations of zooplankton on the western side of the
Bay as compared to the eastern side of the Bay. Short-term variability in the
cross-Bay distribution of zooplankton can occur as a result of pycnocline
"tilting" in response to wind events.
97
-------
Understanding the Estuary: Advances in Chesapeake
Bay Research. Proceedings of a Conference. 29-31
March 1988. Baltimore, Maryland. Chesapeake Research
Consortium Publication 129. CBPITRS 24/88.
Simulating the Vertical Motion of Nekton in the Estuarine
Environment - Scientific Considerations and Speculations
Charles Bostater and Robert Biggs
University of Delaware
College of Marine Studies
Newark, Delaware 19716
ABSTRACT
The vertical and horizontal distributions, and resulting
patterns of living particles ('plankton and nekton1) are
modulated by biological and physical characteristics within
the estuary. The transition reach in estuaries is the
location where spawning of anadrornous species occurs. This
compartment within the estuary is also the location where
wind, river flow, the upper estuarine salinity regime, and
tidally induced vertical mixing and advection interact with
radiative transfer processes, in the optical and thermal
regions of electromagnetic energy spectrum, to provide an
environment for nekton growth and development. These
environmental processes are conceptualized as the stimuli
affecting the statistical characteristcs of nekton swimming
(speed and direction) during their early life stages.
Particular processes are also related to providing the
environment for sustaining food sources for the nekton. We
discuss the formulation and demonstration of an approach to
describe the vertical motion of nekton in estuaries with
reference to the Upper Chesapeake Bay. Specifically, a
simulation model with deterministic and stochastic
components is applied with reference to Mo rone sax at i1i s
eggs and larvae. Typical model output and an approach for
sensitivity analysis are presented which provide insight
into selected environmental processes and swimming
characterisitcs which may be related to larval retention
92
-------
and recruitment. Preliminary results suggest that
statements concerning "active versus passive transport" of
living particles can be refined to "horizontally passive and
vertically active" in particular marine environments. This
conceptual refinement can be assessed within the marine
environment by examining a dimen si on less number we call Pm
(a particle number), which is obtained from scaling and
dimensional considerations using a form of the backwards
diffusion equation and conservation of probability. The
research methods used in this work may provide insight into
processes that should be monitored and considered in the
development of living resources management strategies which
attempt to protect the spawning habitat for anadromous
species within subestuaries of Chesapeake Bay.
INTRODUCTION
The management of estuarine living resources requires
knowledge of their relative abundance and distributions.
The population dynamics of other species which support and
interact with the target estuarine species must also be
understood in at least a conceptual manner to discern their
role as predator or prey.
The habitat and environmental quality that directly impact
the early life stage success and recruitment of the living
resources also needs to be understood. An important
indicator of habitat quality in estuaries is the underwater
light field (ULF). The changes in the underwater light
field and factors affecting it is sometimes termed
"subsurface light climatology". The characteristics of the
subsurface irradiant light field has been studied bv various
researchers CBurt 1353, Champ 1980, Clarke and Backus 1956
), but there remains considerable uncertainty as to its role
as an "environmental stimulus" which nekton (small
organisms exhibiting motility or swimming) respond to, and
use in their behavioral strategy to live, especially in
turbid estuarine water. How nekton may use specific
charateristics of the temporal and spatial variations in the
optical portion of the electromagnetic spectrum may have
important implications. These may include aspects
concerning vertical motion, migration and patchiness. These
factors may ultimately affect their horizontal
transiocation and/or retention within estuaries or
particular spawning or nursery habitats.
TECHNIQUE DEVELOPMENT
Scientific Considerations
In view of the above, the scientific approach taken is, by
necessity, interdisciplinary and involves the application of
numerical techniques. The mathematical conceptualizations
93
-------
are a result of bio-ecological knowledge, estuarine
environmental processes, applied optics and physical
oceanographic principles.
The techniques employed depend upon laboratory studies which
describe the swimming characteristics of nekton or "living
particles" in terms of speed and direction statistics. The
techniques involve implementation of expertiments in a
laboratory setting as well as in the field, concerning the
settling characterisitcs (Sunby 1983) of the egg stage of
the living particles or nekton. Qualitative information
collected by researchers and resource management agencies
assist in discerning vertical and horizontal distributions
of this class of particles.
At the same time one must recognize the lack of precision
and accuracy of current field collections of distribution
and abundance data during the early life stages. Our current
technology prevent accurate descriptions because montioring
programs are not typically designed to assess the relative
importance of space and time scales which affect early life
stage success or "natural mortality". To do so would be too
costly, therefore resort to research and numerical modeling
techniques is necessary.
Figure 1 indicates the spatial and temporal scales of
selected physical and biological processes in the estuarine
environment. Qualitatively, this shows the spatial and
temporal frequency upon which scientific and management data
should be considered for developing research and monitoring
strategies. Currently, monitoring during "selected windows"
is conducted. For example, no data except seechi disc is
collected on a routine basis concerning the ULF in
Chesapeake Bay by management agencies. This should be given
serious consideration by upper scientific and government
management officials since light energy is a fundamental
physical quantity used by the phytopiankton community which
respire, decay and contribute to low dissolved oxygen. In
addition, associated light attenuation is one, if not a key
parameter used in driving water quality simulations for
management purposes. However, it does exhibit wavelength
dependence.
Research has demonstrated the importance of phytoplankton ,
nekton density or primary particle concentrations upon the
attenuation of light in the marine envr i onrnen t . These
concentrations are amplified or can decrease with migration
from prey (zoopiankton), swimming, and specific growth rates
(Huntley, Mann and Escritor 1987).
94
-------
SPATIAL SCALE
03
0)
W
03
o>
u
o
OJ
c
(0
•O
C
(1)
C
•H
M
03
3
03
01
U
Cfl
o
D.
0)
TD
CO
CO
a
03
U
0)
r-l
0)
CO
3
00
•H
95
-------
Previous Similar Research
Previous researchers such as Kamykowski 1976 have examined
the relationships between small scale circulation processes
such as internal waves and the vertical motion of organisms.
Recent research has resulted in estimates (Kamykowski 1388)
concerning swimming characterisitcs of dinoflagellate
swimming for possible coupling with particle transport
processes.
Other efforts have included the development of simplified
vertical migration mechanisms such as attempts to relate
larval distributions to circulation in the North Sea.
Bartsch 1988 reported on coupling a three dimensional
circulation model of the North Sea with particle tracing
tecnqniques. In his effort, the vertical migration response
(an active transport mechanism') was shown to strongly affect
the predicted horizontal trans-location of herring larvae.
Nithout vertical migration, the final passive particle
locations did not effectively match the field collection
data.
Zooplankton research results presented at this conference
suggested vertical migration and possible interaction of
active vertical transport with internal waves, crossbay
seiches and potential upwelling within the mainstem segments
of Chesapeake Bay. Research at Virginia Institute of Marine
Science has addressed the vertical motion of oyster larvae
and their interaction with frontal dynamics (A. Kuo 1988,
pers. cornm. ) . Previous efforts concerning striped bass
larvae in the Potomac estuary have indicated they maintain
their early life stages in the complex estuarine transiton
reach ( Setzler-Harni 1 ton et . al . 1381).
Current Speculations
One hypothesis that may help to explain the maintenance of
larval densities in or just above the estuarine transition
reach (salinity 0-10 ppt) is vertical migration. One can
invoke the simplified basin equations (Knauss, 1378) and
estimate a dimensionless scale which represents the
difference between the time it would take for a hypothetical
particle to move from the middle of the upper layer, to the
middle of the lower layer and vice versa, under specified
salinity and freshwater inflow conditions. This scale and
its variability can be assesed for different salinity and
freshwater inflow regimes given sufficient field data and
by assuming a migration period. This dimension less scale is
given by T'/T where T' = P-T, and it can be shown that:
T=P/[(Q/A)/(Q'/A')+l] . (1)
96
-------
The primed terms denote mean lower layer values of cross-
sectional averaqe fluxes ( Q, 0') derived from the basin
equations which make use of river inflow and layer average
salinities. P is representative of a migration period and
A, A' are crossectional areas for the upper and lower layers
respectively. These terms can be estimated from vertical
salinity information from shipboard measurements.
Adjustments in annual cycles of swimming behavior has been
demonstrated for other fresh and brackish water finfish
(Eriksson 1984).
Using available field data for the extreme Upper Chesapeake
Bay, one notes the change in this scale during the high and
low flow years of 1984, 1985 respectively, as indicated in
Figure 3. One can speculate that in order for organisms to
maintain a particular longitudinal position in an estuarine
reach and if migration was used as the behavioral mechanism,
this behavior may have to be rather adaptable or the
organisms could be flushed into higher salinity waters that
may affect survival. Alternatively, higher freshwater
inflows following storm events or higher flows are
associated with dramatic changes in the underwater light
climatology as indicated by changes in light attenuation.
The effect upon the ULF can be dramatic and can be asessed
quite readily from field data as well as from satellite
data, where the irradiance reflectance ratio is shown to be
dependent upon light attenuation coefficients CDiToro 1978)
which vary with wavelength. Figure 2 shows the changes in
the effective radiance from three wavelengths in the visible
portion of the light spectrum from the Landsat, Thematic
Mapper rnul t i spectral scanner for the Upper Chesapeake Bay.
The color variations are related to changes in light
attenuation and the estuarine processes of scattering and
absorp t i on .
Swimming Characteristics
Vertical swimming speeds can be assessed from video image
processing techniques as depicted schematically in Figure 4.
Under laboratory conditions, swimming behavior can be
statistically assessed and used to derive speed and
direction information. This framework is used for assessing
the vertical swimming response of nekton. For example,
Figure 5 indicates the vertical swimming or "geotaxis" of
striped bass larvae shortly after thev hatch. At a very
early age striped bass larvae swim vertically, followed by
sinking and then then swimming vertically again. As these
nek tonic organisms grow, they increase their swimming speed
and directions. Their swimming speed distributions can be
assessed by using statistical distributions as shown in
Figure 6. One can then use Monte-Carlo stochastic
simulation techniques
-------
rt
PQ
CN
0)
M
6C
•H
-------
oo
(n
\j /
z
0
t
**.
O 10
^ g
_, P
? 1
m g
. . ^.
v— x o
_, °
S 1
0 ^
2 C3
^
X ,
0 3
CD |
a: §
U >
v_
5
_J
o
F—
in
a
D +
a •*•
o +
a +
4D
a +
a +
a +
•-I i
1 I 1 1 I 1 1 ¥ 1 1 1 1 ' 1
*nc>i^-oo>ooNcom'*-nc>ir-
TOIIDJ SSOIUOISUOLUID^ n i /"i i
\v*«'l«f«-" ™«j«*|fcjvj»vi«^ i*4 li^ J * <*•/ '•!•
e
CO
/-*\
s
o
T «~ >-l
M-l
J=
OC
-2 3
<3-
00
o\
,_4
- o>
00
n' -H
K *
-« i •§
i ^
V t8
UJ O
-- i w *
SlO I* "1
10 Su e CQ
«0 6 PM
>- OJ "H
5 t- AJ a)
N » ^-3
y . co 01
5 O Cu
iTl *H CO
n 4J CO
< ^ <*>
- m 2i o> ,c
UJ > CJ
O 01 M
„ 4-1 Q)
-* g S£
S: M P
3 >, O)
T3 -C
m 4J
- K) CD
4-1 1-1
CO O
M [
co
co co
-CM <1> C
i-l O
C co
O CO
. -HO)
« S ™
*~ OJ g M
*~ EC
•H -rt
D -° §
CO CO
CL
14-1 CO
o
co ^
01 O
[ 1 ,-_(
co UH
e .
•H 3
4-1 0
CO H
W "^
ro
0)
l-i
3
00
•H
fa
99
-------
CO
o
CO
•u
03
OT £
o ~ "5
V
2 o o ® . «
O jc CO 2
-------
CO
CO
to
QJ
O,
CO
c
X
n)
4-1
o
cu
-------
_ 00
8 § §
O O
etf
M
"O
r-l
o
co
T)
in
03
-O
0)
(1)
c.
co
•H
U-
o
c
-S
CO
•H
CO
cj
•H
4J
CO
•H
seouajjnooo jo
CO
X
W
00
•H
702
-------
chosen to use stochastic approaches to model the swimming
process. Nhen these numerical techniques are used in
conjunction with analytical or numerical characteristics of
an estuarine physical environment (vertical flux or vertical
velocties), one can estimate the time rate of change of
abundance or "vertical motion" of the nekton or living
par t i cles.
NUMERICAL SIMULATIONS
Figure 7 indicates a simple example of the model output. As
indicated above, we couple time variable vertical swimming
speed and direction characterisitcs estimated from
application of stochastic Monte-Carlo techniques with
estimates of vertical water motion. These latter estimates
can be derived from a variety of methods in estuaries
(Bostater 1987). Our current understanding needs to be
dramatically improved in order to make better predictions as
well as to improve scientific understanding concerning the
effects of bottom topography and channel characterictics
(including bottom stress), river inflow, wind stress and
estuarine fronts upon vetical mass flux. In any event,
existing scientific judgement suggests that vertical
velocity scales are small (depending upon the transport
process and the averaging period used).
Typical estimates (Officer 1975, Dyer 1975, Pritchard 1953)
indicate velocity scales on the order of 0.1 near fronts to
0.0001 cm/sec in open water environments. Recent attempts
to measure vertical velocites using acoustic doppler
profilers (Bostater, et. al . 1987) indicate the difficulties-
involved, however this technique provides promise for future
studies of the interaction of small scale physical and
biological processes. In fact, a very important scientific
question is: to what degree does living particle patchiness
influence recent atempts to measure small scale physical
processes that make use of acoustic doppler techniques,
which are based upon acoustic particle backscatter ?
In any event, this estuarine process model starts with an
initial distribution of particles. In this case the
particles represent striped bass eggs and larvae with
specified particle ages. The physical characteristics of
the estuarine compartment includes parameters such as
temperature, salinity, river inflow, turbulent eddy
coefficient and wind stress. The simulation steps forward
in time and the coupling of vertical swimming, settling
velocities (for the egg stage) and water velocities are
used to estimate the vertical particle motion. In this
simulation example the results of two runs are indicated.
First, the upper layer particles (near the surface) are
repostioned from their uniform horizontal starting positon
103
-------
OC
CO
UJ
h-
2
UJ
O
rr
UJ
o
a
M
.J
CO
CD
h-
^
UJ
2
* i S
.. J V I
\» j CD
•. • ;. " oj
• A
". 'U
h-
Q.
lu
a
_-*... t _.._»-.,.
1
-j
OJ
vH
CO
OJ
OJ
o
OJ
CD
•*HJ
ID
OJ
•^
O
•rt
CD
ID
T
OJ
O
^c*
-2.
\s
JU
UJ
1-
<
-7-
^—
M
a
cr
o
o
CJ
_J
cc
LU
h-
-I
CO
vj
01
4J
a>
E
in
M
CO
0)
C
,e
4—*
a
0)
ID
TD
•H
E
4J
TO
•a
01
4-1
CO
o
o
rH
>.
r^
rH
CO
•H
4-1
•H
C
•H
CO
o>
rH
CJ
•H
4J
M
CO
PL,
.
4-1
3
CL
4-1
3
O
rH
01
T3
O
C
O
•H
4-1
CO
1
3
E
•H
CO
C
O
4J
^
Ol
c
0)
rH
a
g
CO
X
w
r^.
0)
M
3
60
•rH
h
f^
CO
s
J=
4-1
CO
a
c
o
•H
4-1
CO
)-l
60
,_J
E
Ol .
f.
4-1 CO
6C
C 60
O OJ
a,
3 14-1
o
60
C 60
•H C
•O -H
C rH
01 4-1
C- 4J
01 01
T3 CO
4-1 CO
C CO
01
E rH
4-1 rH
(-1 0)
co S
a
e co
O CO
o
•a
Ol 0)
C 4J
•H CO
t-i O
CO -H
3 -a
4-1 C
CO -H
01
o
oi co
J3 rH
4J CO
C CO
•H i-(
*c
4J C
•H O
* Tl
CO CO
TJ VJ
(-1 60
CO -H
s e
S rH
O CO
•O rH
0)
VJ 4J
O CO
co
T3
)-l •
co -a
> 01
Du -H
3 4H
•H
0) 0
> 0)
o ex
E co
104
-------
near 5 meters. The active particles numerically move
(swim) upwards in the water column due to a stimulus such
as light. Swimming can be initiated by defining the
"quantum swimming yield" (Sq), a dimensionless number we
define as a "threshold" for the swimming response. This
ratio is the light level at a given depth to the light level
that stimulates a swimming response. The ratio can be
estimated from knowledge of downwelling light above the
water surface and which is attenuated below the surface due
to light attenuation processes which are a function of
suspended sediment concentrations, inorganic and organic
substances such as chlorophyll.
Figure 7 also indicates the change in the centroid of
particle location. Particles are relocated very close to the
bottom and to the right side of the estuarine compartment or
segment. This response is again produced by defining a
migratory pathway. We have also developed an algorithm
which simulates swarming, simple schooling or "particle
aggregation". This figure also demonstrates the lateral
movement across the estuary which is produced by the
horizontal swimming characteristics which can be
conceptualized as larval or juvenile _migration towards
shallow water environments. The group of 8 particles in the
middle of the figure demonstrate particles which only
settle. The model can simulate more than one type of
species or particle type.
Sensitivity Analysis
We are currently exploring optional modes of running the
model processes to simulate particle motion in estuarine
compartments or segments. We calculate a sensitivity
analysis parameter from:
(0'- Os)/Os
S(0,P)= , (2)
(.?•'- Ps)/Ps
where S, the sensitivity scale, is a measure of the change
in vertical distribution of particles. 0' is the same
measure for a particular model run and Os is the standard
output (from a comparison run') of the same measure. P' is a
parameter value for a specified model run and Ps is the
value of P for the standard or comparison run.
This parameter is helping us to explore various scaling and
dimensional analysis functions derived from basic partial
differential equations which can be used to estimate the
relative importance of advective and diffusive processes in
estuaries for passive and active particle dynamics. The
current research has helped to derive a set of particle
705
-------
transport scales which indicate the conditions that may
control particle motion. These conditions include
parameters which describe advection and/or turbulent eddy
motion, amplitude of the tidal current, tidal (M2) height,
topographic characteristics such as water depth, settling
and active transport stimulated by environmental processes.
Our results indicate a biophysical regime where nekton may
be considered as being horizontally passive and vertically
active when one scales a mathematical description of the
conservation of probability. This regime is described by
Pm, a dimensionless number.
SUMMARY AND FUTURE RESEARCH
The role nekton (swimming organisms) play in maintaining and
sustaining economically important finfish and shellfish
species cannot be over emphasized. This class of biota
forms an ecological organization above the plankton.
Established procedures have previously been developed by
scientists for modeling various aspects of plankton in order
to support the development of management strategies. This
newly conceptualized area of process modeling and research
may help to provide a missing link in the refinement of
ecological models for practical use. This type of modeling
can help to assess questions concerning: (a) early life
stage distributions, (b) larval retention mechanisms, (c)
the effect of living and non-living particles upon the
irradiance reflectance ratio, (d) transport of particles in
estuaries, and (e) statistical sampling protocols for
sampling the vertical distributions of living particles in
the water column. He have developed a method to couple
circulation models with "active particles" and have shown
how laboratory testing procedures (taxis studies) can begin
to be implemented in order to investigate estuarine
porcesses. We believe practical implications from this
research will be utilized as 3 dimensional circulation
models of Chesapeake Bay are applied in order to enhance our
understanding of particle dynamics in estuaries.
706
-------
References
Bartsch, J., 1987, Numerical Simulation of the Advection of
Vertically Migrating Herring Larvae in the North Sea,
Reports On Marine Research (Berichte der Deutschen
Nissenschaftlichen Kommission fur Meeresforschung),
December, Paper Submitted to AAAS Annual Meeting, February
12, 1988, Boston., 15 pp.
Burt, W. V., 1953, Extinction of Light by Filter Passing
Matter in Chesapeake Bay Waters, Science, 118:386-387.
Bostater, C. R., Smullen, J., Biggs, R., 1987, Vertical
Velocities Measured by an Acoustic Doppler Current Profiler
and Comparisons to Analytical Models of Estuarine
Circulation, EOS, 68C44):1305.
Clarke, G., Backus, R., 1956, Measurements of Light
Penetration in Relation to Vertical Migration and Records of
Luminescence of Deep-Sea Animals, Deep-Sea Research, 4:1-14.
Champ, M. A., et.al., 1980, Characterization of Light
Extinction and Attenuation in Chesapeake Bay, August, 1977,
V. S. Kennedy, ed., In: Estuarine Perspectives, Academic
Press, pp. 263-277.
DiToro, D. M., 1978, Optics of Turbid Estuarine Waters:
Approximations and Applications, Water Research, 12:1059-
1068.
Dyer, K. R., 1973, Estuaries: A Physical Introduction,
Wiley-Interscience, pp. 140.
Erikson, T. 1984, Adjustments in Annual Cycles of Swimming
Behaviour in Juvenile Baltic Salmon in Fresh and Brackish
Water, Trans. Arner . Fisheries Society, 113:467-471.
Kamykowski, D., 1976, Possible Interactions Between Plankton
and Semidiurnal Internal Tides. II. Deep Thermoclines and
Trophic Effects, Journal of Marine Research, 34:499-509.
Kamykowski, D., McCollum, S. A., Kir pat rick, G. J., 1988,
Observations and a Model Concerning the Translational
Velocity of a Photosynthetic Marine Dinoflagellate Under
Variable Environmental Conditions, Lirnnol. Oceanoqr . ,
33:766-78.
Huntley, M. E., Mar in, V., Escritor, F., 1987, Zooplankton
Grazers as Transformers of Ocean Optics: A Dynmaic Model,
Journal of Marine Research, 45:911-945.
Knauss, J. A., 1978, Introduction to Physical Oceanography,
Prentice Hall, p. 99.
707
-------
Nelson, N., 1982, Applied Life Data Analysis, John Hley
Sons, New York, 634 pp.
Officer, C. B., 1976, Physical Oceanography of Estuaries,
Niley-Interscience, New York, 465 pp.
Pritchard, D. N., Kent. R. E., 1953, Tech. Rept. UI ,
Chesapeake Bay Institute, Johns Hopkins University.
Ripley, B. D., 1937, Stochastic Simulation, John Niley &
Sons, New York, 237 pp.
Setzler-Harni 1 ton , E., Boy n ton, N. R., Mihursky, J. A.,
Polgar, T. T., Mood, K. V. , 1981, Spatial and Temporal
Distribution of Striped Bass Eggs, Larvae, and Juveniles in
the Potomac Estuary, Trans. Amer. Fisheries Society,
110:121-136.
Sun by , S., A One-Dirnensi onal Model for the Vertical
Distribution of Pelagic Fish Eggs In the Mixed Layer, Deep-
Sea Research, 30:645-661.
708
-------
Understanding the Estuary: Advances In Chesapeake Abstract only
Bay Research. Proceedings of a Conference. 29-31
March 1988. Baltimore, Maryland. Chesapeake Research
Consortium Publication 129. CBP/TRS 24/88.
Effects of Natural Environmental Fluctuations on
Defense-related Oyster Hemocyte Activities
William S. Fisher and Marnita M. Chintala
University of Maryland
Horn Point Environmental Laboratories
P. O. Box 775
Cambridge, Maryland 21613
Epizootic diseases caused by Haplosporidium nelsoni and Perkinsus
marinus have inflicted serious mortalities on Chesapeake Bay oysters
(Crassostrea virginica). Although present in the Virginia portion of
the Bay for thirty years, recent drought conditions have allowed the
agents to invade normally low salinity regions of Maryland.
Blood cells, or hemocytes, are the primary line of internal defense for
oysters. To phagocytose and encapsulate parasites and disease agents,
hemocytes must be able to recognize foreignness, spread to an ameboid
shape, and locomote to the intruder. Hemocytes of oysters from
estuarine and oceanic habitats differed in their defense-related activi-
ties after acute, short-term, and annual changes in salinity and
temperature.
Monitoring and laboratory studies have detailed two major responses of
hemocytes to acute salinity change regardless of ambient salinity; 1)
increased salinity retards activities, and 2) decreased salinity has no
effect until 12 ppt or less. Acclimation to a new salinity can require
several hours or weeks, depending on the activity measured. Hemocytes
from estuarine oysters in a low (1984) and a high (1985) salinity year
were retarded with acute increases in salinity. Yet, acute decreases
to 6 ppt retarded activity much more in 1985 than 1984, probably due to
the higher ambient salinity. Oyster hemocytes from the oceanic habitat,
with a consistent ambient salinity, responded identically in both
years.
709
-------
All heraocytes were more active after an acute increase in temperature,
except that estuarine oyster heraocytes showed a summertime high-
temperature stress (retardation) which was amplified when the oysters
spawned in June. Oceanic oyster heraocyte activities were not retarded
until they spawned in July, even though temperature regimes at both
sites were nearly identical. This indicated that spawning and/or high
temperature and low salinity conditions reduced defense-related activi-
ties of oyster hemocytes.
770
-------
Understanding the Estuary: Advances in Chesapeake
Bay Research. Proceedings of a Conference. 29-31
March 1988. Baltimore, Maryland. Copyright by
Chesapeake Research Consortium 1988.
Evidence for Loss of Suitable Benthic Habitats for Oysters
in Tributaries of the Chesapeake Bay
H. H. Seliger and J. A. Boggs
The Johns Hopkins University
McCollum Pratt Institute and Department of Biology
Baltimore, Maryland 21218
INTRODUCTION
During the 1980 and 1981 spawning and spat setting seasons of the oyster
Crassostrea virginica , a combined study of the physical hydrography, oyster larvae
distributions and phytoplankton distributions in the Choptank River system was made in
order to understand the general mechanism for the transport and retention of planktonic
oyster larvae in the tributary estuaries of the Chesapeake Bay (Seliger et.al., 1982). We
were fortunate in that these years corresponded with atypically high concentrations of
larvae and specific spat sets (spat per bushel of dredged oyster shell) in the seed bed areas
of the Choptank River and its tributaries, Broad Creek and the Tred Avon River. The con-
comitant measurements of biological, physical hydrographic and meteorological
parameters, combined with the focus upon quantification of the three dimensional
distributions of the larvae of a single species, C. virginica , in a specific river system were
essential to developing a description of the complex interplay of factors involved in the
success of this species in the estuary.
A relation between water circulation patterns and the delivery and retention of
meroplanktonic oyster larvae has been proposed and investigated by a number of authors
(J.Nelson 1911; Pritchard 1951; 1952; Manning and Whaley 1954). The fact that high
concentrations of larvae and high spat set success were still possible in the river system,
although the latter had not been observed over the previous decade (see Table 1 of Seliger
et. al., 1982), suggested that oyster success in the system was much more sensitive to
physical and climatic forcing functions than in the past. Based on older charts and anecdotal
descriptions of the extents of oyster harvesting areas, there was an obvious and significant
reduction of the areal extents where commercial oyster harvesting was being carried out. It
appeared important to determine the relationship between high larval concentrations giving
rise to high specific spat sets on the one hand and the availability of suitable bottoms which
these successful planktonic stages could colonize and grow (ca 3 years) to mature adults. It
was conceivable that in addition to the effects of eutrophication, toxic chemicals and oyster
111
-------
Table 1. Summary of oyster bar areas in 1980 and in 1912
Bar No.
1
2*
3
Total
1
2
3*
4
5
6*
7*
8
9*
10*
Total
1*
2*
3
4
5
6
7
Total
* Site
km2 in 1980
0
0.5
0
0.5
0.81
0.1
0.91
0.06
0.02
0.36
0.11
0.02
0.02
0.5
2.91
0.38
0.08
0.02
0.03
0.11
0.1
0.08
0.8
Chester River
Bar No. km2
1
2
3
Total
Broad Creek
21
20
19
18
17
16
1 1
3
Total
Tred Avon River
22
21
20
19
18
17
16
in 1912
10
10
2.74
2.07
1.48
0.97
2.33
0.88
0.39
0.12
10.98
1.49
1.07
0.34
0.82
0.6
0.74
0.51
5.57
5% Remaining
26% Remaining
14% Remaining
of State subsidized cultch deposition
Table 1 Summary of the previously measured areas of oyster bars (1912), the present
(1980) shell areas, and the percent remaining, for the Chester R., Broad Creek and the
Tred Avon R.
772
-------
disease on the long term success of oyster harvests in the tributaries of the Chesapeake
Bay, the gradual silting over of significant areas of suitable hard bottoms for oysters had
resulted in an areal limitation of oyster production.
In the present paper we report a quantitative acoustical and SCUBA diver
examination of oyster bed areas in the Broad Creek and Tred Avon R. tributaries of the
Choptank R. in the central bay as well as in the Chester R. of the northern bay. We report
major siltation over previously described oyster beds and a correlation between strong
bathymetric gradients and the residual locations of exposed shell. Based upon this small
sampling it appears that elimination of oyster beds by sedimentation may be the major cause
of the decline of oyster populations in the Chesapeake Bay.
EXPERIMENTAL
The study areas of the Chester R. and the Choptank R. system are shown in Figure
la and Ib respectively. The echo sounder was a Simrad EY-N operating at 70 kHz,
developed for measurement of fish concentrations by quantitation of swim bladder echoes.
It was calibrated for bottom sediment composition by comparison of recorder signals with
samples both dredged and collected by SCUBA diver. The transmitter produced a 0.6 ±
2% msec pulse of 75 Watts with a cone angle of 27°. The receiver gain control had fixed
steps of 3 ± 0.3 dB and an adjustable time-varied gain (TVG) which provided a linear
increase in gain as a function of time subsequent to the initiation of the transmission pulse.
The TVG had the effect of correcting for transmission losses and therefore of making echo
trace intensities from reflecting surfaces independent of depth. The option of using
"Dynaline" recording enables precise bottom discrimination by blocking strong echoes
momentarily after the return from the sea bottom; the profile appears as a sharp line on the
echogram followed by a white gap. A ceramic transducer, protected by a streamlined PVC
blister, was suspended from the ship on a V-fin depressor which transmitted the pulse
perpendicular to the bottom. The pitch and roll effects of a hull mounted system were
thereby minimized. A portable oscilloscope (Telequipment D 32) was used to check signal
levels and to avoid saturation f the output signals. The primary and echo pulses were
recorded on magnetic tape simultaneously (2 channels) during the soundings.
The bathymetries for the Chester R. and the Broad Creek-Tred Avon R. study
areas were determined from cross channel transects at 930m and 250m separations
respectively. Each transect was initiated and terminated at the 2m depth contour. The
transects were located by Loran C, Raydist coordinates (Radio Navigation System), radar
fixes and compass bearings, and agreed with NOAA navigational charts of the areas. The
ship speed and echo sounder chart speed were held constant throughout the transects and
were checked separately against measured courses. Each sounding was made at constant
gain and recorder settings.
During the transect anchor buoys were dropped overboard without stopping at 1-
meter depth increments determined from the chart recorder and marked on the chart by an
event recorder pen. Subsequent to the complete transect, bottom grab samples were taken
at each buoy location with Van Veen and Ponar type samplers, depending upon the
hardness of the bottom. The collected sediments were mixed and stored in labelled jars for
silt/sand analysis. The silt/sand compositions of the sediments were determined with a 63
H.m mesh (Tyler No.230) sieve. Sand was arbitrarily defined as particles > 63 |im in any
2-dimensions (Wentworth 1922). A uniformly mixed portion (10-20 g) of each grab
sample was sieved until the distilled water used for washdown remained clear. The silt and
remaining sand were transferred to pre-weighed beakers, dried at 76° for 2-3 days, then re-
weighed. The horizontal scale for plotting bathymetry was calculated from the ship speed
and the recorder chart speed and converted to 1:40,000 horizontal scale for contour
plotting. Observations of locations of commercial oyster tongers were marked, as well as
locations of oyster dredge sampling.
113
-------
CHESAPEAKE (BAY
Fig. l(a) The Chester River System, with Study areas enclosed is boxes.
114
-------
Fig. l(b) The Choptank River System, with study areas enclosed is boxes.
115
-------
The areal determinations of the previously defined oyster bars and the present shell
beds were measured from the contour charts with a Graf/Pen, GP6-50, (Science
Accessories Corp.) digitizer. Bathymetry gradients were calculated from isobaths (depth,z)
by measuring the projected distances (r) normal to the isobaths, expressing the gradient as
(dz/dr) x 103.
RESULTS
Calibrations were carried out over an active oyster bar in the Chester R, (Buoy
Rock Bar) whose dimensions are shown in Figure 2a. S—>N and W—>E echogram
transects were made directly across the oyster bed (Figure 2b) and a S—>N echogram
transect just west of the oyster bed (Figure 2c). The water surface and the bottom depths
are indicated by the arrows labelled "Surface" and "Dynaline", respectively. From Figure
2b the echograms of hard sand and soft mud relative to oyster shell can be distinguished in
terms of the densities of their respective acoustic returns. Viable oyster beds can also be
identified by the distinctive light space in the secondary reflection, shown by the rectangles.
The echogram in Figure 2c shows the absence of shell bed. The gradations in densities of
acoustic returns were correlated with silt/sand or mud composition of sediments
determined by physical sampling and more precisely by SCUBA diver. In our small study
area which contained one of the two remaining commercially active oyster bars in the
Chester R., less than 0.5 km2 of oyster shell area, of an original 10 km2 remain (Yates
1912; Merritt 1977). The original extent of Buoy Rock Bar, 2.6 km2 has been reduced to
0.5 km2. From acoustic assay and corroborated by SCUBA diver, Blunt Point Bar (SE of
Buoy Rock Bar) and Long Point Bar (SW of Buoy Rock Bar) have been completely silted
over. The maximum depth for the oyster bar appears to be correlated with the depth of the
bay pycnocline during summer (Seliger et.al., 1980; 1984), because of anoxia that
develops in waters below the pycnocline during summer.
Two W—>E cross channel echogram transects, 500m apart, encompassing an
actively worked oyster bar (Great Bar) in Broad Creek are shown in Figure 3a and 3b,
illustrating from west to east: a) mixed silt/sand (thin primary and secondary traces); b)
soft mud (broad primary trace); c) mixed silt/sand; d) shell bed (white space, see rectangle);
broad tertiary trace); e) location of a commercial oyster tonger and f) hard sand (thin
secondary and tertiary traces) on the eastern shore.
The locations and extents of present exposed oyster shell bars in Broad Creek and
the Tred Avon R. are shown by the cross hatched areas in Figure 4a and 4b respectively.
The dotted lines represent the 6 foot (ca. 2m) and the 18 foot (ca 6m) depth contours.
Peripheral areas where sediment has partially encroached upon shell, i.e., Bed 1 of Figure
4a and Bed 7 of Figure 4b, are labelled "patchy". Areas of soft mud are usually at depths >
6m. In general the beds follow the 6 foot (2m) depth contour and, where they extend into
deeper water, are sharply delineated by the 18 foot (6m) depth contour.
The present oyster shell locations (cross hatched) in the Broad Creek and the Tred
Avon R., together with the areas where the bathymetric gradient (Dz/Dr) x 103 (solid
black) are shown in Figure 5a and 5b respectively. In most instances these solid black
areas are contiguous with the locations of present oyster shell areas. The only saving
feature for the presently remaining area (18%) of Buoy Rock Bar is the very steep
bathymetric gradient at its southern end (not shown in the figures). The previous areas of
active oyster beds in these three tributaries measured by Yates (1912) are shown in Figure
6a, 6b and 6c. Table 1 summarizes the previous bars, the present areas in km2 and the
percent of shell area remaining. In the Chester R. and the Tred Avon R., oyster bed areas
have been reduced to 5% and 14% respectively, of their previous measured extents.
116
-------
'.A
•39° 00'
; Stony Bar Bluff :<
39° 00'
Fig. 2 (a). Chan of the Chester R. study area showing the location of Buoy Rock Bar.
The bathymetry measured is shown in 2m increments (solid lines). The transects occupied
are shown as dashed lines.
777
-------
METERS
(U
-------
METERS
O 2
METERS
f
?
o
u
II
o o
OH
0)
0)
X)
o
o
o
CD
o
(A
c
o
0>
0)
TJ
O
O
t_
CD
o
CD
UJ
CVJ
o
a>
c
o
"Si
.tt O
o
c M
•- -a
b C
to
is IN
M^-5
C.£?'og.S
Bsl'CJ
^ o -^--
wS^"§
;r 2 ^ J3
PL, J3 l/l M
779
-------
h-
c c
I g
°
-------
i *
i en
X
N»
--J —
^^ 8
•a «3 u
*
D u
f-« ^-" *^H
<*- S3 5
o P &
e "
awe
1-SS
U c 6
"S ȣ
rt o*0
O O. en
• *> c3
GJD f^i f)
uu oo -2
121
-------
76° 14'
Narrows Pt.
Cedar Pt,
I
76° 12'
Hail Pt.
N
—39.°oo'
Chester River - Oyster - Bar Locations *- Yates, 1912
Fig. 6(a) Previously measured extent of oyster bars on the Chester River.
722
-------
Leadenham
Cr.
Broad Creek-Oyster-Bar Locations- Yates, 1912
Fig. 6(b) Previously measured extent of oyster bars on the Broad Creek.
123
-------
Tred Avon River-Oyster-Bar Locations-Yates, 1912
Fig. 6(c) Previously measured extent of oyster bars on the Tred Avon River.
724
-------
DISCUSSION
It is obvious that biotic as well as abiotic factors influence the success of sessile
oyster populations and their planktonic and meroplanktonic life cycle stages. Only a small
portion of these factors has been addressed in the present paper. From the success of the
planktonic and settling stages in 1980 and 1981 it appears that water circulation patterns
and food sources for the larval stages were still adequate. The successful commercial re-
harvesting in the early 1980's of seed oysters "planted" in previous years on both Buoy
Rock Bar and Ferry Bar through the State-subsidized re-seeding program is inferential
evidence that, barring climatic extremes, there were still sufficient food sources for the
growth of adult oysters. It appears therefore that for each of the oyster bars investigated,
sedimentation (the covering over of previously viable oyster beds) has been the major
factor in the reduction of oyster harvests. Many formerly productive oyster bottoms along
the Atlantic coast of the United States have been destroyed by high sedimentation rates
(Galtsoff 1964). The filling of embayments with sediments is a general phenomenon
along the Texas coast and is particularly pronounced in Laguna Madre and near the
Colorado R. in Matagorda Bay, where approximately 24-28 km2 of oyster reefs are under
14 feet of mud (Norris 1953). In the York R. in the Chesapeake Bay faulty erosion control
practices,deforestation and population growth were identified 50 years ago as increasing
erosion of soil from the watershed into the river (Brown et. al., 1939). In the upper
Chesapeake Bay aproximately 38000 m2 of shoreline (0.33 x 106 metric tons) are eroded
and enter the bay annually (Schubel 1968).
The data reported in this paper represent the application of a rapid profiling and
bottom composition analysis technique to identification and areal assay of oyster beds. Our
previous physical hydrographic and biological study of transport and retention of oyster
larvae (Seliger et.al. 1982) indicates that the planktonic life cycle stages of oysters have not
been affected as much as the sessile benthic stages. Silting over of upstream seed bed areas
and viable oyster beds along the shorelines may be the major factor in the loss of oyster
harvests in the Chesapeake Bay. The study has not been applied to examine conditions
outside of the mouths of the tributaries. However from the high larval concentrations and
high spat sets measured in the Choptank R. system in 1980 and 1981, and from the
absence of expected mature oyster harvests in the years following the high spat sets (1983,
1984), it appears that a limiting factor to oyster production in the tributaries of the
Chesapeake Bay has been the silting over of suitable hard bottoms.
It is apparent from Table 1 and comparisons of Figures 2a,4a and 4b with Figure 6
that there have been severe reductions in all three tributaries in exposed shell areas suitable
for the setting of spat. The pattern of sedimentation and consequent reduction of exposed
shell bed area is most obvious in the Tred Avon R. )Figure 4b), where the entire central
channel is now essentially soft mud. The present percentages of shell bed remaining are
actually higher than they would be in the absence of the Oyster Management Program of the
Maryland Department of the Environment. It has annually subsidized extensive deposition
of fossil oyster shell cultch in all three tributaries in order to assist commercial harvesters.
The association of the remaining shell areas with steep bathymetric gradients
(Figures 5a and 5b) is presumed to represent areas of highest tidal frictional turbulence and
therefore the areas least likely to be impacted by siltation. The locations of the steep
bathymetric gradients also represent areas where tidal shear fronts are observed. It follows
that the areas immediately contiguous to these steep bathymetric gradient areas (Figure 5)
would represent ideal locations for the deposition of artificial and fossil cultch, in order to
increase the areas of oyster beds with minimum siltation effects. Since the major
streamflow and sediment runoff into the tributaries occurs during the spring freshets, it
would appear more efficient to deposit cultch material for new beds during early June, just
725
-------
prior to the oyster spawning season, in order to minimize siltation on these new surfaces
for spat settling.
In a companion paper in this volume (Seliger and Boggs, "Long Term Pattern of
Anoxia in the Chesapeake Bay"), we present an hypothesis that excess sedimentation in
spring runoff results in extremes of light limitation in the upper (Maryland) portion of the
bay. This inhibits photosynthesis and thus nutrient assimilation, permitting runoff
nutrients in the surface plume to be delivered to the southern (Virginia) portion, where
blooming occurs. A similar process occurs in each of the tributaries of the bay; excess
sediment results in a transfer of production to downstream regions. Concomitant with the
effects upon primary production in the water column this same sediment loading results in
the progressive deposition of sediment upon and the loss of upstream seed bed areas for
oysters and the shallow areas outside of the mouths of the tributaries. These areas are
specifically the regions of the tributaries which are not subject to transient incursions of
anoxic bottom waters from the central bay during summer. In addition the salinities in the
upstream seed bed regions of the Maryland tributaries are low, minimizing the incursions
of high salinity-requiring oyster diseases. The loss of the upstream seed bed areas by
siltation removes from the system the resevoir of potential production, i.e., of larvae for
re-colonization subsequent to extremes of anoxic events or of disease. The loss by siltation
of the shallow oyster beds contiguous with the shorelines removes from the system the
mature oyster resevoir least impinged upon by anoxia of bay bottom waters. What remains
are oyster bed areas in the deeper waters, areas which are most sensitive to anoxic
incursions when, as in 1984, anoxia in bay bottom waters rose to a depth of 6 meters
(Seliger et. al., 1985). Under these conditions oyster populations are not replenished
sufficiently even following years when spawning and setting conditions are excellent.
Much of this sediment deposition is irreversible. Management procedures for optimizing
oyster spat settlement during good years can be inferred from the paper: a) implementation
of severe restrictions on further sedimentation; b) building up of cultch and artificial hard
bottoms in specific contiguous areas of steep bathymetric gradients; c) deposition of both
cultch and seed oysters after the spring freshets, thus avoiding major siltation; d) utilization
of acoustic techniques for following the progress of oyster bed replenishment.
726
-------
Brown, Carl B., Seaney, L.M. and Rittenhouse, G. Advance report on an investigation of
silting in the York River, Virginia, October 25-November 5, 1938. Sedimentation Studies,
Division of Research, SCS-SS-32, U.S. Department of Agriculture, Soil Conservation
Service, Washington, D.C. 12 pp (1939).
Galtsoff, Paul S. The American Oyster, Crassostrea virginica (Gmelin). Fishery Bulletin
64: 480 pp (1964).
Manning, J.H. and Whaley, H.H. Distribution of oyster larvae and spat in relation to some
environmental factors in a tidal estuary Proc. Natl. Shellfish Assoc. 45, 56-65 (1954).
Merritt, D.W. Oyster spat set on natural cultch in the Maryland portion of the Chesapeake
Bay (1939-19975). UMCES Special Report No. 7 (1977).
Nelson, J. Report of the Biologist. Ann. Rept. Dept, Biol. for 1910, N.J. Agric. Coll.
Exper. Sta. pp!83-218 (1911).
Norris, R.M. Buried oyster reefs in some Texas bays. Journal of Paleontology 27 (4):
569-576 (1953).
Pritchard, D.W. Distribution of oyster larvae in relation to hydrographic conditions, Proc.
Gulf and Caribbean Fish Inst. Nov. 1952 pp 1-10.
Pritchard, D.W. The physical hydrography of estuaries and some applications to biological
problems, Trans. 16th N. Amer. Wildlife Conf. Wildlife Management Inst., Wash. D.C.
pp 368-376 (151).
Schubel, J.R. Suspended sediment of the northern Chesapeake Bay. Chesapeake Bay
Institute Technical Report 35 Reference 68-2 (1968).
Seliger, H.H., Boggs, J.A. and Biggley, W.H. Catastrophic anoxia in the Chesapeake
Bay in 1984, Science 228, 70-73 (1985)
Seliger, H.H., Boggs, J.A., Rivkin, R.B., Biggley, W.H. and Aspden, K.R.H. The
transport of oyster larvae in an estuary, Marine Biology 71, 57-72 (1982).
Seliger, H.H., McKinley, K.R., Biggley, W.H., Rivkin, R.B. and Aspden, K.R.H.
Phytoplankton patchiness and frontal regions, Marine Biology 61, 119-131 (1981).
Wentworth, C.K. A scale of grade and class terms for clastic sediments, Journal of
Geology 30, 377-392 (1922).
Yates, C.C. Surveys of oyster bars, U.S. Coast and Geodetic Survey Publication, U.S.
Government Printing Office (1912).
727
-------
Understanding the Estuary: Advances in Chesapeake
Bay Research. Proceedings of a Conference. 29-31
March 1988. Baltimore, Maryland. Chesapeake Research
Consortium Publication 129. CBP/TRS 24/88.
Stabilized Coal Ash as Substratum for Larval Oyster Settlement:
A Pilot Field Study
Kent S. Price, Karolyn Mueller, Joel Rosenfeld,
and Thomas Warren
University of Delaware
College of Marine Studies
Lewes, Delaware 19958
INTRODUCTION
Chesapeake Bay is one of the primary oyster production areas in the
United States. In recent years oyster production has dropped
dramatically due to dis_ease, poor water quality (eutrophication,
pollution), and a lack of suitable substrata for oyster spat set. This
study funded by Baltimore Gas & Electric Company and the Delaware
Research Partnership examines the feasibility of using stabilized coal
waste materials in the form of pucks as a suitable substratum for
oyster spat set and growth.
METHODS
Stabilized coal-ash was constructed by the engineering firm of KBK,
Inc. of Atlanta, Georgia. Bottom and fly ash were supplied to KBK by
Baltimore Gas and Electric Company; the stabilized ash was shaped in
the form of 3-inch and 6-inch diameter discs approximately two inches
thick and allowed to cure. The substrata was then shipped to each of
two sites: 1. the College of Marine Studies, University of Delaware in
Lewes for placement in the Broadkill River, and 2. the eastern shore
of Maryland for placement in the Little Choptank River and Broad Creek,
prime oyster producing tributaries of the Chesapeake Bay. Stabilized
ash and oyster shell control substrata were placed at the Maryland
sites in early July, 1987; the Delaware portion was placed in holding
tanks to which hatchery reared oyster larvae were added. After oyster
settlement was ascertained, the Delaware substrata were placed in the
Broadkill River.
128
-------
The Chesapeake study was conducted at two sites in the eastern
Chesapeake Bay area, one in Broad Creek near Neavitt, Maryland, and
the other in the Little Choptank River near Madison, Maryland. Both
sites are known to be oyster spawning grounds and were recommended by
Dr. George Krantz, Maryland Department of Natural Resources in Oxford,
Maryland.
Two substrata were used at each site to study oyster spat set; 1) coal
ash/concrete 6 and 3 inch diameter pucks, and 2) oyster shell. Coal
ash/concrete pucks were made of a mixture of bottom ash, fly ash, and
cement in a ratio of 59:33:8. Oyster shell was dredged fossil oyster
shell provided by the Langenfelter Company which supplies the State of
Maryland shell for replenishing oyster setting areas.
At each site three plots were set up. Each plot contained nine squares
2 meters by 2 meters, and each plot was separated by the length of one
plot (Figure 1). Coal ash pucks and oyster shell were the treatment
variables in this replicated randomized block design, and each variable
appeared once in each block. The "natural bottom" squares were
included in the design to provide access to the "ash" and "shell"
squares and were not sampled.
The replicated randomized block design was used here to account for any
variability between blocks possibly due to currents, patchiness of
oyster larvae in the water column, bottom differences, etc. (thus our
one random blocking effect might be any of these unknown factors).
The same replicated randomized block design was used at both sites and
each square was coded for identification of sample origin. Plots of
both sites were measured and staked out before substrata planting.
Substrata were planted at the Broad Creek site on 7/2/87 and at the
Little Choptank site on 7/7/87.
Substrata were ferried to the sites in burlap sacks by the oyster boat
"Miss Molly" and transferred to two motor boats as needed for planting.
To avoid confusion one boat planted only oyster shell and the second
only pucks. Sacks of the substrata were opened and poured from the
boats evenly into designated squares and each square was raked with a
clam rake to ensure even substratum deposition.
Twelve sacks of oyster shell were deposited in each "shell" square, and
six sacks of 6" pucks and six sacks of 3" pucks were deposited in each
"puck" square. Sacks were chosen haphazardly for depositing, and the
order in which the squares were planted was also haphazard. Three
sample bags of each substratum (chosen haphazardly) were volumed to
determine the approximate amount of substratum deposited in each
square.
Four 10 cm x 10 cm asbestos plates were hung at each site as an
indicator of intensity of spat set.
Sampling of substrata began on 7/15/87 and continued on a weekly basis
until 8/19/87 after which sampling was monthly until 11/16/87 when
sampling ended for the winter. One 6" puck was sampled from each
"puck" square and at least four shells from each "shell" square.
729
-------
Figure 1; Arrangement of the Latin Square design plots
at both sites on the Chesapeake Bay.
2m
2m
6 meters
Coded Latin Square design plots for the
Broad Creek site and the Little Choptank site.
nb» natural bottom
Broad Creek Site
1.1.1
puck
1.1.4
shell
1.1.7
nb
1.1.2
shell
1.1.5
nb
1.1.8
puck
1.1.3
nb
1.1.6
puck
1.1.9
shell
1.2.1
puck
1.2.4
nb
1.2.7
shell
1.2.2
nb
1.2.5
shell
1.2.8
puck
1.2.3
shell
1.2.6
puck
1.2.9
nb
1.3.1
nb
1.3.4
puck
1.3.7
shell
1.3.2
puck
1.3.5
shell
1.3.8
nb
1.3.3
shell
1.3.6
nb
1.3.9
puck
Little Choptank Site
2.1.1
puck
2.1.4
shell
2.1.7
nb
2.1.2
shell
2.1.5
nb
2.1.8
puck
2.1.3
nb
2.1.6
puck
2.1.9
shell
2.2.1
puck
2.2.4
nb
2.2.7
shell
2.2.2
nb
2.2.5
shell
2.2.8
puck
2.2.3
shell
2.2.6
puck
2.2.9
nb
2.3.1
nb
2.3.4
puck
2.3.7
shell
2.3.2
puck
2.3.5
shell
2.3.8
nb
2.3.3
shell
2.3.6
nb
2.3.9
puck
130
-------
Substratum from each square was placed in a coded Ziploc plastic bag,
sealed, and placed in ice in an ice chest and delivered to the College
of Marine Studies, University of Delaware, Lewes, Delaware for
analysis.
Each substratum was sampled haphazardly from its square by diving until
11/16/87 when sampling was done from the boat using olam rakes.
Sampling of the squares was according to the code of the sites (square
1.1.1. sampled first, then 1.1.2., etc.). On each sampling date the
spat collectors were retrieved and stored in cool bay water and brought
to the lab for analysis and replaced by clean spat collectors until
9/15/87, the end of the spawning season.
Analysis of the samples consisted of examining both flat sides of the
puck (the rim surface was not included in the analysis) and both sides
of a shell under an American Optical stereo microscope at I4x to 84x
and noting oyster spat and measuring spat length from hinge to outer
lip to the closest millimeter. Beginning with the 9/15/87 samples
dead oyster spat were noted separately from live oyster spat. Length
and width of each shell substrate was measured to the nearest
millimeter and spat set per square centimeter was calculated. For
direct comparison spat set per square centimeter was also calculated
for "puck" surfaces and asbestos spat collectors. Density data (spat
per centimeter squared) was loaded into LOTUS worksheets and analyzed
statistically using STATGRAPHICS and MYSTAT on an IBM PC computer.
Oysters from both the Maryland eastern shore sites and the Delaware
site were harvested for metals analysis from the coal-ash pucks and
oyster shell controls, frozen at -70 C, freeze-dried for 48 hours and
ground to powder consistency. Samples were then prepared for analysis
via atomic absorption spectroscopy as follows: sample aliquots were
weighed out, pre-digested in 10 ml concentrated nitric acid at room
temperature for 24 hours, digested at 65 C for 4 hours and diluted to
50 ml with \% nitric acid. Digests were then filtered through
quantitative ashless filters and analyzed for metal concentrations
using a Varian SpectrAA-20 atomic absorption spectrophotometer (flame
mode). Blanks were included in the digestion procedure; no standard
oyster reference material was available from the National Bureau of
Standards therefore no outside tissue standard was included in these
preliminary analyses. We are attempting to find a source of standard
tissue for future analyses. Where necessary, samples were
appropriately diluted to a measurable concentration.
Data Analysis
Two-way mixed model ANOVAs were computed for the Broad Creek data and
the Little Choptank data separately with "substratum" as the fixed
factor and "block" or "plot" as the random blocking factor (Sokal and
Rohlf, 1981). Least significant differences and residuals of the
means of the factors in the ANOVA were plotted. All statistical tests
were run at an alpha level of 0.05.
131
-------
RESULTS
Statistical comparisons of setting intensity on shell versus coal ash
show that shell is slightly but not overwhelmingly selected for setting
by oyster larvae. Although the setting density is about 2:1 in favor
of shell the results are not statistically different based on samples
collected during the summer and on November 16, 1987 (Table 1). Growth
and mortality are statistically equal on shell and coal ash based on
this and other samples taken during the growing season. By comparison,
growth was statistically greater on coal ash in two of three grow out
situations in the Delaware Broadkill River experiment. Mortalities
and setting densities were statistically indistinguishable on shell
and coal ash in the Delaware experiment (Table 2).
Live oyster abundance was quite low on shell and coal-ash substrata
retrieved from the Maryland sites. Tissue for metals analysis was
harvested and pooled in an attempt to obtain at least 0.1 g final dry
weight per sample. Three shell-grown and three coal-ash-grown oysters
samples were available from the Little Choptank, and only one shell-
grown and one coal-ash-grown sample were available from the Broad
Creek site. The paucity of tissue was not apparent until shells and
pucks had been returned to the laboratory and individual oysters were
actually shucked. Many apparently "live" oysters were filled with
sediment; had this been known during the November collection period,
more substrata would have been collected.
Tissue availability from Broadkill River, Delaware oysters was
conversely, quite good. Both shell-grown and coal-ash-grown oysters
were readily abundant; tissue was harvested and pooled as above. A
total of eight shell-grown and eight coal-ash-grown samples have been
analyzed to date.
Broadkill River oysters
Analyses of the following metals have been completed: zinc, iron,
copper, and cadmium. Copper and iron concentrations were higher in
Broadkill River shell-grown oysters than in coal-ash-grown oysters
(Table 3). Non-parametric statistical analysis (Mann-Whitney U test,
= 0.05) found both metals to be present in significantly higher
concentrations in shell-grown oysters than in coal-ash-grown oysters.
Zinc, on the other hand, was present in significantly higher
concentrations in coal-ash-grown oysters than in shell-grown oysters.
Cadmium concentrations were not detectable using flame atomic
absorption methods and it is possible that the digestion procedure
employed allowed for the volatilization of cadmium from the samples;
quality control experiments are being conducted to ascertain if
volatilization does occur during the digestion procedure.
Little Choptank River and Broad Creek oysters
No statistical tests were employed to analyze these data due to the
small number of replicates. The range of concentrations of copper,
zinc, and iron in these oysters is presented in Table 3- The values
for zinc and iron concentrations seem to be quite different from the
concentrations determined for the Broadkill River oysters. These
differences are most probably due to a difference in water quality
and food supply at the Maryland sites.
132
-------
CN
co
CO
•—4
.a
o
oo
rH
OO
i-H
^
CO
i
CO
o
2Z
c
o
•rj
co
o.
fj=
CO
0
0
CO
L.
o
3:
o
— 4
CO
C3
*
L—
CO
•— H
an
^
4-H
s^
•a
o
o
03
•
oo
cn
^^
VD
i— I
.
>
O
z
o
"rj
CO
I
o
CO
o
a
•a
^i
§
CO
0
CO
a.
0
CO
0)
u
CO
co
4-<
s
c
c
o
o
CO
Q
•o
a
c
o
4-1
oo
+1
c
o
FS
CO
o
o
4-1
o
•a
4^
r"
O)
CO
X
_ -^
r^
CO
j=
00
c
c
o
o
•*— «
I
•a
c3
^5
c
O
oo
•fl
c
0
a>
o
o
4~>
O
4-1
C
0)
1— t
CO
X
1— 1
F— ^
co
oo
rH •
0
o
v'
o.
_«;
r^
s
C/)
t-
^4»
CD
O
CO
co
<4_
1—4
^3
4->
C
o
o
<4_
•V— 1
c
cn
CO
CO
CO
Q
f—
CO
•o
*
o
<*-
CO
o
c
CO
CO
1—4
T3
+J
O
0
t^_
1— t
O)
.(— *
w
CO
co
Q
r"
CO
•o
*
co
^
CO
4-1
gs
1—4
PH«
— M
•Tg
•
^•^
CO
CO
co
CO
4—1
c^
CO
o
a.
c
c
•
^^
co
0)
CO
£^
4-1
C
CO
o
a.
c
c
LA
O
V
a.
^^
— ^
at
CO
ts
*•••
ti
z
.— f
^*
ii
3
2
C-
!••
3
1-
3
U
a
rH
ID
cn
LA
+1
rH
rA
rH
^^
a*
oo
CN
*•*
tA
1 —
rH
+ 1
CN
cn
•
a-
^— *
CN
rH
^^>
^»
CN
00
-rl
0
CN
O
rH
_J
UJ
X
00
^y
OO 4->
Q * O
i- >i co
CD 4J L.
> —• CD
-+ V) 4->
on c. co
CD >.
— a o
p«H
—« O) T3
^ c o
o —< a>
O 4-> ^3
O 4-1
1_ CD 4-
03 00 O
LA IA LA
CN
CN
rH CN OO U3 CN LA
IA a- CN
CO
o
CO
Q_
C
o
o
O
CD
co
4-J II
0)
CO
55
§
o
CO
oo
o
0
o
^ co
•H oo a-
i-H tA
CN rH
O
cn
a- LA
CN rA
CN CN
CD
00
CD
CD
O
o _
O 00
O
O
CD
o
o
i
o
D)
O)
I
oo
cn
— o
0 O
c: CD
CD CO
o
Q)
> Q-
o
T3 —
CO CN
CO
o
o ~
o
4->
o
4->
'•s.
^ CO
4-> L.
O 0)
Q. 4->
oo co
o
03
CO
o
co
Q.
o
CO
co.
4-J O
^- CO
CO tj
£
cn
co
oo
4-1
C
4-J
i
4_)
cz
CO
o
L_
co
Q.
>
Vi
CO
a
rj
C
4-1
CO
00
a-
cn
LA
II
r— 4
«— 1
CO
f~
oo
CN
CN
rH
O
•
II
(— H
i
00
LD
r-s
II
— ^
o
o
LA
f*s.
O
o
•
II
o
3
a-
ID
II
1— 1
— (
QJ
.C
CO
CO
CM
8
-.
—4
i
oo
f^
oo
n
,—<
0
O
rH
rA
o
•
II
1
T3
O
O
CQ
CO
-------
Table 3. Metal concentrations in tissue of oysters (Crassostrea vlralnlca)
grown on stabilized coal-ash and oyster shell substrata, Analyses were
performed using atomic absorption spectroscopy, flame mode, Concentrations
are ug/g dry weight of tissue; Broad Creek, Maryland values are ranges,
while Broaaklll River, Delaware are means ± one standard deviation,
ELEMENT
SITE
TISSUE
CONCENTRATION, (uo/g)
Zinc
Iron
Copper
Broadklll R,
(n=8/treatment)
L, Choptank
and Broad Creek
(n=4/treatment)
Broaakill R,
(n=8/treatment)
L, cnoptank
and Broad Creek
(n=4/treatment)
Broaaklll R.
(n=8/treatment)
L, Choptank
and Broad Creek
(n = a/treatment)
shell-grown
coal-ash-grown
shell-grown
coal-ash-grown
shell-grown
coal-ash-grown
she11-grown
coal-ash-grown
she11-grown
coal-ash-grown
she11-grown
coal-ash-grown
x = 2437 + 406
K = 2901 + 232
875 to 1103
0 to 945
K = 286 + 48
X = 218 + 16
307 to 505
304 to 2058
X = 116 + 11
5? = 96 + 11
55 to 92
49 to 118
134
-------
There are several metals yet to be determined in these samples; the
volatile elements (arsenic, selenium, and mercury) will require a
separate digestion procedure than the one currently used. Quality
control experiments are being conducted to determine the most
appropriate method for these elements. Also, several elements are
present in such low concentrations that the atomic absorption furnace
mode of detection will be required.
DISCUSSION
The setting, growth, and mortality data from the Chesapeake ad Delaware
studies indicate that stabilized coal ash is an acceptable alternative
substratum for oyster settlement and growth; these results concur
with preliminary laboratory oyster setting studies (Price, 1987).
Differences do exist; however, between the three sites chosen for
this pilot study. The Broadkill River, Lewes, Delaware is,
historically, a "good" area for growing oysters; there is a plentiful
supply of food for filter feeding bivalves and tidal flushing provides
a constant source of necessary nutrients. The suspended sediment in
the River has also been implicated in stimulating/enhancing feeding
in oysters (Ewart, 1985). The differences in oyster setting, growth,
and mortality between the two Chesapeake sites most probably are due
to water and food quality at each site. As corroborative evidence,
our calculated setting intensity on the shell and coal-ash substrata
at Broad Creek was similar to the setting intensity determined for
Broad Creek by researchers at the Horn Point Laboratory of the
University of Maryland.
Preliminary metals analyses provide evidence that there is a difference
in metals concentrations between coal-ash growth and shell-grown
oysters. Coal-ash is known to adsorb ions from water and may be
depleting the water in the immediate vicinity of a filtering oyster
(Andren et al., 1980); given that such a depletion may occur, there
is no evidence at present to determine that differences in tissue
metals concentrations results in deleterious effects in juvenile
oysters.
In conclusion, stabilized coal ash appears to be an acceptable
alternative cultch material for the oyster Crassostrea virginica.
The second year of this study will provide additional setting, growth,
and mortality data, and hopefully, additional oyster tissue for metals
accumulation analyses.
135
-------
LITERATURE CITED
Andren, A., M. Anderson, N. Loux and R. Tabbot. 1980. Element
flow in aquatic systems surrounding coal fired power plants.
U.S. Document EPA-600/3-80-076, July 1980.
Ewart, J. W. 1985. Effect of suspended silt on growth of
Crassostrea virginica and ultrastructural condition of ingested
algal food in feces. Master's thesis, University of Delaware,
College of Marine Studies, Lewes, Delaware.
Price, K. S. (ed.). 1987. Project ASHREEF. A report on a
stabilized coal waste fish reef on Delaware subaqueous lands.
Electric Power Partners Program, College of Marine Studies,
University of Delaware, Lewes, Delaware.
Sokal, R. R. and F. J. Rohlf. 1981. Biometry, Second Edition.
859 pp. W. H. Freeman and Company, New York, New York.
136
-------
Understanding the Estuary: Advances in Chesapeake Abstract only
Bay Research. Proceeding sofa Conference. 29-31
March 1988. Baltimore, Maryland. Chesapeake Research
Consortium Publication 129. CBP/TRS 24/88.
The Sedimentary History of Submerged Macrophytes
in the Upper Chesapeake Bay
Grace S. Brush and William B. Hilgartner
The Johns Hopkins University
Department of Geography and Environmental Engineering
Baltimore, Maryland 21218
Seeds preserved in sediments show spatial and temporal distributions of
12 species of submerged macrophytes in 11 Upper Bay embayments and
tributaries. Three of the species are known only from the seed record. The
time spanned in the sediment cores includes a warm interval from 1000 to 1200
A.D., when temperatures are believed to approximate the mean annual
temperature from 1930 to 1960; a cool period estimated to be a degree C lower
than the 1930 to 1960 mean annual temperature extending from the 13th century
into the 18th century; and European occupation, beginning in the mid-17th
century and accompanied by extensive land clearance. For at least 1000 years,
the geographic gradients of species composition and diversity have remained
constant. Seeds of Vallianeria americana are most common where salinities are
<1%, and seeds of Zannichellia palustris where salinities are >3%. Species
diversity decreases latitudinally as sedge and cordgrass marshes become more
extensive, with no macrophytes present in those tributaries surrounded by
Cyperaceae and Spartina alterniflora. Similarly, seeds are not found in the
Pocomoke River which is also surrounded by Taxodium dist^chum swamps. In one
area representing the longest record, species diversity increased from a
community consisting only of Zannichellia palustris when water levels were low
to a community of 7 species, as water depth increased with rising sea level.
Significant increases and decreases in populations of individual species
occurred within the broad geographic gradients, with the majority of increases
occurring during an interval immediately after European settlement and the
most significant decline within the last two decades. The post-European
increase is interpreted as a response of the plants to increased nutrient
input with deforestation and agriculture. However, as turbidity increased
with erosion and sedimentation, increases in populations were erased and
sporadic declines culminated in the most recent near extinction of all species
in the majority of upper Bay tributaries.
137
-------
Understanding the Estuary: Advances in Chesapeake
Bay Research. Proceedings of a Conference. 29-31
March 1988. Baltimore, Maryland. Chesapeake Research
Consortium Publication 129. CBPlTRS 24188.
Alternative Sampling Strategies for a Survey of Submerged
Aquatic Vegetation in the Chesapeake Bay
Paul H. Geissler
U.S. Fish and Wildlife Service
Patuxent Wildlife Research Center
Laurel, Maryland 20708
ABSTRACT
Complete aerial censuses of Submerged Aquatic Vegetation (SAV) in the Chesapeake Bay
including digitizing and mapping SAV beds have been conducted in 1978, 1984, 1985, 1986,
and 1987. This approach provides the best information, but it is expensive. Digitizing and
mapping a fraction of the Bay each year provides a less expensive alternative for monitoring
SAV. Annual estimates of the total area of SAV in the Bay would be available from sampling
for evaluating Bay recovery. For example, a 20% sample of the Bay could estimate the total
SAV area for the whole Bay within ± 15% and a 50% sample within ± 2%. The SAV beds in
the entire Bay could be periodically digitized and mapped for reviewing wetland permit
applications and for following changes on specific areas.
BACKGROUND
SAV is important to Chesapeake Bay because it provides important habitat for many species,
enhances water clarity, binds sediments, and enhances nutrient cycling (Orth, et a/. 1986?).
Complete aerial censuses in 1978, 1984, 1985, 1986, and 1987 have provided detailed
information but annual censuses are expensive (about $160K per year) and quality control of
the large amount of data is difficult. The total area of SAV beds and the area of four SAV
density classes are compiled for the Bay, for 3 zones (upper, middle, and lower bay), for 21
sections (e.g. watersheds) and for 164 U.S. Geological Survey quadrangles (135 with SAV for at
least one year). Complete SAV maps are also used to evaluate wetland permit applications and
to follow SAV changes for specific areas. Species composition is also estimated.
This paper will develop alternative sampling strategies for estimating total SAV area. Total
SAV area is the most important value and is widely believed to be an excellent indicator of
138
-------
water quality. I will discuss alternative survey designs, the definitions of sampling units and
the advantages of stratification, and show the relevance of each to an effective sampling
strategy. Confidence interval widths are projected for several sample sizes and alternative
stratifications. Statistical methods for designing and analyzing sample surveys of SAV are
presented and illustrated with a simple example.
ALTERNATIVE SURVEY DESIGNS
Sampling a portion of the Bay at a fraction of the cost of a complete mapping will allow
annual estimates of the total area of SAV and of changes in total area over time. Each year
digitizing, mapping, and reporting costs about SlOOK. This cost could be proportionally
reduced by sampling. Because aerial photographs are relatively inexpensive ($60K) compared
to the cost of digitizing and mapping SAV beds, the entire Bay could be photographed each
year. Then non-sampled areas in any given year could be digitized and mapped if the need
arose in the future. I will refer to a sample of part of the Bay as a survey and refer to a
complete coverage as a census. Three alternative sampling designs will be considered:
1. Complete annual censuses.
2. Annual sample surveys with rotating sampling units. Quarter quadrangles (for example)
are randomly divided into say five groups. One group is digitized and mapped each year on
a rotating basis so that the entire Bay will have been mapped at the end of five years. A
two year, three year or other rotation could also be used.
3. Annual sample surveys with fixed sampling units. A random sample of quarter
quadrangles (for example) is selected and the same quadrangles are digitized and mapped
each year.
Design 1 provides the best information but is the most expensive. Current SAV maps are
always available to evaluate wetland permit application and to follow small scale SAV
changes. The large amount of data makes quality control difficult. Designs 2 and 3 also
provide annual estimates of the total area of SAV for evaluating Bay recovery, but they are
less precise than the estimates from design 1. Designs 2 and 3 are equivalent for estimating the
annual total SAV area. They differ in the estimation of changes in the total SAV area
between two years. Design 2 is analogous to a unpaired t-test while design 3 is analogous to a
paired t-test. Design 3 is more precise than design 2 for estimating changes in total SAV area
between years. Design 2 has the advantage of providing a composite map for evaluating
wetland permit applications although different parts of the map would be digitized in different
years. Design 3 requires a periodic complete survey to provide maps. Estimated changes
would still be based on the fixed sampling units.
A ratio estimate should probably be used to analyze a SAV sample survey (designs 2 and 3)
because it exploits the correlation between the current sample survey and a previous complete
census to increase the precision of the estimates. The correlations among the annual SAV
areas for quadrangles ranged from 0.98 to 0.74, based on the 1978, 1984, 1985, and 1986
complete SAV area censuses.
SAMPLING UNITS AND STRATIFICATION
Sample surveys require a careful definition of sampling units and strata. Sampling units are
the plots on which the SAV areas are measured. Sampling units must not overlap and must
completely cover all areas of the Bay where SAV may occur. Sampling unit boundaries must
be unambiguous, identifiable, and unchanging from year to year. I have used quadrangles as
the sampling unit for this paper because data from previous surveys were not easily available
for smaller units.
Estimates can be reasonably made for the three zones using quadrangle sampling units but the
large size of quadrangles makes it impractical to estimate SAV areas for the 21 sections of the
139
-------
Bay. A minimum of two sampling units are needed in each of the 21 sections to estimate the
variance. This requires a minimum of 42 quadrangles (31% of the Bay) to obtain section
estimates. These estimates would not have the advantage of stratification on the SAV density
which would increase the precision of the estimates. Two SAV density strata would require 82
quadrangles (4 per section for 62% of the Bay). Quarter or ninth quadrangles may be better
sized sampling units. Even if section estimates are not required, quarter or ninth quadrangles
would provide more flexibility for stratification. Other sampling units are possible.
Stratification is the subdivision of the Bay into homogenous regions that are called strata. It
can dramatically increase the precision of the estimates with the same sample size or decrease
the cost of a survey by reducing the sample size. The precision of the SAV estimates for the
Bay or its subdivisions is estimated by comparing the SAV area of the sampling units in the
same stratum. Because differences among strata do not contribute to the sampling error,
grouping sampling units into homogenous strata can greatly increase the precision and decrease
the variance of the estimates. For example, the variance of 1, 2, 3, 11, 12, and 13 is 31 with 5
degrees of freedom. Dividing the numbers into two strata, [1, 2, 3] and [11, 12, 13], reduces
the variance to 2 with the loss of 1 degree of freedom. The stratification could be based on the
area of SAV in each sampling unit as measured during a recent census. If accurate
measurements are not available for the sampling units, they can be assigned to strata on the
basis of a visual estimate without biasing the estimates. This stratification reflects the current
distribution of SAV. Alternative stratifications based on area of water with the appropriate
depth or other criterion may better reflect the historic distribution of SAV and may better
reflect future SAV distribution.
The strata do not have to be equally sized or have the same number of samples. An optimal
allocation of samples to strata increases the precision of the whole Bay estimates. The number
of samples in a stratum should be proportional to the size of the stratum, proportional to the
standard deviation in the stratum, and inversely proportional to the square root of the cost to
measure the stratum. At least two sampling units must be selected from each stratum in order
to estimate the variance, but the sampling rate in unimportant strata may be decreased by
increasing the size of those strata. Important areas can be designated as separate strata and
completely sampled each year.
SURVEY ESTIMATES
Statistical methods for designing and analyzing sample surveys of SAV are presented in the
"survey estimates," "sample size," and "example" sections. These sections are intended as a
tutorial for those who may have to implement a sample survey and may be skipped by those
who wish an overview.
With a complete census (design 1) the total SAV area is directly measured. There is no
sampling error and consequently the variance and confidence interval widths are zero. The
following discussion relates to sample surveys (designs 2 and 3). The total SAV area for the
Bay or one of its subdivisions is the sum of the SAV area in each stratum (combination of zone
and SAV density strata)
A = £At = £Ntafc (1)
where A is the mean per unit estimate of the total SAV area for the Bay, Ak is the SAV area
in stratum k, Nfc is the number of sampling units in stratum k (including those that were not
n*
measured), and at = £at,/nt is the mean SAV area for the nt measured sampling units in
1=1
stratum k. With design 2 (rotating sampling units), separate estimates are made each year
using (1) and compared independently by observing if their confidence intervals overlap. With
design 3 (fixed sampling units) annual estimates of the total SAV area are also made
140
-------
independently using (1), but tests for changes in the total SAV area between two years uses the
differences in the SAV areas between the two years similar to a paired t-test. In this situation,
at in (1) is defined to be the mean difference in SAV area instead of the mean SAV area.
The variance of the total SAV estimate is the sum of the variances for each stratum
v(A) = £ v(A») = £ Nj
The estimate of the stratum variance for a total is N2. times the stratum variance of a mean
-sr- where s2. = -.—~? 53(ati-at) . Use the ffn_l definition on a calculator to estimate
* (.nJTVi=i\ I
the sample standard deviation sk. The estimate also includes the finite population correction
(l-M^V A 95% confidence interval for A is A ± tQ Q5 f ^v(A), where tQ 05 f is the value from
a t-table for 5% significance level with f degrees of freedom,
(3)
If f is not an integer, use the next lowest integer.
A ratio estimate exploits the correlation between the current sample survey and a previous
complete census to increase the precision of the estimates. The total SAV area from an earlier
complete census A' is multiplied by the average change per sampling unit. That average
change is estimated by the ratio R of the mean SAV areas from the current sample survey a^
to the mean areas on the same sampling units for the earlier complete census a't.
(4)
The variance of the ratio estimate is
'(A') = £
(5)
where s\ is the standard deviation of SAV areas for the earlier complete census using the same
sampling units as the current sample survey, and pk is the correlation between the SAV area
on the current survey and on the earlier complete survey. More information on survey
estimates may be found in Steel and Torrie (1980), Snedecor and Cochran (1980), and Cochran
(1977).
SAMPLE SIZE
I have projected the precision of estimates for the total area of SAV in the Bay and in each
zone for several sample sizes and stratifications using data from the 1978, 1984, 1985, and 1986
censuses provided by the Annapolis Field Office, U.S. Fish and Wildlife Service. I have used
the mean per unit estimate instead of the ratio estimate for planning because the correlations
between an earlier complete survey and the current sample survey may decrease over time,
decreasing the precision of the ratio estimate. Four SAV density strata were arbitrarily defined
for the sample size projections presented in this paper using the mean SAV area in the
quadrangles for the four censuses (density 1: 0-9 ha/quad, 2: 10-99, 3: 100-499 and 4: 500 and
more). Density strata 3 and 4 were combined in zone 2 because density stratum 4 had only a
single quadrangle. Quadrangles for which SAV areas were not available for any of the four
surveys are excluded.
141
-------
The number of sample quadrangles in each zone and density stratum was determined using an
optimal allocation which minimizes the variance of the estimate given the total sample size n:
nt = n
(6)
I have treated all quadrangles as if the cost to measure the area of SAV were the same. If
estimates of the cost to measure the SAV in a quadrangle are available (perhaps proportional
to the area of SAV) an alternate allocation formula is available. Say the total cost is
C = c° + £ ck nt (7)
where c°= fixed cost, and ct= cost of measuring the SAV area in a quadrangle in stratum k.
Then
nfc = (C-O
N
(8)
For planning estimates of SAV area, I estimated v(at) separately for each survey year and
used the mean of the four estimates. The precision of estimates of the annual total SAV area
will be the same for designs 2 and 3. The variance for changes between two annual total SAV
estimates with design 2 (rotating sampling units) is twice the variance of an annual estimate.
For changes with design 3 (fixed sampling units), I formed the differences between successive
years (1978-1984, 1984-1985, and 1985-1986), estimated v(a]t) separately for each difference,
and used the mean of the three estimates. Confidence interval widths are expressed as the
percentage of the estimate that must be added to and subtracted from the estimate
v(A)
100 where t, is the tabular t value for f degrees of freedom (f = number of
quadrangles minus the number of strata).
EXAMPLE
To illustrate the use of these equations, consider the following example with 5 quadrangles in
the first stratum and 6 in the second. The SAV area from one year is given in the column
headed "Complete Census." The SAV area for a sample of the quadrangles for the next year is
listed in the column headed "Sample Survey."
First Stratum
No. of
Quad.
10
15
18
16
27
Complete
Census
11.60
10.10
0.00
12.89
0.00
34.59
Sample
Survey
9.72
7.70
0.52
Second Stratum
No. of
Quad.
37
33
36
14
9
26
Complete
Census
223.91
97.90
346.69
132.99
439.96
586.96
1828.41
367.88
42177
Sample
Survey
36.57
164.37
369.54
295.21
216.42
21570
0.8615
35701
Mean* 7.23 5.98
Variance* 40 23
Correlation* 0.9958
Variance 41
* using only those sampling units that were measured in the sample survey.
Estimates for the sample survey are:
A! = 5 * 5.98 = 29.90 A2 = 6 * 216.42= 1298.52 A = 29.90 + 1298.52 = 1328.42
v(AJ = 52 (l-|) y = 77 v(A2)= 62 (l-J) 21|Zfi = 64710 v(A) = 77 + 64710 == 64787
f = f 77 + 64710 V / U^ -I- SiDJl?I _ 3 degrees of freedom.
142
-------
95% confidence interval for A is 1328 ± 3.182 ^J 64787 = (518 - 2138)
The ratio estimate, using the total SAV area from the complete survey 34.59 + 1828.41 =
1863, is
A* = 1863 *T'O ' = 1863 * 0.5921 = 1103
5*7. 1
« '
v(A°) =T \ V (23 + 0.59212*40 -
• (21570 + 0.59212*42177 - 2*0.5921*0.8615>|21570>|42177) = 16760
95% confidence interval for A* is 1103 ± 3.182 ^16760 = (691 - 1515) with 3 degrees of
freedom. In this example, the ratio estimate A° is much more precise than the mean per unit
estimate A because of the large correlations between the surveys.
With equal costs, the optimal sample size for the first stratum using a total sample of 7
quadrangles is n,= 7 — , — ^ — , = 0.03. There must be at least 2 samples in each
5 ^41 + 6^35701
stratum to estimate a variance, so take nt=2 and n2 = 7 - 2 = 5.
RESULTS AND DISCUSSION
Design 1, a complete census, provides the best estimates but is the most expensive. Because
the whole bay is digitized and mapped, there is no sampling error. Consequently, confidence
intervals have zero width. The annual total SAV area estimates have the same precision with
design 2 (sample survey with rotating sampling units) and design 3 (sample survey with fixed
sampling units). Designs 2 and 3 differ in the precision of the estimated change between two
annual estimates (Table 1).
An 18% sample of the quadrangles will provide estimates of the annual total SAV area with a
95% confidence interval width of ± 27.2% of the estimate with designs 2 and 3 (Table 1). Not
using SAV density strata greatly increases the width to ± 66.0% with a 19% sample. The
width is decreased to ;£ 15.1% with a 19% sample if zone estimates are not required.
Confidence intervals are quite wide for zone estimates (± 46.4%, ± 132%, and ± 40.4% for an
18% sample with four SAV density strata).
A 48% sample of the quadrangles will provide estimates of the annual total SAV area with a
95% confidence interval width of ± 2.0% of the estimate with designs 2 and 3 (Table 1).
Using two instead of four SAV density strata increases the width to ± 4.6% with a 48%
sample. Not using SAV density strata greatly increases the width to ± 29.8% with a 49%
sample. There is little change in the width without zone estimates (±. 2.2% with 49% sample)
because there are adequate samples in each zone. Confidence intervals are wider for zone
estimates (± 5.8%, ± 11.1%, and ± 1.4% for an 48% sample with four SAV density strata).
An 18% sample of the quadrangles will estimate changes between two annual total SAV areas
with a confidence interval width of ±. 38.5% of the annual estimate with design 2 (rotating
sampling units) and ± 19.2% with design 3 (fixed sampling units). Design 3 is more precise
because changes on the same sampling units are followed over time, but a periodic complete
census is required to digitize and map all parts of the Bay. Not using SAV density strata
increases the widths to ± 93.3% and ± 35.8% (19% sample) with designs 2 and 3 respectively.
The widths are decreased to ± 21.4% and ± 16.0% (19% sample) if zone estimates are not
required.
143
-------
Table 1. Optimal allocation of sample quadrangles to upper (1), middle
(2), and lower (3) zones of the Bay and four SAV density strata.
Projected 95% confidence interval widths are expressed as the
percentage of the estimate that should be added to and subtracted from
the estimate. Confidence intervals widths are given for estimates of
annual total SAV area (same for designs 2 and 3) and for the change in
SAV area between two successive annual total SAV areas with designs 2
(rotating sampling units) and 3 (fixed sampling units).
Den. SAV Var. SAV Number of Sample Quadrangles
Zone str.mean mean total Quads 16% 18% 24% 32% 40% 48% 59% 69%
115 50 96 19 2 2 2 2 2 2 5 9
1 2 35 2319 277 822224788
1 3 216 30014 1298 622356666
1 4 1108 672258 2216 222222222
213 26 111 35 2 2 2 2 2 3 7 13
2 2 29 782 261 9 2 2 2 2 2 5 9 9
2 3 292 54915 2041 722477777
313 29 52 16 2 2 2 2 2 2 3 6
3 2 41 1556 405 10 2 2 2 2 4 8 10 10
3 3 248 16345 3716 15 2 2 5 10 15 15 15 15
3 4 940 236342 7518 824888888
Bay total quadrangles 135 22 24 33 44 54 65 80 93
95% conf. int. ± % (total) 37.227.213.3 6.5 3.4 2.0 1.0 0.6
95% conf. int. ±% (change 2) 56.2 38.5 18.8 9.2 4.8 2.8 1.4 0.8
95% conf. int. ± % (change 3) 21.319.213.1 6.7 3.8 2.6 1.4 0.8
Zone 1 total quadrangles 35 8 8 8 11 14 17 21 25
95% conf. int. ±% (total) 46.4 46.4 46.4 19.2 9.4 5.8 2.7 1.7
95% conf. int. ± % (change 2) 65.6 65.6 65.6 27.2 13.3 8.2 3.8 2.4
95% conf. int. ± % (change 3) 62.2 62.2 62.2 23.2 10.5 6.9 3.6 2.3
Zone 2 total quadrangles 51 6 6 8 11 11 15 23 29
95% conf. int. ±% (total) 132 132 60.9 18.9 18.9 11.1 5.2 3.3
95% conf. int. ± % (change 2) 187 187 86.1 26.7 26.7 15.7 7.4 4.7
95% conf. int. ±% (change 3) 89 89 44.5 21.9 21.9 13.9 7.3 4.7
Zone 3 total quadrangles 49 8 10 17 22 29 33 36 39
95% conf. int. ±% (total) 66.3 40.4 13.8 7.8 2.9 1.4 0.8 0.4
95% conf. int. ± % (change 2) 93.857.119.511.0 4.1 2.0 1.1 0.6
95% conf. int. ±% (change 3) 32.5 24.0 11.2 6.9 3.1 1.8 1.2 0.8
144
-------
Table 1. Continued.
Combined
Den.
Zone str.
1 12
1 34
2 12
2 34
3 12
3 34
Totals
95% conf.
95% conf.
95% conf.
SAV density strata.
SAV var. SAV
mean
15
439
10
292
22
488
int.
int.
int.
mean total
1035 411
291304 3514
365 452
54915 2041
1117 561
209085 11234
± % (total)
± % (change
± % (change
Quads
27
8
44
7
26
23
135
103
2) 146
3) 42
9%
2
2
2
2
2
2
12
.5 49
.4 69
.3 24
Number of Sample
13%
2
3
2
2
2
7
18
.2 27
.6 39
.6 16
21%
2
6
2
2
2
14
28
.6
.0
.2
30%
2
8
2
3
2
23
40
13.9
19.7
12.3
39%
5
8
5
7
5
23
53
6.7
9.5
6.5
Ouadranales
48%
9
8
9
7
9
23
65
4.6 3
6.5 4
4.4 3
59% 70%
14 19
8 8
14 18
7 7
14 19
23 23
80 94
.2 2.4
.5 3.4
.0 2.2
79%
23
8
22
7
23
23
106
1.8
2.5
1.6
Without SAV density strata.
Den
Zone str.
1
2
3
Totals
95% conf.
95% conf.
95% conf.
SAV
mean
119
58
278
int,
int.
int,
1
2
3
4
Totals
95% conf.
95% conf.
95% conf.
4
35
250
973
int
int
int
Var.
mean
98059
20732
168239
SAV
total
4167
2981
13642
Number of
Quads
35
51
49
135
9%
3
2
7
12
19%
7
5
14
26
29%
11
7
21
39
Sample Quadrangles
39%
15
10
28
53
49%
19
12
35
66
59%
23
15
42
80
70%
27
18
49
94
79%
35
23
49
107
Without zones.
Den. SAV
Zone str.mean
± % (total) 115.1 66.0 49.2 37.8 29.8 23.0 16.6 11.0
± % (change 2) 162.8 93.3 69.6 53.5 42.1 3.2.5 23.5 15.6
+ % (change 3) 63.6 35.8 27.8 21.8 18.2 15.2 12.4 9.2
Var. SAV Number of Sample Quadrangles
mean total Quads 11% 19% 30% 39% 49% 60% 70% 79%
40 278 70 2 2 2 4 8 16 29 42
1577 952 27 2
29156 7013 28 5
280806 9734 10 6
17978 135 15
±% (total) 30.615.1 7.8 3.6 2.2 1.0 0.6 0.4
+ % (change 2) 43.321.411.0 5.1 3.1 1.4 0.8 0.6
± % (change 3) 28.216.0 8.6 4.2 2.6 1.4 1.0 0.6
2
12
10
26
5
23
10
40
11
28
10
53
20
28
10
66
27
28
10
81
27
28
10
94
27
28
10
107
145
-------
A 48% sample of the quadrangles will estimate changes between two annual total SAV areas
with a confidence interval width of ±_ 2.8% of the annual estimate with design 2 (rotating
sampling units) and ± 2.6% with design 3 (fixed sampling units). Using two instead of four
SAV density strata increases the widths to ± 6.5% and ±_ 4.4% for designs 2 and 3 respectively
(48% sample). Not using SAV density strata increases the widths to ± 42.1% and ± 18.2%
(49% sample). The widths are about the same (± 3.1% and ± 2.6% with a 19% sample) if
zone estimates are not required.
CONCLUSIONS
* Sampling can substantially reduce the cost of SAV surveys for estimating annual total SAV
areas and changes in total SAV area between years.
* Quadrangles are too large as sampling units to estimate SAV areas for sections of the Bay
and reduce the precision when zone estimates are required. Quarter or ninth quadrangles
appear to be better sized sampling units because they allow greater opportunity for
stratification and larger sample sizes.
* Precision is greatly increased by stratification on SAV density.
* Designs 2 and 3 provide equally precise estimates of the annual total SAV area.
* Design 3 (fixed sampling units) provides more precise estimates of the change in total SAV
area between years than design 2 (rotating sampling units) but requires a periodic complete
census to provide maps for reviewing wetland permit applications and for following changes
on specific areas. Design 2 does not require a periodic complete census, but the composite
map contains parts mapped in different years.
* A ratio estimate should probably be used to estimate total SAV area and changes in total
SAV area.
ACKNOWLEDGEMENTS
J. Booth, L. Hurley, and F. Seavy provided the data. B. Brun, C. Bunck, N. Coon,
J. Hestbeck, and W. Link provided helpful suggestions on the manuscript.
REFERENCES
Cochran, W.G. 1977. Sampling techniques. Wiley, New York, 428pp.
Orth, R., J. Simons, J. Capelli, V. Carter, L. Hindman, S. Hodges, K. Moore, and N.
Rybicki. 1986? Distribution of submerged aquatic vegetation in the Chesapeake Bay and
tributaries - 1985, EPA. Final Report. 296 pp.
Snedecor, G.W. and W.G. Cochran. 1980. Statistical methods. Iowa State Univ. Press,
Ames, 507 pp.
Steel, R.G.D. and J.H. Torrie. 1980. Principles and procedures of statistics.
McGraw-Hill, New York, 633 pp.
146
-------
Understanding the Estuary: Advances in Chesapeake Abstract only
Bay Research. Proceedings of a Conference. 29-31
March 1988. Baltimore, Maryland. Chesapeake Research
Consortium Publication 129. CBPlTRS 24/88.
The Status of Yellow Perch Populations in Five
Chesapeake Bay Tributary Streams
H. Greening, A. Janicki, and P. Sounders
International Science & Technology, Inc.
11260 Roger Bacon Drive, Suite 201
Reston, Virginia 22090
Yellow perch commercial fishery landings have declined significantly in
many Chesapeake Bay tributaries since the early 1970's. The proposed partial
closings of the yellow perch commercial and recreational fishery by the
Maryland Department of Natural Resources reflect the concern by State
officials that yellow perch populations may be declining in many areas of the
Chesapeake Bay. Although commercial statistics and anectodal information from
recreational fishermen suggest declining yellow perch populations, a more
accurate assessment of yellow perch abundance is needed to determine the
status of this important anadromous fish species in Chesapeake Bay waters.
As part of a study to examine effects of the addition of calcium
carbonate on freshwater tributary stream chemistry and biota, yellow perch
spawning runs, ichthyoplankton abundances, and juvenile abundances were
examined in five Chesapeake Bay tributaries during 1986 and 1987. Results
indicate that the abundances of spawning adults in four historically important
spawning streams near Annapolis are very low, with fewer than 50 spawning
individuals observed in Bacon Ridge Branch, North River, and Magothy Run, and
fewer than 200 spawning adults in Severn Run. The majority of the spawning
adults observed in these streams are large, relative old (6-8 year)
individuals nearing the end of their lifespan. In contrast, results from
Mattawoman Creek (a tributary to the Potomac River) show that the yellow perch
spawning run is relatively strong, with more than 10,000 individuals collected
during the 1987 spring run. The majority of spawning yellow perch in
Mattawoman Creek were estimated to be 2 to 5 years old, with few individuals
older than 6 years of age collected.
As expected from the results of the spawning survey, ichthyoplankton
abundances in Bacon Ridge Branch and North River were lower than in Mattawoman
Creek. Yellow perch larvae were present in ichthyoplankton collections for
these three streams throughout most of the 1987 study period (March through
June). The occurence of yolk-sac larvae long after the observed spawning
period in Bacon Ridge Branch and North River suggest the possibility of
resident yellow perch populations in these streams.
147
-------
As with the spawning survey, greater numbers of juvenile yellow perch
were observed in the Mattawoman Creek estuary collections (126 individuals)
than in the other estuarine seine collections. No yellow perch were collected
in the 1986 or 1987 South River, the 1986 Magothy River, or the 1986 Severn
River seine collections, and only 5 individuals were observed during the 1987
Severn River juvenile survey. Results of the 1986-1987 seining studies are
compared with results of seining surveys conducted at the same sampling
locations in the mid-1970's.
These results suggest much lower current yellow perch populations in the
Annapolis region study streams (Bacon Ridge Branch, North River, Magothy Run,
and Severn Run) than had been estimated in the 1960's and 1970's. In
addition, the current status of yellow perch populations may be different in
different streams or regions of the Chesapeake Bay.
148
-------
Understanding the Estuary: Advances in Chesapeake Abstract only
Bay Research. Proceedings of a Conference. 29-31
March 1988. Baltimore, Maryland. Chesapeake Research
Consortium Publication 129. CBP/TRS 24/88.
Bay Anchovy Ecology in Mid-Chesapeake Bay
T. A. Nevberger and E. D. Houde
University of Maryland
Center for Environmental & Estuarine Studies
Chesapeake Biological Laboratory
Solomons, Maryland 20688-0038
E. J. Chesney
Louisiana University Marine Consortium
Star Route Box 541
Chauvin, Louisiana 70344
Bay anchovy Anchoa mitchilli is the most abundant and ubiquitous fish in
the Chesapeake Bay. It is believed to be a major link in pelagic food chains
through its role in converting planktonic biomass into available forage for
piscivorous fishes. Its population ecology, including abundance,
distribution, age, growth, reproduction, and trophic significance, is being
investigated in mid-Chesapeake Bay. A trawl sampling program was conducted on
a transect from inside the Patuxent River mouth to four kilometers offshore,
from March 1986 to November 1987. Mean among month Catch per Unit Effort
(CPUE) (numbers per 10 minutes of trawling) varied ten-fold in each year.
CPUE was highest in September when catches were dominated by 0+ recruits.
July to November CPUE was more than 5 times higher in 1986 than in 1987. Ten-
fold annual variability in abundances have been observed in the past 30 years,
based on Maryland Department of Natural Resources index surveys in Chesapeake
tributaries. Annuli in otoliths indicate that there are four age groups.
Maximum age is 3+ when anchovies reach 85 mm fork-length and 5 g wet-weight.
Most growth-in-length is completed by age 1+, but large weight increases occur
in older fish. Size-at-age is variable and in part attributable to a
protracted spawning season and prolonged recruitment period.. Based on a
gonosomatic index, the reproductive season extends from May through August.
Apparently all mature females spawn each night between 9 p.m. and I a.m.
Batch fecundities range from 514 to 2,026 eggs and are directly related to
female weight. Spawned eggs were more abundant in 1987 on a transect off the
Choptank River than off the Rappahannock River, but larval anchovies were more
common on the Rappahannock transect, indicating a differential mortality
between the two areas. Eggs and larvae were relatively common below the
pycnocline on the Rappahannock transect but rare on the Choptank transect,
suggesting that low oxygen limits the depth distribution of anchovy eggs and
larvae. Future studies will include estimates of mortality rates of early
life stages and examination of trophic' relationships, energetics, and oxygen
tolerances of all life stages to understand better the role of bay anchovy in
pelagic, estuarine communities.
149
-------
Understanding the Estuary: Advances in Chesapeake Abstract only
Bay Research. Proceedings of a Conference. 29-31
March 1988. Baltimore, Maryland. Chesapeake Research
Consortium Publication 129. CBPlTRS 24/88.
Hatchability and Viability of Striped Bass Eggs:
Effects of Tributary Source and Female Size
E. D. Houde, C. E. Zastrow, and E. H. Sounders
University of Maryland
Center for Environmental & Estuarine Studies
Chesapeake Biological Laboratory
Solomons, Maryland 20688-0038
Variability in hatching success, survival of larvae to three days
posthatch (i.e. viability), and the quality of eggs and larvae were examined
in relation to spawning tributary in which females were captured, and size of
spawning female. Fertilized eggs from 26 females collected during 1986 in the
Nanticoke River, Patuxent River, and C&D Canal were tested. Most variables
that were examined differed significantly among female spawners but were not
related to spawning tributaries from which females originated. Mean
hatchability of 30-36h post-fertilization eggs was 21% lower for spawns of <
10 Ib females than for spawns of > 10 Ib females. Larval survival to 3 days
posthatch was essentially independent of the probability of hatching. Egg dry
weights, egg yolk and oil globule volumes, and weights per egg of protein and
lipid all varied by 1.5- to 3-fold among spawns from individual females. The
smallest eggs and least amounts of constituents were from < 10 Ib females.
Eggs and 5-day posthatch larvae from < 10 Ib females weighed, on average, only
68% and 67% as much, respectively, as those from larger females. Results of
the analysis imply that young females from the most recently-matured year-
classes (i.e. 1981-82) did not produce eggs of high quality in 1936, compared
to those produced by older, larger females. The ultimate probability of
survival to recruitment of small larvae that hatched from eggs of small
females cannot be determined from our analysis, but size seemingly would
increase vulnerability to both starvation and predation-related mortality. It
is probable, although not certain, that 1981-82 year-class females will
produce larger eggs of higher quality when they grow to larger size.
Implications for striped bass management and hatchery production may be
important. Aside from low fecundity and egg yield, small females (i.e. < 10
Ib) from Maryland tributaries produce relatively poor quality eggs. Our data
quantify the relationships, document variability in egg constituents, and
allow estimation of hatchability and survival rates, as well as egg
characteristics that will be useful in future hatchery applications. Because
hatching varies in relation to adult female size and age, results also can aid
development of age or size-specific egg production models for Chesapeake Bay
striped bass.
750
-------
Understanding the Estuary: Advances in Chesapeake Abstract only
Bay Research. Proceedings of a Conference. 29-31
March 1988. Baltimore. Maryland. Chesapeake Research
Consortium Publication 129. CBP/TRS 24/88.
Current Research to Improve the Understanding of Habitat Use,
Distribution, and Population Status of Canvasbacks
on the Chesapeake Bay
G. Michael Haramis and Dennis G. Jorde
U.S. Fish and Wildlife Service
Patuxent Wildlife Research Center
Laurel, Maryland 20708
Abstract: The U.S. Fish and Wildlife Service is currently conducting a
5-year research project to study canvasbacks using the Chesapeake Bay during
migration and winter. The objectives of the study are to estimate daily
survival rate/ identify causes of mortality, evaluate habitat use, determine
distribution and abundance of key aquatic foods, and estimate energy and
nutritional value of natural foods. Radio-telemetry is being used to evaluate
survival, movement patterns, and habitat use of juvenile female canvasbacks
wintering on the Chesapeake Bay. Aquatic food availability and distribution
will be determined using currently available data bases and field sampling of
nutritional aspects of the most important aquatic foods will be evaluated
during feeding and reproduction of captive canvasbacks at Patuxent Wildlife
Research Center.
757
-------
Understanding the Estuary: Advances in Chesapeake Abstract only
Bay Research. Proceedings of a Conference. 29-31
March 1988. Baltimore, Maryland. Chesapeake Research
Consortium Publication 129. CBP/TRS 24/88.
Patterns of Post Larval Availability and Settlement in the Blue Crab:
Effects of Time, Space, and Habitat
Eugene J. Olmi III, Jacques van Montfrans,
Romuald N. Lipcius, and Robert J. Orth
Virginia Institute of Marine Science
The College of William and Mary
Gloucester Point, Virginia 23062
One focus of a long term program to elucidate blue crab (Callinectes
sapidus) recruitment dynamics in Chesapeake Bay is the resolution of post
larval availability and its relationship to settlement patterns in time and
space. Postlarvae and early juveniles were collected on three dates from
three habitats (plankton, artificial settlement substrates and submersed
vegetated bottom) within each of two sites at two geographic locations in the
York River. Relationships of postlarval and early juvenile abundance varied
among time, location and habitat. These complex patterns suggest the
importance of fine-scale processes in settlement dynamics.
752
-------
Understanding the Estuary: Advances in Chesapeake Abstract only
Bay Research. Proceedings of a Conference. 29-31
March 1988. Baltimore, Maryland. Chesapeake Research
Consortium Publication 129. CBPITRS 24/88.
Regulatory Mechanisms of Postlarval Blue Crab Recruitment:
Settlement, Metamorphosis, and Developmental State
Romuald N. Lipcius, Eugene J. Olmi HI,
and Jacques van Montfrans
Virginia Institute of Marine Science
The College of William and Mary
Gloucester Point, Virginia 23062
As part of a long-term program on blue crab, Callinectes sapidu3r
recruitment dynamics in Chesapeake Bay, we quantified the developmental state
(molt stage: proximity to metamorphosis) of recruiting postlarvae in relation
to time, habitat and geographic location. Developmental state was assessed by
molt staging the uropods of freshly-collected postlarvae, and verifying the
staging procedure with laboratory cultures. Geographic locations included two
sites each at the mouth of the York River and 10 km upriver; habitats included
plankton, artificial settlement substrate, and submersed vegetated bottom.
Temporal sampling comprised daily and monthly variation. Various main and
interaction effects of time, habitat and geographic location were significant,
and yielded a complex model of the interrelationships between postlarval
developmental state, the scale and nature of spatio-temporal factors, and
regulatory mechanisms of recruitment in the blue crab.
153
-------
Understanding the Estuary: Advances in Chesapeake Abstract only
Bay Research. Proceedings of a Conference. 29-31
March 1988. Baltimore, Maryland. Chesapeake Research
Consortium Publication 129. CBPlTRS 24/88.
Variation in Postlarval Blue Crab Settlement on Artificial
Substrates in the York River, Virginia
Jacques van Montfrans and Robert J. Orth
Virginia Institute of Marine Science
The College of William and Mary
Gloucester Point, Virginia 23062
Annual and temporal variation in blue crab settlement on artificial
substrates was examined during the period of primary postlarval ingress into
the York River (Aug. - Dec.) over a three year period (1985-1987). Daily
records indicated 1-3 day settlement events correlated with lunar phase;
settlement was strongly associated with full moon and less so with new moon.
Significant differences existed in the magnitude of settlement between years.
Juvenile populations of blue crabs which settled or recruited into a seagrass
bed approximately 12 km nearer the mouth of the York River exhibited similar
annual variations in abundance. Settlement dynamics must be examined on
various scales ranging from days to years in order to understand the
regulation and patterns of recruitment in the blue crab.
154
-------
CONCURRENT SESSIONS
AND
POSTER SESSION:
NUTRIENTS
Chairs:
George Simmons
Virginia Polytechnic Institute and
State University
Carl Cerco
U.S. Army Corps of Engineers
-------
Understanding the Estuary: Advances in Chesapeake
Bay Research. Proceedings of a Conference. 29-31
March 1988. Baltimore, Maryland. Chesapeake Research
Consortium Publication 129. CBP/TRS 24/88.
Radionuclide Concentrations in Susquehanna River and
Chesapeake Bay Sediments •- Implications for Transport
and Distribution of Particle-Reactive Pollutants
R. 1. McLean and S. L. Domotor
Maryland Department of Natural Resources
Power Plant Research Program
Tawes State Office Building, B-3
Annapolis, Maryland 21401
J. K. Summers and V. A. Dickens
Versar, Inc.
ESM Operations
9200 Rumsey Road
Columbia, Maryland 21045
C. R. Olsen
Environmental Sciences Divison
Oak Ridge National Laboratory
P.O. Box X
Oak Ridge, Tennessee 37830
The Peach Bottom Atomic Power Station (PBAPS) has contributed measurable
quantities of radioactivity to the lower Susquehanna River and Chesa-
peake Bay. Since 1980, we have monitored, in spring and fall, concen-
trations of plant-related radionuclides in sediments. Mass balance
estimates derived from grab samples indicate that less than 20% of
particle-reactive radiozinc, radiocobalt, and radiocesium is present
in surface sediments «10 cm) of the Conowingo Reservoir, an impoundment
of the lower Susquehanna. The remaining release inventory is assumed to
be transported to the Chesapeake Bay. Significant seasonal variations
in radionuclide trapping efficiency by the reservoir are not apparent.
Core samples confirm that some, but not all of this surface sediment
radionuclide inventory, remains within the reservoir — trapped in dis-
crete locations by subsequent sediment accumulation. The detection of
PBAPS-related radionuclides in sediments of the upper Chesapeake Bay
(above Baltimore, MD) confirms transport of these radionuclides from the
Susquehanna. Radionuclide concentrations in sediments were generally
undetectable south of the Sassafras River. This is attributed to dis-
persion of radionuclide-labeled particles, dilution by unlabeled
particles from other upper-bay tributaries, and possibly, desorption of
particle-bound radioactivity.
INTRODUCTION
The Susquehanna River is the principal source of fresh water and fluvial
sediments in the Upper Chesapeake Bay (Schubel 1972). Because many
pollutants, including trace metals, radionuclides and trace organics
157
-------
have a high affinity for association with particles suspended in the
water column, information regarding the transport and environmental
distribution of particulates is important in determining the fate of
Susquehanna-derived pollutants in the Upper Chesapeake Bay.
Since 1980, the Power Plant Research Program has conducted environmental
monitoring in the Susquehanna River and Upper Chesapeake Bay to assess
the radioecological impact of radioactivity released by the Peach Bottom
Atomic Power Plant. Located approximately 5 km north of the Pennsylvania-
Maryland border, the plant has discharged low levels of radiocesium
(Cs-134 and Cs-137), radiozinc (Zn-65) and radiocobalt (Co-60) to the
river since initial operation in 1975. Low levels of these radionu-
clides have consistently been detected in biota and sediments of the
lower Susquehanna and Upper Chesapeake Bay (McLean and Domotor 1988).
Because these radionuclides are particle-reactive, they may be scavenged
by particulate matter suspended in the water column, ultimately to be
deposited on the river bottom, or transported downriver as partlculates
or resuspended sediment. Because sediments serve as ultimate sinks for
these radioactive metals, we have extensively monitored radionuclide
concentrations in the sediments of the Conowingo Pond, the Susquehanna
Flats, and the Upper Chesapeake Bay. Results of this sediment monitor-
ing program have provided information useful in describing the transport
and fate of other particle-reactive pollutants introduced into the
Susquehanna River.
METHODS
Surface sediments «10 cm) were collected twice annually from 1981
through 1986 — once in the spring (Apr/May) and once in the fall (Sep/
Oct) — when Susquehanna River flow is at its maximum and minimum,
respectively. Collection sites in the Susquehanna River and Upper
Chesapeake Bay are shown in Figure 1. Sampling locations in the Cono-
wingo Pond are shown in Figure 2. Radionuclide concentrations (pCi/kg)
in Conowingo Pond sediments were converted to mass units (curies) using
sample volumes and sediment densities. Radionuclide mass in the upper
10 cm of sediment was estimated for 22 areas (cells) within the Cono-
wingo Pond sampling grid by extrapolation of each sample to the sediment
volume (to 10 cm) of the respective cell. Radionuclide inventories in
surface sediments of the Conowingo Pond are summations of the radionu-
clide masses within each cell. Deep cores (ca. 200 cm) were taken at
selected locations. Radionuclide mass balance budgets were estimated
using monthly radionuclide release quantities reported by the power
plant operator. This source term was adjusted for radioactive decay.
RESULTS AND DISCUSSION
As indicated in Figure 3 and Table 1, less than 12% of the decay-
adjusted Co-60, 21% of the Zn-65, and 8% of the Cs-134 released by
Peach Bottom was found on any sampling date within the upper 10 cm of
Conowingo Pond sediments. Significant differences in seasonal (spring/fall)
radionuclide inventory are not apparent (Table 1). This implies, that
approximately 80% of the particle-reactive radionuclides released to the
Susquehanna by Peach Bottom is either, or both: 1) transported in solu-
ble or particulate-associated form beyond the Conowingo Dam, or
755
-------
I lollwood fleservoif
Figure 1. Sediment collection locations in the Susquehanna River and
Upper Chesapeake Bay.
759
-------
Figure 2. Sediment collection locations in the Conowingo Pond,
760
-------
16
15 '
14 •
13-
12 '
LLJ 11 '
5 101
- 9 '
< 8 '
K 7'
6 '
5 '
4 '
3 '
CO-60 .
60-
50-
40
30'
20-
10'
0 1
ZN-65
10-
9
CS-134
* *
1981
1982
1983
1984
1985
1986
1987
Figure 3. Percentage of decay-adjusted radionuclide release quantities
present in surface «10 cm) sediments of the Conowingo Pond, spring and
fall, 1981-1986. Solid line describes mean estimates. Asterisks are
upper limit estimates derived from 2a counting uncertainties.
161
-------
Table 1. Percentage of decay-adjusted radionuclide release quantities
present in surface «10 cm) sediments of the Conowingo Pond, spring and
fall, 1981-1986.
Co-60
RADIONUCLIDE
Zn-65 Cs-134
Cs-137
1981
Spring
Fall
1982
Spring
Fall
1983
Spring
Fall
1984
Spring
Fall
1985
Spring
Fall
1986
Spring
Fall
Mean,
+/- 1 S.E.
Spring
Fall
4.3
6.2
8.5
6.0
11.8
11.1
5.1
3.5
5.1
8.8
6.3
11.6
6.9
+/-2.S
7.9
+/-3.2
5.0
4.8
14.0
12.0
20.0
18.0
16.5
11.0
5.0
20.0
13.0
21.0
12.3
+/-6.1
14.5
+7-6.3
1.8
2.8
2.5
2.5
2.6
7.8
1.6
1.5
2.9
5.1
6.5
3.9
3.0
+/-1.8
3.9
+7-2.3
8.8
10.5
9.0
7.7
9.5
10.3
6.4
6.5
6.3
6.7
5.3
4.7
7.6
+/-1.8
7.7
+7-2.3
762
-------
remains trapped within the reservoir and is buried below the 10 cm sam-
ple depth by subsequent sediment accumulation. Our data suggest that,
with the exception of selected locations, appreciable burial of these
radionuclides within Conowingo Pond does not occur. Data presented in
figures 4 and 5 indicate that these radionuclides are consistently
available in sediments collected from the immediate downstream vicinity
of the Peach Bottom discharge on the western shore of the Conowingo
Pond. The two nearest sampling locations (LYH-1 and BC-1) account for
about 70% of the Conowingo Pond inventory. These figures present mean
percentages of the available radionuclide mass for each sampling cell
within the Conowingo Pond. The percentage of the total mass found
within each cell during each sampling period (not presented) indicates
that there is not a significant variation in percentage over time.
This fact argues against appreciable radionuclide burial, as subsequent
samplings would reflect higher percentages in later collections —
assuming sediment accumulation does not exceed 10 cm in the 5-6 month
period between spring and fall samplings.
Because surface grabs indicated that LYH-1 and BC-1 were areas of great-
est radionuclide concentration within the Conowingo Pond, deep core
samples were collected at these locations. Cores were also collected
from two other locations on the eastern shore downstream of the Peach
Bottom discharge (BC-3, CON CK-3). These locations were regarded as
representative of locations downstream of BC (i.e., CON CK and DAM
transects), given the general similarity in surface sediment inventories
at these locations (see Figures 4 and 5).
Table 2 compares, for the same collection period, estimates of the per-
centage of decay-adjusted radionuclides within cells, derived from sur-
face grabs and deep cores. Core samples were sufficiently deep as to
account for all buried Peach Bottom-derived radioactivity (the deepest
penetration for Cs- 134 was about 60 cm). It is apparent that some bur-
ial of radionuclides occurs at the LYH-1 and BC-1 locations, but little
occurs at the other two sites. These data indicate that mass balance
estimates derived from surface grabs alone underestimate the percent-
age of radioactivity retained within Conowingo Pond. The underestima-
tion appears, however, to apply only to the BC-1 and LYH-1 locations as
comparisons of deep core and surface estimates are not significantly
different at the other locations. Therefore on these sampling dates,
the buried increment at BC-1 and LYH-1 combined would result in an
increase in mass balance for Conowingo Pond of 16% for Co-60, 5% for
Cs-134, and 10% for Zn-65. Given the maximum percentages presented in
Table 1, adjusted mean mass balances for Conowingo Pond are 28% for
Co-60, 13% for Cs-134, and 31% for Zn-65. It was not possible to obtain
cores for all locations, however, we feel the data collected provide a
reasonable estimate of the degree of retention of these radionuclides
within Conowingo Pond.
Although there is no significant net difference between spring and fall
mass balances for the Conowingo Pond as a whole (Table 1), there appear
to be seasonal differences in the distribution of radioactivity among
sampling grid cells (Figures 4 and 5). During both spring and fall,
western shore locations account for higher percentages of total avail-
able mass than raid-river or eastern shore Locations. Collections made
163
-------
CO-60
a
0
a
0
a
0 0)
£7
0 It))
a
0 OS
O
I Mil
ZN-65
CS-134
aaaoooo
0 / 0 / 0 Old / O.ll) / 0 I)J / 0 lit / 0.0«)J
a a o
T7i
),» / 2.01 / OH)/ 0 I)]
a
0
3 la laoaooo
0 / 0 01 / 0) / 0.211 / 0 OM / O.IJJ
a a a
O
0000
0.! / O.OJ1 / O.HJ / O.lli
I
00
1 / l.n / O.OM / 0.!
a/o/a/oaaa
0 / 0 / 0 OIM / O.OM / 0.0« / 0 Oil / 0.«)
2/0 o o I Q I a / o la
« 13 / 0 01) / O.OS / 0 OS
1/0/0/0
0/0/0
Q I W I O I O
OH / O.JIt / 0.15 / 0 0»)
COLD MJI PB LYH BC CON DAM
CABIN
CK
Figure 4. Mean percent of decay-adjusted radionculide release quanti-
ties in each Conowingo Pond sampling location, spring collections, 1981-
1986.
164
-------
CO-60
a
0
3 /O/0/O/O/O/O/O
' / 0 / 0 05 / 0 ?8J / 0 15 / o I / 01
O
0
a a Q o a
o.ojjj / o iiu / o 5«iJ / e njj / »i
a a
¥71
OOJJ Jill
o a
t M / o !IM
ZN-65
O
a
J.J O.OM
0 IU
£7
0
a
0
O
0
O
O 0
0.15 O.HJ t.JJJ OOJJ
0/cr/c7/m//w/G/O
0 / 0 / 0 / I.H / ! 5t( / ».JW / O.W
CS-134
I
0/0/0/0/0/0/0
0 / ».OI« / 0.05 / O.lll / O.IU / 0.01) / 0.08J
a I o I a o a a a
O.OIll / O.DJ / 0 !M / O.IJJ / 0.05
a a o
0/0/0
a a
I OM / I.M) / O.ZJJ / 0 15
COLD MJI PB LYH BC CON DAM
CABIN CK
Figure 5. Mean percent of decay-adjusted radionuclide release quanti-
ties in each Conowingo Pond sampling location, fall collections, 1981-
1986.
765
-------
Table 2. Percentage of decay-adjusted radionucllde quantities at
selected sampling locations for deep core (ca. 200 cm) and surface grabs
(10 cm).
Date
MAYS 5
MAYS 5
MAYS 5
MAYS 5
OCXS 5
OCT85
OCT85
OCT85
OCT85
OCT85
OCT85
OCT85
OCT85
OCT85
OCT85
OCT85
Location
LYH
LYH
LYH
LYH
BC
BC
BC
BC
BC
BC
BC
BC
CON CK
CON CK
CON CK
CON CK
Station
1
1
1
1
1
1
1
1
3
3
3
3
3
3
3
3
Nuclide
CO-60
CS-134
CS-137
ZN-65
CO-60
CS-134
CS-137
ZN-65
CO-60
CS-134
CS-137
ZN-65
CO-60
CS-134
CS-137
ZN-65
% of
Available
(Core)
15.10
3.88
8.41
13.23
5.58
3.26
37.04
2.99
0.15
0.51
13.85
0.21
0.01
0.03
4.84
0.12
% of
Available
(Surface)
3.06
0.80
0.91
3.34
1.38
1.41
1.17
4.53
0.34
0.56
0.61
0.00
0.03
0.06
0.09
0.00
Delta %
(Core-
Surface)
12.04
3.08
7.50
9.89
4.20
1.85
35.87
-1.54
-0.19
-0.05
13.24
0.21
-0.02
-0.03
4.75
0.12
766
-------
in the fall, however, exhibit a more equitable distribution of radionu-
clide mass throughout the reservoir. This suggests that the relatively
greater Susquehanna River flow during the spring may scour and remove
radionuclide-labeled sediments from channel and eastern shore stations
within the Conowingo Pond, and that during the fall, lower river flows
allow for greater cross-sectional migration of particulates and sedi-
ments within the reservoir.
There is little question that the Conowingo Pond serves as an efficient
trap and deposition area for particulates suspended in the Susquehanna
River. Radionuclide profiles in our core data indicate that average
sediment accumulation rates range from about 3 cm/yr to greater than 6
cm/yr at the sampled locations (McLean et al., unpublished). Gross et
al. (1978) have estimated that one-half to two-thirds of the particulate
load of the lower Susquehanna is trapped behind the Conowingo Dam and
two upriver dams (Holtwood and Safe Harbor). Olsen et al. (1981) have
also shown accumulation rates to be high based on radiocesium (Cs-137)
profiles in a Conowingo Pond sediment core. However, although Cs-137 (a
weapon test fallout product) was observed at great core depths (>80 cm),
PBAPS-derived Cs-134 was confined to the upper 10 cm. This information
supports our data, further suggesting that, although Conowingo Pond
serves as a sediment trap — and does trap Peach Bottom radionuciides —
there is a turnover within the reservoir of surface sediments. Our data
indicate that radionuclide-labeled particulates deposited in the reser-
voir — with the exception of discrete areas in the downriver vicinity
of the PBAPS discharge — are not appreciably buried by subsequent sedi-
ment accumulation. Instead, they are continuously or periodically
eroded, resuspended, and transported beyond the Conowingo Dam. The
concurrent dilution and replacement of PBAPS-labeled sediments by par-
ticulates transported from upriver sources reasonably accounts for the
relatively high rates of sediment accumulation observed by us and others
in Conowingo Pond.
Radionuciides discharged by Peach Bottom, which are transported beyond
the Conowingo Dam, are found in sediments on the Susquehanna Flats and
in the Upper Chesapeake Bay (Tablet ). Because sediment collection
locations were selected primarily to define the extent of down-bay
detectabllity of PBAPS radionuciides, they are not considered to repre-
sent surface sediments beyond the immediate sampling vicinity. There-
fore, mass balance estimates of PBAPS-derived radioactivity in sediments
were not calculated for this area. Outside the Conowingo Pond region,
Peach Bottom radioactivity in sediments is most frequently detected and
found at highest concentrations on the Susquehanna Flats. Radionuclide
concentrations diminish with distance down-bay and become generally
undetectable south of the Sassafras River. These diminishing concentra-
tions are a function of particle dilution from other Upper Bay tribu-
taries, and likely as well, desorption of particle bound radioactivity
with increasing salinity (Olsen et al., in prep.).
The information derived from mass balance estimates of radionuciides
discharged into the Conowingo Pond suggests, that given average annual
river flows similar to those experienced during our study, less than 20%
of particle-bound pollutants introduced into the Susquehanna River are
deposited within the reservoir. This percentage is probably conserva-
tive (i.e., maximal) as it involves a radionuclide source situated on
767
-------
Table 3. Mean, maximum, and minimum concentrations (piCu/kg) and number
(N) of detectable concentrations (of 12 collections) for radionuclldes
in sediments collected from Susquehanna Flats and Upper Chesapeake Bay
locations, 1981-1986.
Location
SF-1
SF-2
SF-3
SF-6
SF-7
SF-8
SF-9
UB-10
UB-11
UB-11A
UB-11B
UB-12
UB-13
UB-14
Co-60
N Mean Min Max
7 19 5 46
2 10 8 12
2434
2426
1444
5 7 3 10
8 8 2 15
2546
4 16 2 34
4 18 11 30
3 23 13 31
6 37 21 45
0000
0000
Zn-65
N Mean Min Max
4 27 4 81
2 23 11 35
3 83 13 192
0000
0000
2 18 4 42
1 15 15 15
1 12 12 12
0000
1 16 16 16
0000
2 26 7 46
0000
0000
Cs-134
N Mean Min Max
11 34 6 93
11 18 3 37
8 14 3 52
6 14 7 27
7 12 3 24
9 23 9 37
12 40 13 85
9 29 10 61
8 25 11 39
6 12 5 33
4 31 14 59
7 22 6 43
0000
1444
168
-------
the reservoir which discharges almost daily; a fact which would likely
enhance local deposition. Most of the pollutant-labeled particulates
which are deposited within the reservoir, are likely resuspended and
transported beyond the Conowingo Dam. A small fraction of deposited
pollutants may ultimately become buried within the sediment record of
the reservoir.
REFERENCES
Gross, M.G., Karweit, M., Cronin, W.B., and Schubel, J.R. 1978. Sus-
pended sediment discharge of the Susquehanna River to Northern Chesa-
peake Bay, 1966-1976. Estuaries, 1:106-110.
McLean, R.I., and Domotor, S.L. 1988. Environmental radionuclide
concentrations in the vicinity of the Peach Bottom Atomic Power Plant,
1981-1984. Maryland Power Plant Research Program PPRP-R-9, Annapolis,
Maryland.
Olsen, C.R., Larsen, I.L., Cutshall, N.H., Donoghue, J.F., Bricker,
0. P., and Simpson, H.J. 1981. Reactor-released radionuclides in Sus-
quehanna River sediments. Nature, 294:5838; 242-244.
Schubel, J.R. 1972. Distribution and transportation of suspended
sediments in Upper Chesapeake Bay. In; The Environmental Framework of
Coastal Plain Estuaries, B.W. Nelson, Editor. Geol. Am. Memoir No. 133;
151-167.
769
-------
Understanding the Estuary: Advances in Chesapeake
Bay Research. Proceedings of aConference. 29-31
March 1988. Baltimore, Maryland. Chesapeake Research
Consortium Publication 129. CBP/TRS 24/88.
Quantifying Pollutant Sources to Rock Creek Estuary: The Patapsco River,
Sediment Remineralization, and Non-point Source Runoff
Roger S. Copp and Robert M. Summers
Dames and Moore
7101 Wisconsin Avenue, Suite 700
Bethesda, Maryland 20814-4870
INTRDDUCnCW
Rock Creek is a 353 hectare tidal creek in northern Anne Arundel
County, Maryland that is tributary to the Patapsco River. The mean
depth of the estuary is 3.0 meters. The length of the estuary is 5
kilometers, and the mean low tide volume is 9.98 million cubic meters.
The watershed drainage area is 1022 hectares, of which 51.6% is
urbanized, 5.4% is agricultural land, and 43% is forested. Rock Creek
has serious water quality problems including wide fluctuations in
dissolved oxygen, fish kills, and noticeable odors from hydrogen
sulfide emissions, and dense algae blooms.
Water quality investigations were conducted in 1987 and 1988 by
Anne Arundel County and the Maryland Department of the Environment
(MDE). The purpose of the studies was to identify the major causes of
the observed water quality problems and to formulate feasible
management strategies to improve water quality. This paper will
describe the method of -determining the major causes of the water
quality problems and quantify their relative impact on the Creek.
WATER QUAKETY STUDIES
Field studies were conducted in 1987 to provide data for a
nutrient mass balance for the estuary. Anne Arundel County funded
studies included estuary water quality sampling on eight dates at five
stations, measurements of organic sediment thickness in headwater
areas, sediment chemistry, stormwater runoff monitoring, hydrodynamic
investigations utilizing current meters, and in-situ measurements of
sediment oxygen demand and nutrient remineralization. MDE field
studies included estuarine water quality sampling on 15 dates at 16
stations, sediment thickness measurements, sedimentation rate studies,
benthic community surveys, and hydrodynamic investigations utilizing
dye studies. Figure 1 presents the estuarine water quality monitoring
stations and the tidal current meter location. Further details
regarding the methods utilized in the water quality studies and the
results are provided in Dames & Moore (1988) and MDE (1987).
Highlights of the study results are provided below.
Several key study results indicate the severity of the water
quality problems in Rock Creek estuary. Dissolved oxygen (DO)
concentrations throughout the study area were measured in the early
morning and late afternoon by MDE. A sample plot of this data is
170
-------
-------
presented in Figure 2. Morning surface DO concentrations witluLn Rock
Creek on June 30 averaged 4.16 mg/1 (s = 2.2) while afternoon values
averaged 11.32 mg/1 (s = 3.03). Morning and afternoon DO
concentrations and sectional estuary volumes were used to 'Compute
average daytime production rates. The average change in DO mass from
the morning to the afternoon for Rock Creek inside Fairview Point
(stations 1 to 11 shown in Figure 1) was 10,140 kilograms (Kg) or 0.35
mg/l/hr.
Chlorophyll a concentrations were also highly variable.
Chlorophyll a data for August 25 are presented in Figure 3. On the
plot the values for stations 2 and 3 are truncated. The laboratory
results indicated concentrations in excess of 800 ug/1 for these
stations and no quality assurance problems were noted for the samples.
The values were truncated in order to provide sufficient resolution on
the plot.
Phosphorus and nitrogen concentrations are also interesting as
shown in Figure 4. These concentrations are volume weighted average
concentrations for Rock Creek inside Fairview Point. Orthophosphate
concentrations are usually close to detection limits, while total
phosphorus concentrations are often in excess of 0.2 mg/1. Dissolved
nitrogen forms are also close to detection limits in June and July
while total nitrogen concentrations during the summer exceed 2 mg/1,
with concentrations in excess of 3 mg/1 common. Peak total
concentrations were observed in late July and August. Hydrogen sulfide
odors were severe on June 22 and August 3. The importance of the
relationship between odors and nutrient concentrations will be
discussed later in this paper.
Sediment oxygen demand (SOD) and nutrient remineralization was
measured near station 3 and near station 9 (see Figure 1). Duplicate
measurements were made at station 3A on three days and at station 9A on
one day. The duration of the experiments ranged from 4 to 6 hours. The
results are presented in Table 1. Methane production rates were
consistently high with somewhat lower rates when overlying waters were
anaerobic. The average C:N:P molar ratio for releases of methane,
ammonium, and orthophosphate for station 3A was 301:30:1. The methane
production deviates most from the Redfield ratio of 106:16:1. Nitrogen
remineralization is also elevated relative to phosphorus. The
deposition of iron from the SOD chambers may explain this relative lack
of phosphorus remineralization.
Sulfide production was expected during the experiments, however
all samples analyzed were below the detection limit of 0.066 mg/1. The
resulting sulfide release rate was therefore less than 6.73
mmoles/m2/day. This result in itself is significant. Bacterial
production of methane was much higher than sulfate product ion, which
suggests that the sediments were depleted of available sulfate and that
methane production was the dominant bacterial reduction process. Low
sulfate production rates have been reported when high methane
production rates were measured (Martens and KLump, 1984).
DESCRIPTION OF THE SEASONAL MASS BALANCE
In order to evaluate the relative importance of the principal
nutrient sources to Rock Creek, a seasonal mass balance was prepared
for the period June 17 to August 25, 1987. This period was selected
because tidal current meters were in place for the period June 19 to
172
-------
FIGURE 2. ROCK CK. DISSOLVED OXYGEN
DATE-30JUN87 WHEN-AM
20"
0
|lt1
0 16'
L
V 141
E
0 12-
Y
G 8-
E
N 61
M
G
21
01
D
3 4 5 6 7 8 9 10 11 12 13 14 15 16
STATION NUMBER
LAYER D O D B -•—<—»• S
ROCK CK. DISSOLVED OXYGEN
DATE-30JUN87 WHEN-PM
o201
J..1
Si.
L
V 141
E
0 12
I .
E
N 6
M
G
a
a
-I—.—|—i—|—i—|—i—|—i—|—i—|—i—|—i—|—i—i i | i | < r~
3 4 5 6 7 8 9 10 11 12 13 14 15
STATION NUMBER
LAYER a Q D B H—'—* S
j)
77?
-------
hj
ffi
O
o s
"I N
I—I II
H-r-l UJ
^ 5
o s
o
o
o
CO
UJ
cc
D
2
\L
n> ^J
cc +
Ul
03 +
h ON
z
o
CD
D
D
D
cc
LJ
I- CO
h- CM
O
O X -J
774
-------
FIGURE 4. AVERAGE NITROGEN AND PHOSPHORUS CONCENTRATIONS IN ROCK CREEK
u
i
s
o
4.00
3.50 -
3.00 -
2.50 -
2.00 -
1.50 -
1.00 -
0.50 -
ROCK CREEK
0.00
I I I I I I
13-Apr 03-Moy 23-Moy 12-Jun 02-Jul 22-Jul 11-Aug 31-Aug 20-S«p 10-Oct
NH4
DAY/MONTH, 1987
+ N023 « DIN
A TN
8
W
0.40
ROCK CREEK
0.05 -
0.00
13-Apr 03-Moy 23-Moy 12-Jun 02-Jul 22-Jul 11-Aug 31-Aug 20-S«p 10-Oct
DAY/MONTH, 1987
D PO4 + TOP O TP
775
-------
i
in
1
|
t-i
|
§
!
2
1
u
V
1
OT
M£
!
cu
6
ll
2
"5
o£
1 -x
I
^
TJ
•a
^
S a
•w
I
Station
4
2
£
cn en 10 r-<
m cn •"* rH t^ CO
CO ^ CM CM
•H C*« O) iH CJ> CO
!•" CO ^ P s \
Cn & 0> 0> CH 0>
r-t
8
CM
0>
CM
S
CO
f-»
in
CM
CM
CO
S
o
cl
h.
1
S
(0
CM
1
CO
0
<0
8
o
rH
ll
9/4/87
rH
O
1
y
91
S
r-4
cn
10
rH
O
CM
ol
1-
00
X.
•*
^
en
J J
776
-------
August 21 and nine water quality surveys were conducted between June 17
and August 25. Mass balances for total phosphorus (TP) and total
nitrogen (IN) were developed that explicitly considered nutrient inputs
fron Patapsco River inflows, sediment remineralization, stontiwater
runoff, septic systems, and atmospheric deposition. Daily inputs from
these sources were summed, outputs due to settling algae and Patapsco
River outflows were computed, and the net masses of TN and TP in Rock
Creek inside Fairview Point were compared to nutrient masses computed
from monitoring data.
Components of the mass balance are discussed in more detail below:
Patgpsco River
Hydrodynamic forces cause a small but consistent bottom inflow
(average velocity = 0.9 cm/sec) from the Patapsco River to Rock Creek.
The confluence of Rock Creek and the Patapsco is constricted by
Fairview Point and this constriction provided a convenient point to
monitor the velocity of the inflows. The velocity, temperature, and
salinity were measured every two minutes for approximately two months
utilizing two Endeco 174SS meters placed 1.2 meters above the bottom
and 1.2 meters below the surface. Total water depth was 5 meters.
This constriction was used as the seaward boundary of Rock Creek. The
width is 427 meters and the mean low tide cross section area is 1,211
m2. Based on salinity data collected from field surveys and from the
current meters, it was assumed that the bottom cross sectional area was
594 m2. Average daily bottom tidal velocity was multiplied by the
cross sectional area and observed bottom concentrations of TN and TP to
compute the nutrient input from the Patapsco to Rock Creek.
The outflow from Rock Creek to the Patapsco was calculated as the
sum of the hydraulic inputs minus evaporation. The nutrient output was
calculated utilizing measured concentrations of nutrients in surface
waters near Fairview Point (Station 13 shown in Figure 1).
In addition, the daily change in tides was considered in the mass
balance. If the tide level at the end of the day was higher than at
the beginning of the day, there was a net daily inflow from the surface
layer. Similarly, if the tide level was lower at the end of the day,
there was a net outflow of nutrients from the estuary. The surface
area at different tide elevations was measured, and the tide
inflow/outflow was computed as the average of the top and bottom areas
multiplied by the change in tide height.
S^diTrent Remineralization
The phosphorus and nitrogen release rates were assumed to be
constant throughout the study period. The selected phosphorus and
nitrogen release rates were 35 and 490 mg/m2/day, respectively. These
rates were selected based on remineralization rates measured in Rock
Creek in September, 1987. Studies have shown that summer
remineralization rates may be twice as high as fall rates (Boynton et
al, 1986), therefore these measured rates of TN and TP remineralization
may be considered as a lower bounds estimate. A reasonable upper bound
for sediment remineralization could be as high as 70 and 750 mg/m2/day
for phosphorus and nitrogen respectively. These upper and lower bounds
for sediment remineralization were tested in the mass balance to
determine the possible range of influence of sediment remineralization
on the supply of nutrients fueling algal growth.
177
-------
Stormwater Runoff
In order to estimate the runoff contribution to the nutrient
loading of Rock Creek during the study period it was necessary to
estimate the runoff volume and water quality. Stormwater runoff
monitoring was conducted in the fall of 1987. Samples were collected
from two stations in the watershed. The land use at one station was
56% residential, 27% forest, 3% commercial, and 14% construction, while
the other watershed was 100% residential. Two storms were monitored in
the mixed land use watershed while one storm was monitored in the
residential watershed. The volume weighted average TP concentration
for all monitored events was approximately 0.5 mg/1. The IN
concentration was more variable, ranging from 1.6 to 14 mg/1. The
median TP and TN concentrations (0.7 and 2.9 mg/1, respectively) for
these monitored storms were used to represent Stormwater quality in the
mass balance analysis.
The summer runoff volumes into Rock Creek were estimated by
comparing runoff volumes for monitored storms with runoff volumes for
USGS gaging stations in nearby watersheds. The volumes of runoff for
the two storms monitored in the mixed land use watershed were very
similar to runoff volumes at the USGS gaging station at Bacon Ridge
Branch at Chesterfield. Accordingly, runoff volumes (in centimeters)
for the Bacon Ridge Branch gage for the summer of 1987 were used to
estimate runoff to the Rock Creek watershed.
Septic Systems
The number of houses in the watershed was estimated utilizing Anne
Arundel County topographic maps and 1985 aerial photographs. Estimates
of per capita TP and TN loads were obtained from Metcalf and Eddy
(1979) and it was assumed that there were 2.3 people per household.
Further, it was assumed that all houses with in 152 meters of the
estuary or tributary streams contributed nitrogen and phosphorus to the
estuary. From these houses, it was assumed that all nitrogen and 50%
of the phosphorus reached the estuary. The assumption regarding
nitrogen is well supported in the literature (Canter and Knox, 1986;
Sikora and Corey, 1976). Most references regarding phosphorus uptake
in soils report higher removal rates than assumed in this mass balance
(Sikora and Corey, 1976; Hansel and Machmeier, 1980). Accordingly, the
assumption regarding phosphorus should be considered as an upper bounds
estimate.
Atmospheric Deposition
Rainfall nutrient inputs were computed by utilizing rainfall data
from the National Weather Service Baltimore Washington International
Airport weather station and assumed TN and TP concentrations of 1.5 and
0.035 mg/1, respectively.
178
-------
Settling
Settling rates of IN and TP for the mass balance were adjusted to
provide a reasonable fit to the observed nutrient mass measured in Rock
Creek. A realistic lower bound for the settling rate was selected
partially based on settling trap data obtained from station R-64 in the
Chesapeake Bay (Boynton et al, 1986)and partially based on deposition
rates that would be necessary to support methane production rates
observed in Rock Creek.
MASS BALANCE RESULTS
Starting with a known mass (based on measured concentrations from
water quality surveys) of total nitrogen and phosphorus in the water
column of Rock Creek on June 17, 1987, the various nutrient inputs and
outputs discussed above were summed on a daily time step to obtain a
value for the change in nutrient mass expected for the subsequent
survey dates. With the settling loss term set to average TN and TP
values of 26 and 1.2 mmole/m2/day, the predicted mass of nutrient in
the water column is much higher than observed during the early
summer(Figure 5). Since there was no reason to expect the settling
rate in Rock Creek would be constant, settling was varied on a daily
basis as a calibration parameter in order to match the predicted mass
of nutrients in Rock Creek to the observed mass.
The following observations made in Rock Creek provide some
justification for expecting very high sedimentation rates in the Creek.
1) There was high variability in phytoplankton populations during
the summer of 1988. On monitoring dates June 17, June 24, June 30,
July 7, July 22, August 3, August 5, and August 26, phytoplankton
varied between dinoflaggelates, blue green algae, and diatoms. From
June 24 to August 3, a different phylum was dominant on each monitoring
date. Dinoflagellates again dominated in August. The rapid changes in
phytoplankton dominance, and the disappearance of the previously
dominant species during late June and July support the hypothesis of
higher sedimentation during this period.
2) Measured sediment production of methane was very high. The
carbon deposition rate necessary to provide organic carbon for the
methane production would be in the range of 600 mmoles/m2/day.
Assuming the Redfield C:N:P ratio of 106:16:1, the TN and TP deposition
rate would be approximately 91 and 5.7 mmoles/m2/day (1365 and 175
mg/m2/day), respectively. The maximum TN and TP settling rates
necessary in the mass balance to provide a close match between the
observed and predicted nutrient masses were 75 and 2.2 mmoles/m2/day,
respectively. Therefore, the methane production rates from Rock Creek
sediments are consistent with the maximum release rates necessary to
calibrate the mass balance.
3) The maximum settling rates required in order to obtain a mass
balance with variable settling are much higher than maximum settling
obtained at station R-64 of the Chesapeake Bay by Boynton et al (1986).
The maximum TN and TP settling rates observed at R-64 were 17 and 0.81
mmole/m2/day. The mass balance derived settling rate for TP in Rock
Creek is not significantly higher than the R-64 rate, however the TN
settling rate in Rock Creek is much higher than the R-64 rate. This is
779
-------
FIGURES. Mass Balance Results for Phosphorus
1.1
+ Const. Settling
200
Time, Day»
240
o Vor. S«ttnng
180
-------
not unexpected since the average ammonium remineralization rate of 30
mmole/m2/day measured from Rock Creek sediments was more than double
the highest rates (14 mmole/m2/day) reported by Boynton et al. (1986)
for the Chesapeake Bay and other estuarine systems.
The resulting phosphorus and nitrogen mass balances are summarized
in Table 2. It is evident from this analysis that the major sources of
both nutrients are the Patapsco River and sediment remineralization.
All other sources are quite minor relative to these two sources. As
discussed earlier, the septic system loads and the stormwater
phosphorus loads represent upper bounds estimates. The volume of runoff
could be doubled and the phosphorus load would still remain a small
source. The nitrogen concentration and the volume of runoff could be
increased with negligible effects upon the mass balance.
Since there was some uncertainty concerning the exact magnitude
of the nutrient exchange with the Patapsco, a range of exchange rates
was tested for a reasonable upper and lower bound. The cross sectional
area of the inflowing bottom layer of the Patapsco River was decreased
to test the sensitivity of the mass balance to the Patapsco River
influence. This change reduced the Patapsco River nutrient load from
58% to 48% of the total load.
The other major nutrient source is sediment remineralization. As
mentioned above, there is reason to expect that summer rates, under
higher temperatures and presumably higher sedimentation, could be as
much as two times the rates measured in September. In order to
evaluate this question, sediment remineralization rates were increased,
and the impact on the mass balance was evaluated. This resulted in an
increase from 41% to 60% for TP and 48% to 59% for IN for the portion
of the nutrient inputs attributable to sediment remineralization.
The magnitude of the nutrient inputs to Rock Creek are high. The
overall TP load during the summer period was estimated to be 6058 Kg or
5.6 g/m2. Jaworski (1981) has concluded that eutrophic conditions can
be prevented in an estuary if the annual phosphorus load can be
maintained below 1.0 g/m2/yr. A similar situation exists for nitrogen.
The summer load to Rock Creek was estimated to be 74264 Kg or 68.7
g/m2. The acceptable nitrogen load according to Jaworski (1981) is 5.4
g/m2/yr. The phosphorus and nitrogen loads are 6 times and 13 times
their respective acceptable loads.
SUMMARY
The mass balance approach described in this paper was a useful
tool for determining the magnitude and relative importance of nutrient
inputs to Rock Creek estuary from the Patapsco River, sediment nutrient
remineralization, non-point source pollution, and other sources. The
most important sources were the Patapsco River and sediment nutrient
remineralization. A number of questions were raised from this
investigation which can only be resolved by further study of this
highly eutrophic system.
1) The settling rates of organic material in Rock Creek are
essentially unknown. The mass balance analysis suggests that
settling rates are extremely high. Settling rate measurements in
Rock Creek are necessary before these values can be verified.
2) The observed methane release rates are extremely high compared to
any other measurements reported in the literature. More
757
-------
measurements are needed to determine if these rates are generally
representative of methane release in Rock Creek.
3) Sediment nutrient concentrations are not unusually high in Rock
Creek. This suggests that benthic-pelagic coupling must have a
very rapid turnover rate in order to support the observed nutrient
remineralization rates. Additional nutrient remineralization
measurements would be helpful in verifying this assumption.
TABLE 2
Summary of Seasonal Mass Balance for Rock Creek
June 17 to August 25, 1987
Phosphorus Nitrogen
Source Kfr % Kg %
Patapsco River 3,511 57.9 41,886 56.4
Daily tide flux -324 -5.3 -5,954 -8.0
Sediment reminerali-
zation 2,572 42.4 36,004 48.5
Stonnwater runoff 108 1.8 414 0.6
Base flow 5 0.1 250 0.3
Septic systems 180 3.0 1,353 1.8
Rainfall 6 0.1 311 0.4
Total 6,058 100 74,264 100
182
-------
REFERENCES
Dames & Moore, 1988. The Rock Creek Estuary Study. Revised Draft,
April 20, 1988, Bethesda, MD.
Boynton, W.R., et al., 1986. Ecosystem Processes Component Lsvel 1
Data Report No. 3r July 1985-May 1986, CBL No. 86-56.
Canter, L.W., and R.C. Knox, 1986. Septic Tank System Effects on
Ground Water Quality. Lewis Publishers, Inc.
Hansel, M.J., and R.E. Machmeier, 1980. "Qn-Site Wastewater
Treatment on Problem Soils," Journal of Water Pollution Control
Federation. Vol. 52, No. 3.
Jaworski, N.A., 1981. "Sources of Nutrients and the Scale of
Eutrophication Problems in Estuaries," Estuaries and Nutrients. Human
Press.
Metcalf and Eddy, 1973. Wastewater Engineering. 2nd edition,
McGraw-Hill.
Sikora, L.J., and R.B. Corey, 1976. "Fate of Nitrogen and Phosophorus
in Soils Under Septic Tank Waste Disposal Fields" Transactions of
American Society of Agricultural Engineers, Vol. 19, No. 5.
183
-------
Understanding the Estuary: Advances in Chesapeake
Bay Research. Proceedings of a Conference. 29-31
March 1988. Baltimore, Maryland. Chesapeake Research
Consortium Publication 129. CBP/TRS 24188.
Nutrient Regeneration Rates in Chesapeake Bay
Bottom Sediments Based on Bulk Properties
James M. Hill and Jeffrey Halka
Maryland Geological Survey
2300 Saint Paul Street
Baltimore, Maryland 21218
INTRODUCTION
The bottom sediments in any estuarine system constitute the largest
reservoir of trace elements and nutrients. It is therefore imperative
that the interaction between the bottom sediments and the water column
be understood. Currently fluxes from the sediment are measured by
either pore water profiles, dome studies, or the incubation of
sediments. Although these methods provide the best information on the
processes occurring in the sediment, they are costly and time-consuming,
thus are limited spatially and temporally. As a complement to these
measurements, an upper limit average yearly rate of nutrient regenera-
tion can be estimated. These estimates are independent of fluid
analyses and are based solely on physical properties and composition of
the sediments, coupled with sedimentation rates. Because these analyses
are relatively simple and inexpensive wide spatial coverage can be
obtained. Such a data set exists for the Chesapeake Bay.
The method is based on a steady-state approximation of the
sedimentary processes occurring, and can be written as follows:
dC
dt • 0 - Fin - (Pout + ^buried) O)
where: C - the concentration of either Carbon, Nitrogen, or
Phosphorus species
Fin ~ tne flux of nutrient, as a component of organic
matter, into the sediment
Fout ~ tne flux of nutrient out of the sediment, equal to
the rate of regeneration
^buried " tne flux °f organic material lost from the system
due to burial of non-reactive organic matter
184
-------
Equation (1) can be rewritten as:
Fin " fout * Fburied
or- " Fin (2)
where: fr and fn are the reactive and non-reactive fraction of the
input organic carbon. The reactive component input equals
the rate of nutrient regeneration (Fout) and the
non-reactive component equals the rate of burial of
non-reactive organic matter (Fburied).
If it is assumed that the sediment at a site in the Bay is
characteristic of the material input to the sedimentary reservoir at
that site, then the rate of nutrient regeneration can be written as:
Fout " «Ps fr C0 (3)
where: w - the sedimentation rate (cm/yr)
Ps - the density of solids in the sediments (g/cm^)
fr " the reactive fraction of the organic carbon
C0 - the initially deposited reactive carbon
Evaluation of this equation, using existing'data, requires several
conditions to be satisfied. These are:
1. The sediment type and porosity are .constant at a given
location, within the shallow sediment reservoir where the
reactive carbon is depleted;
2. The average sedimentation rate has been constant, at any site,
within the shallow sediment reservoir;
3. The amount of initially deposited carbon can be determined;
1. The properties of the organic matter are relatively uniform
throughout the main stem of the Bay., i.e. the reactive and
non-reactive components of C0 are the same everywhere,
5. Redfield's ratio (Redfield et al., 1966) can be applied to
determine Nitrogen and Phosphorus fluxes, and;
6. Nitrification/denitrification and Phosphorus mineralization are
not accounted for - thus the model yields the total potential
flux of Nitrogen and Phosphorus.
these conditions above will be elaborated upon in the following text.
SEDIMENT DISTRIBUTION
Data on the bulk properties of the bottom sediments in the mainstem
Chesapeake Bay were derived from the work of Byrne and others (1982) and
Kerhin and others (1988). Sample stations were located on a one
kilometer square grid In Maryland waters and a 1.1 kilometer offset grid
in Virginia, a total of 5921 samples. Figure 1 shows the sample
locations and segments used in this paper's analysis. At each sampling
location the top 5 centimeters of sediment were collected with a grab
sampler and analyzed for grain size, bulk density and percent water.
Approximately one third of these samples were analyzed for carbon and
sulfur content. Details of the analytic methodology may be found in the
aforementioned reports. A sediment textural distribution map of the
mainstem Bay, based upon these data and utilizing the ternary diagram
and classification scheme of Shepard (1951), is presented in Figure 2.
As shown in Figure 2, sands, represented by the lower left corner
of the ternary diagram, cover a much larger portion of the mainstem Bay
than has been generally recognized. Over half of the bottom is composed
185
-------
SAMPLING LOCATIONS
V* ^'^^
*>• ">*:••
jT'Av <"'*'•''"» /
*
^ /
'
Figure 1: Location of surficial
sediment samples used for the
determination of bulk properties,
adapted from Byrne and others (1982)
and Kerhin and others (1983). Also
shown are the physiographic segments
within which sedimentation rates
were determined.
186
-------
SA.VDY \ SILTY
CLAY /\ CLAY
Figure 2: Textures of surficial
sediments based upon the
classification scheme of Shepard
(1954). Adapted from Byrne and
others (1982) and Kerhin and others
(1983) .
757
-------
of sediments in which sand sized particles constitute more than 15% of
the grains by weight (Table III). The area represented by the lower
left three fields of the ternary diagram; the sand, silty sands, and
clayey sands in which sand sized particles comprise more than half of
the grains; covers 66$ of the mainstem Bay. The importance of this
large proportion of sand will be dealt with in the discussion section.
The bottom sediments in the deeper basins and axial channels
largely consist of finer grained silts and clays (Figure 2). Throughout
the Maryland portion of the Bay (segments 1-7) silty clays predominate
with slightly coarser clayey silts occurring proximal to the Susquehanna
River mouth, in segment 1. The finest grained clays are present only in
the upper central portion of the Bay (segment 4). The nearly continuous
field of silty clays extends to the south of the confluence with the
Potomac River into Virginia where they grade into coarser clayey silts
northeast of the Rappahannock River (segment 8). South of the
Rappahannock River sands and silty sands dominate the bottom sediments
even in the deepest waters (segments 9-12).
Within the dominant silty clay field extending between the Bay
Bridge and Rappahannock River (segments 3-8) a wide variety of sediment
types occurs in the deeper portions of the Bay (Figure 2). These
include nearly all of the ten classes represented in the ternary diagram
and have little if any relationship with water depth. Of particular
note are the isolated pockets of sands and clayey sands which are
located between the Patuxent and Potomac Rivers (segment 6). These
pockets occur in over 12 meters of water depth and are separated from
nearshore sources. The method of estimating potential nutrient fluxes
describe herein incorporates these variations in textural distribution.
From the sediment data the density of solids at each sampling
station was derived from the percent water using the formulas similar to
those of Bennett and Lambert (1971). In addition the weight percent
clay at each station was used to derive the originally deposited carbon
content, and the weight percent of sand was used to modulate the
sedimentation rates (see the following sections).
SEDIMENTATION RATES
Sedimentation rates for the mainstem of the Chesapeake Bay were
determined by the method of bathymetric comparisons, with the initial
data derived from Byrne and others (1982) and Kerhin and others (1988).
Using the original NOAA survey sheets the water depths within cells 6
seconds on a side (a150 by 200 meters) were averaged for the earliest
(circa 1850) and the most recent (circa 1950) surveys for which there
was an adequate density of data points. The results denote the changes
in the height of the water column over the time interval spanning the
two surveys. The data were rectified to the same mean low water datum by
applying correction factors for eustatic sea-level rise (1 mm/yr) and
the estimate of recent crustal warping for the Bay region. The latter
ranged from over 2.6 mm/yr subsidence in the north to 1.0 mm/yr near the
Bay mouth (Holdahl and Morrison, 1974).
The bathymetric comparison technique was utilized to estimate
sedimentation rates because of its ability to provide measurements for
the sandy sediments which have been shown to cover a large proportion of
the Bay floor. Due to the nature of the data available from Byrne and
188
-------
others (1982) and Kerhin and others (1988) the Bay was divided into 12
segments based upon basin geomorphology (Figure 1), with the averaged
sedimentation rate determined within each segment. In Maryland the
average rate of accumulation per year was calculated for both the muddy
and sandy sediments within each segment, in Virginia only an average
sedimentation rate could be calculated for each segment due to the
characteristics of the original data set. It was felt that the
bathymetrically determined rates provided a better estimate of the
sedimentation rate than radiometrically determined rates'due to the
large areal expanse of sandy sediments, especially in the Virginia
mainstern Bay.
The sedimentation rates for each of the geomorphic segments are
shown in Figure 3 with rates for both muddy and sandy sediments
determined in the Maryland segments and a single rate in the Virginia
segments. The highest rate of nearly 0.8 cm/yr occurs in muddy
sediments of the northern Bay adjacent to the Susquehanna River mouth.
The rates decline southward from this high value to a minimum
(approximately zero) in the upper middle bay.(segment H). Continuing
south, rates rise again for both the muddy and sandy sediments to maxima
of between 0.5 and 0 .6 cm/yr in segments 7, 8, and 10. It is
interesting to note that the shape of the curve is similar to that
reported from radionuclide dating (Officer et al., 198*0 with the
minimum bathymetrically determined rate occurring approximately 60 km
further upbay than the radiometrically determined minimum. The rate for
each segment was input to the model equation for each of the bottom
sampling stations. In Maryland waters where two end member rates were
determined, sands and muds (as shown in Figure 2), the rate utilized in
the model equation at each of the stations was calculated using a linear
approximation based on the percent sand present at each location.
INITIALLY DEPOSITED CARBON (CQ)
One of the most difficult parameters of prime importance to
determine in estimating nutrient regeneration from bulk sediment data is
the concentration of initially deposited carbon (Co). In the literature
to date there has been no independent determination of this parameter.
Generally when examining sediment cores uniform deposition is assumed
and C0 is assigned the value of the carbon content at the sediment-water
interface; carbon values down-hole are subtracted from this assigned C0
to provide the amount of carbon oxidized by bacterial action, thus the
amount of nutrient regeneration (Berner, 1980; Martens et al., 1978).
This method cannot be readily applied to the estuarine sediments of the
Bay because virtually all of the carbon data is from surficial samples
(i.e. the interval 0-5 cm; no C0 can be assigned at the sediment-water
interface), secondly much of the deposition in the Bay is episodic, not
uniform, and finally, there is a significant terrigenous (coal)
component of the carbon content in the northernmost portion of the Bay.
An alternate approach to determine C0 uses the strong relationship of
carbon to the clay-sized particle content of the sediment (Hennessee et
al., 1986; Hobbs, 1983) coupled with a modification of Berner's (1972)
and Sweeney's (1972) approach to describing the relationship of Sulfur
to Carbon in marine sediments.
189
-------
1.0-
. 2 , 3 , 4
Segment Number
8 , 9 . 10, 11. 12 ,
**
(0
QC
0.4-
CO
**
c
9
g 0.2-
:5
«
(A
0.0-
300
250
I
200
I
150
100
50
I
0
Distance from Bay Mouth (km)
Figure 3: Average sedimentation rates (heavy lines) within
each of the geomorphic segments delineated in Figure 1, connected
by light lines to show trends. In segments 1-7 (Maryland)
rates for muddy sediments are shown by solid lines and for
sandy sediments by dashed lines. Only a single overall rate
could be calculated in segments 8-12 (Virginia). All rates
in cm/yr.
190
-------
Sulfur in the sediments of the Bay is found in the form of metal
sulfides formed during microbial sulfate reduction (Hennessee et al . ,
1986). This sulfur is deposited in the sediment at the expense of the
oxidation of organic carbon. A skeletal diagenetic reaction can be
written as follows:
- + 9CH20 + 4Fe(OH)3 •» HFeS + 9HC(>3~ + 10H2Q CO
Based on this reaction the amount of sulfur found in the sediment at any
given time is:
S(t) - ct(C0 - c(t)) (5)
where: S(t) - the bulk sulfur concentration measured at time t (g
S/g dry wt sed; all percentages are based on dry
sediment weight)
C0 - the initially deposited carbon
C(t) - the organic carbon content at time t
a - a constant containing the stoichiometric reaction
factor and the conversion of carbon mass to sulfur mass
This is the approach taken by Berner (1972) and Sweeney (1972). Two
major difficulties arise when trying to apply this equation. First, Co
must be independently defined. The second, and more important reason,
is that not all of the sulfur generated by sulfate reduction is bound in
the sediment. A significant fraction of the sulfide is recycled in the
sediment and not mineralized. Jorgensen (1977) showed that only 10J of
the sulfide generated was preserved in the sediments in a Norwegian
fjord. Since Jorgensen's work the recycling of sulfur has been well
documented (see Jorgensen, 1983).
Consequently, in order for Equation (5) to be validly applied to
sedimentary systems recycling must be taken into account. If at any
location in a sedimentary basin it is assumed the sediments are in
steady state, then the fraction of generated sulfide preserved is a
constant, designated as p (range 0-1). Equation (5) then becomes:
S(t) • o p (C0 - C(t)) (6)
or, C0 - C(t) + S(t)/ap (6a)
Within the Chesapeake Bay, south of the Maryland Bay Bridge
(latitude «39°05'), there is a strong correlation between the percent of
the clay-sized grain fraction and organic carbon; see Table I. It
follows that since C(t) is strongly related to percent clay, that Co is
also strongly related to percent clay. Co can be calculated using the
measured values of C(t) and S(t), the stoichiometric/conversion factor
o, and an average p of 0.52 (Hill, 1987; Hill, in prep). The results of
the linear regression of C0 as a function of percent clay are shown in
Table 2 for comparison with the C(t). versus percent clay fit. The
goodness of fit, in both Maryland and Virginia, is greater for C0 than
C(t), and is quite good, thus allowing the use of percent clay to
calculate Co in the mainstem of the Bay south of the Maryland Bay
Bridge. However there is some question as to the applicability of C0,
determined by the clay content, in the region effected by terrigenous
carbon input (segments 1 & 2) .
191
-------
In order to establish whether the percent clay to C0 relationship
is valid for segments 1 and 2, the amount of terrestrial carbon,
principally coal, can be calculated as follows:
Cter(o) - (C(t) + S(t)/op) - Co (clay) (7)
where: C0(ciay) is Co calculated by the relation in Table 2
p is approximately equal to one (Hill, 1987; Hill, in prep.)
This equation in theory subtracts out the planktonic component of the
organic carbon (C0(clay)), leaving only the terrestrial or coal
component. Figure ^ plots Cter(o) as a function of percent clay for
segments 1 & 2; where Cter(o)-0 (shown by the solid line) the organic
carbon in the sample is solely planktonic. There are a number of samples
that contain solely planktonic carbon, however most of the samples show
a linear decrease in carbon content with increasing clay content; the
lowest values, at high clay contents, are approximately equal to
Cnr(o)-0. This is the expected behavior if the non-reactive carbon was
coal dust which would be associated with the coarser size fractions.
Consequently, percent clay will be used throughout the mainstem of the
Bay to calculate C0,
REACTIVE CARBON COMPONENT OF CQ
The next component of Equation (3) to be evaluated is the fraction
of C0 which is reactive. This can be determined using Berner'a (1980)
multi-G model, applied to core data. Berner's model is based on the
idea that there are several component fractions of organic matter in
sedimentary environments. Each of the components decay at a different
rate, but each follow a first-order decay mode. This can be written as:
C(x) - JC1e-k1(x/w) (8)
where: C(x) - the total measured organic content at depth x
Cj - the initial concentration of the itn component
kj - the first-order rate constant of the 1th component
x - depth into the sediment
u - is the sedimentation rate
This equation can be rewritten to deal with fractional components (fA)
of C0:
C(x) - C0z(Ci/Co)«"ki(x/w) (9)
or - e-xu (9a)
However, this is not directly applicable to the Bay because Co and u>
vary with depth due to changes in sedimentation. Changes in C0 with
depth can be determined by clay content, however little data is
available on detailed down-hole sedimentation rates (Brush and Davis,
198M). The best data that exists use radiometric sedimentation rates
which assume constant sedimentation. Equation (9) becomes:
fT(x) . C(x) . Efie-ki(x/u) (10)
where: fj(x) - the total fractional amount of C0(x) at depth x, which
ranges in value from 1 @ x - 0 to fnr § x - «.
co(x) - C0(clay) at depth x
to - the sedimentation rate as determined by Pb^lu techniques.
192
-------
Table 1. Linear regression fits of C(t) and C0 as a function of percent
clay.
C(t) - m($clay) + b
Segments
3-7 (MD)
C(t)
co
8-12 (VA)
C(t)
N
518
518
956
m
0.0361
0.0601
0.0102
b
0.262
0.375
0.229
R2
81$
85$
70$
P
-
0.52
-
956
0.0662
0.320
76$
0.59
Table 2. Fractional components (fj) of Co and first order rate
constants calculated in the regression fit of Equation ( 11);
along with the sedimentation rates used, and the goodness of
fit.
Station
Number
55
62
63
61
83
85
'1
0.33
0.22
0.18
0.21
0.55
0.36
f?
0.36
0.11
0.28
0.28
0.20
0.29
f nr
0.32
0.39
0.26
0.51
0.21
0.39
fr
1.01
1 .02
1.02
1.00
0.99
1 .01
'r
(fi+f?)
0.69
0.63
0.76
0.19
0.70
0.65
076T±0.08
Station
Number
55
62
63
61
83
85
ki(yr'1)
0.076
0.170
0.081
0.923
0.010
0.103
K5(yr"1)
0.0071
0.0071
(-0.00021)
0.011
0.00096
0.00599
u
ui(cm/yr)
0.87
1.26
0.66
1 .12
0.19
0.37
R2($)
82
70
97
100
91
92
Sedimentation rates from Helz et al., 1982.
193
-------
7
6
5
4
Cter 3
2
1
n
\j
- 1
-2
i • i ' i • i • i
X
\
X ' '
X X.
Xp - X
N • • X
. . • . \ ts • . .. X
• ' -x.. •••-;•;:•:.:>,
- .X .. '. !,..'*N r
0 20 40 60 80
% Clay
Figure 4: Plot of terrigenous carbon versus percent clay.
The solid horizontal line represents planktonically derived
carbon with the dotted lines showing the spread of data
in segments 3-7. The linearly decreasing values of C
(encompassed by the heavy dashed lines) represent the coal
component of organic carbon in segments 1 and 2.
194
-------
Figure 5: Potential flux of organic
carbon out of the sediments (20
g/m2~yr contour interval).
195
-------
The parameter fx(x) is used In order to normalize for down-hole
variability, thus allowing the computation for fif tne various component
fractions of the organic carbon.
Equation (10) has been applied to six cores taken throughout the
mainstem of the Bay for which grain size, carbon, sulfur and Pb210
sedimentation rates were measured (MGS unpubl. data; Helz et al., 1982).
A Marquardt-style regression (Draper and Smith, 1966) was used to fit
the data to a function in the form:
fl(x) = fie-ki + f2e-k2(x/w) + fnr (n)
where: x, w, and fT(x) are measured parameters
The reactive carbon component, fr, is the sum of f-\+f2; the average
value of fr is 0.65±0.08. This value is within the range observed by
others; Nixon (1981) noted in coastal marine waters 25-50% of the
organic matter is fixed in the sediments (compared to 35% in this
model). The variation in fr is quite small (RSD = 12$) indicating that
the planktonically derived organic carbon is relatively uniform
throughout the Bay. This is to be expected due to the uniform behavior
of C0 to clay throughout the Bay.
DISCUSSION/RESULTS
The results of the model are given in Figure 5 and Table 3. Figure
5 is a flux map of the mainstem of the Bay, Fout; in contours of 20 g of
carbon/m2-yr. The flux map mimics the sediment distribution (with the
higher fluxes corresponding with the higher clay contents) but is
strongly modulated by sedimentation rates. The highest fluxes occur
where clay contents and sedimentation rates are both high. This is the
case north of the Bay Bridge (segments 1 & 2) where the potential carbon
flux locally exceeds 80 g/m2-yr in the large field of silty clays. To
the south of the bridge in segments 3 and 4 the potential flux of the
muddy sediments decreases corresponding to the decrease in sedimentation
rate even though the percent clay content remains high. In segment 6,
the potential carbon flux in reduced to less than 20 g/m2-yr in the
isolated deep water coarser sediments. Conversely, the lowest potential
carbon fluxes occur in the coarser grained sandy sediments located
around the Bay's periphery and in the vicinity of the mouth, and in
those areas where the sedimentation rate is lowest (segments H and 9).
Table 3 is a more quantitative statement of the model results. It
shows the integrated flux within each sediment type in the Bay. The
integrated flux was calculated by: obtaining the average flux in each
sediment type in each Bay segment; this average flux was in turn
multiplied by the areal extent, in square kilometers, of each sediment
type in the segment, and; the results were totalled to give Table 4. The
potential fluxes of Nitrogen and Phosphorus are calculated by using
Redfield's ratio (Redfield et al., 1966) of planktonic matter
composition.
There are several interesting features to note in this table. The
first is that sandy sediments provide over 3>Q% of the total flux. This
is somewhat surprising because the sandy sediments are quite low In
organic carbon. However, the low carbon content is offset by the large
areal expanse of these sediments, >57% of the total bottom area, and the
high solids density («1.3 g/cc for sandy sediments as opposed to sO.^
196
-------
Table 3. Calculated potential flux of carbon, divided according to
sediment type. For comparison to EPA estimates, Redfield's
ratio is employed to calculate the Total Nitrogen (TN) and
Phosphorus (TP) flux.
Sediment
Type
SANDY CLAY
SAND
SILT
CLAY
SAND-SILT-CLAY
CLAYEY SAND
CLAYEY SILT
SILTY CLAY
SILTY SAND
SANDY SILT
Area
(km2)
67.9
3715.0
1.9
73.1
327.6
67.9
496.6
1207.3
546.8
87.0
6526.6
Area
1
57
0
1
5
1
8
18
8
1
Flux C
(kg/yr)
x 10°
0.092
74.956
0.009
3.122
12.941
1 .084
22.679
65.592
14.003
2.031
196.509
Flux
0
38
0
1
7
1
12
33
7
1
EPA ANNUAL INPUT
Total » 196.512
34.604
4.784
(C)
(N)*
(P)*
Total
(TN) 137.3
(TP) 13.7
Benthic
14.6
3.4
Based on Redfield's ratio of 106:16:1 (C:N:P)
797
-------
g/cc for silty clays). As a result, sandy areas contribute signif-
icantly to the overall benthic flux and should not be discounted when
detailed flux measurements are to be made in the Bay.
The other interesting feature to note arises from the comparison of
the EPA flux estimates (EPA, 1982) to the flux estimates of the model.
The model predicts input values significantly higher than the EPA
estimates, even though the model only accounts for the main stem of the
Bay, while the EPA estimates are for the Bay plus tributaries. It would
be expected that the model estimates would be higher than measured
fluxes because the mineralization of Phosphorus and denitrification
processes can not be taken into account. If the EPA estimates of
nutrient regeneration are comparable then 58$ of the ammonium released
by diagenetic processes is converted to N2 by nitrification/denitri-
fication processes, and 29$ of the Phosphate liberated is mineralized in
the sediments (most of the mineralization probably occurs in the
northernmost portion of the Bay segments 1-3). However, since the model
does not include the tributaries, which almost doubles the area
involved, the model estimate would be substantially increased.
Denitrification removes 5-25$ of the nitrogen originally incorporated
into the sediment (Nixon, 1981); up to approximately 50$ of the
regenerated Nitrogen. Consequently, either denitrification accounts for
the removal of much greater than 50$ of the Nitrogen in the sediments,
or the EPA estimates of the total annual benthic input is quite low
based on this model.
In summary, this model provides an independent means of determining
upper limit flux estimates using relatively simple, inexpensive
techniques. Better estimates would require more data from cores, with
accompanying sedimentation rate data.
198
-------
Bennett, R.H. j Lambert, D.N.; Rapid and reliable technique for
determining unit weight and porosity of deep sea sediments; Marine
Geology, v. 11, pp. 201-207; 1971.
Berner, R.A.; Sulfate reduction, pyrite formation, and the oceanic
sulfur budget; (in) Nobel Symposium 20; The Changing Chemistry of
the Oceans (eds. Dyrrsen, D. and Jagner, D.), pp. 347-361; 1972.
Berner, R.A.; A rate model for organic matter decomposition during
bacterial sulfate reduction in marine sediments; Biogeochimie de la
Matiere Organique A'L'Interface Eau-Sediment Marin, Collogue
Internationaux du C.N.R.S., pp. 34-44; 1980.
Brush, G.S.; Davis, F.W.; Stratigraphic evidence of human disturbance in
an estuary; Quaternary Research, 22 pp.; 1984.
Byrne, R.S. ; Hobbs, C.H.; Carron, M.J.; Baseline sediment studies to
determine distribution, physical properties, sedimentation budgets,
and rates in the Virginia portion of the Chesapeake Bay; Virginia
Institute of Marine Science, Final Report to the U.S. Environ-
mental Protection Agency, 155 pp.; 1982.
Draper, N.R.; Smith, H.; Applied Regression Analysis; New York: Wiley;
1966.
Environmental Protection Agency; Chesapeake Bay Program Technical
Studies: A Synthesis; U.S. Government Printing Office: 1982-
509-660; 635 pp.; 1982.
Helz, G.R.; Sinex, S.A.; Setlock, G.H.; Cantillo, A.Y.; Chesapeake Bay
sediment trace elements; Final Report to the U .S. E.P.A., 202 pp.;
1982).
Hennessee, E.L.; Blakeslee, P.J.; Hill, J.M.; Distributions of organic
carbon and sulfur in surficial sediments of the Maryland portion of
Chesapeake Bay; Jour. Sed. Pet., (in press); 1984.
Hill, J.M.; Sedimentary carbon and sulfur as indicators of biogeochem-
ical processes; Geological Society of America annual meeting
(Phoenix, Arizona); Abstracts with Program vol. 19, p. 703; 1987.
Hill, J.M.; The behavior of carbon and sulfur in the sediments of
Chesapeake Bay; (in prep.).
Hobbs, C.H.; Organic carbon and sulfur in sediments of the Virginia
Chesapeake Bay; Jour. Sed. Pet., v. 53, pp. 383-393; 1983.
Holdahl, S.R.; Morrison, N.L.; Regional investigation of vertical
crustal movements in the U.S., using precise relevelings and
mareographic data; Tectonophysics 23: 373-390; 1974.
Jorgensen, B.B.; The sulfur cycle of a coastal marine sediment
(Limfjorden, Denmark); Limnology and Oceanography, v. 22, no. 5,
pp. 814-832; 1977.
Jorgensen, B.B.; The microbial sulfur cycle; (in) Microbial Geochemis-
try; (ed.) Krumbein, W.E.; Blackwell Scientific Publications, pp.
91-124; 1983.
Kerhin, R.T.; Halka, J.P.; Wells, D.V.; Hennessee, E.L.; Blakeslee,
P.J.; Zoltan, N.; Cuthbertson, R.H.; The surficial sediments of
Chesapeake Bay, Maryland: Physical characteristics and sediment
budget; Maryland Geological Survey, Report of Investigations No.
^8, 160 pp.; 198?).
Martens, C.S.; Berner, R.A.; Rosenfeld; J.K.; Interstitial Water
Chemistry of Anoxic Long Island Sound sediments. 2. Nutrient
Regeneration and Phosphate Removal; Limn. Oceanogr., v. 23 (4), p.
604-617; 1978.
Nixon, S.W.; Remineralization and nutrient cycling in coastal marine
ecosystems; (in) Neilson, B.J., Cronin, L.E., eds., Estuaries and
Nutrients, The Humana Press; pp. 111-138; 1981.
199
-------
Officer, C.B.; Lynch, D.R.; Setlock, G.H.; Helz, G.R.; Recent
sedimentation rates In Chesapeake Bay; (in) Kennedy, V.S., ed.,
The Estuary as a Filter, Orlando, FL: Academic Press, p. 131-157;
1981.
Redfield, A.C.; Ketchum, B.H.; Richards, F.A.; The influence of
organisms on the composition of seawater; (in) The Sea; Vol. II
(ed. Hill); Wiley-Interscience, New York; 1966.
Shepard, F.P.; Nomenclature based on sand-silt-clay ratios; Jour. Sed.
Pet., v. 21, pp. 151-158; 1951.
Sweeney, R.E.; Pyritization during diagenesis of marine sediments;
(unpubl. Ph.D. dissertation): Los Angeles, Univ. of CA, 181 pp.;
1972.
200
-------
Understanding the Estuary: Advances in Chesapeake
Bay Research. Proceedings of a Conference. 29-31
March 1988. Baltimore, Maryland. Chesapeake Research
Consortium Publication 729. CBP/TRS 24/88.
Delineation of Regional Sediment Resuspension Potential in Chesapeake Bay,
with Implications for Bottom Sediment Management
Richard W. Faas
Lafayette College
Department of Geology
Easton, Pennsylvania 18042
INTRODUCTION
Chesapeake Bay, the largest estuary on the east coast of North
America, 1s a partially-mixed system 1n which seaward flowing, low-
salinity water overlies landward flowing, high-sal Inlty water with
mixing occurring between the two water layers. Nearly 50X of Its
freshwater comes from the Susquehanna River which 1s also the primary
source of Its sediment. Sediment Input depends upon Susquehanna flow
rates, but 1n the decade 1966-1976, approximately 50 million metric
tons were delivered to the upper Bay from the Susquehanna (Gross et al
1978). Estimates of sediment retention for the Bay vary between 90 to
greater than 100% (Includes sediment Introduced from offshore) (Meade
1982; Biggs and Howell 1984; Schubel and Carter 1984). Sediment
accumulation appears to be greatest in the northern part; lesser but
significant accumulation occurs 1n the southern part; and minimal
accumulation Is found 1n the middle portion of the Bay (Officer et al
1984).
Geochemlcal studies have demonstrated the existence of near-
bottom suspensions of fine-grained sediments, In excess of 60 cm
thick, extending northward from the mouth of the Potomac (Station 8,
Fig. 1) to Tolchester Beach (Station 15, Fig. 1) (Nichols et al 1981).
Sampling has generally been limited to the Bay axis, consequently, the
area! extent of these suspensions is unknown. They have been called
"fluid mud", "sling mud", "fluff", "creme de vase", and "slib" when
occurring in concentrations >10 g/1 (Wells and Coleman 1981). No
201
-------
Resuspension
Potential
1. Location of stations occupied for hydrographlc and sediment
sampling In Chesapeake Bay. Map also shows regions of high, interme-
diate, and low resuspension potential determined from this investiqa-
tlon. 3
-------
universal definition 1s yet available for this phenomenon Inasmuch as
Its recognition depends on the techniques used (Parker and K1rby
1982). In this paper, "fluid mud" 1s considered to be a suspension
ranging between 10 to 480 g/1, corresponding to a density from 1.03 to
1.30 Mg/m3. Kirby and Parker (1983) classified fluid muds as
stationary and mobile suspensions which may show discontinuities or
density stratifications, termed "lutoclines". Similar features have
been observed 1n situ In suspensions from a dredged channel 1n the
James River (NTc"hol s et al 1978; Nichols 1985). These suspensions
possess the rheologlcal properties of a viscous fluid (Krone 1963;
M1gn1ot 1968; Pazwash and Robertson 1971; Bryant et al 1980; Faas
1981; Dyer 1986) and are believed to play an Important, but as yet un-
known role in estuarlne transport processes (Officer 1981; Nichols
1985).
MATURE OF THE PROBLEM
River-borne sediment enters the estuary, mixes with sediment
brought up from seaward reaches and, through processes not fully
understood, settles through the water column to accumulate on the
bottom (Schubel 1968, 1971, 1982; Zabawa 1978; Kranck 1975, 1986;
G1bbs 1983; 1985). During passage from fresh into saline water, clay
particles react with dissolved ions of various types, e.g., trace
elements, nutrient Ions, and toxic compounds, and, by settling, remove
some of these materials from the water. This results 1n chemical
enrichment of bottom sediments which when consolidated, serve as a
quasi-permanent reservoir of these materials (Harris et al 1980; Gibbs
1982, 1986). Fluid mud suspensions tend to remain potentially mobile
under dynamic tidal conditions and have been observed moving along the
bottom as well-defined, cohesive sediment flows at velocities between
15-25 cm/s (Ingllss and Allen 1957; Kirby and Parker 1977).
The potential for an estuarlne fluid mud to be resuspended 1s
dependent upon many factors, e.g., morphologic, hydrographic, and
rheologic. These latter properties include the "apparent" viscosity
of the suspension, Its rheologlcal behavior throughout the range of
shear rate and shear stress experienced during an accelerating tide,
and its tendency to develop a yield stress during hindered settling.
A simple measure of resuspension potential should be the difference
between the yield stress of the suspension and the maximum tidal shear
stress to which the suspension has been exposed during a tidal cycle.
The purpose of this study is to determine the rheologlcal behavior of
fluid mud suspensions under simulated estuarlne shear rate and shear
stress condltons as a predictor of their potential for resuspension.
Two aspects of the fluid muds seem critical to this problem: 1) their
hindered settling behavior and density development during a slack
water Interval; and 2) the corresponding development of yield stress
and viscous flow characteristics during the same interval. It 1s
suggested that these properties interact 1n such a way as to either
Inhibit resuspension or determine when resuspension will occur in
the shear stress field.
203
-------
Background
In muddy estuaries, fine sediment particles settle out of the
water column during slack water and may accumulate to form stationary
suspensions which, If undisturbed, may consolidate Into settled mud
(Parker and K1rby 1982; K1rby and Parker 1983). Little 1s known about
the development of time-dependent properties of short-term (1-4 hr)
consolidation of fine sediment suspensions. Michaels and Bolger
(1962) studied the settling rates and volumes of flocculated kaolin
suspensions with respect to plastic and structural properties. M1g-
nlot (1968) showed that fine sediment suspensions from different
rivers consolidated at different rates, dependent upon salinity and
sol Ids concentration. Owen (1970) demonstrated that a thin 1.1 cm mud
layer, capable of withstanding erosion could be deposited from a
dilute sediment suspension in four hours. Creutzberg and Postma
(1979) showed that muds from the southern North Sea did not stabilize
in consolidation periods shorter than 1.5 hours and would be resuspen-
ded by current velocities less than 12 cm/s. Hawley (1981) demon-
strated that a thin mud layer, capable of resisting erosion at veloci-
ties as great as 14 cm/s could develop 1n 1.25 hours settling from a
dilute (4 g/1) suspension 1n water having a salinity of 15 g/1.
Studies of the viscosity of naturally occurring clay/water
suspensions were made by Krone (1963) who determined the viscosity of
San Francisco Bay muds; Migniot (1968) who studied the relationships
between viscosity, salinity, and sol Ids concentration for muds from a
variety of marine, estuarine, and fluvial environments; and, several
studies by Dutch workers (NEDECO 1965, 1968; Allersma 1982) on the
muds from Surinam and the Port of Bangkok, Thailand. Wei Is (1978)
measured viscosities of muds from the central Surinam coast, and Faas
(1981) observed different forms of viscous behavior which appeared to
control resuspension patterns in the Rappahannock Estuary, and record-
ed similar behavior patterns from fluid muds on the NE Brazilian
continental shelf (1985, 1986). Bryant et al (1980) found that muds
from Rotterdam and Brisbane behave as Newtonian fluids in low concen-
trations (<10 X) and exhibit non-Newtonian behavior as concentrations
Increase. This same phenomena is observed 1n stationary and mobile
suspensions from the Severn Estuary and Inner Bristol Channel (K1rby
and Parker 1983).
Methods
Samples were collected in a north-south transect along the
Chesapeake Bay axis 1n October 1982, using the RV Ridqely Warfield.
Bottom samples were obtained with a standard grab sampler and were
kept refrigerated at 10 ° C until analysis. The grab samples were
sampled for natural water content onboard ship Immediately upon
retrieval. Grain size analysis was performed in the laboratory with
the 152H Bouyoucos hydrometer (Bouyoucos 1962; Kaddah 1974), Organic
matter content was determined by loss on Ignition (Davles 1974).
204
-------
Although fluid mud was recovered at several localities, the
amounts were too small to be useful for experimental work and
suspensions for analysis were generated directly from the upper 5 cm
of a bottom grab sample. In Chesapeake Bay and other similar muddy
environments, fine-grained material enters the system from multiple
sources, e.g., the watershed, the offshore, the atmosphere, from
erosion of the tidal flats and estuary margin, and is constantly being
regenerated from the bottom sediments through the activities of infau-
nal and epifaunal organisms (Rhoades 1963, 1974; Haven and Morales-
Alamo 1968; Rhoades and Young 1971) and resuspended by tidal scouring
(Schubel 1968, 1971; Nichols and Biggs 1985), and wind-waves (Anderson
1972). Consequently, it seems likely that suspensions generated from
the bed will closely resemble the suspensions from which the bed
accumulated.
Suspensions analyzed for density and viscosity were prepared from
bottom samples by placing an amount of wet sample, equivalent to 50 g
dry sediment, 1n a Waring blendor and adding room temperature water,
with salinity adjusted to that existing at the time of sampling.
After blending for 10 minutes, the sample was transferred to a 1-
Uter graduated cylinder, shaken thoroughly to ensure complete disper-
sion, and allowed to settle. Settling of the sharp interface that
formed between the water and the sediment was measured for a two-hour
Interval. Each 50 g sample had the same initial density (1.031 Mg/m3)
and changes in density were calculated at 10,- 30, 60 and 120 minuter
as the Interface settled. Grain density was assumed to be 2.65 Mg/m3.
Viscosity analyses were performed on 20 ml subsamples of fluid
mud taken from the settling tubes. Analysis was done with the Brook-
field 8-speed RVT rotational viscometer, equipped with the UL adaptor
for low viscosity fluids. Measurements consist of rotating a cylin-
drical spindle at eight different speeds in the suspension in a cup,
the radius of which is only slightly larger than that of the spindle.
The shear stress and shear rate at each speed can be calculated pre-
cisely (due to the mathematical ly correct dimensions of the cup and
spindle). Inasmuch as behavior is usually non-Newtonian, viscosity is
considered to be "apparent" as it varies with changes 1n shear rate
and shear stress (Van Wazer et al 1963). Yield stress was determined
by recording the highest value of shear stress initially achieved at
the lowest shear rate. Each sample was analyzed through an accelera-
ting and decelerating shear rate cycle to simulate the shear stresses
of a tidal cycle. Shear rates ranged between 0.61 to 122.36 s ,
corresponding to those customarily experienced in estuaries (Bryant et
al 1980; Dyer 1986). Data were plotted as flow diagrams (log shear
rate vs. log shear stress) and rheograms (shear rate vs. log "appa-
rent" viscosity - F1g. 2a and 2b). After analysis, each sample was
transferred to a pre-welghed alumunum moisture can and placed 1n a
105°C drying oven to determine the percentage of solid material.
205
-------
Shear Strata (Pa)
•dh
O
o
O
o.
JO
3 =r
oo re
o
_
01 o» •"•
— "* 2
Srt- *
(A
a
a
*< -J
o»
-3
<
Vt tQ
If
•si
a *
co
«o
O M Ol
2 3
(-• O
-•.•a
3 »
<•* m
ro g
••
r
O
O
a
O
z
_k
CO
I
O
I
rn
co
>
•D
m
CO
o
O
5
"
> IL a. {/>
—•• __. rt> T
_^ —•» ^ J
.fT _ 3
-------
DISCUSSION OF RESULTS
Hindered Settling and Density Development
Hindered settling 1s exhibited by flocculated suspensions 1n
concentrations >10 g/1 and 1s characterized by a sharp -'-terface
between the overlying clear water and the flocculated material which
travels downward as interstitial water is expelled upward through the
flocculent structure. The process has been suggested to be the mode
of accumulation of fluid mud (Einstein and Krone 1962) and 1s des-
cribed 1n detail by Been and 511 Is (1981) and Sills and Been (1984).
It is believed that hindered settling occurs 1n Chesapeake Bay
during slack water by sediments which flocculate in the low salinity
reaches of the upper Bay (Schubel and Kana 1972; Zabawa 1978; Gibbs
1985). It is understood that there will always be some residual
circulation during most slack water intervals. However, the very
presence of fluid muds indicates that the shear stresses generated by
this residual circulation are too weak to keep the fine particles
separated and in suspension. Creutzberg and Postma (1979) have
demonstrated experimental ly that there is a critical deposltlonal
velocity below which North Sea silts become deposited. This velocity
1s 12 cm/3, measured 15 cm above the bottom. It seems likely that a
similar relationship exists for the sediments of Chesapeake Bay.
Figure 3 shows the changes 1n density with settl1ng time of the
suspensions from each sample site. Three regions, each demonstrating
different modes of hindered settling, are shown. Rapid settling
occurs at Station 7 and seaward. Intermediate settling is observed at
stations landward of Station 16" Stations- 17 and 24 settle to a
slurry density of 1.30 Mg/m3 in about 1-hour, hence fluid mud should
not be found abundantly at these sites. Stations 18 and 19, while not
achieving a density of 1.30 Mg/nr, also settle to high density slur-
ries. Slowest settling is seen between Stations 10 to 16, with set-
tling increasing slightly toward Station 16. At Station 14, a density
of 1.19 Mg/m3 was achieved after 1-hour of settling.
In a different view, Figure 4 shows the Increase in density
during 1-hour of settling of each suspension sample. Stations 10
through 16 show a maximum density increase of 12% as compared to 20-
40X increase further up and down the estuary. It is clear that the
least dense, most water-rich sediments will be found between Stations
10 to 16. Fluid mud should occur in significant quantities 1n this
reach and has been documented by Nichols et al (1981).
Viscosity Characteristics
The suspensions exhibit non-Newtonian shear thinning (pseudoplas-
tic) flow behavior, typical of dilute clay/water systems (van Olphen
1963). In physical terms, this means that the "apparent" viscosity of
the suspensions decreased as the shear rate (defined as the velocity
gradient, I.e., dv/dy. between the top and basal surfaces of the
207
-------
Suspension Density (Mg/m )
333
3. Longitudinal profile showing density Increase with settling time
for all stations. Minimal Increase occurs at Stations 10 to 16
with greater Increase occurring landward of Station 16 and seaward of
Station 10.
205
-------
Percent Increase
en
r\>
o
tn
O
z
ro
.*
(£>'
CD
O)
Ul
ro
o
ro
ol
(£)
CO
xl
O>
4. Percent Increase of suspension density 1n 1-hour settling time for
all stations. Lowest Increase 1s observed to occur between Stations
10 to 16.
209
-------
layers being sheared) Increased. The reverse phenomena, shear
thickening (dilatant) flow occurred 1n samples that had settled long
enough to attain a 10% solids concentration (usually 5 hours). Flow
diagrams and rheograms are shown 1n Fig. 2a and 2b for Station 13.
The sample was analyzed after 1, 2, and 6.5 hours of settling. During
this time, the density of the suspension increased from 1.04 Mg/rtr
(7.0% solids) to 1.054 Mg/m3 (8.7% solids) to 1.069 Mg/m3 (11.1%
solids). A tendency toward shear thickening was observed at the
lowest shear rate 1n the longest settled sample - however, shear
thickening was not common in these suspensions. These behavioral
patterns are important sedimentologically inasmuch as shear thinning
results 1n a decrease in "apparent" viscosity which, under high shear
rates, would encourage resuspenslon by reducing the thickness of the
laminar layer of the benthic boundary layer. Shear thickening results
in an increase 1n "apparent" viscosity which, 1f occurring at some
critical higher shear rate, would increase the thickness of the lami-
nar layer of the benthic boundary layer and Inhibit resuspenslon
(Faas 1985, 1986). A dramatic increase 1n the thickness of the
laminar layer was observed in the drag reduction experiments of Gust
(1976) and Gust and Walger (1976) when fine sediment was introduced
Into the flow field.
In order to visualize the pattern of viscous change throughout
the Bay,"apparent" vlscositleswere plotted for 1 and 2-hour settling
times for samples from each station (Fig. 5). Suspensions possessing
high "apparent" viscosities exist seaward of Station 10 and landward
of Station 14. Low "apparent" viscosities occurred in suspensions
between Stations 11 and 14, with an anomalous intermediate "apparent"
viscosity suspension existing at Station 12.
Figure 6 shows the relationship between suspended solids and
"apparent" viscosity for most samples. Each was measured at 1 and 2-
hour settling intervals, and shows the associated density increase.
Two distinct trends are noted. Samples from the upper Bay (Stations
15, 16, 18, 19, and 24) show a large increase of "apparent" viscosity
with suspended solids concentration whereas the Increase 1n middle Bay
sediments is much less. Upper Bay samples are less viscous than those
from the middle Bay since they require a greater solids concentration
(2X) to achieve the same "apparent" viscosity. The distinction 1s
quite abrupt and results from the fact that upper Bay sediments are
generally coarser-grained than middle Bay sediments. However, since
upper Bay suspensions achieve high densities more rapidly during the
same settling interval than those from the middle Bay (Fig. 3), their
"apparent" viscosities are also greater. This Is best Illustrated by
the differences 1n "apparent" viscosity between the two groups after
one hour of settl 1ng (F1g. 5).
One of the time-dependent properties often found in clay/water
systems is yield stress. This phenomena results from an Internal
strengthening of the suspension and requires that a certain amount of
shear stress be applied to the system before flow 1s Initiated.
Mlgniot (1968) found that yield stress increased rapidly 1n concen-
270
-------
Apparent Viscosity (cP) «t 6.12 s
t O*> GO O
> .. ?
-1
ro
o
o>
o
)
I
o
z
»
t\>
o
K
m
5. Apparent" viscosity Increase with settling time for all stations.
Lowest viscosities are found at Stations 11, 13, 14, and 20. Greatest
viscosity Increases are found landward of Station 14.
277
-------
900 H
s
400-
^
«
CM
•
r»
• 300-i
e
o
o
>200-
o
|
100-
Lower & Middle Bay
Upper Bay
10 16
Suspended Solids (%)
20
• r
25
6. "Apparent" viscosity versus percent of suspended sol Ids for all
stations. Left side of dashed line shows "apparent" viscosity
Increase for one and two-hour settled samples from lower and middle
Bay stations (6-14), measured at 1.22 s"1 shear rate. Low "apparent"
viscosity (about 100 cPs) occurs 1n one hour of settling, generally
doubling after two hours. Right side of dashed line shows the same
data for upper Bay samples (15-24). "Apparent" viscosity after one
hour 1s at least twice that of lower and middle Bay samples, and also
doubles after two hours of settling. Station 17 behaves as an upper
Bay sample at lower suspended sol Ids concentrations as does Station 8
from the lower Bay. Station 6 (lower Bay) Illustrates middle Bay
behavior. Upper Bay suspensions require greater densities to achieve
the same "apparent" viscosities as middle Bay suspensions. However,
due to rapid settling and density Increase, upper Bay suspensions
become more viscous than middle Bay suspensions in equivalent settling
times.
272
-------
trated mud suspensions (>200 g/1), as did Malherbe et al (1986) 1n the
muds from the harbor of Zeebrugge (Belgium). However, Bryant et al
(1980) observed little Indication of yield stress development in their
study of the fluid mud of the Severn Estuary and Bristol Channel.
Figure 7 shows that yield stresses as great as 2 Pa (1 Pa = 10
dynes/cmO may develop 1n Chesapeake Bay suspensions at densities to
1.15 Mg/m3. This stress (2 Pa) corresponds to a critical shear velo-
city (U*) of 2.5 cm/s (M1gn1ot 1968). Calculations, fol lowing Ter-
windt and Breusers (1972, 1982) Indicate this U* can be achieved by a
current of 67 cm/s, measured at a distance of 1 m above the fluid mud
interface. Development of. a greater yield stress in 60 minutes of
hindered settling is unlikely to occur at Stations 8, and 10 to 16
(possibly excluding 14 and 15) inasmuch as densities through this
reach Increase very slowly. Pritchard (1971) Indicates that the net
non-tidal current (up-estuary in the lower layer) is one-fifth the
magnitude of tidal currents (25 to 100 cm/s). Mean flood tidal velo-
city 1n the lower layer is about 40 cm/s and the mean net non-tidal
bottom flow 1s approximately 10 cm/s up-estuary. Resuspension of
these sediments could occur at current velocities in excess of 67
cm/s, measured 1 m above the fluid mud interface; however, since these
stations are al 1 in water depths greater than 12 m, such velocities
seem unlikely.
Stations 17 to 24 reach suspension densities between 1.20 - 1.30
within 60 minutes and presumably develop correspondingly greater
yield stress. However, they lie at depths less than 10 m and are
1ikely affected by strong tidal currents. Schubel (1969) measured
flood tidal velocities of "75 cm/s and ebb tidal velocities of "55
cm/s at 50 cm above the bottom in the upper Bay (corresponding to
Station 18) with concentrations of suspended sediment as high as 279
mg/1.
Nichols (1986) relates resuspension to a simple relationship
between energy capacity "e" and sediment concentration "c" and
indicates resuspension will occur whenever "e" exceeds "c". This
concept falls to take into account sediment responses associated with
density-dependent phenomena such as shear thinning and shear thick-
ening behavior and time-dependent phenomena such as yield stress. In
addition, the geochemical setting may also be a determining factor.
In oxidizing environments, when the sediment concentration approaches
200 g/1, shear thickening usually occurs at some critical shear rate
and shear stress and resuspension may be restricted (Faas 1985, 1986).
However, in anoxlc environments, the rheological behavior of the
sediments 1s quite different. Studies of acoustic properties of gassy
sediments generally Indicate-a loss of a rigid framework resulting in
absorption and attenuation of the acoustic energy (Wood and Weston
1964; Schubel and Schiemer 1973; Anderson 1974). Figure 5 Indicates
that low densities and low apparent viscosities occur in suspensions
throughout this reach. In these sediments, shear thinning dominates,
yield stress 1s minimal, and a greater potential for resuspension 1s
predicted.
213
-------
2.0 n
1.5 •
« 1.0
s^
CO
CO
lil
oc
H
co
o
_l
UJ
0.5 -
1.00
1.05
1.10
1.15
DENSITY (Mg/m )
7. Yield Stress Increases with density during two-hour settling
Interval of middle Bay samples (Stations 10 to 16).
214
-------
CONCLUSIONS
Chesapeake Bay can be rather sharply divided Into three regions
of varying resuspenslon potential (Fig. 1). Lowest potential extends
from the mouth of the Bay northward to Station 9 off the Patuxent
River. This low potential Is due primarily to high "apparent" visco-
sity and rapid density development due to lesser amounts of clay-sized
material 1n bottom-generated suspensions. Greatest potential extends
from Station 10 to Station 14 off the mouth of the Patapsco River 1n
the middle Bay. Suspensions from this reach show the lowest "appa-
rent" viscosity and lowest yield stress development of any suspensions
within the Bay. Settling rate and density Increase with time are very
low which contributes to the low "apparent" viscosity. Intermediate
resuspension ootential exists in suspensions from Station 14 northward
to Station 24 at the mouth of the Susquehanna. "Apparent" viscosity
increases sharply, reaching its maximum at Station 19, then decreases
slightly. The percent of density increase during one-hour of hindered
settling is lowest at Station 16, then increases to about 20% for the
remaining stations. The middle reach of Chesapeake Bay (Stations 10
to 14) contain dense suspensions of fluid mud (Nichols et al 1981).
This reach experiences severe anoxia for several months each year
(Officer et al 1984; Seliger et al 1985). Such conditions lead to a
weakening of sediment bonds (Faas and Wartel 1977; Wartel et al 1985)
and may be a factor contributing to the development of fluid mud in
this reach.
Since this work dealt specifically with a laboratory characteri-
zation of the physical properties of suspensions generated from Chesa-
peake Bay bottom sediment, actual jjn situ resuspenslon behavior re-
mains untested. The fact that laboratory experiments reveal distinct
patterns of sediment behavior and a range of parametric variability
which corresponds with Independent field observations, e.g., distribu-
tion of fluid mud, sediment textural variation, and current circula-
tion patterns, appears to substantiate the interpretations expressed
and to provide an additional framework within which to analyze past
and future data. However, 1t remains to be determined at which of the
sampling stations nearbottom fluid mud suspensions actually exist, how
much area they occupy, their persistence throughout the year, and
their rheologlcal behavior through a lunar tidal cycle.
ACKNOWLEDGEMENTS
I am grateful to F. BMnckmann, W. Blair, W. Ivorsen, and G.
01 sen of the National Bureau of Standards for their Interest and
cooperation during the acquisition and analysis of the samples. I
also thank the Captain and crew of the RV Rldqely Warfleld for their
hospitality and cooperation 1n the field and their efforts in
obtaining the samples. K. Swider assisted in sampling and analysis. I
appreciate the efforts of J. Rucker and J. Wei 1 s 1n a review of this
paper. Their suggestions were most helpful. This work was supported,
1n part, by an NSF-IRP grant for research at the National Bureau of
Standards. The Brookfleld vlscometer was provided by the Chemical
Engineering Department, Lafayette College.
275
-------
REFERENCES
ALLERSMA, E. 1982. Mud 1n estuaries and along coasts. Delft Hyd. Lab.
Pubi 270: 663-685.
ANDERSON, A. L 1974. Acoustics of gas-bearing sediments. App. Res.
Lab. Rept. ARL-TR-74-18. The University of Texas at Austin, 162p.
ANDERSON, F. E. 1972. Resuspenslon of estuarlne sediments by small
amplitude waves. Jour. Sed. Pet. 42: 602-607.
BEEN, K., AND G. C. SILLS. 1981. Sel f weight consol 1dat1on of soft
soils. Geotechnique 31: 519-535.
BIGGS, R. B., AND B. A. HOWELL. 1984. The estuary as a sediment trap:
alternate approaches to estimating its filtering efficiency. In:
Kennedy VS (ed) The estuary as a filter, Academic Press, New York,
pp 107-129.
BOUYOUCOS, G. J. 1962. Hydrometer method Improved for making particle
size analysis of soils. Agron. Jour. 54: 464-465.
BRYANT, R., A. E. JAMES, AND D. J. A. WILLIAMS. 1980. Rheology of
cohesive suspensions. In: Collins MB, Banner FT, Tyler PA, Wake-
field SJ, James, AE (eds) Industrial embayments and their environ-
mental problems. A case study of Swansea Bay. Pergamon Press, New
York, pp 279-287.
CREUTZBERG, F., AND H. POSTMA. 1979. An experimental approach to the
distribution of mud 1n the southern North Sea. Neth. Jour. Sea.
Res. 13: 99-116.
DAVIES, B. E. 1974. Loss-on-1gnit1on as an estimate of soil organic
matter. Soil Sd. Amer. Proc. 38: 150.
DYER, K. R. 1986. Coastal and estuarine sediment dynamics. John
Wiley and Son, Ltd. 342p.
EINSTEIN, H. A., AND R. B. KRONE. 1962. Experiments to determine
modes of cohesive sediment transport 1n salt water. Jour. Geophys.
Res. 67: 1451-1461.
FAAS, R. W. 1981. Rheological characteristics of Rappahannock estuary
muds, southeastern Virginia, U.S.A. Spec. Publ. Int. Assoc.
Sedlmentol: 505-515.
. 1985. Time and density-dependent properties of fluid mud sus-
pensions, NE Brazilian Continental Shelf. Geo-Mar. Lett. 4: 147-152
. 1986. Mass-physical and geotechnical properties of surficial
sediments and dense nearbed sediment suspensions on the Amazon
continental shelf. Cont. Shelf Res. 188: 189-208.
216
-------
FAAS, R. W., AND S. WARTEL 1977. The effect of gas bubble formation
on the physical and engineering properties of recently deposited
fine-grained sediments. Geol. en Mijnb. 56: 211-218.
GIBBS, R. J. 1982. The differentiation of toxic metals in estuaries.
Thai. Jugo. 16: 181-190.
. 1983. Coagulation rates of clay minerals and natural sedi-
ent. Jour. Sed. Pet. 53: 1193-1203.
. 1985. Estuarine floes: their size, settling velocity and
density. Jour. Geophys. Res. 90, C2: 3249-3251.
. 1986. Segregation of metals by coagulation 1n estuaries.
Mar. Chem. 18: 149-159.
GROSS, M. G., M. KORNWEIT, W. B. CRONIN, AND J. R. SCHUBEL. 1978.
Suspended sediment discharge of the Susquehanna River to northern
Chesapeake Bay, 1966 to 1976. Estuaries 1: 106-110.
GUST, G. 1976. Observations on turbulent-drag reduction 1n a dilute
suspension of clay in sea-water. Jour. Fluid Mech. 75: 29-47.
GUST, G., AND E. WALGER. 1976. The infl uence of suspended cohesl ve
sediments on boundary-layer structure and erosive activity of tur-
bulent seawater flow. Mar. Geol. 22: 189-206.
HARRIS, R., M. NICHOLS, G. THOMPSON, J. BONACKI, AND G. VADAS. 1980.
Heavy metal inventory of suspended sediment and fluid mud in Chesa-
peake Bay. Va. Inst Mar. Sci. Spec. Sci. Rpt 99, 133 p
HAVEN, D. S., AND R. MORALES-ALAMO. 1968. Occurrence and transport of
faecal pellets 1n suspension in a tidal estuary. Sed. Geol.
2: 141-151.
HAWLEY, N. 1981. Flume experiments on the origin of flaser bedding.
Sedlmentology 28: 699-712.
INGLISS, C. C., AND F. H. ALLEN. 1957. The regimen of the Thames
estuary as affected by currents, salinities, and river flow. Pro.
Inst C1v. Eng. 7: 827-868.
KADDAH, M. T. 1974. The hydrometer method for detailed particle-size
analysis. 1. Graphical interpretation of hydrometer readings and
test of method. Soil Sc1. 118: 102-108.
KIRBY, R., AND W. R. PARKER. 1977. The physical characteristics and
environmental significance of fine-grained sediment suspensions in
estuaries. Estuaries, Geophysics and the Environment, Nat. Acad.
Sc1., Washington, D.C. pp 110-120.
217
-------
. 1983. Distribution and behavior of fine sediment 1n the
Severn Estuary and Inner Bristol Channel, U. K. Can. Jour. F1sh.
Aquatic Sd. 40: 83-95.
KRANCK, K. 1975. Sediment deposition from flocculated suspensions.
Sedimentology 22: 111-123.
. 1981. Partlculate matter, grain size characteristics, and
flocculatlon 1n a partially-mixed estuary. Sedimentology 28:
107-114.
. 1986. Settling behavior of cohesive sediment. In: Mehta AJ
fed) Estuarine cohesive sediment dynamics. SpMnger-Verlag, New
York, pp 151-169.
KRONE, R. B. 1963. A study of rheologic properties of estuarlal sedi-
ments. Tech Bull 7, Comm on Tidal Hydr, U.S. Army Corps of Engr,
WES, Vlcksburg, MI.
MALHERBE, B., P. DEWOLF, AND B. PAQUOT. 1986. Nautical bottom re-
search and survey for optimization of maintenance dredging 1n mud
areas. Symp. on advances in underwater technology, ocean science
and offshore engineering, 6, Brighton, U.K., pp 427-441.
MEADE, R. H. 1982. Sources, sinks, and storage of river sediments in
the Atlantic drainage of the United States. Jour. Geol. 90: 235-
252.
MICHAELS, A. W., AND J. C. BOLGER. 1962. Settling rates and sediment
volumes of flocculated kaolin suspensions. Ind. & Engng. Chem.
Fund. 1: 24-33.
MIGNIOT, C. 1968. A s-tudy of the physical properties of various forms
of very fine sediments and their behaviour under hydrodynamlc
action. La Houllle Blanche 7: 591-620.
NEDECO. (1965). A study of the siltatlon of the Bangkok Port Channel,
Vol 2, The field Investigation. Neth. Engrng. Consult., The Hague,
474 p.
, 1968. Surinam transportation study. Neth. Engrng. Consult., The
Hague, 293 p.
NICHOLS, M. 1985. Fluid mud accumulation processes 1n an estuary.
Geo-Mar. Lett. 4: 192-198.
. 1986. Effects of fine-sediment resuspenslon in estuaries. In:
Mehta AJ (ed) Estuarine cohesive sediment dynamics. Springer-Ver-
lag, New York, pp 5-42
NICHOLS, M., AND R. B. BIGGS. 1985. Estuaries. In: Davis R (ed) Coas-
tal sedimentary environments, Springer-Verlag, New York, pp 77-186.
278
-------
NICHOLS, M., G. S. THOMPSON, AND R. W. FAAS. 1978. A field study of
fluid mud dredged material: Its physical nature and dispersal. U.
S. Army Corps of Engineers, WES Tech. Rept. D-78-40, 74 p.
NICHOLS, M., R. HARRIS, 6. S. THOMPSON, AND B. W. NELSON. 1982. Metal
Inventory and fate of suspended sediment 1n Chesapeake Bay. Vat
Inst, Mar. Sc1. Spec. Sc1. Rept. 106, 97 p.
OFFICER, C. B. 1981. Physical dynamics of estuarlne suspended sedi-
ments. Mar. Geol. 40: 1-14.
OFFICER, C. B., D. R. LYNCH, G. M. SETLOCK, AND G. R. HELTZ. 1984
Recent sedimentation rates 1n Chesapeake Bay. In: Kennedy VS (ed)
The estuary as a filter, Academic Press, New York, pp 131-157.
OFFICER, C. B., R. B. BIGGS, J. L. TAFT, L. E. CRONIN, M. A. TYLER,
AND W. R. BOYNTON. 1984. Chesapeake Bay anoxia: origin, develop-
ment, and significance. Science 223: 22-27.
OLPHEN, H., VAN. 1963. Introduction to clay-colloid chemistry.
Intersdence Publ., New York, 318 p.
OWEN, M. W. 1970. Properties of a consolidating mud. Hyd. Res. Sta.
Rept. INT 83, 90 p.
PARKER, W. R., AND R. KIRBY. 1982. Time dependent properties of co-
hesive sediment relevant to sedimentation management - European
experience. In: Kennedy VS (ed) Estuarlne comparisons, Academic
Press, New York, pp 573-589.
PAZWASH, H., AND J. M. ROBERTSON. 1971. Fluid-dynamic consideration of
bottom materials. Jour. Hyd. Dlv. ASCE HY 9: 1317-1329.
PRITCHARD, D. W. 1971. Chemical and physical oceanography of Chesa-
peake Bay. The estuarlne environment. Estuaries and estuarlne
sedimentation. AGI short course lecture notes, IVA -1 to 28.
RHOADES, D. C. 1963. Rates of sediment reworking by Yoldla llmatula 1n
Buzzards Bay, Massachusetts, and Long Island Sound. Jour. Sed.
Pet. 33: 723-727.
. 1974. Organism-sediment relations on the muddy sea floor.
Oceanogr. Mar. B1ol. Ann. Rev. 12: 263-300.
RHOADES, D. C., AND D. K. YOUNG. 1971. Animal-sediment relations 1n
Cape Cod, Massachusetts II. Reworking by Molpadla oolltlca (Holo-
thuroldea). Mar. Blol. 11: 255-261.
SCHUBEL, J. R. 1968. Suspended sediment of the Northern Chesapeake
Bay. The Johns Hopkins Univ. Chesapeake Bay Inst. Tech. Rept. 35.
Ref. 68-2, 264 p.
279
-------
. • 1969. Size distributions of the suspended particles of the
Chesapeake Bay turbidity maximum. Neth. Jour. Sea. Res. 4: 283-309.
. 1971. Tidal variation of the size distribution of suspended
sediment at a station 1n the Chesapeake Bay turbidity maximum.
Neth. Jour. Sea. Res. 5: 252-266.
. 1982. An eclectic look at fine particles In the coastal ocean.
In: Klmrey K, R. Burns R (eds) Proc. of a pollutant transfer by
participates workshop. Seattle, MA. pp 53-142.
SCHUBEL, J. R., AND T. M. KANA. 1972. Agglomeration of fine-grained
suspended sediment 1n northern Chesapeake Bay. Powder Tech. 6: 9-
SCHUBEL, J. R., AND E. W. SCHIEMER. 1973. The cause Of acoustically
Impenetrable or turbid character of Chesapeake Bay sediments. Mar.
Geophys. Res. 2: 61-71.
SCHUBEL, J. R., AND W. H. CARTER. 1984. The estuary as a filter for
fine-grained sediment. In: Kennedy VS (ed) The estuary as a filter,
Academic Press, New York, pp 81-105.
SELIGER, H. H., J. A. BOGGS, AND W. H. BIGGLEY. 1985. Catastrophic
anoxia 1n the Chesapeake Bay 1n 1984. Science 228: 70-73.
SILLS, G. C., AND K. BEEN. 1984. Escape of pore fluid from consolida-
ting sediment. In: Parker WR, Kinsman DJJ (eds) Transfer processes
1n cohesive sediment systems, Plenum Press, New York, pp 109-125.
TERWINDT, J. H. J., AND H. N. C. BREUSERS. 1972. Experiments on the
origin of flaser, lenticular and sand-clay alternating bedding.
Sedlmentology 19: 85-98.
. 1982. Flume experiments on the origin of flaser bedding. Sedl-
mentology 29: 903-907.
VAN WAZER, J. R., J. W. LYONS, K. Y. KIM AND R. W. COLWELL. 1963.
Viscosity and flow measurements. A laboratory handbook of rheology.
Intersclence Publ. New York, 301 p.
WARTEL, S. 1973. Variations In concentration of suspended matter 1n
the Scheldt estuary. Bull. Inst. r. Scl. nat Belg. 49: 1-11.
WARTEL, S., G. VERREET, AND W. VIAENE. 1985. Internal structures 1n
mud sediments. (The occurrence of gas and Its Implications). In:
Van Grleken R, Wollast R. (eds) Progress 1n Belgian oceanographlc
research, Proc. of Symp, Brussels. The University of Antwerp Press,
pp 220-22.
WELLS, J. T. 1978. Shallow-water waves and fluid mud dynamics, coast
of Surinam, South America. Coastal Studies Inst. Tech. Rept. 257,
Louisiana State Unlv, Baton Rouge, LA. 56 p.
220
-------
WELLS, J. T., AND J. M. COLEMAN. 1981. Physical processes and fine-
grained sediment dynamics, coast of Surinam, South America. Jour.
Sed. Pet. 51: 1053-1068.
WOOD, A. B., AND D. E. WESTON. 1964. The propagation of sound 1n mud.
Acustlca 14: 156-162.
ZABAWA, C. F. 1978. Mlcrostructure of agglomerated suspended sediments
1n Northern Chesapeake Bay Estuary. Science 202: 49-51.
227
-------
Understanding the Estuary: Advances in Chesapeake
Bay Research. Proceedings of a Conference. 29-31
March 1988. Baltimore, Maryland. Chesapeake Research
Consortium Publication 129. CBP/TRS 24/88.
Comparison of Sediment Landscapes in Chesapeake Bay
as Seen by Surface and Profile Imaging
Robert J. Diaz and Linda C. Schaffner
Virginia Institute of Marine Science
The College of William and Mary
Gloucester Point, Virginia 23062
Contribution No. 1468 of the Virginia Institute of Marine Science.
INTRODUCTION
The sediment-water interface 1s the boundary layer between the water
column and sediments. It 1s Involved In virtually all processes and
cycles within aquatic and estuarine ecosystems. Interactions and
reactions at the sediment-water Interface are of particular importance
in regulating processes Involving nutrient regeneration-
remlneralization (Boynton and Kemp 1985), fate of toxicants (01 sen,
Cutshall and Larsen 1982), development of hypoxia-anoxia (Garber
1987), sediment mixing (Schaffner et al. 1987a, b), and sediment
transport (Wright et al. 1987). Much effort has and is being expended
to provide details of these processes which will eventually be used 1n
management plans for water quality, sediment quality, and fisheries
resources.
Generally, field methods for investigating sediment-water interface
processes or fluxes are time and labor Intensive. Complementary
methods are needed to suoport detailed studies and allow for better
comprehension of these dynamic processes. Rhoads and Cande (1971)
proposed the use of sediment profile cameras as a means of quickly
collecting data on the character of the sediment-water interface.
Rhoads and Germano (1986) outlined a scheme using sediment profile
cameras to assess the character of the sediment-water interface
relative to benthic community succession. Day, Schaffner, and Diaz
(1n press), in addition to using a sediment profile camera, also
advocated the use of bottom surface cameras in conjunction with the
profile camera to provide a more complete evaluation of the sediment-
water Interface.
Sediment profile and bottom surface cameras provide a unique in situ
view of the sediment-water Interface yielding both qualitative ali3
quantitative data on its biological, chemical, and physical character.
This in situ photographic approach and subsequent image analysis can
quickty ThlTcost effectively cover large areas of bottom defining
biological, sediment fabric, and energy gradients or other spatial
patterns. Natural or anthroprogenic events (i.e. storms, high flows,
dredged material disposal) through time can also be easily followed
and recovery rates measured.
222
-------
In this paper we will demonstrate the utility of using a surface and
profile Imaging camera system to provide a broad characterization of
the sediment-water interface from selected tributaries and mainstem of
the Chesapeake Bay. Emphasis will be placed on defining the redox
potential discontinuity and Its depth in the sediment relative to
biological and geochemical factors.
METHODS AND MATERIALS
A modified Benthos model 3731 sediment profile camera and Benthos
model 371 standard camera and 372 standard flash were combined into a
photographic system for evaluating sediment quality and benthic
habitat complexity. The sediment profile camera provides images of
the sediment column 15 cm wide and up to 20 cm deep. The profile
camera does not provide comprehensive resolution of surface features,
particularly if the prism penetration exceeds the optical axis of the
camera lens. The standard camera 1s used to provide information on
the surface by photographing an area approximately 20 x 30 cm in front
of the profile camera. In combination this Surface and Profile
Imaging (SPI) camera system provides a high resolution quick look Into
the character of the sediment water interface. The configuration of
cameras in the SPI system can be seen in Figure 1.
Data from 359 SPI Images collected 1n the Patuxent River, York River,
and Lower Chesapeake Bay (Fig. 2) between April 1986 and February 1988
were used in this evaluation of sediment landscapes. Each image was
analyzed using an International Imaging Systems 125 image processor
interfaced to a Prime 9955 computer. Of the 14 major parameters
measured from each image (Table 1) surface relief, depth of apparent
RPD, void area, and sediment grain size were selected for evaluation.
Surface relief 1s maximum point of prism penetration minus the minimum
point across the 15 cm width of the prism face plate. Apparent RPD
depth 1s the area of the Image visually discerned as being aerobic
divided by the width of the analyzed image.' We use the term apparent
in describing this parameter because no actual measure is made of the
redox potential. An assumption is made that, given the complexities
of iron and sulfate reduction-oxidation chemistry, the reddish-brown
color tones in sediments are Indications of sediments that if not
aerobic are not intensely reducing. This 1s In accordance with the
classical concept of RPD depth which associates 1t with sediment color
(Fenchel 1969). The area of an image occupied by voids and the type
of voids are good indications of subsurface biological and physical
processes. Void area is expressed as a percent of the total analyzed
Image area. All images are then standardized to a constant 15 cm
prism penetration to avoid over or under weighting images that were
less than or greater than 15 cm. Sediment grain size was estimated by
comparing each image to sediments of known grain size. Sediment types
followed the Wentworth classification as described 1n Folk (1974) and
represent modal class for each image.
The entire data set was stratified a posteriori by sediment type (as
described above), salinity at each Tocation (from Stroup and Lynn
1963), and depth (recorded at time of collection) (Table 2).
Broadscale patterns and trends were then evaluated using SPSSX (SPSS
1986).
223
-------
00
o
a;
-------
37 •
"oo1
NAUTICAL MILES
OS 10 IS 20 23
10 20 30 40 90
KILOMETERS
77« - 00'
Figure 2. Location of areas around the Chesapeake Bay from which SPI
data were collected.
225
-------
Table 1. Image analysis measurements from sediment profile camera photographs.
Measurement
Method
Usefulness
a - Depth of Penetration
b - Surface Relief
c • Digitized Image Statistics
1. Pixel densities for
total Image
2. Pixel densities for
areas of Interest
d - Depth of apparent RPD
Layer
e - Color Contrast of apparent
RPD
f - Area of Anoxlc Sediment
g - Area of 0x1c Sediment
h - Voids
1 - Other Inclusions
(Methane Bubbles, Mud
Clasts, Shells)
J - Burrows
k - Surface Features
1. Tubes
2. Ep1fauna
3. Pelletlzed Layer
4. Shell
5. Mud Clasts
1 - Sediment Grain Size
m - Dredged Material or other
Layers
Average of maximum and minimum
distance from sediment surface
to bottom of prism window.
Maximum minus n1m1mum depth of
penetration.
Actual range of densities the
digitizing camera detects from
the sediment profile Image.
Area of apparently oxlc layer
(g) divided by width Image.
Maximum and minimum distance
from sediment surface to top
of RPD layer are also measured.
Contrast between oxlc and
anoxlc layers 1s determined
from light Intensity level
density slicing of digitized
and specially enhanced Image.
Select desired pixel density
for boundary between oxlc and
anoxlc, count anoxlc pixels,
and convert to area.
As 1n f, except use oxlc
pixel count.
Number counted, depth from
surface of each measured,
area of each delineated.
Number counted, depth from
surface of each measured,
area delineated.
Number counted, area delineated.
Counted and spedated.
Counted and spedated.
Thickness and area delineated.
Qualitative estimate of coverage.
Qualitative estimate of coverage.
Determined from comparison of
Image to Images of known grain
size.
Measure thickness above original
sediment surface and area
delineated.
Penetration depth 1s a good
Indicator of sediment compaction.
If the camera 1s level, this 1s a
good measure of small scale bed
roughness, on the order of 15mm
(prism window width).
For cross comparisons of Images, 1t
Is necessary to have measurements
relying upon Image pixel density
done on a similar Intensity range.
Gives a good Indication of DO
conditions In the bottom waters
and the degree of blogenlc
activity In muddy sediments. In
sands will be related to porosity
and turbulence.
Establishes boundary of RPO.
Depending upon whether the RPD 1s
straight or convoluted will be of
use 1n understanding the biologic
and physical process.
When calculated to a constant depth
of penetration and combined with
oxlc layer area a good understanding
of RPO dynamics can be obtained.
When calculated to a constant depth
of penetration and combined with
anoxlc layer area a good under-
standing of RPD dynamics can be
obtained.
Presence of oxlc voids 1s a good
Indicator deep living fauna and
high blogenlc activity.
Often other Inclusions such as
methane or mud Clasts are Indicative
of certain processes and are helpful
1n understanding recent events.
Burrow presence Is a good Indica-
tion of deep living fauna and high
blogenlc activity.
Presence of these features 1s
Indicative of recent biological and
physical processes.
Provides modal estimate of grain
size and sediment layering.
Location of dredged material and
measuring Its thickness provide
quantitative measure for relating
Impacts to the benthos of any
disposal project.
226
-------
Table 2. £. posteriori strata definition by sediment type, salinity,
and depth.
Sediment strata (Wentworth Size Classes)
Clayey Mud
S1lty Mud
S1lt
S1lty Sand
Fine Sand
Fine-Medium Sand
Medium Sand
Salinity range (ppt)
0 to 5
5 to 15
15 to 20
20 to 25
>25
Depth Interval (feet)
15-30
30-45
45-60
>60
227
-------
RESULTS
The a posteriori stratification of Image data by sediment type,
salinity, and depth showed that most of the variation 1n surface
relief, apparent RPO depth, and percentage of void area could be
explained by sediment type alone. For example, the pattern of the
apparent RPD depth was similar with regard to sediment type by
salinity range (Fig. 3). Therefore, the data were restratified and
reanalyzed by only sediment type.
Surface relief
Surface relief tended to increase with increasing grain size (Fig. 4).
From clayey mud to silty sand the increase in surface relief was due
to biogenic activities of the benthic fauna. In sands the surface
relief was due to current generated bed forms. The magnitude of
surface relief in fine sediments averaged 0.7 cm in clayey mud to 1.1
cm in sllty-sand. This corresponds to surface slopes of 2.7° and
4.2°, respectively. Bed forms in sands averaged 1.4 to 1.7 cm in
height, or 5.3° to 6.5° in slope.
Apparent RPD depth
The depth of the apparent RPD, as measured by brown and reddish-brown
color tones of the sediment, tended to increase with increasing grain
size (F1g. 5). The higher mean value for RPO in clayey mud over silty
mud was due to several highly reworked low salinity stations. Median
values for the apparent RPO were the same for both of these sediment
types (0.5 cm). The increase in RPD depth in silt and silty sand was
due to biogenic reworking of sediments by Infauna. In sand sediments
porosity was the major determinant of RPD depth.
The thin apparent RPD depths 1n clayey and sllty mud sediments were
clearly defined from the grey color tones of the subsurface sediments.
Apparent RPD layers less than 1 cm thick in muddy sediments, while not
smooth, were more uniform than deeper RPD layers. The complexity in
the form of the RPD was highest in silt and silty sand sediments from
biogenic activities of infauna. In sands the apparent RPD was
simplest in form being close to a uniform surface between aerobic and
anaerobic sediments.
Percent void area
The average and median percentage of void area, standardized to 15 cm
of prism penetration, was low. Void area in fine and predominantly
fine grained sediments averaged 1.3 to 2.1% with median values being
much less at 0.0 to 0.8% for the same sediments (Fig. 6). In sands
voids were not major subsurface features. At times voids do occur 1n
sands, but they tend to be small. In fine'sediments about 15% of the
Images have voids that were much larger than average, being up to 22%
of the sediment area. The majority of these large voids appeared to
be active biogenlc structures from subsurface deposit feeding. Except
in clayey muds many of the largest voids resulted from physical
cracking of the sediment caused by the camera prism.
DISCUSSION
Sediment landscapes in the Chesapeake Bay exhibit broadscale patterns
related mainly to sediment grain size and secondarily to salinity,
which are a primary determinant of the character of infaunal
228
-------
MEAN OF RPD DEPTH
oo
(u o
9 m
0,-a
rt
w D*
n
o, o
H. Hi
fD rt
3 Sf
rt (0
rt to
n to
• n
ID
D
rt
O
V
M>
m
o
09
(C
CO
M
0)
N
O
g
(D
H
N
O
m
CO
o
o
o
CJi-
!i
N
CJi
01
»
_^ O>
CJI •
o
•*
10
°I
_' a
V 8
ro»
en 3
m
m
D
229
-------
SURFACE RELIEF (CM)
00
Hi
fD
fD W W
3 0) C
Q, H H
Hi
cr H- fu
&) c» n
i^ ft
re M
to Cu ft
l-t H- H
ft (U H-
3 ft
rt - HI
(u
3
era
re
H,
H
O
g
3 O
• Hi
cr1
O
(D
n
Q>
§
fD
H
(U
C (JQ
w m
l-t CO
rt »
K*
M CT1
fD •<
H CO
cu rti
3 a
OQ H-
ro a
• ro
3
Qi rt
3
O. rt
O
no
GO
CO
m
m
m
'%.
230
-------
DEPTH OF RPD (CM)
o
00
to
H
n>
rt
0
rt
to
M
H
§
OQ
IB
•
Figure 5. Depth of apparent RPD, from profile camera images, by sediment type
Bar is median, x is mean, box is interquartile range, and end bars
V
V
O)
•MMB ^^^ '
5\.
H
\
\.
%.
-------
PERCENT VOID AREA
o
no
00
3 n m
a 3 H
(D O
n rt n
3 H 3
O. to rt
rt 01
cr H-TO
to o ro
M 3 ^
CO • O
Hi
0)
H W H-
n> to B
i-l to
rt 00
o H- m
rt ca
ro
a.
H>
3 M Pf
TO
n>
3 3*
• 0)
CO CO
a <
(B O
(U H-
3 CL
« CO
«
o en
X re
0>
H- 3
CO &
to
n N
H (0
J3 O.
to ft
H O
rt
H- M
M tn
ID
Hi
8*0
oo M
(B H.
' §
C/)
m
_t i ft I I
CD
PD
03
OD
no
o
2J2
-------
communities. Within each salinity zone, as defined, the basic
patterns of surface relief and apparent RPO depth were similar by
sediment type. At salinities above 5 ppt patterns in void area by
sediment type were also similar. At salinities less than 5 ppt
functional groups of infauna capable of producing subsurface feeding
voids are limited in abundance (Schaffner et al. 1987a). See Figure 7
for representative images.
In sediments ranging from mud. to silty sands, the complexity of
surface relief and apparent RPD depth increases with increasing grain
size. This is due mainly to the increasing dominance of infauna in
sediment mixing processes along this sediment gradient (Schaffner et
al. 1987b). With the transition to sand sediments physical forces
dominate surface relief and RPO depth. In sands, bed forms are the
predominant surface relief and the apparent RPO layer tends to be more
uniform, not following the surface contours provided by bed forms.
Apparent RPO layers in clayey and silty muds tended to be broadly
uniform, following the contour of the surface sediments, upon which a
smaller scale (on the order of mm's) convolution is superimposed. In
silts and silty sands the apparent RPO is most complex and convoluted
providing a greatly increased biologically reactive interface.
The degree of biogenically-induced structural complexity in Chesapeake
Bay surface sediments, as documented by surface and profile imagery,
might have important effects on cycling of dissolved and particulate
substances at and through the sediment water interface. For example,
consider the processes associated with geochemical cycling across the
RPD layer. While flux rates are typically based on simple area!
measurement and the RPD is considered to be a simple contact plane
between aerobic and anaerobic environments (Fenchel 1959), over most
of the Chesapeake Bay's sediment landscape this assumption would lead
to an underrepresentation of the actual area of the RPD layer. The
results of numerous studies clearly demonstrate that biogenic
structures are regions of enhanced biological and geochemical activity
(Aller 1982, Aller and Yinst 1978, Aller and Aller 1986) and that the
activities of infaunal organisms can increase flux across the oxic-
anoxic sediment interface (Henriksen, Hanson and Blackburn 1980, Aller
and Yinst 1978). Our documentation of the apparent RPO layer, a
complicated surface much greater in actual area than a simple areal
measurement would estimate, strongly suggests the need for further
evaluation of the effects of infaunal benthos on sediment-water
interface flux processes in the Chesapeake Bay.
CONCLUSIONS
There are broadscale patterns in the sediment landscapes of the
Chesapeake Bay with regards to data collected by surface and profile
imaging. General trends noted are:
- Biogenic voids are common and an integral part of sediment
structure, except in sand and tidal freshwater and oligohaline
habitats.
- Surface roughness increases concordant with increasing grain size.
In fine grain sediments roughness is primarily biogenic and best
developed in silts and silty-sands. In sands roughness is from
current generated bed forms.
233
-------
Figure 7. Examples of sediment profile images. Scale is IX. S -
sediment water interface, T - worm tube, V - feeding void.
a. 2 cm high bed form in medium sand at 5 m depth off Cape
Charles in the Lower Chesapeake Bay.
b. Muddy sediments off Broome's Island, Patuxent River,
showing thin (less than 1 cm) apparent RPD. Notice
highly mottled appearance of subsurface sediment which
may result from biogenic mixing. Also notice
polychaete tubes at surface of sediment.
c. Silty sediments along Eastern Shore south of Cape
Charles at 22 m depth. Apparent RPD is deeply
convoluted and along the right of the image it extends
down below the penetration of the camera prism. This
type of apparent RPD is due to biogenic reworking by
deep dwelling fauna. Surface relief in this image is
all from biogenic activities. Notice small polychaete
tubes at the surface.
d. Silty sediments near York River entrance channel at 10
m depth. Apparent RPD is deep in sediments and
convoluted from biogenic activities. Large void is
from head down deposit feeding of maldanid polychaetes.
234
-------
-------
7B
-------
7C
-------
7D
-------
- The biologically reactive interface, as represented by the apparent
RPD, is greater than predicted by surface area alone. Deepest and
most complex RPO's are found in silts and silty-sands at meso- and
polyhaline salinities.
- Except when very thin (< 0.5 cm) and there is no deep biogenic
activity, or in sand sediments, the apparent RPO layer is not a
simple contact plane between aerobic and anaerobic environments.
The actual RPO area could be many times that described by simple
surface area.
REFERENCES
Aller, R. C. 1982., The effects of macrobenthos on chemical
properties of marine sediments and overlying waters. In P. L.
McCall and M, 0. S. Tevesz (eds.), Animal-Sediment Relations:
The biogenic alteration of sediments. Plenum Press, NY. p. 53-
102.
Aller, R. C. and J. Y. Yinst. 1978. Biogeochemistry of tube-
dwellings: a study of the sedentary polychaete Amphitite ornata
(Leidy). Journ. Mar. Res. 36:201-254.
Aller, J. Y. and R. C. Aller. 1986. Evidence for localized
enhancement of biological activity associated with tube and
burrow structures in deep-sea sediments at the HEBBLE site,
western North Atlantic. Deep-Sea Res. 33:755-790.
Boynton, W. R. and W. M. Kemp. 1985. Nutrient regeneration and
oxygen consumption by sediments along an estuarine salinity
gradient. Mar. Ecol. 23:45-55.
Day, M. E., L. C. Schaffner, and R. J. Diaz, (in press). Long Island
Sound sediment quality survey and analyses. NOAA Tech. Memor.
NOS OMA, 113 pp.
Fenchel, T. 1969. The ecology of marine microbenthos IV. Structure
and function of the benthic ecosystem, its chemical and physical
factors and the microfauna communities with special reference to
the ciliated Protozoa. Ophelia 6:1-182.
Folk, R. L. 1974. Petrology of sedimentary rocks. Austin, Texas,
Hemphill's. 170 p.
Garber, J. H. 1987. Benthic-pelagic coupling in the Chesapeake Bay.
In M. P. Lynch and E. C. Krome (eds.), Perspectives on the
Chesapeake Bay: Advances in estuarine sciences. CRC Pub. No.
127. Chesapeake Research Consortium, Gloucester Pt. VA. p. 11-
34.
Henriksen, K., J. I. Hansen and T. H. Blackburn. 1980. The influence
of benthic infauna on exchange rates of inorganic nitrogen
between sediment and water. Ophelia. Suppl. 1:249-256.
Rhoads, D. C. and S. Cande. 1971. Sediment profile camera for in
situ study of organism-sediment relations. Limnol. Oceanogr.
TFTTlO-114.
239
-------
Rhoads, 0. C. and J. 0. Germane. 1986. Interpreting long-term
changes in benthlc community structure: a new protocol.
Hydrob1olog1a 142:291-308.
Olsen, C. R., N. H. Cutshall and I. L. Larsen. 1982. Pollutant-
particle associations and dynamics 1n coastal marine
environments: A review. Mar. Chem. 11:501-533.
Schaffner, L. C., R. J. D1az, C. R. Olsen and I. I. Larsen. 1987.
Faunal characteristics and sediment accumulation processes 1n the
James River estuary, Virginia. Estuarlne Coast. Shelf Scl.
25:211-226.
Schaffner, L. C., R. J. Diaz and R. J. Byrne. 1987b. Processes
affecting recent estuarlne stratigraphy op. 534-599. In Coastal
Sediments '87, Waterways 01v., ASCE, New Orleans, LA.
SPSS. 1986. SPSSX user's guide, 2nd ed. SPSS Inc., Chicago, IL, 988
pp.
Stroup, E. 0. and R. J. Lynn. 1963. Atlas of salinity and
temperature distributions 1n Chesapeake Bay. Chesapeake Bay
Institute Report 2, 410 pp.
Wright, L. 0., 0. B. Prior, C. H. Hobbs, R. J. Byrne, J. 0. Boon,
L. C. Schaffner and M. 0. Green. 1987. Spatial variability of
bottom types 1n the lower Chesapeake Bay and adjoining estuaries
and Inner shelf. Estuarlne Coast. Shelf Sd. 24:765-784.
240
-------
Understanding the Estuary: Advances in Chesapeake
Bay Research. Proceedings of a Conference. 29-31
March 1988. Baltimore, Maryland. Chesapeake Research
Consortium Publication 129. CBP/TRS 24188.
Use of In Situ Nutrient Addition and Dilution Bioassays to Detect
Nutrient Limitation in the Tidal Freshwater Potomac
R. Christian Jones
George Mason University
Department of Biology
Fairfax, Virginia 22030
INTRODUCTION
Nuisance blooms of blue-green algae have regularly Invaded the tidal
freshwater Potomac and its embayments since 1983. Many of these blooms
have been concentrated in the river stretch between Marshall Hall (River
Mile 22) and Quantico (RM 40). Due to the expense of major upgrades of
waste treatment in the Washington area, these unpredicted blooms have
been a major cause of concern in water quality management. A blue-ribbon
panel commissioned following the 1983 bloom concluded that the river
system was phosphorus-limited and that the 1983 bloom was due to
phosphorus that became available not from current waste discharges, but
from an unexpected source, the river's sediment (Thoraann et al. 1985).
The mechanism for this release was hypothesized to be enhanced pH due to
photosynthetic activity, a positive feedback mechanism. Similar, but
more limited blooms have been experienced in 1985 and 1986.
Fundamental questions remain to be answered. Is phosphorus always
limiting algal growth in the river? Are carbon or light limitation
important? What about the role of climatic factors? If the system is not
limited by phosphorus, how much phosphorus must be removed before
phosphorus limitation and subsequent water quality improvement takes
place?
Nutrient addition bioassays have been used by numerous workers as a
method of determining growth limiting factors in nutrient-limited
situations. In these experiments soluble nutrients (usually N and P) are
added to closed containers of receiving water containing either native
algae or representative cultures and growth is measured over a period of
time against unspiked controls. Experimental designs have ranged from
laboratory studies with single species growing in filtered, autoclaved
water (Miller et al. 1978) to field designs in which nutrients are added
to containers of native phytoplankton and incubated in the field (Goldman
1963, 1978). In very eutrophic systems where nutrients are not limiting,
241
-------
little information can be gleaned from nutrient addition bioassays.
Paerl and Bowles (1987) have proposed that nutrient dilution bioassays
are needed in these rich waters to determine levels to which nutrients
must be diluted before nutrient limitation can occur. It is only after
nutrient limitation is attained that incremental removal of nutrients
will result in decreases in algal crops.
The phytoplankton is a community, that is an assemblage of species.
Each species has its own growth characteristics and each may respond
differently to environmental characteristics. In addition, they compete
for scarce resources and as these resources change so too do the
competitive relationships among the species. Based on these
considerations it is not hard to visualize that complex responses could
occur when communities are confined in enclosures and pulsed with
nutrients. Thus, in a study such as this it is useful to consider not
only the aggregate responae (chlorophyll, photosynthetic rates, total
cell density), but also the responses of various species.
The purpose of this study was to conduct nutrient addition and
nutrient dilution bioassays on Potomac river water from 2 sites (Gunston
Cove and Indian Head) on two dates in September 1986. The method used
was a modification of that of Paerl and Bowles (1987) with changes as
noted below. Three questions were to be addressed:
(1) Are phytoplankton at these sites nutrient limited?
(2) If not, how much nutrient must be removed (in this case, diluted)
to attain nutrient-limited status?
(3) Is there a difference in response among different major groups or
even species of phytoplankton and do these differences help
account for the aggregate response of the community? Do
"nuisance" species and "desirable" species of phytoplankton differ
in their response?
METHODS
Study sites were located in the tidal fresh portion of the Potomac
River below Washington, D.C. One site identified as GC was located in
Gunston Cove, a shallow embayment on the Virginia side of the river about
13 miles downstream of the Woodrow Vilson Bridge. The second site was
located in the river raainstetn off Indian Head about 19 miles downstream
of the Wilson Bridge. This station was situated near the river channel
in a stretch of the river in which blue-green algal blooms have developed
in recent years* Experiments were conducted at each station during early
and late September 1986.
At each site twelve polyethylene containers (5-gallon Cubitainers)
were used. Six treatments were established in duplicate as follows:
control, three levels of phosphorus addition (+P, +10P, +11P), 10%
dilution with synthetic river water (-10%), 50% dilution with synthetic
river water (-50%). Phosphorus additions were to be made by adding 20 ml
of a stock solution containing 0.56 g K2HP04/L (+P, +11P) and/or 6.07 g
K2HP04/L (+10P, +11P) giving a final added P concentration of 0.1 mg P/L
(+P), 1.1 mg P/L (+10P), or 1.2 mg P/L (+11P). Dilutions were performed
using a synthetic river water made up with ACS grade laboratory chemicals
to simulate the major ionic composition of Potomac River water (Jones
1987).
242
-------
Nutrient and synthetic water additions indicated above were made
prior to filling of the bags with river water. In the field river water
was pumped from a depth of 0.3 m (1 ft) directly into the bags in the
following order: CTRL, +10P, +P, +11P, -10%, -50%. After filling and
mixing the contents of each bag well, water was poured from each for
analysis. Three samples were taken from each cubitainer: one for
nutrient and pH analysis, a second for chlorophyll and photosynthetic
rate determination at GMU, and a third for phytoplankton analysis. The
phytoplankton sample was stored in an amber glass bottle to which 2 ml of
acid Lugol's iodine solution was added, while the other two samples were
held in 500 ml polyethylene bottles in a dark, insulated container at
ambient temperature.
After sampling each bag was sealed with its plastic screw lid,
lowered Into the water, and attached with a metal clip to an eye screw on
a floating wooden frame. After two days, each bag was removed from the
water, mixed thoroughly, and sampled as on Day 0. During the first
experiment several bags were lost at the Indian Head site due to exposure
of the apparatus to constant wave action. The site was also a long run
from the launching ramp in Gunston Cove and we could not monitor it every
day. Thus, for the second experiment bags were filled at Indian Head,
covered, and moved to a station in the outer part of Gunston Cove near
the Coast Guard station for incubation.
Samples for chlorophyll and photosynthetic rate were returned to
George Mason University within a few hours of collection. Chlorophyll was
measured by filtration, extraction with acetone and DMSO, and
fluorometric determination of pigments (Jones 1987). Light saturated
photosynthetic rate was measured using C-14 uptake in the lab at ambient
temperature under artificial illumination (Jones 1987). Phytoplankton
were enumerated by species -using the inverted microscope-settling chamber
technique (Lund et al. 1958, Jones 1987). At least 300 cells (1500 in GC
samples) were counted. To minimize variation between counters, one person
counted all GC samples (Ms. Vicki Andrele) and another person all IH
samples (Dr. Claire Buchanan). Biovolumes were determined for dominant
species by measuring cell dimensions and approximating cell shape to an
appropriate geometric solid for which volume can be easily calculated.
It should be noted that cells may vary in size and that small changes in
dimensions lead to large changes in cell volume. Nutrient analysis was
conducted by Lower Potomac Pollution Control Plant lab.
Changes in each parameter were subjected to one-way analysis of
variance to determine if treatment means were significantly different
from one another. The Tukey HSD test was used to compare treatment means
with the control. The SYSTAT statistical package was used to conduct all
statistical tests.
RESULTS
Gunston Cove
At the beginning of the September 2-4 experiment in Gunston Cove
enclosures reflected the large algal populations present in the cove.
Initial chlorophyll levels in undiluted treatments (+P, -HOP, -flip, and
CTRL) were 233-295 ug/L indicating bloom conditions. Photosynthetic
243
-------
rates were 1079-1367 ug C/L/hr in undiluted treatments. Chlorophyll and
photosynthetic rate were proportionately lower in the -10% and -50%
dilutions. Blue-green algae were the dominant phytoplankton group in
terms of cell density comprising 84% of all cells counted. The most
abundant species was Merismopedia tenuissima comprising 60-80% of the
total count in initial samples. Chroococcus dispersus and Oscillatoria
planctonica were found in substantial numbers (2.8-19.8%). The
bloomformer Microcystis aeruginosa was present, but not abundant (0.4-
2.1% of total cells). Cryptophytes composed about 7% of the total cell
density. When biovolumes were computed, cryptophytes became dominant
comprising about two-thirds of the total biovolume. This was due to
their much larger size when compared to the blue-greens.
After two days cell density increased in all treatments (Figure la).
Change in cell density was significantly different among treatments, but
no treatment was significantly different from the control. The increase
was greatest in the control and +P treatments and least in the +10P
treatment. Changes in biovolume were not significant among treatments.
After two days chlorophyll values had decreased in all treatments. In
the control, chlorophylls decreased by 22%. Similar decreases -were
observed in all other treatments except -50% where little change in
chlorophyll was observed. None of the responses in chlorophyll were
significantly different from the control. Photosynthetic rates in the
control decreased 24%. All treatments to which P had been added showed
increases in photosynthetic rate of from 4 to 44% which were
significantly greater than the control. Dilutions showed decreases in
photosynthetic rate which were not different from the control.
Green algae increased in most treatments, but showed no significant
differences among treatments (Figure Ib). Diatoms showed little change
under most treatments, but were significantly enhanced in the -50%
dilution. Cryptophytes increased slightly in the control and decreased in
all three nutrient addition treatments, but we.re much higher in the -50%
dilution. Although not significantly different than the control, the
-50% dilution did result in significantly more cryptophytes than the
nutrient addition treatments. The blue-greens were the group most
strongly responsible for the observed increase in cell density.
Merismopedia tenuissima continued to be the overwhelming dominant
comprising 45-62% of cells counted in all treatments. Chroococcus
dispersus and Oscillatoria planctonica remained subdominant with 7-26% of
cells counted. Microcystis aeruginosa density increased at about the
same rate as the community as a whole retaining a 1-2% share of total
cells.
Initial conditions at Gunston Cove during the September 23-25
experiment were similar to those observed earlier in September. Initial
chlorophyll levels in the undiluted treatments were even higher than
those found in early September ranging 317-358 ug/L. Photosynthetic rates
were also higher: 1845-2041 ug/L. Chlorophyll and photosynthetic rate
were proportionately less in dilution treatments. Initial phytoplankton
densities in the late September experiment were similar to those observed
in early September (300,000-400,000 cells/mL) as was species composition.
Blue-greens were again the most abundant group at the beginning of the
experiment comprising 82% of total algal cells. As in early September
Mertsmopedia tenuissima was the dominant species accounting for 44-63% of
all cells counted. Chroococcus dispersus and Oscillatoria planctonica
244
-------
GUNSTON COVE SITE
140
120
80
60
40
20
0
-40
0.
SEPTEMBER 2-4, 1986
NS 0.05 O.01
L
-
I I I I I I I I I I I I I I
TREATMENT
220
180
ft 140
5 100
i—
5 eo
tf 20
-20
NS
O.O5
O.O5
JL05
TREATMENT
Figure 1. Response of phytoplankton in in situ bioassays. Gunston Cove,
September 2-4. 1986* a. (upper) Percent change in cell density (cells/mL),
biovolume (um3/L)> chlorophyll (ug/L), photosynthetic rate (ug/L/hr). b.
(lower) Percent change in density of green algae, diatoms, cryptophytes, and
blue-greens (cells/ml). Six bars for each parameter indicate response in
control, +P, -HOP, +11P, -10% dilution, and -SOX dilution respectively.
Numbers across top of graph indicate significance of one-way ANOVA among
treatments for each parameter. + and - connote treatment means significantly
different from the mean for each parameter.
245
-------
continued to be subdoralnant (5-25%). Again Microcystis aeruginosa was
present, but not particularly abundant (0.5-1.3%). In terms of algal
biovolume, cryptophytes were again the most important group comprising
71% of the total.
Cell density increased significantly in two of the three nutrient
addition treatments; nutrient dilution treatments showed little change
relative to the control. Biovolurae increased in most treatments, but was
significantly greater than the control only in +10P. Chlorophyll levels
decreased in almost all treatments during the late September GC
experiment. Decreases of 14-16% were observed in the controls with
similar declines in most other treatments except -50% where a small
increase was measured. Photosynthetic rate decreased in all treatments.
The decrease in the controls was 26-29% while that in the nutrient
additions was 2-18%. More substantial declines were observed in the
nutrient dilution treatments.
Greens increased in all treatments especially those with large
additions of P with positive changes of up to 200% being found (Fig. 2b).
Scenedesmus btjuga was the most abundant chlorophyte accounting for as
much as 2% of the total count. Diatoms did poorly in all treatments,
decreasing 45-80% in all nutrient addition and control treatments and
somewhat less in most dilution treatments. Cryptophytes declined 0-40%
in all treatments except -50% dilution, where a significant increase
(about 80%) was registered. Blue-greens remained the dominant group and
followed trends similar to those in total cell counts. Increases were
significantly greater than the mean in two of three nutrient addition
treatments. Merismopedia tenuissima remained the dominant comprising 34-
54% of all cells counted, a slight drop in dominance. Chroococcus
dispersus and Oscillatoria planctonica were subdominant (6-30%).
Microcystis aeruginosa increased slightly in dominance to 1.7-2.7%.
Indian Head
Initial conditions for the September 23-25 experiment at Indian Head
reflected much lower algal standing crop. Chlorophyll levels were
similar in all undiluted samples at the beginning of the experiment
ranging from 29-34 ug/L. Photosynthetic rates were varied from 290-370 ug
C/L/hr. Dilution treatments showed proportionately lower chlorophyll and
Photosynthetic rate values. Initial cell densities were 60,000-90,000
cells per mL, blue-greens comprising 46%. Diatoms, greens, and small
flagellates were also important. Merismopedia tenuissima (15-36%),
Chroococcus dispersus (9-13%), and Oscillatoria planctonica (4-12%) were
the dominant blue-green species. Discoid centric diatoms were also a
dominant group numerically (20-24%). Diatoms were dominant in biovolume
(39%) followed closely by the unidentified flagellates.
Phytoplankton densities increased strongly in all treatments (Fig.
3a) with highest increases of about 500% in the +P treatment
(significantly greater than the control) and lowest increase of 100-200%
(significantly less than the control in -50% dilution. Phytoplankton
biovolume followed a similar trend with increases found in all treatments
and significant relationships with the control Identical to those in
densities. The greatest increase in biovolume was 600% in +P and the
least was slightly less than 200% in the -50% treatment. After 2: days
chlorophyll increased in all samples, but none were statistically
246
-------
aiNSTON COVE SITE
SEPTEMBER 23-25, 1986
O.OOS 0.05
1UU
80
40
2U
-20
-40
£/S
—60
K
-
-
•
-
•
1
f
§
I
i
1
•
RJ
I
i
1
t
I
I
1
1
•
+
I | +
gg Lb I %
JW"}
t
&
1 1 1 1 1 1 1 [ 1 1 1 1 1 1 1 I 1 1 1 1 1
mm*
\
\
T
TREATHCMT
200
160
N 120
O 80
I 48
ij
0
-40
-80
0.001 0.05
0.05
0.01
UDa_.
TREATKCNT
Figure 2. Response of phytoplankton in In situ bloassays. Gunston Cove,
September 23-25, 1986. a. (upper) Percent change In cell density (cells/nL),
biovolume (um3/L), chlorophyll (ug/L), photosynthetic rate (ug/L/hr). b.
(lover) Percent change in density of green algae, diatoms, cryptophytes, and
blue-greens (cells/mL). Six bars for each parameter indicate response in
control, +P, +10P, -HIP, -10Z dilution, and -50t dilution respectively.
Numbers across top of graph indicate significance of one-way AMOVA among
treatments for each parameter. + and - connote treatment means significantly
different from the mean for each parameter.
247
-------
INDIAN HEAD SITE
EPTCMBER 23-25. 1986
600
Cf\f\
500
I iJ A r"\A
$ 4OJ
JUU
1—
6 ^00
& 1 UO
0
A />f"\
-100
•
1
i
1
+
1
!
1
1
i
•
1
1
1
i
£
s
1
s
•
1
1
••
5
i
i
•
i
•
i
.<
'i
1
I
•H
KJ
i
«,
^
$
^
I
i
1^ nnmPln
•i ^ ^ P/ ,; /
HI H ^ii ^
K9 fifl
1 1 1 ! 1 1 1 1 1 1 1 1 1 1 ! 1 1 1
CELL DEN
E40VOL
CHIOP
PSYN
TREATMENT
a:
0.001
600
500
400
300
0
1iY>
0.001 0.01
-1
1
i
n
-
g
ii
r-
., IL
NS
B-cn—
;
i
:
-
i
GPEEf-iS
DIATOMS
FLAG
BGA
TFEATMENT
Figure 3. Response of phytoplankton in in situ bioassays. Indian Head,
September 23-25, 1986. a. (upper) Percent cTumge in cell density (cells/ml),
biovolume (um^/t), chlorophyll (ug/L), photosynthetlc rate (ug/L/hr). b.
(lower) Percent change in density of green algae, diatoms, small flagellates,
and blue-greens (cells/mL). Six bars for each parameter indicate response in
control, +P, +IOP, +UP, -10X dilution, and -SOX dilution respectively.
Numbers across top of graph indicate significance of one-way ANOVA among
treatments for each parameter. * and - connote treatment means significantly
different from the mean for each parameter.
248
-------
different from the control. Photosynthetlc rate decreased markedly in
the controls and in the -10% dilution treatment. Other treatments showed
little consistent change from the initial readings; no changes were
significantly different from the control.
All major groups increased in all treatments (Fig. 3b). Increases in
chlorophytes were much lower at +10P and -10% dilution than in other
treatments, being significantly less than the control. Diatom increases
were highest in +P (500%, significantly greater than the control) and
were generally 200-400% in other treatments. Discoid centrics increased
as overall numbers climbed maintaining 23-27% of cell numbers and
probably a greater percentage of overall biomass. Blue-green increases
were similar to those in total phytoplankton with greatest increase in +P
(significantly greater than the control) and lowest increase in -50%
dilution (100-200%).Mertsmopedia tenuissima remained the dominant blue-
green (29-39%); Chroococcus dispersus (7-11%) and Oscillatoria
planctonica (5-10%) were subdominant cyanophytes.
DISCUSSION
Nutrient addition bioassays In various forms have been used by
numerous investigators to determine which nutrient element is limiting
algal growth in natural waters. These tests assume that if X is the
limiting nutrient, then addition of X to a water sample containing algae
will cause an increase in algal growth and/or standing crop. If X is not
limiting, then no response will occur. In hypereutrophic waters, all
nutrient elements may be present in excess. In these cases algal growth
is probably limited by carbon or light and nutrient levels in receiving
waters may have to be reduced markedly before any detectable change in
algal activity is observed. Thus, expensive nutrient removal schemes may
be implemented without a detectable response in algal bioraass. Nutrient
dilution bioassays (Paerl and Bowles 1987) offer a means of ascertaining
to what levels nutrients must be reduced to achieve nutrient limitation;
at this point further nutrient reductions will have proportional effects
on algal biomass.
In this work nutrient addition and nutrient dilution bioassays were
both utilized in situ to detect the degree of nutrient limitation at
Potomac River embayment and mainstem stations subject to algal blooms.
During the experiment an algal bloom was in progress at the embayment
site in Gunston Cove with chlorophyll levels in, excess of 200 ug/L. The
bloom in was composed of cyanophytes Chroococcus and Merismopedia with a
substantial component of green algae and cryptophytes compared with
previous Gunston Cove blooms dominated by Microcystis aeruginosa. At the
Indian Head site chlorophyll levels (30-50 ug/L) were well below bloom
levels observed there in some previous years. Merismopedia, Chroococcus,
and Oscillatoria were the dominant blue-greens. Discoid centric diatoms
were very important in the river. Other parameters reflected the
magnitude of algal standing crops. Initial pH at Gunston Cove was highly
elevated (9.0-9.6) compared with Indian Head (7.6-7.9).
It is important to recall the differences in depth at the two sites.
At the GC site with water only 1.5 m deep, the algae circulate only
through a shallow, near-surface water column. At the IH site,
phytoplankton circulate through 5-10 m of water column. Thus the IH
249
-------
phytoplankton spend more time in the dark deeper waters and thus would be
more likely to be light-limited. Light limitation could occur even at
the GC site if large phytoplankton populations build up. One result of
enclosing phytoplankton in bags is that they spend more time near the
surface and probably experience higher light levels. Another result is
that being enclosed, turbulence is decreased and increased sedimentation
may occur. In addition decreased turbulence may lead to an increase in
the boundary layer around individual cells and more difficulty in
obtaining nutrients. Finally, enclosure with only a small amount of air
and decreased surface turbulence may lead to carbon limitation as carbon
dioxide is depleted from the enclosed water and air.
Phytoplankton communities at Indian Head responded very favorably to
enclosure. Chlorophyll increased in all treatments and photosynthetic
rate responded favorably in the first experiment. Total cell density and
biovolume and the density of all major groups also increased in all
treatments at Indian Head. This response probably reflects increased
light availability when the algae are held in an enclosure near the
water's surface as opposed to circulating over the entire water column.
Since algal crops were not excessive and final pH was less extreme (less
than 9.4), carbon would be a less serious limiting factor than at Gunston
Cove.
Indian Head communities responded positively to +P addition,
although the response to +10P and +11P was not statistically significant.
Cell density, biovolume, diatoms, and blue-greens all increased
significantly faster in the +P treatment than in the controls. Indian
Head communities responded negatively to -50% dilution with cell density
and biovoluroe significantly less than the controls. In addition to data
reported here, particulate phosphorus and organic nitrogen also increased
faster in nutrient addition treatments than in the control (Jones 1987).
Phytoplankton communities at Gunston Cove did not respond as
favorably to enclosure as did those at Indian Head. In almost every
case, chlorophyll levels decreased during the incubation period. The
water column at Gunston Cove is much shorter and thus algae are already
spending a large portion of their time near the surface. Negative
factors associated with enclosure may have overwhelmed any slight light
advantage. Due to the higher standing crops already present in the bags,
dissolved C02 was exhausted much more rapidly in the enclosure. At its
extreme in this study, dissolved C02 reached a low of 0.034 umoles/L in
GC enclosures (as compared with a low of 0.65 uM at Indian Head). This is
well below the level thought to end photosynthesis by green algae and
would be expected to severely stress even blue-greens (King 1970). On the
other hand cell densities increased in virtually all treatments in both
GC experiments. This increase was largely accounted for by very small
cyanophytes. When algal biovolume was computed, this strong positive
effect of enclosure disappeared. Large-bodied algae, particularly
cryptophytes and diatoms, were negatively affected by enclosure with
nutrient addition and caused biovolume to decline in many cases- Numerous
other studies have shown that low dissolved C0£ concentrations and high
pH inhibit photosynthesis by greens, diatoms, and other algae and enhance
the dominance of cyanophytes (Shapiro 1973, King 1970, Tailing 1976, Moss
1973). Interestingly, cryptophytes and diatoms were actually stimulated
by nutrient dilution in the Gunston Cove samples. This may also indicate
the effect of carbon limitation which is alleviated by the high
250
-------
alkalinity and moderate pH of the synthetic river water and the lower
photosynthetic activity in the diluted medium. Chlorophyll behaved
similarly to large-bodied forms while photosynthetic activity mirrored
changes in small cyanophytes. This is consistent with the general
observation of a negative correlation between size and growth rate
(Reynolds 1984).
Photosynthetic rates in GC experiments often increased more or at
least decreased less in nutrient addition treatments compared with the
controls. In the first experiment a strong increase was found with
nutrient addition when compared with the controls. In the second
experiment the decrease in photosynthetic rate was less in the nutrient
additions than in the controls. Chlorophyll levels were less supportive
of the nutrient stimulation hypothesis particularly in the second
experiment. Cell densities were consistently higher in nutrient enriched
cultures in the first experiment, but not the second. Cell biovolume was
negatively affected by nutrient addition in the second experiment; no
trends were obvious in the first experiment. Increases in particulate
phosphorus and organic nitrogen were consistently and often significantly
greater in nutrient addition treatments than in the control (Jones 1987).
The stronger response to nutrient addition in the second GC experiment
may have reflected an increase in P limitation as the bloom proceeded. As
discussed above a clear response to nutrient addition was obstructed by
carbon limitation which severely inhibited the dominant cryptophytes.
In summary, the phytoplankton of the freshwater tidal Potomac
responded to ir» situ nutrient bioassay in a manner consistent with
phosphorus limitation of growth. Responses of major groups of algae
differed markedly especially when high algal crops rendered carbon
limitation a problem in the enclosed containers. Future studies should
ensure against carbon limitation by using open enclosures, shorter
incubations, and/or inorganic carbon additions.
ACKNOWLEDGEMENTS
I wish to thank Hans Paerl for providing a prepublication draft of
his methods paper on nutrient dilution bioassays and for advice on
setting up these experiments. Vicki Andrle and Claire Buchanan provided
phytoplankton enumerations and Sue Touart and Ann Powel assisted in
laboratory analyses. Allan Hide provided field support. Fairfax
County's Lower Potomac Pollution Control Plant provided nutrient
analyses. Funding for this study was provided by the Metropolitan
Washington Council of Governments. The assistance of Wendy Chittenden
and Stuart Freudberg is greatly appreciated.
251
-------
LITERATURE CITED
Goldman, C.R. 1963. The measurement of primary productivity and limiting
factors in freshwater with C-14. Pages 103-113 in M.S. Doty (ed.)
Proceedings of the conference on primary productivity measurement,
marine and freshwater. U.S. Atomic Energy Commission T.I.D.-7633.
Goldman, C.R. 1978. The use of natural phytoplankton populations in
bioassay. Mitteilungen. Internationale Vereinigung fur Theoretische
und Angewandte Limnologie 21: 364-371.
Jones, R.C. 1987. A study of nutrient limitations to algal growth in the
Potomac River Estuary. Final Report submitted to Metropolitan
Washington Council of Governments. Washington, D.C.
King, D.L. 1970. The role of carbon in eutrophication. Journal Water
Pollution Control Federation 42: 2035-2051.
Miller, W.E., J.C. Greene, and T. Shiroyama. 1978. The Selenastrum
capricornutum Printz algal assay bottle te'st: experimental design,
application, and data interpretation. EPA-600/9-78-018.
Moss, B. 1973. The influence of environmental factors on the distribution
of freshwater algae: an experimental study. II. The role of pH and
the carbon dioxide-bicarbonate system. Journal of Ecology 61: 157-
177.
Paerl, H.W. 1983. Factors regulation nuisance blue-green algal bloom
potentials in the lower Neuse River, N.C. Water Resources Research
Institute of the University of North Carolina. Report 188.
Paerl, H.W. and N.D. Bowles. 1987. Dilution bioassays: their
application to assessments of nutrient limitation in hypereutrophic
waters. Hydrobiologia 146: 265-273.
Reynolds, C.S. 1984. The Ecology of Freshwater Phytoplankton. Cambridge
University Press.
Shapiro, J. 1973. Blue-green algae: why they become dominant. Science
179: 382-384.
Tailing, J.F. 1976. The depletion of carbon dioxide from lake water by
phytoplankton. Journal of Ecology 64: 79-121.
Thomann, R.V., N.J. Jaworski, S.W. Nixon, H.W. Paerl, and J. Taft. 1985.
The 1983 Algal Bloom in the Potomac Estuary. Prepared for the
Potomac Strategy State/EPA Management Committee.
252
-------
Understanding the Estuary: Advances in Chesapeake Abstract only
Bay Research. Proceedings of a Conference. 29-31
March 1988. Baltimore, Maryland. Chesapeake Research
Consortium Publication 129. CBP/TRS 24/88.
Numerical Tagging of Phosphorus in the James Estuary
Wu-Seng Lung
University of Virginia
Department of Civil Engineering
Charlottesville, Virginia 22901
In estuarine eutrophication analysis, one of the key questions often
asked is to where nutrients from particular source (s) would be transported.
For example, in the James Estuary, phosphorus input to the upper estuary could
be incorporated into the biomass of phytoplankton in the water column,
deposited into the sediments, or transported to the lower estuary. Perhaps a
more meaningful question is: how much phosphorus in the peak algal biomass is
from a given source? A simple component analysis is not appropriate for the
eutrophication modeling analysis simply because of the nonlinear relationship
of the phytoplankton growth-nutrient limitation dynamics in the system. A
numerical tracer was added to a water quality model of the James Estuary to
determine the fate of phosphorus in the system. That is, a source or sources
of phosphorus was numerically labeled and added to the James Estuary. The
model was then used to quantify the amount of such labeled phosphorus in
different components of the water column: organic phosphorus, orthophosphate,
and algal biomass. Results of the analysis using the water quality data
collected in September 1983 indicated that municipal wastewater discharges in
the upper estuary (i.e., from Richmond, Falling Creek, Proctors Creek, and
Hopewell plants) contributed to about 75% of phosphorus in the peak algal
biomass (as chlorophyll a) in the water column. Upstream (nonpoint) input
provided another 15% of phosphorus in the peak algal biomass. Industrial
wastewaters played a very small role in contributing to the algal biomass in
the James Estuary. Finally, the Appomattox River which receives wastewater
discharges from the City of Petersburg contributed an insignificant amount of
phosphorus to the peak algal biomass in the mainstream of the estuary.
253
-------
Understanding the Estuary: Advances in Chesapeake Abstract only
Bay Research. Proceedings of a Conference. 29-31
March 1988. Baltimore, Maryland. Chesapeake Research
Consortium Publication 129. CBPlTRS 24/88.
Some Response Patterns of Phytoplankton to Nutrification Based on
Enrichment Experiments and Apparent Influence of Silicon
T. J. Smayda and T. A. Villareal
University of Rhode Island
Graduate School of Oceanography
Kingston, Rhode Island 02881
Nutrient-spike experiments evaluated the effect of various
concentrations of N+P and N+P+Si on phytoplankton community structure,
speciation, biomass/ growth rate and primary production. Seven test levels,
including maximal NH^ doses equivalent to secondary effluent, were used and
attenuated in a graded series to simulate a nutrient gradient. Enrichment
with NH^, PO^ and Si02 stimulated carbon production rates, chlorophyll biomass
yield and species growth rates. Lag effects, mortality, suppression of growth
and production, and differential species responses were commonplace, however.
Enrichment did not trigger blooms of nuisance species or algal groups during
the 48-hour experiments.
Enrichment with various N+P+Si concentrations tended to be more
stimulatory than enrichment with N+P alone. In an experiment designed to
evaluate kinetics, a highly significant yield-dose relationship occurred
between chlorophyll production over a 48-hour period and initial NH4
concentrations (Y - 5.17X0.62; r2 - 0.94). The relationship between
chlorophyll yield and NH4 uptake was described by Y - 3.33 + 2.06X (r2 -
0.85). There was a significant correlation between NH4 uptake and its
concentration, with the relationship influenced by Si02- The presence of Si02
in excess of 5 mg-at m~3 stimulated NH4 uptake by about 60% at NH4
concentrations greater than 12.5 mg-at m~^ (r2 «• 0.98). In the absence of, or
at < 5.0 mg-at m~3) the relationship was described by Y - 0.93X0.91 (r2 -
0.94) . Between 5.0 and 18.5 mg-at m~3 NH4, £ 90% of the NH4 was taken up over
the 48-hr period. Above 18.5 mg-at m~3 NH4, the percentage uptake fell off
rapidly, with the percentage utilization influenced by Si02 concentration.
This apparent influence of Si02 on the various responses of natural
phytoplankton communities to nitrogen enrichment was an unexpected finding
which warrants further study to evaluate its potential role as a regulator
and/or mediator of phytoplankton community responses to estuarine
nutrification.
The enrichment experiments were carried out using summer phytoplankton
communities from Massachusetts Bay in outer Boston Harbor. The results would
appear to be relevant to Chesapeake Bay nutrification issues; hence,
submission of this abstract.
254
-------
Understanding the Estuary: Advances in Chesapeake
Bay Research. Proceedings of a Conference. 29-31
March 1988. Baltimore, Maryland. Chesapeake Research
Consortium Publication 129. CBP/TRS 24/88.
The Importance of Submarine Groundwater Discharge to Nutrient Flux
in Coastal Marine Environments
George M. Simmons, Jr.
Virginia Polytechnic Institute and State University
Biology Department
Blacksburg, Virginia 24061
INTRODUCTION
The purpose of this investigation was to examine the role of submarine grounclvvater
discharge (SGWD) on nutrient flux into the ncarshorc area of two coastal marine
environments on Virginia's eastern shore. The movement of water across
sediment/water interfaces is important biologically because it sets the microclimatic
conditions for sediment inhabiting micro and macro organisms. SGWD, therefore, is a
site specific phenomenon. Such movement also contributes to bcnthic-pclagic coupling
(Nixon 1981; Rowe et al. 197.5; Rowe and Smith 1977).
Historically, nutrient flux across sediment/water interfaces has been measured by two
methods: in situ and diffusion gradient methods (Zcit/schcl 1979). This paper reports
the movement of water and dissolved material across sediment/water interfaces in
marine environments by a third method, that of bulk flow, using seepage meters and
mini-piezometers (Lee 1977, 1980; Lee and Cherry 1978).
Initial interest in groundwatcr discharge into marine habitats focused on the relationship
between fresh, groundwater inflow and/or scawatcr intrusion into fresh, groundwatcr
supplies (Rcilly and Goodman 1985). Over a period of several decades, numerous
mathematical equations and models have been used to relate the interactions between
these two systems and their interface properties. Glover (1959) developed a
mathematical expression to describe the interface between fresh and salt groundwatcr in
a coastal aquifer that accounted for the movement and discharge of freshwater. Cooper
(1959) developed a hypothesis to explain the mixing /.one, or /one of dispersion, the
continuous circulation of scawatcr observed in various field studies, and he attempted
to quantify the amount of mixing due to tidal fluctuations. Henry (1959, 1964)
advanced the concepts further by using an advcction - diffusion equation to account for
255
-------
hydrodynamic dispersion. Kohout (I960, 1964) was one of the first investigators to
quantify and suggest the continuous cycling ofseawatcr as a result ofSGWD. The
important summary to these developments from an ecological perspective is the
realization that fresh, groundwater does move into coastal marine environments and
that seawater cycles through the sediments as a result of the SGWD (Fig. 1).
Kohout (1966) and Kohout and Kolipinski (1967) appear to be the first investigators to
study the importance of freshwater seepage into shallow marine ecosystems. Their
studies, conducted along the shore of Biscayne Bay, Florida, showed a definite
relationship between biological /onation and SGWD into the bay. Johannes (1980) also
presented significant information on the ecological significance of SGWD. While
acknowledging the fact that SGWD to the sea is widespread, "... overlooking the fact
could lead to serious misinterpretation of ecological data in studies of coastal pollution,
of benthic Donation and productivity, and of the flux of dissolved substances between
bottom sediments and overlying water." Harden Jones (1980) has even suggested
groundwater seepage as a landmark for identifying spawning grounds for plaice.
In the past, direct measurements of bulk water flow across sediment/water interfaces in
marine environments would have been cost prohibitive due to drilling requirements.
However, with the development of seepage meters and mini-piezometers by Lee (1977)
and Lee and Cherry (1978), such information can be obtained easily and cost effectively.
Studies dealing with seepage flux have been estimated in lakes and streams (Lee 1977;
Lee and Ilynes 1978; Lee and Cherry 1978; Lock and John 1978; Lee 1980; Erickson
1981), estuarys (Valiela et al. 1978; Bokuniewic/ 1980; Zimmerman et al. 1985; Capone
and Bautista, 1985), and coral reefs (D'Elia et al. 1981; Oberdorfer and Buddemeicr
1985, 1986; Lewis 1987; Simmons and Love 1987; Simmons and Netherton 1987;
Simmons in press).
METHODS
Two sites were used in this study and both were located on Virginia's eastern shore
(Fig. 2) in conjunction with land-based wells established by the U.S. Geological Survey.
The first site, located off Chincoteage Island, represented a high density human
population area with large volumes of freshwater input to the groundwater system due
to human consumption and septic tank usage. Steclman's Landing represented an
agricultural site in which SGWD to Magothy Bay was buffered by woodland and
marshland areas (Fig. 3) each ~350 m wide at the research site. In general, seepage
meters were established ~10-I5 cm beneath the water's surface at low tide, and 10m and
I00m offshore the low tide mark. Samples of SGWD were collected in acid washed bags
that were rinsed in deionized/distilled water. Water samples were then transferred to
acid washed plastic bottles, kept on ice, and returned to the laboratory for analyses.
Standards for nutrient analyses were made in substitute ocean water. Nitrate was
measured by cadmium reduction following E.P.A. Method 353.3 (E.P.A. 1983).
Ammonia determinations were made using the phenate method (E.P.A. Method 350.1)
(E.P.A. 1983 and American Public Health Assoc. et al. 1976). Total and dissolved
phosphate were measured by the ascorbic acid method (E.P.A. Method 365.2) (E.P.A.
1983).
The specific gravity of samples was measured with a hydrometer read to the fourth
decimal place. Corresponding density/salinity tables were used to convert hydrometer
measurements to salinity values (American Public Health Assoc. et al. 1976).
Temperature measurements were made with calibrated long stem thermometers.
Oxygen measurements were made using cither a macro or micro-winklcr techniques and
titration with the appropriate normality of thiosulfatc.
256
-------
Artesian (confined) aquifer
Figure 1. Schematic diagram of the interface between fresh and salt groundwatcr
systems in a coastal environment (from Johannes 1980).
VIRGINIA'S EASTERN
SHORE
CHESAPEAKE
BAY
\
."•Chincoteague Site
ATLANTIC OCEAN
Steelman'e Landing
Site
•CALt
Figure 2. Study sites for submarine groundwatcr discharge on Virginia's eastern shore.
257
-------
SCALE* I MILE-1.6KM
Figure 3. Study site at Steelman's Landing.
258
-------
The seepage meters used in this study were 0,25 m2. The flux of SGWD was measured
with seepage meters. The volume of water collected was divided by the time interval of
collection and multiplied by 4.0 to approximate the amount of water passing across one
square meter of bottom. The quantity of SGWD collected was expressed as
L-m~2'day~'. Knowing the concentration of a particular nutrient and multiplying this
by the seepage Ilux resulted in nutrient flux expressed as mg'm~2>day~l.
RESULTS
Seepage meter discharge rates for the two sites are summarized in Table 1 and show that
discharge rates were higher at the Chincotcague Site than the Stcelman's Landing Site.
The SGWD at the Chincotcague Site showed a classical decrease in discharge rate with
distance from shore, but the highest discharge at Stcclman's Landing varied with
distance from shore. Tidal fluctuation off Chincotcague was fairly consistent during the
study periods, but off Stcelman's Landing, changes in wind direction often left seepage
meters exposed during low tide, or inundated to the extent that sample recovery at high
tide was not feasible. The inability to keep the shallow meters submerged and to collect
at needed time intervals may have contributed to some of the variability at the
Steelman's Landing Site.
Salinity measurements from the two sites arc summarized in Table 2. Hven though all
wells sampled from Chincoteague are summarized in Table 2, only one well represented
salinity values of groundwatcr that discharged into the research site. This well had a
mean salinity of 1.3 ppt. The salinity of SGWD in both seepage meters and piezometers
indicated the presence of fresh, groundwater entering the channel at least to a point I Cm
offshore the low tide mark. The influence of SGWD off Stcclman's Landing was equally
apparent in both seepage meters and piezometers. Further evidence of fresh,
groundwater seepage entering Magothy Bay was found in water pools above the high
tide mark in the marshland area (Fig. 2). Woods Well was located on the edge of the
woodland area next to the marshland and had a mean salinity of 2.2 ppt. Marsh Well
was located ~85m into the marshland area and contained a mean salinity of 26.7 ppt.
Water sampled from pools between the Marsh Well and the Bay contained a salinity of
2.1 ppt. During a field trip in February at the Steelman's Landing site, a small, shallow
stream of water (~lcm deep by 5m wide) was observed flowing from the marshland into
the bay. The time of collection was almost at mean low tide and the flowing water was
believed to be residual Bay water left in the marsh from the previous high tide. The
salinity of this water, however, was 4.8 ppt.
Concentrations and flux of ammonia are summarized in Table 3. No nitrate was
measured in any of the water samples from the Chincotcague Site. All of the inorganic
nitrogen was in the form of ammonia. Wells were anaerobic and the well nearest the
study site had a mean ammonia concentration of 74.01 mg/L (± 70.0, N= 3). In
contrast to the Chincoteague Site, the distribution of nitrate and ammonia was very
different at the Stcelmanjs. Landing Site. No ammonia was measured in any of the wells,
but nitrate was present (X = 11.55 ± 5.22; N= 7). Moreover, these well waters were
highly oxygenated (8-10 mg-L"1). Conversely, nitrate was not found in any of the
SGWD samples, but ammonia was present. Table 3 shows that ammonia levels in the
SGWD was considerably lower than that measured off Chincotcague, but
concentrations were higher than that measured in ambient bay water. In two of the
wells sampled, Road and Field Wells, concentrations of nitrate at times exceeded 10
mg'L"1. On the February collecting trip, a concentration of 22.91 mg-L ' was
measured in the Road Well. High concentrations also were recorded in the pond water
and to a certain extent in the stream water (X = 4.68 ± 3.30; N = 6). Interestingly,
Woods Well and Marsh Well did not show any nitrate presence, but did show ammonia
presence on only one occasion. A greater number of piezometers were used off
259
-------
TABLE 1
SEEPAGE METER DISCHARGE RATES
Chincoteaguc Site - 1987
SITE MONTHS
Steclman's Landing Site • 1987-88
MONTHS
SHORE
Mean
S.D.
N
MAR
20.26
16.53
6
APR
28.16
13.71
19
MID (~ 10m offshore)
Mean 13.23 9.83
S.D. 22.11 15.21
N 3 14
DEEP (~100m offshore)
Mean 1.71 1.84
S.D. 0.40 0.69
N 3 3
MAY
13.59
14.98
11
8.83
7.26
5
DEC
2.27
1.02
5
1.47
0.97
3
JAN FEB
5.95 1.16
2.33 11.08
3 6
6.59
4.11
6
9.33
7.68
9
2.08
1.21
8
2.00
1.43
12
260
-------
TABLE 2
SALINITY (ppt)
Chincoteague Site
Reference Water
Wells Chincoteague Sound
Mean 9.7 28.7
Range 0.4-23.7 25.0-30.3
N 3 3
Submarine Groundwater Discharge
Shallow Mid
SM Piez. SM
Mean 21.2 16.7 30.4
Range 8.5-30.1 6.2-27.1 29.4-30.8
N 11 2 4
Steelman's Landing Site
Reference Water
Field Wells, Pond, Creek Woods Well Marsh Well Magothy Bay
Mean 0.7 2.2 26.7 31.5
Range 0.0-1.9 1.6-2.9 21.7-35.4 30.4-32.8
N 12 3 37
Submarine Groundwater Discharge
Shallow Mid Deep
SM Piez SM Piez SM
Mean 30.8 30.9 25.5 29.7 31.0
Range 29.8-32.4 25.6-32.7 24.7-27.5 27.2-31.4 29.8-33.5
N 8 12 6 89
267
-------
Chincoteague Site:
TABLE 3
Concentration and Flux of Ammonia (NHa-N)
Reference Water (mg-L~')
Mean
Range
N
Shallow
Mean
Range
N
Mid
Mean
Range
N
Deep
Mean
Range
N
Steelman's
Wells
52.83
4.61-150.66
6
Submarine Groundwater
Conc(mg'L-')
61.95
2.09-188.83
9
1.70
0.31-3.60
4
4.8
3.1-6.5
2
Landing Site:
Chincoteague Channe
None Detected
Discharge
Flux (mg-m-May- ')
Mean 787.98
S.D. 801.68
N 8
Mean 14.93
S.D. 13.68
N 4
Mean 9.10
S.D. -
N 2
Reference Water
Mean
Range
N
Surface
Mean
Range
N
Mid
Mean
Range
N
Deep
Mean
Range
N
Wells
None Detected
19
Submarine Groundwater
Conc(mg-L-1)
0.19
0.00-0.44
7
1.32
0.00-1.94
4
0.33
0.00-0.67
5
Magothy Bay
0.19
0.00-0.42
5
Discharge
Flux (mg-nr'-day-1)
Mean 1.27
S.D. 1.47
N 4
Mean 15.84
S.D. 13.33
N 3
Mean 3.24
S.D. 3.36
N 5
262
-------
Steclman's Landing and they showed a higher concentration of ammonia than the
seepage meters. An extreme example occurred in December when the mean amonia
concentration from six pic7.ometcrs, one meter into the sediments, was 4.03 ± 2.16
rng-L ', and no ammonia was detected in the seepage meters even after their pumping
for over 24 hrs at a rate of ~ 1-5 L-m"2'day~1.
Total phosphate levels at the Chincoteaguc Site are summarized in Table 4. The mean
concentration of total phosphate in the well closest to the research site showed a mean
concentration of 0.34 mg/L (± 0.33; N= 3). Dissolved phosphate was ~80-95% of the
total phosphate measured in well and seepage meter samples from Chincoteaguc. As
with the ammonia samples, too few piezometer samples were taken off Chincoteague to
be meaningful. However, samples were within the range measured by the seepage
meters suggesting little dissolution from the sediments.
The amount of total phosphate measured in well, surface, and seepage water samples
ofTStcelman's Landing was considerably lower than measured at the (Chincoteaguc Site
except for the deep site (Table 4). Furthermore, the mean concentration of dissolved
and total phosphate in seepage meter samples from the mid and deep water sites was
lower than piezometer samples from the same depths (.10 vs .17 mg-L~', respectively).
This suggests that 1) dissolved phosphate was being fixed at the sediment/water interface
and 2) the presence of the seepage meters did not cause dissolution of phosphate from
the sediments. In contrast to the Chincoteague Site, the proportion of dissolved
phosphate was —30% of the total phosphate concentration.
Due to water conditions, only a few hydraulic head readings could be taken which were
considered to be reliable. At the Chincoteague Site, mean values of 8.4 and 12.1 cm
were measured with a manometer during the March and April field trips. At the
Steelman's Landing Site, a hydraulic head of 12.5 and 22.5 cm was measured on the
January and February collecting trip, respectively, using water level height from flowing
piezometers at low tide.
DISCUSSION
The data collected off Chincoteague and Steelman's Landing show that water below
ambient salinity did move into shallow, ncarshore marine environments. The discharge
values reported here are similar to those reported by Bokunicwicz (1980) and
Zimmerman et al. (1985). Moreover, it could be demonstrated that the SGWD was
being driven by a positive hydraulic head. Other investigators have expressed caution
and concern when using nutrient data obtained from seepage meters (Brock ct al. 1982),
and such concern between the real and experimental world is valid in this study.
However, nutrient concentrations were usually higher in piezometer water than seepage
meter water, and dissolved oxygen concentrations taken from ports in the seepage
meters indicated an aerobic environment. Therefore, an assumption has been made
that, in these cases, there was little or no seepage meter effect,
In comparing nutrient flux data, the problem arises with the units of expression. For
comparative purposes, I have converted other reported values to mg-nv^day-' for the
elemental species and tabulated these in Table 5. For those investigators familiar with
the original citation values, these numbers also arc listed for comparative purposes.
Nutrient transport into marine environments has been reported by other investigators
using several methods. A partial summary of some of these studies has been presented
by Nixon (1981). In addition, Phoel, et al. (1981) studied nutrient flux in the York River
at three different depths. They reported ammonia to be the major inorganic nitrogen
species moving into the water column through diffusion flux and their flux rates ranged
263
-------
TABLE 4
Concentration and Flux of Total Phosphate (PO4-P)
Chincoteague Site:
Reference Water (mg'L~')
Mean
Range
N
Shallow
Mean
Range
N
Mid
Mean
Range
N
Deep
Mean
Range
N
Steelman's
Wells
0.37
0.01-0.07
6
Submarine Groundwater
Conc(mg'L-')
0.62
0.19-1.42
12
0.79
0.00-1.62
8
0.06
0.05-0.06
4
Landing Site:
Chincoteague Channe
0.04
0.00-0.07
11
Discharge
Flux (mg-nr'-day-1)
Mean 15.82
S.D. 17.84
N 9
Mean 3.55
S.D. 2.98
N 8
Mean 0.09
S.D. 0.03
N 4
Reference Water (mg-L~' )
Mean
Range
N
Shallow
Mean
Range
N
Mid
Mean
Range
N
Deep
Mean
Range
N
Wells
0.06
0.02-1.60
19
Submarine Groundwater
Conc(mg'L~')
0.08
0.04-0.14
9
0.10
0.05-0.19
9
0.10
0.05-0.13
9
Magothy Bay
0.11
0.08-0.12
5
Discharge
Flux (mg-nr^day-')
Mean 0.25
S.D. 0.24
N 9
Mean 0.59
S.D. 0.66
N 8
Mean 0.98
S.D. 0.92
N 8
264
-------
STUDY
Callender and
Hammond
(1982)
Capone and
Bautista (1985)
Nixon (1981)
Phoel, et al.
(1981)
TABLE 5
Comparative Data — Other Nutrient Flux Studies
ORIGINAL CITATION VALUE CONVERTED VALUES
0.6-6.5mMole-m-
1-21
1 N113 9.6-88.8
-' NH3 144-285.6
1 NII3-N
1 NH3-N
10/zM NO3
0.12-600 mg-m-'-day-1 NO3-N
n-2-hr-' NII4f 4.4-84.0mg-m-2-day"! NH4-N
3m:22-191 ugatN-m-^hr1 NII3 7.2-648mg-nr2-day-1 NH3-N
7-1678 ugatN-m 2-hr'l NH3 2.4-564.0 nig-m^-day"1 NH3-N
9m:162 ugatN-m-2-hr~l NII3 55.2
NH3-N
16m:511 ugatN-m-2-hr-' NH3
288ugatN-nr2-hr-' NM3
It ugatN-m-'-hr-1 NO3
172.8mg-m-2-day-1 NH3-N
96.0mg-trT2-day- ' NH3-N
4.8 mg-m-2-day~1 NO3-N
Callender and .02-0.3mMole-m-2-day-1 PO4-P .72-9.36mg-nr2-day-1 PO4-P
Hammond 0.1-2.0mMole-m-2-day ' PO4-P 3.12-62.4mg-m-2-day-1 PO4-P
(1982)
Nixon (1981) 2-50 ^Moles-m-2-hr ' PO4-P
Zimmerman,
etal. (1985)
29-70 x JO' 6g-m-2-day-1 DRP*
DRP*
7.8xlO-3g-m-2-day ! DRP*
1.44-37.7mg-m-2-day-' PO4-P
.02-.07mg-m-2-day-1 PO4-P
.003-.02mg-m-2-day-1 PO4
7.92 mg-m-2-day-f PO4-P
*DRP - Dissolved Reactive Phosphate
265
-------
between 2.4 - 172.8 mgday-') and the highest values to occur at the 3m depth
under maximal light conditions (564.0mg-m 2-day '). Capone and Bautista (1985)
reported a mean concentration of 0.120 mg NO3-N at the sediment water interface in
Great South Bay, New York, and when coupled with the discharge rates of Bokuniewicz
(1980), the flux rate should be ~0.12 - 6.00mg-rrr3-day-'.
Phosphate is another important nutrient species that has been investigated. Again,
some of the highest values reported arc those by Callender and Hammond (1982) (285.6
mg«m J'day-'). Zimmerman ct al. (1985) conducted a study in a nearshore estuarine
environment and concluded that >99% (7.92 mg-nr'-day-') of the dissolved reactive
phosphate entering the water was due to groundwater seepage.
Several investigators have attributed nutrient (lux to invertebrate irrigation, and there
is no doubt these animals play an important role in sediment irrigation and nutrient
transport. However, the data to date suggest that SGWD in shallow environments plays
the dominating role. In other studies o/T Key Largo, Florida I have observed seepage
meters to become partially anaerobic, benthic invertebrates to succumb to the lowered
oxygen levels and expire, and yet the discharge rate remained the same (unpublished
data). The relationship between SGWD with and without the presence of invertebrates
deserves closer attention, particularly in estuarine environments.
The data reported here show that ammonia flux into the shallow /one ofTChincotcaguc
is the highest reported thus far in the literature. I lighcr flux rates for phosphorus (DRP)
have been reported by other investigators (Table 5). The important comparison is
between the Chincoteague and Steelman's Landing sites. Chincoteaguc represents an
environment where fresh, groundwater is pumped from the Wallop's Island area for
human consumption in Chincoteague (Fenncma and Newton 1982). The water, after
being used, is discharged into septic tanks and then into the groundwater on the island.
The water is then able to make its way directly into the bay. At our study site, there
were no woods or marsh buffer and the phosphate and ammonia data collected were
correspondingly high. At Steelman's Landing, groundwater beneath cropland showed
high nitrate levels (22.0 mg/L), but these high concentrations did not translate into high
ammonia (or nitrate) levels in the SGWD. This possibly may have been due to the
buffering activity of the plant communities in the woodland and marshland areas.
Furthermore, the fact that the concentration of the different nutrient species from the
piezometers was generally greater than that collected from the seepage meters suggests
that the microbial community at the sediment/water interface plays an important role
in converting the nutrients to biomass. Submerged rooted aquatic vegetation would
play the same role unless the concentration of nutrients bathing the root system were
toxic.
266
-------
SUMMARY
1. Two sites on Virginia's eastern shore were studied over a short time period with
regard to the role of SGWD as a vehicle for nutrient flux. One area represented a
site of high density human population with little or no plant buffer of groundwater
entering the bay. The other represented an agricultural area buffered from the bay
by a woodland and marshland area.
2. The data showed that water of lower salinity, under a positive hydraulic head, did
enter the nearshore bay environment with varying concentrations of nutrient
species. OffStcelman's Landing, water of low salinity also was found on the surface
in the marsh habitat. This suggests that much of the fresh, groundwater may be
exiting in the marshland.
3. Higher concentrations of ammonia and phosphate were measured off Chincotcague.
The data suggest that more attention should be given to the role of SGWD on
nearshore coastal environments when considering the establishment of high density
human populations.
4. Even though the total amount of nutrients entering shallow marine environments
may be less than that carried in by rivers, the effect of SGWD is site specific at the
sediment/water interface. Moreover, SGWD also could be a vehicle for synthetic
chemicals and toxic metals.
ACKNOWLEDGEMENTS
The author acknowledges his appreciation for the cooperation of the U.S. Geological
Survey, Richmond, Va.; the Nature Conservancy; and, Virginia's Marine Resource
Commission for providing access to wells, land, and water for this project. This research
was funded by a Virginia Tech Environmental Water Resources Research Grant
(#2087490). The author also recognizes the field assistance of Messers. Jake Waller, Ed
Lamoureux, Karl vonSchmidt-Pauli, Bob Cummins, and Dr. Ken Elliott. Appreciation
also is extended to Dr. G. F. Ocrtcl, Barrier Island Program, Old Dominion University
for permission to use their facility at Oyster, Va.
REFERENCES CITED
American Public Health Association, American Water Works Association, Water
Pollution Control Federation. 1976. Standard methods for the examination of
water and wastewater. Fourteenth Edition. American Public Health Association,
Washington, D.C. xxxix + 1193 p.
Bokuniewicz, H. 1980. Groundwater seepage into Great South Bay, New York. Estuar.
Coast. Mar. Sci. 10:437-444.
Brock, T.D., D.R. Lee, D. Janes, and D. Winek. 1982. Groundwater seepage as a
nutrient source to a drainage lake; Lake Mendota, Wisconsin. Water Res.
16:1255-1263.
Callender, E. and D.E. Hammond. 1982. Nutrient exchange across the sediment - water
interface in the Potomac River estuary. Estuarine, Coastal and Shelf Sci.
15:395-413.
Capone, D.G. and M.F. Bautista. 1985. A groundwater source of nitrate in nearshore
marine sediments. Nature 313:214-216.
Cooper, II.II. 1959. A hypothesis concerning the dynamic balance of freshwater and
saltwater in a coastal aquifer. J. Gcophys. Res. 64:461-467.
267
-------
D'Elia, C.F., K.L. Webb, and J.W. Porter. 1981. Nitrate-rich ground water inputs to
Discovery Bay, Jamaica: a significant source ofN to local reefs? Bull. Mar. Sci.
31:903-910.
Environmental Protection Agency. 1983. Methods for chemical analysis of water and
wastes. OPA-600/4-79-020.
Erickson, D.R. 1981. A study of littoral groundwater seepage at Williams Lake,
Minnesota using seepage meters and wells. M.S. Thesis, the University of
Minnesota, ix + 135 p.
Fennema, R.J. and V.P. Newton. 1982. Ground water resources of the eastern shore
of Virginia. Plan. Bull. 3332. Commonwealth of Virginia. State Water Control
Board. Tidewater Regional Office. 74p. + Appendices.
Glover, R.E. 1959. The pattern of freshwater flow in a coastal aquifer. J. Geophys.
Res. 64:457-459.
Harden Jones, F.R. 1980. The migration of plaice (Plcuroncctes platessa) in relation
to the environment, p. 383-399. hi Bardach, J.H., J.J. Magnuson, R.C. May and
J.M. Reinhart (cds.) Fish Behavior and Its Use in the Capture and Culture of
Fishes. ICLARM Conf. Proc. 5, 512 p. International Center for Living Aquatic
Resources Management, Manila, Philippines.
Henry, H.R. 1959. Salt intrusion into fresh - water aquifers. J. Geophys. Res.
64:1911-1919.
. 1964. Effects of dispersion on salt encroachment in coastal aquifers. In: Sea
Water in Coastal Aquifers. U.S. Geol. Surv., Water - Supply Pap. 1613-C, pp.
70-84.
Johannes, R.E. 1980. The ecological significance of the submarine discharge of ground
water. Mar. Ecol. Prog. Ser. 3:365-373.
Kohout, F.A. 1960. Cyclic flow of salt water in the Biscayne Aquifer of southeastern
Florida. J. Geophys. Res. 65:2132-2141.
. 1964. The flow of fresh water and salt water in the Biscayne Bay Aquifer of
the Miami area, Florida. Seawater in coastal aquifers. U.S. Geol. Surv., Water
Supply Pap. 1616-C: 12-32.
1966. Submarine springs: a neglected phenomenon of coastal hydrology.
Hydrology and Water Resources Development, Symposium. Central Treaty Org.
(CENTO), Ankara, Turkey. Feb. 7-12, 1966. pp. 391-413.
and Kolipinski. 1967. Biological zonation related to ground water discharge
along the shore of Biscayne Bay, Miami, Florida, pp. 488-499. |n G.I I. LaufT(cd.)
Estuaries. Conf. on Estuaries, Jekyll Island, GA.
Lee, D.R. 1977. A device for measuring seepage flux in lakes and estuaries. Limnol.
Oceanogr. 22:140-147.
. 1980. Groundwater - solute influx. Limnol. Oceanogr. 25:183-186.
and J.A. Cherry. 1978. A field exercise on groundwater flow using seepage
meters and mini-piezometers. J. Geol. lid. 27:6-10.
and II.B.N. Ilynes. 1978. Identification of groundwater discharge zones in a
reach of Hillman Creek in Southern Ontario. Poll. Res. Can. 13:121-133.
Lewis, J.B. 1987. Measurements of groundwater seepage flux onto a coral reef: spatial
and temporal variations. Limnol. Oceanogr. 32:1165-1169.
Lock, M.A. and P.J. John. 1978. The measurement of groundwater discharge into a
lake by a direct method. Int. Revue gcs. Ilydrobiol. 63:271-275.
Nixon,S. 1981. Remincralization and nutrient cycling in coastal marine ecosystems, p.
111-138. hi B.J. Neilson and L. E. Cronin (cd.) Estuaries and Nutrients. Humana
Press, Clifton, N.J.
Oberdorfer, J.A. and R.W. Buddemeier. 1985. Coral reef hydrology. Proc. 5th Int.
Coral Reef Symp. 3:307-312.
and . 1986. Coral reef hydrology: Field studies of water
movement within a barrier reef. Coral Reefs 5:7-12.
268
-------
Phoel, W.C., K.C. Webb, and C.F. D'Klia. 1981. Inorganic nitrogen regeneration and
total oxygen consumption by the sediments at the mouth of the York River,
Virginia, p. 607-618. In B.J. Ncilson and L.E. Cronin (cd.) Estuaries and Nutrients.
Humana Press, Clifton, N.J.
Reilly, T. E. and A.S. Goodman. 1985. Quantitative analysis of saltwater - freshwater
relationships in groundwater systems - a historical perspective. J. Hydrol. 80:
125-160.
Rowe, G.T. and K.L. Smith, Jr. 1977. Bcnthic - pelagic coupling in the mid - atlantic
bight, p. 55-66. In B.C. Coull (cd.) Ecology of Marine Benthos, Univ. of South
Carolina Press, Columbia, S.C.
, C.H. Clifford, and K.L. Smith, Jr. 1975. Bcnthic nutrient regeneration and its
coupling to primary productivity in coastal waters. Nature 255:215-217,
Simmons, G.M., Jr. Groundwater discharge quality in a deep coral reef habitat. NOAA
Technical Rept. Ser., Off. Ocean and Coastal Resource Mgt., Sanctuary Programs
Div. (In Press).
and J. Netherton. 1987. Groundwater discharge in a deep coral reef habitat -
evidence for a new biogeochemical cycle?, pp. 1-12. In C.T. Mitchell (ed.) Diving
for Science - 86. Proc. Amer. Assoc. Underwater Sci. Amer. Acad. Underwater
Sciences, 947 Newhall St., Costa Mesa, CA 92627.
and F.G. Love. 1987. Water quality of newly discovered groundwater
discharge into a deep coral reef habitat, pp. 155 - 164. hi R.A. Cooper and A.N.
Shepard (eds), Science Applications of Current Diving Technology on the U.S.
Continental Shelf. NOAA Syinp. Ser. Undersea Res. 2(2):7-16. NOAA Undersea
Research Program, Rockville, MI).
Valiela, I., J.M. Teal, S. Volkmann, D. Shafer, 11.3. Carpenter. 1978. Nutrient and
paniculate fluxes in a salt marsh ecosystem: Tidal exchanges and inputs by
precipitation and groundwater. Limnol. Oceanogr. 23:798-812.
Zeitzschel, B. 1979. Sediment - water interaction in nutrient dynamics, p. 195-218. In
K.R. Tenore and B.C. Coull (ed.) Marine Benthic Dynamics. Univ. of South
Carolina Press, Columbia, S.C.
Zimmerman, C.F., J.R. Montgomery, and P.R. Carlson. 1985. Variability of dissolved
reactive phosphate flux rates in ncarshore estuarine sediments: effects of
groundwater flow. Estuaries. 8:228-236.
269
-------
Understanding the Estuary: Advances in Chesapeake Abstract only
Bay Research. Proceedings of a Conference. 29-31
March 1988. Baltimore, Maryland. Chesapeake Research
Consortium Publication 129. CBP/TRS 24/88.
Nitrogen Cycling in Chesapeake Bay Sediments: Balance
Between Regeneration and Denitrification
W. M. Kemp, P. A. Sampou, J. M. Caffrey,
and M, S. Mayer
University of Maryland
Horn Point Environmental Labs
P. O. Box 775
Cambridge, Maryland 21613
K. Hendriksen W. R. Boynton
University of Aarhus University of Maryland
Department of Ecology and Genetics Chesapeake B iological Lab
Aarhus, Denmark Box 38
Solomons, Maryland 20688
Nitrogen transformation and recycling processes were investigated at two
stations (10 m, 25 m) in the mesohaline region of Chesapeake Bay in relation
to seasonal patterns of phytoplankton production and bottom water Q£
depletion. Intensive sampling at 1-2 wk intervals was done in spring 1986,
and additional measurements were made during spring, summer and mid autumn of
1986-1987. Fluxes of DIN across the sediment-water interface were measured
with core incubations, and nitrification (NI) and denitrification (DN) were
estimated using specific inhibitors in intact cores. Annual cycles of DIN
(NH4 plus N03~) flux from sediments to over-lying water followed the seasonal
temperature cycle, with summer rates ca. 5-fold greater than those in early
spring. A mid-spring peak in DIN flux corresponded to an earlier event of
particulate organic nitrogen (PON) deposition, with a temporal lag related to
ambient temperature. Pools of NH4+ in sediment pore-waters increased by more
than an order-of-magnitude from Apr. to Aug. A decline in macrofauna and
associated bioturbation during this period may have contributed to the NH4+
accumulation, which was quantitatively equivalent to ca. 10% of the mean DIN
flux across the sediment-water interface. Seasonal patterns of NI and DN were
opposite that observed for DIN flux, with highest rates in spring and fall.
NI rates, which were correlated with sediment redox, approached zero during
incipient hypoxia (< 1 mg Q£ l> ) of bottom waters in spring. Hates of both
NI and DN were low in summer even at the shallower station which experienced
hypoxia only rarely. It is likely that the reduced depth of $2 penetration
into sediments and accumulation of toxic anaerobic metabolites (e.g., sulfide)
both contributed to loss of NI and DN. Prelininary budgets of N inputs and
outputs to and from the sediment surface revealed that in spring inputs of PON
exceeded outputs of both NH^ and ^, which fluxes were themselves similar to
one another. By Aug., recycling flux of NH4 to overlying water was greater
270
-------
than the input of PON, and DN was virtually eliminated, especially at the
deeper (anoxic) station. The importance of NH4+ recycling in support of
phytoplankton production is demonstrated, as is the effect of DN in reducing
NH4+ availability for algal growth. It is suggested that any increase in
bottom water 02 concentration in summer (through, for example, reduced
eutrophication) would have the effect of enhancing DN and further decreasing
NH4+ availability for phytoplankton.
277
-------
Understanding the Estuary: Advances in Chesapeake
Bay Research. Proceedings of a Conference. 29-31
March 1988. Baltimore, Maryland. Chesapeake Research
Consortium Publication 129. CBP/TRS 24/88.
Variability of Groundwater Nitrate Concentrations in
Non-Agricultural Ecosystems
T. B. Parkin, E. E. Codling, J. J. Meisinger,
and J. L. Starr
U.S. Department of Agriculture-ARS
Environmental Chemistry Laboratory
Beltsville Agricultural Research Center
Beltsville, Maryland 20705
Mention of brand names does not constitute endorsement of products by the USDA or the recommendation of this
product over other suitable products
ABSTRACT
Results from several recent watershed-scale
studies suggest that removal of fertilizer NC>3~ from
laterally moving, shallow groundwater occurs in forest
and marsh lands adjacent to agricultural systems, thus
mitigating the impact of N03~ pollution of both ground-
and surface waters by agriculture. To date, the precise
mechanisms responsible for this observation remain
unclear. This study was initiated to investigate the
removal of groundwater N03~ in non-agricultural
ecosystems and to study the mechanisms controlling this
process. A site was located on Maryland's Eastern
Shore, and contained an established agricultural field,
abutted by a grass buffer strip, a forest, and a marsh,
all adjacent to the Wye River. Each ecosystem was
sampled with groundwater wells located along transects
which followed the downgrade of the surface topography,
with the agricultural field at the highest elevation and
the marsh at the lowest elevation. Nitrate analyses of
these wells indicated a general trend of decreased
groundwater N03~ concentration under the
non-agricultural ecosystem. However, in 3 of the 8
wells located in the forest decreased N03~
concentrations were not observed. Predicted NC>3~
concentrations in these wells based on Cl~
concentrations indicate that, even in a relatively
272
-------
homogeneous sandy aquifer, preferential flow paths of
groundwater may exist, which contribute to the
variability observed. Therefore, caution must be used
in defining the role of non-agricultural ecosystems in
mitigating N03- of groundwater when data are based on
only a few well transects.
INTRODUCTION
The fate of nitrogen fertilizers applied to soil
is an area of primary importance in agricultural
research, and there is an increasing concern over the
impact of nutrients derived from agricultural systems on
surface and groundwater quality. Nitrates from
agriculture have been implicated in the declining
quality of the Chesapeake Bay (U.S. Environ. Protection
Agency, 1983). These conclusions have come primarily
from surface runoff studies of heavily manured
agricultural lands. Less attention has been given to
investigating the impact of groundwater nitrate loading
to the Chesapeake Bay from adjoining cropland.
Several recent studies have shown that
non-agricultural ecosystems can act as filters in
removing N03~ from shallow groundwater draining
agricultural lands. Nitrate movement from agricultural
fields into forested land was monitored along nine well
transects in a 1568 ha. watershed by Lowrance, et al.
(1984). Seasonal mean concentrations of N03~N were
always found to be significantly higher in field wells
than in the non-agricultural areas. No conclusive
evidence was given regarding the process by which the
riparian forest reduced the N03~ levels, however,
decreasing ratios of N03~N:C1 from the field to the
streams indicated that biological processes (plant
uptake and denitrification) were important in the
removal of N03~N.
Along two groundwater well transects from a corn
field into a forested area, PeterJohn, et al. (1984)
found N03~ concentrations of shallow groundwater inside
the forest to be about an order of magnitude lower than
in the corn field. These N03~ losses were estimated to
be about 1/3 uptake by forest vegetation and 2/3 by
denitrification, however, direct measurements of
denitrification were not done.
In another study Jacobs and Gilliam (1985)
measured N03~N concentrations of 7 to 8 mg/L in the
subsurface drainage water at the edge of a forested
area. Apparent passage through about 16 to 47 m of the
forested area resulted in a drop to <0.1 mg/L N03~N. It
was speculated that denitrification was the primary
mechanism of N03~ removal, rather than N uptake by
vegetation or dilution by deep seepage water.
273
-------
Other studies have also reported similar findings
of decreased groundwater NC>3~ levels in riparian zones
(Schnabel, 1986; Cooper, 1986; Davidson et al., 1986).
Despite the general similarity of results of th€;se
previous studies, the precise mechanisms have not been
clearly determined. Possible mechanisms responsible for
these observations may include: i) dilution of shallow,*
high NC>3~ groundwater by upwelling of deeper, low NC>3~
groundwater, ii) dilution by low NC>3~ recharge percolate
in the forest area, iii) microbial denitrification in
the shallow groundwater, and iv) plant uptake of NC>3~.
The objective of this report is to describe the extent
of variability of groundwater N03~ concentrations in
non-agricultural ecosystems bordering agricultural
land. Our data are preliminary but are important
because of their implications on future sampling
strategies of this and other similar studies.
MATERIAL AND METHODS
The study site is located adjacent to the Wye
River in the Eastern Shore region of Maryland at the Wye
Research and Education Center near Queenstown, Md. The
agricultural soil is a well-drained Matapeake silt loam
(fine-silty, mixed, mesic, Typic Hapludult) underlaid by
sand. The site consists of an agricultural field,
abutted by a narrow grass buffer strip, forested land
and a marsh which is located on the river's edge. The
agricultural land had been cropped to corn in 1984
(moldboard plow tillage) and has been in soybeans since
1985. The grass buffer strip is used as an access road,
and the forest ecosystem is a small patch of deciduous
woods which is bordered by pine trees on the eastern and
western edge and a Phragmities marsh to the south (Fig.
1). In 1985 a preliminary investigation of the site was
initiated. Piezometers were installed in each of the 4
ecosystems (field, grass, forest, and marsh) at the top
of the water table (located ca. 0.5 to 1.5 m below the
soil surface). These piezometers consisted of ceramic
candles attached to 1.5 m lengths of 3/4" polyvinyl
chloride (pvc) pipe. Groundwater samples were collected
for N03~ analyses by drawing a vacuum on the pvc pipe
and subsequently collecting groundwater which
accumulated in the pipe with a tube attached to a
syringe. In 1987 a network of groundwater wells was
installed at the site. These wells were pvc pipe (3/4"
diameter, screened 60-120 cm) which were installed below
the surface of the water table. Water samples were
collected from the wells in the summer and fall of 1987
and in the winter of 1988 for N03~-N and Cl~ analyses.
Groundwater heights were also determined at these
sampling times. Groundwater N03~N (N03 + N02) levels
were determined by an automated cadmium reduction method
(Technicon Autoanalyzer II, Industrial Method # 100-70W,
1973). Chloride was determined using the colormetric
ferricyanide procedure (Am. Pub. Health Assoc.. 1985).
274
-------
RESULTS AND DISCUSSION
From preliminary sampling of the study site,
conducted in 1985, it was observed that at the edge of
the agricultural field groundwater N03~N concentrations
were approximately 13 mg-N/L, and concentrations
decreased markedly in the adjoining grass buffer strip
where levels of ca. 6 mg-N/L were observed. The water
below the forested area contained very low N03~ levels
(ca 0.5 mg-N/L), and groundwater NC>3~ concentrations
below the marsh were undetectable. These results
corroborate earlier work which indicates that
non-agricultural ecosystems may act as buffers to remove
agriculturally derived NC>3~ in shallow, laterally moving
groundwater (Jacobs and Gilliam, 1985; Lowrance et al.,
1984; PeterJohn, 1984).
Based on these preliminary results, the site was
extensively instrumented with groundwater wells in
1987. Wells were established in a pattern which roughly
followed the surface topography from the field, grass,
forest, and marsh (Fig. 1).
PINE
TREES
CORN/SOYBEAN
F2 •
GRASS
STRIP
w
S <--* N
*V
E
DECIDUOUS
U7 • W3»
W2«
Wl •
•FS
F6 •
•F3
PINE
TREES
Figure 1. Map of study area indicating groundwater well locations.
The general slope of the land surface topography
along with water table heights at different times of the
year are shown in Fig. 2. Groundwater gradients were
found to follow the general direction of the surface
275
-------
3
C/)
OC
UJ
h-
UJ
2:
-1
SURFACE ELEVATION
QC
UJ
H-
UJ
z:
-i
cr
LU
K
UJ
-1
MARSH
^SURFACE
MARSH
FOREST GRASS
FIELD
GROUNDWATER HEIGHT (Summer; June 22, 1987)
GRADIENT • 0.84X
GROUNDWATER HEIGHT (Fall; Sept, 11, 1987)
GRADIENT - 0.51X
i
en
DC
UJ n
1- u
UJ
-1
i i
GROUNDWATER
GRADIENT -
o
MARSH
i i
i i i i i
HEIGHT (Winter; January 29, 1988)
0.18X
FOREST GRASS FIELD
i i i i i
15 30 45 60 75 90 105
DISTANCE FROM RIVER (METERS)
120
Figure 2. Surface elevation of land topography and seasonal
water table heights relative to the height of the marsh.
topography, however, neither water table heights nor
gradients were constant over the year. In the summer,
groundwater gradients were greatest (0.8 %) and a low
point in the water table was observed in the second
series of wells located in the deciduous forest (wells
W5, W6, and W7). During the Fall, groundwater levels in
all wells (except the marsh wells), dropped, and the low
water table zone in the forest wells became more
pronounced. In the winter, recharge water caused the
water levels in the wells to rise and the gradient from
276
-------
field to forest decreased to 0.18%. It is hypothesized
that the low forest water table zone observed during the
summer and fall resulted from low rainfall (1987
precipitation was 30 cm less than the 30 year average)
combined with high transpiration rates by the deciduous
trees. Also, in the forested area, laterally moving
groundwater may be coming from both the agricultural
lands and the river.
Chloride concentrations in the wells support this
hypothesis (Fig. 3). In the center and west edge of the
sample site Cl~ levels in the forest wells ranged from
600 to 1000 mg C1~/L which are approximately 20 to 40
fold higher than Cl- concentrations in the field
(average concentration 26 mg/L). However, along the
east edge of the site, Cl- concentrations in the forest
wells were similar to those observed in the agricultural
field. Chloride concentrations in the river were the
same as those of the marsh wells.
3000
60
75
90 105 120
DISTANCE FROM RIVER (METERS)
Figure 3. Chloride concentrations of groundwater along the East
edge, West edge, and center of the study site.
Average N03~ concentrations in the shallow
groundwater below the agricultural field are presented
in Table 1. In this area NC>3~ exhibited a higher degree
of spatial variation than temporal variation.
Coefficients of variation in this area ranged from 34 to
46% on any given sample date, but variability over time
for a given well was low (CVs 8-19%).
277
-------
Table 1. Nitrate concentrations of individual
agricultural field wells for each sampling date.
Sampling date
Wella 6/22/87 9/11/87 10/25/87 1/29/88 Mean
PI
P2
F3
P4
P5
P6
Mean
%CV
8.6
9.5
3.5
3.1
8.5
4.5
6.3
46
7.9
8.6
5.4
, 3.2
8.3
dry
6.7
34
7.8
9.8
4.7
3.0
8.1
dry
6.7
42
N/L —
7.0
8.2
4.3
3.3
10.8
4.0
6.3
46
7.8
9.0
4.5
6.5
8.9
4.3
8.5
8.3
18
19
15
-
aWell designations refer to those indicated on Fig 1.
bCoefficient of variation (%).
Variability of NC>3~ in the grass strip was lower
than the field and was relatively constant both
temporally and spatially with CVs in the range of 25%
(Table 2).
Table 2. Nitrate concentrations of individual grass
strip wells for each sampling date.
Sampling date
Well3 6/22/87 9/11/87 10/25/87 1/29/88 Mean %CV*>
Gl
G2
G3
G4
Mean
%CV
mm .
3.7
4.0
5.5
3.1
4.2
24
4.5
5.1
5.0
3.9
4.6
12
5.1
3.7
2.7
4.0
3.9
25
2.7
3.1
3.7
2.5
3.0
17
4.0
4.0
4.2
3.4
25
21
31
21
*Well designations refer to those indicated on Fig 1.
^Coefficient of variation (%).
Groundwater N03~ concentrations in the forest
showed a higher degree of spatial variability (Table
3). Coefficients of variation for the two Fall sample
dates were ca. 20%, however in the summer and winter,
CVs exceeded 100%. The reason for this appears to be a
spatial heterogeneity in the forest area. The wells in
the center and along the west edge of the deciduous
forest area were dry during the fall (wells W2, W3, W5,
W6, W7). In the summer and winter, when water was
present, these wells typically contained low NC>3~
levels. However, the wells located on the east edge of
the forest area (wells Wl, W4, W8) contained water on
all sample dates and supported nitrate levels similar to
those observed in the agricultural field.
Groundwater N03~ in the marsh were undetectable on
all sample dates (detection limit 0.1 mg-N/L).
275
-------
Table 3. Nitrate concentrations of individual forest
wells for each sampling date.
"Sampling date
Well3 6/22/87 9/11/87 10/25/87 1/29/88 Mean %CVb
Wl
W2
W3
W4
W5
W6
W7
W8
Mean
%CV
„„ _ „ /T
5.1
0.3
1.2
8.2
0.2
0.2
0.2
6.0
2.7
120
8.6
dry
dry
6.2
dry
dry
dry
6.0
6.9
20
9.5
dry
dry
5.5
dry
dry
dry
6.5
7.2
29
y-n/u -•
7.1
0.2
0.1
6.1
3.7
0.7
0.3
5.8
3.0
100
7.6
0.3
0.7
6.5
0.3
0.5
0.3
6.10
24
-
-
19
_
-
-
4.9
*Well designations refer to those indicated on Pig 1.
bCoefficient of variation (%).
Groundwater N03~N concentrations at the site are
inversely correlated to chloride levels (Fig. 4). This
correlation indicates that water from the river may be
entering the groundwater wells located in the forest,
thus, diluting the N03~.
-2
-1 0 1
LOG NITRATE
Figure 4. Correlation between N03~ and Cl~ concentrations for
individual wells in the study site. Regression equation;
log(Cl-) - -1.03 * log(N03-) + 5.01.
To investigate the magnitude of this dilution
effect, predicted N03~-N concentrations were calculated
for the wells based on the average Cl~ concentration of
279
-------
each well and the N03-N concentrations observed in the
agricultural field (Fig. 5). This calculation corrects
the N03-N concentrations in the wells for dilution by
river water.
IS 30 45 60 75 90 105
DISTANCE FROM RIVER (METERS)
120
Figure 5. Measured (circles) and predicted (squares)
concentrations along the Eastern edge, Western edge, and
in the center of the study site. Vertical bars are 95%
confidence intervals about the measured average N03~
concentrations. Sample wells indicated correspond to
locations designated in Fig. 1.
280
-------
Along the east edge of the site, predicted
concentrations roughly corresponded with measured
concentrations, indicating that nitrate is essentially
unaltered as it moves from the field through the
forest. Implicit in this interpretation is the
assumption that the forest wells are actually being fed
by groundwater which contains NC>3~ levels similar to
those observed in field wells Fl and F3. An expanded
well network is needed to confirm or refute this
assumption. In the center and west edge of the field a
discrepancy exists between actual and predicted NC>3~
concentrations. Since dilution of agriculturally
derived NC>3~ by salt water intrusion from the river has
been accounted for, the difference between actual and
predicted NC>3~ concentrations are due to either : i)
denitrification, ii) plant uptake, or iii) dilution by
low N(>3~ recharge. Studies are currently underway to
differentiate the relative importance of these
mechanisms.
SUMMARY
Along the west edge and center of our study site
we observed decreased groundwater NC^-N concentrations
in the grass and forest ecosystems. This observation is
similar to the pattern of decreased N03~N concentrations
in non-agricultural ecosystems reported in previous
studies. However, along the east edge of the site,
N03-N concentrations were similar in the field, grass,
and forested areas, indicating that non-agricultural
ecosystems had little effect on groundwater N03~N
levels. The reason for this discrepancy is unclear but
is is possible that: i) along the east edge of our site,
the forest wells are being fed with groundwater which is
higher in NC>3~ than that present in the agricultural
field or ii) nutrient cycling processes are different in
the pine forest. In any case it is evident that our
system is not as simple as indicated by previous studies
reported in the literature. Past studies may have may
have relied on inadequate sampling or, on the other
hand, a real difference could exist between our site and
others. Before any firm conclusions can be drawn
regarding the mechanisms of N03~ removal by the
non-agricultural ecosystems at our site, the reasons for
the spatial heterogeneity of groundwater NC>3~ in the
forest ecosystem must be elucidated. This will require
a more detailed hydrological description of the site.
ACKNOWLEDGEMENTS
The authors wish thank The University of Maryland,
Wye Research and Education Center for providing access
to the study site.
287
-------
BIBLIOGRAPHY
American Public Health Association. 1985. Standard
methods for the examination of water and wastewater.
16 ed.
Cooper, J.R., J.W. Gilliam, and T.C. Jacobs. 1986.
Riparian areas as a control of nonpoint pollutants.
Jin David L. Correll (ed. ) Watershed Research
Perspectives. Smithsonian Inst. Press. Wash., D.C. pp.
166-192.
Davidson, E.A., and W.T. Swank. 1986. Environmental
Parameters regulating gaseous nitrogen losses from two
forested ecosystems via nitrification and
denitrification. Appl. & Env. Microb. 52:1287-1292.
Jacobs, T.C., and J.W. Gilliam. 1985. Riparian losses of
nitrate from agricultural drainage waters. J. Env.
Qual. 14:472-478.
Lowrance, R. Richard, Robert L. Todd, and Loris E.
Asmussen. 1984. Nutrient cycling in an agricultural
watershed: I. Phreatic Movement. J. Env. Qual.
13:22-27.
PeterJohn, Williams., T., and David L. Correll. 1984.
Nutrient dynamics in an agricultural watershed:
Observations on the role of a riparian forest.
Ecology 1466-1475.
Schnabel, R.R. 1986. Nitrate concentrations in a small
stream as affected by chemical and hydrologic
interactions in the riparian zone. In David L.
Correll (ed.) Watershed Research Perspectives.,
Smithsonian Inst. Press. Wash., D.C. pp. 263-282.
U.S. Environ. Protection Agency, 1983. Chesapeake Bay
Program technical studies: A Synthesis. U.S.
Government Printing Office, 606-490, pp. 634
252
-------
Understanding the Estuary: Advances in Chesapeake
Bay Research. Proceedings of 'a Conference. 29-31
March 1988. Baltimore, Maryland. Chesapeake Research
Consortium Publication 129. CBPlTRS 24/88.
Chesapeake Bay Sediment Monitoring for Water Quality Model Development
Lewis C. Linker
EPA Chesapeake Bay Program
410 Severn Avenue
Annapolis, Maryland 21403
INTRODUCTION
The Chesapeake Bay Program has a comprehensive modeling strategy consisting of three
mathematical water quality models: the Watershed Model, the Steady State Eutrophication
Model, and the Time Variable Eutrophication Model. The Watershed Model covers the
entire 64,000 square miles of the Bay drainage basin and simulates pollutant loads
delivered to the Bay from various land use, population, and point source treatment
scenarios. Essentially complete, the Watershed Model is undergoing a series of
refinements which will be concluded in 1989. The Steady State Model is designed to give
an initial estimate of the relationships among nutrients, eutrophication, and anoxia, and to
provide an initial evaluation of proposed nutrient control strategies. The Steady State Model
was completed in the spring of 1987 and is fully successful in its application. The Time
Variable Model will improve nutrient control strategy evaluation by projecting the degree
and timing of the Bay response to control actions. The Time Variable Model will be
capable of short-term simulations of critical episodic events (e.g. pycnocline tilting) and
long-term simulations of about 30 years. Work was initiated on the Time Variable Model
in October 1987, and will be completed in 1991.
Steady State Model results provided important guidance for the development of the Time
Variable Model (HydroQual, 1987). Among the findings of the Steady State Model:
o The decline in dissolved oxygen (DO) in the bottom waters of the Bay
between 1984 and 1985 was due to increased sediment oxygen demand (SOD),
phytoplankton resperation, and bacterial oxidation.
o Bottom sediments were the largest source of dissolved inorganic phosphorus
(DIP) and ammonia nitrogen during the summers of 1984 and 1985 (the simulated
time period).
o Bay DO and algae are controlled largely by SOD and sediment nutrient flux.
283
-------
o Only management actions that reduce SOD and sediment nutrient flux
improve Bay water quality to any significant degree.
The strong linkage between Bay sediments and water quality requires a multilayer sediment
model to be incorporated into the Time Variable Model framework.
The sediment submodel will have three components: net deposition of particulate organic
matter (POM), diagenesis of POM to dissolved inorganic components within the sediment,
and nutrient flux, the movement of the dissolved inorganic nutrients from the sediment to
the water column (Figure 1). Detailed modeling of Bay vertical processes requires an
intensive sediment monitoring program to provide necessary data for model formulation,
calibration, and verification. The sediment monitoring program described below will begin
in April 1988 and continue for one year.
The sediment monitoring program is a cooperative effort of the Chesapeake Bay Program's
Modeling and Monitoring Subcommittees with expert assistance from HydroQual, Inc.,
the participants of the Sediment Processes and Sediment Modeling Workshop, and the
U.S. Army Corps of Engineers. The generous cooperation of the principle investigators
participating in the sediment monitoring program is gratefully acknowledged. They are:
Walter Boynton, Michael Kemp, Johnathan Garber, Peter Sampou, and Jeff Cornwell,
University of Maryland; Richard Wetzel, Larry Hass, and Bruce Neilson, Virginia Institute
of Marine Science; David Burdige, Old Dominion University; and Grace Brush, Johns
Hopkins University.
EXISTING SEDIMENT DATA
Since 1984, the Maryland Department of the Environment (then the Office of
Environmental Programs) has supported an integrated sediment, water column, and
phytoplankton monitoring program called SONE (Sediment Oxygen and Nutrient
Exchange) (Boynton et al, 1987). SONE focuses on the exchange of material between
sediment, water column, and phytoplankton in the upper and mid-Bay (Figure 2).
Incubated sediment cores are used to measure SOD and nutrient flux, and vertical arrays of
sediment traps are used to measure movement of material between the sediment and water
column. This ongoing study is foundational to the sediment monitoring program. The
SONE study allows the use of an existing data set for a large portion of the Bay, and
informed judgment as a guide for the sediment monitoring program. Other important
studies of Bay vertical processes can be found in an excellent review and synthesis by
Garber (1987).
REQUIRED SEDIMENT DATA
The close linkage between data collection and model development requires correct
anticipation of data needs. Effort is concentrated on the mainbay and on lower estuary sites
of major tributaries.
Sediment station locations
The sampling plan has a total of 25 stations located along the mainstem channel axis and in
the lower tributaries (Figure 2). Stations 2, 4, 5, 7,9, 10, 11, 12, 13, and 14 are existing
SONE program stations (Boynton et al., 1987). Stations 6,7, and 8 are the mid-Bay
transect stations and include shallow lateral stations and a deep water station of a previous
study (Malone et al., 1986). Stations 20, 21, and 22 are lower Bay transect stations.
Lateral transect stations are used in the mid-Bay and lower Bay to capture aspects of the
vertical exchanges between nutrient generating deep waters and adjacent biologically
284
-------
UJ U
<0
^ o
o
UJ
V)
en
00
o
en
(9
X UJ
-
U.Z
I
en —
uj oT
z •
UJZ
(9 •
cn
— J
u.
CO
CO
UJ
o
5
CO
CO
O)
O
O
DC
Q
0
O
•o
o
c
0)
1
So
0)
£
•5
tf»
"c
o
o
a
o
O
ALGAE
_j
i-
t-
DETRITAL
POM
t_
RE SUSPENSION
4V NET DEI
"* *• V OFALG/
SEDIMENT (C N P
FOCUSINO K-N.r
O
© 2
O
UJ
z
vu
o.
'5
ol
£
o>
285
-------
Figure 2. Location of sediment monitoring stations.
| = SOME station. Sone stations have four flux sample
periods, except for stations 2 and 7.
CBP station. CBP stations have two
flux sample periods .except for station 21.
High frequency stations; six flux sample
periods.
/\ « Sediment trap.
256
-------
productive shallow shelves (Malone et al., 1986). The mid-Bay transect stations are
approximately centered in the area of summer anoxia. Southern transect stations are in an
area of summer hypoxia (U.S. EPA, 1983).
There is very little existing information on lower Bay sediment/water column exchanges.
Lower Bay stations 20,21, 22, 24, and 26 are located according to homogeneous sediment
types identified by Wright et al. (1986) as: estuary mouth shoals and spits, Bay-stem
plains, Bay-stem channels, lower estuary or shallow bay muddy beds, and Bay mouth
shoals respectively. The remaining eight stations are located relative to the 2-D model
segmentation scheme. All sediment stations are located close to CBP water column
monitoring stations.
Net Deposition
The net deposition component of the model simulates the input of inorganic and organic
matter to the sediments. Remineralization of organic matter within the sediment provides
material which contributes to SOD and dissolved nutrient fluxes. Inorganic matter
contributes to sedimentation and advects organic inputs down into the sediments.
Measurements of the net input of organic and inorganic matter to the bottom will be used to
calibrate net deposition.
Two vertical array sediment traps will measure the net flux of solids across various water
column layers and to the bottom. Each vertical array trap has three pairs of particle
collectors located above the pycnocline, just below the pycnocline, and just above the
sediment, as in the SONE program (Boynton et al., 1987). Particulate organic nutrients
(particulate organic carbon [POC], paniculate organic phosphorus [POP], paniculate
organic nitrogen [PON], particulate organic silica [POSi]), total solids, BOD, chlorophyll a
and pheophytin, and general algal classification will be determined from the collected
particulates. Algal identification of three broad functional groups (diatoms, non-diatom
eucaryotes, and picoplankton), will provide information on phytoplankton occurrence, cell
size, and settling rate. Sediment traps are located at two stations: station 7, an existing
SONE station in the mid-Bay transect, and station 21 in the lower Bay transect. The
sampling schedule results in 25 sampling periods a year with intensive weekly sampling in
summer months between July and mid-September. Data from the sediment traps will be
used for calibration of net deposition.
Sediment particulate organic profiles will be measured to determine the particulate organic
material (POC, POP, PON, POSi, particulate organic sulfur [POS]) deposited over the
years modeled in the long term simulations. The vertical distribution of bulk POM will be
matched with coincident determinations of average sedimentation rates in duplicate cores.
Radionuclide (14C) and pollen dating techniques (Brush, 1984; Brush et al., 1982) will be
used to determine average sedimentation rates, with particular emphasis on the profile
between 1950 and the present. From this, long term average net deposition of refractory
organic matter will be determined. The vertical profiles of particulate organics will be
determined at all stations. Vertical profiles of pore water concentration will further
characterize sediment composition at most sediment stations (Table 1).
Diagenesis
The diagenesis component of the model simulates the transformation of POM inputs to
dissolved inorganic nutrierits. There are three fractions of diagenic material: a labile
fraction, a refractory fraction, and an inert fraction. These are empirical classifications
based on a fraction that is fast reacting (labile) and is in thermodynamic equilibrium, a
257
-------
TABLE 1. SEDIMENT PLAN ACTIVITIES, STATIONS, AND SAMPLING
FREQUENCY.
ACTIVITY
Nutrient Flux
Denitrification
Sediment
Traps
Particulate
Organic Profile
Pore Water
Profile
Recent Rates of
Sedimentation
Long Term
Diagenesis
METHODS
ambient
shipboard
anaerobic
transitive
acetlyene
blockage
nitrification
(N-serve)
15N
vertical array
particle traps
depth profile
of POM
pore water
concentration
pollen dating
sulfate
depletion
STATIONS
3,43,6,73,8,16,17,18,
20,22,23,24,25,26
4*,5*,9*,10*,11*,12*,
13*,14*
2*,7*,2
same as above
2*,7*,21
3,4a,7a,16,18,20,21
4,14
2,7,21
2,7,21
2 stationsA
7*,21
all stations
2,3,4,4a,5,6,7,7a,8,9,10,
11,12,14,16,18,20,21,22
all stations
2,3,4,4a,6,7,7a,8, 14,
16,17**,18,20,21,22,
23**>24**,25**,26**
SAMPLING
FREQUENCY
2 periods
4 periods
6 periods
same as above
6 periods
2 periods
4 periods
6 periods
6 periods
2 periods
25 periods
1 period.
1 period
r
1 period
1 period
* SONE stations
A Not Determined
** two sample periods for surface sediments
255
-------
fraction that is slow reacting (refractory) but has kinetic rates that are important to the
model, and a fraction (inert) that does not react within the time frames of the model. The
kinetic rates for these three fractions must be determined for the diagenesis component of
the model. Organic material is transformed at rates which are a function of temperature, the
degree of anoxia, the amount of organic matter present, the presence of other chemical
constituents, and the sedimentation rate.
The sediment in the Bay contains a considerable amount of the refractory organic fraction
that has not undergone complete decomposition. This heterogeneously distributed organic
portion of the sediment continues to exert considerable SOD and contribute to nutrient flux
from the sediments. Long term diagenesis will be determined from main-Bay and southern
Bay tributary stations. Diagenesis rates will be determined by long-term sulfate depletion
studies on sediment slurries from three sediment depths: a surface (0-2 cm) sample of
recently deposited material, sediment of a "medium" age collected at a depth between 6 and
8 cm, and older deep sediments collected between 12 and 14 cm. Sediments will be
incubated for 50 to 250 days. Total carbon dioxide (iCXh), NHt, NO2, NO3, PO4, Si,
SO3, ZH2S, CHU, and pH will be measured over time.
Nutrient flux
The flux component of the model completes the sediment cycle by returning inorganic
nutrients to the water column. Model calibration requires extensive temporal and spatial
coverage of SOD and nutrient flux measured under ambient bottom water conditions.
Shipboard measurements of incubated intact cores will be the primary method of data
collection. Cores will be collected and maintained at ambient conditions. Some
observational measurement will be made of the effects of bioturbation. Fluxes of O2,
NH4, NO2, Np3, PO4, Si, SO3, ICO2, CH4, and iHiS will be measured. Hydrogen
sulfide flux will be measured at stations with overlying water DO < 1.0 mg/L.
Nutrient flux studies have three sampling frequencies. Eight stations (4, 5,9,10, 11, 12,
13,14; the SOME stations) have a sampling frequency of four periods. Three stations
(2,7, 21) have a high sampling frequency of six sampling periods. The remaining
stations (3, 4a, 6, 7a, 8, 16, 17, 18, 20, 22, 23, 24, 25, 26) have a low sampling
frequency of two periods. Denitrification measurements have the same sampling
frequencies as nutrient flux measurements but are limited to nineteen main Bay stations.
The sampling scheme is not systematic with respect to time; rather, sampling periods are
established to coincide with an annual cycle of benthic processes. Sampling frequencies
are based on the experience gained from the SONE program. All sediment stations with
four or more sampling periods are sampled according to the following SONE sampling
periods: "(1) a period (April-May) when the early spring phytoplankton bloom occurs, and
nutrients (particularly nitrate) are high in the water column, (2) a period influenced by the
presence of a large macrofaunal community (spring-early summer), (3) a period during
which macrofaunal biomass is low but water temperature and water-column metabolic
activity are high and anoxia is prevalent in deeper waters (August), and (4) a period in the
fall when anoxia is not present and the macrofaunal community biomass is low but
reestablishing" (Boynton et al.,1987).
To improve temporal coverage of sediment nutrient flux and denitrification, three high
frequency stations have two additional sampling periods. Additional sampling periods
include: (1) a period in the winter (December - early March) when anoxia is not present,
metabolic activity is low and nitrate is increasing in the water column, and, (2) a period in
the early fall (September) just prior to the break-up of anoxia .
259
-------
The majority of the stations (14 out of a total of 25 stations) have a low sampling frequency
of two periods; one sample period when anoxia is prevalent in deeper waters (August), and
one when anoxia is not present (spring-early summer). Location and frequency of the
various field and laboratory measurements are described in Table 1.
Paired with the measurement of ambient nutrient flux are measurements of anoxic fluxes.
Anoxic fluxes are measured at all stations with the same frequency as ambient fluxes.
Anoxic fluxes are designed to measure rapid, short-term changes in flux due to the die-off
of benthic infauna and the rapid chemical changes caused by decreasing oxygen
concentrations and oxidation-reduction potential. Pore water concentrations of NH*, NC>2,
NO3, PO4, Si, SO3, ICO2, CH4, ZH2S, Fe, Mn, and pH will be measured with short-
term anoxic incubations of surficial (0-2 cm) sediment.
Denitrification is a major sink for nitrogen in the Bay and has seasonal and region-specific
properties which must be delineated for successful sediment model development (Twilley
and Kemp, 1985). Three methods for measuring denitrification will be used. At all
stations sampled, an acetate inhibition method will be used with the same sampling
frequency as for nutrient flux. Acetate inhibits the final reaction (with N2 the product) of
the denitrification reaction path. This allows the analysis of an intermediate product
concentration without the problem of background contamination. Nitrification potential
will be measured at a limited number of stations by N-serve treated control sediment
slurries, an inhibition technique that prevents nitrification in the treated slurries. Control
slurries will be compared to untreated test cores. Calibration of the denitrification and
nitrification measurements will be with 15N labeled nitrate and ammonia respectively.
Details of sampling periods and times are in Table 1.
CONCLUSION
Work on acquiring data for calibrating the sediment submodel will be initiated in April,
1988. The sediment monitoring program is based on the consensus recommendations of
the expert panel of the Sediment Processes Workshop, and the Modeling and Monitoring
Subcommittees of the Chesapeake Bay Program. The sediment data outlined above are
essential to the Time Variable Model of the Chesapeake Bay.
290
-------
REFERENCES
Boynton, W.R.; Kemp, W.M.; Garber J.; Barnes, J.M. Sediment and water column
interaction in the Chesapeake Bay. In: State of the Chesapeake Bay Second Annual
Monitoring Report. 1987.
Brush, G.S.; Martin, E.A.; DeFries, R.S.; Rice, C.A. Comparisons of 210Pb
and pollen methods for determining rates of estuarine sediment accumulation. Quater.
Resear. 18:196-217; 1982.
Brush, G.S. Stratigraphic evidence of eutrophication in an estuary. Wat. Res. Resea.
20(5):531-541; 1984.
Garber, J.H. Benthic-Pelagic Coupling in Chesapeake Bay. In: Perspectives on the
Chesapeake Bay: Recent Advances in Estuarine Sciencies. Eds: M.P. Lynch and E.C.
Krome, Chesapeake Research Consortium. Gloucester Point, VA. 1987.
HydroQual, Inc. Development of a coupled hydrodynamic/water quality model of the
eutrophication and anoxia process of the Chesapeake Bay. EPA Contract No. 68-03-3319.
Annapolis, MD. 1987.
HydroQual, Inc. Workshop Number 2: Sediment processes and sediment modeling
workshop, December 2-3, 1987. USCOE Contract No. DACW39-88-C-0004. U.S. Army
Corps of Engineers, Baltimore District. Baltimore, MD. 1988.
Malone, T.C.; Kemp, W.M.; Ducklow, H.W.; Boynton, W.R; Tuttle, J.H.; Jonas, R.B.
Lateral variation in the production and fate of.
Twilly, R.R.; W. M. Kemp. Preliminary results on the significance of sediment
denitrification to the fate of nitrogen in the Chesapeake Bay. Final report to the EPA
Chesapeake Bay Program. Annapolis, MD. 1985.
Wright, L.D.; D.B. Prior, C.H. Hobbs; R.J. Byrne; J.D. Boon; L.C. Schaffner; M.O.
Green. Spatial variability of bottom types in the lower Chesapeake Bay and adjoining
estuaries and inner shelf. Estuarine, Coastal and Shelf Science, Vol. 24, P. 765-784.
1987.
291
-------
CONCURRENT SESSIONS
AND
POSTER SESSION:
PHYSICAL PROCESSES
Chairs:
William Boicourt
Horn Point Environmental Laboratory
University of Maryland
Evon P. Ruzecki
Virginia Institute of Marine Science
College of William and Mary
-------
Understanding the Estuary: Advances in Chesapeake Abstract only
Bay Research. Proceedings of a Conference. 29-31
March 1988. Baltimore, Maryland. Chesapeake Research
Consortium Publication 129. CBPtTRS 24/88.
Hampton Roads Circulation: The Combined Effects
of General and Meso-Scale Features
Evon P. Ruzecki and David A. Evans
Virginia Institute of Marine Science
The College of William and Maryand
Gloucester Point, Virginia 23062
Historical data sets from the Hampton Roads region of the James River
estuary and the James River Hydraulic Model obtained in 1964 and 1968
respectively are re-examined with respect to hypothesized net cyclonic
transport in the lower James and a meso-scale circulation feature associated
with a more recently examined (1987) frontal feature in this region. The 1964
data illustrate the development and decay of an upstream-directed subsurface
jet-like feature during flood tide which coincides with the newly described
front while subsequent hydraulic modal data indicated net cyclonic transport
in the lower James over several tidal cycles. The combined effects of both
circulation features are examined with respect to recirculation of oyster
larvae spawned upstream and the upstream directed injection of salt and
municipal/industrial contaminants from the Hampton Roads region. Effects of
these circulation patterns are illustrated by sequential 'openings' of
upstream shellfish beds which were closed as a result of a recent (1985)
sewage treatment plant incident in the lower James.
294
-------
Understanding the Estuary: Advances in Chesapeake Abstract only
Bay Research. Proceedings of a Conference. 29-3 J
March 1988. Baltimore, Maryland. Chesapeake Research
Consortium Publication 129. CBPlTRS 24/88.
A Theory of Tidal Intrusion Front and its Practical Application
Albert Y. Kuo and Robert J. Byrne
Virginia Institute of Marine Science
The College of William and Mary
Gloucester Point, Virginia 23062
Paul V. Hyer
Kentron Inc.
303 Butler Farm Road
Hampton, Virginia 23666
A one-dimensional analysis for a two-layer fluid was formulated to
describe the characteristics of a tidal intrusion front. The analytical
result relates the vertical transport through the front (i.e./ the diving
depth of the heavier fluid) to the densimetric Frounde number of the
approaching flow and the depth change across the front. It also defines a
critical Froude number, above which no steady state solution exists.
Interpretation of the result provides information on the movement and maximum
transport capacity of the front.
The theory was used to interpret the characteristics of the front
observed off Newport News Point in the lower James/Hampton Roads of Virginia.
Because of the phase difference between the tidal currents on the two sides of
Newport News Point, a convergent zone is formed at early stage of flood tide.
The density difference between the two water masses is large enough to make
the more saline Hampton Roads water dive beneath the fresher water of the
lower James, thus form a tidal intrusion front. The front moves upriver as
flood current strength increases, slows down and intensifies as it encounters
a steep drop of bottom elevation. The observed diving depths of heavier water
were explained by the theoretical result.
The theory was used to predict the impact of a proposed man-made island
on the frontal characteristics, particularly its ability to entrain oyster
larvae to the lower portion of water column in which the net transport is
upriver toward seed oyster beds. A 400 acre island was proposed to be
constructed on the Hampton Flats, a shallow embayment on the north side of
Hampton Roads, which is a broad water body at the mouth of the James River.
The proposed island would be located immediately downriver of the front. Its
effect on the flood current approaching the front was quantified with a two-
dimensional (in horizontal plane) numerical model, when combined with
inferences from oyster larvae studies, indicate that the transport capacity of
the front would be markedly reduced by island construction at the proposed
sites.
295
-------
Understanding the Estuary: Advances in Chesapeake
Bay Research. Proceedings of a Conference. 29-31
March 1988. Baltimore, Maryland. Chesapeake Research
Consortium Publication 129. CBP/TRS 24/88.
Changes in Circulation and Salinity from Increased Channel Depth
in the Baltimore Harbor
Peter Olson
The Johns Hopkins University
Department of Earth and Planetary Sciences
Baltimore, Maryland 21218
Vincent Grano
Space Telescope Sciences Institute
Baltimore, Maryland 21218
1. INTRODUCTION
This paper presents results of numerical simulations of the circulation
within the Baltimore Harbor system, in an effort to determine the
quantitative effects of channel enlargement on the circulation and salt
distribution within the Harbor, subject to the external environmental
conditions found at various times throughout a normal year. The
incentive for this study comes from the plan to dredge the principal
navigation channels in the Patapsco River and in certain segments of the
main stem of Chesapeake Bay. Within the Baltimore Harbor system, the
dredging program calls for enlarging the navigation channels from their
present dimensions, approximately 12.8 meters deep and 244 meters wide
in the main ship channel, to the planned dimensions of approximately
15.3 meters deep and 244 meters wide. The long term average flow
pattern of Baltimore Harbor includes a three-layered gravitational
circulation (Stroup et al. , 1961; Boicourt and Olson, 1982; Olson et
al., 1982). The magnitude of this circulation depends upon the depth of
the channel and on the vertical stratification imposed on the Harbor at
its mouth by the main stem of Chesapeake Bay, both of which increase
with dredging.
Figure 1 is a location map, showing points referred to in this paper.
The mouth of the Harbor, which is taken to be the cross section at which
the Patapsco River joins the main stem of Chesapeake Bay, is chosen to
be the line joining Cedar Point on the south shore of the Patapsco with
the North Point on the north shore. Ranked in terms of volume, the
Harbor system has four major branches: Bear Creek, Curtis Creek, Middle
Branch and Northwest Branch. We designate the Middle Branch as being
the head of the Harbor. The mean low water volume of the Harbor system
thus defined is 468xl06 m3. The largest Harbor tributaries, Bear and
Curtis Creeks, have volumes of 40xl06 and 26xl06 m3, respectively
(Cronin and Pritchard, 1975). If we exclude all harbor branches, then
the main channel of the Harbor through Middle Branch is 21 km long and
296
-------
\
Figure 1: Reference map of the Baltimore Harbor system showing major
navigation in channels, plus locations referred to in the text.
-------
has a volume of approximately 364xl06 m3. In the outer portion of the
Harbor, beyond Bear Creek, the cross sectional area is approximately
20,000 m2. The average surface width in the outer portions of the
Harbor is approximately 3 km and the average depth is 7m. However,
average depth is not a good measure of the Harbor bathymetry, because
the greatest depths occur within relatively narrow ship channels.
Baltimore Harbor circulation has been the subject of study for several
decades (Hachey, 1934; Garland, 1952; Stroup et al., 1961; Wilson, 1970;
Hansen and Rattray, 1972; Hansen and Festa, 1974; Long, 1977; Boicourt
and Olson, 1982; Olson et al., 1982). In addition to the interest in
Baltimore Harbor circulation generated by commercial and environmental
concerns, it has received considerable attention from physical
oceanographers because it represents a mean flow induced primarily by
small scale mixing processes. The essential physics of the three layer
circulation is represented in the diagrams in figure 2. The low
salinity surface water and the high salinity bottom water from nearby
portions of the Chesapeake Bay enter the Harbor, where they are
partially mixed by the action of the tides and winds. The resulting
water is of intermediate salinity and density. It discharges outward
into the Chesapeake Bay as a mixed layer, at a depth intermediate
between the two incoming layers. Under equilibrium conditions the total
salt content is the same for every cross section, and there is zero net
salt and volume flux through any cross section. Nevertheless, there is
a large volume transport in each of the individual layers. Field
measurements by W. Boicourt (Boicourt and Olson, 1982) indicate three
layer flow velocities range from 2 to 8 cm/sec when averaged over 3 to 7
day intervals.
The classical two-layer estuarine circulation, which is found within the
main stem of the Chesapeake Bay, is evidently not present to any
significant degree within the Baltimore Harbor system. This circulation
is induced by partial mixing between salt water and fresh water derived
from terrestrial run-off (Pritchard and Carpenter, 1960). The absence
of an important contribution from this type of circulation can be
explained by the low volume transport typical of the Patapsco River
drainage. The average discharge from the Patapsco River is approximate-
ly 2.6 m3/sec and the discharge from the Jones Falls into the Northwest
Branch is significantly smaller than this. If we project the Patapsco
discharge onto the mean cross sectional area of the outer portion of the
Harbor, the resulting mean velocity is only .013 cm/sec. Thus, even
allowing for the enhancement of a two layer circulation by mixing (which
can amplify the circulation in each layer by a factor 10 or 20) the flow
induced by the Patapsco River discharge, on average, will amount to
significantly less than 1 cm/sec. For this reason, it is justifiable to
neglect the influence of Patapsco River discharge on the laterally
averaged circulation. The exception to this would be brief times
following storm run-off.
2. NUMERICAL MODEL
The numerical model is a finite difference representation of the
governing equations for laterally-averaged flow in a stably stratified
shallow channel. We fix a coordinate system with the horizontal x-axis
parallel to the ship channel as shown in figure 1, and the vertical
z-axis positive upward. The origin of coordinates is fixed by mean sea
level at the head of Middle Branch. The domain consists of Middle
Branch plus the main stem of the Patapsco estuary; we exclude from
consideration all other tributaries and branches. The conservation of
horizontal momentum, in laterally averaged form, is
a-
d_B_
'dz
(1)
295
-------
<
CD
O
O
O
—H
CD"
Harbor Head
depth
O_
5'
CD
^ >>\
V
\
\
\ \
k
i
1
'
1
f 1 1
J
si i '
cO
(A
§: 1
-• 1
^f 1
5' i i
5, ' 1 '
o
* 1
1
) 1
•
v
1
I
/
f 1
}
1
I
1
1
, f
1
'a 1 Is-
x ' $
^ 1 S
5 i i:
CD i <*£
"1 t~
0 , 5
1 E M±
' ^ ' o
0 | *
* '
1
1
i
H
u~
CD
CD
Q
CD
CD
Q.
Tl
O
O
CD
3
Q
o'
Harbor Entrance
Figured: A schematic diagram showing essential physical processes for
the three-layer density driven circulation. Upper Diagram: Low salinity
surface water and high salinity bottom water from Chesapeake Bay enter
Baltimore Harbor, partially mix, and discharge out of the Harbor as a
mixed layer. Lower Diagram: Typical profiles for salinity and
horizontal velocity.
-------
The laterally averaged conservation of salt is
(2)
dz
The conservation of mass is
(iuB) = 0 (3)
which has the depth integrated form of
d d C*
— [rjB(x,0)] +— I uBdz = 0 (4)
Finally, the equation of state connecting salt to density may be written
n
— = (l+/?s) . (5)
Po
In equations (l)-(5), B(x,z) is lateral width, D is depth, g is gravity
ft is the saline coefficient, PO is freshwater density, k is the drag
coefficient, (Kz, K,) and (N2, Nx) are (vertical, horizontal) mixing
coefficients for salt and momentum, respectively. Representation of
vertical mixing by diffusion coefficients is consistent with mixing by
sidewall boundary layer processes (Phillips, Shyu and Salmun, 1987).
The horizontal diffusion coefficients provide numerical stability and do
not strongly influence the solution.
The appropriate boundary conditions are:
i) On the channel bottom, z - -D
w°rz=0' ".fr*"i«i <6a)
ii) On the surface, z - 0
»-£-0 (66)
iii) At the Harbor head, the freshwater influx is ignored, as it makes
a negligible contribution to the circulation. Therefore we have x - 0
iv) At the Harbor mouth, x - L, we specify the surface elevation
n = nm(n • (6d)
The salinity of inflowing water is specified
s = sm(^,f) if u<0, (6e)
and we assume an advective balance for outflow
£(s5) = ^(usB) if u>0 . (6/)
ol ox.
In addition to the driving provided by mouth stratification and
fluctuations in sea level, we specify the longitudinal component of the
wind stress
at z = 0 . (7)
300
-------
The data necessary to initialize a run consist of the initial
distribution of salinity and surface elevation
o *~ o f "v *y t ^ O i
o Q o ^_ -A. , ,c , t — \J )
Calculations are made for three domains:
i) The base configuration, consisting of the main branch of the
Patapsco estuary plus the Middle Branch, both with a dredged channel
12.8 m deep, 244 m wide, 21 km long, and having a total volume of 264 y.
106 m3.
ii) The plan configuration, which is identical to the base
configuration except that the channel is deepened everywhere to 15.3 m,
and 244 m wide. The total volume of the plan configuration is 276 K 106
m3.
iii) The ideal rectangular channel. A sequence of calculations is
carried out in a rectangular channel 12.8 m deep, 1 km wide and 21 km
long. The purpose of these simulations is to check our parametric
representation of the three layered flow in an idealized channel, in
which topographic effects are minimized.
The grid used for both base and plan domains consists of 22 equally
spaced cross sections, with 20 grid points in the vertical at each cross
section. The details of the computational algorithm can be found in
Wang and Kravitz, 1980 and Olson et al., 1982. We have carried out the
set of calculations summarized in tables 1 and 2. Table 1 gives the
pertinent data for the steady state calculations, which are driven by
imposing at the Harbor mouth the top and bottom water salinities, and an
astronomical tide. The astronomical tide data was taken from the Fort
McHenry tidal station records. Top and bottom water salinities were
taken from the published results of the Chesapeake Bay Hydraulic Model
experiment (Granat and Gulbrandsen, 1981, plate no. 78) in which the
effects of channel dredging in the main stem of Chesapeake Bay were
modelled. We assume these results are indicative of salinity changes
expected in the main stem, and we use them as boundary conditions for
the plan geometry. Table 2 lists the parameters for the model runs
which include meteorological forcing. The applied wind stress was
calculated from hourly wind data recorded at Baltimore Washington
International Airport during the appropriate season. Nontidal sea level
fluctuations during the same time interval come from the recordings at
the Fort McHenry tide station.
A number of useful parameters can be computed which characterize the
overall Harbor response, including volume flux, residence (flushing)
time and stratification in the Harbor interior. For the three layer
flow, four quantities prove to be diagnostic measures of the Harbor
response: the mixed layer volume flux Q, the residence time T, the
interior stratification 4S» and 6, the length scale of the flow. These
are defined as
i
uBdz x = L, u>0
-D
dx
301
-------
Dimensional Analysis
In order to apply the numerical results to cases whose external
conditions differ from those in table 1, it is desirable to establish
scaling laws which correctly determine the Harbor response to arbitrary
changes in both channel depth and mouth salinities.
For the three layered flow, in steady state, we introduce the following
dimensionless variables (denoted by primes)
(x'.z') =(x/L,z/D)
(u'.uT) -(Lw.Dul/K,
s' =s/Asm (10)
B' = 5/5
where 2 is the depth- averaged width, related to the cross -sectional area
A by
Equations (1) and (2) become
^ ' dz^ ' \_dz' dz' A2dx' dx'
with parameters
P =N
A -L/D
IN,KK =NX/NZ,KX/KZ (12)
In dimensionless form, the response parameters (9) become
Q' - fu'B'dz'; x'-l,u'>0
T- V l
T = --
DAQ'
As\ =s'(0,l)-s'(0,0) (13)
'lu'(x',- 1/2)
502
-------
The response of the three-layered flow depends strongly on the Rayleigh
number R, which measures the relative importance of stratification
versus mixing. Its value for Baltimore Harbor can range between 106 and
1012, depending on external conditions. By contrast, the dependence on
other parameters is weak. The aspect ratio A is fixed by Harbor
geometry. Our representation of momentum and salt mixing by constant
diffusivities requires, for internal consistency, that the turbulent
Prandtl number P has a value near unity. The three-layered flow is
insensitive to values of mixing anisotropy paramenters I^I^ less than
105.
3. STEADY STATE CALCULATIONS
In this set of calculations the flow is driven only by the astronomical
tide and the vertical stratification of Chesapeake Bay imposed at the
Harbor mouth. The surface and bottom water salinities used as boundary
conditions are listed in table 1.
With no meteorological forcing present explicitly, model results
indicate that the mean circulation settles into a tidally-averaged
steady state three-layered flow within a period of 10-30 tidal cycles,
depending on the specific initial conditions used. The structure of the
three-layered circulation is most sensitive to the value assigned to the
vertical mixing coefficient K . Numerical experiments indicate that a
vertical mixing coefficient lying within the range
0.1 < Kz < 1.0 cmz/sec
best matches the observed Harbor salinity cross sections, as measured by
Boicourt and Olson (1982). For purposes of comparison, we carry out all
calculations with two values: Kz-0.3 and Kz-1.0. These values result in
best agreement with observed salinity patterns under weak and vigorous
mixing conditions, respectively.
When the freshwater discharge into the Chesapeake .Bay reaches its
seasonal low in summer and early autumn, the average salinity in
Baltimore Harbor is highest, and the stratification is weakest. A weak
three-layered flow is expected under these conditions. Results of
calculations for base and plan channel configurations are shown in table
3, and in figures 3-4. The salinity patterns are essentially the same
for both base and plan configurations; the isohalines diverge from the
Harbor mouth in a fan pattern which is the characteristic signature of
the three-layered circulation. The important difference between plan
and base is found in the magnitude of the stratification. In the base
configuration, the inner two-thirds of the Harbor is essentially
homogeneous, while in plan configuration a stratification of 2-3ppt
exists as far up as Key Bridge (point b). The stronger stratification
found in the plan configuration is reflected in the magnitude of the
three-layered flow. For the base configuration, a weak three-layered
circulation (with velocities less than 2 cm/sec) is found near the
Harbor mouth only; the gravitational circulation inside the Harbor is
negligible. By contrast, in the plan configuration a vigorous
three-layered flow is found at the mouth (with velocities up to 7
cm/sec) and a weak but still significant mean circulation exists within
the Harbor interior.
Equivalent tendencies are seen in figure 4, which shows the same
calculation carried out with a mixing coefficient of Kz-1.0. The
enhanced mixing results in more homogeneity than the case illustrated in
figure 3, but the relationship between base and plan response is
preserved. In particular, the strength of the three-layered flow at the
Harbor mouth is essentially the same for both choices of vertical mixing
coefficients.
Table 3 summarizes the comparison between base and plan configuration
response, in terms of the response parameters defined by equation (9).
The increase in channel depth plus the concomitant increased mouth
stratification increases the outflowing volume flux within the middle
layer from approximately 87 m3/sec to 210 m3/sec. This translates into
a decrease in residence time from approximately 48 days to 20 days, a
303
-------
Table 1: Calculation Parameters - Steady State
Channel Low Flow High Flow
length depth ^)S duration length depth ^S duration
(mouth) (mouth)
base 21 km
plan 21
ideal 21
12.8 m 1.8 ppt 50 td cy 21 km
15.3 m 4.85 50 21
12 many variable
12.8 m 9.5 ppt 50 td cy
15.3 15.75 50
Table 2: Calculation Parameters - Variable Forcing
Channel
Low Flow
High Flow
length depth AS duration length depth
-------
15
20km
•14.
10
mi
R = 3.09x10'°
cm/s
-10 0 10
AS = l.80%o
K2 = 0.3 cme/s
10-
m
20km
10
m
18-
R=l.43x10"
cm/s
-10 0 10
AS= 4.85%o
K7 = 0.3 c mvs
10-
m
Figure 3: Comparison of base configuration (upper) versus plan
configuration (lower) at steady state under low flow conditions. Shown
are laterally averaged salinity cross sections contoured in ppt from
Middle Branch to the Harbor mouth, plus three representative velocity
profiles. Nominal locations: a-Middle Branch, b-Key Bridge-Curtis Bay,
c=North Point. Calculations made with a vertical mixing coefficient K2
- 0.3cm2/sec.
305
-------
5
0-
m
5 10 15 20km
\4
- -**"" .--IS""""
hx -*^"^
X /
1 1 / /I
R= 2.78x10*
cm/s
-10 0 10
to-
rn
A S = 1.80%
00
Kz = I.OcmVs
20km
R = l.29x10
10
cm/s
•10 0 10
10-
m
AS = 4.85 %o
Kz = I.OcmVs
Figure 4: Same as figure 3 except that calculations were made with Kz
1.0 cmz/sec.
306
-------
reduction of 60%. The interior stratification, measured within the
Middle Branch, increases by approximately 1 ppt. The horizontal length
scale of the three-layer flow, the distance of penetration into the
Harbor, is essentially the same for base and plan configurations.
High flow conditions
As the freshwater discharge into the Chesapeake Bay reaches its seasonal
high--in late winter and early spring--the average salinity in Baltimore
Harbor is lowest and the mouth stratification is at a maximum. Under
these conditions the strength of the three-layer flow is also at a
maximum.
Results of calculations for both base and plan channel configurations
are shown in table 4 and in figures 5 and 6, for high flow conditions.
From figure 5 it is apparent that the effect of increasing Chesapeake
Bay stratification is to stratify the entire Harbor; this is especially
prominent for the planned configuration. For example, within the Middle
Branch (point a) a vertical stratification of 2.5 ppt occurs in the base
configuration experiment, and a stratification of over 6 ppt occurs in
the plan configuration. As can be seen from the velocity profiles, this
enhanced stratification results in a stronger three-layered flow
throughout the Harbor system, when compared to weak stratification
conditions. This is particularly true for the enlarged channel. At
high flow, in the base configuration, three-layered flow velocities
approach 8 cm/sec at the mouth and drop to less than 2 cm/sec past Key
Bridge, while in the plan configuration, at high flow, velocities reach
17 cm/sec at the mouth and remain as large as 7 cm/sec in the inner
portion of the Harbor.
As was found for low flow conditions, channel enlargement significantly
affects the Harbor response parameters. Table 4 is a summary of these.
The outward volume flux within the mixed layer increases from
approximately 300 m3/sec in the base configuration to 500 m3/sec in the
plan configuration. This translates into a decrease in residence time
from approximately 14 to 8 days. The stratification near the Harbor
head increases by 3-6 ppt as a result of the deepened channel, depending
on the strength of turbulent mixing processes. For a vertical mixing
coefficient of Kz - 1.0 cm2/sec (strong mixing), the increase in the
stratification in the Harbor interior is approximately 3 ppt, while for
Kz - 0.3 cm2/sec (weak mixing) the increase is by 6 ppt.
Parametric description of three layer flow
In order to make predictions about Harbor response to arbitrary changes
in channel depth and Chesapeake Bay salinities, it is desirable to
generalize these results in the form of simplified scaling laws. For
the three-layered flow in a tidally-averaged steady state, the analysis
of Section 2 indicates the critical dimensionless parameter is the
Rayleigh number R. To further our understanding of how the
three-layered flow depends on R, we have conducted an additional series
of numerical experiments using the rectangular version of the Harbor,
consisting of a channel 1 km wide, 12 m deep and 21 km long. We have
varied both the mouth stratification and the mixing coefficient Kz in
order to vary the Rayleigh number between the limits
108 < R < 3xl012
Figures 7-13 show the results of these simulations. The first three
figures, 7-9, illustrate the transition in salinity and velocity fields
as R is increased. At relatively small values of R (figure 7) mixing is
relatively strong while stratification is relatively weak and a
significant stratification occurs near the Harbor mouth only. The three
layer flow exists only in the outer one-third of the Harbor; internal to
that is a weak two layer flow. As the Rayleigh number is increased to
moderate values (figure 8) the stratification and the three layer flow
fill the outer two-thirds of the Harbor. At large values of the
Rayleigh number (figure 9) both the stratification and the three layer
flow extend throughout the whole length of the Harbor.
307
-------
Table 4: Critical Response Parameters, High Flow Conditions
Channel
Base
Plan
Residence
Time
15
13
8.
8.
days
8
1
Outward Interior Length
Flux Stratification Scale
273 m3/sec
324
491
535
3.
0.
9.
3.
3 ppt
3
2
0
7.
4.
7
8
0 km
2
.0
.6
Mixing
Coefficients
0.
1.
0
1
3
0
.3
.0
cm2/sec
Table 5: Scaling Laws for Three Layered Flow
Parameter Scaling Power Law
Baltimore Harbor Rectangular Channel
Mixed Layer
(Outward) Flux
Residence Time
Interior
***
D2/KZ
JS
Q'
T'
-------
R = l.64x 10"
cm/s
-10 0 10
10-
m
10
15
20km
AS = 9.50%o
Kz= 0.3 cmz/s
20km
R= 4.64x10"
cm/s
-10 0 10
AS = 15.75 %o
Kz= 0.3 cmvs
Figure 5: Comparison of base configuration (upper) versus plan
configuration (lower) at steady state, under high flow conditions.
Calculations made with a vertical mixing coefficient Kz - 0.3cm2/sec.
309
-------
10
m
10
15
20km
R = 1.48x10
10
cm/s
-10 0 10
10-
m
AS= 9.50%o
Kz= I.OcmVs
R = 4.18x1 0
10
cm/s
-10 0- 10
10-
m
AS= I5.750/*
'00
K = I.OcmVs
Figure 6: Same as figure 5 except that calculations made with K - 1.0
cmz/sec.
310
-------
5-
10- /
m i /
a
5 10 15
y • 'V
r
1
20km
V 1
6^
r^
b c
R = 6.41 x 10* AS = 5%o Kz =
cm/s
-10 0 10
.. .
.. /
m 1
Figure 7: Steady
*/\0 cme/s
/ ^^
> l
/ )
a [ b f c
state response of idealized rectangular
channel 1 km
wide, 12 m deep, 21 km long, to imposed stratification of 5 ppt at R
6.41xl08. Shown is the laterally-averaged salinity cross section plus
three vertical profiles of horizontal velocity.
0
15
20 km
10
m
R= 1.28 x 10
cm/s
10
-10
10
10-
m
L
AS= I0%o
Kz= 1.0 cmVs
Figure 8: Steady state response of idealized rectangular channel to
imposed stratification of 10 ppt at R - 1.28xl010.
311
-------
10
m
a
R= 1.42 x 10"
cm/s
-10 0 10
10 -
m
III
a
10
20 km
AS = io%o
K2 = 0.3 cmVs
Figure 9: Steady state response of idealized rectangular channel to
imposed stratification of 10 ppt at R - 1.42xlOn.
312
-------
N
"b
0>
0)
>\
o
cc
1-0
00
O
O 1-1 tfl
Or-l «
MS o
I> a) oi
9)
)-(
bl
•H f fl -H
fc,
313
-------
°0
CM
."b
_o
£
3
JC
o .?
"b o>
o
o:
CM
b
«
o
-------
o1""
O
N<
'o
(M
"b
o ^
0>
0
o>
>%
o
o:
C M >-i M
O fl O 05
r-l 4-1 rH
^ (N j
3
-------
o °o b
q 46 u 9 n I DJ 6 9 j u i
CM
o
L~2
^
\
O
£
o»
O
c o> s
O H cfl
•i-l O i-H
4J S
03 >H ij
^ -U 3
CO
0) p^ i-l
O
c e
•H * 3
4J.-H «U 0)
J-
4J
o>
BMW)
MO
•H|i-l
-------
The dependence of the response parameters on R is shown in figures
10-13. Plotted on each figure are the results using the idealized
rectangular channel, plus the results from the simulations already
described in this section for plan and base configurations of Baltimore
Harbor. Although there exists some scatter, it is clear that the
response parameters for the three -layered flow can be scaled in terms of
simple power law functions of the Rayleigh number, over most of the
range of Rayleigh numbers considered. In figures 10-13 we have fit the
results to power laws of the type aRn; the values for a and n are given
in Table 5.
As the Rayleigh number approaches and exceeds 1011 a transition occurs
and the response parameters diverge from simple power law behavior.
This transition occurs when the inflowing layers penetrate all the way
to the Harbor head without appreciable mixing. In that limit the
parameter ^Sh' asymptotes to unity, and Q' and T' asymptote to constant
values. Thus, the power law regime applies to circumstances in which
the stratification at the head is much weaker than at the mouth. All
available data indicates that this condition is met in Baltimore Harbor,
that the appropriate Rayleigh number is less than 1011, and that the
flow lies in the power law regime.
From figures 10 and 11, and table 5, it is evident that the outward flux
in the mixed layer of the three- layered flow, at the mouth of Baltimore
Harbor, obeys a scaling law of
Q' - 1.4X10-2R1/2
and similarly the flushing time obeys
T' - 9xlO*R-1/2
Expressed in terms of dimensional quantities the above formulas become
Q -
T
Both volume flux and residence time depend only on mouth salinity and
channel dimensions. Importantly, they are independent of mixing. The
fact that they do not depend on the strength of vertical mixing makes
these scaling laws additionally robust.
The scaling law for outward volume flux given in table 3 was first
proposed by Long (1977) on the basis of a simplified analytical model.
For a rectangular channel, Long obtained
Q' - R1/2^
Long's coefficient 1/8 is greater than our result by a factor of about
3, but there is agreement on the Rayleigh number dependence.
4. CALCULATIONS WITH TIME VARIABLE METEOROLOGICAL FORCING
The influence of local and non-local winds on circulation in Chesapeake
Bay and its tributaries has been extensively documented (see Wang and
Elliott, 1978; Wang, 1979a,b; Grano, 1982). These studies leave little
room for doubt that wind strongly affects the circulation within the
Chesapeake Bay system over time scales of two to ten days. The
experiments performed in this Section are similar to those presented in
Section 3, consisting of a direct comparison between the base and plan
Harbor channel configurations, both under low flow and high flow
conditions. In addition, the experiments presented in this section
include the effect of surface wind stress and meteorological tides. In
all the simulations presented in this section, a value of the vertical
mixing coefficient K2 - 1.0 cm2/sec was used. The duration of each of
these experiments was 50 tidal cycles.
317
-------
The variability in the circulation induced by meteorological forcing is
best summarized as time histories of overall response parameters as
shown in figures 14 and 15. The first two traces show the variation in
the applied wind stress and sea level at the mouth over the last 30
tidal cycles. During low flow conditions (figure 14) the wind is
generally weak, and when strong wind events do occur they usually come
from the south and southeast. In response to wind events from that
direction the Chesapeake Bay tends to fill, resulting in an increase in
sea level at the Harbor mouth. Thus there exists a negative
corrrelation between sea level at the Harbor mouth and the local
direction of the wind stress.
While the wind stress is generally weak during this portion of the
hydrological cycle, the stratification within is sufficiently weak also
so that the wind driven component of the circulation can dominate flow
within the Harbor. The last two traces on figure 14 show the Harbor
response, in terms of the horizontal velocity as a function of time, at
three specified depths near the Harbor mouth. The upper graph is the
response of the base configuration, while the lower graph is response of
the plan. It is clear from these traces that the response is basically
baroclinic, with the surface layer following the wind and the middle and
deep layers opposing the wind. In the base configuration, velocities in
the near surface layer show excursions up to 15 cm/sec with an average
velocity over the 50 tidal cycles of approximately 5 cm/sec, directed up
the estuary. The middle and deep layers in the base configuration show
velocity excursions up to 20 cm/sec, and over the 50 tidal cycles the
average velocity is approximately 5 cm/sec. In the plan configuration,
the near surface layer exhibits transient excursions up to 15 cm/sec and
the average velocity in that layer over the 50 tidal cycles simulation
is approximately 6 cm/sec directed upestuary. In the middle and deep
layers, velocity excursions of up to 25 cm/sec occur, and the average
velocity over the duration of the experiment is approximately 7 cm/sec
in both bases, directed upestuary in the bottom layer and downestuary in
the middle layer.
The applied wind stress and the mouth sea level under high flow
conditions as a function of time for 30 tidal cycles are shown at the
top two traces in figure 15. During this part of the hydrological cycle
wind stress is relatively high compared to its yearly average, and
strong wind events generally come from the north. In spite of the fact
that wind stress tends to be high, the stratification found in the
Harbor is also at a yearly maximum, and is sufficiently strong that the
three-layer component of the circulation tends to dominate over the
transient components. This dominance is illustrated in the last two
traces which show the horizontal velocity at three depths near the
Harbor mouth as a function of time in both the base and plan
configurations.
In the base configuration, velocity excursions of up to 15 cm/sec occur
in the surface layer directed both up and down channel. The 50 tidal
cycle average velocity is approximately -5 cm/sec, consistent with that
associated with the three-layer circulation. The middle layer also
shows excursions up to 25 cm/sec, but the velocity is directed
down-channel during nearly all of the 50 tidal cycle experiment. Its
average value is approximately 8 cm/sec. In the bottom layer, velocity
excursions of up to -15 cm/sec occur, and the average velocity is
approximately -9 cm/sec.
The pattern of response in the plan configuration is very similar to the
pattern of response in the base configuration, the difference being in
the magnitude. In the plan configuration, excursions in the surface
layer reach -18 cm/sec and the average velocity is approximately -7
cm/sec. In the middle layer, velocity excursions of up to 25 cm/sec
occur and the average velocity is approximately 10 cm/sec. In the deep
layer, velocity excursions of up to -22 cm/sec occur and the average
velocity over the duration of the experiment there is approximately -15
cm/sec.
5. CONCLUSIONS
318
-------
"e 0.2
-20-1
20
30 40
Tidal C ye !•<
50
Figure 14: Harbor response to meteorological forcing at low flow
conditions. Shown are the following time histories (from top to
bottom): (i) longitudinal wind stress component; (ii) meteorological
tide at Harbor mouth; (iii) volume flux at mouth; (iv) horizontal
velocity at three depths for base configuration (solid-near surface,
dash—middle, dots-near bottom); (v) horizontal velocity at three depths,
plan configuration. The volume flux was essentially the same for both
plan and base configuration.
-------
-20J
20
30 40
Tidal C ye I* $
50
Figure 15: Same as figure 14, for high flow conditions
320
-------
The results of numerical modeling presented in Sections 3 and 4 lead us
to the following conclusions concerning changes in circulation within
the Baltimore Harbor system resulting from channel enlargement,
A change in channel depth from 12.8 m to 15.3 m will increase the
magnitude of circulation in Baltimore Harbor but will not change its
pattern. This conclusion applies if the channel is maintained at a
uniform depth in the longitudinal direction, without sills or barriers.
The three-layered density driven flow, which for timescales longer than
5 to 10 days is a major circulation component within the Baltimore
Harbor system, will be enhanced by channel enlargement. The enhancement
is due to combined effects of increased water column depth plus the
accompanied increase in bottom water salinity which is likely to occur
as the channel taps deeper waters in adjacent portions of the Chesapeake
Bay. This will result in decreased residence (flushing) times within
the main stem of the Harbor. The magnitude of the changes for
three-layered flow resulting from increased channel depth are best
summarized by two parameters: the volume flux (transport) in the
out-flowing, mixed layer Q and the harbor residence time T. During the
low run-off portion of the hydrological cycle, projected increases in
vertical stratification at the Harbor mouth from 1.0 to 4.85 ppt as a
result of increased channel depth will cause an increase in volume flux
in the mixed layer from approximately 88 m3/sec to 210 m3/sec and a
corresponding decrease in residence time from approximately 48 to 20
days. During the high run-off portion of the hydrological cycle, a
projected increase in vertical stratification at the harbor mouth from
9.5 to approximately 15.75 ppt as a result of increased channel depth,
will cause an increase in volume flux in the mixed layer from
approximately 300 m3/sec to 500 m3/sec, and a corresponding decrease in
the residence time from approximately 15 days to 8 days.
Calculations including with time variable meteorological forcing
indicate that the channel enlargement will increase the magnitude of
this component of the circulation by a small amount--at the 10% level,
approximately. Dredging will influence the magnitude of this component
in the circulation only; there is no indication of any change in the
qualitative character of the circulation as a result of the channel
enlargement.
References
Boicourt, W.C.; Olson, P. A hydrodynamic study of the Baltimore Harbor
System; Pt. 1: Observations on the circulation and mixing in
Baltimore Harbor. Chesapeake Bay Inst. Tech. Rep.; 1982.
Cronin, W.B.; Pritchard, D.W. Additional statistics on the dimensions
of the Chesapeake Bay and its tributaries: cross-section widths and
segment volumes per meter depth. Chesapeake Bay Inst., Special
Report 42, The Johns Hopkins University; 1975.
Elliott, A.J.; Wang, D-P; Pritchard, D.W. The circulation near the head
of Chesapeake Bay. J. Mar. Res. 36: 643-655; 1978.
Garland, C.F.; A study of water quality in Baltimore harbor. Maryland
Board of Natural Resources, Rep. 96: 132pp; 1952.
Granat, M.A.; Gulbransen, L.F. Baltimore Harbor and Channels Deepening
Study; Chesapeake Bay Hydraulic Model Investigation. U.S. Army
Engineer Waters Experiment Station, Technical Report HL-81; 1981.
Grano, Vincent. On the non-tidal circulation of the Upper Chesapeake
Bay. Ph.D. Thesis, The Johns Hopkins University; 1982.
Hachey, H.B. Movements resulting from mixing of stratified water.
Journal Res. Bd. Canada 1: p. 133; 1934.
Hansen, D.V.; Rattray, M. Jr. Estuarine circulation induced by
diffusion. J. Mar. Res. 30: 281; 1972.
321
-------
Hansen, D.V.; Festa, J.F. Inlet circulation reduced by mixing of
stratified water masses. Rapp. Pv. Cons. int. Explor. Her.:
163-170; 1974.
Long, R,R. Three-layer circulation in estuaries and harbors. J. Phys.
Ocean. I' 415-421; 1977.
Olson, Peter; Boicourt, W.C.; Najarian, T.O. A Hydrodynamic Study of
the Baltimore Harbor Circulation. Chesapeake Bay Inst. Tech. Rep.;
1982.
Phillips, O.M.; Shyu, J-H; Salmun, H. An experiment on boundary mixing:
mean circulation and transport rates. J. Fluid Mech. 173: 473-499;
1987.
Pritchard, D.W.; Carpenter, J.H. Measurements of turbulent diffusion in
estuarine and inshore waters. Bull. Inter. Assoc. Sc. Hydrol. 20:
37-50; 1960.
Stroup, E.D.; Pritchard, D.W.; Carpenter, J.H. Ches. Bay Inst. Tech.
Rep. 27, Ref. 61-5: 77 pp.
Wang, D-P. Subtidal sea level variations in the Chesapeake Bay and
relations to atmospheric forcing. J. Phys. Ocean. 9: 413-421;
1979a.
Wang, D-P. Wind-driven circulation in the Chesapeake Bay, Winter 1975.
J. Phys. Ocean. 9: 564-572; 1979b.
Wang, D-P; Elliott, A.J. Non-tidal variability in the Chesapeake Bay
and Potomac River: Evidence for non-local forcing. J. Phys. Ocean.
8: 225-232; 1978.
Wang, D-P.; Kravitz, D.W. A semi-implicit two dimensional model of
estuarine circulation. J. Phys. Ocean. 10: 441-454; 1980.
Wilson, R.E. A study of the dispersion and flushing of water-borne
materials in the Northwest Branch of Baltimore Harbor, and the
relationship between these physical processes and the water quality
of the inner harbor. Ches. Bay Inst. Tech. Rep. 64, Ref. 70-3:
lOlp.; 1970.
522
-------
Understanding the Estuary: Advances in Chesapeake
Bay Research. Proceedings of a Conference. 29-31
March 1988. Baltimore, Maryland. Chesapeake Research
Consortium Publication 129. CBP/TRS 24188.
A Numerical Investigation of Circulation and Salt Distribution
in the Patuxent River Estuary
Chris Kincaid and Peter Olson
The Johns Hopkins University
Department of Earth and Planetary Sciences
Baltimore, Maryland 21218
Harry Wang
Maryland Department of the Environment
201 W. Preston Street
Baltimore,.Maryland 21201
1. INTRODUCTION
In this study we apply a two dimensional numerical model of laterally
averaged circulation and salinity to the Patuxent River estuary, shown
in figure 1. Both steady state and time variable simulations are made
to determine the interaction between salinity gradients, astronomical
and meteorological tides, river discharge and surface wind stress and
their effects on the circulation.
In comparison with other Chesapeake Bay tributaries, the Patuxent River
estuary exhibits large variations in channel depth. Figure 2 is a
longitudinal profile of the estuary from its intersection with the Bay
upstream 80 km to Nottingham, MD 65 km from cross section 0. The
bathymetry consists of two basins, each bounded on its downstream side
by a shallow sill. The sill separating the lower basin from the main
stem of Chesapeake Bay is 10 m deep; the average lower basin depth is
approximately 18 m and includes a 43 m deep hole. The inner sill
separating the upper and lower basins is approximately 3 m deep and the
upper basin behind it averages 7 m depth. Circulation and salinity in
the Patuxent are strongly influenced by two of the river's bathymetric
features: (1) the inner sill, located between Benedict and Lower
Marlboro, MD and (2) the 40 m depression located at Pt. Patience. The
most striking piece of evidence for topographically controlled
hydrodynamics is the presence and persistence of a nearly vertical salt
front located about the inner sill. Throughout the lower part of the
estuary from the surface to approximately 10 m depth, the isohalines
slope upstream in the manner characteristic of two-layer circulation in
a partially mixed estuary. But beginning near cross section 20 a
pattern of vertical isohalines is evident in the salinity cross sections
from all four seasons. The change in salinity across the front varies
from about 9 ppt in the fall to about 4 ppt in the spring (figure 3).
The width of the front and the fine structure within the frontal region
are not well resolved by the Maryland Office of Environmental Programs
seasonal surveys (which report salinity profiles at 5 km intervals).
323
-------
CHALK POINT ;
BENEDICT ----
PATUXENT
RIVER BASIN
>
<
\ LOWER MARLBORO
UJ
UJ
Q.
UJ
SHERIDAN X
POINT 0
^•"'-SOLOMONS
SCALE IN MILES
Figure 1: The Patuxent River basin. The estuarine portion begins about
5 km above Mataponi. The tidal limit is a few kilometers below Bowie.
524
-------
DEPTH-FEET
o o o o o o
"0 -H
— <\J C\J rO rO
Sa313IAI-Hid3Q
-------
-15
i
-5
i
15
25
RIVER KILOMETER
35 45 55
65
75
85
95
105
i
-15 -5
I t_
5
15
25
35
45
55
65
75
85
95
105
45 J
Figure 3: Patuxent River estuary salinity cross sections (a)
seasonally averaged for spring 1984, from OEP Tech Rep. #7, (b) seasonal
average for fall 1984, from OEP Tech Rep. #7.
326
-------
Topographically controlled frontal structures are known to be important
in estuaries (Huzzey (1982); Hibiya (1986)). Similarly, permanent small
scale fronts are often present on continental shelves, especially near
the mouth of large estuaries (Garvine, 1974, 1979). It is certain that
the interaction between the upper and lower basins in the Patuxent is
strongly affected by the dynamics in the frontal region.
The dynamic conditions within the deep hole at Pt. Patience are also not
well understood. On the basis of available data (the maximum depth of
OEP surveys is 23 m) , one would infer that the circulation within the
deep hole has a seasonal variability. In spring and winter and probably
during most of the fall, the deep hole is only weakly stratified to 23 m
depth. This indicates that there is a substantial circulation to at
least 23 m for 8 months of the year or more. In the summer the OEP 1984
(figure 15a) survey (in this case, data from a single cruise, rather
than a seasonal average) showed a well-developed halocline in the
interval between 5 and 10 m. The halocline intersects the outer sill
and the upstream wall of the lower basin, in effect putting a lid on the
deep hole. In summer, the deep hole may at times become unventilated
and practically stagnant.
A variety of modelling approaches have been used on coastal plain
estuaries, from simple one dimensional, semi-analytical models of salt
wedge dynamics (Prandle, 1981; Oey, 1984) to laterally-averaged two
dimensional models such as used in this study (Festa and Hansen, 1976,
1978; Blumberg, 1978; Wang and Kravitz, 1980) to fully three dimensional
models (Oey, Melor and Hires, 1986; Gordon and Spaulding, 1987). At the
present time, 3-D models are used in situations where the density-driven
component of the circulation is subordinate to wind or tidally induced
flow. In those circumstances, the horizontal pattern of depth-averaged
flow can be solved without reference to the density distribution, and
the vertical structure of the flow can then be determined locally at
each grid point. However when density gradients are as important in
driving the circulation as tide and winds, as is the case in the
Patuxent River estuary, depth averaging does not work. In this case it
is more appropriate to lateral average, as we have done. Having
computed lateral averages, it is possible to then compute the flow in
lateral sections, using the equations of motion for that plane. We have
not pursued such an approach for modelling the Patuxent, because we
judge lateral flows to be of secondary importance in this case.
Most of the earlier two-dimensional models were designed for steady
state conditions, or for simulating slowly evolving flows associated
with variations in boundary salinity or river discharge. However, it
has been demonstrated throughout the Chesapeake Bay system that the
circulation is strongly time variable, not only from astronomical tides,
but also in the subtidal frequency band, 0.1-1.0 cpd (Wang, 1979;
Elliott, 1978; Wong and Garvine, 1984; Olson, 1986). The subtidal
variability is due to the combined effects of meteorological tides
(non-local forcing) and locally applied surface wind stress (local
forcing). These components often dominate the circulation on timescales
shorter than about 5 days. Only when the circulation is averaged over
time intervals of 10 days or more does the density driven flow prevail.
The measurements reported by Boicourt and Sanford, 1987 confirm this to
be true for the Patuxent River estuary as well. It is necessary to
resolve frequency components of the circulation ranging from the
semi-diurnal tide at the high end, to seasonal variability at the low
end, in order to construct an accurate picture of dynamics in the
Patuxent.
2. NUMERICAL MODEL
The model geometry has its origin at mean sea level at the head of the
estuary, oriented with x,y,z-axes in the downstream, cross stream and
vertical directions respectively. The maximum channel depth in any
lateral section is denoted by H, and the sea surface elevation, relative
to its mean value, is denoted by n. The laterally-averaged velocity
components in downstream and vertical directions are u and w,
respectively. Channel geometry is specified by the width function
327
-------
B(x,z). The laterally-averaged shallow water equations of motion are as
follows. The momentum balance in the vertical direction is hydrostatic,
and the other governing equations are
o
-
at
dx
d
die
dz
_d_
+ dl
ox
dx
dz
dz
—
dx
~
dx
—
dz
f —
zdz
(3)
^ }
(4)
representing conservation of mass, horizontal momentum and salinity,
plus a linear equation of state. Here t is time, S is salinity, p is
density and g is gravity. The parameters N and Nz are horizontal and
vertical turbulent viscosities, and ^ and Kz are turbulent mixing
coefficients for salt in the horizontal and vertical directions
respectively. The boundary drag coefficient is Cp, and in the equation
of state P, denotes fresh water density and e is the coefficient of
saline contraction, 7.29 x 10~A, ppt"1.
The appropriate boundary conditions are the following. On the channel
bottom, z--H, the normal velocity and salinity flux both vanish. In
terms of the local channel bottom slope, these are
dH
w -- u = 0,
dx
= -H
(5)
and
KxdHdS
- -- T^T-— = 0,
dz Kzdxdx
accurate to the order of the bottom slope.
is , to the same order
Z--H, (6)
The bottom stress condition
-H.
(7)
Since the slope of the sea surface is small compared to the bottom
slope, it is neglected in applying the upper surface conditions. The
surface equivalents of (5) and (6) are then
3 9
—
oz
0,
and the surface stress condition is
(8)
(9)
where r is the laterally averaged longitudinal component of the wind
stress .
Entrance and exit conditions are specified as follows. At the head of
the estuary, the Patuxent River freshwater discharge QR is specified as
a time series and distributed uniformly through the water column as a
horizontal current UR, satisfying the condition
328
-------
j BuRdz-QR(f), x = 0 (10)
J — H
and the salinity is set to zero
5 = 0, x = 0 (11)
At the estuary mouth, x-L, where the Patuxent joins the main stem of
Chesapeake Bay, the exit velocity condition is just
a
— (ufi)-O x-L (12)
O vX
while the sea surface conforms to the sea surface of the Bay. This
condition is expressed in terms of a time series r?m (t) for the Bay
surface elevation
T] = rim x = L (13)
Conditions on salinity at the mouth are constructed to simulate the
exchange processes between the Patuxent and the main stem of the Bay.
If the transport is out of the estuary, we prescribe a purely advective
balance
— (Sfi) = - — (uSB), x-L,if u>0 (14)
at a X
When the transport direction changes , we expect there to be a time
interval during which water from the preceding outflow is carried back
into the Patuxent. As inflow continues, this pool is exhausted and
Chesapeake Bay water begins to enter. This sequence of events is
represented by a relaxation condition of the form
. 5 = SC at t = tc; x-L, (15)
where tc is the time of current direction change, t* is the relaxation
time constant, an adjustable parameter, S. is mouth salinity at tc, and
Sm is the Bay salinity. The data required to initialize the calculation
are an initial salinity distribution and sea surface elevation pattern
(relative to mean water level) .
The model contains five friction and mixing parameters, the boundary
drag coefficient CD, plus mixing coefficients for turbulent momentum, Nz
and Nx, and salinity S, and S_. Estimates of drag coefficient values
lie in the range 1-2 x" 10"3 (Phillips, 1980); and we choose CD - 1.6 x
1CT3.
The largest source of error in estuarine modeling comes from trying to
parameterize the turbulent mixing of salt and momentum. The mixing
parameters are known to have a functional dependence on the Richardson
number
The exact functional relation, however, between the mixing parameters
and the Richardson number is not known. An alternative method, and that
used in this study, is to assign values to these parameters based on
direct comparison with observed field measurements.
In partially mixed estuaries such as the Patuxent, the most important of
the mixing parameters are Kz and Nz, controlling the vertical flux of
salt and momentum, respectively. Often in estuaries the horizontal
mixing coefficient for salt K^ is unimportant; however in the Patuxent
this is not the case, as it controls the structure of the salinity
front. By contrast, the horizontal mixing coefficient Nx is dynamically
329
-------
insignificant; usually it is assigned a nominal value large enough to
allow the horizontal dispersion term to contribute a small amount to the
momentum balance, for purposes of numerical stability.
The vertical eddy viscosity, Nz-5.0 cm2 s"1, was selected on the basis of
comparisons between observed and calculated velocity profiles and
results of a similar study of the Baltimore Harbor. Preliminary
solutions with horizontal eddy viscosity, Nx < 1.5 x 106 era2 s"1 were
unstable to numerical oscillations. The value of Nx - 2xl06 cm2 s"1 was
chosen to prevent numerical instabilities from developing.
The mixing parameters for salt, FL and K2, were chosen on the basis of a
sensitivity analysis. The salt front is a first order feature of the
Patuxent River, appearing with only minor deviations in position and
gradient in all four of the 1984 OEP seasonally averaged surveys. The
position and gradient of the front is one piece of data we require our
model to successfully simulate and is the basis of the sensitivity
analysis. With an unsmoothed geometry file any value ^ < 1.25xl06 cm2
s"1 proved to be unstable to numerical oscillation. The calculated
front, after seven days, with K^ - 1.25xl06 cm2 s"1, is more diffuse,
As/AX - .27 ppt/km than the observed front, AS/AX = .4 ppt/km. Clearly
such large values of K^ do not accurately model the front. In an
attempt to stabilize the solution, the geometry file was run through a 3
point smoothing routine. With a smoothed version the lower limit on tL
reduced to Y^. > 7.5x10*. A constant value of K^ - 1.25xl05 was adopted
for the numerical experiments. The value chosen for vertical mixing Kz
- 0.2 cm2 s"1 produces solutions which match observed surface salinity
and allow for vertical structure in isohaline patterns.
Equations (1) through (15) are solved on two-dimensional grids using
finite differences. The time-stepping procedure is the semi-implicit
method used by Wang and Kravitz (1980), in which the sea surface height
is advanced with an implicit time step, while the interior variables
(salinity, horizontal and vertical velocity) are advanced using an
explicit time step. The details of the computational algorithm can be
found in Wang and Kravitz, 1980. Diffusion terms are represented by
central differences, advectipn terms by upwind differences. We found
that it is necessary to use very small time steps, on the order of 100
s, to insure numerical stability.
In order to include in a realistic way the extreme variations in channel
depth, we have used the sloping bottom approximation. Let AZ and AX
denote the grid spacing in the vertical and horizontal directions,
respectively. According to this approximation, variations in channel
depth are represented by increasing or decreasing the number of grid
points by one in neighboring columns. The local channel slope is then
either *AZ/AX or zero. We have constructed a grid with resolutions
(grid spacings) of 2 m in the vertical and 640 m in the horizontal, so
that the inclined segments have slopes of *0.003125. This permits us to
simulate all the major topographic structures in the Patuxent, including
the deep hole at Pt. Patience, where the grid extends to 30 m depth.
The slope of the model topography closely approximates the average
gradients found in the estuary. In order to correctly apply horizontal
differential operators at these points, we have inserted image or
"buffer" points onto the grid rows within sloping sections.
The bathymetric grid used in this study consists of 164 columns of w and
S points and 103 columns of u points, extending over 104 km (AX - 638
m). Vertical resolution is 2 m. In the model the estuary head is at
river kilometer 100, at the US Rt. 50 bridge crossing, near the tidal
limit. The mouth is defined at river kilometer -4, about 4 km
downstream from Drum Pt., approximately 8 km downstream from Solomons
Is. Major topographic structures incorporated into the model grid
include: (1) the outer sill at the model-defined mouth; (2) the deep
hole at Pt. Patience (to 30 m); (3) the inner sill above Benedict; (4)
the upper basin around Lower Marlboro; and (5) the narrow and shallow
upper reach above Jug Bay.
330
-------
6
12
18
24
30
-44 12 2O 28 36 44 52 6O 68 76 84 92 1OO
Figure 4a: The initial salinity profile for spring 84 steady state
simulation. Distribution is uniformly stratified. The contour interval
is 1.0 ppt. Right is estuary head, left is estuary mouth. Positive is
left and down.
6
12
18
24
30
_-1OC
1O 5 O -3 -1O
VELOCITY.CM/S
LOWER MARLBORO
10
10 5 o -a -10
VELOCITY.CM/S
-44 12 2O 28 36 44 52 6O 68 76 84 92 1OO
DISTANCE (KM)
Figure 4b: Contours of salinity and velocity profiles 520 hours into
the run. River flux and mouth salinity held constant. No tidal or wind
forcing. Coefficients K^ - 1.25xl05, Kz - . 2 Nx - 2xl06, Nz - 5.0.
331
-------
SURFACE SALINITY, PPT
Figure 5: Comparison of observed spring 1984 seasonally averaged
surface salinity, plotted diamonds with ± 1 ppt error bars, vs.
calculated solutions for spring steady state run. Shown is the
progression from initial, linear profile, solid line, to the solution at
400 hours, long dashed line, to the steady state profile at 520 hours,
short dashed line.
332
-------
3. STEADY STATE CALCULATION
A steady state simulation of circulation and salt distribution within
the Patuxent River estuary has been run for the purpose of modeling the
gross features of the spring 1984 seasonally averaged salinity profile
(figure 3a). The most prominent feature of the Patuxent River salinity
distribution is the salt front located near the inner sill, between
Benedict and Lower Marlboro. The goal of this simulation is to produce
the frontal structure from random initial conditions.
To achieve a steady state solution a constant river flux 5-20 m3/s was
imposed at the head of the estuary. A salinity profile, increasing
linearly from 7 ppt at 1 meter depth to 9 ppt at 9 meters depth was
imposed at the mouth of the estuary. The values were taken from spring
1984 seasonally averaged data. Figure 4a shows the uniformly stratified
initial salinity distribution. The salinity gradient across the front
is .06 ppt/km. Figure 5 illustrates how the model accurately predicts
both the position and gradient of the salt from random initial
conditions. The solid line shows the linear surface salinity profile at
t-0. The 1 ppt isohaline is initially at kilometer 86. After 400 hours
the 1 ppt isohaline has moved to 50 km and the frontal gradient is .27
ppt/km. After 520 hours the solution has reached a steady state. The 1
ppt isohaline is at 48 km and the frontal gradient is .29 ppt/km which
compares well with the observed value of .3 ppt/km. Calculated surface
salinities fall within the ± 1 ppt error bars of all the observed data
points. Figure 4b shows the steady state salt distribution and velocity
profiles within the upper and lower basins. The velocity profiles at
Lower Marlboro and Pt. Patience have remained invariant over the past
100 hours. The steady state profile at Lower Marlboro is 3 cm/s surface
outflow balanced by a -1.7 cm/s inflow. At Pt. Patience values for
surface outflow and bottom inflow are 3.0 cm/s and -2.0 cm/s
respectively.
The calculated steady state salinity distribution for the spring of 1984
compares well with the observed seasonally averaged profile shown in
figure 3a. The salt front in both profiles is centered over the inner
sill. The observed salinity gradient across the front is AS/ax- .3
ppt/km which is very close to the calculated value of .29 ppt/km. The
steady state simulation accurately models the gross salinity features of
the estuary given random initial conditions. The predicted circulation
is a two layer shear flow with an average 3.0 cm/s surface velocity and
a -2 cm/s bottom velocity.
4. TIME VARIABLE CALCULATIONS
Two intensive survey periods were selected to study time variable
circulation under two distinct forcing regimes. The spring survey
covers the time interval May 15 - June 15, 1986 when river flow is high
and the estuary is less saline. The second, or fall, survey runs
between September 3 - September 30, 1986 and is characterized by lower
river flow and higher salinities in the adjacent Chesapeake Bay. During
each survey period tidal heights were measured from the N.O.S. gauge
located at Solomons Is. River flow data was collected from the U.S.G.S.
stream gauge located at the Rt. 50 bridge. Time variable mouth salinity
was taken from data collected by Boicourt and Sanford, 1987 at station
PL 1, located just upriver from Drum Point (see figure 2). Hydrographic
surveys and moored current meters provide measurements of salinity
distribution and velocity profiles respectively.
The spring period can be divided into 3 regimes based on mouth salinity
and river flow. The mouth stratification expressed as 6s-s4-sj, where SB
is mouth salinity in ppt at 9 meters depth and ss is mouth salinity in
ppt at the surface, determines the strength of density driven
circulation. The first 10 days are characterized by strong stratifica-
tion, 6s- 1 ppt, and weak river flow, the second 10 day period by both
weak stratification, 6s- 0, and river flow and the last 10 days by
strong stratification, 6s- 2, and river flow. The initial salinity
distribution was derived from the May 9 slack water survey.
333
-------
The simulation begins at midnight on May 15, 1986 and ends at 744 hours
on June 16, 1986. During this period direct comparisons between
observed and calculated salinity fields can be made on May 21, June 3
and June 9. Figure 6a is a plot of instantaneous surface salinity at
various times within the May 21 tidal cycle. The tidal excursion in
figure 6a ranges from a minimum of 5 km near the mouth to a maximum 10
km near the inner sill. When comparing calculated, tidally averaged
solutions with observed data, it is important to consider this tidal
excursion because a data collection survey may take up to 6 hours to
complete or 1/2 the period of an astronomical tide. During this time
isohalines are being advected with the tidal flow. If the survey on May
21 began at the mouth, when the tidal height was .23 meters, and ended
six hours later when the tidal height was -.15 meters, the calculated
solution would exactly match observed data. A second comparison on June
3, 460 hours into the simulation, figure 6b, shows that the tidally
averaged solution, though slightly more diffuse than the observed field,
lies within the error bars at all but one point. A final comparison is
made, figure 6c, on June 9, 20 days into the simulation. Again the
solution matches observed values to an acceptable level. The average
frontal salinity gradients based on calculated and observed data taken
from kilometers 35 and 45 at 5 meters depth are .4 ppt/km and .35
ppt/km. The comparisons indicate the numerical model is accurately
predicting the salt balance within the estuary during the month long
simulation.
Table 1 lists the range in tidal velocities and the ratio between
observed and calculated ranges. It is found that, on average, observed
values exceed calculated values by a factor of 1.6. This relationship
is expected, however, because current meters, located in the middle of a
river measure the maximum of a velocity profile. Calculated velocities
are all scaled by estuary width and, therefore, represent lateral
averages rather than lateral maximums. Assuming a parabolic distribu-
tion, the maximum is 1.5 times the lateral average, which accounts for
the difference between calculated and measured values.
Figure 7 is a month long time series of velocity profiles for cross
sections at PL1 and PL4 (located just down river from Broomes Is., see
figure 2), showing the persistence of a two layer shear flow. The
average profile at PL1 is a 5 cm/s surface outflow balanced by a deeper
5 cm/s inflow. Figure 7 also illustrates the dominant driving
mechanisms in the estuary by comparing events in the mouth salinity and
river flow records against model responses in the PL1 and PL4 velocity
profiles. During most of the survey period river flux is small, 5- 5.0
m3/s. On June 12, 684 hours into the simulation, river flux increases
by more than 100% of the average to 11.2 m3/s. A pulse in the velocity
profile at PL1 is seen after a 12 hour time lag. The response to the
flood event is seen throughout the entire PLl record. Surface
velocities increase while deeper return flow stalls.
Comparing time series of mouth salinity and velocity shows that the
largest fluctuations in velocity ±4 cm/s, or an 80% fluctuation in mean
lower layer velocity, correlate with changes in mouth stratification.
Table 2 shows that increases in £s cause positive deviations in average
PLl velocity, D,3> at 13 meters depth. There is a 36 hour time lag
between corresponding peaks in 6s and DI3 at station PLl.
The gross circulation patterns within the estuary are best shown using
streamfunction contours. Streamfunction distributions, within two of
the three 6s regimes outlined in Table 2, were calculated by vertically
integrating the horizontal velocity field from the expression
/.
UBdz' (17)
-H
Figure 8a shows tidally averaged Streamfunction on May 20 when 5s - 1.0
ppt and ij - 5.0 m3/s . The most vigorous circulation appears in the
lower estuary and is particularly strong near Pt. Patience. The deep
hole is being flushed by a deep return flow and is clearly not stagnant.
334
-------
I-
Q.
(L
f%
LJ
u
<
Li-
ft:
D
MAY 21, 1986
Figure 6a: Plots of instantaneous surface salinity for half a tida
cycle on May 21, 1986 vs. observed data, plotted diamonds with ± 1 pp
error bars. The solutions for this range of tidal heights, .23, -.0
and -.15 meters, plotted as solid, short dashed and long dashed line
respectively, outline the tidal excursion within the estuary
Coefficients same as those in figure 3.
JUNE 3, 1986
Figure 6b: Tidally averaged surface salinity, solid line, vs. observe
data plotted diamonds with * 1 ppt error bars 470 hours into spring rut
June 3, 1986.
-4
JUNE 9, 1986
22
48
74
100
DISTANCE, KM
Figure 6c: Tidally averaged surface salinity, solid line, vs. observe'
data, plotted diamonds with * 1 ppt error bars, on June 9, 613 hour,
into the spring run.
-------
Table 1 Tidal Velocity Range
Observed Calculated Ratio
obs./calc.
1.5
1.7
0.8
1.1
2.1
1.0
1.0
1.6
2.2
1.4
1.5
1.5
2.1
1.3
Station Depth (ft) range (cm/s)
PL1
P2
P3
PL4
P5
P6
P8
P9
8
45
8
25
8
80
8
45
8
28
8
22
8
15
Average correlation factor -
Time
Table 2
period (hours)
0-240
240-480
480-720
43
46
'26
36
67
65
30
28
27
32
60
45
92
55
1.58
range (cm/s)
29
26
30
32
31
30
30
28
24
23
40
21
50
43
Mouth Salinity Reeimes
57 (ppt)
0.96
0.1
2.0
DI3 (cm/s)
5.5
4.0
8.0
336
-------
12
6
0
SPRING RUN - 5/15-6/15
384
768
18 _ 1-9 METERS
19
12
384
768
12
0
-12
12°
1 METER
S METERS
768
-12
O
12 _
9 METERS
768
-12
13 METERS
768
-12
I Z MM
PL1 PL4
17 METERS
768
-12
48 96 144 192 24O 288 336 384 432 48O 528 576 624 672 72O 768
TIME INTERVAL,! HOUR
Figure 7: Time series of tidally averaged velocity profile at
1,5,9,13,17 meter depths at stations PL1 and PL4 vs. time series of
river flux. (m3/s) and mouth salinity, ppt, for 1,3,5,7, and 9 meters,
from spring intensive survey.
337
-------
Figure 8a: Contour of tidally averaged streamfunction/1000, cm2/3• on
May 20 during the spring intensive run. Contour interval is 8. Average
river flux 3-5 m3/s and average mouth stratification 53 - 1 ppt.
I
Q.
LJ
Q
30
-4 4 12 20 28 36 44 52 60 68 76 84 92 100
DISTANCE (KM)
Figure 8b: Contours of tidally averaged streamfunction/1000, cm2/s, on
June 13 during spring intensive run. Average river flux 5- 11 m3/s and
average mouth stratification 53- 2 ppt.
338
-------
In examining circulation patterns it is interesting to follow the zero
streamline near the bottom boundary in the lower estuary. The
streamline rises vertically off the bottom near river kilometer 40 and
returns, along the surface, to the estuary mouth, indicating the strong
density driven circulation is, on average, confined to the lower
estuary. Smaller, less vigorous two layer circulation patterns are seen
just above the inner sill near Lower Marlboro and further upstream near
kilometer 60 where the transition between one and two layer flow occurs.
Figure 8b shows the streamfunction distribution on June 13 during an 11
m3/s flood event. The density driven flow in the lower estuary is now
very weak compared to surface outflow. The circulation is split into
two weak cells, one along the upslope between Pt. Patience and Sheridan
Pt. and the other in the deep hole. The zero streamline now rises off
the bottom at kilometer 10 and forms a closed cell. The density driven
circulation in the deep hole does not feel the strong mouth
stratification, 5s- 2 ppt, because the river flow is advecting the salt
out of the estuary. The strong surface outflow has sealed off the deep
hole. Figure 8b and figure 7, at 684 hours, both show that circulation
within the deep hole vanishes during the flood event.
The second time variable simulation, the fall run, covers the time
period Sept. 3 - Sept. 30, 1986 when, in comparison to spring
conditions, river flow is small and Bay salinity is high. The
simulation, as in the spring run, is driven by time series of tidal
height, mouth salinity and river flow. In addition to these data, a
time series of local wind speed was recorded at the Patuxent Naval Air
Station near Solomons Is. Hydrographic surveys on Sept. 15, Sept. 22
and Sept. 30, 1986 provide a basis for comparison between measured and
calculated salinity distributions. A comparison between the observed
velocity profile at Station PL1 (from Boicourt and Sanders, 1987) and
the numerical solution at PLl is made.
The mixing parameters for the fall simulation were the same constant
values used in the spring simulation. The initial salinity profile was
taken from a digitized version of a Sept. 2, 1986 slack water salinity
profile. The wind stress was computed using data from P.N.A.S. and a
quadratic law relating wind stress to wind velocity via a friction
coefficient, taken here to be 1.6 x 10~3. Wind speeds measured on land,
because of increased frictional resistance, are often less than those
acting on the estuary. In light of this observation, two fall runs were
made: 1) fall run 4, which uses observed wind speeds and 2) fall run 5,
which uses twice the observed wind speed. Results of both simulations
will be presented below.
Comparison of observed surface salinity from the first hydrographic
survey on Sept. 15 (figure 9a) shows the solution from run 4 to be
within the error bars of all but one station, whereas the surface
salinity for run 5 fits observed data in the lower estuary but not in
the upper estuary. Figures 9b and 9c show comparisons made between
calculated and observed surface salinity on Sept. 22 and Sept. 30.
Again, the solution from run 4 fits the data throughout the estuary.
The calculated gradient, .41 ppt/km, in salinity across the front is
close to the observed, .45 ppt/km ± .1 ppt.
Run 5 does not compare well with the observed salinity distribution.
The frontal structure in all three comparisons is displaced upriver and
the average frontal gradient is .3 ppt/km. Based on comparisons of
measured and calculated fields, run 4, with observed wind speeds, gives
a more accurate solution for salt distribution within the estuary.
339
-------
18_
12 _
SEPT. 15, 1986
SEPT. 30, 1986
0
-4
22 48
DISTANCE, KM
74
100
Figure 9: Plot of calculated, tidally averaged surface salinity, ppt,
vs. observed data, pl-otted diamonds with * 1 ppt error bars on (a) Sept.
15, 1986, 312 hours into the run, (b) Sept. 22, 1986, 470 hours into the
run and (c) Sept. 30, 1986, 650 hours into the run. Fall runs have time
variable mouth salinity, river flux and tidal heights. Fall run 4,
solid line, uses observed time variable wind speed. Fall run 5, dashed
line, uses 2x observed wind speed.
340
-------
Figure 10 shows a comparison of measured and computed average velocity
profiles at station PL1. Observed data was taken from Boicourt and
Sanders, 1987 and the calculated profile from run 4. The velocity
profiles are similar in both flow structure and magnitude. Both
profiles are ± 5 cm/s 2-layer shear flows. Values for maximum, observed
inflow and outflow velocities are approximately 1.2 times calculated
values. The level of no motion for observed and calculated profiles is
at 7 and 6 meters respectively.
Figures lla and lib are time series of wind speed versus PL1 and PL4
velocity profiles for both fall runs. A comparison of the profiles from
the two runs illustrates the effects of increased wind stress on
circulation and explains the intrusion of salt into the upper basin in
fall run 5. Both runs show the same qualitative response to surface
wind stress, but because wind speed is related to wind stress by a
quadratic law, the quantitative differences between runs 4 and 5 are
significant. During a two day period, September 10-11, an average -2.5m
s"^ wind in fall 4 causes the surface velocity to go to zero. During
the same period, in fall run 5, an average -5 cm2 s"1 wind event causes
surface velocity to reverse direction. The resulting -6.5 cm2 s"1
surface flow, a 200% fluctuation in mean velocity, advects salty lower
basin water over the sill into the upper basin, resulting in the
displaced frontal structure recorded in figure 9.
In general, fall 4, using observed wind speeds, shows a consistent two
layer density driven shear flow. Wind driven circulation produces
fluctuations in mean flow of * 100% in the surface layer and •* 30% in
the lower layer. There is one event, in the 13 and 17 meter records, 10
days into the simulation, in which the calculated fluctuation is 100% of
the mean flow. This is probably due in part to a 1.4 ppt increase in 6s
which is coincident with a 4 m/s wind event. An identical 4 m/s wind
event at 312 hours produces only a 25% perturbation in the average
inflow at 13 meters depth, suggesting that as much as 75% of the
fluctuation mentioned above, at 240 hours into the simulation, is
attributable to changes in mouth stratification.
Fall run 5, with double the observed wind speed, shows fluctuations of *
80% of the mean lower layer inflow and as high as * 200% of the mean
surface outflow. Fluctuations of this size compare well with data from
Boicourt and Sanders (1987), though the timing of the responses does not
exactly coincide. The high values for wind speed necessary to produce
such large model fluctuations, however, as shown above, generate
salinity fields which do not match observed data.
5. TOPOGRAPHICALLY CONTROLLED CIRCULATION
The influence of variable channel depth on circulation and salt
distribution within the Patuxent River, primarily near the inner sill
and within the 43 m deep hole, remains unresolved by observation alone.
The code used in this study has been modified to handle variable
bathymetry in an attempt to further the understanding of topographically
controlled circulation. The most direct piece of evidence for
topographic control is the persistence of a nearly vertical salt front
at the inner sill in OEP surveys for all four seasons.
The dynamic effect of the sill is clearly seen in current meter records.
Time series from stations P8 and P9, located to either side of the sill,
show a 90" phase difference in maximum tidal velocities. The model
successfully matches this phase difference in tidal velocities and
predicts a low frequency sill-induced circulation. Figure 12 shows one
month averaged streamfunction plots for fall and spring forcing regimes.
The two fields, which are clearly similar, show two density driven
circulation cells, one in the lower basin and one in the upper basin,
divided in the middle by the inner sill.
341
-------
DEPTH, M
K) -* -»
O tt O tH O
O
•H
§ <*
N
O
3
^
r
in o
r
O
o
J
*^
0
£ !
CO
1
___>.
1 11 1
1 !
i !
s g
ra O "H
"IS / r
13-1 ' 1
6 1 // r
/ / z
// H
// G
// z
X * C^
/ • —
/ ** ^
// m
7-'' i
/ i
^"^ -**' *n
^^"^vx 7 r
N"""
•MM
O
Figure 10: Survey averaged velocity profiles at station PI
. Fall run
4 plotted as a short dashed line. Observed data (from Boicourt and
Sanders, 1987) plotted as a dashed line. Both profiles show a 2-layer
shear flow.
342
-------
^ 12
rfj Q
§ -12
* C
2O
o
-20
20 <
0
-20
t/> r
S* 20
* 0
8 —2O
» 20C
o
20
20 C
0
—20
C
^ 12
=1
1 -12
* C
20
Q
—20
2o'
Q
—20
^C. <
• o
I -20
0
JS 20
0
—20
2O
o
—20
Figure 11
1 ~+. -ijjYW, . /r*HA~ . AIM A., Jfl . A*A* -,rA >
"vyv • y \*/*\/ - •>"• • ifv./ -v^1 v*v/"* v w v^Vv^v*"1
i i
338 871
,--, 1 METER
) 672
8 METERS
1
) 672
~ • METERS
Iff. r,- — ~_^— ^.^.^^..^.^ -»— -~» — -»^~ -X^^»- — «_•"
> 672
~* 13 METERS
** i
> 672
PL1 PL4 17 METERS
I 1 1 1 1 I I 1 1 1 1 1 1 t
) 48 96 144 192 24O 288 336 384 432 480 828 S76 624 672
TIME INTERVAL, 1 HOUR
L Vl< ifWV ""^ ."n/ V ^f ^^ ,
) 336 871
/*-.. /"*• .- /'\ -i;.1 METER
.-- .£'—"•* ^ V / V ^-. / <• x"**v A-. . ^•*>-i*. ,** <« /
v- ^v- ^ \j v- " "-N./ _
J 672
B METERS
3 672
? METERS
r ^ A , ^.^-v
••»-•* — *-^^' V^..--— ""^^ ^ ' " "*-" ^"•v-/
3 872
13 METERS
. ^.^ - A u- ^>N
0 672
- ...PL1 PL4 17 METERS
, l ***l 1 >••" 1 1 1 1 I L 1 1 1
0 48 98 144 192 24O 288 336 384 432 48O 828 876 624 672
TIME INTERVAL. 1 HOUR
: Time series of tidally averaged velocity profile at 1,5,9,!
and 17 meter depths at stations PL1 and PL4 vs. (a) a time series of
observed wind speed, fall 4 and (b) a time series of twice the observed
wind speed, fall 5.
-------
0
6
12
t 18
24
30
FALL4
iw i i i i i i i i i i L
-4 4 12 20 28 36 44 52 60 68 76 84 92 100
DISTANCE (KM)
Figure 12: Streamfunction contours calculated from (a) spring survey
averaged and (b) fall 4 survey averaged velocity fields, cmVs/1000.
Contour interval is 5.
344
-------
Figure 13 shows the large distortions in instantaneous surface elevation
which occur in the vicinity of the sill, where both width and depth
decrease. The surface slope changes sign to either side of the sill.
The effect of this sign reversal on velocity is illustrated in the
profiles at kilometer 32 and 54 located on either side of the sill. The
phase difference in tidal velocity records at P8 and P9 indicates the
phase of the tide changes near km 44, because of the constriction.
Figure 14 shows contours of salinity variance (<*') calculated for both
transient runs from the expression
(18)
where N is the total number of entries at point x,y in the time domain.
The time average salinity at a point is expressed by
• N
The largest variance for the spring survey occurs between the inner sill
and Lower Marlboro, in association with a local maximum in tidal
excursion. Of more interest is the absence of a minimum variance within
the deep hole near Pt. Patience. The values of a' - .3 and a' - . 5 for
spring and fall runs suggest that, on average, the deep hole is not
stagnant. During both steady state and transient simulations the hole
remains well flushed with average velocities of 4 cm s"1. The hole may,
however, be subject to transient short term periods of stagnation.
Salinity surveys made in the lower basin have documented that well
developed halocline structures are sometimes present over the deep hole
section, particularly during summer (low flow) conditions. An example,
from summer 1984, is shown in figure 15a. The processes leading to
halocline formation are not fully understood, nor do we know how
frequently such events occur and how long they last. These may well be
very significant events, however, because they may produce stagnant
conditions in the deep hole region and in smaller holes throughout the
lower estuary. From figure 15a it appears the halocline in the Patuxent
forms in response to increased stratification in adjacent parts of
Chesapeake Bay. This stratification could result from either a decrease
in surface salinity, an increase in bottom water salinity or both. The
pattern of isohalines near the Patuxent mouth in figure 15a indicates at
least the first alternative occurred.
In order to determine whether this scenario can account for halocline
formation, and also to determine the consequences for circulation in the
deep hole region, we conducted a numerical experiment of the model's
response to increased stratification at the mouth. The initial
conditions corresponded to the final state of the spring intensive
simulation. On that state we have imposed a pattern of decreasing
surface salinity at the mouth, producing an increased mouth
stratification.
The deep hole response to several days of increased stratification is
shown in figures 15b. The salinity pattern is very similar to the data
in figure 15a. Low salinity surface water from the Bay has intruded
about 5 km into the Patuxent, and between 4 and 10 m depth the
isohalines throughout the lower estuary are nearly horizontal. The
topographically controlled front separating upper and lower basins has
an enhanced gradient. Velocities throughout the water column in the Pt.
Patience deep are uniformly small, less than *2 cm2 s"1. The velocity
profile at Lower Marlboro remains a normal two layer flow while
circulation within the deep hole has nearly vanished. It is clear that
low salinity, surface water intrusions from the Bay can produce
transient salinity patterns like that seen in figure 15a.
345
-------
I
O
(0
O
O
CM
0)
CO
(0
00
(0
00
CM
o
CM
CM
O
(0
I
^
10 -*-a^\^
z--a
0
h-
Q
Id ' '
1
o
o
(/to in o'
CO
2
o
M
o
0
.J
Ld
>
n
co
W'HId3Q
Figure 13: Plot of instantaneous surface elevation and velocity
profiles at 159 hours, during the May 21 tidal oscillation, for the
spring intensive run. Plot shows the abrupt sign reversal in surface
slope which occurs across the inner sill at 44 kilometers and the phase
difference in tidal velocity between profiles at 32 and 54 km.
346
-------
30
-4 4 12 20 28 36 44 52 60 68 76 84 92 100
DISTANCE (KM)
Figure 14: Contoured salinity variance calculated over (a) entire
spring intensive and (b) entire fall run 4. Contour interval .1.
347
-------
-15
I
-5
40-
45-1
IS
25
i
RIVER KILOMETER
35 45 55
65
75
I
85
95
105
i
Figure 15a: Patuxent River estuary salinity cross section, from a
single slack water cruise, summer 1984, from OEP Tech Rep. #7.
PT. PATIENCE
LOWER MARLBORO
6 2—2-6 -1O
VELOCITY.CM/S
10 6 2-2 -6 -1O
VELOCITY.CM/S
12 2O 28 38
52 60 88 76 84 92 1OO
DISTANCE (KM)
24 _
30
Figure 15b: Salinity pattern and velocity profiles over seventh tidal
cycle of summer intrusion experiment. Contours are ppt.
-------
CONCLUSIONS
Steady state simulations indicate that the seasonal average flow in the
lower basin and much of the upper basin of the Patuxent Estuary is a two
layer density driven flow, with velocities of ± 3 cm/s in the lower
basin. A stationary, nearly vertical salinity front is present in all
model calculations, centered on the inner sill separating the two
basins. The front forms regardless of initial conditions. Its location
is dictated by topography and its width is governed by the balance
between river discharge, the strength of lower basin density driven
circulation, and the magnitude of the horizontal turbulent diffusivity.
Transition to unidirectional (one layer) flow occurs well upstream from
the front, in the neighborhood of river kilometer 55. The best fit to
the observed seasonal average salinity distribution, based on a
comparison with data from spring 1984 (high discharge conditions), is
found for eddy mixing and viscosity coefficient values Kz - 0.2 and Nz -
5.0 cm2s"1.
Comparison between field data taken during the spring 1986 intensive
survey and the model calculation for the same period shows that the
model accounts for all of the major features seen in the data.
Calculated surface salinities are generally within the error estimates
of observed values throughout the whole estuary. The optimal value of
the horizontal mixing coefficient K^ is near 1.25xl05 cm*/s. This value
produces a salinity front on the inner sill above Benedict with a 0.35
ppt/km gradient, compared with a mean value of 0.4*.l ppt/km estimated
from the data. Comparison between measured and computed tidal
velocities shows generally good correlation throughout the estuary, with
an average observed/calculated ratio of 1.6. This indicates the lateral
distribution of tidal velocities is nearly parabolic. The range of
calculated tidal velocities is 20-50 cm/s. The estuary is dominated by
a 2-layer »5 cm/s shear flow. On average, the transition from one to
two layer flow occurs near river kilometer 60 but is very sensitive to
freshwater runoff. Three layer flows are not seen. Low frequency
variability, *50% of fluctuation of mean surface velocities and ± 80%
mean lower layer velocities, during this period is due primarily to
variations in boundary salinity.
Our comparison of the observations made during the fall intensive survey
reinforces the conclusions drawn from the spring survey comparison, and
in addition, highlights the wind driven component of the circulation.
Calculations using values of wind speed measured remotely match the
observed surface salinities whereas the calculation using twice the
measured winds does not.
As was found for the spring period, the two layer density driven flow
predominates in long term (>10 day) average. Average two layer flow
velocities at the mouth are calculated to be +5 and -6.25 cm/s, with the
level of no motion at 6-7 m depth. These agree with average values at
station PL1, corrected to give zero net flux. Three layer flows occur
only as brief transients, the result of combined wind and density
forcing. Wind driven fluctuations of *5-6 cm s"1 are seen within the
surface layer. Fluctuations in mean lower layer velocity are due to
both wind events, *2 cm s"1, and variations in mouth salinity, as much
as *4 cm/s. The magnitude of the wind driven component in the
calculations agrees with the observed magnitude. However, the phase of
wind driven fluctuations near the estuary mouth does not agree well with
the data from station PL1, as presented by Boicourt and Sanford (1987).
The deep hole and basins at 36 and 54 kilometers contain recirculatory
circulation cells which persist through both spring and fall
simulations. The basin at 68 kilometers maintains two layer circulation
during fall or low flow conditions but is flushed during spring or high
river flow conditions. There is no minimum in salinity variance within
the hole as would be expected were it stagnant. The deep hole is, on
average, well flushed by a vigorous two layer circulation to at least 30
m depth. The hole is subject to brief periods of stagnation in
association with halocline development in the lower estuary and or high
runoff events. Simulations of summertime conditions indicate that the
349
-------
deep hole at Pt. Patience can stratify and become stagnant in response
to sharp increases in Chesapeake Bay stratification. During these
episodes, a sub horizontal halocline develops over the deep hole and
circulation velocities below the halocline nearly vanish.
350
-------
References
Boicourt, W.C.; Sanford, L.P. A hydrodynamic study of the Patuxent
River estuary. Part A: Observations of current, temperature,
salinity, tidal height and wind. Part B: Hydrography of the
Patuxent River estuary. October 1985 to October 1986 (preprint).
Blumberg, F.A. The influence of density variations on estuarine tides
and circulations. Estuarine Coastal Mar. Sci. 6: 209-215; 1978.
Elliott, A.J. Observations of the meteorologically induced circulation
in the Potomac estuary. Estuarine Coastal Mar. Sci. 6: 285-299;
1978.
Festa, J.F.; Hansen, D.V. A two-dimensional numerical model of
estuarine circulation: the effects of altering depth and river
discharge. Estuarine Coastal Mar. Sci. 4: 309-323; 1976.
Festa, J.F.; Hansen, D.V. Turbidity maxima in partially mixed
estuaries: a two-dimensional numerical model. Estuarine Coastal
Mar. Sci. 7: 347-359; 1978.
Garvine, R.W. Dynamics of small-scale oceanic fronts. J. Phys. Ocean.
4: 557-569; 1974.
Garvine, R.W. An integral hydrodynamic model of upper ocean frontal
dynamics: Part I. Development and analysis. J. Phys. Ocean. £:
1-18; 1979.
Gordon, R.B.; Spaulding, M.L. Numerical simulations of the tidal- and
wind-driven circulation in Narragansett Bay. Estuarine Coastal
Shelf Sci. 24: 611-636; 1987.
Hibiya, T. Generation mechanism of internal waves by tidal flow over a
sill. J. Geophys. Res 91: 7697-7708; 1986.
Huzzey, L.M. The dynamics of a bathymetrically arrested estuarine
front. Estuarine, Coastal and Shelf Science 15: 537-552; 1982.
Oey, L.-Y. On steady salinity distribution and circulation in partially
mixed and well mixed estuaries. J. Phys. Ocean. 14: 629-645; 1984.
Oey, L.-Y.; Melor, G.L.; Hires, R.J. A three-dimensional simulation of
the Hudson-Raritan Estuary. Part I: Description of the Model and
Model Simulations. J. Phys. Ocean. 15: 1676; 1985.
Olson, P. The spectrum of subtidal variability in Chesapeake Bay
circulation. Estuarine, Coastal Shelf Sci. 2_3: 527-550; 1986.
Patuxent Estuary Water Quality Survey: 1984 Data Summary. MD Office of
Environmental Programs Tech Rep #7, 1985.
Phillips, O.M. The Dynamics of the Upper Ocean. Cambridge: Cambridge
University Press; 1980.
Prandle, D. Salinity intrusion in estuaries. J. Phys. Ocean. 11:
1311-1324; 1981.
Wang, D.P. Wind-driven circulation in the Chesapeake Bay, winter 1975.
J. Phys. Ocean. 9: 564-572; 1979.
Wang, D.P.; Kravitz, D.W. A semi-implicit two-dimensional model of
estuarine circulation. J. Phys. Ocean. 10: 441-454; 1980.
Wong, K-C; Garvine, R.W. Observations of wind-induced subtidal
variability in the Delaware estuary. J. Geophys. Res. 89:
10589-10597; 1984.
557
-------
Understanding the Estuary: Advances in Chesapeake
Bay Research. Proceedings of a Conference. 29-31
March 1988. Baltimore, Maryland. Chesapeake Research
Consortium Publication 129. CBP/TRS 24/88.
Lagrangian Drift Model of Suspended Sediment Transport
in Chesapeake Bay
Kurt W. Hess
National Environmental Satellite, Data and Information Service
National Oceanic and Atmospheric Administration
INTRODUCTION
Suspended sediment transport in Chesapeake Bay is simulated
with a numerical model of circulation and Lagrangian drift,
and modeled distributions are compared to in situ and
remotely sensed data. Suspended sediment, which is
important in itself for transporting chemicals and for
limiting photosynthesis by blocking light, also acts as a
tracer for circulation patterns and can be used in
understanding transport mechanisms. More importantly,
suspended sediment is a variable which can potentially be
used to calibrate and verify circulation models, and unlike
water currents, it can be remotely sensed, making data
acquisition rapid and inexpensive. Model results also offer
the possibility of interpolating conditions in space and
time when only limited observational data is available.
The circulation model used in this study has been used to
hindcast tidal, wind-driven, and density currents in the Bay
and the local continental shelf (Hess 1985, 1986). The
Lagrangian drift model (drifters move at the speed of the
surrounding water) was originally developed to simulate the
motion of free-moving larval stages of marine organisms
(Johnson 1987) and was applied to blue crab larval drift
(Johnson et al. 1986, 1987; Hess and Johnson 1988). That
drift model has been modified to include settling under the
force of gravity and the multiple injection of drifters to
352
-------
simulate a continuous source. A settling velocity which
represents suspended sediment particles in Chesapeake Bay
(Schubel 1972) was chosen and applied uniformly to all
drifters. When settling takes the particle to the bottom,
the particle is removed from further advection.
Identification of areas of deposition can also be made.
A convenient test case for the simulation is the high
turbidity occurrence in the upper Bay during the high-
precipitation event in March 1979. Remotely-sensed
(Stumpf 1988) and in situ (Cronin 1982) observations of
water turbidity in the Bay are compared to model simulation
results.
THE NUMERICAL CIRCULATION MODEL
Hess (1985, 1986) has developed a general three-dimensional,
free-surface numerical circulation model, called MECCA
(Model for Estuarine and Coastal Circulation Assessment).
MECCA uses finite difference approximations to the momentum,
continuity, and temperature and salinity equations to
simulate time-varying water currents, salinities, and
temperatures in a shallow water domain at time scales from a
few minutes to several months, and space scales from a few
kilometers to a few hundred kilometers. The model is
designed to simulate circulation driven by tides, winds,
water density gradients, and atmospheric pressure gradients.
A fully three-dimensional, time-variable approach was
decided upon because of the importance of knowing surface
currents (as opposed to layer-averaged currents), and
because of the rapid changes in currents expected to occur
when tides and winds are important.
The major feature of the circulation model is the use of
split-mode velocity equations. The external, or barotropic,
velocity mode (vertically-averaged velocity) is subtracted
from the total velocity to get the internal, or baroclinic,
velocity mode. Because the internal mode stability
requirement in the numerical solution scheme is less
stringent than that for the external mode, the internal mode
can be updated less often, resulting in significant savings
in computer time. For these simulations, the external-mode
time step used was 6 minutes, and the internal-mode time
step was 30 minutes. The model uses a dimensionless
vertical coordinate, also known as a sigma coordinate. This
terrain-following coordinate improves the representation of
the bathymetry.
All modeled mass and momentum diffusivities are functions of
local velocity gradients, and therefore may vary over time
and space. In addition, vertical diffusivities are reduced
in the presence of vertical density gradients as a function
of the local Richardson number.
353
-------
The differential equations are approximated by finite
differences on a grid mesh of 434 square and triangular
elements 11.2 kilometers on a side (Fig. 1) . Placement of
the variables in the horizontal plane is staggered to
prevent spurious numerical solutions. Vertical staggering
at the 10 model levels allows for better resolution of
vertical gradients. Semi-implicit grid-row manipulations
are employed in both the vertical and horizontal to augment
computational stability. The grid mesh allows for the
resolution of narrow inlets in barriers and for variable-
width channels. These channels allow for rivers, which
generally have widths much narrower than a typical grid
cell, to be explicitly included in the numerical grid
scheme.
Tidal forcing at the deep-water boundaries was represented
by the sum of constituent tides with amplitudes and phases
adjusted to reproduce the tides at the Hampton Roads gage
and the currents at the entrance to Chesapeake Bay. Twenty
constituents are included and the offshore amplitude is
updated each 6 minutes.
Daily river flows at five major rivers (Susquehanna,
Potomac, Rappahannock, York, and James) are supplied at grid
cells representing the approximate location of the fall
lines. These values are assumed to represent midday
conditions; quadratic interpolation is used to estimate
flows at other times.
Numerous model calibration and verification studies have
been carried out over the past few years (Hess 1986) .
Validation studies have shown that tides and tidal currents
in the mouth are simulated to the acceptable level of
accuracy of 10 to 20 cm/s, or 15 percent of full scale (Fig.
2), and that tidal currents throughout the Bay as a whole
are of similar accuracy (Fig. 3). Further tests (Hess 1985)
indicate that wind-driven and density-driven currents are
modeled to a similar level of accuracy. In sum, MECCA
produces reasonable approximations to the actual currents
when historical input data are used. More importantly, it
is a useful tool for the comparison of scenarios covering
long periods, as in high runoff events modeled here.
Several simplifications of the model were made for these
simulations. No winds were applied, a linearized bottom
friction was assumed, and water density was assumed to vary
only with salinity. A single-component tide with an
amplitude of 0.4 meters and a period of 12 hours was used to
drive the currents. Constant horizontal and vertical
viscosities were used.
THE LAGRANGIAN DRIFT MODEL
Johnson (1987) has developed a Lagrangian drift model called
LARTREK. LARTREK simulates the movements and distribution
354
-------
NORTHERN
BOUNDARY
DEEP WATER
BOUNDARY
SOUTHERN
BOUNDARY
Figure 1. The circulation model grid mesh for Chesapeake
Bay and the adjacent coastal waters showing the location of
oceanic boundaries, and river inputs. Grid cells measure
11.2 x 11.2 km. Cells marked with an "X" along the deep-
water boundary require input water level data. Cells along
the northern and southern boundaries marked with a "«" use
a radiation outflow boundary condition. Cells marked with a
">" at the heads of rivers require flowrate boundary
conditions.
355
-------
0
Figure 2. Modeled and NOS-predicted tides and currents near the
mouth of Chesapeake Bay. Flood currents are positive.
0
D
E
L
E
0
12
10
oa
06
04
02
00
MEAN TIDAL RANGES
(HETERS)
00 02 04 06 08 10 12
OBSERVED
rt
0
0
E
I
TIDAL
CURRENT
SPEEDS
(n/s)
04 Oft
OBSERVED
(a)
(b)
Figure 3. Modeled and NOS-observed (a) mean tide ranges and (b)
current amplitudes at several locations in Chesapeake Bay.
356
-------
of Lagrangian drifters using the three dimensional
velocities calculated by MECCA, and algorithms to simulate
beaching and deposition. LARTREK is run concurrently with
MECCA.
The simulation of suspended sediment transport with discrete
Lagrangian drifters offers several advantages over the
solution of the advective-diffusive equation. Large
concentration gradients at a sub-grid scale can be produced
with Lagrangian drifters. Identification of individual
drifters allows for explicit representation of the time the
sediment has been water-borne or deposited, and for
determination of the point of origin. An inhomogeneous
sediment can be represented by multiple settling velocities.
The Lagrangian approach also offers advantages when applied
to biological organisms which have swimming or migratory
behaviors.
Horizontal advection
After each external-mode time step of MECCA (6 minutes of
simulated time), LARTREK calculates new positions for each
drifter by three-dimensionally interpolating the u, v and w
components of velocity from MECCA. Because the horizontal
velocity field, v(x), may vary over the distance a drifter
moves within a time step, the drifter's position is advanced
in the horizontal plane by a predictor-corrector method.
For a drifter initially at point xo, the first estimate of
the spatially-averaged velocity, v1, is v(x0). Then an
estimate of the updated position, Xp, is computed by
xp = x0 + v'At (1)
where At is the circulation model's external mode timestep.
An improved estimate of the spatially-averaged velocity over
the path is then computed by
V = [v(x0) + v(xp)]/2 (2)
Another estimated end position is then recomputed from (1);
four iterations of this procedure have been shown to be
sufficient (Johnson et al. 1986).
A small random drift velocity is computed using the
relationship
v" = [2Ah/At]1/2 (3)
where A^ is the local horizontal turbulent eddy viscosity
(Csanady 1973) . The velocity v" is added to the spatially-
averaged velocity at each timestep in a randomly chosen
direction.
In this numerical drift model, the interaction of a water-
borne drifter with the shoreline is carefully controlled.
In the present version of the model, drifters are reset to a
357
-------
distance of 100 meters from land if they encounter a land
boundary. Drift parallel to the boundary is not restricted.
In some areas of the MECCA grid system, the land-water
boundary occurs adjacent to a triangular grid. Because of
the differences in geometry between square and triangular
grids, separate but analogous land interaction schemes and
velocity interpolation schemes are invoked for the
triangular grid cells.
Application to Suspended Sediment
A settling velocity is added to the drifter's vertical
velocity to simulate sediment settlement and deposition. A
velocity of 10~5 m/s (approximately 1 meter per day) was
used in all simulations (Schubel 1972). When a drifter
touches bottom it is deposited and removed from further
motion; there is no resuspension. Diffusion in the vertical
direction is approximated by introducing a random motion.
Suspended sediment entering the Bay through a river is
simulated by the simultaneous release of five Lagrangian
drifters in the MECCA cell which represents the river's
uppermost reach. The drifters have the same horizontal
coordinates and are spaced uniformly over the vertical,
starting from the surface. Injection occurs at specified
time intervals following a hydrodynamic spinup period.
To insure a constant concentration, an injection rate which
depends on the river flow is used. Given a reference flow
(Qr) and time interval (tr), the injection time interval
(tj_) for any other flow (Q) is
ti = (Qr/Q)tr - (4)
Here Qr=2000 m3/s and tr=6 hours. Although there is no
convenient way to accurately estimate concentration from the
distribution of drifters, one could simply divide the number
of drifters in each MECCA cell by the cell's volume to get a
relative concentration.
RESULTS OF SIMULATIONS
General sensitivity tests
Several scenarios were run to test the relative importance
of different factors. Early tests showed that the non-
linear advective terms have little effect in determining
sediment distribution.
A series of runs was made to determine the time required to
come to at state of equilibrium (i.e. when the rate of
injection equals the rate of removal by deposition) for
suspended sediment entering the Bay through the Susquehanna
River (Fig. 4a). For a Susquehanna flow of 4000 m3/s, the
time was 7.5 days (Fig. 4b); for a flow of 8000 m3/s, the
time was 11 days (Fig. 4c); for a flow of 12,000, the time
was 12 days (Fig. 4d). The difference in the required times
358
-------
SUSOUCMAMU I.
IALTIMM
CHCSTO «.
I I
(A)
(B)
^
(C)
(D)
Figure 4. Simulated equilibrium positions of Lagrangian
drifters in the upper Chesapeake Bay for various Susquehanna
River (A) flows. Modeled position of drifters (shown as a
small square drawn at the drifter location) in the uppermost
4 meters at equilibrium for flows of (B) 4,000 m3/s, (C)
8,000 m3/s and (D) 12,000 m3/s.
359
-------
is due to the advection of drifters into deeper water when
the flow is increased.
Tests were made to determine the effect of the vertical
random velocity. Computing a value from (3), but with the
vertical diffusivity (Av=0.003 m2/s) rather than the
horizontal diffusivity (An), gave the diffusive velocity
w'=4 x 10~3 m/s. This speed caused rapid deposition of
drifters and was not suitable for representing diffusion.
An alternative diffusive velocity, 10~4 m2/s, was used
instead.
March 1979 flood simulations
The high flow event in March 1979 is an ideal test case for
the simulation model because shipboard and satellite data on
suspended sediment distributions are available. Daily mean
river flowrate data for the Susquehanna River (Fig. 5) were
taken from U.S. Geological Survey records (USGS 1979) for
the period of the simulation (1 February - 31 March, 1979).
Annual mean river flowrates were used for the Potomac,
Rappahannock, York, and James Rivers.
Observed suspended sediment concentrations in the
Susquehanna River (Lang 1982) varied with flow, increasing
to a peak on 7 March 1979, and diminishing thereafter; there
is no data before 5 March (Fig. 5). The modeled
concentration was assumed to be zero until 5 March, then to
vary directly with the flow until 13 March, when it again
becomes zero. The injection rate which approximately fits
this assumption is
ti = (Qr/Q)2tr (5)
Observations of surface concentrations are available for
four March dates. Landsat images showing Bay-wide surface
water turbidity distribution with a resolution of less than
a kilometer are available for 9 and 17 March (Stumpf 1988).
Ship cruise data is available for 12-13 and 26-27 March
(Cronin et al. 1982).
The observed and simulated distribution for 9 March 1979 is
shown in Fig. 6. The remotely-sensed area of maximum
distribution extends from the Susquehanna River entrance
southward to latitude 39.00 N. The modeled region of
maximum sediment in the upper 4 meters does not extend as
far south.
The observed and simulated distribution for 12-13 March 1979
is shown in Fig. 7. The shipboard data's area of maximum
distribution extends from the Susquehanna River entrance
southward to about'20 kilometers below its previous
(9 March) position. The modeled region of maximum sediment
in the upper 4 meters extends only as far south as the
northern entrance of the Chester River.
360
-------
14000
12000
10000
8000
6000
4000
2000
0
FLOW
CONCENTRATION
50
60
<
oe.
o
o
Figure 5. Daily flow and suspended sediment concentration in
the Susquehanna River for I February - 31 March, 1979. Flow
is shown in the units m3/s and concentration is in units
ing/1.
361
-------
UPPER
CHESAPEAKE BAY
09 MAR 79
ASSAFRAS
ft.
3
*
(A)
(B)
Figure 6. (A) Observed (satellite) distribution of surface
suspended sediments (dark areas indicate high
concentration), and (B) simulated positions of Lagrangian
drifters (squares) in the upper 4 meters in the upper
Chesapeake Bay model grid for 9 March 1979.
362
-------
(A)
(B)
Figure 7. (A) Observed (shipboard measurement) distribution
of surface suspended sediments (dark areas indicate high
concentration), and (B) simulated positions of Lagrangian
drifters (squares) in the upper 4 meters in the upper
Chesapeake Bay model grid for 12-13 March 1979.
363
-------
The observed and simulated distribution for 17 March 1979 is
shown in Fig. 8. The remotely-sensed area of maximum
distribution extends from the Sassafras River entrance
southward to about 10 kilometers south of latitude 39.00 N.
The modeled region of maximum sediment in the upper 4 meters
does not extend as far south.
The observed and simulated distribution for 26-27 March 1979
is shown in Fig. 9. The shipboard data's shows two areas of
high concentrations, one near the Bush River and the other
just north of the Patapsco River. The modeled region of
maximum sediment in the upper 4 meters shows some
patchiness; areas of high concentration are located near
Annapolis and the entrance to Chester River.
Many of the simulations showed that modeled sediment was not
advected as far to the south as the observations would seem
to indicate. To test the influence of the settling velocity
on the distribution, another simulation of the March flood
was carried out with a settling rate of 0.5 x 10~5 m/s, half
the previously used value. There was only a small increase
in the area of maximum near-surface concentration.
CONCLUSIONS
These results of preliminary computer simulations of
suspended sediment transport demonstrate the potential
usefulness of the technique. Although the sediment.
generally followed the expected patterns, in some cases
modeled suspended sediment did not move a far south in the
Bay as observations would indicate. The simulations may not
be unrealistic, however, given the assumptions of no wind
and no sediment release before 5 March.
Further investigations should focus on the effects of grid
resolution on the currents, influence of density gradients,
the representation of vertical diffusion, and the role and
modeling of resuspension.
ACKNOWLEDGEMENT
The author would like to thank his colleague Dr. Richard
Stumpf for supplying data for this study and for his
valuable discussions about remote sensing, modeling, and
suspended sediment transport.
364
-------
(A)
(B)
Figure 8. (A) Observed (satellite) distribution of surface
suspended sediments (dark areas indicate high
concentration), and (B) simulated positions of Lagrangian
drifters (squares) in the upper 4 meters in the upper
Chesapeake Bay model grid for 17 March 1979.
365
-------
26-27 MAR 79
*
(A)
(B)
Figure 9. (A) Observed (shipboard measurement) distribution
of surface suspended sediments (dark areas indicate high
concentration), and (B) simulated positions of Lagrangian
drifters in the upper 4 meters in the upper Chesapeake Bay
model grid for 26-27 March 1979.
366
-------
REFERENCES
1. Cronin, W. B.; Taylor, W. R.; Mallonee, M. Suspended
sediment distribution and related data, September 1978
- October 1980, Part I. Chesapeake Bay Institute, Johns
Hopkins University, Baltimore , Md., CBI Open File
Report no. 27; 1982.
2. Csanady, G. T. Turbulent diffusion in the
Environment. Reidel, Boston; 1973.
3. Hess, K. W. Assessment model for estuarine circulation
and salinity. NOAA Technical Memorandum NESDIS AISC 3,
Washington, DC: U.S. Department of Commerce, National
Oceanic and Atmospheric Administration; 1985; 39 p.
Available from: Marine Environmental Assessment
Division, 1825 Conn. Ave., NW, Washington DC 20233.
4. . Numerical model of circulation in Chesapeake Bay
and the continental shelf. NOAA Technical Memorandum
NESDIS AISC 6, Washington, DC: U.S. Department of
Commerce, National Oceanic and Atmospheric
Administration; 1986; 47 p. Available from: Marine
Environmental Assessment Division, 1825 Conn. Ave., NW,
Washington DC 20233.
5. Hess, K. W.; Johnson, D. F. Numerical simulations of
blue crab larval drift and larval recruitment to
Chesapeake Bay. Marine Environmental Assessment
Division, Washington, DC: U.S. Department of Commerce,
National Oceanic and Atmospheric Administration; 1988;
34 p. Available from: Marine Environmental Assessment
Division, 1825 Conn. Ave., NW, Washington DC 20233.
6. Johnson, D. F. A model for the simulation of larval
drift. NOAA Technical Memorandum NESDIS AISC 8,
Washington, DC: U.S. Department of Commerce, National
Oceanic and Atmospheric Administration; 1987; 21 p.
Available from: Marine Environmental Assessment
Division, 1825 Conn. Ave., NW, Washington DC 20233.
7. Johnson, D. F.; Hess, K. W.; Pytlowany, P. J.
Interdisciplinary synoptic assessment of Chesapeake Bay
and the adjacent shelf. NOAA Technical Memorandum
NESDIS AISC 5, Washington, DC: U.S. Department of
Commerce, National Oceanic and Atmospheric
Administration; 1986; 90 p. Available from: Marine
Environmental Assessment Division, 1825 Conn. Ave., NW,
Washington DC 20233.
8. . Circulation modeling as an aid to management of
the blue crab fishery in Chesapeake Bay. Proceedings of
the Tenth National Conference of the Coastal Society,
12-15 October 1986, New Orleans, LA. 1987: 111-118.
367
-------
9. Lang, D.J. Water quality of the three major tributaries
to the Chesapeake Bay: the Susquehanna, Potomac, and
James Rivers, January 1979-April 1982. USGS Water
Resources Investigation 82-32.
10. Schubel, J. R. Distribution and transportation od
suspended sediment in upper Chesapeake Bay. Geological
Society of America, Memoir 133, 151-167; 1972
11. Stumpf, R. P. Sediment transport in Chesapeake Bay
during floods: analysis using satellite and surface
observations. Jol. Coastal Res., 4, 1-15; 1988.
12. United States Geological Survey, 1979. Water Resources
Data, Maryland and Delaware, Water Year 1979, Water
Data Report MD-DE-79.
368
-------
Understanding the Estuary: Advances in Chesapeake
Bay Research. Proceedings of 'a Conference. 29-31
March 1988. Baltimore, Maryland. Chesapeake Research
Consortium Publication 129. CBP/TRS 24/88.
How to Estimate the Thickness of Benthic Boundary Layers
in Estuaries With and Without Tides
Sergei A. Kitaigorodskii
The Johns Hopkins University
Department of Earth and Planetary Sciences
Baltimore, Maryland 21218
The benthic boundary layer (BBL) is the fluid layer adjacent to the sea
bottom. It is usually characterized by strong shear and intensive
vertical mixing. In the past the BBL has been modelled as an Ekman
layer. However, as was recently shown by the author, in presence of
imposed background stable stratification the Ekman boundary layer model
failed and the height of BBL is determined by the scale Lj, - u* / N,
where u» is the friction velocity based on the bottom stress and N is
the buoyancy frequency, which is chosen to characterize the background
stratification (can be considered also as initial stratification). It
is interesting to apply this idea to the vertical mixing in estuarine
systems in presence of strong tides. Some estimates relevant to the
description of the variability of heights of BBL in such conditions will
be presented.
The purpose of this short paper is to show that the benthic boundary
layer (BBL) generated by tidal waves can be strongly influenced by the
presence of imposed background stable stratification, since the
thickness of BBL is much less than the wavelength, the Prandtl model for
laminar BL can be described by the equation
du. d2U dam
V - = (1)
dt dz* dl ^ J
where u. is the free stream velocity which in the case of tidal wave on
finite depth can be presented as
(2)
369
-------
where v,-aa/shkD, a is wave amplitude, D is depth, k and a are
correspondingly wave number and frequency. The thickness of laminar
benthic boundary layer f> in such case is given by the usual expression
where v is kinematic viscosity, T is wave period. It is well known
since the work of Collins (1963) that in natural conditions the BBL is
always turbulent even if wave amplitude a doesn't exceed a few
centimeters. Since the period of tidal wave T and the characteristic
length x-2«/A: exceeds the typical time and length scales of turbulence
in the BBL, the tidal motion can be considered as a continuous sequence
of stationary realizations for the BBL, or at least the parameters of
BBL can be treated as slowly varying functions of time and position
(compare with T and x-2*/*). This justifies the approach when to
estimate the height h of a turbulent benthic boundary layer we can
introduce an effective value for eddy viscosity Kh in such a way that
instead of (3) we can write
\ 1/2 / ,, \ 1/2
For effective turbulent viscosity in shear driven BBL we can use the
simplest expression
Kh*u.h (5)
Here we avoid a numerical constant of proportionality being of order
one. More precisely the expression (5) can be written as
Kh*>HU,h (6)
where x-o.4. With (6) the expression (4a) can be rewritten as
h = 2*L0 (7)
where the scale L, is defined as
u,
L.-- (8)
The scaling length L, corresponds to the scale of the well mixed region
in neutrally stratified tidal BBL and is a fundamental estimate for the
thickness of turbulent BL generated by periodic tidal motion, the
application of the scale L, to the estimates of the thicknesses of
benthic boundary layers need first of all its comparison with internal
Ekman scale i,-u./n (n-Coriolis parameter, which is a good measure of the
height of stationary turbulent boundary layer. Since as a rule
we can conclude that BBL generated by tidal motion in neutrally
stratified environment can easily reach Ekman height Le. A typical
range of values of u* in tidal flows with different bottom roughness
conditions are 0.7-4 cm/sec (see Kagan, 1968). For this range of values
of u* the semidiurnal tide will produce BBL of the thicknesses which are
comparable with the Ekman height Le. However some recent observations
of the turbulence close to the sea bottom (see for example Ozmidov,
1987) and also the numerical simulation of the formation of BBL in
initially continuously stratified fluid subject to a suddenly imposed
barotropic pressure gradient demonstrate that the thicknesses of the
370
-------
well mixed turbulent region close to the bottom of the sea are
sufficiently less than the scales L. and L,. Following recent work by
the author (S.A. Kitaigorodskii, 1988) it can be explained by taking
into account the effect of initial stable stratification N(z,x), where N
is Brunt-Vaisals frequency (can be considered as slowly varying function
of time and position). In this case the growth of BBL is limited, since
the flux Richardson number Rfh in the vicinity of the entrainment zone
must be less than its limiting value Rfi
-------
which demonstrates that if the tidal wave amplitudes are very small the
BBL can become very thin compared both with Ekman height Le and height
hN - b Lu. The detailization of (17) can be done by assuming some form
for the bottom stress u*2. If we'll use the classical quadratic law
u'.~u'.~a'o' then
a aa
- = — -const (17)
L. g U. •
Even though aa/u.-io" the square root of it in (17) permits still to
write that
h*LN-La^ (18)
Therefore in the case of imposed background stratification we can simply
write that the thickness of equlibriura turbulent boundary layer are
defined only by two scales 1^, and i. , so that
(20)
where
The two asymptotic expressions for Y (v,~) will correspond to AI,,-"»
(neutral BBL) when h~Lt and v, -*o when ti~L*.
References
Collins, J. I. Inception of turbulence at the bed under periodic
gravity waves. J. Geophys : Res. 68. : No. 21; 1963.
Kagan, B. A. Fluid dynamical models of tidal motions in the sea.
Leningrad: Hydroraeteorological Publishing House; 1968.
Kitaigorodskii, S. A. A note on similarity theory for atmospheric
boundary layers in presence of background stable stratification.
Tellus: [1988 in press].
372
-------
Understanding the Estuary: Advances in Chesapeake Abstract only
Bay Research. Proceedings of a Conference. 29-31
March 1988. Baltimore, Maryland. Chesapeake Research
Consortium Publication 129. CBP/TRS 24/88.
High-Resolution Thermistor Chain Observations
in the Upper Chesapeake Bay
Charles C. Sarabun, Jr.
The Johns Hopkins University
Johns Hopkins Road
Laurel, Maryland 20707
Selected results from four years (1984-1987) of thermistor chain data
and coincident CTD, current meter and acoustic backscatter measurements are
presented. The data shown exemplify a number of super-tidal features which
were ubiquitous during the measurement periods. The features are subsurface
intrusions with mixing and high-frequency internal waves on their surfaces,
estuarine surface fronts, monochromatic, high-amplitude internal wave trains,
breaking internal waves and broader-band internal wavefields. These features
are discussed in light of their effects on mixing and transport, and their
implications for monitoring sampling strategies and interpretation of results,
373
-------
Understanding the Estuary: Advances in Chesapeake
Bay Research. Proceedings of a Conference. 29-31
March 1988. Baltimore, Maryland. Chesapeake Research
Consortium Publication 129. CBP/TRS 24/88.
Satellite Observations of Turbidity Variations in
Chesapeake Bay, Spring-Summer 1987
Richard P. Stumpf
National Oceanic and Atmospheric Administration
National Environmental Satellite, Data, and Information Service
Assessment and Information Service Center
Washington, DC 20235
INTRODUCTION
The transport and distribution of suspended sediments can have a strong
influence on the living resources of Chesapeake Bay through their effect
on turbidity and siltation. The suspended matter shows considerable
spatial and temporal variability--particularly with changes in river
discharge. As these variations may occur within a few days within
portions of the Bay system; frequent, synoptic observations are needed
to evaluate the changing conditions.
Satellite sensors may provide routine synoptic coverage of the turbidity
or sediment loads in the surface waters of the Bay. Although satellites
have been used as a means of routine monitoring of chlorophyll and
temperature in the ocean (e.g. Brown et al. 1985), they have not been
used for routine studies of estuaries. The experimental Coastal Zone
Color Scanner (CZCS) had some potential for this purpose, however some
of the bands did not have the dynamic range to handle turbid water. (It
was also shut down in June, 1986.) Landsat and SPOT sensors provide
only 1-2 images per month, therefore the frequency of usable (cloud-
free) images is too low to monitor changes caused by episodic events.
In contrast to these other sensors, the Advanced Very High Resolution
Radiometer (AVHRR) can provide daily images with usable radiometric
information; therefore it appears to be the best sensor presently
available to investigate estuaries.
Currently, this sensor can provide reliable estimates of sea surface
temperatures (Strong and McClain 1984). It has been used to detect
chlorophyll blooms in turbid water (Stumpf and Tyler 1988), and through
374
-------
the calculation of water-reflectance from the sensor data, Stumpf (1987,
1988a, 1988b) has shown the ability of this sensor (and others) to
provide estimates of suspended sediments (seston) or turbidity measures
such as the Secchi depth or attenuation coefficient. This paper will
show the utility of this sensor in monitoring short-term and seasonal
changes in turbidity, as defined by reflectance, in Chesapeake Bay using
data from 1987.
METHODS
Satellite Characteristics
The AVHRR is onboard the NOAA TIROS-N (Television and Infrared
Observation Satellite) platform, which is polar-orbiting and sun-
synchronous. NOAA-6, 8, and 10 overpasses occur at about 0800 and 2000
local time. NOAA-7 and NOAA-9 overflights occur about 1430 and 0230
local time. During 1987, the NOAA-9 and NOAA-10 satellites were
operational. The satellites have approximately a 9-day cycle and track
from west to east, producing about 6-7 potentially usable daytime scenes
during each cycle.
The AVHRR scans orthogonally to the direction of travel over an angle of
+55.4° from nadir. The usable scene width is about 2000 km. The sensor
has a field of view (pixel size) of 1.4 milliradians, which corresponds
to a ground diameter of 1.1 km at nadir. The satellite collects 360
scanlines per minute.
The AVHRR has either 4 or 5 bands as shown in Table 1. Channels 3, 4,
and 5 are used to calculate sea surface temperature. Channels 1 and 2
are used here in the calculation of water reflectances. The
specifications of the AVHRR are described in detail in Kidwell (1985)
and Lauritson et al. (1979). Stumpf (1987) and Everdale (1986) present
example treatment and applications of the data to coastal oceanography.
Table 1. AVHRR Spectral Bands
channel 1234 5*
wavelength .58-.68 .72-1.0 3.5-3.9 10.5-11.3 11.5-12.5
(Von)
description red near-IR thermal - infrared
*Available only on NOAA-7 and NOAA-9.
Processing
The images were obtained as Level IB digital data tapes from NOAA's
Satellite Data Services Division. The images were mapped to a Mercator
projection with a pixel size of 1.18 km at 38°N using a nearest neighbor
routine, and missing data points were filled with a 3x3 average. An
additional linear correction produced a positioning accuracy of 1 pixel
375
-------
for the shoreline. Both reflectances and SST's were calculated for each
image. The temperatures, not presented here, were calculated using the
equations in Kidwell (1985).
The reflectance is the ratio of irradiance leaving the water to the
irradiance entering the water. This is estimated using
R(X) - -- (1)
E0(X) cos 90 exp{-[Tr(X)/2 + Toz(X)]/cos 9O)
where
- LA(A)] / TL(X) (2)
R is the above water reflectance, X is the spectral band, L^ is the
radiance from the water, L* is radiance at the sensor, LA is the
atmospheric path radiance, T^ is the transmission coefficient from the
earth to the sensor, Eo is the solar irradiance outside the atmosphere,
90 is the solar zenith angle, Tr is the Rayleigh optical depth, TOZ is
the gaseous (e.g., ozone) absorption optical depth (Stumpf 1988b) .
Equation (2) is an atmospheric correction assuming uniform atmosphere
over the scene area. This acts as a bias correction for both Rayleigh
scattering radiance and minimum aerosol radiance (Stumpf 1988b) .
An additional correction is performed to remove variable haze and clouds
(Stumpf 1988b) . This correction is
RD - R(l) - YR(2) (3)
where Rp is the cloud-corrected reflectance; R(l) and R(2) are
determined from (1); and Y is a constant determined by the aerosol
characteristics. This correction has the advantage of removing
atmospheric contamination pixel-by-pixel. Here, Y is set equal to 1.0,
which applies to aerosols having relatively large particles, such as
clouds, that produce achromatic haze. Heavy overcast and some (colored)
haze types cannot be removed and generally produce elevated
reflectances. Overcast was masked in white using a combination of
temperature and reflectance data.
RJJ has a precision within about +0.004 depending on the presence of
uncorrectable haze. The accuracy cannot be readily determined, owing to
uncertainties in the sensor calibration. There is evidence that the
NOAA-9 AVHRR channel 1 and channel 2 detectors have shown a decrease in
sensitivity since launch, however the sensor has no onboard method of
calibrating this change. Currently, the calibration has not been
corrected for changes in sensitivity, therefore reflectances shown in
this report may be somewhat less than the true reflectances (or those
observed when the NOAA-9 was first launched in 1985). Within the 7-
month period investigated here, changes in the sensor should remain
slight.
Stumpf (1987) has shown that the reflectance can be related to the
376
-------
suspended sediment concentration through the relationship
•33 bbs*
S* + ax/ns
(A)
JU
where ns is the sediment concentration; bbs is the specific backscatter
coefficient for the sediment; S* - as* + bbs*. where as* is the specific
absorption coefficient of materials related to the sediments; and ax is
the absorption coefficient for water, dissolved and algal pigments.
Constant values of bbs and S can be used to describe an estuary over a
range of data. Variations in ax, such as would occur in strong blooms,
may affect the relationship, and may reduce the value of Rp somewhat
below the true value (Stumpf 1988b). The form of this relationship
causes R to vary approximately with the logarithm of ns.
Monthly mean reflectances were determined using four weekly scenes per
month. The spatial means were found using all the pixels within a
geographic area. The upper Chesapeake extends from 39°27'N to 39°02'N,
the middle Bay from 39°02'N to 38°22'N, the lower Potomac from 77°01'W
(Morgantown) to the entrance, and the lower James from 76°41'W to the
entrance.
River flow data was obtained from USGS gaging station data. The
stations are at Conowingo Dam, Maryland for the Susquehanna, near Chain
Bridge at Washington, DC for the Potomac, and Cartersville--about 60 km
above the fall line at Richmond--for the James.
RESULTS
River Flow
The three major tributaries to the Bay, the Susquehanna, Potomac, and
James Rivers, showed their highest flows of the year in March and April
1987 (Figure 1). Peaks occurred about March 3 in the Potomac and James.
The Susquehanna had high flow about March 11. In April, the Susquehanna
and Potomac Rivers had high flow events somewhat early in the month,
both rivers peaked about April 6 and the Potomac had its maximum peak
for the year about April 18. The James peaked twica later in April,
with flood conditions about April 18 and high flow on April 26. Flows
from June to September corresponded to those in May, with no high flow
events through the summer.
The satellite imagery shows variations in reflectance consistent with
the expected influx of sediments during these events. As an example,
Figures 2 and 3 show reflectances on April 1 and April 10. The
Chesapeake and lower Potomac have low to moderate reflectances (<0.025)
on April 1, with the James showing moderately high reflectances (0.03).
On April 10, after the Potomac and Susquehanna had flow peaks about
April 6, their estuaries showed high reflectances. The Susquehanna
plume extended into the middle Bay below Annapolis (marked A in the
figures). The turbidity plume from the Potomac extended into the main
Bay, with reflectances of double those of April 01, even though the flow
peak was not exceptionally large for this river. In contrast the James
estuary had maintained approximately the same reflectance, consistent
377
-------
aooo -i
eooo-
SUSQUEHANNA RIVER
O
4OOO-
200O
sooo n
4OOO -
,3OOO •
20OO-
1OOO-
400O-I
30OO •
o
2OOO-
1000 •
POTOMAC RIVER
JAMES RIVER
Figure 1. River discharge for the Susquehanna, Potomac, and James
Rivers, Spring of 1987.
378
-------
Figure 2. Rp reflectances for Chesapeake Bay found from NOAA-9 data on
01 April 1987.
379
-------
Figure 3. RD reflectances for Chesapeake Bay found from NOAA-9 data on
10 April 1987.
380
-------
with the lack of a strong peak in discharge.
The high flow events clearly carry material into the lower portions of
the estuaries (Figure 4). The- early March event apparently caused
increased mean reflectances in the lower Potomac and James. In
contrast, the March 5 and March 11 peaks in the Susquehanna did not
carry sufficient material into the middle Bay to affect the mean
reflectances. The April 6 peak in the Susquehanna (highest of the year)
did carry sediment well into the middle Bay (Figure 3). The lower
Potomac also had increased reflectances resulting from the April 6 and
April 18 peaks.
Reflectances in the lower James peaked later in April, consistent with
the later peak in the river discharge. The Potomac and Chesapeake
reflectances decreased to summertime levels by mid-May. The lower James
finally decreased to a mean reflectance of 0.015 by late June.
Monthly Mean Reflectance Variations
The variations in mean reflectances over the Bay during the spring and
summer followed the seasonal changes in river flow. Total discharge
into the Bay was high in March, increasing in April and decreasing
through the summer (Figure 5). The mean reflectances were moderate in
March and increased in April in the lower portions of the major
tributaries and the middle and upper Bay (Figure 6). The means in these
areas decreased into May-June. The James, which started with far higher
reflectances, showed continued decreases into July-August.
Tangier Sound and a turbidity zone along the Virginia western shore
below the Rappahannock maintained moderate reflectances throughout the
period. As these two areas are not directly associated with any rivers,
other factors beside discharge would determine their turbidity levels.
CONCLUSIONS
The reflectance data collected from the AVHRR can show variations in the
turbidity or suspended sediment loads of the Bay. The imagery
identifies the positions of turbid plumes caused by high river flow and
reveals fronts and small scale temporal changes. Examination of the
series of images shows the effects of individual high flow events.
Moderate discharge can increase the turbidity of the estuarine portions
more than 60 km from the fall line.
As the estuaries may take several weeks to clear, the timing of these
events may be critical in determing the productivity of these portions
of the Bay system during the spring and early summer. Continued
analyses of these events, particularly in comparison with long-term
averages, could establish the frequency and importance of the high
turbidity events.
381
-------
.05 n
.04-
.03
.02-
.01 -
JAMES
POTOMAC
CHESAPEAKE
Figure 4. Variation in spatially averaged RD reflectances in the middle
Chesapeake Bay, lower Potomac, and lower James, 1 March - 28 May 1987.
382
-------
8000 n
CHESAPEAKE BAY - 1987
6000-
co
UJ
O 4000
o
co
o
2000-
APR
MAY
JUN
JUL
AUG
SEP
Figure 5. Mean monthly flow rate into the Chesapeake Bay in 1987.
•04 -i
MEAN REFLECTANCE
MARCH APRIL MAY-JUN
JAMES
UPPER BAY
POTOMAC
MID BAY
SEP
Figure 6. Monthly mean RJJ reflectances the lower James, lower Potomac,
upper and middle Bay, for March through September, 1987.
383
-------
REFERENCES
Brown, O.B.; Evans, R.H.; Brown, J.W.; Gordon, H.R.; Smith, R.C.; Baker,
K.S. phytoplankton blooming off the U.S. East Coast: A satellite
description. Science 229:163-166; 1985.
Everdale, F.G. Satellite Oceanography - Volume I: NOAA-n AVHRR Digital
Data. 63 p. 1986; Technical Memorandum NESDIS AISC 7, National
Oceanic and Atmospheric Administration, Washington, DC.
Kidwell, K.B. NOAA Polar Orbiter Data (TIROS-N, NOAA-6, NOAA-7, NOAA-8,
NOAA-9, NOAA-10) Users Guide. 1987; National Oceanic and Atmospheric
Administration, National Environmental Satellite, Data, and
Information Service, Washington, D.C. 20233.
Lauritson, L., Nelson, G.J., Porto, F.W. Data Extraction and
Calibration of TIROS-N/NOAA Radiometers. 1979; Technical Memorandum
NESS 107, with amendments, National Oceanic and Atmospheric
Administration. Washington, DC.
Strong, A.E.; McClain, E.P., Improved ocean surface temperatures from
space--comparisons with drifting buoys. Bull. American Meteorological
Society, 65:138-142; 1984.
Stumpf, R.P. Application of AVHRR Satellite Data to the Study of
Sediment and Chlorophyll in Turbid Coastal Water. 1987a, Technical
Memorandum NESDIS AISC 7, National Oceanic and Atmospheric
Administration, Washington, DC, 50 p.
Stumpf, R.P. Sediment transport in Chesapeake Bay during floods:
analysis using satellite and surface observations. Journal of Coastal
Research, 4(1):1-15; 1988a.
Stumpf, R.P., Remote sensing of suspended sediments in estuaries using
atmospheric and compositional corrections to AVHRR data: 1988b, Proc.
21st International Symposium on Remote Sensing of Environment, ERIM,
Ann Arbor, Michigan, October 26-30, 1987, in press.
Stumpf, R.P.; Tyler, M.A. Satellite detection of bloom and pigment
distributions in estuaries. Remote Sensing of Environment, in press;
1988.
384
-------
Understanding the Estuary: Advances in Chesapeake
Bay Research. Proceedings of a Conference. 29-31
March 1988. Baltimore, Maryland. Chesapeake Research
Consortium Publication 129. CBP/TRS 24/88.
Evaluation of Conowingo Reservoir Release for Controlling Salinity
in the Upper Chesapeake Bay
Bernard B. Hsieh
Maryland Department of Natural Resources
Tawes State Office Building
Annapolis, Maryland 21401
ABSTRACT
A dynamic system model with both control and predict phases
is developed to achieve the effect of controllable input on
an output variable combined with other uncontrollable inputs.
This system uses variable coefficients and a self-tuning
scheme, recursive computation algorithms, and small computa-
tional and storage requirements. Testing of present regula-
tions will evaluate alternative release policies for potential
use. This model will demonstrate the impacts of regulated
riverflow from the Conowingo Reservoir on salinity in a tidal
dynamic Chesapeake Bay system. Sub-tidal heights and local
wind stress are considered as the uncontrollable inputs in
this model. The effects on fishery spawning and drinking
water are evaluated by this system as well.
INTRODUCTION
Riverflow from the Susquehanna River is discharged into the
Chesapeake Bay after being regulated by a multi-objective
reservoir, the Conowingo Reservoir. The circulation patterns
of the Bay, typical of an estuary, are dominated mainly by
tidal fluctuations, seasonal freshwater discharge, and wind
385
-------
forces on the water surface. Maintaining water quality in the
Bay, especially salinity, is a major concern for operation of
this reservoir. The Upper Chesapeake Bay and Elk River are
important rockfish (Morone saxatilis) spawning areas.
Salinity levels favorable to rockfish larvae spawning and
growth are between 0.0 and 5.00 ppt. Their spawning period
usually occurs in the spring, which is classified as a high
flow season in this area. The salinity level (chloride) of
the Susquehanna mist remain 250 ppm or less to meet drinking
water standards for the city of Havre de Grace, downstream.
New release strategies can be implemented when the required
riverflow is determined. This estimate is derived by testing
specified salinity levels in certain critical areas. Recent
research (Hsieh 1987) indicates that the Susquehanna river-
flow, Chesapeake Bay sub-tidal signals and the cross-Bay
component of wind stress from the Delaware Bay mouth are the
most significant forcing functions that influence daily
salinity levels in the Elk River area. A multiple-input
control and prediction model with self-tuning scheme and
variable forgetting factor modules is used to evaluate this
water resources management problem. The results of this study
could be linked to a regional control center as one of the
sub-systems; In this way, all specialized control loops could
be used as a whole to address the management needs of the
entire river basin.
There are several advantages of using the system model
approach: When only select critical areas need to be evaluated
for environmental effects, a model which connects mechanism
sources can simplify the complexity of the problem; The system
model reduces data requirements for model calibration and
verification, and simulation costs for designing long-term
engineering plans; The self-tuning scheme can be used to
install a telemetry computer system for designing future
automatic control facilities.
The design for a tidal water quality control system is shown
in Figure 1. In this design, the system output, salinity, is
the function of three separate inputs. The controller, com-
bined with the cost function and system model, calculates the
required riverflow when a desired or set-point salinity is
proposed. System parameters and coefficients of the con-
troller are adjusted by a recursive algorithm for each time
step, after initial conditions are given. The system output
shows a strong non-linearity when compared with each system
input. Usually, a non-linear transfer function can be used to
simulate the approximate system coefficients. However, in a
noisy water environment, the order selection is problematic.
The higher-order system model is not practical for this
calculation, whereas the lower-order model results in less
accuracy. This problem can be solved by using recursive
processes. Kalman's estimation method includes variable
forgetting factor (weighting factor), and therefore enables
the parameter estimates to follow both gradual and sudden
386
-------
a
•i-i
s
60
C
• H
u
u
o
tj
3
4-1
CO
C
O
01
73 .
u
s c
O D
w C
CO O
•r-l M
1-1 •!-!
CO >
> C
01
•H W
C CO
•H 3
l-t 4J
CO 0)
C/5 O)
0)
VJ
00
557
-------
changes in the dynamic system (Fortescue et al. 1981). Use of
a variable forgetting factor with the correct choice of
information bounds can avoid difficulties associated with
constant exponential weighting of past data. Recursive
estimation methods are no longer required to generate an
optimal solution but the initial covariance can still be
accurately estimated using this process.
MUI1E1VAKLATE CCNUOL AND PREDICT MDDELS
Astrom (1970), Astrom and Wittenmark (1973), and Clarke and
Gawthrop (1975) introduced and developed a self-tuning regu-
lator or controller, which uses a stochastic difference
equation as a system identification algorithm, and a feedback
control law to minimize the variance of the output function.
The basic self-tuning controller structure has been expanded
and is now applicable to multivariate cases (Borisson 1975,
1979; Kbivo 1980; and Bayoumi et al. 1981). Ganendra (1978)
used a self-tuning controller with real-time control of
release from a river-regulating reservoir. Ganendra (1976)
also used a self-tuning predictor to develop a riverflow
forecasting model.
To satisfy alternative release policies and downstream water
quality control needs, a self-tuning predictor is used in
conjunction with a self-tuning controller to generate the
observed output (salinity). The generated observed output and
original uncontrollable inputs, coupled with an alternative
release policy, produce the optimal required riverflow through
the controller. The individual system designs for the self-
tuning controller and predictor are shown in figures 2a and
2b. A self-tuning controller with unknown parameters and
variable forgetting factor is used to define the estuary's
water quality for this multiple-input case. The governing
equations for this model are summarized as:
H(q-1)Tc(t-Kc)
eft) (1)
where q"1 = backward shift operator.
y = system output (salinity).
u = system input/control variable (riverflow).
Tc = sub-tidal heights from the Chesapeake Bay mouth.
WS = wind stress from the Delaware Bay mouth.
e = a sequence of independent random numbers (noise).
A = 1 + a^q"1 4 ... + anq~n
B = b0 + biq~l 4... +
C = 14 c^cfl + ... +
H = h0 4- h]_q + ...+
M — ni0 4- m-^q""''- 4- .. .4-
388
-------
Uncontrollable Input
1
Controtabto Input I System
Optimal
ControtaM*
Input
Model
8y«t«m Output
Self-Tuning
Feedback
Controller
Target Output
figure 2(a) Self-Tuning Feedback Control System
with Multiple Inputs.
389
-------
a
fj
a
•a
a -a
*• a
a >
"3 a
e a
— n
en a
XI
a
c
o
U -H>
3
> a
o c
2 •-•
C
O
o
U
•a
o
o *•
•a a
-------
KL = pure time delay between riverflow and salinity.
KC = pure time delay between tidal heights and
salinity.
^ = pure time delay between wind stress and salinity.
Using the above information and concepts of minimum square
error predictor and minimum cost control, we obtain the
control variable :
(2)
u(t) = -l/E(Gy(t)-Cw(t H
where E - BF + Cfl-q"1) = BQ + e^"1 + ... + e^"11 (3)
F = 1 + f^"1 + ...+ f!_£ q1'*! (4)
G — gg + g^q + ...+ 9h-iq (^)
To simplify the notation, data and parameter vectors are
introduced:
(6)
, w(tfkl-l),...,
w(t+Ki~n) .. .TC/t+ki-k,-;),...
(7)
...-cn,J10/ ...Jln/
where J1=HF and J2=MF.
The final required riverflow is computed using the formula:
*
u(t) = -l/eo(4>(t)0(t) - enu(t) - w(t+kx)) (8)
The predicted output for the self-tuning predictor is
computed using a similar procedure except that the miriimum
loss function is calculated by the predictor instead of
calculating the cost function with the controller. The new
parameter vector is presented as:
4>(t) = [-Y(t+K1-l)f...f-y(tfK1-n)/u(t),..,u(t-nfl),Tc(t+K1
,TC/t+K1-Kc-n+l) ,WS (t+Ki-I^),... ,WS (t+Ki-K^-n+1),
(9)
391
-------
where £(t) = prediction error at time t.
= prediction estimate at time t.
The predicted output at time step (t+K^) is:
=4>(t)fl(t) (10)
SMJIATICN EXPERIMENTS AND APPLICATIONS
A salinity station in the Elk River area near Town Point, MD
was selected as the study site. This location was chosen
because of the ease in expressing transport mechanisms from
several directions of forcing as well as the abundance of
existing data. The flow station is located downstream from
Conowingo Dam, Susquehanna River.
The subtidal height is calculated by removing the semi-
diurnal and diurnal components from hourly tidal records at
Hampton Road, Va. The wind data from Wilmington, Del. is
converted to local cross-Bay wind stresses. The daily mean
data set is obtained by subtracting several deterministic
components from the hourly readings and taking the daily
average. The model time step was selected as one day.
Estimation of coefficients for the self-tuning controller is
made by supplying an initial coefficient vector and initial
error covariance, pure delays, and the stabilizing factor.
However, these values can only be rough estimates since this
system is assumed to be unknown. The lower-order system is
used for simplicity. The sum of square errors and minimum
forgetting factor can be estimated by the degree of noise.
Usually calibration is done by assuming that the desired
salinity is equal to the observed salinity, then adjusting kj,
kc, and J^ until the most optimal and stable condition is
attained. The forgetting factor responds to the sensitivity
of the estimated error for each time step. During the
calibration process, a minimum value of forgetting factor is
needed to manage the most dynamic condition. The required
flow would nearly equal the measurement flow in an optimal
selection of these system parameters. The final selection for
this study is kj, V^, and k^ equal to 1, and equal to 0.20
(n=2) . The self-tuning predictor is calibrated by using the
sum of square errors as the single indicator.
Fishery Spawning and Drinking Water Concerns (Control Phase)
Data from the Upper Chesapeake Bay hydrographic survey
(MD-DNR) were used to estimate the difference between Havre de
Grace and Old Town Point when salinity information in the
Havre de Grace area was not available. This difference can be
used as an approximate basis for the simulation. Under this
392
-------
prior consideration, eight sets of simulation tests were
conducted. The target assumptions were: (1) Target values of
salinity in the Elk River were selected as 0.1 and 2.0 ppt,
during the spawning season; and (2) Target salinities in Elk
River were 2.0 and 4.0 ppt during the low flow season. The
second assumption will result in salinity levels of about 1.3
ppt and 2.5 ppt in the Havre de Grace area.
The difference between required flow and observed flow, is
evaluated for excess or deficient conditions. The daily
results of these simulation tests are averaged on a monthly
basis (Table 1) to examine how current release policies, (5000
cfs minimum release from 15th of April to 15th of September)
meet the needs for each month. The simulation shows that the
current policy provides adequate water quality for striped
bass spawning and larvae under the conditions during the
springs of 1982 and 1983. However, greater flows are required
during the fall season to meet drinking water standards,
especially during October when the policy is not in effect.
.Storage Reallocation Plans (Prediction Phase)
The prediction phase of the model evaluates the effects of
storage reallocation at the lower Susquehanna River basin on
the Upper Chesapeake Bay. Two plans at the mouth of the
Susquehanna River's USGS station in Marietta are provided by
the US Army Corps of Engineers (Baltimore District), both of
which meet the flow target of 5000 cfs and 7400 cfs. The flow
difference between each plan and the historical flow will be
used to add or subtract from the flow at the USGS Conowingo
Station. The new flow, considered as the new controllable
input, simulates salinity levels produced by the reservoir's
operation. The two uncontrollable inputs remain the same.
The corresponding salinity difference (original and simulated)
and flow difference (original and simulated) for each plan are
shown in figures 3a-3d. The maximum difference occurs in
October 1983 (0.4 ppt), when the mean salinity level is about
6.2 ppt. The effects of both storage reallocation plans are
insignificant. However, more salinity stations are required
to obtain more definitive conclusions.
OCNCIIJSIONS AND OMffiHES
(1) For a dynamic estuarine environment, a multivariate
control system with variable forgetting factor (control
and prediction phases) is capable of simulating sudden
changes in input and output functions. This scheme
eliminates non-linearity effects in a dynamic, noisy
system.
393
-------
41
M
C
CO
JS
o
M
dl
t*4
IM
• H
O
.
t>4
•U
•>
t-i
41
W
JD
O
(u
TJ
4)
b
•r4
tr
41
orf
.
CO
"O
41
>
U
4)
CO
.0
o
.
CO
u
41
U
Ifl
H
J3
U
C
a
u
:«
O O CO CM \O NO
"•• co NO r"* co r~*
O rx in NO CM -*
1 1 1
§00 co ON *j r^
r- m >o r- ON
NO en CO r~ ON CM
in — in o — en
en en vo in m
?-7-7
— •• ~~ ~-» — — ^t Q\
oo ao en en in in
CM CM ON ON — —
in m en en en en
in m CM CM m m
r^ r-. ON en co NO
en \o CO CM r-» CM
en ON CM en *o NO
m o in CM >» o
in in CM CM in in
— — -$• s^ en en
~~ •••• co oo oo oo
O o o o o o
O O O O O O
— o — o — o
O CM O CM O CM
l»4
.* u
u x a
o. a 3
< S ">
CM CM CM
00 CO CQ
ON ON ON
NO o ON o — en
in o in •" in NO
• •••••
—
1 — CM —
vo NO O O en en
-» ~r o o -* •*
m m o o CM CM
>» O m vo — en
o — —
•^ O CM vO CM en
NO in oo vo en —
••* •"• ON ON " • ""
r* p^ p^ r*- CM CM
— — >»-»•»•»
O O O O O O
O O O O O O
CM >y CM s» CM ^t
CL. u >
41 U O
W O Z
CM CM CM
CO 00 00
ON ON ON
w
00
rl
O
M
•
(K
•
U-4
>M
• H
a
•
lu
"
41
>
U
41
10
J3
O
•
lu
•e
41
U
• r4
&
^
_
co
T3
41
>
U
u
jO
O
.
C/5
iJ
4>
0
M
R
6-*
J=
i.
C
o
£
M
>i
en oo ^^ ON ON "•
CM o ~» m vo NO
CM r-» o r~ o oo
1 1 1
— r~ o NO ON —
in oo ON en o co
oo r— co — ON en
O vO O ON CM VO
oo in CM oo CM oo
CM en CM
1 1 1
m m CM CM co oo
\o vo ^y *^ ^y •*
CM CM CM CM CM CM
•a vtf — — en en
en en m m en en
«j r- — o ao co
r*» ON in in r^ -^
O NO *a en *^ en
r* en ••" r** en O
en en in -t en en
en en 10 NO en en
in m — — N B
a. to 3
< Z "5
en en en
co oo oo
ON ON ON
— o m co — o
-» m o CM m en ,
— en CM in ** NO
— — m CM i ;
i
— r~ en *o in •*
r> r^ — o en 00
—••*-» O st oo
CM CM IO 00 CM CM
NO r-* NO r*« o *^
1 en — — —
oo oo O O — —
vo NO -* -» oo oo
en en O O *o *o
m in r~ i-» CM CM
CM CM
•™ »» o o m CM
oo -*
41 U O
W O X
en en en
00 00 CO
ON ON ON
>-,
U
• H
C
•H
(J ^
O rd
M-l U>
CQ 4J
4J OJ
C OC
01 >-J
& fO
•H U
(-1
01 C
a. oi
x >
^ »iH
00
C
O CO
• H
•u C
CO 01
r-1 J=
3 3
•H i— 1
CO 01
"O
S o
0 6
t-t
U-l I—I *J
O 01
3 )-i w
O -u
^H C U3
t-l O -i-l
1-4 U
01 4-1
> Oi C
•H 4-1 -H
P CO O
•H CM
T3 M
01 CO C
l-i > 3
•H -H O
g." H
CT r-'
OJ 3 *-*
DC: S co
01
rH
4S
CO
H
394
-------
2000-
1500-
1000
500 4
I -500-
4
* -1000-
-1500-
-2000-
4-4
JUL82
JUL83
JAN84
Date
Figure 3(a) Flow change at Conowingo gauge (USGS)
due to 31,000 acre-feet maximum storage
plan(U.S. Army Crops of Engineers) at
Mare itta gauge.
s
I
d
0.2-1
0.1
0.0
a
n
-0.1-
-0.2-
—A—viA-
I
JUL82
JAN83
JUL83
JAM84
Date
Figure 3(b) Simulated salinity change due to flow
variation of 3(a).
395
-------
500(M
I
o
w
d -
a
11
5000-
-10000:
-15000-
JUL82
4AN83
JUL83
J/W84
S
a
0.54
0.3
0.1
U.
a
w
2 -0.3
-0.5
Date
Figure 3(c) Flow change at Conowingo gauge (USGS)
due to 280,000 acre-feet maximum storage
plan(U.S. Army Crops of Engineers) at
Mareitta gauge.
JAMS 3
JUL83
JANS4
Date
Figure 3(d) Simulated salinity change due to flow
variation of 3(c).
396
-------
(2) The self-tuning scheme is potentially useful for
telemetry systems. This system can be linked with
multi-objective reservoir operation systems to
optimally utilize water resources.
(3) The pure delay between the output function and the
control variable is the most critical factor affecting
the stability of this system. Proper selection of the
minimum forgetting factor is related to the system's
stability when sudden changes in input variables
occur.
FURfflER GGNSZEERATICNS AND KPPLLCKTICNS
(1) A regulation evaluation system is currently being
conducted by coupling the control phase with the
prediction phase. A reservoir management model using
reliability programming methods can determine an
optimal solution and the highest reliability for the
entire system.
(2) Target stations other than Town Point are needed
to verify the model and examine the Upper Bay
variations before a final decision is made.
(3) With some modification, this model can be used for
water quality and target flow control systems.
ACKNOWLEDGEMENTS
Support from the Water Resources Administration and the
Tidewater Administration for this project are sincerely
acknowledged.
397
-------
Astrom, K.J. Introduction to stochastic theory. Academic
Press, N.Y. 1970.
Astrom, K.J.; Wittenmark, B. On self-tuning regulators.
Automatica, Vol 9:185-199, 1973.
Bayoumi, M.M.; Wong, K.Y.; El-Bagoury, M.A. A self-
tuning regulator for multivariable systems. Autonatica 17:
575-592, 1981.
Borisson, U. Self-tuning regulators-industrial application and
raultivariable theory. Report 7513, Dept. of Automatic
Control, 1975. Lund Inst. of Tech., Lund, Sweden
Borisson, U. Self-tuning regulators for a class of
multivariable systems. Automatica 15: 209-215, 1979.
Clarke, D.W.; Gawthrop, P.J. Self-tuning controller.
Proc, I.E.E., 1975. Vol 122, no.9:929-934.
Fortescue, T.R.; Kershenbaum, L.S.; Ydstie, B.E.
Implementation of self-tuning regulators with variable
forgetting factors. Automatica 17: 831-835, 1981.
Ganendra, T. A self-tuning predictor applied to river flow
forecasting. Real-time forecasting/control of water
resource system, 1976. Selected papers from an IIASA
Workshop, Eric F. Wood, Editor.
Ganendra, T. A self-tuning controller applied to river
regulation. Proceedings of modeling, identification and
control in environment systems, 1978. Vansteenskists,
Editor.
Hsieh, B.B. A system approach to analysis of water quality
due to transport mechanisms in the water environment.
Advances in water pollution control, 1987. Vol 3:109-118.
Pergamon Journals Ltd., London.
Roivo, H.N. A multivariable self-tuning controller.
Automatica 16: 351-366, 1980.
398
-------
Understanding the Estuary: Advances in Chesapeake
Bay Research. Proceedings of a Conference. 29-31
March 1988. Baltimore, Maryland. Chesapeake Research
Consortium Publication 129. CBP/TRS 24/88.
Controlled Energy Dissipation from River Inflow as a Factor
in Managing Estuarine Water Quality
T. A. Was tier
Office of Marine and Estuarine Protection
U.S. Environmental Protection Agency
Washington, DC 20460
INTRODUCTION
River inflows into estuarine systems may carry with them large
amounts of energy depending on their size and volume, and
the constant tidal action provides an additional input of energy
carried by the rhythmic movement of water into and within an es-
tuary. The potential and kinetic energy of rivers is often
tapped for hydroelectric power during theJr passages toward the
sea; and where tidal range and velocity are sufficient, the
energy of tidal action has been used to generate electrical power.
As the flow from a river and tidal flow from the sea meet in an
estuary they usually are moving in different directions, with the
kinetic energy of their passage also moving in opposite direc-
tions. Upon the meeting of two such inflows the kinetic energy
is rapidly dissipated, sometimes with dramatic visible effects.
For example, with an incoming tide there may be a zone of
significant wave action, i.e., a "tide rip", over a limited area
during a time of very light winds. Conversely, on an ebb tide
there may be a zone of very smooth water with a clearly visible
line of demarcation between two water bodies as the kinetic
energy from river and tide move smoothly along together. In
either case the kinetic energy from both sources is dissipated in
an uncontrolled manner with effects generally on the mixing char-
acteristics of the system, particularly on stratification and
resuspension of sediments. 599
-------
If the dissipation of this energy in the Chesapeake Bay could be
controlled in some manner, it might very well be possible to con-
trol the extent of stratification in parts of the Bay so as to
enhance the aquatic environment for JivJng resources management
and public health. This investigation is a preliminary investiga-
tion of the technical feasibility of doing this in the Chesapeake
Bay. It does not address the broad issues of administrative
feasibility, social desirability, and Jegal responsibility, nor
the even broader scientific issue of whether or not we know
enough about the physical, chemical, and biological processes of
the Bay to be able to do this in a responsible manner.
The specific questions to be addressed in this study are:
1. Has a significant change in estuarine stratification and water
quality resulting from a river Jnflow change ever been
documented?
2. What physical conditions would be necessary to do this in the
Chesapeake Bay and do they presently exJst?
3. Can the propagation of kinetic energy and its dissipation in
the Bay be demonstrated?
4. Is it feasible to develop a mathematical description of the
kinetic energy flow in the Bay sufficiently quantitative to serve
as the basis for regulating river discharges into the Bay?
5. Since such a management scheme would involve real time
control of river flows into the Bay, what information on
conditions in the Bay would be needed as a basis for using such
an approach?
G. Is this approach worth pursuing further, and, if so, what
must be done?
The following sections of this paper address each of these
questions in order.
HAS A SIGNIFICANT CHANGE IN ESTUARINE STRATIFICATION AND WATER
QUALITY RESULTING FROM A RIVER INFLOW CHANGE EVER BEEN
DOCUMENTED?
In the early 1940's at Charleston,S.C., a hydroelectric dam wets
built near the head of tidewater on the Cooper River, the major
river input to Charleston Harbor. To feed this dam, the flow of
the Santee River was diverted to the power pool behind the
structure. The net result was an increase in mean river flow to
Charleston Harbor from 300 cfs to 15,000 cfs. The first notice-
able effect of this change was an increase in the cost of dredg-
ing from $400,000 per year to $4,000,000 per year over a period
of several years; even at this cost it was impossible to keep
any slips perpendicular to the main channel clear.
A subsequent investiga,tion(Wastler and Walter 19G9) showed that
the increased river flow had changed the main part of the Harbor
from an unstratified to a salt wedge estuarine circulation
system. This study concluded that reduction of the mean flow to
less than 8000 cfs would break the stratification and reduce the
input of sediment to the Harbor. The flow of the Santee was in
-------
part rediverted to its original waterway to reduce the mean flow
to these levels, and the stratification was indeed broken.
The Harbor at Charleston is once again a natural deepwater system
with a significant reduction in annual dredging costs. Sedimen-
tation was reduced to nearly its original levels and other as-
pects of water quality were also improved, in large part due to
considerable improvements in waste treatment, which were also re-
commended because of the tremendous decrease in flushing time
that would occur with the breaking of the salt wedge.
This example involved only an overall reduction of mean river
input into the system, and no effort was made to control the
river input in synchronization with tidal flow. However, this
example does demonstrate that dramatic changes can be caused by
changes in stratification resulting from changes in river flow and
that these changes could be quantified and predicted, at least in
the example given.
WHAT PHYSICAL CONDITIONS WOULD BE NECESSARY TO DO THIS IN
THE CHESAPEAKE BAY AND DO THEY PRESENTLY EXIST?
In the Charleston Harbor situation there was a single major
river inflow into the Bay and it clearly was a primary factor
in controlling the extent of stratification. There was also
a dam at the head of tidewater which had the capability of
regulating the river discharge into the estuary. While it might
be possible to visualize other physical conditions with which
it might be possible to regulate river flow so as to alter strat-
ification, is clear that these conditions are suffucient.
In the Chesapeake Bay system, the major single river inflow is
the Susquehanna, which provides about 87 percent of the riverine
input to the Bay above the confluence of the Potomac and 50 per-
cent of the entire riverine input to the entire Bay. Conowingo
Dam is about 9 miles above the head of tidewater and receives the
flow from over 90 percent of the Susquehanna drainage basin. The
pool behind Conowingo Dam is not sufficient for long-term storage
but the dams operated in the Basin by the Corps of Engineers do
have a total amount of storage sufficient to supply a minimum
flow of about 30,000 cfs over a several month period with proper
routing. Thus, the physical structure necessary to provide river
flow regulation does exist at the head of tidewater.
The question of whether or not the Susquehanna does have a suffi-
cient impact on stratification of the main stem of the Bay to al-
low effective regulation, particularly at low flows, must also be
addressed. To do this, the results of the main stem monitoring
program were examined in regard to the extent of stratification
of the main stem throughout the year arid how well this correlated
with the annual cycle of river flow. Table I exhibits the ratios
of salinity in the Main Stem of the Bay for the full period of
reported data for the present monitoring program, i.e., June 1984
to July 1987, at approximate mile points closest to the routine
sampling stations. The geographical area covered is from Havre de
-------
Date
Table I
Salinity Ratios in the Main Stem of the Chesapeake Bay
(June 1984 - July 1987)
Miles from Havre de Grace
20
1984
June
July
Aug
Sept
Oct
Nov
Dec
1985
Jan
Feb
Mar
Apr
May
June
July
Aug
Sept
Oct
Nov
Dec
198G
Jan
Feb
Mar
Apr
May
June
July
Aug
Sept
Oct
Nov
Dec
1987
Jan
Feb
Mar
Apr
May
June
July
0,
0,
0,
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
.94
,41
-
-
-
.79
.52
.78
.70
.85*
.00
.79
-
.42
-
-
-
-
-
.71
.92
.75
.07*
-
.09
-
.77
.08
.45
.08
.45
-
30
0.
0.
0.
0.
-
0.
0.
0.
0.
0.
-
0.
0.
-
0.
-
0.
0.
0.
0.
0.
0.
0.
-
0.
0.
0.
0.
0.
0.
0.
22
22
97
94
40
20
53
GO
59
82*
80
09
G2
20
29
42*
52
03
09
70
47
-
74
70
-
G9
,78*
02
40
50 GO
70
80 90 100
0.42 - 0.03 - 0.05 - 0.78
0.30 0.30 0.40 0.45 0.49* - 0.57
0.33 0.38 0.03 0.02 0.41* 0.38
0.08* 0.70* 0.81 0.74 0.71 0.71 0.84*
0.82* 0.82* 0.72 0.08 0.78* 0.70 0.09
O.GG 0.74* - - 0.74* 0.78* 0.83*
0.75* 0.79
0.83 - - - 0.85* 0.88 0.85
-
0.
0.
0.
0.
0.
0.
0.
-
-
0.
-
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
02
52*
71*
70*
01*
07*
83*
03
47
24
25*
51
50*
08
GO*
88*
07*
-
58
-
81
-
24
GG
0.00
0.
72
-
O.G7' 0
0.54 0
0.08 0
0.07 0
0.05 0
0.72* 0
0.82* 0
0.90 0
0
0.80
-
0.58*
0.45*
0.43*
0.52*
0.07*
0.07
O.G8
0.85
0.74*
-
0.03
-
0.79
0.74
0.57
0.72
0.73
0.03
0
.00 0
.02* 0
.07* 0
.70 0
.75 0
.70 0
.74 0
.80 0
.91
0
-
-
0.08
0.41
O.G7
O.G1
O.GO
0.08*
0.71
0.72
0.8G
-
0.70
-
0.78
0.04*
O.G2
0.09*
0.71
.84
.82 0.70
.01*0.08*
.78*0.05
.72 O.G8
.08*0.75*
.07*0.70
.78 0.77
.88 0.81
0.8G
.74* -
0.84
0.51* -
0.52 0.72
0.57 0.55*
0.5G*O.G8
0.70*0.07
O.G9 0.04
0.70*0.70
0.79*0.74
0.74 0.75*
- 0.85
0.71
- 0.82
0.79
O.G8 0.77
0.51*0.80
0.70 O.GG*
0.79 0.09
0.04 0.78
-
0.82* 0
0.74* 0
0.77* 0
0.70* 0
0.70* 0
0.71* 0
0.70 0
0.92 0
0.8G
0
-
O.G8
0.70
0.57*
O.GO
0.70
0.09
0.80
0.88
0.74
-
0.80
-
0.79
O.G7
0.70
0.71
0 . 75
0.08
-
.85
.08
.82
.79
.78-*
.70*
.75
.99
-
.84
-
0.73
0.09*
O.G3
0.02*
0.89
0.07*
0.78
0.80
0.74
-
o.ai
-
O.G4
0.72
O.G5
0.73
0.78
-
Data Source: Chesapeake Bay Program monitoring data
-------
Grace to the mouth of the Potomac, the reach of the Bay in which
the Susquehanna accounts for nearly 90 percent of the river flow
entering the Bay. Data are presented on a monthly basis and each
value represents only one, or at the most an average of two sam-
ples .
There are several striking features of the data presented in
Table I as far as the impact of the Susquehanna flow on strati-
fication is concerned. First, the data show absolutely no evidence
of any type of seaonal overturn and redevelopment. There are
changes in the degree of stratification during a year, but these
do not appear to be related to any kind of seasonal regime, nor
to any obvious environmental scenario. Stratification in 1984 ap-
pears to be stronger than in either 1985 or 198G; however, there
is less than half a year of record in 1984, so it is difficult to
make an equitable comparison. Second, the change of the degree of
stratification down the Bay from Havre de Grace does not appear
to to be very great for a specific year or season, which suggests
that whatever forces are causing the degree of stratification ob-
served persist far down the Bay. Third, the degree of stratfica-
tion is extreme throughout the Bay and for much of the year. At
salinity ratios of around O.G, Charleston Harbor behaved as a
salt wedge estuary. However, the size and structure of the Chesa-
peake are different from Charleston, and it is not known whether
the use of salinity ratio as a surrogate for estuarine behavior
is valid. Fourth, the asterisks (*} by certain values in the
table indicate times and places where the Bay exhibited more than
one pycnocline. These are quite common in the data set, and sug-
gest the existence of a consistent phenomenon responsible for the
condition.
Table II presents some data which may offer an insight into the
reasons for the existence of multiple pycnoclines in the Bay as
well as indicate a source of some of the organic material respon-
sible for the depleted oxygen levels in the bottom layer of the
water column. This Table compares water column densities in the
Potomac near its mouth with those in the Bay itself above, at,and
below the confluence of the Potomac with the Bay. For ease in
reading the density values have been adjusted so that they appear
in the table as small numbers; however, what is significant are
their relative values, and these are unchanged. The data
presented are for the entire year of 1985, and what is important
is to note the relative densities of the Potomac water and those
of the Bay at and close to the confluence.
Remember that water(e.g., Potomac water) tends to ride over other
water it meets of higher density(e.g..usually Bay bottom water)
and slide under water it meets of lesser density(e.g., usually
Bay surface water). The data show that in nearly all cases the
surface water in the Potomac would tend to ride out over or mix
with the surface waters of the Bay, although on two occasions,
3/19/85 and 5/G/85 the relative densities were such that the
Potomac water would tend to slide under the surface water of the
Bay. If these conditions persisted for any length of time, a
significant amount of Potomac surface water could be introduced
-------
Table II
Comparison of Water Densities at the Mouth of the Potomac River
with Those in the Main Stem of .ae Chesapeake Bay Above and
Below the Potomac Confluence*
Date
1/14/85 U
M
L
2/11/85 U
M
L
3/4/85 U
M
L
3/19/85 U
M
L
4/8/85 U
M
L
4/22/85 U
M
L
5/G/85 U
M
L
5/20/85 U
M
L
G/3/85 U
M
L
G/17/85 U
M
L
Bay above Potomac
Potomac Mouth
(CBS. 2) (LE2.3)
1.278
1.45G
1.391
-
1.515
1.09G
-
1.455
1.171
1.288
1.435
1.155
1.44G
1.708
0.9G9
1.119
1.392
0.978
1.189
1.5G3
1.025
1.150
1.394
0.979
-
1.504
1.007
1.175
1.434
1.254
1.444
1.265
-
1.451
0.977
1.08G
1.204
1.21G
-
1.411
1.109
1.2GO
1.488
0.95G
1.090
1.28G
1.044
1.321
1.511
0.898
LOGS
1.3G4
0.9G7
-
1.273
0.882
1.042
1.2GO
Bay at
Confluence
(CB5.3)
1.320
1.552
1.457
-
1 . 585
l.OOG
-
1.45G
1.241
-
1.451
1.198
1.G2G
1.85G
0 . 95G
1.227
1.5G9
1.053
1.389
1.513
1.15G
-
1.450
1.040
1.3G5
1.616
1.054
-
1.402
Bay belc
Potomac
(CBS. 4)
1.350
1.429
1.490
1.320
-
1.4G7
1.145
-
1.308
X
X
X
X
X
X
1.059
1.305
1.513
1.092
1.449
1.G91
1.244
1.316
1.502
0.973
1.333
1.5G1
1.082
1.248
1.434
-------
Table II (Continued)
Date
7/8/85 U
M
L
7/22/85 U
M
L
8/0/85 U
M
L
8/19/85 U
M
L
9/9/85 U
M
L
9/23/85 U
M
L
10/7/85 U
M
L
11/12/85U
M
L
* Values In
readable
U - upper
M - middle
L - bottom
x - no data
Bay above Potomac
Potomac Mouth
(CBS. 2)
0.991
1.35G
1.611
0.97G
1.418
1.G09
1.03G
-
1.3G9
1.034
1.399
1.G05
1.003
1.240
1.429
1.220
1.429
1.GG9
1.347
-
l.GGO
1.422
-
1.733
table are
numbers .
layer
layer ( -
layer
(LE2.3)
0.991
-
1.284
0.939
-
1.301
1.034
-
1.159
1.012
-
1.193
0.879
-
1.487
1.211
-
1.300
1.287
-
1.340
0.852
-
1.4G4
(Density x
means there
Bay at
Confluence
(CB5.3)
1.034
-
1.727
1.03G
1.102
1.4G5
1.112
-
1.419
1.099
1.25G
1.559
0.82G
1.07G
1.389
1.2G7
-
1.7G7
1.320
-
1.G49
1.044
1.247
1.G27
1000) - 1000,
Is no middle
Bay below
Potomac
(CBS. 4)
1.054
-
1.58G
1.220
-
1.550
1.215
-
1.5G7
1.105
-
1.284
1.003.
-
1.432
x
X
X
1.38G
-
1.G39
x
X
X
which gives more
layer)
-------
into the bottom waters of the Bay where it would decay without
benefit of reaeration, thus causing possibly severe oxygen deple-
tion. For most of the year it appears that the surface layers
of the Potomac tend to end up in the surface waters of the Bay
but that a midlayer representing some kind of blending of Potomac
bottom water and Bay bottom water may very well form in the Main
Stem of the Bay. The small but significant differences between
the densities of the bottom and midlayers in the Bay suggest the
existence of such a mechanism.
This look at the salinity structure as shown by the monitoring
data suggests that there is • no obvious, direct relationship
between the Susquehanna River flow and stratification in the Bay.
A more detailed look at the salinities of the bottom layer
indicated that these did tend to vary consistently with river
discharge, but a quantitative relationship could not be estab-
lished with the data available. Nevertheless, the wide range of
values of salinity ratios in no coherent pattern over a year sug-
gests that the Bay responds quite rapidly to changes in river
discharge and that samples taken a minimum of two weeks apart are
not adequate to establish the nature of the response. More de-
tailed measurements are needed to resolve this problem.
CAN THE PROPAGATION OF KINETIC ENERGY AND ITS DISSIPATION IN
THE BAY BE DEMONSTRATED?
The quantity of kinetic energy in a flowing body of water is a
function of its mass and the square of its velocity. In this
case there are two bodies of water involved, that associated with
the river discharge into the Bay and that associated with the
tidal flow. Since water behaves as an incompressible fluid, the
interaction of the two bodies should be measurable as a
perturbation of the observed tide height by the river discharge.
The methods described by Wastler(Wastler 19G9) and used by
Wastler and Walter in the Charleston Harbor analysis allow the
estimation of the energy present at the dominant periods
exhibited in each record at each sampling point. These methods
involve the calculation of cross-spectra from river discharge and
tide height records and the interpretation of the results in
terms of the physical parameters involved. The effect of the pre-
sence of some potential energy due to the slope of the estuary
from its head toward the sea is eliminated by using only
deviations from the mean tide height at each tide gage; the
potential energy of the water would, however, be small in
relation to the kinetic energy in any event.
For the present case records of hourly discharges from Conowingo
Dam were available for the first nine months of 1983. These were
run against hourly tide gage records for the same period at Havre
de Grace, Matapeake, and Solomons, affording a coverage from the
head of tidewater nearly to the mouth of the Potomac. A range of
river discharges based on daily means from 5000 cfs to 103,000
cfs were covered by these records, thus incorporating an excellent
range of flows from the Susguehanna. The river discharge records
had an extremely strong diurnal component typical of hydroelectric
-------
dam operation. The analysis showed that the propagation of ener-
gy from the river discharge was extremely rapid and consistent at
all river disacharges. This amounted to 44 hours to the Solomons
gage, or roughly two days from Havre de Grace to the mouth of the
Potomac. This is, of course, the speed of energy propagation, not
the time of water travel. It shou]d be noted that the rapidity
of energy propagation from the river flow could have some bearing
on the similarity of the degree of stratification down the Bay,
as shown in Table I. The diurnal component of river flow was
still present at Solomons with about 25 percent of the original
energy still present. Long-period components of the river
discharge (i.e., those of periods greater than 7 days) were not
large in the river record, and these were not distinguishable in
the records at Solomons. This suggests that the long-period energy
of the river discharge is dissipated in the upper part of the
Bay.
These results indicate that energy from the Susquehanna flow
makes its way at least halfway down the Bay in a quantifiable
form, that it tends to be distributed throughout the upper part
of the Bay in its presently uncontrolled regime, and that its
time of propagation is quite rapid. Thus it appears that there
is enough energy available to use if there is a means of
controlling it.
IS IT FEASIBLE TO DEVELOP A MATHEMATICAL DESCRIPTION OF THE
KINETIC ENERGY FLOW INTO THE BAY SUFFICIENTLY QUANTITATIVE TO
SERVE AS THE BASIS FOR REGULATING RIVER DISCHARGES INTO THE BAY?
Kinetic energy balance equations for estuarine flow and for river
discharges are standard .textbook exercises, but apparently no one
has ever put the two togather. As part of this investigation a
synthesis of the two was attempted, not to develop an exact solu-
tion, but to establish the nature of the equation and the para-
meters necessary for its solution. A harmonically varying river
discharge was introduced to determine its effect.
The results indicated that a mathematical description of
the process is feasible, but there are some problems. First,
there appear to be a large number of acceptable mathematical
solutions depending on a number of unknown quantities which can
only be measured in the field. That is, it would be necessary to
experimentally vary the discharge from Conowingo in a preset pat-
tern and measure what happens. Second, with opposing harmonic
flow patterns, there is a real possibility of a feedback process
developing with the formation of a disastrously large standing
wave in the Bay. Again, this m-ay be a mathematical artifact, and
field investigation may demonstrate that such is the case.
In short, yes, it is feasible to develop a quantitative mathemat-
ical description of what would be required, but it must be
based on accurate field studies in its development stage, and it
must be carefully and thoroughly evaluated in the field before it
is used in an operational mode.
-------
SINCE SUCH A MANAGEMENT SCHEME WOULD INVOLVE REAL TIME CONTROL
OF RIVER FLOWS INTO THE BAY, WHAT INFORMATION ON CONDITIONS IN
THE BAY WOULD BE NEEDED AS A BASIS FOR USING SUCH AN APPROACH?
Before such a management scheme could be implemented, it would
be necessary to have a very clear understanding of what changes
in river discharge would do to the salinity structure of the Bay.
It would also be necessary to understand what salinity structure
was needed at specific points in the Bay to provide the suitable
degree of stratification to protect and enhance the habitat for
living resources and to maintain a predetermined level of water
quality. With these considerations in mind, it can then be
stated that high-frequency real time measurements of salinity are
the only water quality information really necessary. However,
the only cost effective technique to obtain this type of
information presently available is automatic remote sensing
buoys. The major cost of such devices is in the buoy itself and
in the data transmission system, so it would be reasonable to
collect other useful information such as Temperature, Dissolved
Oxygen, Turbidity, pH, and other parameters of interest to
management authorities and responsible investigators.
IS THIS APPROACH WORTH PURSUING FURTHER, AND,IF SO, WHAT MUST
BE DONE?
This paper has emphasized the unknowns and the problems involved
in developing and implementing a management approach based on
flow regulation for positive control of the aquatic environment.
These problems and unknowns are real in the context of this
management approach certainly; but they are no less real in the
context of any other management strategy.
A major problem identified has been that of establishing a firm
relationship between river flow and stratification. Certainly it
is necessary to know this to differentiate between "wet years"
and "dry years" in any quantitative sense for any type of model.
Establishing such a relationship is a first and major step in any
further investigations, whether of this management approach or
any other approach toward water quality control.
For example, it would be impossible to develop a complex water
quality model of the Bay without accounting for the daily pattern
of flows out of Conowingo, the effect of such variations in flow
on water quality, the development of multiple stratification
layers in the Bay, and the occasional underflow of the Potomac
into these layers. The development of any such model would go
far toward evaluating the feasibility of controlled river flow
regulation as a management tool.
This approach should be examined further, not as a unique and
separate entity, but as part of the total array of management
techniques that might be feasible.
-------
BIBLIOGRAPHY
Wastler,T.A. Spectral analysis: applications in water pollution
control. Washington,DC: U.S. Department of the Interior; 19G9.
Wastler,T.A.; Walter.C.M. Statistical approach to estuarine
behavior. J. of the Sanitary Engineering Division, ASCE. 94:
No.SA C, Proc. Paper G311, 1175-1194; 19G8.
-------
Understanding the Estuary: Advances in Chesapeake
Bay Research. Proceedings of 'a Conference. 29-31
March 1988. Baltimore, Maryland. Chesapeake Research
Consortium Publication 129. CBP/TRS 24/88.
Application of a Multiple Linear System to Identify Tidal Signals
in the Upper Chesapeake Bay
Bernard B. Hsieh
Maryland Department of Natural Resources
Tawes State Office Building, B-3
Annapolis, Maryland 21401
ABSTRACT
The Upper Chesapeake Bay has a dynamic environment of physical
characteristics due in part to incoming tidal fluctuations from the
Delaware Bay via the C & D Canal, and the mouth of the Chesapeake Bay.
The tidal propagation signals from the two bay systems are a function
of most of the Upper Bay area because of the difference in distance
traveled from the mouth. The high volume flow from the Susguehanna
River might result in some water level change as well. Although local
effects are also important factors, this study focuses on non-local
sources of impact.
In order to understand energy transport due to tidal forcing and
combined river runoff, it is useful to utilize a multiple transfer
function model which can be used to indicate the dominant forcing
factors and discern the tidal direction for specified areas. Three
important frequency bands—semi-diurnal, diurnal, and 3-20 days, are
used to perform this investigation from hourly sea level readings.
The model's structure is defined by river flow and tidal signals
from both bays as inputs, and one Upper Bay location as an output.
INTRCfflOCTION
Studying tidal propagation phenomena is probably the most reveal-
ing means of understanding the dynamic behavior of water movement in
the Chesapeake Bay. The Upper Chesapeake Bay receives tidal fluctu-
ation signals from the Delaware Bay through the C & D Canal, and from
410
-------
the mouth of the Chesapeake Bay. Dronkers (1972) indicated that the
propagation of a tide is modified by its degree of penetration into
the continental shelves, coastal areas, and estuaries. The distortion
of the tidal wave takes place at this point. This change is primarily
caused by : the friction of the bottom, changes in tidal velocity due
to variation in depth, and the shape of the estuary. Tidal prediction
is primarily employed at harbors situated at coasts where tidal re-
cords include various impacts resulting from meteorological effects.
Tidal prediction is difficult due to variation in the amount of river
runoff and additional disturbances which may change the water level
considerably. Abnormal weather changes such as tsunami may generate
high tidal waves or intensive rainfall, which result in great amounts
of river runoff. This directly affects the properties along the
coast. Rather than focusing on single-point tidal prediction, this
study uses non-local effects such as tidal signals from the two bay
systems and the Susquehanna River flow, as the basic mechanisms for
determining the sea-level change for several selected locations
Reedy Point, which is located at the east end of the C & D Canal, is
an interesting area for researchers because of the possibility of
recording any signals from the Chesapeake Bay which are transported
there. Havre de Grace, near the mouth of the Susquehanna River, and
the Baltimore Harbor, the most important commercial port in the Bay,
are included in this study as well. Two significant questions are
addressed in this investigation : (i) Which bay contributes more
tidal signals to the Upper Bay area ? (ii) Do any river runoff
signals travel from the Susquehanna solely to the Chesapeake Bay ?
This approach is characteristic of multiple inputs/single output
systems. The multiple transfer function modeling technique, which is
based on the frequency domain process, produces multiple frequency
response functions. These functions use a Fourier transformation
which drives this system from time-domain to frequency-domain, and
simultaneously estimates their partial response factor and calculates
the multiple coherence. A recent study (Hsieh 1985), showed that the
partial multiple coherence is a powerful technique to use for separ-
ating the relative magnitude of each factor. This technique, due to
linear relationships in the system, represents only a fraction of
output power at any given frequency. The difficulties in computing
the inverses of complex matrices when a multiple linear system is
applied, are successfully solved by the Gaussian elimination method
and the partitioning approach.
TIDAL CCNSTITUEKC5 FOR SELECTED STATIONS
Data containing hourly observations of sea-level elevation for 1983
were obtained from five stations of the National Ocean Service. Each
time series, containing 8,760 records, was linearly detrended prior to
cyclical analysis. Every tidal record indicates a very weak linear
trend. The amplitude and phase angle can be computed for any desired
period by conducting harmonic analysis. However, this technique fails
to show the importance of phenomena for which the period does not ex-
actly coincide with a specified harmonic. One method to reduce this
disadvantage is to use very small frequency bands in the analysis.
Budgell (1981) stated that the main harmonic component in the tide
will tend to be semi-diurnal due to the two bulges created by tidal
411
-------
forces. The main lunar and solar semi-diurnal constituents that
correspond with this phenomena are the tidal components M2 and S2.
When the tide enters the area over a continental shelf and penetrates
into an estuary, its propagation is effected by the reduced depth. In
mathematical terminology, this means that non-linear terms in the
hydrodynamic equation will produce new harmonics which are a combina-
tion of the two original frequencies. This study will examine the
tidal constituents for four categories: (1) semi-diurnal components,
(2) diurnal components, (3) shallow water components, and (4) 3-20
days components. This method of classification covers the tidal
period from two hours (M12) to about 350 hours (Msf).
An investigation was made by removing M2 tides from each tidal
record. A subtidal fluctuation typical of Reedy Point, Delaware, is
shown in Figure 1. The rest of the stations reveal less subtidal
fluctuations (Hsieh 1985). One useful technique in evaluating the
fluctuation of the tide is by using a low pass filter to show its
longer term variations. The longer term fluctuations for five tidal
stations are shown in Figure 2. The water level at Havre de Grace is
the highest while the lowest is found at Reedy Point. The trend com-
ponent for each station is small enough to disregard. The freshwater
inflow at Conowingo Dam accounts for approximately 50 percent of the
total input of freshwater entering the Bay, with more than 85 percent
of the freshwater entering the Bay above the mouth of the Potomac
(Schubel 1972). In 1983, the hourly range was between 200 cfs and
90,000 cfs. Figure 2 suggests a strong seasonality in freshwater
inflow and approximates an eight-months cycle.
To determine the basic structure of each of the tidal records,
harmonic analysis of the above-mentioned categories is used. None of
the significant components for shallow water constituents are found in
the frequency band between a 2-hour and 6-hour period. Significant
tidal components for semi-diurnal and diurnal periods are shown in
Table 1. The M2 tide accounted for more than 75 percent of total
variance at Reedy Point, but only about 20 percent of the total vari-
ance at the Baltimore station. A low percentage for the M2 tide was
also found at Havre de Grace. The low percentage and small amplitude
of the M2 tide could be attributed to its loss of tidal wave energy
from the Chesapeake Bay entrance, due to the long distance traveled
and the complex geometry of the Bay. Therefore, the local effects and
low variation of the tide in the Bay's inner area are important
factors to consider. However, the 60 percent of total variance at
Baltimore results from longer than two days fluctuations. The
periodgram reveals that most of the Baltimore signals have inregular
variability. The station at Havre de Grace might receive some tidal
signals from the Delaware Bay due to the effect of standing waves at
the end of the estuary. The S2 component is generally more signifi-
cant than the N2 component, but this phenomenon does not apply to
these Upper Bay stations. In inner Bay stations such as Havre de Grace
and Baltimore, the S2 component is even less than the KL and 01
components.
FREQUENCY MDEEIP1E TRANSFER FUNCTION MDOEL
In order to determine the impulse response functions from the x(t)
and y(t) data in the frequency domain, it is necessary to transform
412
-------
o -
4 -
S
u
b 2 H
t
I
d 0 -
e
< -2-
t
»
) -4-1
-J
0 1000 2000 3000 4000 5000 0000 7000 8000 9000
Hour
Figure 1. Hourly subtidal height at Reedy Point, Del.
10-
9
a •
7 -
6 •
5-
4 -
3-
2 -
1 -
0-
A=" Harve De Grace Tide
B= Hampton Road Tide
C= Lewes Tide
D= Baltimore Tide
E= Reedy Point Tide
F= Conowingo Dam Flow
r I I 1 I
1000 2000 3000 4000 5000
Hour
0000 7000 8000 9000
o
w
<
L
0
G
Fighre 2. Low frequency variations of tidal hight and
freshwater inflow for selected Upper Bay
stations.
413
-------
ffO
ro^rnojoot-i ^ r-
•H^OCOOO corom
4J kl * . .
JJ £ JJJ o o o r- fH o m o (NQOO
O Qj O»-IOQO*-IO fNO»HO
OOOOOO OOC3O
[^•\oni/ioo*H inffii/">«H
(fl . • POmiT»»-fi-tf-( OOU5O
O«HOOvrnO «-tOOO
p-
r-rgocoOCTN c\oofN..
t.nmcN^'ino rsiofNo
O O O CN O O OOOO
r^Oo^tsivor*- ror^-fNr*.
0lOrHirv0GO C\OOOCO
rHjrs(fM n*irrtri_
IH »-4«Ht-f»-i»Ht-* psirsjfvjrM
CM (N i
• H (0
CO PQ
-------
the input and output functions from time-domain into frequency-domain.
The initial solution is obtained by using formulas 1-12 (Huthman 1978;
Enochson 1968) .
co
Y(t) = Jh(T)x(t -t) dT+ e(t) (1)
-66
disregarding e(t) we obtain :
/»
X(f) =Jx(t) e -D2Xft dt (2)
-00
Y(f) =.Jy(t) e -J2Kft dt (3)
and : -<*
H(f) =ph(t) e -D dt
the Fourier transform H(f) of the impulse response function h(t) is
called the frequency response function, which is :
H(f) = Y(f)/X(f) (5)
A model of a linear system responding to multiple inputs is developed
by expanding the above relationship. (Figure 3)
Y(f) = Z Yl(f) = £ Hly(f)Xl(f) (6)
1=1 1-1
This relationship can be expressed more concisely in matrix notation.
Refine a N-Dimensional frequency response function vector to obtain :
H(f) = (Hly(f), H2y(f) ..... HNy(f)) T (7)
Next, define a N-dimensional cross-spectrum vector of the output y(t)
with the input X(t) :
Gxy(f) = (Gly(f),Gzy(f ),...., GNy(f)) T (8)
Finally, define the N x N matrix of the power and cross spectra of
all the inputs X(t) by :
'cil(f) G12(f)..GlN(f)\
Gxx(f) =/ G21(f) .... (9)
• » » » • I
GNl(f) GN2(f)..GNN(f)/
Use the system of linear equations to obtain the least squares
solution for equation (6), which provides the matrix equation :
Gxy(f) = Gxx(f) .H(f)
The final solution to this calculation is :
H(f) = Gxx-^f) Gxy(f) (11)
The linear relationship of the multiple system can be expressed as
the multiple coherence function estimates for the discrete frequen-
cies by the equation :
415
-------
c:
cd
CO
ex
•H
...J
a
(U
1-1
IT) •
IU 4-1
.5 g.
ill
3
cx v
•H t-H
J-l 60
••-< c
3 -H
v
M
00
476
-------
Kyx2(f) = (Hl(f)*Gyl(f) +.. .+HN(f)*GyN(f \/Gyy(f) (12)
where : Gyx(f) = (Gyl(f),...,GyN(f)) T
Gyy(f) = output power spectrum
In order to separate the relative importance for each input on out-
put for a particular frequency band, use equation (12) to obtain
the partial multiple coherence.
Kyx2(f) = HGly + ... + HGNy (13)
where : HGly = Hl(f) *Gyl(f)/Gyy(f)
define : TGH = HGly2 + ... + HGNy2 (14)
to attain the partial multiple coherence for input 1 :
Kyxl(f) = HGly2 *Kyx2(f)/THG (15)
TIDAL SIGNALS MDCEL
The multiple transfer function model, based on a frequency domain,
can be applied to the tidal signals model. This model assumes that
sea-level changes at selected Upper Bay locations are multiple func-
tions of tidal propagation from Lewes, Delaware, Hampton Road, Va.,
and runoff from the Susquehanna River. This linear system is solved
by using the Gaussian elimination method and the partitioning tech-
nique to compute the inverse of matrix Gxx(f). Since the cross-
spectrum consists of a real part and an imaginary part, the condition
number of matrix Gxx(f) is an important factor in determining whether
or not the solution is accurate.
The Reedy Point station is selected as an example for demonstrating
this approach. The multiple response function is obtained by using
detrended tidal records and non-seasonal Susquehanna River flow.
Detrended tidal records are used instead of subtidal height fluctua-
tions because the greatest variation of tide is derived from the semi-
diurnal and diurnal components. The final partial multiple coherence
is able to demonstrate the dominant signal for each particular fre-
quency band. The frequency band which accounts for larger variance
is more significant than the others. For example, if the partial
multiple coherence of M2 tide for the Delaware signal at Reedy Point
has a greater variance, then the Delaware Bay contributes larger M2
fluctuations to the Upper Bay. Huthmann (1978) indicates that the
multiple coherence function will be unity over the entire frequency
range under ideal, noise-free conditions in which there is a true
linear relationship in the multiple input/single output system. The
multiple coherence function is less than unity in a natural system.
In this study, two data files are constructed. One file combines the
hourly tidal height with the subflow data by taking the log function
and removing the 8-month significant cyclic component. The second
file is constructed by daily means from the first data file. The
multiple coherence over semi-diurnal components at Reedy Point is
close to unity (Figure 4). Since the noises occur in part of the
frequency range, the smoothing curve is obtained by using a
417
-------
M 1.0
u
I
t o.a
i
P
I 0.6
e
C 0.4 •
o
h
e 0.2
r
e
n 0.0-
c
a .070
.074 .078 .083
Frequency (cph)
.087
T
.091
Figure 4. Multiple coherence over semi-diurnal
components at Reedy Point, Delaware.
1 .0
0.8-I
0.6
P
M
C 0.4
0.2
0.0
.070
A = Delaware Bay signals
B = Chesapeak Bay signals
C = Susquehanna River signals
,074 .078 .083
Frequency (cph)
.087
.091
Figure 5(a). Partial multiple coherence for semi-
diurnal tidal components at Reedy Point,
Delaware.
418
-------
1 .0
0.8 -
0.6
P
M
C 0.4
0.2 •
0.0-
.032
A = Delaware Bay signals
B = Chesapeake Bay signals
C = Susquehanna River si1
—i—y ...i.,.
,034 .030
.038 .039 .041
Frequency (cph)
.043 .045
I .0
0.8
0.6
P
M
C 0.4
0.2-
0.0-
Figure 5(b) . Partial multiple coherence for diurnal
tidal components at Reedy Point, Del.
A = Delaware Bay signals
B = Chesapeake Bay signals
C = Susquehanna River signals
IIIIilllfIT
0.00 0.05 0.10 0.15 0.20 0.25 0.30 0.35 0.40 0.45 0.50
Frequency (cpd)
Figure 5(c). Partial multiple coherence for daily
tidal components at Reedy Point, Del.
419
-------
30-frequency band moving averages method (for a total of 200 frequency
bands). The partial multiple coherence in Figures 5a-5c separates the
signals from the Delaware Bay, Chesapeake Bay and Susquehanna River
flows. The Susquehanna River signal is very weak compared to the
other two signals. The most significant peak in Figure 5a indicates
that the N2 tide (12.66 hr) from the Chesapeake Bay is more signifi-
cant than the Delaware signals. The frequency band is so close that
certain peaks are smoothed by insignificant neighboring frequency
bands. Therefore, the wider frequency range of the dominating source
is selected (Figure 5b). For diurnal components, the Reedy Point
sea-level is mainly dominated by the Delaware Bay signal except for
fluctuations around 27.5 hours, which are provided by the Susquehanna
River signal. For the daily mean set (Figure 5c), the Delaware signal
is the major source, but two frequency bands (3 days and 2.4 days) are
shared with the Chesapeake Bay signals.
DISCUSSION
The Havre de Grace and Baltimore Harbor stations receive the same
inputs. Data from these stations can therefore be used to compare the
variation of signals at different frequency bands due to the change in
location. Several iicportant findings are summarized below.
Semi-Diurnal Components
(i) Baltimore Harbor receives even higher tidal signals from the
Delaware Bay than does Havre de Grace. The Baltimore Harbor area also
has stronger K2 & M2 tidal components than the Reedy Point station.
(ii) Each station has its own higher constituents from the Chesa-
peake Bay signals : Reedy Point has a higher M2, Baltimore receives a
stronger M2, and Havre de Grace contains a higher L2 tidal component.
Diurnal Components
(i) The Baltimore Harbor area and the Havre de Grace station show
similarity in their patterns resulting from the Delaware Bay signals.
Reedy Point is the highest over the entire frequency range, but the
Baltimore station has a stronger Kl tide.
(ii) The Chesapeake Bay contributes very few diurnal components to
Reedy Point. Baltimore receives stronger signals except at the PI
tidal frequency band.
(iii) Havre de Grace receives stronger signals at 27.78 hours from
the Susquehanna River flow.
Daily Mean Component
(i) Baltimore and Havre de Grace receive similar tidal components
from the Delaware Bay signal propagation. Peaks are found at 2.5
days, 4.5 days, and 3 weeks.
(ii) Again, the Chesapeake Bay contributes the same slow variation
between Baltimore and Havre de Grace, but carries only a 3.5-day
signal to the Reedy Point station.
(iii) The Susquehanna River flow contributes about 3-day fluctua-
420
-------
tions both at Baltimore and Havre de Grace, but no significant signals
are found at Reedy Point.
It would be interesting to know if significant tidal components
would respond to these factors more precisely if the frequency band
was specified. The partial multiple coherence for three desired
categories is shown in Table 2. These values provide a relative index
to determine which signals contribute to a major tidal component.
The Susguehanna River responses appear to be a minor factor for every
important tidal component except for the 29-day period. This
phenomenon is also valid for Havre de Grace and Baltimore. A very
significant finding is that the Delaware Bay signals have the
greatest percentage of slow variations (Table 2). The Delaware Bay
signals also show their significance in the Ql and Kl components and
the Chesapeake signals dominate 12 tidal constituents. The partial
multiple coherence and amplitude (total variance) are two major
factors used to calculate the amount of variation attributable to
individual signals.
GCNdDSICNS
The partial multiple coherence technique, generated by taking
fractions of multiple coherence, is a very powerful tool to use in
describing the multiple linear system of tidal propagation signals
from the two bay systems. The Susguehanna River flow plays a minor
role in this case, except at 29-day period fluctuations. The Delaware
Bay signals dominate the slow variation of tidal behavior. The Balti-
more Harbor and Havre de Grace areas show the same pattern signals
over slow tidal fluctuations. For diurnal components, the Delaware
Bay signals are the controllers of the KL and Ql components. The O2
tidal signals from the Delaware Bay are not found in the Chesapeake
Bay. The Chesapeake Bay transports a large number of semi-diurnal
signals to the Delaware Bay. However, the Havre de Grace and
Baltimore Harbor stations detected several significant tidal signals
from the Delaware Bay. This phenomenon suggests that both bays
distribute their own signals for high frequency tidal transport.
421
-------
more
^J
&
)e Grace
0)
t-i
3
33
S
2
>.
in
3
OJ
S
w
r
S
2
10
5
s
•g
-H
S-
o o r^ o o (N
m o o o o o
o o o o o o
O 0 O 0 0 0
00 Gf\ VD *3* r-
in o *-t m o oo
r\i «r p- rg r-- o
o o o o o o
*r 1-4 in *3- «r rn
p* o^ in *r en ^o
m m CM r* oo oo
o o o o o o
o r- ^H o o c^
o o o o o o
o o o o o o
o o o o o o
\O CO O m 00 VO
•^ (T\ CM CO P* ^T
n vo vO vo ^" ^
o o o o o o
oo ^r CJN TT m in
r^- CT^ rn ^-i «-i fH
^o rsi m m m r-t m ro ^
O o ^ o O CM
o o o o o o
o o o o o o
r^ vo (N CM ^r CM
-H ^ OJ ^" CN r-*
co r* ^ rsi r* co
o o o o o o
^ o ^f r- T ^
co o *^ m P* LT»
^-* CNJ m r* tN o
o o o O O o
r^ o o*» CN \o P^
^ O ^H ^T "-O CO
^ fN IN (N rg CM
fN (N (N (N (N| ^> O
vo in oo ^o
o^ r* »H co
0 0 0 O
^H m co r*
o CN »-H m
0 O O O
0 0 0 O
Ch o VD ro
r- p- *r in
ON co r- o
0 0 0 O
co <*o T ^r
<-* co rn in
r* o (N co
O 0 0 O
i-) ^r fH co
O --H O -H
o o o o
0 0 0 O
or\ cr\ T m
TT in in t-*
«-t O O O
o o o o
as m *r co
TP o T r>-
co cr> o*> co
o o o o
m r- oj r-
O*i O CO CO
m ^ in \D
t— t ^-4 -^ »— t
b^ a, o O
fN(fN^Ti-<(N':J1*T^
Oo«— lor'-jincoin
*-HOOOOOino
oooooooo
^rinnioinrnfMT
(M'rco^ococor-m
CNl^rrsiCMOO-Hrs)
OOOOOOOO
-r»»-H(N
momocoo-HO
^rm^r^r^r^rMin
oooooooo
UDOco^mroinr^
o^Or-tofnin^o
oooooo^oo
oooooooo
fM(Nm^H^tHinvD
Cho^cr\\^rsjin^'r-i
—trrrsirsjf-irHr-ti-H
OOOOOOOO
•HOrnm^HfNrrco
(Ncoino*^in>*o^ocrs
^rcsj^^r^^D»-(in
oooooooo
mmcomviJO^o
oooo--t— 'CNro
•HoooocNino
oooooooo
Oojm^O'-iOTin
o^i^^vDinotNr^
OtnfNrHOCNrn^H
oooooooo
•—Ivor"! ^mvOO"\O
co^ro>«-*vDinmin
vomvococo^r— tvo
oooooooo
wwtnwwtnww
>i>i>i>i>i>i>i>i
'O'D'O'D'O'O'D'a
iD(Timco»-imma5
i-l tH rsj (M (N TJ«
I
I
8,
S
01
6
w
>1
1
s
o
a.
B
o
o
t-i
O
•f~i
(0
3
0)
u
c
OJ
O
-------
REFERENCES
Budgell, W.P. A stochastic-deterministic model for estimating
tides in branched estuaries. Manuscript Report Series, No. 10,
Ocean Science and Surveys, Fisheries and Oceans, Canada 1981.
Burlington, Ontario. 189 p.
Dronkers, J.J. Tidal theory and computations. Advances in
Hydroscience. 10:145-226; Academic Press, Inc. 1972.
Enochoson, L.D.; Otnes, R. Programming and analysis for digital
time series data. Frequency response and coherence function
computations. 1968. The Shock and Vibration Information Center,
Naval Research Laboratory. Washington, D.C. p. 185-210.
Huthmann, G.; Wilke, K. Time and frequency domain of water quality
systems, Modeling, identification and control in environmental
systems. 1978. Amsterdam : North-Holland Publishing Company; 557-
583. a
Hsieh, B. Development of multiple coherence via linear system and
Gaussian elimination method. November 1985. Technical Notes, MD.
Dept. of Natural Resources, Annapolis.
Schubel, J.R. The physical and chemical conditions of Chesapeake Bay
an evaluation. January 1972. Chesapeake Bay Institute Ref. 72-1.
423
-------
Understanding the Estuary: Advances in Chesapeake
Bay Research. Proceedings of 'a Conference. 29-31
March 1988. Baltimore, Maryland. Chesapeake Research
Consortium Publication 129. CBP/TRS 24/88.
Time Variations of Bottom-Water Inflow at the Mouth of an Estuary
Glenn A. Cannon
National Oceanic and Atmospheric Administration
Pacific Marine Environmental Laboratory
7600 Sand Point Way NE
Seattle, Washington 98115
ABSTRACT
Puget Sound is a fjord-like estuary, but its 30-km long entrance sill, Admiralty Inlet, has
characteristics very similar to coastal plain estuaries. The replacement of bottom water in
Puget Sound has been studied for many years, because it is a dominant process responsi-
ble for flushing some contaminants. Previous studies showed bottom-water inflow
increased during neap tides when mixing was minimal over the entrance sill. Recent
observations show the increased inflow starts before minimum neap tides, and simple
model calculations with these data demonstrate this is an effect of variations in the
horizontal density gradient at the mouth of the estuary caused by salinity variations
outside the mouth. This time-dependent process may be responsible for changing inflow
characteristics at time scales between wind effects and seasonal effects, and it may be
important in other estuaries such as Chesapeake Bay.
INTRODUCTION
Puget Sound is the southernmost glacial carved estuary in western North America and is
surrounded by major urban centers. The entrance sill to this estuary is topographically
complex (Figure 1) and plays a major role in regulating the replacement of water inside
the estuary below the sill. This process is important for removal of some contaminants.
Salinity (density) at sill depth outside the estuary is always greater than inside at or below
the sill, but water from outside does not flow continuously into the deeper water. Over
the sill, the flow is two layered, and salinity is horizontally stratified, closely resembling
coastal plain estuaries. Previous studies showed bottom-water inflow events during neap
tides when mixing is least over the entrance sill (Geyer and Cannon 1982). Other studies,
however, had implied that inflow also might occur on very large spring tides when the
tidal excursion could transit the sill on a single flood tide. This paper describes new
observations to resolve this discrepancy and to determine whether the onset of the
intrusions could be predicted. Other recent observations have shown this circulation
424
-------
Figure 1. Topography of the Admiralty Inlet entrance sill connecting Puget Sound with
the Strait of Juan de Fuca showing locations of moored instruments outside and inside the
sill. Inset shows track across the sill connecting the moorings.
425
-------
process is exceeded only by wind effects in the amount of non-tidal energy
(Bretschneider et al. 1985).
OBSERVATIONS
Currents and salinity of the deep water just inside the entrance sill both show pronounced
increases at about fortnightly intervals which is the dominant characteristic of inflow of
new bottom water (Figure 2). The onset of these intrusions occurred in all cases before
minimum neap tidal currents, and the maximum salinity occurred at about this minimum.
Inflow during the largest intrusions is sufficiently large that the tidal currents do not
reverse at the bottom. Spring tides all occurred during decreasing salinity in the basin.
Thus, it appears clear that major inflows occur only during neap tides, but the onset
occurs before the minimum neaps. Outside the sill, salinity variations are larger than
inside the estuary, and they are not in phase.
Because of the relatively large variation in outside salinity and because the increase in
inside salinity occurred before neap tides, it was felt that variations in the pressure
gradient across the sill might play a role in the onset of the intrusions. Figure 2 shows a
comparison between conditions outside at sill depth and inside near bottom. Flow inside
the sill greater than 10 cm/sec shows the increase associated with increasing salinity.
Large salinity differences between inside and outside were coincident with intrusions of
new bottom water. A quasi steady-state balance between the pressure gradient and
vertical mixing predicts currents about the same as observed. A change in salinity of 1.4
parts per thousand is about what is required to initiate intrusions.
Given that the density difference across the sill accounts for the onset of bottom-water
intrusions, the cause of the seaward salinity variations needs to be explained. They may
be partly a result of offshore winds affecting the flow in the Strait of Juan de Fuca and
partly a result of mixing over a deeper sill in the Strait. Earlier work showed that coastal
storm winds drive surface water onshore at the mouth of the Strait. This reverses the
surface pressure gradient in the Strait causing inflow at the surface (Holbrook and Hal-
pern 1982) and sometimes reversing the bottom flow and/or suppressing the inflow of
saltier water (Cannon and Bretschneider 1986). The present study is the furthest into the
Strait that the coastal wind effects have been seen both in the surface and bottom layers,
and it appears they have an effect on salinity variations which regulate intrusions of
bottom water from the Strait of Juan de Fuca into Puget Sound. The details of this work
are described in Cannon et al. (1988).
DISCUSSION
Bottom-water intrusions are one of the major circulation features of Puget Sound, and
they play a major role in flushing some contaminants. New observations show the onset
of the intrusions occurs before minimum neap tides, and simple model calculations
demonstrate the importance of the density gradient across the sill caused by salinity
variations outside the sill. These variations may be further complicated because the
entrance sill is not at the coast, and connects Puget Sound to the Pacific Ocean through
another estuary, the Strait of Juan de Fuca. Wind events on the Pacific coast may be
capable of causing significant salinity changes more than 135 km from the coast.
However, the spring-neap cycle effect on mixing over the sill shown by Geyer and
Cannon (1982) still must play the major role. Although Puget Sound primarily is a fjord,
the entrance sill resembles a coastal plain estuary, and the process described here may be
important for variations in the magnitude of inflow in the bottom layer of a coastal plain
estuary. Because this is a time-dependent process, these results are presently being used
in guiding the development of a time-dependent laterally averaged model of the Puget
Sound estuary which includes this process as well as others.
426
-------
-60 ;
Dec
83
10
20
1
Jon
84
10
20
1
Fab
84
10
20
1
Mor
84
10
20
Figure 2. Observed salinity and currents at the entrance to Puget Sound showing the
relationship between the salinity (density) difference across the sill and the onset of
inflow of new bottom water into the estuary (vertical line at Feb. 6 shows example):
a) salinity inside at the bottom starts increasing;
b) salinity with tidal signal removed ouside at sill depth (upper) starts increasing
before inside at the bottom (lower);
c) flow inside at the bottom increases inward, shaded greater than 10 cm/sec inward
(plus is seaward);
d) salinity difference across the sill has increased, shaded greater than 1.4 parts per
thousand (ppt);
e) tidal flow shows inflow during neap (smaller) currents.
REFERENCES
Cannon, G.A., J.R. Holbrook, and DJ. Pashinski, 1988. Forcing the onset of bottom-
water intrusions over the entrance sill of a fjord. To be submitted.
Cannon, G.A., and D.E. Bretschneider, 1986. Interchanges between coastal and fjord
circulation. In: Contaminant Fluxes Through the Coastal Zone, G. Kullenberg,
ed., Rapp. P-v. Reun. Cons. int. Explor. Mer, 186, 33-48.
Bretschneider, D.E., G.A. Cannon, J.R. Holbrook, and D.J. Pashinski, 1985. Variability
of subtidal current structure in a fjord estuary: Puget Sound, Washington.
J. Geophys. Res., 90, 11949-11958.
Geyer, W.R., and G.A. Cannon, 1982. Sill processes related to deep-water renewal in a
fjord. J. Geophys. Res., 87,7985-7996.
Holbrook, J.R., and D. Halpern, 1982. Wintertime near-surface currents in the Strait of
Juan de Fuca. Atmosphere-Ocean, 20, 327-339.
427
-------
CONCURRENT SESSIONS
AND
POSTER SESSION:
TOXICS
Chairs:
Harriette L. Phelps
University of the District of Columbia
Robert C. Hale
Virginia Institute of Marine Science
College of William and Mary
-------
Understanding the Estuary: Advances in Chesapeake Abstract only
Bay Research. Proceedings of a Conference. 29-31
March 1988. Baltimore, Maryland. Chesapeake Research
Consortium Publication 129. CBPtTRS 24188.
Dynamics of Organic Pollutants in Blue Crabs, Callinectes sapidus,
Collected from the Lower Chesapeake Bay Region
Robert C. Hale
Virginia Institute of Marine Science
The College of William and Mary
Gloucester Point, VA 23062
The blue crab, Callinectes sapidus, is an abundant and widely
distributed species in the Chesapeake Bay. It also supports a valuable
fishery. Little information concerning concentrations of toxic organic
compounds in crustaceans of the southern Chesapeake Bay is currently
available. As a consequence, a study to determine the tissue burdens
and behavior of lipophilic polycylic aromatic compounds (PACs) in these
organisms was undertaken. Identification of compounds was by capillary
gas chromatography/mass spectrometry. Quantitation was accomplished by
flame ionization detection, using 1,1'-binaphthyl as an internal
standard. Highest concentrations of PACs were detected in
hepatopancreas, followed by ovarian and muscle tissues. Extractable
lipid levels in the tissues were positively correlated with organic
xenobiotic concentrations. The major contaminants detected in blue
crabs sampled from the southern bay were alkylated PACs, as opposed to
unsubstituted polynuclear aromatic hydrocarbons which have been
reported to predominate in molluscs and sediments of the bay. This
dichotomy may be due to differences in contaminant bioavailability or
in the relative abilities of the organisms to eliminate xenobiotics.
Crabs from both heavily industrialized and hypothetically "clean"
areas, showed evidence of exposure. Chromatographically unresolved
complexes were observed in 39% of the hepatopancreas samples. Adult
female crabs generally contained higher concentrations of lipophilic
pollutants than other intermolt groups.
Evidence that ecdysis in crustaceans may affect the disposition of PACs
was observed in laboratory experiments. Crabs were exposed to
radiolabeled benzo(a)pyrene (BaP) for 36 h and allowed to depurate for
252 h. Newly molted and adult intermolt female crabs retained greater
concentrations of radiolabeled material in hepatopancreas than adult
intermolt males or juvenile intermolt females. A variety of oxidized
biotransformation products were detected in the hepatopancreas of the
crabs by high performance liquid chromatography.
430
-------
Understanding the Estuary: Advances in Chesapeake
Bay Research. Proceedings of aConference. 29-31
March 1988. Baltimore, Maryland. Chesapeake Research
Consortium Publication 129. CBPITRS 24/88.
Bioassay for Phytotoxicity of Toxicants to Sago Pondweed
W. James Fleming and Jeffrey J. Momot
U.S. Fish and Wildlife Service
Patuxent Wildlife Research Center
Laurel, Maryland 20708
M. Stephen Ailstock
Anne Arundel Community College
Environmental Center
101 College Parkway
Arnold, Maryland 21012
INTRODUCTION
Submerged aquatic vegetation (SAV) occupies an important niche
in many aquatic systems, including freshwater lakes, estuaries, and
marine coastal waters. As primary producers, SAV captures much of
the energy that enters the system, making it available to support a
complex biological community. SAV provides food and habitat to
numerous organisms common to shallow water ecosystems and maintains
homeostasis in the system by buffering changes in water quality
through the removal of nutrients, toxics, and particulates. The
effects of the loss of SAV from an estuary can have a dramatic
effect on life in the estuary, as evidenced by changes that have
occurred in the Chesapeake Bay.
The decline of SAV in the Chesapeake Bay during the previous 2
decades has been substantial and well documented (Orth and Moore
1983; 1984). This decline was widespread, affecting all portions of
the Bay and all SAV species (Munro and Perry 1982; Orth and Moore
1984).
Synthesis documents prepared for the U.S. Environmental
Protection Agency (Kemp et al. 1982; Wetzel et al. 1982) cited water
quality problems as the most likely causes of the decline. Two
general mechanisms through which water quality stresses SAV were
identified. These were phytotoxicity and a reduction of light
energy due largely to heavy phytoplankton blooms in response to
nutrient enrichment. The reduction of light energy was considered to
431
-------
be the most important factor, however the effects of contaminants,
alone or in combination with other stressors, are poorly
understood.
To better understand the role of contaminants on SAV
populations, three types of information are particularly useful:
measurements of the types and concentrations of contaminants that
enter the estuary; knowledge of the environmental conditions
necessary for the survival of SAV; and identification of the
concentrations of contaminants which negatively impact growth of SAV
either directly or in combination with other environmental
perturbations.
Analytical chemistry determinations have provided excellent
data on the quantity of specific toxic compounds found in the
Chesapeake Bay. However, such determinations are costly for routine
water quality testing and have limited value for the development of
water quality standards. Bioassays are more useful indicators of
the presence of toxic substances and provide a more direct link with
the resource to be protected than analytical chemistry data. The
biological interpretation of such chemical data is often limited by
a lack of knowledge of effect levels. Bioassays can be used in the
laboratory to establish toxicity levels for specific contaminants or
to detect adverse effects on biota of complex effluents entering the
bay. The use of bioassays in conjunction with analytical
determinations can be a powerful tool for providing the information
necessary for legislative restrictions to improve and or maintain
the quality of aquatic resources.
Optimal value of a bioassay is obtained when the culture system
is well defined, has predictable variability, and is capable of
supporting the test organism through all stages of its life cycle.
The establishment of a bioassay system for submerged aquatic
angiosperms is difficult because of the complexity of their life
cycles and the paucity of specific information on the factors
regulating reproductive growth. Microcosms have been used for
conducting toxicity tests on SAV, but their usefulness is limited
because of difficulties in maintaining cultures for long periods of
time in a balanced system (Correll and Wu 1982). To avoid these
problems, attempts have been made to to establish axenic cultures of
submerged angiosperms. Growth in the axenic systems often has been
poor or the cultures were not truly axenic (Thursby 1984; Durako and
Moffler 1987).
Our laboratory has established axenic cultures of sago pondweed
(Potamogeton pectinatus) collected from several locations around the
Bay Stem, the Potomac River, and tidal impoundments. Sago pondweed
is a perennial, narrow-leafed, submerged aquatic macrophyte found in
temperate climates throughout the world (Fernald 1932). It is
tolerant to mildly alkaline waters and to a wide range of salinities
(between zero and about 10 ppt Yeo, 1965). Stewart (1962) reported
that sago pondweed was common and abundant in fresh water and
brackish portions of the Chesapeake Bay. Martin and Uhler (1939)
stated that sago pondweed is one of the more important waterfowl
432
-------
food plants in North America. It has a palatable rootstock, turions
(tubers), leaves, stems, and seeds. Stewart (1962) and Munro and
Perry (1982) reported substantial use of sago pondweed by several
waterfowl species through 1959. More recently, it only rarely is
found as a food item of waterfowl from the Chesapeake Bay (Munro and
Perry 1982; Perry and Uhler In press). The decline of sago pondweed
as an important waterfowl food in the Chesapeake Bay is probably
related to the overall decline of this and other species of SAV in
the Bay.
In our axenic cultures, sago pondweed grows rapidly on a
defined medium and exhibits normal vegetative growth. Turions are
occasionally formed in cultures placed under various environmental
stresses including light deprivation and nutrient depletion. When
plants are removed from the culture system they can be easily
established in a variety of soils. These plants, after additional
growth under long day conditions, have produced both flowers and
turions. The system thus provides a convenient source of large
numbers of axenic plants which are morphologically similar to plants
collected from the field or grown from naturally occurring
propagative structures.
While our current culture system holds many advantages, two
limitations are noteworthy. First plant size is limited by the
culture vessel and attainment of plants large enough to readily
produce flowers may require substantial culture modification.
Second and most significant, our axenic system is not capable of
supporting plants through autotrophic growth. This limitation is a
result of an undefined, limiting, environmental factor and not a
genetic abnormality; this is confirmed by the plant's high
photosynthetic activity when placed on bicarbonate enriched media,
and rapid growth when transferred to soils. Pending modifications
of the culture system to accommodate autotrophic growth, use of
axenic cultures for bioassays requires a two tier bioassay to
evaluate the effects of photosynthetic inhibitors in chronic
toxicity tests. This paper reports on limited tests with plants
produced or tested in axenic cultures.
EXPERIMENTS WITH ATRAZINE
In order to validate the use of axenic cultures in bioassays,
we conducted a series of tests to determine how closely toxicity
data derived from plants grown in axenic cultures corresponded to
results from plants grown in normal, non-sterile microcosms. Our
initial work was with the herbicide atrazine, we selected atrazine
because of the large data base on its effects on SAV, compared to
what is known of the phytotoxic effects of other herbicides or
toxicants.
First, we examined the effects of atrazine on photosynthesis.
After a preliminary rangefinding study, we selected geometrically
arranged concentrations of atrazine to bracket a level that we
expected to inhibit net photosynthesis by 50% (IC50). Plants
produced in axenic culture were removed from their culture vessel
433
-------
and placed in clean, non-sterile 300 ml biological oxygen demand
(BOD) bottles containing atrazine plus distilled water purged to
about 3 pg/1 oxygen by bubbling with nitrogen; 2.5 g of sodium
bicarbonate was added to each liter of test solution to provide an
inorganic carbon source for photosynthesis. BOD bottles were
incubated for 5 h in a 21 C waterbath under daylight^ spectrum
fluorescent lights (Vital ite®) producing about 130 ^Bin/m /s
at the neck of the BOD bottles. Net photosynthetic activity was
determined by measuring oxygen produced. Oxygen was determined with
an Orion oxygen electrode following the manufacturer's directions.
Photosynthesis in control plants resulted in an average of 3.5 +
0.64 mg of oxygen produced per g dry weight per hour. Our estimate
of the IC50 for atrazine was 29 /ig/1 (95% CI= 20-40 /jg/1 ; Figure
1). Atrazine's ICSOs for photosynthesis in aquatic plants has
previously been reported to range from 55 to 117 /ig/1 (Jones and
Winchell 1984; Jones et al . 1985; Kemp et al . 1985) which is similar
to our finding, given the differences in testing protocols. Correll
and Wu (1982) reported that photosynthesis of sago pondweed was
stimulated at 75 M9/l> with respiration exceeding photosynthesis in
plants exposed to atrazine at 650 ^g/1 ; why their results differed
substantially from ours and those of others is not known.
In the second test, we compared the responses of vegetative
plugs of sago pondweed originating from our axenic stock to that of
plants growing from tubers collected from the wild, when both were
exposed to atrazine at 0, 100, and 1000 ^g/1 under non-axenic
conditions. Using a split plot design, we had five replicates of
each treatment level. A replicate consisted of three plant plugs
and three tubers placed in a 19 1 bucket. Each bucket contained a
sand, peat, shell substrate and 14 1 of the atrazine-water
solution. Weights of plants and tubers were recorded pretreatment
and again after 32 days, at which time the experiment was
terminated. Water was added to each bucket to compensate for
evaporation. Buckets were kept under daylight spectrum fluorescent
lights 24h per day (about 75 /iEin/m /s ) at ambient room
temperatures (about 22 C). Data were analyzed by analysis of
variance and Tukey's test for mean separation.
In the control group, proportional increases in biomass were
similar between plugs and tubers (426% vs. 440%; Table 1). Compared
to these controls, the 1000 ^g/g atrazine solution depressed growth
(P < 0.05) an average of 64% for plants started from plugs, and 50%
for those started from tubers; the response of tubers to atrazine
was not different from that of vegetative plant plugs (P > 0.05).
Growth in the 100 (ig/1 atrazine group did not differ from controls
(P > 0.05).
The third and most critical part of our test sequence was to
determine if plants exposed to atrazine while growing under axenic
conditions would respond in a similar way to plants we tested in the
buckets. This test was conducted in two parts. In Part 1, we used
very young plants, grown for about two weeks from a single rhizome
tip. At this time, the tips were just beginning to differentiate
into new leaves and additional rhizome tips and weighed <0.05 g.
434
-------
w
u
£
•rf
C
ZT>
n
o
w
e
£
a.
e
L
•-<
C
O
u
o
*t
100
80.
y = -
50.97 logx - 28.41
R-squared:O.81
5 10 50
Concentration of Rtrozine
Figure 1. Net photosynthesis of sago pondweed exposed to atrazine,
Photosynthesis was determined by measuring changes in dissolved
oxygen.
Table 1. Biomass of sago pondweed grown for 32 days in 19 1 buckets
containing atrazine. Plants introduced into buckets were of two
forms, rhizomous tubers and vegetative plugs produced from axenic
cultures.
Plugs
Tubers
Pretreatment
Posttreatment
Pretreatment
Posttreatment
Atrazine
0
0.60+0.16Aa
2.56+1.19A
1.00+0.62A
4.40+2.64A
Concentration
100
0.70+0.19A
2.40+0.96A
1.01+0.66A
4.19+2.69A
(«Q/1)
1000
0.68+0.10A
0.91+0.20B
1.06+0.69A
2.20+1.40B
aMeans followed by different letters within the same row are
significantly different (P < 0.05).
435
-------
These small plants were placed into 125 ml Erlenmeyer flasks
containing 50 ml of filter-sterilized media to which atrazine was
added (Table 2). Plants were grown for 7 weeks. Control media and
culture media containing the atrazine was replaced with a similar,
freshly made media midway through the study. Plants were maintained
under full spectrum fluorescent lights at about 20-22 C, with a
14L.-10D light cycle.
The results of this test were inconclusive. Whereas our bucket
experiment demonstrated a depression of growth at 1000 ^g/1
atrazine, the data from the flask seemed to show stimulatory
responses to low levels of atrazine and no depression of growth at
1000 jig/1. Stimulation of growth has been previously reported to
occur in Hyriophyllum spicatum exposed to 5 pg/1 atrazine, but at 50
^g/1 growth was significantly inhibited (Kemp et al. p!985).
Further examination of our data indicated much variability within
the control and treatment groups and that the growth of the plants
was not normally distributed. This variability was related largely
to the failure of many of the plants to thrive, resulting in a
highly skewed distribution of plant weights. Believing this to be
related to the small, relatively morphologically undifferentiated
plant material that we started with, we ran a small study (Part 2)
using larger, more developed plants, averaging about 0.32 g each.
Part 2 was conducted under the same axenic and ambient
conditions as Part 1, but the test period was 8 weeks and the media
was not replenished. Variability within treatment groups was
greatly reduced in controls from Part 2 compared to Part 1
(Coefficient of variation * 33% vs. 130%). Controls from Part 2
increased their biomass 10-fold during the 8 week test. Results of
this experiment were promising (Table 2), but still not convincing
as we worked only with high atrazine concentrations in a effort to
elicit a plant response. However, 1000 ^g/1 atrazine under these
conditions inhibited growth by a 57%, which is similar to that
produced in the bucket experiment.
These growth data compare favorably with those of Forney and
Davis (1981) who reported that atrazine's IC50 to Potamogeton
perfoliatus was 907 ^g/1 as determined by final dry weights. They
also reported ICSOs of 1104 j*g/l to M. spicatum and 163-532 Atg/1 to
Vallisneria americana where leaf or plant length was the variable.
Correll and Wu (1982) reported that 120 ^g/1 caused 100% mortality
in V. americana. and Cunningham et al. (1984) found that 130 M9/1
significantly reduced growth of P. perfoliatus during a 4 week
exposure period. Kemp et al. (1982; 1985) reported that increase in
biomass was affected at considerably lower concentrations of
atrazine, with ICSOs of 30-130 ^g/1 to P. perfoliatus and 91 ^g/1 to
M. spicatum. The wide range of toxicity values for atrazine to SAV
probably reflects the absence of a standard toxicity testing
protocol and difficulties in maintaining SAV in non-axenic cultures
for long periods of time.
436
-------
O1
c
•t—
•O J=
a; en
QJ IO ••-
5 Q. QJ
T3-- X
C *->
O 1 •>
CLOC CO
4->
O T3 C
CD O) re
re -4-> r—
«/> re o.
•t«-
t|_ 4J QJ
0 C S.
0) 3
t . % .
*-^ »•. +•» •
t/> o> re >
O.H- E -^
.,-<»_ 0> ^-.
4-> T- 0) O) f—
' -0 S- X \
of c o en
«— 3 E oo a.
«/> >> - s-
Q.r— CM O l»
•r- 0) «4- C
+J > •»-> O
•r- l_ C -r-
o> +-> re X -t->
E re o. o re
Or— I- S-
M QJ c en 4->
•r- fc. •—< C
_e QJ QJ
I. - I. U
i— c . QJ C
**- v> X O
O 4-> -* O
4. O> ••
s_ re QJ i/> QJ
QJ Q. X C C
.Q O -t-
E C r-. -F- N
3 i— I +J «X)
C I. 3 J-
O f— +J
-o • <+- o <
C QJ )
re c v>
•r- C •»-»
^ N O V>
+j re ••- o>
.C S. •!-> 4->
CT4J 3
•r- re t— tt>
0) O -C
X -c t/» +->
4->
.C "- (U C
to x c •«-
o> -t-
i- OJ N T3
t- t. re re
• r— « r—
3 Q.
V» O Q)
V) J2 C
re u •<-> o>
E .r- .c
o
o
o
in
I-H
O
O
o
I-H
o
o
r—l
CD
t-H
1— 1
O
0 C C X
•i- 0) -r-
co x en
re c:
§co
.
CM •»- S. O
cn
QJ C -M
r— X O) 3
J3 O S- O
« S- Q> -Q
1— cn X re
.0
1 1
1 1
CO
CO O
^
vo
o
4-1 .-H
CM -HI
in vo
O CM
CO
<
CO
^s.
t»H
vo
t— 1 •
•H CM
^- 4-1
r- *»•
• •
t-H VO
<
et
CM
rH VO
• •
CM in
4-1 -H
CO CM
r~-
• t-H
CM t-H
CO
O
VO CO
CO
co
o
4-1 CM
CO +|
cn o
• •
o m
CO
v_>
CM CO
cn
in
o
•H CM
r-" 4-1
t-» CM
• •
O CO
>
(/> V)
^H| re o.
E T-
+-> o +->
t- ••- 1
re co of
o.
CO
^- CO
CO
«a-
o
4-1 ^H
VO +|
00 O
* •
o **•
CO
CO CO
in
CO
o
4-1 —•
0 4-1
in CM
t-H ^
i i
i i
i i
i i
i i
i i
<
<
**•
l-H
^»-
l-H •
4| CM
r>- 4-1
^J- 0
• •
co cn
VI
V) Vt
CM| re CL
E ••-
•4-> O *->
$_ -t- 1 •
re co ac.
o.
m
o
o
VI
Q-
~,_,-
4->
C
0)
S-
0)
(4-
4-
•r—
•o
>)
^
«0
w
u
•r-
U-
•r—
C
o>
•r-
>
0)
S-
re
X
O
J-
(/)
0) C
E 0
re •.-
fcO 4->
re
0) i-
j= -i->
•<-> c
QJ
C O
•t- C
-C O
•t-> U
'x OJ
c
«« -r-
&- N
0) re
4-> i-
-t-> -t->
o> re
P^
0)
4-> V)
C QJ
QJ ^
1- -l->
QJ
(f- 4J
**- re
•r—
•o -o
QJ
>> 4->
J3 U
3
T3 -0
QJ C
X o
0 U
i— 4J
0 0
*»- c
00 CO
C •!->
re oo
QJ QJ
ac h-
re ^
437
-------
CONCLUSIONS AND FUTURE RESEARCH
Axenic culture techniques for sago pondweed can produce an
abundant amount of healthy plants year-round for toxicity testing.
Plants produced in axenic cultures showed photosynthetic responses
that were similar to published results for SAV exposed to atrazine.
Plants produced under these conditions and grown in buckets
responded no differently to atrazine exposure than did plants
produced by non-axenic methods. Growth of plants exposed to
atrazine in axenic cultures varied depending on the starting size of
the plants. Larger, more differentiated plants exposed to atrazine
in axenic cultures demonstrated a depression of growth similar to
those exposed under non-axenic conditions. Plants in our axenic
culture system grow far more rapidly than those grown in buckets.
This could reduce the amount of labor required to maintain the test
plants, and also compress the time required to detect effects. We
emphasize this point because "he U.S. Environmental Protection
Agency requires the use of fresh effluent and frequent renewal of
newly collected fresh effluent in many of their aquatic bioassays.
Thus, the practicality of a bioassay is partially dependent on the
duration of exposure required for each test protocol.
The use of axenic testing procedures for determining the
effects of toxicants on plant growth appears to be promising, but is
clouded by recent findings that indicate that plants in our axenic
culture system appear to be primarily heterotrophic, not
autotrophic. We are attempting to refine the current culture
technique in an effort to produce an axenic, autotrophic culture
system. In the interim, we are continuing to develop protocols for
a bioassay in which individual pollutants and effluents from various
sources in the Chesapeake Bay area will be examined for
phytotoxicity. In this approach, effluents will be first tested for
inhibition of photosynthesis. If no inhibition is found, filter
sterilized effluents will be tested for their effects on plant
growth and morphology in axenic cultures. If photosynthesis is
found to be inhibited by an effluent, the phytotoxicity of the
effluent will be examined in our nonaxenic bucket "microcosms".
While continuing to explore the use of axenic cultures in
toxicity bioassays, we are maintaining contact with State and
Federal regulatory agencies concerning requirements for standard
protocols for SAV toxicity testing. We hope to begin a trial program
of effluent screening in 1988 to evaluate the usefulness and
sensitivity of our test system and protocols.
438
-------
LITERATURE CITED
Correll, D.L. and T.L. Wu. 1982. Atrazine toxicity to submersed
vascular plants in simulated estuarine microcosms. Aquat. Bot.
14:151-158.
Cunningham, J.J., W. M. Kemp, M.R. Lewis and J.C. Stevenson. 1984.
Temporal responses of the macrophyte, Potamogeton perfoliatus L.,
and its associated autotrophic community to atrazine exposure in
estuarine microcosms. Estuaries 7:519-530.
Durako, M. J. and M. D. Moffler. 1987. Nutritional studies of the
submerged marine angiosperm Thalassia testudinum I. Growth
responses of axenic seedlings to nitrogen enrichment. Amer. J.
Bot. 74(2):234-240.
Fernald, M.L.
Potamogeton,
17(Part 1):1
Forney, D.R. and D.E.
of herbicides on
29:677-685.
1932. The linear-leaved North American species of
Section axillaries. Mem. Am. Acad. Arts Sci.
Davis. 1981. Effects of low concentrations
submersed aquatic plants. Weed Sci.
Jones, T.W. and L. Winchell. 1984. Uptake and photosynthesic
inhibition by atrazine and its degradation products on four
species of submerged vascular plants. J. Environ. Qual.
13:243-247.
Jones, T.W., W.M. Kemp, P.S. Estes and J.C. Stevenson. 1985.
Atrazine uptake, photosynthetic inhibition, and short-term
recovery for submersed vascular plant, Potamogeton perfoliatus
L. Arch. Environ. Cont. Toxicol. 15:277-283.
Kemp, W.M., W.R. Boynton, J.J. Cunningham, J.C. Stevenson, T.W.
Jones and J.C. Means. 1985. Effects of atrazine and linuron on
photosynthesis and growth of the macrophytes, Potamogeton
perfoliatus L. and Myriophyllum spicatum L. in an estuarine
environment. Marine Environ. Res. 16;255-280.
Kemp, W.M., J.C. Means, T.W. Jones and J.C. Stevenson. 1982.
Herbicides in Chesapeake Bay and their effects on submerged
aquatic vegetation. Pages 503-567 in Chesapeake Bay Program
|