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
>       
o

 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 
•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