^ f a
$  «  \       UNITED STATES ENVIRONMENTAL PROTECTION  AGENCY
S - k»«—i - 5
        ?                         WASHINGTON, D.C.  20460

                                                                     905R87104

                                          I5I987
                                                                     OFFICE OF
                                                              RESEARCH AND DEVELOPMENT
        SUBJECT:  Transmittal  of  ORD Internal  Report Entitled Report on
                  Methodology  Protocol  for  Field Exposure/Effects Assessment
                  of a Great Lakes  Area of  Concern  (Deliverable 5884-A)
        FROM:      James  W.  Falco/-/
                  Director,  Office  of Environmental Processes
                    and  Effects and Research  (RD-682)
        TO:        Howard  Zar
                  Acting  Director,  Great Lakes National Program Office

             The  attached copy  of the subject ORD Internal Report is being
        transmitted to  your office  in response to the Great Lakes National
        Program Office's  need for methods to deal with complex water quality
        issues  in the Great Lakes "Areas of Concern."  The report summarizes
        research  study  which included development and application of chemical,
        biological, toxicological,  and mathematical modeling approaches.

             The  study  attempted to deal with toxic substances from both a
        chemical-specific basis and a toxicity basis.  Toxicity was defined
        and  measured by a number of bioassay techniques using water from control
        site effluents  and points of impact.  Bioassay results were converted
        to toxic  units  which put data on a common basis for intercomparison.
        This toxicity based approach measures aggregate toxicity of all chemical
        constituents present but does not lead to an understanding of what is
        causing the effects.  Zinc  and copper correlated well to effects in
        different trophic levels and specific ecosystem functions.  Copper and
        zinc concentrations, corrected for hardness, exerted the greatest in-
        hibition  on zooplankton grazing, whereas zinc in combination with alkalinity
        appeared  to inhibit zooplankton reproduction.  Growth rates of gizzard
        shad and  emerald  shiners exhibited an inverse relationship to zinc,
        chromium, and copper in sediment.

             Although PCS concentrations were^present in fish and accumulated
        rapidly in caged  organisms, they did not appear to be significant in
        describing any  of the observed toxicity in the Cereodaphnia reticulata
        reproduction and  survival.  Finally, a mass balance was attempted for
        PCBs and  other  chemicals.   It was estimated that an unknown source of
        PCBs, about 100-200 grams,  exists within the study.  The most likely
        source  is from  ground water.

-------
     The report is provided to you for possible application to other
"Areas of Concern" and for possible use in developing a "Remedial
Action Plan" for the lower River Raisin.

     In an attempt to evaluate the utility, timeliness and quality
of the information this office provides to the Program Offices through
research products (deliverables), a "critique sheet" has been developed
and is attached to this transmittal memo.  I would appreciate your
providing the information requested and returning the completed
form to me at your earliest convenience.  In order to obtain a
meaningful indication of our performance, this request will continue
to be made for all deliverables transmitted to you through December
31, 1987.  Your cooperation in this request is appreciated.

Attachments

-------
                            OEPER Deliverable

                              Critique Sheet


     The Office of Environmental Processes and Effects Research (OEPER)
has recently completed work on a project for or of interest to you.  As
a means of ensuring the quality of the products delivered by OEPER and to
determine if we are meeting the user's needs, your answers to the following
questions, and any additional comments you may have, are requested:

Deliverable Title:   Report on Methodology Protocol for Field Exposure/Effects
                     Assessment of a Great Lakes Area of Concern (Deliverable
                     5884-A)

Date Received:
Timeliness:  	 Ahead of schedule,  	 On schedule,  	Late

Did the product provide the information desired?  	 Yes, 	 No*

Quality of information contained in product:  	 Excellent,  	 Good,  	 Fair,
                                              	 Poor*

Product Format:  	 Excellent,  	 Good,  	 Fair,  	 Poor*,  	 Unacceptable*

Was an interim product provided?  	Yes,  	 No

Was the interim product useful?  	 Yes,  	 No

How will the product be used?



Comment s:
Name:                                       Phone:
     Please return this sheet to:  Director, OEPER (RD-682).  Thank you for
your time and assistance.

*Please explain

-------
     This draft of "Sumary Report:   An Integrated Approach to a Study



of Contaminants and Toxicity in Monroe Harbor (River Raisin), Michigan,



A Great Lakes Area of Concern" fulfills obligations under Delivery



#5884A.
                              June 24, 1987

-------
SUMMARY REPORT:  AN INTEGRATED APPROACH TO A STUDY OF CONTAMINANTS
      AND  TOXICITY  IN MONROE HARBOR  (RIVER  RAISIN), MICHIGAN,
                   A GREAT  LAKES AREA OF CONCERN
                                by

     The  Staff,  Contractors,  Grantees  and  Collaborators  of the
                   Large Lakes Research Station
               U.S. Environmental Protection Agency
                    Grosse  lie,  Michigan   48138
                        EPA Grant Numbers:
     810232,  810775,  810776,  810779,  810808,  811578 and 811731

                       EPA Contract Number:
                            68-01-7170
         This study was conducted in cooperation with the
             Michigan Department*of Natural Resources
                 ENVIRONMENTAL RESEARCH LABORATORY
                OFFICE OF RESEARCH AND DEVELOPMENT
               U.S.  ENVIRONMENTAL PROTECTION AGENCY
                      DULUTH,  MINNESOTA   55804

-------
                       NOTICE

This document has been reviewed in accordance with
U.S. Environmental Protection Agency policy and
approved for publication.  Mention of trade names
or commercial products does not constitute endorse-
ment or recommendation for use.

-------
                                   FOREWORD
     There are many complex,  unresolved environmental  issues impeding
rational  management of the Great Lakes ecosystem.   These issues range from
the long-term, lake-wide ecosystem risk associated with persistent toxic
substances to the acute effects of short-lived contaminants in estuaries,
harbors,  and nearshore areas.  To help address some of these issues for the
Great Lakes, the U.S. Environmental  Protection Agency, Office of Research and
Development sponsored a two year, site specific study of the River Raisin
near Monroe, Michigan.  This study involved a multidisciplinary team of EPA,
state, contractor, and academic researchers who focused on developing
experimental methods and a project design for identifying contaminants,
determining their sources, and evaluating their effects using this "Area of
Concern"  as a representative system.

     This report summarizes the methods and tools  that were developed and
tested during this project.  The purpose is to document these so that they
might be applied to other, similar areas around the Great Lakes and for any
other location facing similar issues.   If more information is required, the
reader is encouraged to request the related project reports, data bases, and
journal articles from the U.S.E.P.A. Large Lakes Research Station or directly
from the Principal Investigators.
Gilman D. Veith, Ph.D.                 William L. Richardson, P.E^
Director                               Chief
ERL-Duluth                             Large Lakes Research Station
                                     11

-------
                                    PREFACE
     This report synthesizes several project reports which represent an
integrated study of contaminants and their sources,  fates, and effects using
Monroe Harbor as a representative site—one of 42 Great Lakes "Areas of
Concern".  The methods and results presented in all  of the project reports
are documented here so that this summary report may  be useful in designing
future research studies of similar ecosystems.  The  primary purpose of the
study was to demonstrate a research approach to understanding the problem of
complex mixtures of contaminants and their possible  effects on aquatic
ecosystems.

     There are many scientific and technical challenges to be faced in
research of this kind.  Not surprisingly, these studies have raised questions
which require lengthy scientific discussion and analysis.   Such detailed
information will be found in the final technical report of each project.
Collectively, these technical reports will represent the complete project
documentation.

-------
                                   ABSTRACT
     Traditionally, water quality assessment has relied largely on chemical
analyses for specific toxic constituents.  However,  in areas like Monroe
Harbor which have been affected by complex mixtures  of industrial pollutants,
it is usually not possible to identify all toxicants that are present or to
know how they interact.  In areas such as this,  the  aggregate toxicity of
these constituents can be measured by using test organisms.  In this study,
performed during 1983-1984, both the chemical-specific and toxicity-based
approaches were used to assess water quality in  effluents and receiving
waters of Monroe Harbor, Michigan, an estuary of the River Raisin on western
Lake Erie.  Monroe Harbor is designated as an "Area of Concern", one of 42
such areas with degraded water quality in the Great  Lakes Basin.  This site
served as a representative system for demonstrating  the two water assessment'
approaches.

     Toxicity was defined and measured by a number of bioassay techniques
using water both from control sites and from likely points of impact in the
system.  In various assays, toxicity was expressed as inhibition of bacterial
decomposer activity, phytoplankton productivity, zooplankton reproduction and
grazing efficiency, as survival of zooplankton and larval fish, and as
contaminant accumulation by invertebrate and fish species.

     All bioassay results were converted to toxic units which put data on a
common basis for making intercomparisons.  In most cases, the greatest
relative toxic or inhibiting effects found in the River Raisin were observed
in the vicinity of the Monroe Wastewater Treatment Plant and lakeward from
it.  Zooplankton and larval fish toxicity were simulated in the River Raisin
using a probabilistic transport and fate model.   Two assumptions of the model
were that toxicity was additive and conservative.  Model predictions based on
these assumptions were consistent with observations.

     PCBs are common in the River Raisin ecosystem,  and bioavaiTable to
aquatic life.  A number of common carp, emerald shiners, small mouth bass,
gizzard shad and mirror carp exceeded th.e USFDA action level of 2.0 mg/kg
total PCBs.  PCB bioaccumulation studies were conducted using caged clams,
fathead minnows, and channel catfish.  Although the thirty-five day exposure
was not long enough to reach equilibrium, the studies demonstrated that the
PCB homolog pattern of the water was rapidly matched by the organism.  In
addition, phytoplankton were found to take up significant amounts of
sediment-bound hexachlorobiphenyl in just 24 hours.
                                     IV

-------
     Chemical analyses supported all of the biological effects studies and
provided data to define the present environmental status of the system.
Despite some interference problems, it was possible to carry out high
resolution gas chromatography of most PCB congeners and other organochlorines
in over 250 water, sediment, and biota samples.  An undetermined loading of
PCBs that approached 200 g/day was found in the lower River Raisin by an
input-output mass balance model.  Copper, zinc, and chromium concentrations
were well below the EPA Water Quality Criteria (1976) for drinking water.
However, sediment concentrations of the three metals were found to be
elevated when compared to sediment data from open water areas of the Great
Lakes.  Statistical analysis of the data suggests that the source of the
metals was local.

     An important area of research in the project was an attempt to integrate
the toxicity-based and the chemical-specific approaches.  Statistical
analysis indicated that zinc and copper appeared to be toxic to different
species and to affect specific ecosystem functions, either singly or in
combination.  Copper and zinc concentrations, corrected for hardness, exerted
the greatest inhibition on zooplankton grazing, whereas zinc corrected for
alkalinity appeared to inhibit zooplankton reproduction.  Growth rates of
gizzard shad and emerald shiners exhibited inverse relationships to
concentrations of zinc, chromium, and copper in the sediment.  Although PCBs"
were important in the bioaccumulation studies, they did not appear to be
significant in describing any of the observed toxicity in the Ceriodaohnia
reproduction and survival in the seven-day Mount-Norberg life cycle test.

-------
                                   CONTENTS
FOREWORD	     ii
PREFACE	    iii
ABSTRACT	     iv
FIGURES	    xii
TABLES	    xvi
ABBREVIATIONS AND SYMBOLS  	  xviii
ACKNOWLEDGMENTS  	     xx

1.0  INTRODUCTION  	      1

     1.1  Purpose	      1
     1.2  Organization of the Summary Report 	      2

2.0  CONCLUSIONS 	      4

     2.1  Integration of the Whole-effluent and Chemical-specific
          Approaches 	      4
     2.2  Whole-effluent Approach  	      4
     2.3  Chemical-specific Approach 	      5

          2.3.1  Bioaccumulation/Bioavailability Studies 	      5
          2.3.2  Fate and Transport Model of Chromium, Zinc, and
                 Copper	*	      6

     2.4  Related Studies	„.  .      6

          2.4.1  Descriptive Biology 	      6
          2.4.2  Water Chemistry 	      7
          2.4.3  Sediment Chemistry  	      8
          2.4.4  Data Base Documentation	      9
          2.4.5  PCB Input-Output Mass Balance Model  	      9

3.0  RECOMMENDATIONS	     10
                                       %

     3.1  Biology	     10
     3.2  Water Chemistry	     13
     3.3  Water Quality Modeling 	     13
     3.4  Data Base Documentation	     14
                                     VI

-------
                              CONTENTS  (CONT'D.)
4.0  BACKGROUND	   15

     4.1  Basin Description  	   15
     4.2  Hydrology	   15
     4.3  Industrial Development 	   17
     4.4  Industrial Landfills 	   17

5.0  PROJECT OBJECTIVES AND APPROACH 	   18

     5.1  Objectives	   18
     5.2  Approach	   19

          5.2.1  Water Quality Assessment  	   19

                 5.2.1.1  The Whole-effluent (Toxicity-based)
                          Approach	   19
                 5.2.1.2  The Chemical-specific Approach 	   22

          5.2.2  Integration of the Toxicity-based and
                 Chemical-specific Approach  	   23

     5.3  Survey Strategies  	   24

6.0  WATER QUALITY MODELS OF METALS, TOXICITY,  AND PCBs  	   28

     6.1  Modeling Objectives  	   28
     6.2  Transport Considerations 	   30

          6.2.1  Hydrological Description of the System  	   30
          6.2.2  Resuspension Considerations for the 1983 Surveys  ...   33

     6.3  Twenty-Two Segment Fate and Transport Model  of Copper,
          Chromium, Zinc, and Toxicity 	   35

          6.3.1  Approach	   35
          6.3.2  Results	   36

                 6.3.2.1  Modeling Conservative Substances 	   36
                 6.3.2.2  Modeling Copper, Chromium, and Zinc  	   36
                 6.3.2.3  Modeling Toxicity as  a State Variable  ....   42

          6.3.3  Evaluation	   46

     6.4  Input-Output Mass Balance Model for PCBs	   47

          6.4.1  Approach	   47
          6.4.2  Results and Evaluation	   48


                                    vii

-------
                              CONTENTS  (CONT'D.)



7.0  COMPONENT STUDIES	   61

     7.1   Water Chemistry	   61

          7.1.1  Survey Approach	   61

                 7.1.1.1   Rationale and Strategy 	   61
                 7.1.1.2  Survey Schedule and Stations 	   62
                 7.1.1.3  Sampling Plan	   62

          7.1.2  Physical/Chemical Parameters  	   63

                 7.1.2.1   Methods  	   63
                 7.1.2.2  Results  	   63
                 7.1.2.3  Evaluation 	   63

          7.1.3  General  Chemistry	   65.

                 7.1.3.1   Methods  	   65
                 7.1.3.2  Results  	   65
                 7.1.3.3  Evaluation 	   65

          7.1.4  Organic Contaminants   	   67

                 7.1.4.1   Methods  	   67
                 7.1.4.2  Results  	   68
                 7.1.4.3  Evaluation 	   68

          7.1.5  Metallic Contaminants  	   71

                 7.1.5.1  Methods	\.  .  .   71
                 7.1.5.2  Results  	   72
                 7.1.5.3  Evaluation 	   72

          7.1.6  Quality Assurance:  Results and Evaluation   	   75

                 7.1.6.1  Physical/Chemical Parameters 	   76
                 7.1.6.2  General Chemistry  	   76
                 7.1.6.3  Organic Contaminants 	   76
                 7.1.6.4  Metallic Contaminants  	   77

     7.2  Biological  Studies  in Monroe  Harbor  (River  Raisin),
          Michigan	   78

          7.2.1  Introduction	   78
          7.2.2  Bacterial Decomposer Bioassay  	   82
          7.2.3  Phytoplankton  Carbon Uptake Bioassay  	   85


                                    vi ii

-------
                         CONTENTS  (CONT'D.)
     7.2.4   Zooplankton Grazing Bioassay  	     88
     7.2.5   Zooplankton Reproduction Bioassays  	     91
     7.2.6 ' Zooplankton Reproduction and Fish Lethality
             Bioassays of Sediment 	     95
     7.2.7   Fathead Minnow Toxicity Bioassays 	     97
     7.2.8   Fathead Minnow Flow-Through Bioassay  	     98
     7.2.9   In-Situ Bioaccumulation of PCBs in Clams,  Fathead
             Minnows, and Channel  Catfish  	     98

             7.2.9.1  Approach 	     98
             7.2.9.2  Methods  	    100
             7.2.9.3  Results  	    103
             7.2.9.4  Evaluation 	    105
             7.2.9.5  Additional  In Situ Bioaccumulation
                      Experiments	    109

     7.2.10  Resident Larval Fish Studies  	    109
     7.2.11  Contaminant Body Burdens of Resident Fish
             Populations	    121
     7.2.12  Evaluation of Biological Studies  	    125

7.3  Sediment Chemistry  	    128

     7.3.1   Physical Characterization 	    128

             7.3.1.1  Approach 	    128
             7.3.1.2  Methods  	    129

                      7.3.1.2.1' Field 	    129
                      7.3.1.2.2  Laboratory  	    129

             7.3.1.3  Results and Evaluation	.* .  .    134

     7.3.2   Metallic Contaminant Characterization 	    137

             7.3.2.1  Approach 	    137
             7.3.2.2  Methods  	    137

                      7.3.2.2.1  Field 	    137
                      7.3.2.2.2  Laboratory  	    138

             7.3.2.3  Results and Evaluation 	    138

     7.3.3   Organochlorine Contaminant Characterization 	    140

             7.3.3.1  Approach 	    140
             7.3.3.2  Methods  	    140

-------
                              CONTENTS (CONT'D.)
                          7.3.3.2.1   Field  	    140
                          7.3.3.2.2   Laboratory 	    140

                 7.3.3.3  Results and Evaluation  	    141

          7.3.4  Adsorption/Desorption Experiments  	    145

                 7.3.4.1  Approach  	    145
                 7.3.4.2  Methods 	    150

                          7.3.4.2.1   Metals 	    150
                          7.3.4.2.2   Organics 	    151

                 7.3.4.3  Results and Evaluation  	    154

                          7.3.4.3.1   Metals 	    154
                          7.3.4.3.2   Organics	    154.

8.0  DATA BASE DOCUMENTATION SUMMARY	    159
9.0  REFERENCES	    162

APPENDIX A - PCB MASS BALANCE MODELS, 1983	    169
APPENDIX B - DATA BASE DOCUMENTATION	    174

     B.I  Physical Data	    174

          B.I.I  Flow Data	    174

                 B.I. 1.1  Flow and Water Chemistry Data From the
                          Second Low Head Dam, Monroe 	    174
                 B.I.1.2  Waste Water Treatment Plant, WWTP .  .  .  \ .  .    174
                 B.I. 1.3  USGS Daily Flow Data	    174

          B.I.2  River Current Data	    174
          B.I.3  Wind Data	    174
          B.I.4  Lake Level Data	    175
          B.I.5  Hydrolab  Profile Data	    175
          B.I.6  Transect Bottom Depth Data	    175
          B.I.7  Precipitation Data	    175

     B.2  Water Quality Data	    176

          B.2.1  Grosse lie Data System (GIDS)	    176

                 B.2.1.1  Monroe Section	    176
                 B.2.1.2  MHSPEC Section   	    176

-------
                         CONTENTS  (CONT'D.)
           .B.2.1.3  MHTCHM Section  	   180
            B.2.1.4  SEDCHM Section  	   180

     8.2.2  Organics Analytical Data Base	   180
     B.2.3  Point Source Data	   181

B.3  Biota Data	   181

     B.3.1  In-House Ceriodaohnia and Fathead Minnow Data  ....   181
     B.3.2  Ohio State University Larval  Fish Data	   181
     B.3.3  University of Minnesota Biota Data 	   182

            B.3.3.1  Ceriodaohnia Bioassays  	   182
            B.3.3.2  Phytoplankton Productivity Bioassays  ....   182
            B.3.3.3  Zooplankton Grazing Bioassays 	   182
            B.3.3.4  Bacterial Productivity Bioassays	   182.

     B.3.4  Other In-House Biological Data 	   182

B.4  Sediment Data	   182
B.5  Other Data	   182

-------
                                    FIGURES



Number                                                                   Page

 4.1      Monroe Harbor - River Raisin study  area 	   16

 5.1      Example demonstrating the relationship between exposure
         (dilution fraction of effluent)  and response (reproduction)
         of an organism	21

 5.2      Monroe Harbor and River Raisin sampling stations in
         1983-84	25

 5.3      Lake Erie sampling stations in 1983-84	26 "

 6.1      Illustration of toxicant fate	29

 6.2      Schematics of net estuarine circulation patterns in the
         River Raisin, Monroe Harbor 	   31

 6.3      River Raisin and nearshore Lake Erie sampling station
         locations for Surveys 1-3 (July, August and October,
         1983)	32

 6.4      Log probability plots of variables  as indicated for
         Survey 1 (July 1983), Station 2 	   37

 6.5      Comparison of observed and model computed alkalinity and
         hardness for Survey 1 (July 1983) 	   38

 6.6      Comparison of model results with July, September and
         October data for suspended solids 	   39

 6.7      Comparison of model results with July, September and
         October data for hardness 	   40
                                        \

 6.8     Comparison of model results with July, September and
         October data for alkalinity 	   41

 6.9     Comparison of model results with July (Survey 1) data for
         total copper (top), total chromium (middle) and total
         zinc  (bottom)	43
                                    xn

-------
                               FIGURES (CONT'D.)
Number                                                                   Page

 6.10    Comparison of model results with September (Survey 2) data
         for total copper (top), total chromium (middle) and total
         zinc (bottom)	   44

 6.11    Comparison of composite model results with data for
         Ceriodaphnia fecundity toxicity (top), emerald shiner
         (middle) and gizzard shad (bottom) growth toxicity, in
         toxic units	   45

 6.12    Monroe Harbor chloride mass balance for the upper reach of
         the study area in September, 1983 (Survey 2)	   50

 6.13    Monroe Harbor hardness mass balance for the upper reach of
         the study area in September, 1983 (Survey 2)	   51

 6.14    Monroe Harbor total suspended solids mass balance for the
         upper reach of the study area in 1983 (Surveys 1-3)	   53

 6.15    Monroe Harbor total PCB mass balance for the upper reach of
         the study area in July, 1983 (Survey 1)	   54

 6.16    Monroe Harbor total PCB mass balance for the upper reach of
         the study area in September, 1983 (Survey 2)	   55

 6.17    Monroe Harbor total PCB mass balance for the upper reach of
         the study area in October,  1983 (Survey 3)	   56

 6.18    Unaccounted loading of PCB homologs in the upper reach of
         the study area in July, 1983 (Survey 1)	.\. .  .   57

 6.19    Unaccounted loading of PCB homologs in the upper reach of
         the study area in September, 1983 (Survey 2)	   58

 6.20    Unaccounted loading of PCB homologs in the upper reach of
         the study area in October,  1983 (Survey 3)	   59

 7.1      Monroe Harbor water chemistry:, PCB homolog percent
         composition at selected stations  	   69

 7.2     River Raisin (master and plume study stations) and western
         Lake Erie station array, 1983-1984  	   80

 7.3     River Raisin station array, 1983-1984 	   81
                                    xiii

-------
                               FIGURES (CONT'D.)
Number                                                                   Page

 7.4     Bacteria"! uptake rate expressed as percent stimulation/-
         inhibition for 3 temperature ranges and 4 dilutions 	    83

 7.5     River Raisin primary production for three dilutions,
         July, 1983	    86

 7.6     River Raisin primary production before and after weighting
         for initial algal biomass, 1984 	    87

 7.7     Zooplankton grazing for three dilutions in the River
         Raisin, 1983-1984 	    90

 7.8     Cumulative probability of Ceriodaohnia fecundity in the
         River Raisin	    93

 7.9     Ceriodaohnia reproduction and grazing in the vicinity of
         the Monroe Wastewater Treatment Plant 	    94

 7.10    Cage designs for Monroe Harbor bioaccumulation studies  ....   101

 7.11    Uptake of total PCB (wet weight) by clams, fathead minnows
         and channel catfish	106

 7.12    Uptake of lipid content - corrected total PCBs by clams,
         fathead minnows and channel catfish 	   107

 7.13    River Raisin larval fish sampling stations, 1983-1984 	   Ill

 7.14    Instantaneous growth rates of River Raisin emerald
         shiners, 1983	   113

 7.15    Cumulative probability of median fish length in the River
         Raisin, 1983	114

 7.16    Spatial distribution of copper, zinc, Ceriodaohnia. and
         larval fish toxicity, River Raisin  	   115

 7.17    Location of Monroe Harbor sediment stations and transects  ...   130

 7.18    Total PCB concentrations in Monroe Harbor surficial
         sediments  (1983)   	   142

 7.19    Total PCB concentrations in Monroe Harbor surficial
         sediments  (1984)   	   144
                                     xiv

-------
                               FIGURES (CONT'D.)
Number                                                                   Page

 7.20    Total  PCB concentrations at depth intervals in a Monroe
         Harbor, Station 4 sediment core 	   146

 7.21    Total  PCB concentrations at depth intervals in a Monroe
         Harbor, Station T43/S sediment core (1984)   	   147

 7.22    Total  PCB concentrations at depth intervals in a Monroe
         Harbor, Station T47/C sediment core (1984)   	   148

 7.23    PCB homolog patterns in surficial  sediments from selected
         Monroe Harbor stations  	   149

 7.24    Flowchart of metal  adsorption/desorption experiments  	   152

 7.25    Sorptive and desorptive partitioning of HCB in Cvclotella
         systems	156

 7.26    HCB purged from Microcvstis sp. as a function of time
         compared to the total amount of HCB sorbed  (CO) initially
         (t = 0)	157

 A.I      Pritchard's models  for River Raisin/Monroe  Harbor 	   170
                                     xv

-------
                                    TABLES
Number                                                                   Page

 5.1     Monroe Harbor - Lake Erie Survey Schedule,  Objectives and
         Stations in 1983-84	   27

 6.1     Comparison of Zinc and Copper Partition Coefficients From
         Monroe Harbor and Other Freshwater Systems   	   46

 7.1     Monroe Harbor Water Chemistry:  Physical/Chemical
         Parameters	   64

 7.2     Monroe Harbor Water Chemistry:  General Chemistry 	   66*

 7.3     Monroe Harbor Water Chemistry:  Organic Contaminants  	   70

 7.4     Monroe Harbor Water Chemistry:  Metallic Contaminants 	   73

 7.5     Precision Estimated for Pooled Samples in Total and
         Dissolved Metals Analyses 	   78

 7.6     Decomposer. Ecosystem Function:  Relative Inhibition/-
         Stimulation (Percentage Change Relative to the Control)
         of Bacterial Uptake of Acetate by 100% River or Outfall
         Water	   84

 7.7     Inhibition/Stimulation of Zooplankton Grazing by Addition
         of 10, 25, and 50% River/Outfall (River Raisin) to Control
         Water (Lake Huron)	   89

 7.8     Ceriodaohnia Bioassay, Monroe Harbor, Cruise 3  	   92

 7.9     Total PCB and PCB Homolog Accumulation in Clams After 25
         Days of Exposure	t	104

 7.10    Abundance of Larval Fish Collected in the River Raisin,
         1983	116

 7.11    Ranking of Species Abundance in the River Raisin,
         1982-1984	117
                                     xvi

-------
                                    TABLES
Number                                                                   Page
 7.12    Simple Growth Rates of Gizzard Shad In the River Raisin,
         1983	118
 7.13    Metal Concentrations and Larval  Gizzard Shad Growth in
         River Raisin, 1983	119
 7.14    Mortality Rates of Monroe Harbor Gizzard Shad,  1983 and
         1984	119
 7.15    Histopathologic Lesions in Monroe Harbor Larval  Gizzard
         Shad, 1983-1984	120
 7.16    PCBs in River Raisin Fish, 1983-1984  	   122
 7.17    PCBs in Water, Sediment and Larval Gizzard Shad, River
         Raisin, 1983	125"
 7.18    Qualitative Sediment Survey - MDNR	131
 7.19    Physical and Chemical  Characteristics of Monroe Harbor
         Sediment Samples  	   135
 7.20    Sorption and Desorption Partition Coefficients  for
         Hexachlorobenzene (HCB) and Hexachlorobiphenyl  (HCBP) to
         Phytoplankton Harvested From Steady-State Chemostats  	   158
 B.I     Monroe Harbor Parameters and Codes  	   177
 B.2     Monroe Harbor Stations Sampled by Survey  	,.  .  .   178
                                    xvn

-------
                           ABBREVIATIONS AND SYMBOLS

AAS           Atomic absorption spectrophotometry
AOC           Area of Concern
BBL           Benthic boundary layer
BHC           Benzenehexachloride
cfs           Cubic feet per second
cms           Cubic meters per second
CSC           Computer Sciences Corporation
ODD           Dichlorodiphenyldichloroethane
DDE           Di chlorodi phenyldi chloroethylene
DDT           Di chlorodi phenyltri chloroethane
DO            Dissolved oxygen
ECD           Electron capture detector
EC50          Toxicant concentration causing sublethal biological  effects  to
                50 percent of the exposed organisms at a  specific  time  of
                observation
FI            Fish fillet
GC            Gas chromatography
GIDS          Grosse lie Data System
HCB           Hexachlorobenzene
HCBP          Hexachlorobiphenyl
IOC           International Joint Commission
                                    xvm

-------
                      ABBREVIATIONS AND SYMBOLS (CONT'D.)

Kj            Desorption partition coefficient
Ks            Sorption partition coefficient
LC50          Toxicant concentration killing 50 percent of exposed organisms
                at a specific time of observation
MDNR          Michigan Department of Natural Resources
NOEL          The highest effluent concentration at which no unacceptable
                effect will occur even at continuous exposure
PCB           Polychlorinated biphenyl
PPB           Parts per billion
PPM           Parts per million
QA            Quality assurance
QC            Quality control
TSS           Total suspended solids
TIC           Total inorganic carbon
TOC           Total organic carbon
TU            Toxicity units
USEPA/LLRS    United States Environmental Protection Agency, Large, Lakes
                Research Station
USFDA         United States Food and Drug Administration
WH            Whole fish
WWTP          Waste water treatment plant
                                    xix

-------
                                ACKNOWLEDGMENTS
     The U.S. Environmental Protection Agency staff, Computer Sciences
Corporation, and Cranbrook Institute of Science at the Large Lakes Research
Station contributed broadly to the project.  The EPA staff participated in
the planning, coordination, and execution of this multidisciplinary project.
EPA staff members William Richardson (Chief), David Dolan (presently with the
International Joint Commission), John Filkins, Michael Mull in, and Mary
Gessner (presently with the National Oceanographic and Atmospheric
Administration) also contributed as co-authors in many of the project
reports.  Computer Sciences Corporation (CSC), especially Kevin McGunagle and
David Griesmer, created the data files and managed the data base.  Elliott
Smith (presently with Raytheon Service Company) and staff of Cranbrook
Institute of Science provided much of the sample collection and analytical
support and helped with overall project planning.
SUMMARY REPORT

Primary authors of the Summary Report are listed below:
     Sections 1.0, 2.0
       3.0, 4.0, 6.2.1,
       6.3

     Section 5.0
     Section 6.2.2


     Section 6.4


     Section 7.1
Kenneth R. Rygwelski
Kenneth R. Rygwelski
William L. Richardson
James L. Martin
Kenneth R. Rygwelski
Abed R. Houssari
          %

V. Elliott Smith
William A. Frez
Stephen G. Rood
Computer Sciences Corp.



Computer Sctences Corp.
U.S. EPA/Large Lakes
  Research Station

U.S. EPA/Large Lakes
  Research Station

Computer Sciences Corp.


Raytheon Service Co.
                                     xx

-------
     Section 7.2             Russell G. Kreis         U.S. EPA/Large Lakes
                                                        Research Station
                             Joseph E. Rathbun        Raytheon Service Co.

     Sections 7.3.1,         Joseph V. DePinto        Clarkson University
       7.3.2, 7.3.4          Thomas C. Young

     Section 7.3.3           John C. Filkins          U.S. EPA/Large Lakes
                                                        Research Station

     Section 8.0             Kevin P. McGunagle       Computer Sciences Corp.
       Appendix B

     Appendix A              Kenneth R. Rygwelski     Computer Sciences Corp.


Editors of the Summary Report:

     Kenneth R. Rygwelski
     V. Elliott Smith
     Essential to the preparation of the final document was the dedicated
word processing support provided by Debra Caudill of Computer Sciences
Corporation.
PROJECT REPORTS

     Many persons and agencies have contributed their time and energy to this
study through their detailed project reports, for which the U.S.
Environmental Protection Agency is very appreciative.  All of these reports
are cited in the References (Section 9.0).  The following list identifies
these contributors, areas of their research, and sections that contain
summaries of their work:
     - University of California (EPA Grant No. 811578)

             Wllbert J. Lick, S. Maclntyre, and C.H. Tsai
             Analysis of sediment entrainment in River Raisin (Section 6.2)

     - Clarkson University (EPA Grant No. 810776)

             Joseph V. DePinto, T.C. Young, R.L. Autenrieth, and T.W. Kipp
             Physical and chemical characterization of sediments
                (Section 7.3)
                                     xxi

-------
     - Cranbrook Institute of Science (EPA Grant  No.  810232)

             V.  Elliott Smith1,  Susan P.  Hendricks2,  J.E.  Rathbun1,
                S.G.  Rood1, and  K.R.  Rygwelski3
             Metals,  organics, and general water  chemistry (Section  7.1);
             Zooplankton and bioaccumulation bioassays (Section 7.2)

     - Manhattan College (EPA Grant No.  810779)

             Dominic  M. Di Toro, J.P. Connolly, R.P.  Winfield,  S.M.  Kharkar,
                U.A.  Wolf, C.J.  Pederson,  J.R. Blasland,  and  J.O.  Econom
             Exposure and toxicity modeling (Section  6.0)
             Evaluation of bioassays (Section 7.2)

     - University of  Minnesota (EPA Grant No. 810775)

             Donald C. McNaught, S. Bridgham, and C.  Meadows
             Bacterial, phytoplankton, and zooplankton bioassays
                (Section 7.2)

     - Ohio State University

         Civil  Engineering Department (EPA Grant  No.  811731)
             Keith W. Bedford, C.M. Libicki, R. Van Evra,  0.  Wai,  and M.
                Abdelrhman
             Hydrodynamics (Section 6.2)

         Center for Lake Erie Area Research (EPA  Grant No. 810808)
             Laura A. Fay, P.C.  Stromberg, and J. Hageman
             Toxicity effects on larval  fishes (Section 7.2)

     Others have contributed to the project by providing support in the
field.  The Michigan  Department of Natural Resources  assisted in sample
collections and conducted a qualitative sediment  survey of the study area.
Their staff collected resident adult fish for analysis and participated in
the water sampling.  WAPORA, Incorporated of Cincinnati, Ohio conducted
surveys of the River  Raisin and provided the project  with velocity readings
and Hydrolab  data.
^Currently with Raytheon Service Company.
^Currently with the University of Michigan.
3Currently with Computer Sciences Corporation.
                                    xxn

-------
                                  SECTION 1.0

                                  INTRODUCTION
1.1  PURPOSE

     The evaluation and control of toxicants In the aquatic ecosystem Is an
Important concern of the U.S. Environmental Protection Agency (USEPA).
Currently, the assessment and control of toxicants In effluents focuses
primarily on individual toxicants.  This chemical-specific approach to toxics
control has involved the use of laboratory-generated water quality criteria
or standards to limit specific toxicants directly.  In addition,
concentrations of chemical-specific toxicants have been described both
spatially and temporally in receiving waters using fate and transport models
in order to evaluate the impact on the receiving water.  Toxicity, as
measured through toxicity testing of effluents using aquatic test species, is
an alternative to the chemical-specific approach for establishing water
quality criteria.

     The USEPA Office of Research and Development in cooperation with the
Michigan Department of Natural Resources and the USEPA Great Lakes National
Program Office chose the lower River Raisin, Monroe, Michigan to serve as a
representative system to evaluate the two generic approaches to water quality
assessment.  This water resource is used for fish and wildlife habitat, sport
fishing, and recreation.  Because of their proximity to pollutant inputs from
industrial and municipal sources, nearshore areas such as this are often
adversely affected.  The International Joint Commission (IJC) identified the
following major pollutant problems and resulting consequences in the River
Raisin (Great Lakes Water Quality Board, 1985):

         - Conventional pollutants
         - Heavy metals
         - Toxic organics
         - Contaminated sediments
         - Fish consumption advisories *
         - Biota impacted
         - Aesthetic deterioration

Due to the degraded condition of the water and sediments and to high
concentrations of PCBs in the fish, the IJC has designated the lower River
Raisin as an "Area of Concern".

-------
     The primary goal of the Monroe Harbor (River Raisin) study has been to
investigate the feasibility and development of methods that address toxicity
in the effluents and in the receiving water based on toxicity testing using a
wide range of test species.  Also, since the area was known to be impacted by
toxic substances like PCBs and heavy metals, these parameters and others were
analyzed in various media, the hypothesis being that part of the observed
toxicity could be explained by these chemical-specific parameters.
Consistent with recommendations in the Technical Support Document for Water
Quality-Based Toxics Control (USEPA, 1985), an important goal in the study
was to demonstrate an integration of the whole-effluent or toxicity-based
approach and the chemical-specific approach during the study period
1983-1984.


1.2  ORGANIZATION OF THE SUMMARY REPORT

     Sections 2 and 3, Conclusions and Recommendations, focuses on the
     toxicity-based and chemical-specific approach.

     Section 4, Background, briefly describes the River Raisin drainage
     basin, hydrology, local industrial activity, and industrial landfills.

     Section 5, Pro.iect Objectives and Approach, states project goals and
     provides information that is helpful to understanding toxicity-based and
     chemical-specific approaches.  The section also serves as an
     introduction to the various study areas.

     Section 6, Water Quality Models of Metals. Toxicitv. and PCBs.
     demonstrates the feasibility of predicting toxicity in the receiving
     water by treating toxicity measured in effluents as a state variable.
     The section also addresses the behavior of zinc, copper and PCBs in the
     receiving water.  These parameters were used in some of the
     "chemical-specific" approaches to evaluate toxicity in the water.

     Section 7, Component Studies, is divided into water chemistry*
     biological studies and sediment chemistry:

          Section 7.1, Water Chemistry, describes the water quality surveys,
          sampling procedures and stations  sampled.  It also includes a
          description and evaluation of the analytical methodology used.
          Particular emphasis is put on describing areas in these
          methodologies that require special consideration either due to the
          complex nature of the industrial  effluent  samples or problems of
          trace analyses in a field situation.

          Section 7.2, Biological Studies  in Monroe  Harbor (River Raisin),
          Michigan, describes the following bioassays used in the toxicity
          approach of evaluating the water  quality:  bacterial substrate
          uptake, phytoplankton photosynthesis  inhibition, zooplankton
          grazing rate, zooplankton reproduction, and larval fish studies.

-------
     Several  biological studies addressed toxic endpoints for a specific
     chemical (PCB) in clams,  fathead minnows and channel catfish by
     conducting in-situ bioaccumulation research.

     Section 7.3,  Sediment Chemistry, provides insight to the degree of
     heavy metal  and organochlorine contamination observed in the
     sediments and some discussions on the adsorption/desorption
     kinetics that govern the potential for these bound chemicals to be
     released to the overlying waters.  Sorption/desorption experiments
     were performed to determine equilibrium partition coefficients and
     the rate of and extent of desorption of hexachlorobenzene and
     hexachlorobiphenyl with several species of phytoplankton common to
     the lower River Raisin.

Section 8, Data Base Documentation Summary, describes important
considerations when undertaking the job of storing and retrieving the
diverse sets of data that are produced in a multidisciplinary study such
as the Monroe Harbor (River Raisin) study.

-------
                                  SECTION 2.0

                                  CONCLUSIONS
     A number of bioassays and aquatic test species were used to assess water
quality in Monroe Harbor.  Also,  chemical  analyses provided data that allowed
relationships to be drawn between toxicity-based water quality assessment and
specific toxic chemicals.  Although the results are specific to Monroe
Harbor, the methodologies and approaches may have application in assessing
water quality in similar systems.


2.1  INTEGRATION OF THE WHOLE-EFFLUENT AND CHEMICAL-SPECIFIC APPROACHES

     1.  Zinc, copper, and chromium appeared to be toxic to different test
species and affected specific ecosystem functions, either singly or in
combination.  Chromium appeared to be the most influential factor in
suppressing phytoplankton productivity.  Copper and zinc, corrected for
hardness, exerted the greatest inhibition on zooplankton grazing, whereas
zinc, corrected for alkalinity, appeared to inhibit zooplankton reproduction.

     2.  Growth rates of gizzard shad and emerald shiners were greatest in
the upper and mid-portions of the River, with considerably lower growth rates
downstream.  Growth rates exhibited an inverse relationship to concentrations
of zinc, chromium, and copper in sediment.


2.2  WHOLE EFFLUENT APPROACH

     1.  The attempt to  integrate the conventional chemical and biological
methods of water quality evaluation has demonstrated the need to place the
bioassay and in situ observations on a common basis so that the
intercomparison can be made.  Multiple types of bioassays appear to offer a
method of examining various expressions of toxicity.  It is interesting that
the zooplankton, larval  fish assay methods, and in situ observations yielded
similar toxic unit distributions.  This Suggests that in fact toxicity exists
and can be quantified in the River Raisin.

     2.  Toxicity in the receiving water is usually not conservative or
additive.  However, as a first approximation, a probabilistic, deterministic
fate and transport model assumed that zooplankton and larval fish toxicities
were conservative and additive.  These model predictions were consistent with
observations in the River Raisin.  Without further analysis of the underlying

-------
mechanisms and processes involved that affect toxicity, it may not be
appropriate to assume that the conservative and additive nature of these
toxicities extend into Lake Erie or elsewhere.

     3.  In most cases, the greatest relative toxic or inhibiting effects
were observed in the vicinity of the Wastewater Treatment Plant and lakeward.

     4.  Comparative results of zooplankton reproduction and grazing
bioassays indicated that reproductive inhibition was a direct result of
physiologic exposure to toxicants rather than as a result of grazing
suppression.

     5.  Zooplankton grazing and reproduction assays were two of the most
informative bioassays conducted.  Results could be interpreted spatially and
were reproducible.  Typically, dose-response functions exhibited increasing
toxic effects with increasing dose concentrations.


2.3  CHEMICAL-SPECIFIC APPROACH

2.3.1  Bioaccumulation/Bioavailabilitv Studies

     1.  Common carp, emerald shiners, smallmouth bass, gizzard shad, and
mirror carp (in various numbers for each species) exceeded the USFDA action
level of 2.0 mg/kg total PCBs.  Based on these results, the Michigan
Department of Natural Resources re-issued a fish consumption advisory for the
lower River Raisin.

     2.  Eighteen of twenty whole common carp from the River Raisin exceeded
the USFDA action level of 2.0 mg/kg for PCBs.  Additionally, seven of eleven
common carp fillets exceeded the action level.  Generally, analyses of whole
fish exhibited greater concentrations than did fillet analyses.

     3.  Bioavailability experiments, although preliminary, demonstrated that
a significant amount of sediment-bound HCBP was bioaccumulated by algae
maintained in contact with the sediments for up to 24 hours.  This exchange
of material, which has major implications relative to organic contaminant
transport and bioaccumulation, was attributed to the preferential sorption of
nonpolar organics to high organic content particulate material.

     4.  Chlorinated organics partitioning experiments using three species of
phytoplankton revealed that partition coefficients for Scenedesmus (green
alga) and Microcvstis (blue-green alga) .were similar in magnitude to the
experimental partition coefficients values obtained using River Raisin
sediments; however, Cvclotella (centric diatom) partition coefficients were
approximately an order of magnitude higher than those of the other algae and
River Raisin sediments.

     5.  Freshwater bivalve mollusks were suitable organisms for i_n situ PCB
bioaccumulation studies.  Their accumulation rates and capacities in Monroe

-------
Harbor were comparable to those of fathead minnows and channel  catfish,
especially when the data are corrected for tissue lipid content.

     6.  Clams caged on the sediments did not accumulate more total  PCBs than
did clams caged in the water column.   This suggests that the PCBs associated
with the bedded sediments were much less bioavailable than those present, in
dissolved or part'iculate phase, in the water column.

     7.  Thirty-five days of exposure to Monroe Harbor water or sediments in
1984 was not sufficient for clams, fathead minnows, and channel catfish to
accumulate total PCBs to equilibrium concentrations.  PCB homologs patterns,
however, were accumulated to apparent equilibrium after only two to four days
of exposure.

     8.  The ability of clams to acquire PCB homolog patterns resembling
those in their immediate environment suggests that clams may be useful for
identifying sources of PCBs which exhibit distinctive homolog patterns.

2.3.2  Fate and Transport Model of Chromium. Zinc, and Copper

     1.  The 22 segment model assumed the metals to be conservative.  This
was a valid, first-cut assumption since no systematic deviation was observed
between model calculated and observed spatial profiles.  This indicated that
no other significant transport phenomena (such as resuspension) or sources
and/or sinks were present.

     2.  Prediction of dissolved metals or exposure concentrations based on
total metals data alone would be difficult because of the wide range of
partition coefficients observed.


IA  RELATED STUDIES

     Although the following studies were conducted in part to support the
evaluation of the two approaches, they have stand-alone significance in
establishing the present environmental status of the Monroe Harbor.
Information on data management considerations is also included.

2.4.1  Descriptive Biology

     1.  Gizzard shad was the dominant larval fish species in the River
Raisin during 1983-1984.  Emerald shiners, common carp, white bass/white
perch, and freshwater drum were also common.

     2.  Histopathological examination of larval gizzard shad indicated that
lesion incidence was about 90% in the River Raisin; however, a  similar lesion
incidence was observed in gizzard shad samples collected in nearshore Lake
Erie.  Several types of  lesions were observed in populations; epithelial
necrosis was the most abundant lesion afflicting numerous organs and tissue
types.

-------
2.4.2  Water Chemistry

     1.  The project goal of attributing toxic effects to specific water
quality factors in Monroe Harbor was an ambitious one, and the results were
only partly successful.  Success depended on selecting a few contaminants for
study which were-both measurable by the LLRS and were capable of causing or
influencing certain biological effects.

     2.  The operational work of describing spatial and temporal gradients of
water quality in Monroe Harbor required a good deal of flexibility on the
part of field crews, equipment and methods.  The surveys were labor-intensive
and, therefore, relatively expensive.

     3.  Matrix problems were encountered in the organochlorine analysis of
Monroe Harbor water.  Unknown interferences in some samples were not removed
by our clean-up method, and sometimes prevented accurate quantitation of
PCBs.

     4.  Filtration of large-volume water samples proved to be an inefficient
method of collecting particulate fractions for organochlorine analysis.
Cross-contamination appears to have occurred between some samples due to
inadequate cleaning of the filtration system.

     5.  Despite some interference problems, it was feasible to carry out
high resolution gas chromatography of most PCB congeners and other
organochlorines in over 200 water samples.  These results will be useful in
modeling the behavior of individual compounds.

     6.  Concentrations of PCB and metals (chromium, copper, and zinc) were
typically similar within a factor of 10 at the upstream boundary of Monroe
Harbor (Station 1) and in Lake Erie (Station 11).  Seasonal changes in
concentrations at both sites were within a factor of 2 or 3.

     7.  Concentrations of contaminants within Monroe Harbor fluctuated
greatly at times,  in response to changes in loadings, flows or resuspension
events.  Measurements based on grab samples and composites (3/24 hours)
agreed within 10%.

     8.  Metals concentrations were well below the EPA Water Quality Criteria
(1976) values for drinking water.  However, copper and zinc values for the
river were generally within an order of magnitude of those which caused
chronic effects in Crustacea (Mount and Norberg, 1984).  Some relatively high
PCB concentrations were recorded in thevturning basin.  Levels of other
parameters (ammonia, nitrite, chloride and hardness) were also elevated but
not exceptional in Monroe Harbor.

     9.  The quality control measures applied here to chemical analysis of
water proved to be more useful for documenting the quality of results than
for managing the field and laboratory work in progress.  Most of the final
PCB data were not available until after the field surveys were completed.

-------
2.4.3  Sediment Chemistry

     1.  The average values of total  copper,  zinc,  and chromium (112.7,
231.7, and 91.3 yg/g, respectively)  in River Raisin bottom sediments are
elevated when compared with sediment  data from open water areas of the Great
Lakes.  Within the system, transect  42 contained elevated metals levels
relative to other locations.

     2.  In regard to relationships  among River Raisin bottom sediment
parameters, a principle component-factor analysis explained variation among
the data.  The four factors were grouped as follows:

     Factor 1 (35% of common variance) - total metal levels of the sediments,
              pore-water Zn concentration, sediment depth, and inorganic
              carbon content;

     Factor 2 (27% of common variance) - physical properties (moisture,
              density, porosity), loss on ignition, and transect;

     Factor 3 (20% of common variance) - pore water metal concentrations;

     Factor 4 (16% of common variance) - organic carbon content of sediments.

These results suggest that the metals originated from sources in spatial
proximity, and that relatively rapid partitioning to particulate matter
followed by sediment transport and deposition may be primarily responsible
for their distribution within this aquatic system.

     3.  Bottom sediments collected from the River Raisin during this study
indicated a pronounced spatial variation in physical characteristics.  Bottom
sediments upstream of the turning basin were found to consist primarily of
quartz, sand, and gravel with higher-dry densities and lower porosities,
while sediments from the turning basin and downstream were composed
predominantly of silt and fine sand with some clay and organic detritus.
                                                                  \
     4.  Metal adsorption equilibria for River Raisin sediments showed a
linear dependence of total particulate metal content on soluble metal
concentration, with partition coefficients (48-hour equilibration) of ~50,
30, and 25 L/g for Cu, Cr, and Zn, respectively.

     5.  Desorption of Zn was completely and rapidly reversible (24-48 hour)
in contrast to that of Cr, which was much slower; Cu desorption could be
described by assuming the metals were retained in rapidly reversible
("loosely-bound") and slowly-released, resistant ("tightly-bound") forms.
However, overall desorption did not reach a meta-stable equilibrium even when
metals were permitted to desorb for periods of as long as 24 days.

     6.  Values obtained for River Raisin experimental sediment-water
partition coefficients were 14 L/g for HCB and 40 L/g for HCBP.  These values
are within the range of experimental results for other Great Lakes sediments
and are within a factor of two of partition coefficients calculated from


                                      8

-------
empirical regressions based on octanol-water partitioning and organic content
of the sediment.

     7.  PCBs were present in the sediment of the lower River Raisin with the
higher PCB concentrations found in the turning basin and downriver to the
Edison water intake.

     8.  The vertical distribution of PCB concentrations in the sediment
varied from station to station.

     9.  Characteristic of Lake Erie sediment analyzed, the homolog pattern
of sediment collected downriver of the Edison intake suggest an enrichment of
the hexa and hepta component of the total PCB mass.

2.4.4.  Data Base Documentation

     1.  Defining data requirements at the start of a project can save time
and confusion.  Decisions and procedures should be documented.

     2.  It is difficult to maintain multiple copies of the same data.
Inconsistencies will inevitably be found and will have to be corrected.

2.4.5  PCB Input-Output Mass Balance Model

     1.  Unaccounted for loadings of PCBs, enriched in the tri- and
tetrachlorobiphenyl compounds, apparently occur in the Monroe Harbor/River
Raisin.

     2.  PCB loadings from Mason Run and resuspension of PCB contaminated
sediments did not appear to be significant contributors to the undefined
source of PCBs.

-------
                                  SECTION 3.0

                                RECOMMENDATIONS
3.1  BIOLOGY

     1.  All physical, chemical, and biological analyses should be conducted
on the same sample for any given environmental matrix sampled.  When analyses
are coordinated in time and space, the data base generated is more complete
and cohesive and results are more easily interpreted, compared, and
statistically assessed.  In the case of biological analyses and bioassays,
causative or influential factors associated with responses observed can be
more easily delineated.

     2.  For each site, a suite of bioassays should be employed which spans a
range of trophic levels.  Bioassay results may indicate which species exhibit
the greatest sensitivity and, by inference, which trophic levels may be the
most adversely impacted in a given system.  At a minimum, phytoplankton,
zooplankton, and fish should be used as test specimens.

     3.  The suite of bioassays employed should represent a range of
ecosystem functions to be used as assay endpoints.  Functional ecosystem
endpoints may include: metabolism, feeding, reproduction, growth, behavior,
and lethality.  In some systems it may be advantageous to include
mutagenicity and carcinogenicity as assay endpoints.  Both acute and chronic
assays should be applied as well.
                                                                  i
     4.  Bioassays must be carefully selected and pertinent to the system
being investigated; i.e., organic compounds may be the dominant contaminants
in some systems, whereas heavy metals may be in others, thus dictating the
type of assay required.  When contaminant factors are unknown, an array of
bioassays should be employed.

     5.  Bioassays are typically effort-intensive and costly, again requiring
careful selection.  The development of additional rapid and cost-efficient
bioassays is greatly needed.  Comparatively, bioassays are considerably less
expensive than complete organochlorine, metals, and other water quality
analyses used to support a chemical-specific approach.

      6.  Potential control or reference sites must be carefully examined,
validated,  and selected.  Usually all measurements from survey stations are
compared to those at this site to determine differences from an unimpacted
area.  In assays, the control site is doubly critical in that control waters


                                      10

-------
are usually tested as a series of dilutions in the assay, and survey response
values are reported relative to the control response.

      7.  In certain assays, many variables come into play which may affect
the response endpoint relative to a control.  Certain species of
phytoplankton, for example, not only respond to toxicants but may also
respond to the absence or presence of certain nutrients.  Although the
approach is somewhat controversial, monoclonal cultures could be employed
where population characteristics and kinetics are known.  The confounding
influence of nutrients could be avoided by saturating culture media with
primary nutrients and thus limiting the responses observed to those solely
associated with toxicants.  Filtration of all resident phytoplankton
assemblages from the sample to be tested is also recommended.

      8.  The most informative data concerning body burden concentrations of
organic compounds in biota result from a sampling scheme designed in a food
chain context.  Bioaccumulation factors can be calculated for each trophic
level and will indicate the greatest relative contribution of contaminants to
each trophic level.  Similarly, trophic levels used in bioassays should also
be correlated with trophic levels used in food chain studies.

      9.  At least 10% additional effort must be devoted to replication of
bioassays.  Result variability confuses interpretation and bioassay
usefulness.  All field and laboratory procedures must be consistent to
decrease variability in results.

     10.  All bioassays should be conducted using a dilutional series to
generate dose-response functions and to allow the calculation of LC50 and
EC50 values.  These values are critical to derive the probabilistic modeling
and exposure concentrations expressed as toxic units.

     11.  It is very difficult to determine cause-effect relationships and
discern influential factors in response results when investigating complex
mixtures of toxicants in water and/or sediment.  Statistical procedures for
preliminary screening are correlation analysis, multiple regressioi^ analysis,
analysis of variance, and analysis of covariance.  As pointed out previously,
the data bases must be complete and compatible with each other to execute
these statistical analyses.  If this is the case, results will then be
amenable to further analyses using multivariate techniques, such as principal
component analysis, factor analysis or correspondence analyses.  These
statistical techniques in combination with exposure probability analysis
appear to be the most informative in indicating influential effects of
complex toxic mixtures.

     12.  Organisms used in bioassays must be sensitive to environmental
conditions but not so sensitive as to produce sharp threshold responses which
yield only "live" versus "dead" results.  Specifically, the species selected
for bioassays must be sensitive within the range of concentrations tested and
must be discriminating and resolute in their dose-response functions, or else
EC50 or LC50 values cannot be derived.  When sharp dose-response curves are
                                      11

-------
obtained, additional test:control  water ratios in the dilutional  series may
be required.

     13.  When bioassays are conducted over a period of time in some acute
and chronic assays, resupply water (test water) must be added to the test
cultures.  At the-present, no information is available in regard to effects
of sample holding time on resupply water and this should be investigated.

     14.  Contaminant depuration kinetics should be investigated, in addition
to kinetics of accumulation.  This would provide information on organism
detoxification capabilities and on ecosystem responses to remedial  actions.

     15.  Some expression of physiological "stress" should be determined for
organisms caged in severely degraded environments.  To the extent that caged
organisms are stressed, we may be less confident in the bioaccumulation
results.

     16.  The influence of water temperature, organism age, and reproductive
status on bioaccumulative capability should be investigated.

     17.  It seems clear from this work that desorption kinetics can play a  -
role in determining the exposure potential and/or toxicity of sediment-bound
contaminants in aquatic systems; therefore, further research is needed to
define the relevant mechanisms and correctly parameterize the desorption
process.

     18.  In order to determine the spatial and temporal variations in
exposure of water column biota to an in-place contaminant, there are several
physical and chemical processes that must be understood at a mechanistic
level.  These processes include:  the resuspension and subsequent deposition
of the particulate matter; the particle-particle interactions of resuspended
sediments (i.e., aggregation-disaggregation); and the interfacial processes
which determine the phase/speciation of a contaminant either in-place or
in-suspension.  Furthermore, there are several potentially significant
interactions among these physical and chemical process.  For exampTe, it is
well known that chemical properties, such as pH and ionic content, have a
very significant direct effect on metal adsorption; however, these properties
can also affect particle aggregation, thus causing an indirect impact on
metal partitioning by altering sorbent surface area.

     19.  In order to better understand and refine food chain models of
toxicants, additional research is needed  to understand the factors governing
the spatial and temporal distribution of*contaminants among the  various
particulate matter types  in aquatic systems.  For example, abiotic
particulate matter  is primarily allochthonous in origin, is low  in organic
carbon content and has a relatively high  density, while the biotic solids  are
primarily autochthonous primary producers, contain significantly more organic
carbon and have a much lower density.  These property differences have major
implications relative to contaminant transport, bioaccumulation, and
toxicity.  Therefore, more  studies are needed in the following areas:  the
kinetics of contaminant uptake and release from particulate matter as a


                                      12

-------
function of the solid phase properties; the rate and extent to which
sediment-bound contaminants become bioavailable to various trophic levels;
and the field Measurement of the distribution of a contaminant between biotic
and abiotic particulate matter for purpose of model calibration and field
testing.


3.2  WATER CHEMISTRY

     1.  In other such studies, designed to correlate water chemistry with
effects, considerably more initial effort should be spent on exploring the
whole range of water or sediment contaminants which might contribute to
toxicity.  Synergistic effects among pollutants even at lower concentrations
might be significant in systems like Monroe Harbor.

     2.  Future results would be improved by using more efficient methods to
collect particulate matter, extract organochlorines from larger volumes of
water, and clean extracts to remove interfering substances.  These methods
should be established and tested on trial samples before surveys begin.

     3.  In view of elevated PCB concentrations in the Monroe Harbor turning.
basin, sediments and groundwater should also be investigated as possible
sources of contaminants in this system.  The source of high levels of zinc in
the Monroe Harbor Wastewater Treatment Plant effluent should be investigated
as well.

     4.  Sample analysis should follow surveys as soon as possible so that
the quality control results for blanks, duplicates and spiked samples can
influence field strategies and methods in subsequent surveys.  Unexpected
changes in water quality or field conditions may require adjustments in
methods used for collecting or processing samples.

     5.  In the future, preliminary studies should be planned to characterize
the chemistry (and biology) of control station water to be used in bioassays.
Ideally, control water should be similar in matrix to sample waters but have
the lowest possible concentrations of the contaminants in question.


3.3  WATER QUALITY MODELING

     1.  Total toxicity was successfully predicted in the receiving water
(River Raisin) using the transport results of a 22-segment fate and transport
model.  Although this was not attempted,here, it is possible that within the
framework of the same 22-segment model, toxicity, resulting from copper and
zinc only, could be predicted spatially in the Monroe Harbor receiving water
using data already available from the study.

     2.  Modeling of data from a high-flow period should be conducted.
Sediment resuspension potential should again be examined to see if it is
significant.
                                     13

-------
     3.  The PCB model  results suggest that the source of PCBs loadings not
accounted for 1s local.  Since this PCB load has been characterized
chemically to the congener level  of detail, it may be possible to link its
composition to that of suspected  sources by chemically matching PCB congener
patterns.

     4.  A possible source of the unaccounted for PCB loading to the lower
River Raisin may be contaminated  groundwater.  Due to the complicated
hydrology in the area,  a complete understanding of groundwater flow under
different hydrological  events would be a first step in assessing this
possible source.


3.4  DATA BASE DOCUMENTATION

     1.  Project Management should meet with data processing personnel to
define data requirements for a project in the planning stages.  Data
processing procedures should be well defined and documented.

     2.  There should be only one "official" copy of the data.  Different
procedures for processing the information should be available to meet
different requirements of data usage.
                                      14

-------
                                  SECTION 4.0

                                  BACKGROUND
4.1  BASIN DESCRIPTION

     The River Raisin drains an area of 1,070 square miles (2,771 square
kilometers) and discharges into Lake Erie at Monroe, Michigan.  A portion of
Michigan's southeastern lower peninsula and the northeastern portion of
Fulton County, Ohio lie within the boundaries of the basin.  The drainage
basin narrows to a width of 2.5 miles (4 km) for the last 15 miles (24 km) of
the river reach.  The study area included approximately 2.5 miles (4.0 km) at
the end of this strip (Figure 4.1).


4.2  HYDROLOGY

     Monroe County has little topographic relief.  There is a gentle slope
southeastward from a maximum elevation of 730 feet (223 m) in the northwest
corner to 572 feet (174 m) at Lake Erie.  This gradual decline of only 158
feet (48 m) in nearly 26 miles (42 km) explains the low velocities of streams
located in the county (Mozola, 1970).

     Runoff during rain events creates rapid increases in flow and very
turbid waters primarily due to clay till in the drainage basin.
Approximately one-third of the total rainfall runs off through the river
system.  Relative to other areas in Michigan, erosion in the River Raisin
basin is considered to be high.  The U.S. Department of Agriculture estimated
that 8.3 to 10.8 tonnes of topsoil per hectare per year are lost (Michigan
DNR, 1979).

     Much of the area adjacent to the River Raisin is prone to flooding.  A
large portion of the eastern fringe of the city of Monroe was marshland at
one time.  Over the last thirty years, approximately 80% of the marshlands
have been filled in for industrial and recreational uses.  The river banks
and surrounding areas at the mouth of the River Raisin are man-made (Monroe
County Drain Commission, 1984).

     The port of Monroe is served by a dredged channel 15,800 feet (4.8 km)
long, 300 feet (91.2 m) wide and 21 feet (6.4 m) deep from Lake Erie to the
mouth of the River Raisin.  From the river mouth to the turning basin, there
is a dredged channel 8,200 feet (2.5 km) long and 200 feet (60.8 m)  wide
(Michigan DNR, 1979).


                                     15

-------
                 WEST END OF LAKE ERIE
                                         "    '
'   —\
         Figure  4.1.  Monroe Harbor - River Raisin  Study Area.

-------
     Water flow in the lower River Raisin is strongly influenced by Detroit
Edison's water intake and natural forces.  The Detroit Edison plant is
located at the mouth of the River Raisin on the south side of the river.
During low river flow, the entire flow of the river is pumped through the
plant along with water from Lake Erie.  This causes a flow reversal at the
mouth of the river.  Lake level fluctuation due to seiches can also cause net
flow reversals in the river.  The lower River Raisin/Monroe Harbor is also a
fresh water estuary.  Top and bottom flows are often opposite in direction
due to temperature-induced water density differences at certain times of the
year (see Section 6.2.1 for a full discussion).


4.3  INDUSTRIAL DEVELOPMENT

     Monroe, at the River Raisin mouth, is the most populous and
industrialized city in the basin.  Much of the industry is associated with
automobile manufacturing in nearby Detroit and southeastern Michigan.
Additional industries in the area are primary metals, fabrication of metal
products, machinery and transportation equipment, manufacture of paper
products, chemicals, furniture, food processing, and dairy related industries
(Michigan DNR, 1979).

     Industrial dischargers that were sampled during the 1983-1984 surveys
were the Monroe Metropolitan Water Pollution Control Facility and Ford Motor
Company, Monroe Stamping Plant.


4.4  INDUSTRIAL LANDFILLS

     Two industrial waste sites border the study area.  These two sites have
been ranked in Michigan's Sites of Environmental Contamination Priority
List - Act 307 (1987).  The Ford Motor Company waste pile located on the
southeast corner of their property is relatively close to the downriver side
of the turning basin.  A waste lagoon is also included in this contaminated
area ranking.  Copper, zinc, lead, and chromium have been found at» the site
and both soil and groundwater have been affected.  The Port of Monroe
landfill has been linked to PCBs, benzene, xylene, cumene, and ethyl benzene
contamination.  The areas affected are groundwater, soil, and wetlands.
Operation of the Port of Monroe Landfill spanned a period of at least 23
years.  The landfill operation closed in 1975.  Approximately 600 acres of
former swampland were filled to depths averaging 10-20 feet.  One count of
rail disposal activity indicated approximately 8,000 cars per year.  This
site operated for 20 years without significant government oversight.
Numerous complaints were directed toward the landfill operator, Heckett
Engineering (General Disposal Company).  Various reports claim that
considerable amounts of the following illegal materials were disposed of:
household garbage, paint thinners, paint sludges, wood, paper, cutting oil,
coal tar, and "black" and "white" sludge.  The disposal company stated that
the following materials were solicited and disposed of:  foundry sand,
brickbats, broken concrete, cinders, fly ash, glass, construction debris,
dirt, gravel, rock, clay, and slag (MDNR, 1985).


                                     17

-------
                                  SECTION 5.0

                        PROJECT OBJECTIVES AND APPROACH
5.1  OBJECTIVES

     The purpose of the Monroe Harbor (River Raisin) study was to initiate
the development of integrated water quality methods which combine
conventional  single chemical transport and fate modeling with water quality
evaluation procedures.  The latter are based upon biological  evaluations such
as bioassays, in situ population surveys, and current water quality criteria.
The research products include innovations for analyzing and synthesizing
bioassay and population data that provide compatibility with transport and
fate models.  These methods were developed within the context of a joint
multidisciplinary investigation of the River Raisin.

     Specific objectives in the study were:

     - To explore the possibility of modeling toxicity as a state variable.

     - To develop and evaluate bioassays that utilize aquatic test species
       for the purposes of water quality assessment.

     - To develop and test a fate and transport model to predict zinc,
       chromium, and copper exposure.

     The following were research products that were obtained in part to
support the above objectives.  They are also significant stand-alone
contributions in describing the present status of Monroe Harbor.

     - Analysis of chemical residues in water, sediment, and biota for
       surveillance purposes and to update the status of Monroe Harbor as an
       "Area of Concern".

     - To investigate the importance of in-place pollutants (metals and PCBs)
       relative to other sources in a mass balance framework.
                                     18

-------
5.2  APPROACH

5.2.1  Water Quality Assessment

     A chemical discharged to the surface water Is subject to various
pathways of transport and transformation that ultimately result in a
concentration of the chemical in the water,  sediment,  and biotic compartment.
The effects of these exposure concentrations is then often measured in terms
of some impact on the uses of that body of water (bathing, fish consumption,
municipal water supply) and the aquatic ecosystem.  The impact to the
ecosystem may include mortality of key species through direct action of the
chemical on the organism or longer term inhibition of reproduction, growth,
migration pattern, etc.  The following two sections present a brief overview
of the toxicity-based and chemical-specific approaches used in the study to
address the effects that discharged chemicals have on the aquatic biota.

5.2.1.1  The Whole-effluent (Toxicity-based) Approach--

     The EPA Office of Water Enforcement and Permits and Office of Water
Regulations and Standards (1985) define whole-effluent toxicity as an
aggregate of toxic effects of an effluent measured directly with a toxicity  .
test.  An effluent sample is collected and diluted in test chambers; usual
dilutions used are 100%, 30%, 10%, 3%, 1% and a control.  It is often
desirable to understand the interaction of receiving water and effluent, so
dilution water used in the test chambers is often receiving water.  Test
organisms are placed in the test chamber for a specified period of time.  At
various points during the test, the effects of the test samples on the
organism are assessed (the endpoint can be mortality, lower fecundity,
reduced growth rates, etc.).  These lab data are used to establish an
exposure (% effluent in test chamber) and response functional relationship.
Interpolation or extrapolation using this relationship is performed to find
the dilution fraction corresponding to a 50% response in the population.  For
permit limits, it is usually stated either as an LC50 (the effluent
concentration at which 50% of the test organisms are killed) or a No Observed
Effect Level or NOEL (the highest effluent concentration at which no
unacceptable affect will occur even at continuous exposure).  The measurement
of whole-effluent toxicity can then be used to limit discharge of toxicants
in an effluent.  Toxicity itself is a limited parameter in the effluent water
quality criteria.

     An inverse relationship exists between the LC50 or NOEL and the toxicity
of the effluent (the lower the LC50 or NOEL, the higher the toxicity of the
effluent).  For example, an effluent that kills 50% of the organisms at a 10%
dilution of the effluent (LC50 * 10%) is more toxic than an effluent that
kills 50% of the organisms at a 20% dilution (LC50 * 20%).  To put these
toxicity data on a direct relationship to an adverse effect, a toxic unit was
defined:
                                     19

-------
         TU  •   LcVo? NOEL                                    E(*uat1on 5J

         (where LC50 or NOEL is expressed as percent effluent
         in dilution water)


Therefore, an effluent with a LC50 of 10% contains 10 TUs, while an effluent
with a LC50 of 20% contains 5 TUs.  These TUs are further classified as being
either derived from a chronic or an acute test (Office of Water Enforcement
and Permits; and Office of Water Regulations and Standards, 1985).

     The above discussion on TUs was abstracted from the Technical Support
Document for Water Quality-based Toxics Control. 1985,  and was the basis for
the toxicity-based approach in Monroe Harbor.  In addition to defining TUs in
terms of LC50, the sublethal bioassay toxicity data from this study was
represented by EC50.  This term was defined as the dilution fraction of the
sample that causes the response of the experimental organisms to be half that
of the control response.  For example, Figure 5.1 shows that zooplankton
reproduction was functionally related to the sample dilution.  In the
control, 15 young were produced per adult.  Half this response was 7.5 young ~
per adult.  One finds that a 0.2 or 20% effluent dilution fraction yielded a
response of 7.5 young per adult.  Hence, the EC50 was 20% and the number of
TUs would then be 1/0.2 or 5.

     The following bioassays were conducted during the Monroe Harbor study to
evaluate effluent and receiving water quality with a toxicity-based
approach.  These bioassays can be considered to be acute tests except for the
zooplankton reproductive rate assay (Ceriodaphnia). which should be
classified as a chronic test.

     EC50-based TUs were determined for the following bioassays:

     - Bacterial bioassay (acetate uptake)

     - Phytoplankton photosynthesis (^4C uptake)

     - Zooplankton grazing rate

     - Zooplankton reproductive rate (Ceriodaphnia)

     - Fathead minnows (growth)

     LCSO-based TUs were determined for:

     - Fathead minnows

     The Monroe Harbor study also addressed the response of resident
populations within the receiving water (i.e., the River Raisin).  These
organisms were exposed to 100% of the sample.  Only the location of the
exposure varied in these assays.  Some effect (presumed to be a toxic effect)

                                     20

-------
           15
    
-------
can affect such biological functions as larval fish growth.  These
observations can be translated into toxicity units (TUs) by the following
relationship (Di Toro gt aj_., 1985):


                              R
                                                                 Equation 5.2
                1 + exp {[In (CT) - in (!/<}>) ]/g)
where    R - biological response monitored in the bioassay at the
             experimental station.

        R0 - the control response

          - dilution fraction (% effluent) at which R is observed

         3 = sensitivity of the population employed in the bioassay

        Cj = the toxicity of the sample in toxic units (TU).

     If the upstream station response is taken as R0, the control response,  -
and R is the response observed at any other station, then, since the dilution
fraction () is 1 in these exposures, the only two unknowns in the dose
response function are @ and the toxicity, Cj.  For an assumed value of
beta, it is possible to compute Cj, which must have been there to have
caused a reduction of response from that at the upstream station, RQ,
relative to that observed, R.  It was assumed in these calculations that
toxicants primarily influenced the biological response and that the upstream
station was at zero toxicity, which may not be the case.  Toxicity gradients
in the river based on this approach, therefore, show the relative increase of
toxicity from upstream to downstream.

     For the Monroe Harbor study, toxicity in the receiving water was
determined based on observation of toxicity responses for the following:

     - Resident larval fish species distribution

     - Instantaneous growth rates for resident larval fish species

     - Ceriodaphnia - simple mortality.

     An important topic in the study was the prediction of toxicity in the
receiving water, investigated by treating the TUs being discharged as a state
variable.  A toxicity model (Section 6) demonstrates this concept.  The
toxicity results of the bioassays conducted in the river were compared to the
predicted model results.

5.2.1.2  The Chemical-specific Approach--

     The chemical-specific approach is defined in the Technical Support
Document for Water Quality-based Toxics Control (1985) as using laboratory-

                                     22

-------
generated water quality criteria or state standards to directly limit
discharges of specific toxicants.  Unlike the whole-effluent (toxicity-based)
approach described in Section 5.2.1.1, attempts here are made to consider a
range of toxic endpoints including human health impact and bioaccumulation
that are based on specific chemicals.  Once a criterion is developed, the
number is applied, through exposure analysis, as a permit limit to ensure
that the level of that toxicant is not exceeded after discharge.

     The chemical-specific approach to defining water quality in this study
was addressed by analyzing for chemicals that were known to be problems in
the study area.  Copper, chromium, and zinc analyses were performed because
of the known toxicity of these metals to cladocerans (Mount and Norberg,
1984) and other freshwater invertebrates (Hart and Fuller, eds., 1974), and
because of relatively high concentrations of these toxic heavy metals in the
River Raisin sediment.  The Phoenix Memorial Laboratory at the University of
Michigan, Ann Arbor (Jones, 1983) used neutron activation analysis to scan
for 33 elements in sediments collected in the River Raisin turning basin.
Relative to southern Lake Huron sediments (Robbins, 1980), the River Raisin
sediments were elevated in copper, zinc, and chromium.  In addition, The
USEPA Great Lakes Surveillance Branch (1975) recommended that the
contaminated dredged sediments from the navigation channel should not be
disposed of in the open lake.  Among other contaminants, their analysis
revealed high concentrations of copper (1450 mg/kg), zinc (970 mg/kg), and
chromium (530 mg/kg).

     Polychlorinated biphenyls (PCBs) were included in the study of Monroe
Harbor because high concentrations of PCBs in fir'  have been found in the
area.  Gas chromatography/mass spectrophotometry analysis of carp from the
River Raisin indicated high levels of PCBs in carp (Veith gi al., 1980).

     Bioavailability of organics to biota were addressed by the following
studies:

     - Bioaccumulation of PCBs in caged clams, channel catfish, and fathead
       minnows.

     - Adsorption/desorption experiments using hexachlorobenzene and
       hexachlorobiphenyl with three phytoplankton species common to the
       lower River Raisin.

     A secondary goal In this study was the development of a metals and PCB
mass balance model.  In support of the chemical-specific approach, a fate and
transport model of copper, chromium, and zinc was developed.   Important
processes considered in this model were adsorption and desorption of
toxicants to particulates and resuspension.  Also, an input-output model was
developed to determine  if the study area acted as a source or  sink of  PCBs.

5.2.2  Integration of the Toxicitv-based and Chemical-specific Approach

     An  integration of the two approaches was attempted in the Monroe  Harbor
study.   One way in which this was accomplished was via the following


                                      23

-------
empirical relationship that allows one to calculate TUs for a water sample
based on hardness and metal concentration (a chemical-specific approach).


                               a2
         Toxldty  -  a1 [Me] H *                                Equation 5.3
Toxlcity here Is the toxiclty exhibited in the bioassay reported in toxic
units (TU), [Me] is the dissolved metal concentration (yg/L), (H) is the
hardness concentration (mg/L) and a-j and a£ are the constants of
proportionality and the hardness exponent, respectively.  When the bioassay
utilizes 100% test water, the hardness is that of the sample.  However, if
the bioassay employs a dilution sequence then the hardness is that of the
mixture of sample and dilution water at which the toxiclty 1s quantified.
For toxicity tests, the hardness at the dilution at which the response falls
to one-half that of the control Is the hardness of concern.  The toxicity
units derived from this chemical-specific approach using the metals data were
compared to TUs derived from toxicity-based, EC50 bioassays.  The "fit" or
significance of the correlation was demonstrated statistically for copper and
zinc (see Section 7.2).

     Integration of the two approach methods was also attempted by
statistically determining the contribution of certain "cause" variables, such
as pH, metals concentrations, and alkalinity, to observed "effects"
variables, such as Ceriodaohnia survival and reproduction.  Again, an attempt
was made here to describe observed toxicity by us'ng chemical data (see
Section 7.2).
5.3  SURVEY STRATEGIES

     Surveys of Monroe Harbor were conducted in 1983 and 1984.  Location maps
of the sampling stations in the River Raisin and Lake Erie are shown in
Figures 5.2 and 5.3.  Three types of surveys were conducted:  (1) .surveys
that were primarily designed for modeling purposes were intensive five to
seven day surveys with stations longitudinally distributed along the river;
(2) surveys that examined both the toxicity and chemistry of plumes in the
river from the Monroe Wastewater Treatment Plant effluent and in Lake Erie
from the river and power plant effluents; (3) sampling of "master" stations
in 1984 to estimate variability of toxicity and specific chemicals.  Table
5.1 summarizes the stations sampled and objectives of the surveys.  Further
details of sampling methods and strategy are found in Section 7.1.
                                     24

-------
                                      FORD
                                 C\ MOTOR CO.
MASON
  RUN
                                         0) INTAKE
                                         EFFLUENT
        MONROE
        WWTP
                        DETROIT
                        EDISON
INSET:  WWTP  PLUME   STATIONS
     Figure 5.2.  Monroe Harbor and River Raisin Sampling  Stations
                             in  1983-84.
                                25

-------
MICHIGAN
         01234
        0123
    Figure 5.3.  Lake Erie Sampling Stations in 1983-84.
                           26

-------
oo

>-
LU ^~
> 00
a:  i
=3 PO
oo oo
LU OO
^L Z
«C O
_l i—i

 I  «C
   I—
Of. OO
o
ca a
a: z

CC. •-"
   CO
   o
20
— CT"
£
£
*
£
P.
•*
=•
£
if
?
p.
£
1*-.
£
*£
Pv.
r£
EC
^,
IT-
£


E
2
^
k
W
P
F
^
[2
— |C_
gc_
c->-
h

EE.
r
k
r>-
>
I
1.

II
Js
" "- 	 — i '• "" ' "T ' ' T "" ~
J

! !
! i
j , 1
!
X
>•
i x

1 ! x
! -
i x

i ~
; x
X
X
i i x
! x
' : 1 x
! x

! x
x
_^ X J_
X 1
1 ^ j
i ** i
** i
x ; x 'x 1 x
X X X
>. X ' X
x x ( x x
XX X ! X
>« x x
xx x 'X
XX X

X
x , x x x
u 5 « - = 3 • .
• C C - ~ _- ' oc i c
=x-e- — x E ~ . i. =• —
^x w — — , ^* t^3-«- i ^c= 5
i-£-ler-r I.I •= .1 £~c i-:
I^^lll-i' IP^-1 11-,-fl^
fe.i = ^S^;- iislr cSo.*-§fei

*~- \ cc- ' s.
*T __ ; 7 _ J- ^ y
-!c, , - | v
ni 	
>
>
X
^

>^








X

X
X












X



x


x


x
c «.*
' , c f c
C p ^ 3 »
c — » a . b k
•c C u — c — ^
= - c =1 ri
4 «O £ d &
at
er
»-
! "















x

x




















X


X
e. *" «e
*^ |
'C t *"
, i. ^ e
; £8 =
! 1
i *
I 	 i














x

x




















x


X
£ c
C V
"i ^ ~
c > c
c c •»-
f» X —
4.i
6/12/84
















x

x



_j
















x


x
•c
C C 1-
> — k—
w c •**
^ c c
6, — *-"
C* — c
•e
• L. i-
ai
//1 0/84
«
— , ,|


,





1 ,
! 1
|
i

i
x x

X X



j

1



: !



!




'

X l >^


" >




X *T
•^ tr
^ c
c o
                                                   27

-------
                                  SECTION 6.0

               WATER  QUALITY  MODELS OF  METALS, TOXICITY,  AND  PCBs
6.1  MODELING OBJECTIVES

     Mass balance modeling was undertaken to predict the behavior of
toxicants in the receiving water (River Raisin) for both the
chemical-specific and toxicity-based approaches.  The level or concentration
of a toxic compound in the environment depends on the quantity added to the
environment and the processes which influence its fate.  The important
processes affecting compounds discharged are transport and transformation.
Transport processes tend to distribute the compound within and among phases, .
whereas transformation processes chemically alter the pollutant.
Mathematical models provide the framework to incorporate these processes in
order to predict fate.  In the Monroe Harbor study, the two phases considered
were water and sediment.  A simple diagram demonstrating chemical fate is
shown in Figure 6.1.

     In support of current approaches to evaluate and control specific
toxicants in the environment, a goal of this study was to focus on the
development and calibration of a mass balance model to predict exposure
concentrations of copper, zinc, and chromium in the River Raisin  Physical
and chemical data were collected in 1983 to support the model needs.  A
separate model calibrated with specific conductance (a conservative tracer)
defined the transport field that was used in the framework of the metals
model.  See Section 5.1.2.1 for reasons given for selecting the above
parameters.

     Toxicity, measured in toxic units, was predicted in the River Raisin
utilizing the modeling transport framework that was developed above and
calibrated to specific conductance.  Ceriodaphnia fecundity, emerald shiner
and gizzard shad larval growth toxicity, caused by presumed toxicity
gradients in the River, were computed by the model and compared to
observations.  This approach assumed that toxicity acted as a conservative
and additive state variable.

     A third model was developed to investigate the relative loading
significance of PCB point and non-point sources in the River Raisin study
area.  An input-output model was developed to address these loading issues.
This was a first-level type of model used to determine whether or not the
system acted as a source or a sink.
                                      28

-------
                                       C
                                       
-------
6.2  TRANSPORT CONSIDERATIONS

     Any mass balance approach must identify and address the major processes
that transport toxic materials within the modeled system.  In the Monroe
Harbor study, 1t was determined that the two layer, estuarine, circulation
pattern and the resuspension potential were important considerations.  These
processes are addressed in the following two sections.

6.2.1  Hvdroloqical Description of the System

     Major physical features influencing modeling in Monroe Harbor were the
turning basin, the navigational channel, Detroit Edison's cooling water
intake, and the two-layer estuarine flow.  These features complicated the
River hydrology which, in turn, influenced the transport of toxicants.  This
transport was addressed in order to predict exposure concentrations in space.

     During the three 1983 surveys, the river was stratified from the turning
basin to Lake Erie due to temperature-induced density differences between
river and lake waters.  Monroe Harbor is a Great Lakes estuary with the water
surface elevation controlled by Lake Erie.  During the July survey, warm (low
density) river water entered Lake Erie near the surface, causing a return
flow in the bottom layer of denser, colder Lake Erie water.  During the fall
(September and October surveys), colder, higher density river water entered
Lake Erie near the bottom, causing a return flow of warmer Lake Erie water
near the surface.  With this complicated hydrology, it is conceivable that
contaminants could travel upstream.  The transport regime used in the models
was developed using a method of analysis for partially mixed estuaries
(Pritchard, 1969).  Horizontal and vertical circulation patterns for summer
and fall are shown in Figure 6.2.

     River stratification was primarily within the dredged navigational
channel and turning basin.  The channel is maintained by the U.S. Army Corps
of Engineers to a depth of 21 feet (6.4 m) below mean low water datum and to
a width of 200 feet (61 m).  It was observed that freighter activity in this
area upset this stratification on a short-term basis.

     The Detroit Edison Plant located near the mouth of the River Raisin
along the south shore is a coal-fired, steam-electric power plant which has
an exceptionally large, once-through cooling system requiring up to 3000 cfs
(4.9 cms) of water.  Water is discharged through a rock-walled canal which
leads into Plum Creek (Cole, 1978).  Essentially, all flow from the River
Raisin was diverted through Detroit Edison during the three surveys in 1983;
however, this flow met only 12.5%, 8.6%, and 9.1% of Edison's needs for the
July, September and October surveys, respectively.  The remainder of the
cooling flow came from Lake Erie.

     The modeled study area included the extreme lower portion of the 1072
square mile (2.776 knr) drainage basin where the River Raisin flows into
Lake Erie.  The study area included the river reach from the mouth up to the
first low head dam (Dam No. 6) for a distance of approximately 2.6 miles
(Figure 6.3).  This system is characterized by low flow during the

                                     30

-------
Summer Circulation Pattern
                                                      Net Erie and  River
                                                      Flow

                                                       Net Vertical  flow

                                                      Net Erie Flow
                      Turning Basin
Lake Erie
Fall  Circulation Pattern
                                                       Net Erie Flow
                                                        Net Vertical Flow
                        Turning Basin
                                                           Net Erie and
                                                           River Flow
  Lake Erie
   Figure 6.2.  Schematics  of Net Estuarine Circulation  Patterns
                in  the  River Raisin, Monroe Harbor.
                                  31

-------
DAM  (
(575.5)
          Union Camp  uywTP Effluent Plume Stations
    ^-STATION NUMBERS

     • DAM ELEVATION FT ABOVE MSL
    Figure  6.3.   River  Raisin  and  Nearshore  Lake  Erie  Sampling  Station
        Locations  for Surveys  1-3  (July,  August and  October,  1983).
                                    32

-------
summer-fall period with higher flow in the spring.   The historical record
indicates that the median flow is 316 cfs (8.94 cms), and is less than the
seven day, ten year low flow of 41.1  cfs (1.16 cms) less than 1% of the time.
The river flow was low during the 1983 mass balance modeling surveys on the
River Raisin.  The flows for the July, September and October surveys were 374
cfs (10.6 cms), 124 cfs (3.51 cms), and 230 cfs (6.51 cms), respectively.  As
the following section demonstrates, this low flow condition was an important
consideration when resuspension potential was assessed.

6.2.2  Resusoension Considerations for the 1983 Surveys

     Sediment resuspension, transport and deposition may have a considerable
impact on the distribution of sorbed contaminants.   Contaminant adsorption
and desorption during resuspension events may also largely impact their
distribution and fate.  Therefore, quantification of sediment resuspension,
transport and deposition is essential for understanding and predicting
contaminant fate in aquatic systems.

     Three analyses were conducted to quantify sediment resuspension and
transport and to assess its impact on contaminant loadings.  Bedford et a]..,
Parts 1-4 (1986) applied improved acoustic and analytical methods for
parameterizing resuspension.  Ziegler and Lick (1986) developed a
two-dimensional, vertically integrated, numerical model of sediment
resuspension, transport, and deposition and applied the model to the Monroe
Harbor/River Raisin system.  Di Toro et aj.. (1985) performed a sediment mass
balance for the system and determined the contribution of sediment
resuspension and transport.

     Resuspension of bottom sediments is the net result of a wide variety of
different fluid mechanical processes with characteristic time and length
scales that spread over six orders of magnitude.  These effects are most
heavily concentrated in a thin layer adjacent to the bottom called the
Benthic Boundary Layer (BBL) (Bedford and Abdelrhman, 1986).  Sediment
behavior near the BBL may be best characterized through use of in situ
devices.  However, such devices have only recently become available.  These
devices will aid in the direct measurement of sediment resuspension and  in
the validation of predictive techniques.

     Bedford gt al_., Parts 1-4 (1986) used an improved acoustic backscatter
transducer, with associated current meters and other devices to obtain high
frequency, in situ measurements of vertical variations in sediment
concentration near the BBL as well as factors contributing to those
variations.  The devices allowed measurement without entering the flow field
at a frequency in time and space not previously possible.  Devices were
deployed on two occasions  in the turning basin of the River Raisin:  a 5-hour
period on July 5, 1985 and for a 40-hour period ending on July 10.  The
devices were deployed near the downstream southern boundary of the turning
basin where water depth was 3.7 m.  Deployments were primarily intended  for
testing the devices and associated methods.  However, valuable information
was obtained concerning sediment resuspension and flows in the Harbor.   In a
subsequent application to  the Central Long Island Sound (Bedford et al_., Part

                                      33

-------
4, 1986), these methods were used to compare in situ and laboratory predicted
entrainment rates.

     This application (Bedford fit aj.., Parts 1-4, 1986) indicated that no
well defined horizontal flow field occurred in the water column at the
sampling site for the study condition.  Horizontal velocities were
approximately of'the same magnitude as vertical velocities during the
sampling periods, and virtually no mean current was observed.  Observed
irregularities in the flow field may be due, in part, to boat traffic in the
basin as well as boundary reflection and refraction.  Vertical velocities
appeared sufficient to keep fine sediment particles in suspension but did not
appear to be of sufficient intensity to cause large scale sediment
resuspension.  The ambient, steady sediment concentration located
approximately 100 cm above the bottom was 10 mg/L, with relatively steep
concentration gradients below that elevation.  Divers reported a 5 cm layer
of silt covering a pebbly bed at the study site.  Core samples had a density
of approximately 2.5 g/cnr and a porosity of over 85%.

     Ziegler and Lick (1986) developed a two-dimensional (2-D) vertically
averaged, numerical model of sediment resuspension, deposition, and transport
and demonstrated the model using data obtained from the Monroe Harbor/River
Raisin.  The model is based on the numerical solution of two-dimensional
equations of continuity, momentum, and transport.  Hydrostatic pressure is
assumed.  The numerical solution scheme is based upon a second-order
accurate, explicit, and conservative, finite difference scheme which allows
treatment of variable geometry and a wide variety of boundary conditions,
including the open boundary.

     The 2-D sediment model was applied to the Monroe Harbor/River Raisin
using constant inflow loadings (concentrations.of 100 mg/L), and two flow
conditions, approximately 45.4 nr/sec (1603 ftvsec) and 151.3 nr/sec
(5344 ft3/sec).  Hydrodynamics were run to steady-state before computation
of deposition and entrainment.  The model application was intended to
demonstrate the model's capabilities.

     The 2-D model predicted that for lower flow conditions, when using
variable depth, velocities would decrease rapidly as flows entered the
harbor, with a large subsequent impact on sediment transport.  No return
flows, or flow reversals, were predicted and only deposition of sediments
occurred for the simulated conditions.  The high flow predictions were for
increased velocities in the turning basin with no flow reversals.  The major
difference between the two cases was the prediction of sediment entrainment
for the high flow case.  Entrainment contributed to increased suspended
solids concentrations.  Most of the predicted entrainment occurred in the
upstream channel with deposition occurring as flow entered the deeper harbor.
A small area of sediment erosion was also predicted in the shallow section of
the downstream river channel.  Sediment deposition occurred at a faster rate
during the high flow case due to increased concentrations.

     D1 Toro e£ al_. (1985) addressed the question of solids distribution by
modeling the steady-state distribution of suspended solids using the spatial


                                     34

-------
grid and transport coefficients established for conservative substances.
They treated solids as a conservative material (i.e., settling was
neglected).  Model predictions compared favorably with observed
concentrations and accounted for the majority of the observed variability.
The unaccounted for variability was attributed to nonconservative factors
influencing suspended solids concentrations.

     These three investigations independently suggested that resuspension was
not a major source of suspended solids for the flows which occurred during
the three 1983 intensive surveys in Monroe Harbor.  Results suggested that
the system had enough energy to keep particles suspended but not enough
energy to result in appreciable resuspension.  This allowed treatment of
suspended solids as a conservative material to account for the majority of
its observed variability.  Results also suggested that the remaining,
nonconservative, variability may not have been the result of the mean
current, but rather of  high-intensity intermittent events such as ship
traffic.
6.3  TWENTY-TWO SEGMENT FATE AND TRANSPORT MODEL OF COPPER, CHROMIUM, ZINC,
     AND TOXICITY

6.3.1  Approach

     A mass balance approach was used by Manhattan College to predict
toxicity and metals exposure concentrations in the receiving water (River
Raisin) (Di Toro et al_., 1985).  As a first step, the study area was divided
into twenty-two segments of uniform water quality.  Then, a conservative
substance was used to determine transport coefficients and flows in this
system.  The transport regime was then tested on other conservative
substances.  Toxicity and metals were then modeled based on the transport
coefficients determined using conservative substances.

     The complex estuarine circulation pattern of the River Raisin was taken
into consideration when WASTOX, Manhattan College's fate and transport model,
was applied.  A two-dimensional model framework was used with a total of 22
model segments.  The model employed two vertical layers from the turning
basin to the river mouth.  The model was run to steady state for each of the
surveys.  Flow and vertical exchange rates (bulk dispersion coefficients)
between segments were determined using specific conductance data collected at
numerous transects, with vertical readings taken every half meter.  The
steady state flows and exchange rates were determined based on mass balance
of a conservative substance and hydraulic continuity around a segment.

     Methods were also developed which allowed the model computation of
variability as well as mean concentrations.  The ability to compute the
variability was important because recent water quality criteria require that
the probability of exceeding the criteria concentration be evaluated.  Often,
the data themselves, including toxicity in toxic units, are quite variable.
It is the purpose of probability models to compute the variability of
concentration within the context of deterministic models.  The method


                                     35

-------
employed used the same transport and reaction equations as the steady state
WASTOX model.  It computes the variance of the concentrations given the
variances (and cross correlations) of the input loadings and boundary
conditions.  Although the method employed is independent of the underlying
probability distribution of the variables involved, it has been found that
most variables are consistent with the log normal probability distribution.
Figure 6.4 illustrates this fact.  The straight line represent lognormal
probability distributions.

     Survey 1 (July 12-17) and Survey 2 (September 13-18, 1983) were designed
such that point sources and stations longitudinally distributed in the river
were sampled synoptically.  These intensive six-day surveys were conducted
along with bioassays using water collected from selected modeling stations.
Survey 3 (October 25-28, 1983) focused on the chemical and toxicological
impact of the Monroe WWTP effluent plume to the River Raisin.  Sampling
stations for these surveys are given in Table 7.1 and locations are shown in
Figure 5.2.  Point sources sampled during these surveys included the Monroe
WWTP effluent, Ford influent and effluent, and Mason Run.  Additional
sampling station and collection method details are described in Smith et al_.,
1985.  Also, Section 7.1 of this report presents a summary of this topic.

6.3.2  Results

6.3.2.1  Modeling Conservative Substances--

     The circulation pattern based on specific conductance was verified by
computing the distribution of other conservative tracers and comparing them
to observations.  As shown in Figure 6.5, the comparison between the solid
lines representing the mean WASTOX predicted surface segment concentrations
and the surface measurements was quite reasonable for alkalinity and hardness
in the July 1983 survey.  The predicted concentration of conservative
substances for the other 1983 surveys also compared well to the observed.

     Variability was also predicted and the comparison of the computed
median, 16th and 84th percentiles and the data are shown in Figures 6.6-6.8.
The center data symbol is the median, and the vertical line spans the
percentiles.  The model agrees exactly at the boundaries since these data
values were used as model inputs.  Alkalinity and hardness distributions are
well reproduced.  The suspended solids data distribution appears to exhibit
more variability in the middle of the river than was computed by the model.
The reason for this unexplained variability may be due to the intermittent
ship traffic.

6.3.2.2  Modeling Copper, Chromium, and Zinc--

     Metals enter the River Raisin study area from municipal and industrial
wastewater discharges, from upstream River Raisin, and from Lake Erie.  The
Monroe Wastewater Treatment Plant, Ford Motor Company, and Mason Run were
sampled for metals during Surveys 1-3 (July, September and October, 1983).
The Ford discharge and upstream River Raisin provided the largest loads of
copper and chromium to the modeled area.  Zinc loading was shared almost


                                     36

-------
 \
 US
 2
   in'
            ALKALINm

          STA  »  2  - J\'L\
         1   10   &0  90  99
            PROBABILITY
                                          103
           HARDNESS

         STA  »  2  - JULY
\

7
                                          IP2
       1   10   $0   90  99
           PROBABILITY
   102
 \
 U
 C
 v
   10'
           SUSP.  SOLIDS
          ?TA  t  2  - JULY
         1   10   f.O  90  99
            PP.OBABILm
  101
\
0
Z
•
y
"•'
  10°
           TOTAL CU
         STA  »  2 - JULY
        1   10   50  90
           PROBABILITY
 \
            TOTAL CR
          ?TA  >  2  - .H
        1   10   50   90
            PROBABILITY
 PHYTO TOXICITY  - JULY 1983
         EC50  - STA *  Z
 . 102
                                           101
                                           100
        1   10   50  BO  09
           PROBABILITY'
Figure 6.4.   Log  Probability Plots of  Variables as  Indicated
              for Survey  1  (July  1983),  Station 2.
                                  37

-------
               ALKALINITY
          200
      ~   150
          100
           50
                  1  .  1   i  1  i  1
             .5   3  2.5  2   1.5  1
             RAISIN  R.  MILE
         400



         300


     u
     §  200

     i
         100
              HARDNESS
JULY  SURVEY
                                         u
5   9   -. 5 -1
      L. ERIE
 JULY SURVEY
             .5  3  2.5  2  1.5   1
             RAISIN R.  MILE
5   0  -. 5  -1
     L.  ERIE
Figure 6.5.  Comparison of Observed and Model  Computed Alkalinity
            and Hardness for Survey 1 (July 1983).
                            38

-------
              J
              \
              M
             • E
              c
              0
              c
              B

              c 101
              0
                   River Raisin -  Susp.  Solids  -  July 1983
                   3.5
              2.5
 1.5
Miles


                                     -.5
              3    River Raisin -  Susp.  Solids  - Sept  1983

              \102
              C
              5
              •pi
              v
              t
              U
              v
^
•S
3.5
                             2.5
                        1.5
                       Miles
           .5
-.5
              \ 102
              tc

              E
                    River  Raisin  - Susp   Solids -  Oct  1983
C
I
c


0
^
                 id
                    3.5
                        1.5
                       Miles
                    -.5
Figure  6.6.  Comparison of Model  Results with  July, September  and
                 October  Data for Suspended Solids.
                                   39

-------
             J
             \
                    River  Raisin  - Hardness - Julv  1983
»— •
•B
3.5
                            2.5
          1.5
        Miles
           .5
-.5
                    River  Raisin  - Hardness  - Sept  1983
             tt
             E
             0 1Q2


             0
                  3.5
2.5
 1.5
Miles
                                    -.5
             J
             \
                    River Raisin -  Hardness -  Oct 1983
               102
                  3.5       2.5       1.5        .5       -.5
                                    Miles
Figure  6.7.  Comparison  of Model  Results  with July,  September and
                     October Data  for Hardness.
                                  40

-------
                    River  Raisin  - Alkalinity  - July  1983
                102
                   3.5
2.5
 1.5
Miles
.5
-.5
                    River  Raisin  -  Alkalinity  -  Sept  1983
              i 102
                   3.5
2.5
 1.5
Miles
                                                .0
         -.5
                     River  Raisin  - Alkalinitv  - Oct  1983
              : 102
                                      1.5
                                     Miles
                             -.5
Figure  6.8.   Comparison of Model  Results with  July,  September and
                    October Data for Alkalinity.
                                   41

-------
equally between the upstream, Ford, and the Monroe Wastewater Treatment
Plant.

     The transport field from the twenty-two segment WASTOX model that
described conservative tracer distributions for the three 1983 surveys was
the basis for computing transport in the metals models.  Although steady
state VIASTOX spatial profiles of both total and dissolved fractions were run
for zinc, copper, and chromium, only examples of the probability model are
presented here.  The comparison of the computed and observed spatial
distribution of total copper, chromium, and zinc for the July and September
surveys are shown in Figures 6.9-6.10.  The data point just upstream of the
downstream boundary is in the Ford effluent plume and, therefore, does not
represent the laterally mixed concentration at that location, which is what
the model computes.  In general, the medians are reasonably well reproduced,
but the variability appears to be larger than that predicted by the model.
The contribution of internal sources of variability such as resuspension and
deposition may be the cause of the increased variability in the observations.
As mentioned in Section 6.2.2, these resuspension events were most likely
related to intermittent events such as freighter traffic.

     The transport and fate of heavy metals is often affected by sorption
onto suspended solids or sediments.  WASTOX assumes that a local equilibrium
exists between the dissolved and particulate phases and that reaction is
parameterized by use of a partition coefficient that determines the fraction
of total metal concentration that is particulate and dissolved.  The
dissolved metal concentration is an important parameter in exposure studies
because the toxic metal species are contained within this fraction.

     Laboratory and field studies have demonstrated that partition
coefficients often decrease with increasing particle concentrations (O'Connor
e£ al_., 1980; Rygwelski, 1984).  However, this phenomenon was not observed
for copper and zinc in Monroe Harbor.  Furthermore, no unique partition
coefficient could adequately describe the distribution of dissolved
copper or zinc for each survey.  The lack of strong trends or relationships
for partition coefficients in Monroe Harbor complicates metal exposure
predictions for toxicity purposes if predictions are based on total metals
data only.

     A comparison of partition coefficients from Monroe Harbor with those
from other freshwater systems indicates large variations for copper and zinc
(Table 6.1).  It has been suggested by Di Toro et al_. (1985) that one reason
may be the intermixing in the estuary of carbon-rich Lake Erie particles and
low carbon particles from the river.  Carbon-rich particles have a higher
sorption potential for metals compared to low carbon particles.

6.3.2.3  Modeling Toxicity as a State Variable--

     The applications of the probability model to toxicity are illustrated in
Figure 6.11.  The distribution of Ceriodaohnia fecundity, emerald shiner and
gizzard shad larval growth toxicity, caused by the downstream boundary and
                                     42

-------
                           Raimn  - Total Cu  -  Julv 1983
Concentrat
               J
               \
               «
               3
               *i
               t
               u
               V
               C
               v
               o

               0
               U
                                      1.5
                                     Milea
3.5       2.5



 Emr Raimn  - Total Cr  -  Julv  1983
        -.5
3.5
                             2.5
                  1.5
                 Miles
        -.5
                     Rmr  Raimn  - Total 2n  - Julv 1983
jj/
Concentration
°
                                 .
3.5
                             2.5
                  1.5
                 Miles
.5
-.5
Figure 6.9.   Comparison  of Model Results with July (Survey  1)  Data
          for Total Copper (Top), Total  Chromium  (Middle),
                       and Total Zinc  (Bottom).
                                  43

-------
                      J
                      \
                      *
                            River Raisin  - Total Cu - Sept 1963
                        ID'
3.5      2.5
                                           1.5
                                          Miles
.5      -.5
                      ~     River Raisin - Total Cr - Sept  1983

                      at
                           3.5      2.5
                1.5
               Miles
                            River Raisin  - Total Zn - Sept
                      ; 10'
                        100
                           3.5
                15
               Miles
                                -5
Figure  6.10.   Comparison  of Model  Results with  September (Survey 2)  Data
            for Total  Copper (Top), Total  Chromium (Middle),
                           arid Total  Zinc  (Bottom).
                                        44

-------
               ^ Ri^er  Raism - C.  Daphma  Toi  - Composite
               £  -5
                   3.5
         2.5
 1.5
Miles
          .5
-.5
               ^ River  Raism  - Em. Shiner  Toi  - Composite
               c
               t
               u
               0
               u
                  .5
                   3.5
         2.5
 1.5
Miles
          .5
-.5
               ^  River  Raiain  - Giz. Shad Toi  -  Composite
               C
               0
               4«
               e
               u
               v
               C
               c
               0

               0
               U
.5
                   3.5
          2.5
 1.5
Miles
                  -.5
Figure  6.11.  Comparison of Composite Model  Results with Data  for
      Ceriodaphnia  Fecundity  Toxicity (Top), Emerald  Shiner
                 (Middle), and  Gizzard Shad  (Bottom)
                 Growth Toxicity, in Toxic  Units.
                                  45

-------
     TABLE 6.1.  COMPARISON OF ZINC AND COPPER PARTITION COEFFICIENTS FROM
                  MONROE HARBOR AND OTHER FRESHWATER SYSTEMS


                                      COPPER                  ZINC

Monroe Harbor (Suspended            104 - 105-3             104 - 106
  Solids 7-50 mg/L)

Flint River, Michigan1              104-5 -  105-°           104'9 - 105-4
  (Suspended Solids
  3-34 mg/L)

Inner Saginaw Bay, Lake             104 - 105               104'5 - 105-5
  Huron2 (Suspended
  Solids 8-46 mg/L)

Outer Saginaw Bay, Lake             105-2 -  105-7           105'5 - 106-5
  Huron2 (Suspended
  Solids .6-22 mg/L)

U.S. Rivers1 (Suspended             104-4 -  105-1           104'5 - 105J
  Solids 10-50 mg/L)


Ju.S. EPA, 1984.
zSmith et al_., 1977.
Monroe WWTP, are computed and compared to observations.  With the exception
of the farthest upstream point, the profiles are reasonably representative of
the observations.  This application assumes that toxicity is conservative and
additive.  While this may be a good approximation in some cases, it has been
shown that it is not always the case.  Loadings of toxic units (TUs) to the
toxicity model were from the effluents and from upstream and downstream
boundary conditions.

     The model employed to describe transport of toxicity was the same as
that developed for the metals above.  Determinations of toxic units in the
effluents were based on methods described in Section 5.2.1.1, The
Whole-effluent (Toxicity-based) Approach.  Observed toxic units in the
receiving water were determined based on the in situ dose/response
relationship, Equation 5.2, Section 5.2.1.1.  More information on the
bioassays used in the toxicity model can be found in Section 7.2.

6.3.3  Evaluation

     A steady state, mathematical, chemical fate model  of the River Raisin
was constructed.  A two-dimensional segmentation scheme was utilized to
represent this stratified Great Lakes estuary.  A two layer transport regime

                                     46

-------
was calculated using the method developed by Pritchard for the LLRS-Monroe
Harbor Surveys 1-3 (July, September,  and October,  1983).  The model was
successfully calibrated using specific conductance,  chloride, hardness and
alkalinity as conservative tracers.  The model  was used to calculate the
spatial profiles of both total and dissolved heavy metals assuming these
parameters were conservative in the River Raisin system.  This was a valid
first-cut assumption since no systematic deviation was observed between the
model-calculated and the observed spatial profiles.   The latter indicated
that no other significant transport phenomena (such as resuspension) or
sources and/or sinks were present.  Prediction of exposure or dissolved
metals concentrations based on total  metals data alone would be difficult
because of the wide range of partition coefficients observed in this study.

     The application of the probability model,  assuming that zooplankton and
larval fish toxicity is conservative and additive, is consistent with
observations.  However, cases are known where additivity is not observed.
Without further analysis of the underlying mechanisms and processes involved
that affect toxicity, it may not be appropriate to assume that the
conservative and additive nature of these toxicities extend into Lake Erie or
elsewhere.
6.4  INPUT-OUTPUT MASS BALANCE MODEL FOR PCBs

     PCBs and metals were selected as parameters of concern in the study
because of their documented presence (Section 7.1) and toxicity.  Metal
concentrations correlated with observed acute toxicity.  It was, therefore,
desirable to develop a 22 segment model to predict exposure concentrations in
spatial detail.  Since PCB concentrations did not correlate
with acute toxicity (Griesmer, 1987), development of a 22 exposure segment
model did not proceed.  Although acute effects of PCBs were not observed in
the toxicity tests, PCBs are still considered a problem because of the
bioaccumulation observed in fish and clams in the River (Section 7.2.6).  A
bioaccumulation, predictive model was not developed as it was not a primary
objective of the study.  However, the availability of data did permit a look
into a PCB loading analysis of the area.  A description of this analysis
fol1ows.

6.4.1  Approach

     The mass balance models for PCBs in the River Raisin/Monroe Harbor were
based on a simple algebraic input-output model (Pritchard, 1969).  The PCB
model served as a framework to assess the water quality status of Monroe
Harbor by determining the PCBs loading across its boundaries.  The PCB model
was designed to address the following general questions:

     - What is the status of PCBs in the water?

     - Are local point sources contributing significant contaminant loads
       to Monroe Harbor?
                                    47

-------
     - Does the River Raisin/Monroe Harbor act as a sink or a source for
       PCBs?

     - If the point sources are important, what is their rank according to
       loading?

     - Are non-point source loads a concern?

     The PCB model employed two segment layers to represent the stratified
water sometimes present in the lower River Raisin.  The model boundaries were
Station 1 (upstream) and Station 4 (turning basin, downstream) (see Figure
6.3).  The PCB model input loadings included the Monroe VIWTP, upstream
(Station 1) and Lake Erie (Station 4) (see Figure 5.2).  Station 4 at the
turning basin was sampled extensively, both bottom and surface samples were
taken during Surveys 1 and 2 (September and October, 1983).  Surface samples
only were taken during the July survey.  The two segment PCB model did not
provide spatial resolution in the lower River Raisin.  However, it was
considered sufficient to address the question of whether the lower River
Raisin is a source or sink of PCBs.

     The PCBs model assumptions were:

     - A steady-state condition and two-layer stratification

     - PCBs acting conservatively

     - PCBs concentrations uniform within each layer of each segment

     - The pollutant concentrations at the boundary between segments or
       layers equal to the average concentration in the two adjacent
       segments or layers

The net difference between the known input and output total PCB loads was
calculated for each 1983 survey.  Daily mass balances were determined for
each of the five days during the July and September surveys.  Only data
collected on October 25th were abundant enough for modeling purposes during
the October survey.  In addition to modeling total PCBs in the system, each
of the ten PCB homolog groupings were modeled.  Appendix A contains the
solutions to Pritchard's model for each of the three 1983 surveys.

6.4.2  Results and Evaluation

     Events such as high flow, high winds, and extreme lake level
fluctuations did not occur during the modeling surveys.  In addition, since
no significant rainfall occurred, runoff was minimal.  These conditions
helped support the steady-state modeling assumptions.

     Mason Run was not considered to be a major source of PCBs to the system.
A high flow measurement made in the spring of 1984 (April) was 3 cfs.  The
1983 surveys were conducted in a normally low flow time of the year.
Therefore, the flows during these surveys, although not measured, were


                                     48

-------
presumed to be no greater than 3 cfs.   At this low flow,  the concentrations
of PCBs measured in Mason Run were not high enough to cause a significant
load of PCBs to the system.  Calculations of PCB loadings from Mason Run
based on the high flow (3 cfs) and average PCB concentrations for the July
and September surveys yielded 0.000879 kg/day and 0.00256 kg/day,
respectively.  These loadings are only 2.5% and 1.8% of the loadings for July
and September relative to the "unaccounted for" PCB source in the system.

     Another point to consider is that Mason Run is part of a Great Lakes
estuary.  Observations of flow in Mason Run during the July survey indicated
that at times the flow was reversed.  A non-parametric, Mann-Whitney U-Test
(Zar, 1974) was performed on total suspended solids, hardness, total PCB,
chloride, and conductivity for the July and September surveys to determine if
the data from Stations 4 and 8 (Mason Run) had the same probability
distribution or came from the same population.  These results showed
insignificant differences between these two data sets for each of the above
parameters.  This supports the observation that water in Mason Run is often
influenced by the River Raisin.  The reverse is unlikely because Mason Run
with a maximum flow of 3 cfs could not influence water chemistry of the River
Raisin for the parameters evaluated in the statistical test.

     Data collected in September, 1983 (Survey 2), allowed mass balance
modeling from Station 1 to Station 4.  This survey, unlike the July and
October surveys (Surveys 1 and 3), included sampling of the top and bottom
layers.  Surface to bottom differences in chloride, specific conductivity and
hardness occurred at Station 4.  It was necessary, therefore, to determine
bulk flows in the upper and lower layers at Station 4 in order to perform a
mass balance on these conservative substances (Appendix A, Equation G).  The
results of modeling chloride and hardness based on transport coefficients
using specific conductivity are presented in Figures 6.12 and 6.13.  The
unaccounted for term (Einputs - Zoutputs) in these figures is negligible when
compared to the input loads which demonstrate that the conservative input and
output loadings for this model are balanced.

     The Mann-Whitney U-Test applied to PCB and suspended solids data in
September at Station 4 for the top and bottom layers showed that no
significant differences existed in the population probability distributions
for each of these parameters.  In addition, a Student's T-Test was performed
on the September, log transformed, PCB data at Station 4 (top and bottom).
Test results showed that the means of the top and bottom layers were not
statistically different at the 95% confidence level.  The difference of the
two flows (upper and lower layers) was required in order to perform as mass
balance calculation.  This was a simplification and allowed the use of
Equation H in Appendix A.  Although sampling of PCBs and solids did not occur
at Station 4 (top and bottom) during July and October surveys, it was assumed
that the concentrations for these two parameters were also homogeneous in the
vertical distribution.

     Since PCBs are often  associated with solids in the aquatic environment,
transport of PCBs in the system is often related to sediment transport.  The
suspended solids mass balance results in the River Raisin are shown in Figure

                                     49

-------
   o  o o o o
   
                                                               a.
                                                               a.
                                                               O) CM
                                                                  0)
                                                               t. >
                                                               o s-
                                                               <4- 3
                                                                  oo
                                                               QJ >_x
                                                               u
                                                               c n
                                                               « oo
                                                               r— CTV
                                                               W i—
                                                               CD
                                                                   •»
                                                               10 S-
                                                               tn 0)
                                                               ia J3
                                                               z: E
                                                                  QJ
                                                               Q) 4->
                                                               •D Q.
                                                               •»- OJ
                                                               i. 00
                                                               O
                                                               o
                                                              J3
                                                               s_
                                                                3QI801HD
                                                               O)
                50

-------
                                                   O)

                                                   «t
             -_  2 2 Z Z 2
              
                                                   o  s-
                                                   ««-  3
                                                   U
                                                   c ro
                                                   
 l/)  Q.
 a>  a)
 c 
-------
6.14.  These results show that there was some variability in the unaccounted
for loading of total solids.  This variability was discussed in Section
6.3.2.1, which suggested that the observed variability was due to
intermittent disturbances from ship traffic.  In order to rule out the
possibility that these intermittent resuspension events contributed
significantly to-the calculated PCB unknown loading, the following analysis
was performed.  The highest total suspended solids loading unaccounted for in
Figure 6.14 was 10,000 kg/day.  The highest observed PCB sediment
concentration in the turning basin was 0.55 mg/g (Section 7.3.3.3).
Multiplying the sediment loading by the PCB concentration in the sediment
yields a maximum of 5.5 x 10"3 kg PCB/day that left the system due to its
association with particles.  This amount represented only 16% and 4% of the
"unaccounted for PCB loading" for the months of July and September,
respectively.  Therefore, resuspension can be ruled out as a significant
contributing factor to this large unaccounted for loading of PCBs.

     Mass balance results for total PCB are presented in Figures 6.15, 6.16
and 6.17.  The same model and assumptions that were applied to the total PCBs
in the upper reach were applied also to the homolog balance as shown in
Figures 6.18, 6.19 and 6.20.  Results show that the unaccounted for loading
was enriched in the tri- and tetrachlorobiphenyl compounds.  This trend was  -
most obvious in September and October (Surveys 2 and 3).

     PCBs have been manufactured in the United States for only half a
century, and awareness of their toxicity began to develop a little more than
a decade ago.  We still know very little about their long-term toxic and
carcinogenic effects on humans.  We do know, however, of incidents where
human exposure to PCBs has caused severe toxic reactions, and of laboratory
experiments where PCBs have been accountable for a variety of toxic symptoms
and cancers in animals.

     We also know that PCBs are unusually persistent and that, although their
manufacture ceased in this country in 1977, most of the hundreds of thousands
of tons that were manufactured between 1930 and 1977 are still with us and
are uncontrolled in the environment.  Because of the low solubility of PCBs
in water and their high affinity for fatty tissue, it appears that the
primary exposure of humans to existing PCBs is through accumulation in the
food chain.  Thus, it appears that those deposits of PCBs in aquatic
environments which are accessible to fish populations, pose the greatest
threat to human health.  Unaccounted for loadings of PCBs enriched in the
tri- and tetrachlorobiphenyl compounds were found in the Monroe Harbor/River
Raisin receiving water.  Their source is unknown.

     The two-layer transport regime was developed using the method developed
by Pritchard for the Monroe Harbor July, September and October 1983 surveys.
The model was successfully calibrated using specific conductance, chloride,
hardness and alkalinity as conservative tracers.

     Model calculations showed that a source of PCBs exists in the lower
River Raisin.  This unaccounted for loading was calculated for the three 1983
                                      52

-------
     BM
SSI
                                  %
\
                                   \
                                                      -C
                                                       u
                                                       ns
                                                       (U
                                                      C£.
                                                       0)
                                                       Q.
                                                       Q.
                                                       *o co

                                                       1C i—
                                                      CD

                                                       «/> >>
                                                       CO Ol
                                                       W1 CO
                                                      f— PO
                                                       0 00
 (U  C
•a -r-
 c
 0)  
-------
                                                        T3
                                                        O)
                                                        3
                                                       4->
                                                       C/l

                                                        Ol
                                   \

O

O
3Dd 1V101
                                                       0)
                                                       cr:

                                                       s-
                                                       a;
                                                       a.
                                                       CL
                                                       a>
 t. r-
 O
«+-  >»
    0)
 V  >
 O  i-
 C  3
 
-------
                          is a JD CD a
O        O



 AVQ/0> 8Dd
O
 I
                                                     •o
                                                     3
                                                     to

                                                     
i 
(4- 5-

Q. Q.
i OJ
i— OO

'W
4-> C
O ••-
1 (—
• u
O
X)
z

-------
                                                   (O
                                                   O)
            o 2
                                                   •o
                                                   3
                                                   +->
                                                   GO

                                                   O)
                                                   _c
                                                   4->








aftflQnrVyVSnc


















q
•




•0 01 - c
0 0 o c

AVO/O> a















B6GS5S5Q5?*
WYYV$V§^X^J**






'










• — ' — p 	 .
3 — OJ M
' o o e
i i
3d IViOi









%
\
o
\
^>
Ir
q.
^
^
tt



V
^
-*



-^.
t
%=.
^


1
>




0
1 *O
; o>
Q£
^

Q. U
O
^~
t} C
4-> -r-
£
' l_
0
^
s_
(O
3:
cu
°
i Wi
c
: O
2_

;
r*.
! r—
I to
i
1 Qj
: V.
56

-------
                                                  OJ
        ooooooooo
     •O ooooooooo
      £• ooooooooo
      £1,223233322
      Oooooooooo
                 — n •" ^ n
                 E d t-; ^ j
                              ^
                              (N
             °0
OOIOI^OH
0)
                                                  u
                                                  ITJ
                                                  OJ
                                                  O)
                                                  Q.
                                                  0)
                                                     O)
                                                  I/) >
                                                  0> l-
                                                  O 3
                                                  •— CO
                                                  Z 00
                                                    CT»
                                                  CO •—
                                                  O
                                                  Q.  •
                                                  -o
                                                  (O
                                                  o
                                                  TD
                                                  OJ
                                                  O
                                                  O
                                                  o
                                                  (O
                                                  00
                                                  O!
    57

-------
                                     •o
                                      3
s ° °
g>oo
•3 .;
b_

_t
~"^
,
""5

J
' 	 ^
,
t*
•MBHiMaH
— n
j
"••™^™

i i i
%
AoQ/6> 9NIQV01 OOlOhOH 8
o o o o o o o
2 a 3 a a a a
3 O O O O O O














%
Dd



-i-
^^
\
-i-
>
V
^^
^^
-SI
"5
'O
->
^
>
00
>


\

 •

,^_ *
^^^ *•"
3
' O
r— CO
O 00
«•
1 -^
i 00 S.
O 0)
Q- -Q
O ••->
i Q.
, O1 (U
C CO
•o c
ro -i-
O
. _J
T3
(1)
i =
O
U
o
(O
C
3
•

-------
                                                                                       o>
"" SSC2°SSS3
 ^ oocoooooo

 0*222222222
                                                                                       QJ
                                                                     §
                                                                     Q
                                                                  CO
                                               U
                                               (O
                                               OJ
                                              0£

                                               i.
                                               
                                                  s_
                                               I/I 3
                                               O10O
                                               o<—-

                                              §'" ro
                                                  CO
                                                  en


                                              rn  #>
                                              O S_
                                              Q_ OJ
                                                  J3
                                              «*- O
                                               O 4->
                                                  U
                                               cno
                                                                                      T3
                                                                                       (O
                                                                                       O
                                                                                       •o
                                                                                       (U
                          §
                                                                                       3
                                                                                       O
                                                                                       O
                                                                                       o
-C-
        °0
o
OJ
                                                                                       O)
                                                                                       Ol
                                      59

-------
surveys during July,  September,  and October.   This source of PCBs approached
200 g/day in September.
                                     60

-------
                                  SECTION 7.0

                               COMPONENT STUDIES
7.1  WATER CHEMISTRY

7.1.1  Survey Approach

7.1.1.1  Rationale and Strategy--

     The water chemistry surveys of the River Raisin-Monroe Harbor-Lake Erie
system were designed to address multiple objectives.   Data on spatial and
temporal gradients of water chemistry were necessary  for investigations of:  .

     •  Relationships between physical, chemical  and biological effects;

     •  The mass balance of various contaminants;

     •  Seasonal variations in water quality and toxicity;

     •  Effective methods of sample and data collection for these purposes.

(See Objectives, Section 5.1.)

     Whenever possible, physical and chemical measurements of water quality
coincided with sampling for contaminants analysis and for bioassays.  This
ensured comparability of the data.  The first five surveys were designed to
identify spatial gradients of toxicity and water quality in Monroe Harbor and
adjacent Lake Erie waters.  Principal point sources were also characterized
in the same way.  The results proved to be adequate for mass balance
modeling, but a greater range of toxicity measurements were needed for
correlations with water quality factors.  Therefore,  five additional surveys
of only four master stations were conducted during May-August 1984 in order
to look for temporal variations of toxicity and chemistry.

     Throughout this study, the intended strategy was to adjust the field and
laboratory methods to cope with any problems and to meet new requirements as
the work progressed.  In practice, however, the analytical results lagged too
far behind the surveys to have much influence on the project design.  As a
result, the methods used for processing and analyzing water samples for
organochlorines proved to be less than optimal.  These problems will be
discussed in more detail below.
                                      61

-------
7.1.1.2  Survey Schedule and Stations--

     During the two years of field studies, the sampling schedules and
locations were varied to reflect increased understanding of the system, and
to accommodate different program objectives.  Ten surveys were conducted.
The dates and main objectives of these, as well as the 43 stations sampled,
were shown in Table 5.1 and Figures 5.2 and 5.3.  The selected 13 stations
for which water chemistry data are summarized here were those shown in Figure
5.2, with the addition of Station 11 in Lake Erie.

     As the study progressed, various stations or groups of stations were
sampled to provide data on chemical gradients within portions of the system
or within observed plumes.  Certain stations were used to define
concentrations at upper and lower boundaries of the system, in point source
discharges or at control sites.  The latter were chosen by the bioassay teams
as sites where similar water could be found that was unlikely to produce
measurable toxicity.  Chemical and biological characteristics of the control
water were then defined.  Some problems encountered in the selection of these
control sites are discussed further in Section 7.2.

     A few master stations were sampled throughout the year to document
temporal changes in water chemistry.  Selected stations were sampled on
occasion near the surface and near the bottom in order to detect any
stratification that might occur.  In general, plume stations were sampled
only once in connection with a flow event, as in Surveys 3, 4 and 5.  In some
cases, station locations were selected to represent strategic points within
the system where water quality differences might be expected.  Examples are
Station 8 in the mouth of Mason Run and Station 5 at the Ford plant outfall.
All station locations were identified by Loran-C coordinates and water depth.
Most were also located easily by reference points on nearby land.

7.1.1.3  Sampling Plan--

     Jja situ water quality measurements were taken concurrently with samples
analyzed for general chemistry, and organic and metallic contaminants.
During the first five surveys, field laboratories were established at the
turning basin dock aboard the U.S. EPA R/V Bluewater, a 50-foot research
vessel, and within a 30-foot EPA laboratory trailer parked nearby.  This
minimized the holding time of samples prior to filtration, solvent extraction
or chemical analysis.  This was an important aspect of quality assurance in
the field.

     Typically, samples were collected by 3-person crews in small outboard
motor boats and were delivered within an hour to the field laboratory.
Precautions were taken to avoid materials, techniques and conditions which
might degrade sample quality during collection, transport or processing.

     To ensure that daily grab samples of water were representative,
composites of samples taken every eight hours were also obtained during
Surveys 1-3.  In addition to whole water collections, many water samples were
                                     62

-------
separated Into dissolved (filtered) and participate fractions for organics
and metals analyses.

7.1.2  Phvsical/Chemical Parameters

7.1.2.1  Methods-

     Field measurements included standard physical  and chemical  indicators of
water quality at all  stations where contaminant samples were taken.  These
parameters were as follows:

     • Water Temperature (*C) - Temperature meter/probe used ia situ.

     • Dissolved Oxygen (mg/L) - D.O. meter/probe used In situ.

     • pH (Standard Units, SU) - pH meter/probe used in field laboratory or
                                 in situ.

     • Specific Conductance (yS/cm) - Conductivity meter used in field
                                      laboratory or in situ.

     • Total Suspended Solids (mg/L) - Filtration and freezing samples as
                                       field laboratory; drying and weighing
                                       at EPA/LLRS.

All of these were Standard Methods as described in APHA, 14th Edition (1975)
and were adapted to the instrument models used.  Measurements of oxygen, pH,
and specific conductance were corrected for water temperature.  Readings were
taken only at one meter below the surface except when sampling near the
bottom occurred or when vertical profiles were measured at one meter
intervals.

7.1.2.2  Results--

     Results are summarized in Table 7.1 for those 13 stations (Figure 5.2)
that were sampled in more than one survey.  Other stations were sampled only
once in order to characterize chemistry gradients in plumes at the Monroe
WWTP or in nearshore Lake Erie.

7.1.2.3  Evaluation--

     Temperature was subject to seasonal variations.  The lowest and highest
ranges occurred in offshore Lake Erie (3.4-24.9*C at Station 11) and near the
Edison power plant outfall (14.4-33.3*C at Station 29), respectively.

     Dissolved oxygen  (DO) was lowest in effluent waters of the Monroe WWTP
(7.60 mg/L at Station  7) and Ford complex (7.17 mg/L at Station 9), and
highest at Stations 4  and 6 (12.5 mg/L).  However, water temperature affected
the DO levels at saturation.
                                     63

-------
                TABLE 7.1.   MONROE  HARBOR WATER CHEMISTRY:  PHYSICAL/CHEMICAL PARAMETERS
Station Niflber
1 -x + S.D.
(Range)
n
3 7+ S.O.
(Range)
n
4 x"± S.D.
(Range)
n
5 If+S.D.
(Range)
n
6 "x + S.D.
(Range)
n
7 x"+ S.D.
(Range)
n
8 x * S.D.
(Range)
n
9 7+ S.D.
(Range)
n
10 T+ S.D.
(Range)
n
11 TT+S.D.
(Range)
n
25 7 + S.D.
(Range)
n
26 7 + S.D.
(Range)
n
29 7. + S.D.
(Rang*)
n
Temperature
(°C>
•20.3 + 6.6
(6.3 - 28.8)
45
22.9 + 4.6
(11.0 ~ 28.2)
33
19.7 * 6.3
(6.2 " 28.3)
88
22.1 + 3.8
(10.9 - 26.4)
33
22.0 + 3.9
(10.8 " 26.0)
32
21.3 + 2.5
(14.0 - 23.8)
36
21.0 + 5.9
(7.5 ~ 28.2)
36
24.5 + 3.6
(14.0 - 29.2)
33
22.2 + 4.2
(11.0 ~ 26.5)
33
12.3 + 8.0
(3.4 ~ 24.9)
25
N.D.
16.2 + 6.8
(5.1 " 23.4)
42
25.2 * 7.5
(14.4 - 33.3)
14
Dissolved Oxygen
(mg/L)
9.01 + 1.63
(5.8 - 10.8)
19
9.40 + 0.65
(8.5 - 10.6)
9
8.02 + 1.55
(4.5 - 12.5)
49
9.43 + 1.40
(6.7 - 12.2)
10
9.89 * 1.05
(8.7 r 12.5)
10
7.60 + 0.47
(6.9 ~ 8.2)
5
9.32 + 1.61
(6.1 - 10.5)
6
7.17 + 1.03
(6.3 - 8.3)
3
7.80 + 0.20
(7.6 ~ 8.0)
3
10.00 + 0.68
(8.6 7 10.6)
11
N.D.
8.77 + 0.82
(7.5 7 10.7)
35
8.19 + 1.32
(5.9 ~ 10.6)
14
PH
(S.U.)
8.26 + 0.39
(6.20 - 8.65)
42
8.29 + 0.16
(7.89 - 8.68)
33
8.15 + 0.37
(7.30 - 9.10)
121
8.55 + 0.31
(7.9 - 9.1)
63
8.62 + 0.38
(7.95 - 9.16)
33
7.54 + 0.20
(7.24 - 7.95)
36
8.28 + 0.17
(7.99 - 8.56)
36
7.80 + 0.15
(7.58 - 8.17)
33
8.17 + 0.32
(7.70 - 8.81)
33
8.19 + 0.55
(7.40 - 8.89)
35
N.D.
8.05 * 0.26
(7.70 - 8.50)
34
7.76 + 0.18
(7.60 - 8.10)
13
Specific
Conductance
( S/cm)
691 ••• 64.8
(490 - 847)
45
671 + 53.7
(521 ~ 801)
33
544 + 101
(273 ~ 928)
133
279 + 37.6
(220 ~ 357)
63
280 i- 48.2
(223 ~ 342)
33
912 * 62.4
(737 ~ 1060)
36
548 + 68.8
(379 ~ 671)
36
391 + 53.8
(236 ~ 462)
33
305 * 42.7
(251 ~ 362)
33
257 + 28.9
(220 ~ 344)
35
N.D.
334 + 81.3
(254 ~ 497)
42
408 + 109
(288 " 531)
14
Total Suspended
Solids
(ma/Li
32.4 * 14.1
(16.1 - 95.3)
42
20.3 + 31.8
(-152 - 46.3)
33
33.9 * 16.9
(15.0 - 112.0)
102
26.1 + 10.4
(6.7 - 51.3)
62
25.1 + 13.6
(9.5 - 68.3)
33
32.5 * 68.3
(3.5 - 319)
37
37.9 + 15.4
(19.4 - 96.2)
38
15.2 + 9.4
(5.1 - 57.2)
33
43.1 + 8.1
(32.1 - 75.2)
33
15.9 + 10.0
(5.7~ 42.3)
• 22
20.7 + 5.0
(15.0 " 24.3)
3
40.0 + 25.6
(10.4 - 73.8)
6
51.0 + 38.8
(15.8 r 110.0)
6
N.D. 3 Not done.
                                                 64

-------
     The arithmetic means of pH were also lowest (acidity highest) in both
effluents (7.54 at Station 7, 7.80 at Station 9) and in the power plant
discharge (7.76 at Station 29).  This was probably due largely to dissolved
carbon dioxide and/or chlorine, and perhaps to acidic components in the
effluents.  However, pH was similar (within 1 unit or a factor of 10) at all
stations.

     Specific conductance was highest (912 yS/cm at Station 7) in the WWTP
effluent and was most variable downstream below the turning basin (Station
4), where mixing of river and lake water generally occurred.

     The suspended solids concentration was highest below the turning basin
(66.0 mg/L at Station 4) due to a combination of solids loadings from the
WWTP upstream and periodic resuspension caused by ship traffic in the turning
basin.

7.1.3  General Chemistry

7.1.3.1  Methods--

     Sample splits were analyzed for four water chemistry parameters using a-
Technicon Autoanalyzer  II and the following established methods:

     Dissolved Ammonia  (mg/L) - phenate method (NAQUADAT No. 07555) according
to Environment Canada (1979).

     Nitrite (mg/L) - sulfamilimide method (102-70W) according to Technicon
Industrial Systems (1971).

     Chloride (mg/L) -  ferric thiocyanate method (NAQUADAT No. 17206)
according to Environment Canada (1979).

     Hardness (as CaC03, mg/L) - EDTA method (165-71W) according to
Technicon Industrial Systems (1972).

7.1.3.2  Results--

     Results are summarized  in Table 7.2 for 10 of the 13 stations (Figure
5.2) that were sampled  in more than one survey.  The whole suite of general
chemistry parameters were measured only during the first five surveys.
Chromine/chloramine measurements using a Fischer and Porter Titrator were
attempted during all surveys, but since concentrations were generally at or
below detection limit in all but waste water samples, the results are not
reported here.

7.1.3.3  Evaluation--

     Dissolved ammonia  concentrations were relatively high in the Monroe
Waste Water Treatment Plant  (WWTP) effluent  (3.06 mg/L at Station 7) and just
downstream from the WWTP (0.398 mg/L at Station 3).
                                     65

-------
TABLE 7.2.  MONROE HARBOR WATER CHEMISTRY:   GENERAL CHEMISTRY
Station Munber
1 JT+S.ff.
(Range)
n
3 7 + S.D.
(Range)
n
4 7± S.D.
(Range)
n
5 7 + S.D.
(Range)
n
6 7 + S.D.
(Range)
n
7 T+ S.D.
(Range)
n
8 TT+ S.D.
(Range)
n
9 x" + S.D.
(Range)
n
10 7 + S.D.
(Range)
n
11 7* S.D.
(Range)
n
Dissolved Anmonia
(HM/D
0.027 * 0.027
(0.001 ~ 0.127)
21
0.398 + 0.302
(0.125 ~ 1.12
18
0.178 1- 0.065
(0.047 ~ 0.360)
38
0.039 * 0.052
(-0.003 - 0.162)
18
0.028 + 0.046
(-0.005 - 0.136)
18
3.06 + 1.43
(0.71 ~ 6.83)
22
0.167 + 0.055
(0.074 " 0.29)
22
0.089 * 0.076
(0.008 ~ 0.26)
18
0.104 + 0.068
(0.025 ~ 0.233)
18
0.119 + 0.108
(0.0005 ~ 0.265)
10
Nitrite
(mg/L)
0.011 + 0.007
(0.002 - 0.044)
36
0.032 i- 0.028
(0.007 - 0.132)
32
0.024 + 0.015
(0.008 - 0.103)
53
0.045 + 0.045
(0.0001 ~ 0.105)
33
0.046 i- 0.048
(-0.0008 - 0.105)
33
0.158 + 0.156
(0.011 - 0.582)
37
0.031 + 0.020
(0.011 - 0.098)
37
0.040 + 0.033
(0.007 ~ 0.165)
33
0.042 + 0.041
(0.001 - 0.102)
33
0.025 + 0.031
(0.0007 " 0.085)
14
Chloride
(ma/L)
47.7 + 4.6
(32.6 - 52.8)
21
48.1 + 5.5
(39.9 - 56.3)
18
33.8 + 6.6
(19.4 - 48.1)
38
12.9 + 1.5
(10.2 - 14.8)
18
12.5 + 2.6
(10.1 - 21.1)
18
70.4 + 7.6
(59.4 - 94.5)
22
33.8 * 5.5
(21.2 - 43.9)
22
26.7 + 5.6
(19.5 - 41.9)
18
14.1 + 2.1
(9.8 - 19.5)
18
11.4 + 1.6
(9.0 - 15.2)
10
Hardness
(as CaCO,)
(mg/L)
332 + 37
(218 - 401)
36
309 + 59
(85 ~ 390)
32
243 + 50
(142 - 326)
53
139 + 37
(98 r 236)
33
131 + 31
(95 - 200)
33
342 * 56
(197 ~ 528)
37
261 + 53
(169 ~ 373)
37
168 + 30
(115 ~ 235)
33
142 + 30
(103 ~ 209)
33
115 + 36
(29 ~ 173)
14
                             66

-------
     Nitrite concentrations were similarly higher in the WWTP effluent (0.158
mg/L at Station 7), but were relatively constant at lower concentrations
throughout the rest of the system.

     Chloride concentrations at the upstream boundary (47.7 mg/L at Station
1) and 1n the WWTP effluent (70.4 mg/L at Station 7) appeared to be diluted
by intrusions of'low-chloride water from Lake Erie (11.4 mg/L at Station 11).
Intermediate values were seen throughout the Monroe Harbor system.

     Hardness concentrations (as CaC03) followed a similar pattern to those
of chloride, from the upstream boundary (332 mg/L at Station 1) to Lake Erie
(115 mg/L at Station 11).

7.1.4  Organic Contaminants

7.1.4.1  Methods--

     Organic samples (whole water)  were collected in 4 L (1 gallon)
pre-cleaned, brown glass jugs as used commercially for pesticide-grade
solvents.  Caps were lined with Teflon®.  At most river and lake stations,
the jugs were filled by quickly submerging them by hand from a small boat.
At point source stations, jugs were fastened into a weighted metal harness
and submerged at the end of a rope.  At Station 4 only, deeper water samples
(one meter above the bottom) were collected using a Tygon® hose and a
peristaltic pump.  The samples, enclosed in opaque cartons to exclude direct
sunlight, were transported to the field laboratory generally within one hour.
The water samples  (4 to 12 L in volume) were subdivided for solvent
extraction into 1  L portions in additional jugs.

     One of the project goals was to measure contaminants in particulate and
dissolved fractions of water.  Each volume of 3 to 8 L was passed through
several glass fiber filters (4.25 cm diameter; 0.7 ym porosity) in sequence
to isolate particles and filtrate.   This was a slow process because of the
filter changes and handling.  As it was difficult to clean the filter system
properly between samples, the dissolved fraction was sometimes contaminated.
At higher concentrations of suspended solids, adequate volumes of filtrate
were difficult to  obtain.  Therefore, after Survey III only the particulate
fractions were collected and analyzed.  Larger filters (142 mm) of the same
type were used after Survey II, to cut down on filtering time.

     Water samples were batch-extracted twice with 100 mL of dichloromethane
in the field laboratory.  The combined extracts were stored in sealed jugs at
room temperatures  in the dark for up to several weeks.  Filters with
particulate samples were stored with acetone in sealed jars in the dark,
until they were Soxhlet-extracted with hexane:acetone  (1:1).  Both kinds of
extracts were concentrated in Kuderna-Danish assemblies on a steam bath, and
were dried on sodium sulfate columns before final evaporation under nitrogen
to a 10 mL volume.  Final extracts of most water and all particulate samples
were cleaned of lipids by treatment with concentrated  sulfuric acid and by
elution through an activated Florisil® column.  Only acid was used for some
                                     67

-------
water samples.  The cleaned extract, re-concentrated to 2 ml In hexane, was
stored 1n sealed glass ampules until analysis.

     Although no other clean-up procedures were tried in this study, other or
additional treatments might have been more effective with some samples.  A
thorough investigation of clean-up methods for all types of samples to be
collected should'have preceded actual surveys.  Samples from certain stations
such as the Ford effluent (Station 10), presented special clean-up problems,
due to interfering oils or other substances.  These problems were not
resolved in this study, causing some significant gaps in the PCB data.

     Analysis by high resolution, fused silica, capillary gas chromatogcaphy
was performed on a VARIAN Model 3700 gas chromatograph equipped with a 63Ni
electron capture detector (ECO).  A 50-m fused silica column (0.2 mm i.d.)
coated with SE-54 (Hewlett-Packard) was used to separate the PCB congeners.
The chromatographic data were acquired with a Hewlett-Packard Laboratory
Automation System and transferred to a Digital PDP-11/45 computer where the
raw files were analyzed using a series of specialized programs.  Once
analyzed, the data were stored in a final data base and archived on magnetic
tape.

     Further details of the organics sampling, extraction, clean-up and
analysis are given in the Project Report (Smith et aj.., 1985).

7.1.4.2  Results--

     Results are summarized in Table 7.3 for 9 of the 13 stations (Figure
5.2) that were sampled in more than one survey.  Station 9 samples were not
analyzed due to clean-up problems, and no organic samples were collected from
Stations 25, 26 and 29.  The table includes data for selected organochlorines
and total PCB.  Other organochlorines analyzed were alpha-, beta-, and
delta-BHC, hexachlorobenzene, heptachlor epoxide, alpha chlordane,
oxychlordane, alpha-, and gamma-chlordane and 10 unknown chlordane
derivatives, cis-nonachlor, and 4,4'-ODD and DDE.  The average homolog
composition of PCB mixtures at selected stations is presented in Figure 7.1.

     Data for nearly all of the individual PCB congeners were also obtained,
but these results are too volumnious to summarize in this report.

7.1.4.3  Evaluation--

     Polychlorinated biphenyls (PCBs) were the dominant group of
organochlorine substances in Monroe Harbor, with mean concentrations ranging
from 8.5 ng/L at the upstream boundary (Station 1) to 230 ng/L at the mouth
of Mason Run in the turning basin (Station 8).  The mean concentration in
Lake Erie (29 ng/L at Station 11) was markedly higher than at the River
Raisin upstream boundary, possibly due to Detroit River influence.
Relatively high concentrations occurred in the Monroe WWTP effluent (180 ng/L
at Station 7).  The significance of these levels is discussed in Section 6.4
(Mass Balance Model for PCBs).
                                    68

-------
on
pos
                                <_>

                                4->
                                 C
                                 Ol
                                 o

                                 a;
                                a.

                                 en
                                 o

                                "o

                                 o


                                 oo  •
                                 o  i/»
                                 a.  c
                                    o
                                    •^
                                 ..  +j
                                 >> 
                                 •M
                                 <« *J
                                 3 «3

                                 1_
                                 O
                                 J2
                                  a;
                                  o

                                  c
                                  o
               80            O
               w            »-


80d 1V1O1 JO  OO1OWOH %
                                  01
                                  s.
                                  en
69

-------
3
l«
3 ^
r
                  i
                  th
                  «8
CD .
i
                         CM
                       •f 1 1
     °°c>
     + 1 • CM

     »8
     o d
                       ss?
       o
     s*
     00
     +i < cQ
     fO ^3
     dS
     r^ <^
      • CM
     ~-g>
     + 1 I CM
                       MS
                        • <"»

                       S|
                       *i*
                       I*
             8f
             do
             +l > CM
                               00
                                +•,5
             ?i3
             :,fs
                               a?
                               do
             o
             d 10
             2
              • *-* C

             sl
             *|3
                                        + 1 i
                     oo
                     + 1 i CM

                     38
                     CM IO
                     d *Z
                     +i • !
                      o »
                      »- in
                      + 1 i in
                      SJ8
                      d d
                      o w

                      28
                      d d
+ 1 > ft

"IS
"d
                                        •fia
                                                S3
                                                 s
                                                •f i • -*
                              PPJ
                              d d
                              + l > -o
                              28
                                                + 1 • -0
                              SS
                              dd
                              + 1 . »  N.
                                                  O KJ^


                                                  28.'
                                                                       CM «
                                                                       in «»
                                                                       VI


                                                                       Ix'1
                                                      70

-------
     For most of the 24 hour sampling periods,  concentrations based on
composites of three samples were only 10% higher than those based on a single
grab sample.  Therefore, diurnal changes In PCB concentration seemed minor,
although high seasonal variability was observed near the turning basin
(6.7-630 ng/L at Station 4; 19-730 ng/L at Station 8).

     Filtrate PCB concentrations measured only  during the first three surveys
were often higher than the corresponding total  values.   Apparently, the
filtrates were contaminated in the filtration process.   In later surveys,
only whole water and the particulate fraction were analyzed.  These results
are considered more reliable and indicated that approximately 50% of the
total PCB was in the particulate fraction.

     Differences in PCB composition within the  system are evident in the
station profiles of homolog percent composition (Figure 7.1).  The homolog
groups representing compounds of 1-9 chlorines  per molecule may vary in their
proportions due to differences in source Aroclors and/or weathering effects.
At the upstream boundary (Station 1), PCB composition appeared to be
dominated by monochlorobiphenyls, which are probably least stable in the
aquatic environment (Figure 7.1).  Just downstream from the Monroe WWTP and
turning basin (Stations 3, 4, and 5), PCB was relatively enriched in tri- and
tetrachlorobiphenyls.  In contrast the proportion of penta-, hexa-, hepta-
and octachlorobiphenyls appeared to be diminished at Station 3 and Station 4,
relative to Station 1.  In every case, the percentage of nona- and
decachlorobiphenyls was insignificant.

     Difficulties were encountered in routine clean-up of dichloromethane
extracts of some water samples, which contained a petroleum-like material.
Although this material was not removed by acid  or Florisil® columns, some
other method, such as gel permeation chromatography (GPC), might have been
successful.  However, it was not within the scope or technical means of this
study to evaluate clean-up methods.  Specialized clean-up often may be
required in industrialized areas such as Monroe Harbor.

7.1.5  Metallic Contaminants

7.1.5.1  Methods--

     Whole water samples were collected in 0.5  L linear polyethylene (l.p.e.)
bottles with polypropylene caps.  At river and  lake stations, the bottles
were filled by quickly submerging them by hand  from a small boat.  At point
source stations, a larger l.p.e. bottle on a weighted rope was used to
collect sub-surface water to fill the sample bottles.  At Station 4 only,
water from 1 meter above the bottom was also collected using a Tygon® hose
and peristaltic pump.  The samples were transported generally within one hour
to the field laboratory, where half of the sample was passed through a
0.45 ym cellulose acetate filter to prepare the dissolved metals fraction.
Both samples were spiked with nitric acid to keep the metals in soluble form.
All procedures for collecting, processing and storing metals samples were
previously tested under similar conditions and found to give acceptable
blanks.  See Section  7.1.6.4 for precautions taken to ensure sample quality.

                                      71

-------
     All total metals samples were nitric acid-digested using a method
adapted from USEPA (1971).

     Copper and chromium were analyzed with a Perkin-Elmer 460 Atomic
Absorption Spectrophotometer fitted with graphite furnace, a HGA-2200
controller and AS-1 Auto Sampler.  Analytical conditions were determined by
applying the methods in Perkin-Elmer (1977).

     Duplicate injections and analyses were performed on each sample, blank,
and standard.  The mean peak height was used to calculate the concentration.

     Zinc, because of its higher concentrations, was analyzed using a
Perkin-Elmer 603 Atomic Absorption Spectrophotometer in the flame mode.
Standard grade acetylene and air were used to support the flame.  Analytical
conditions were taken directly from the Perkin-Elmer 603 Methods Manual.

     After absorbance readings or peak heights were measured, the data were
processed automatically using a computer program which was written to correct
for blanks, perform linear regressions, plot the standards, and determine the
concentration of the samples.

     Further details of metals sampling, processing and analysis are given in
the Project Report (Smith et aj.., 1985).

7.1.5.2  Results--

     Results are summarized in Table 7.4 for 13 stations (Figure 5.2).  As
indicated in the ranges, negative values occurred in a few cases after blank
corrections were made, and these were included in mean calculations.

7.1.5.3  Evaluation--

     Dissolved metals as a fraction of total concentrations averaged 38% for
copper, 34% for zinc and 34% for chromium.  Clearly these metals appear to be
largely associated with suspended solids in the water column or in effluents.

     Metal concentrations at Station 1 defined the boundary conditions.  Mean
concentrations increased by approximately 50% within and below the turning
basin (Stations 4, 5, and 26), but decreased again to near boundary
conditions in Lake Erie (Stations 6, 11, and 25).  Metal contaminants appear
to exit the River Raisin mainly through the Edison power plant discharge
canal (Station 29).

     The highest mean concentrations of 37.5 ug/L for copper, 61.3 ng/L for
zinc and 24.8 pg/L for chromium were measured in the Ford plant effluent
(Station 9).  These were approximately ten times the boundary concentrations.
Higher concentrations of zinc were seen also in the Monroe WWTP effluent
(mean, 41.0 ug/L).  To some extent the observed differences between metal
concentrations at certain stations can be attributed to seasonal biases in
addition to real water quality gradients.  Not all stations were sampled on
the same surveys.  Other data not presented here defined concentration


                                     72

-------
s
1
§ ^

US
> <*•

"o
at
*
a
§
o <-^
t_ —1
f X
LJ (

(0

o
8
«•>
"•> ;
> j
"o «-
at
at
5
u
IM — 1
en :
o w

1
O "
i:
o "
at
at

Q
I.


O ;
"S >

O
t-





(

o *•>
CM 00
O O
•O
+ 1 . IM
CM O
IM •
o'S

B~
«- f>I
+ 1 i CM

S8 *
• 0
*— N^

ss
«- in
+ 1 i CM

0 °
• CM
CM .

3j
-Nt CM
+ 1 • CM
Kl O
CM •
• IM
•O i
N. «•»
00 NO
+ 1 • IM
in »»
CM •

CM C




«- CO
+,.3
IM Kl


Kl «•»
«&C
• 9
r.I

Ix"
.-


CM *s
«- in
0 0

+ 1 ' «-
o «-
IM •
• o
0 *->

in <•*
O CM
in
+ 1 ' «-

S- CO
in •
• 0


So
IM K
+ 1 • ^

2.K5

CM i

0 0
• in
NO «-
S °.
' CM
•o •-
in ^
Kl ^*
O CM
+ 1 ' 2
Kl S
«— N!
+ 1 i ro

In"^
• 0
ro ^

o o
IM -*
•f 1 >  IP
ON O


CM N_,

~.

• in
CM «-
N. NO


in w
o«C
«" f
•f |O£
|xv
in


r*. ^
0 O

+ 1 ' •-
fM «-

s?

ss
«— O


K) ~*
Kl -*
o •-

+ 1 ' «-
(SK ^~

• O
o —

in eo
«- MS
o
+ 1 > IM

t- Kl
"^ 0


O
-df •
r^ ro
•f 1 i «—

CM 0

IM *r-

>0 0
"~ o
•f 1 i IM
O 0
«— INJ
•* Ci

h- N-
+n 2
**f ^O

• 0
CM ~

NO CM


-^~o
+ 1 I IM
-» O
o •
• CM
Nfl «-
o «i c
CO
+ IQC
\*~
h.


in 'N
o o

+ 1 . IM
•o o
CM •
dS

RR
«- CO
CM
+ 1 i IM

K*
1 ^_
ro ^

§s
•- NO
+ 1 . CM

N. 0
O •

IM T

CMO
*~CM
+ 1 . IM
°.°

^ •L,

«— -o
«— in
+ 1 . CM
FN* O


CM \^

~


CM «^
CM
+ 1 i CM
h- *$•
in •

si C
e> «»
*,!

Ix"
00

73

-------
111
u
.
i
1
52

l!

1
M
o
1

i <-.
U -J
" 1

*•»
o
8

IM

1 **
"o ^

.*
o


•«— ^
N _l
«l
o w
^

i
o **
11
0s*
(A

5
u
O J
o C
- 1
m w
4-t
0

i
*



%£
torn
in
+ 1 . «-

in S
^-

0~

s's^

8§

•* o
«M «
S
25 •
• in
if\ (V
+ 1 • £
8 °;

N. v*

*z
in in
in (M
*o
oo"

«- ro
•O w
OS
N. •
+ 1 • £
O >*

• -o
CM v^
O CM
spi
+i. 2
o ro
in •
r^2

2fc
+i«*
lx~
-
o SC
-S
*'£
Ix"
o

Sc?
o o

+1 • •-
ro o

• o
o *^




•^ **
+ 1 I CM
S*.
• 0


r-«
ro o

CM >O
+ 1 • 2
Kl°
• CM


0 0
• in
in CM
CM
o o
in •
• o



«- Kl
7,T^
00 ^

• o
(M ^
$z

•f 1 • CM
CM CO

• O
CM v*
:f
+ ICX
|x"
-

P 5
o o

+1 . «*
Kl »-
CM •
• O
O ~


sr-

0 Kl
•f 1 i •>*
o in
in •

CM >•«

VO-
CM O

r- N.
+ 1 > >*
in o
CM •

in •^

&S.
• »
O CM

i ^a
""
S. ^
«- in
•f 1 i •*
m ^*

• CM
K> v
• <^ C
+ IOC
|xw
in
CM

SC
0 O

+ 1 i r-
» 0

• o
O ^^


r^ «

»- in
+ 1 . ^
R1".
• o
CM V

r-- o
in •
• o
ro «-
+ 1 i h-
Kl°
• o
Kl ^

0 S
•& '
• CM
•O CM

O O
O •
• CM



in -o
0 CM
+ 1 t r^
to ro
o •

"~
in «•»
CM CO
+ 1 i N.
CO O>
in •

-» +*
• ** C
-

o2


CO «

«- - ro
IM •

ro w

«- o
ro •

-* ^
+1 . r»
ro o

(M i

ss
• Kl
ro ro

00
• «*


rs. ~
ro >r
o in
^f CO

• CM
Kl ^
CM •-
CM «-
+ 1 i N-
O ^

~ 2

2|C
•f lae
Ix"
a
                                         74

-------
gradients observed In discharge plumes near the Monroe WVITP and In nearshore
Lake Erie.  Refer to Surveys 3 and 4 in Table 7.1.
7.1.6  Quality Assurance:  Results and Evaluation
     Quality assurance (QA) was established at the design stage of the water
chemistry program.  To ensure acceptable quality of samples and data
collected in the field work, the written survey plans prescribed
     • A system of identification numbers and labelling for all samples;
     • Instrument calibrations and checks
     • Sampler design, materials and construction
     • Water sampling procedures for survey vessel  and shore crews
     • Types and preparation of sample containers
     • Equipment and methods for handling, filtering and preserving samples
     • Numbers and frequency of field duplicates and blanks
     • Reference materials for precision analysis
     • Design and use of field data sheets
     • Assignments and scheduling of survey staff based on training and
       experience.
     As much as possible, all materials and methods were evaluated under
realistic conditions prior to the main field season; that is, a "shake-down"
survey was conducted.  This became Survey 1 since the results turned out
reasonably well.  However, some procedural changes were made in later
surveys, as already discussed.  The importance of staging field trials of
staff, methods and equipment can not be overstated.  Among the activities to
be included in field trials are:
     • Rehearsals of sampling schedules and staff coordination;
     • Test deployment of field and lab equipment;
     • Testing of methods and procedures;
     • Staff training;
     • Collection and analysis of preliminary samples.
     Details of quality control measures used in the field work are given in
the Survey Plans  and Project Report (Smith et al_., 1985).
                                      75

-------
     The quality assurance program in the laboratory (including the field
lab) was designed to evaluate data quality with respect to:

     • Instrument performance;

     • Blanks and detection limits;

     • Measurement precision;

     • Measurement accuracy;

     • Analyte recovery.

     The QA results and evaluation for water analysis are summarized briefly
below.

7.1.6.1  Physical/Chemical Parameters--

     Hydro] ab® monitoring equipment used for measuring temperature, dissolved
oxygen, pH and conductance was calibrated according to the manufacturer's
instructions before each survey cycle.  Precision and accuracy of pH and
conductance measurement were evaluated as noted in 7.1.6.2 below.

7.1.6.2  General Chemistry

     Background interferences were assessed by analyzing filter and bottle
blanks.  Results from these analyses were all less than the quantitation
limits of 0.013, 0.002, 35, and 0.9 mg/L for NH3, NO?, CaC03, and Cl,
respectively, indicating that the blanks did not contribute significantly to
the observed sample concentrations.

     The precision (expressed as the standard deviation/mean x 100) for
NHj, N03» CaC03, and Cl were 13, 6, 7 and 8 percent, respectively, of
the mean value for replicate analyses.  Precisions for alkalinity, pH,
conductivity and residual chlorine were 3.8, 0.5, 2.9, 4.7 percent,
respectively, of mean values.

     Accuracies were determined for Surveys 1-4 by analyzing EPA
intercomparison samples.  Accuracies for Nl^, NO^, CaCO-j, Cl, pH, and
alkalinity were within their respective 95% confidence intervals as required
by EPA/Cincinnati.

7.1.6.3  Organic Contaminants--

     System blanks were analyzed to determine the background error
contributed by the entire analytical procedure relative to the total PCB
concentration of samples.  Mean concentrations in blanks, regardless of
sample type (i.e., grab, composite, whole water, particulate), were less than
12% of the corresponding sample concentrations.  Additionally, the blank
median values for all analyses were less than 10% of the sample
                                     76

-------
concentrations indicating that, overall, blanks did not significantly affect
the observed total PCB concentrations in samples.

     The precision of field duplicates was determined as the range of the
duplicate measurements divided by the mean (relative range).  Since the
precision was concentration-independent, all  precision data were pooled and
estimated as + 24% of the expected mean concentration for samples.  The
precision of GC analyses, based on duplicate analyses of standards, ranged
from 8 to 13% of the mean concentration.

     The overall accuracy of trace organic analyses of water was not
determined due to the lack of suitable standard reference materials.
Consequently, the best available accuracy measurement for this study was
based on replicate GC analyses of standard Aroclor mixtures.  These averaged
113% of the expected concentration.

     Based on sequential Soxhlet extractions of particulate samples, 96 + 5%
(n - 4) of the total PCB was recovered in the first extraction.  Based on
that result, the extraction procedure for particulates was considered
thorough.  However, none of the particulate PCB results are included in this
summary.  Since the Monroe Harbor work ended, the results of whole water
extraction efficiency tests at EPA/LLRS have indicated that three solvent
extractions of water containing particulates are necessary to achieve nearly
complete recoveries of PCBs.

7.1.6.4  Metallic Contaminants--

     Detection limits for copper, chromium, and zinc were 0.31, 0.25, and 4
ug/L, respectively.  Filter blank and system blank values were less than the
detection limits.  Generally, sample concentrations exceeded those of blanks,
indicating that contamination from the analytical method was not a problem.

     Water samples were pooled to form a whole water QC sample representative
of the range of sample matrices for Monroe Harbor water.  Precision was
determined as the coefficient of variation (CV - Standard Deviation/Mean X
100) of multiple analyses of pooled QA samples.  For concentrations above the
detection limit, the precision for chromium, copper and zinc in whole and
filtered water ranged from 16 to 80% and showed concentration dependence for
whole water analyses (Table 7.5).

     Field precision, determined as the standard deviation derived from
duplicate measurements, was site- and metal-dependent.  However, the
resulting standard deviations were not concentration-dependent and could be
applied to each datum for appropriate sites and metals.

     Accuracies for copper, chromium, and zinc were determined by analyzing
EPA intercomparison samples (WP475 and WS378) and were within the 95%
confidence level as defined by EPA/Cincinnati.
                                     77

-------
        TABLE 7.5.  PRECISION ESTIMATED FOR POOLED SAMPLES IN TOTAL AND
                           DISSOLVED METALS ANALYSES

Chromium


Fraction
Total
Total
Total
Dissolved

..
Samel e
River Raisin
Monroe WWTP
Nearshore L.
Monroe WWTP


cvm
16
26
Erie 25
33
Mean
Cone.
(ua/L)
2.0
2.2
2.3
0.3
Coooer


CV(%)
25
17
18
-
Mean
Cone.
fuQ/L)
3.8
9.6
4.4
~
Zinc


CV(%)
80
5
60
~
Mean
Cone.
(ua/L)
5
41
10
-
     Recoveries, determined by standard additions, for copper, zinc, and
chromium were 105%, 104%, and 133%, respectively, and were independent of the
fraction analyzed (i.e., total vs. dissolved).


7.2  BIOLOGICAL STUDIES IN MONROE HARBOR (RIVER RAISIN), MICHIGAN

7.2.1  Introduction

     Biological studies conducted in the River Raisin were focused primarily
on determining whether chemical exposure exerted inhibitory or toxic effects
on biota.  Additionally, if toxic effects were observed, the goal was to
identify the most likely causative agents involved.  Influential factors are
very difficult to discern when evaluating complex mixtures of contaminants in
water and sediment.  For this reason, several statistical methods have been
applied to evaluate biological results.

     The majority of assays conducted during this study were whole-effluent,
toxicity-based tests, although additional bioaccumulation studies reflected
the chemical-specific approach.  A suite of bioassays were employed in which
test specimens were chosen to span a range of trophic levels represented by:
bacteria, algae, zooplankton, crayfish, insects, snails, clams, and fish.
The rationale for selecting an array of test specimens from varying trophic
levels is found in the varying response sensitivities of certain organisms to
environmental conditions and/or specific toxicants.  Toxicological effects
were addressed in terms of biotic function and, in some cases, structure.
Functional factors usually concern processes and process rates whereas
structure addresses community composition.

     Acute and chronic bioassays assessed inhibition/stimulation of primary
and secondary ecosystem functions.  Metabolic and physiologic processes
examined were organic substrate uptake, primary production via l4Carbon
uptake, grazing, survival, reproduction and fecundity, growth, and mortality.
Also assessed were bioaccumulation rates, body burden concentrations of
organic compounds, and morphological effects.  Variations in these factors
were determined spatially and seasonally.

                                     78

-------
     Field operations were conducted during 10 surveys throughout the
ice-free periods of 1983 and 1984 (Table 5.1).  Stations were longitudinally
distributed within the River and nearshore waters of western Lake Erie
(Figures 7.2 and 7.3).  All biological  samples collected in the field were
accompanied by samples used for physical measurements and chemical  analyses.
For in situ bioassays and bioaccumulation studies, samples or specimens were
re-introduced or introduced, respectively, to the system for various exposure
periods and then recovered.  Chemical analyses or measurements of functional
endpoints were then determined in the laboratory.  Other bioassays were
conducted entirely in the laboratory.

     Selection of control and/or reference stations is generally a difficult
task and is extremely critical.  Control and reference sites are usually
regarded as relatively unaffected areas, and are usually located above point
sources, in headwater reaches of rivers, or at sites where potential effects
are suspected to be at a minimum.  For this study, three control/reference
sites were selected.  The control/reference station selection process
consisted of three steps in this study.  First, sites suspected of being
potentially unimpacted were chosen; second, reconnaissance bioassays and
chemical analyses were conducted; and third, if little or no effects were
confirmed, the control/reference station was adopted for use.

     Station 1, in the headwater area of the River Raisin, was chosen as a
reference for this study because bioassay results indicated little or no
toxicological  effects, and contaminant concentrations in the water were
considerably lower than in downstream areas.  Because the River Raisin is a
complex estuary, there was a possibility of upstream effects during flow
reversals, but was never detected in this upper reach of the River.
Bioassays that required water for sample dilution to generate dose-response
curves utilized offshore, surface (1 m depth) water from Lake Erie and/or
Lake Huron.  These offshore waters had very dilute concentrations of
toxicants and caused little or no effects on test specimens.  The three
stations selected are best regarded as reference stations rather than
controls, but these terms are used interchangeably for this study.

     Bioassays that are conducted over a period of time require renewal water
for completion.  Large volumes of control water from Lake Huron and Lake Erie
were collected numerous times during these studies.  Various holding times
for these samples could not be avoided and there was a potential for chemical
changes during the holding period.  The effects of holding time may be
problematic and must be  investigated in the future.

     To examine toxicological effects and to determine causal factors, a
number of mathematical modeling  and statistical techniques were employed.
For this study, an integrated water quality method was developed which
combined conventional, single-chemical transport and fate modeling with
evaluation procedures based upon biological methods such as bioassays, in
situ population surveys, and water quality criteria.  The resulting new
methods of analyzing bioassay and population data were compatible with
transport and fate model requirements and with extensions of conventional
modeling methods to directly accommodate variability and toxicity.  Further,


                                     79

-------
80

-------
           union Camp  yyWTP Effluent Plume Stations
DAM 6
(573.0)
        -STATION NUMBERS
        DAM ELEVATION FT ABOVE MSL
             Figure 7.3.   River Raisin Station Array, 1983-1984.
                                      81

-------
the observed bloassay and population responses were related to chemical
concentrations to determine the most influential factors.

     To Identify potential causal agents in regard to bioassay results, two
approaches were employed.  In the first approach, correlation, canonical
correlation, and multiple regression analyses were applied.  In these
analyses, significant variables that most likely influenced bioassay results
were identified.  The second approach was founded on converting all chemical
and biological data into a uniform measure of toxicity, "toxic units", which
then formed a basis for comparisons.  A toxic unit is defined as the inverse
of the concentration or dilution fraction eliciting a 50% response in a given
bioassay (i.e., EC50 or LC50) and is determined from a dose-response function
(see Section 5.2.1).

     To evaluate the dose-response curve, a log-logistic model was applied
and dose-response data were reduced to toxic unit concentrations.  These
concentrations were utilized for examining the correlation of toxicity to
chemical concentrations to identify causal agents.  Toxic units were also
used for modeling toxicity directly as a water quality variable, and for
determining exposure probabilities within the context of deterministic,
steady-state models.  For a complete discussion of modeling methods, refer to
Sections 5.0 and 6.0.

7.2.2  Bacterial Decomposer Bioassav

     Decomposition of organic matter by bacterial populations is an integral
ecosystem function in aquatic and terrestrial environments.  Bacterial
degradation of autochthonous and allochthonous materials is critical to the
balance and cycling of energy and matter in aquatic ecosystems.  The primary
goal of this assay is to measure the inhibition/stimulation of heterotrophic,
decomposer activity and to detect toxicological effects on a distributional
basis.

     A modified method of Hobble (1969) was employed by McNaught e_t al_.
(1984) using acetate as the carbon substrate for uptake.  Samples were
innoculated with 1-15 yd of 3H-acetate and incubated in dark bottles for
two hours while being cooled by river water pumped shipboard.  Control
samples (reference samples), composed of 100% lake water (e.g., Lake Erie or
Lake Huron), and a dilutional series, consisting of 1:1, 1:3, and 1:9 ratios
of river water:lake water by volume were used for the assays.  The bacterial
uptake rates of acetate were determined and results were reported as percent
(%) inhibition or stimulation relative to the control.

     Bacterial uptake velocity results were interpreted seasonally,
spatially, and in regard to the dilutional sequence.  The most definitive
result obtained for the River Raisin was that greatest inhibition generally
occurred during low temperature periods and conversely, greatest stimulation
was evident at warm temperatures (Figure 7.4).  Within the study area,
bacterial uptake activity exhibited little or no consistent spatial patterns.
For example, on two consecutive days, two completely different spatial
patterns were observed (Table 7.6).  On the first day, all stations in the


                                    82

-------
300
200

Z
0
§
D
| 100
en
\
O
ffi
I
Z
# o

-100
A ^
A
-
Q
a
a
A
a
O
a
A a
A
.- ~ ~~ 50 100
9 D D
D D
6 8
s
- m m An '
                                                EFFLUENT
                                                  °C
                                               O  6-11

                                               D 12-17

                                                 18-24
Figure  7.4.  Bacterial Uptake Rate  Expressed as Percent
  Stimulation/Inhibition for Three  Temperature Ranges
                 and  Four Dilutions.
                         83

-------
  TABLE 7.6.  DECOMPOSER ECOSYSTEM FUNCTION:  RELATIVE INHIBITION/STIMULATION
       (PERCENTAGE CHANGE RELATIVE TO THE CONTROL) OF  BACTERIAL  UPTAKE  OF
                    ACETATE BY  100% RIVER OR OUTFALL  WATER
STATION                         10 MAY 1984                    11 MAY 1984
1
4
7
8
11
26
29
41
42
43
44
45
46
+ 33%
+ 67%
+ 50%
+ 33%
0%
0%
+1283%
0%
- 67%
+ 33%
+ 83%
0%
- 17%
- 43%
- 29%
- 29%
- 57%
NT
- 29%
- 43%
- 43%
- 14%
0%
NT
- 29%
- 29%
NT - Not tested.
River exhibited stimulation, as did most lake stations as well.  On day two,
all River and lake stations showed inhibition.  Considering the dilutional
series, most stations exhibited a curvilinear response, with inhibition
occurring at 10% dilution, stimulation at 25%, and inhibition at 50%.

     The fundamental process of organic matter degradation is intuitively an
important aspect of the ecosystem for toxicological assessment.  The acetate
method for bacterial activity assessment indicated that inhibition-primarily
occurred during cool water periods and that stimulation occurred at warm
temperatures.  This was reasonable since many process rates are
temperature-controlled.  However, results were extremely variable and spatial
patterns were very difficult to discern.  Similarly, little or no
site-specific consistency was observed and, therefore, no specific sites
which inhibited decompositional activity were identified.  Stimulation was
commonly observed in this study and was probably due to abundant bacterial
populations which were adapted to contaminant substrates.  In dilutional
series experiments, the specific curvilinear responses observed were
difficult to interpret, whereas a dose-response function with increasing
inhibition and increasing river water concentration would be more amenable to
evaluation and statistical analysis.  Although the concept of assessing
decomposer activity is sound, considerable methodological development must be
pursued to reduce variability and enable applications to field studies.
                                     84

-------
7.2.3  Phvtoplankton Carbon Uptake Bioassav

     Photosynthetlc activity and the fixation of carbon by phytoplankton
assemblages is a primary ecosystem function.  Primary producers form the
basis of planktivorous food chains and play a critical role in carbon
cycling, energy transfer, and oxygen generation, which are necessary for
maintaining ecosystem stability.  If autotrophic structure or production is
altered, repercussions may potentially result throughout upper food chain
elements.  The inhibition/stimulation of gross productivity by contaminants
was measured here using the traditional field assay employing the fixation of
radiolabeled bicarbonate (McNaught et aj.., 1984).

     Phytoplankton were collected at stations.longitudinally situated in the
River.  Samples were innoculated with 5 yCi l4C-biocarbonate and
resuspended at the collection depth (1 m) for three to four hours.
l4Carbon uptake rates were determined using a Beckman 1800 liquid
scintillation counter.  Methods used here for primary production assessment
followed the method of Vollenweider (1969).  Control  (reference) samples and
a dilutional sequence of 1:0, 1:1, 1:3, and 1:9 ratios by volume of river
water:lake water were assayed.  Phytoplankton uptake  rates of  I4C were
determined and results expressed as percent inhibition/stimulation relative
to the control.

     Inspection of raw phytoplankton results indicated that dose-response
functions varied greatly and that productivity was stimulated  at most River
stations (Figure 7.5).  Although phytoplankton toxicity was explicitly
evident at some sites and inhibition increased as a greater proportion of
experimental water was added, it appeared that the initial influence of
nutrient concentrations produced stimulation.

     When assay results were corrected for algal biomass using a weighting
factor derived from 100% control and 100% experimental water,  inhibition of
primary production was more apparent (Figure 7.6).  After appropriate
weighings were applied, inhibition was observed  in downstream  and Lake Erie
stations, relative to upstream  stations and controls.  Causative factors in
algal inhibition were not readily apparent.  However, factors  associated with
trophic state (nutrients) and possibly dissolved chromium, were significant
(Rz » 0.48) in analysis of covariance for explaining  the results in these
tests of complex mixtures.

     Results obtained using gross primary production  as an assay endpoint
exhibited site-specific consistency and an expected dose-response
relationship, where greater inhibition was obtained as greater proportions of
experimental water were added.  Although variability  in response was
observed, biomass standardization reduced variability and increased
interpretability (Bridgham et al.., 1987).   It appears that biomass
corrections must be applied due to differing algal concentrations,
biovolumes, and physiological states in the initial experimental and control
samples.
                                     85

-------
160
140
120
100
7 80
£
v. eo
n
E 40
O 20
o»
£
- 100
a! 80
>. 60
03 40
S 20
JULY 13 A, 1983 320
260
240
200


""*
|
UJ

• •«

	 r
11

— — — — . :^-
fl.ii
12345
160
r j 12°
: 80
: 40
JULY 14, 1
-
! * » v
-
.

.
-
6

1



_ t
!
1
!
;
;
983
-.
^
i
|
'

!

• ^ ^ ••




1
1234
|

1
I
1 ;


__





5 6
a
=
z
O
o
C 140
< 120
O
100
80
60
40
20

JULY 14A, 1983

320

280


"" _
- ;
~ .
- •
•
}




§
"B
t
.
:
•
•
3 :



r
[
a r

j.
>
•
. ;
?
1 2 3 4 S
STATION




' V
200
i 160
i 120
80
: : 40
:
JULY 15, 1983

-

-



•
K
v;
. :
•
6


-j
...

„
_ r
4
J

J



g
i
:
":

,








""


'
I |
1
ft*"*
B
Tl





""


123456
STATION
% SAMPLE
ini
10 25 SO
Figure 7.5.   River Raisin Primary Production for Three Dilutions,
                           July, 1983.
                               86

-------
     CD

     X
O
O
         8.0
             Photosynthesis    Station Averages
         6.0
     z
     o
     p   4-°
     0)
     o
     o
         2.0
          2.0
                  * change relative to L.  Huron control

                  * change relative to expected
                   0 --- oSTA 1
'•I.
                       STA 29
              «

                 V-0 STA  4

          «X N XX
          9 — 0- --- OSTA 26
                                     STA 1 .
                                      STA 26
                                      STA 29
              10   25     50            100


                         %  EFFLUENT
Figure  7.6.  River  Raisin Primary Production  Before and After
         Weighting for Initial  Algal Biomass,  1984.
                            87

-------
     Nutrients appear to be a confounding influence in phytoplankton toxicity
bioassays.  The problem Is particularly evident when oligotrophic, control
water is used to dilute nutrient-rich test water.  This may stimulate the
nutrient limited phytoplankton assemblages in the control water.  Another
problem associated with using resident, mixed populations, in phytoplankton
bioassays 1s that it cannot be determined which species are responding to
nutrients and which are responding to toxicants.  The nutrient response
usually masks any toxic response.  In addition, particular species within
these mixed test assemblages may be responding to different stimulants or
inhibitors.

     To avoid the confounding influence of nutrients and mixed phytoplankton
populations, the following recommendations are proposed for developing
phytoplankton bioassays.  Monoclonal cultures of phytoplankton species
representing several physiological groups (i.e., diatoms, green and
blue-green algae) should be used for bioassays.  The cultures should be
spiked with saturating concentrations of phosphorus, nitrogen, and silica and
maintained at steady-state conditions.  Nutrient spikes have been shunned by
some researchers because some phytoplankton taxa exhibit inhibitory effects
only when nutrient-starved, whereas others are inhibited during active
uptake.  Under steady-state conditions, variability is controlled and
parameters such as chlorophyll a concentrations, RCarbon uptake rate, cell  -
number, biovolume, and physiologic state are known.  Resident populations
should be removed from the test water by filtration before addition to the
monoclonal culture.  By controlling as many variables as possible and
negating the influence of nutrients and differential responses of mixed
populations, resultant responses can be more easily attributed to toxicity.

7.2.4  Zooolankton Grazing Bioassav

     Herbivore (rotifers, cladocerans, copepods) grazing is the principal
process involved in converting primary production mass into animal protein
for utilization by upper trophic levels.  Zooplankton grazing Is a secondary
ecosystem function.  Several toxicants have been shown to significantly
reduce zooplankton feeding rates in Great Lakes ecosystems (McNaught 1982;
McNaught gi aj.. 1987).  Inhibition/stimulation of zooplankton grazing was
determined to indicate potential toxicological effects on this secondary
ecosystem function in the River Raisin (McNaught et al_., 1984).

     Phytoplankton assemblages were collected from Lake Erie and incubated in
a Percival culture chamber.  Phytoplankton were cultured as a food source to
simulate densities found in the natural environment.  Control samples and a
dilutional sequence of 1:1, 1:3, and 1:9 ratios of river water:culture medium
were assayed.  Zooplankton were collected from Lake Erie; dominant forms
included Diaotomus sp., Cvcloos bicusoidatus. and Daohnia spp.  All
zooplankters were adapted to the algal food source for four hours prior to
introduction into the test vessels.  Animals were transferred into test
vessels using net carriers which allowed for the passage of water and
phytoplankton.  Zooplankters were allowed to feed for 15 minutes and then
removed.  Initial and final particle densities were determined using an
Elzone particle counter and were expressed as filtering rates (mL/an/time).

-------
Filtering rates were determined for the particle size range of 3 to 40
(equivalent spherical diameter).
                                                          m ESD
     The greatest zooplankton grazing inhibition was typically observed at
Stations 7 and 8 at the Monroe Vlastewater Treatment Plant (Table 7.7).
However, some Inhibition was exhibited at most stations, indicating the
presence of toxicants which inhibit feeding throughout the entire system.
The upstream station, number 1, generally exhibited toxicity at high doses
(low dilutions), but usually to a lesser degree than other sites (Figure
7.7).  Several patterns were observed in the dilutional series; the dominant
response was increased toxicity with increased dose (Table 7.7 and Figure
7.7).  Of the heavy metals investigated, zinc and to a great extent, copper,
may have had inhibitory effects on feeding rates; corrections for hardness
concentrations, which varied with dilution, appeared to be necessary.  At
particular stations, Pi apterous sp. feeding was distinctly more sensitive and
inhibited to a greater extent than that of Daphnia retrocurva.  Grazing
theory predicts that Diaptomus is representative of more oligotrophic
conditions and is probably the more sensitive species of the two (McNaught,
1975).

     This grazing bioassay usually produced consistent and reproducible
results over time and space.  In general, result variability appeared to be


     TABLE 7.7.  INHIBITION/STIMULATION OF ZOOPLANKTON GRAZING BY ADDITION
              OF 10, 25, AND 50% RIVER/OUTFALL (RIVER RAISIN) TO
                           CONTROL WATER (LAKE HURON)

       Filtering Rate (mL Per Hour)  and % Change  Relative to the Control,
                          Water Collected 9 May 1984.
                      Zooplankter Tested was Diaptomus sp.
STATION
CONTROL
Species Mean
10% OUTFALL
25% OUTFALL
               (-26%)
                     (-45%)
50% OUTFALL
1
4
7
8
11
26
29
41
42
43
45
46
0.60
0.46
0.47
0.48
0.29
0.46
0.41
0.39
0.47
0.521
0.41
0.45
0.50 (-17%)
0.30 (-35%)
0.28 (-40%)
0.22 (-54%)
0.20 (-31%)
0.06 (+30%)
0.24 (-41%)
0.31 (-21%)
0.34 (-28%)
0.31 (-40%)
0.33 (-20%)
0.38 (-16%)
0.44 (-27%)
0.11 (-76%)
0.13 (-72%)
0.12 (-75%)
0.19 (-34%)
0.47 (+2%)
0.33 (-20%)
0.21 (-46%)
0.29 (-38%)
0.27 (-48%)
0.21 (-49%)
0.21 (-53%)
0.27 (-55%)
0.09 (-80%)
0.02 (-96%)
0.00 (-100%)
0.16 (-45%)
0.43 (-7%)
0.21 (-45%)
0.07 (-82%)
0.20 (-57%)
0.07 (-87%)
0.12 (-71%)
0.09 (-80%)
                    (-67%)
                                     89

-------

0.5
0.4
0-3
.O
-^ 0.2
•c 0.1
OCT. 28, 1983 % SAMpl£
; H fl I
> ' \ 10 25 50
I,S'
_ \
\
V
...<-'-ii/.FL_.J--
. 0 rfi ij B n
« 1131415161718
i
U APRIL 3, 1984
IMP
< 0.8
K 0.7

O 0.6
5 0.5
N 0.4
K °'3
O 0.2
2 0.1
O
.
_

-
^ '
L -
/
L i
-
- ^
n , , I l i 1: i 1 : i 1;
< '
' /••-.'
1;/'< ' (,'.•
in &i ^
Mi Ha Kb :L
5
T
s
//,<.- / / ; /
//
? /

M s M
L ul 1
1   4   31   32  33   34   36
                                         38
a
0
0
N °7
Oc
O
0.5
0.4

0.3
C.2
0 1



"
— ,f
\
, i
"fc • s

- v
- ^
- ^





"


>
jij





/ '





MAY



,- / /

]

8

9,



, / /





198^





}\
' H
1

t





I
1







~ ;
\
•i ^
rl ^
E4 ^





/


-
|]
            1    4   41    42    43  45   46
                     STATION
Figure 7.7.   Zooplankton Grazing for Three Dilutions
          in the River Raisin, 1983-1984.
                        90

-------
very low.  Dose-response functions were typically the expected
linear-curvilinear response with increasing inhibition with increasing
concentrations of experimental  waters.   The methodological  technique
employed, which controlled phytoplankton and zooplankton numbers, undoubtedly
was the major factor in reducing result variability.

7.2.5  Zooplanktcm Reproduction Bioassavs

     Three assays were conducted to assess potential  toxicological  effects on
zooplankton reproductive processes (Dolan et al_., 1985; McNaught e_t aJL,
1984).  All assays employed species of the parthenogenic cladoceran genus
Ceriodaphnia.  In general, reproductive success may be impacted directly by a
toxicant(s) or may be a result of suppressed feeding activity.  Direct
physiologic effects may reduce the rate of reproduction or the number of
offspring produced.  On the other hand, feeding inhibition may impair
reproduction because organismal energy must be devoted to metabolism,
cellular maintenance, and survival.  Inhibition/stimulation of the
zooplankton reproductive process was assayed in the River Raisin.

     Two 7-day Ceriodaphnia chronic assays were conducted using Ceriodaphnia
affinis/dubia to assess reproductive success.  The primary difference between
the two assays conducted was the dilution factor used.  In one series of
assays, tests were usually performed on 100% experimental water.  In the
other bioassay, a dilutional series of 1:1, 1:3, and 1:9 proportions by
volume of river water:lake water were used.  Controls were used in both
assays and contrasted to experimental results.

     Stock cultures of Ceriodaohnia may be obtained from commercial sources.
Filtered, offshore lake water  (e.g., Lake Huron or Lake Erie) was used for
culturing medium.  Specimens were placed in covered Pyrex dishes with 1.5 L
of filtered surface water.  Four to five cultures were maintained
simultaneously and a new one was begun each week with 100-150 individuals.
Naturally-occurring bacteria were supplemented with 3 mL of Fleishmann's
yeast (5000 mg/1 L distilled water) as a food source every other day.  A
constant temperature of 25°C and a photoperiod of 16 hours/day maintained
healthy females.

     Test procedures followed  the method of Mount and Norberg (1984).  Single
specimens of uniform age were  placed in 30-mL Pyrex beakers containing
15 mL of test water at 25*C; usually ten replicates were conducted.  The test
solution was renewed on days 3 and 5 of the 7-day test by transferring
specimens with an eyedropper.  Individuals were fed yeast suspension daily at
the rate of 250 ug per animal.  Survival of test specimens was monitored
daily.  On test day 5, if young were produced, they were counted, then
discarded prior to adult transfer.  On day 7, the number of original females
surviving was recorded, the number of neonates counted, and the number of
broods per female calculated.  Data were analyzed using the statistical
procedure developed specifically for Ceriodaohnia toxicity tests, which
compensates for the effect of  mortality on fecundity  (Hamilton 1984).
                                      91

-------
     Lowest adult survival and greatest Inhibition of neonate production were
usually observed in downstream stations (particularly near and lakeward from
the Wastewater Treatment Plant), and compared to results at upstream stations
(Table 7.8)  The cumulative probability of mean young/adult for receiving
waters best indicated that toxicity increased from upstream to downstream
stations (Figure 7.8).  Reproduction assays using a dilutional series
generally exhibited decreased reproductive success with increased doses of
experimental water, although some variability was observed.

     Inhibition appeared to track well with the gradient of dissolved zinc in
the effluent of the Wastewater Treatment Plant, Station 7 (Table 7.3).
Multiple regression analysis of variables indicated that zinc, corrected for
alkalinity, was a significant factor in predicting brood size.  Canonical
correlation analysis indicated that, besides zinc, suspended solids,
dissolved copper, and pH may have been causal factors of inhibition.
Additionally, modeling results indicated that zinc and copper toxicity
increased by approximately 2 toxic units from upstream to downstream reaches
of the River (see Section 6.0).  Although brood size may be related to heavy
metal concentrations, the relationships with alkalinity and pH levels may be
synergistic and influential along with metals.  Under differing pH
conditions, metal species may become more available/unavailable to
Ceriodaohnia and this is reflected in results.

     On comparing results obtained in reproduction and grazing assays, an
interesting finding comes to light.  Suppression of both processes was
demonstrated and toxicity inferred, near the Wastewater Treatment Plant and
its plume (Figure 7.9).  Relative to results from the shore opposite to the
Wastewater Treatment Plant, greater reproductive inhibition was observed
           TABLE  7.8.   CERIODAPHNIA  BIOASSAY, MONROE  HARBOR,  CRUISE  3
                        (OCTOBER 31 - NOVEMBER 5, 1983)




STATION
1
7
13
14
15
16
17
18
19
8
%
SURVIVAL
OF
ADULTS
100
70
100
90
100
100
80
100
90
100
NUMBER OF
YOUNG/FEMALE
CERIODAPHNIA
sp.*
16.1
,10.4
15.0
18.1
19.8
19.4
22.2
22.3
21.0
22.9


MEAN NO. OF
BROODS/FEMALE
2.5
1.4
2.5
2.3
2.9
2.8
2.6
3.2
2.8
3.2


ALKALINITY
mq/L
228.0
181.0
188.0
218.0
220.0
225.0
226.0
227.0
225.0
177.0

DISSOLVED
ZINC
uq/L
1.7
32.9
25.0
8.7
6.5
7.3
4.7
4.3
3.3
3.0
*A11 data method
                                      92

-------
        CERIODAPHNIA FECUNDITY
5O
40

30
L_
i^
D
Q
\
0 20
z
D
O

^^



^c
LLJ
S


10

9

8


-

-


•








^

4

4 c
•* <«•
5 7

7
M
l l I I l I III





1

^444
1 4 5

A 5
4 7
_ 7
7
7
5


5

7



i i i


1
, .
44 4 4
4 7
4 5

5757
5 7


7













	 ! nl 	 J,


1 4
.',
5
3



















h 1 1 i i i i t
          10  20     50      80  90
                PROBABILITY
99
Figure 7.8.  Cumulative Probability of Ceriodaphm'a Fecundity
              in the River Raisin.
                   93

-------
40
W
H
- 20
D
O
x 10
0 8
6
6
ff
a
LU
c
1- 2
LU
Z
1
m

^^ • WWTP SHORELINE
0"""" — "^ • OPPOSITE SHORELINE
r \
' \
| \
' \
I \
/ Hx ^-*> _
N "^.^sB
• ' / ^^ _j— ^ "^ —
i / B^-*
j *
jj/
"W ~


i i i i i i i i
13 14  15  16  17  18

      STATION  NO.
                                                   19
(A
t-
Z
3
O
X
0
L
^
a
z
N
<
C
O


20


10
8
6

4



2



m
• WV.TP SHORELI
• OPPOSITE SHO
^_ -. — •'^»»w •
•^^•S. Vx ^
">**>**«. y
^*B "^N^
"*



-


i i i i i i i i
1 131415161718 19
                         STATION NO.
Figure  7.9.  Ceriodaphnja Reproduction and Grazing in the  Vicinity
           of the Monroe Wastewater Treatment Plant.
                               94

-------
compared to feeding inhibition.  This observation leads to the conclusion
that reproductive inhibition at this site is a direct physiologic effect of
exposure rather than a result of grazing suppression.

     Zooplankton reproduction bioassays demonstrated toxicological effects on
adult survival and neonate production in the River Raisin.  Results indicated
that the Wastewater Treatment Plant effluent (Station 7) had dramatic adverse
effects and toxicity increased in a lakeward direction as well.  Results of
statistical analyses suggested that heavy metals, alkalinity, and pH were all
influential factors in inhibition.  Comparative observations of grazing and
reproductive results showed that physiologic effects of toxicants may have
inhibited reproductive success directly rather than indirectly through
grazing suppression at this site.  Renewal water used during chronic tests
remains problematic because of changing biochemical factors associated with
holding time.  This factor can only be assessed through holding-time studies.
These chronic tests also required time-consuming and effort-intensive
maintenance to obtain optimal results; this test cannot be regarded as a
rapid bioassay method.

7.2.6  Zooolankton Reproduction and Fish Lethality Bioassavs of Sediment

     In its "1983 Report on Great Lakes Water Quality:  Appendix" (IJC,
1983), the International Joint Commission Dredging Subcommittee reported that
"In the Great Lakes, the role of sediments as a significant source of fish
contaminants and their effects on human health has not yet been adequately
investigated and/or demonstrated".  The report goes on to say that "the
effects of the contaminated sediments on the aquatic biota in the vicinity of
the overlying waters should be assessed", and that "on-site fish toxicity or
bioaccumulation studies should be undertaken to determine if toxic materials
of concern are leaving the contaminated sediments".

     Preliminary results from the 1983 Monroe Harbor study indicated that a
major problem in the River Raisin might be in-place pollutants.  Sediments
from Station 4 were found to be highly contaminated with  PCBs and heavy
metals.  Because the turning basin sediment can be greatly disturbed by
turning freighters, it was decided to explore whether a simulated
resuspension of sediments, using a sediment elutriate method, would result in
increased toxicity to aquatic organisms.  Three bioassay  experiments with
Station 4 sediment were conducted between July and October 1984, using
Ceriodaphnia affinis/dubia and fathead minnows (Dolan et  al_., 1984).

     The objectives of the first sediment bioassay were to determine if
Station 4 sediments would produce any demonstrable toxic  effects and to
evaluate the feasibility of adapting the larval fathead minnow and C.
affinis/dubia methods to a sediment elutriate test.  The  objective of the
second sediment bioassay was to  simulate a range of sediment perturbations
(resuspensions) to demonstrate a possible dilution effect using different
sediment:water ratios.  The third experiment was designed to determine
whether relationships between elutriate phase toxicity  and elutriate water
chemistry could be established.
                                     95

-------
     The sediment preparations followed a modified liquid phase-elutriate
method developed at the USEPA-Corvallis (Cairns el il_., 1984).  The seven-day
bioassay procedures followed Mount and Norberg (1984) and Norberg and Mount
(1983, draft) for Ceriodaohnia and fathead minnows, as described previously.
Sediment was collected by Ponar grab from Stations 4 and 12 (control), along
with grab water samples.  In experiment II, elutriate from a sediment:water
ratio of 1:4 was "prepared by placing the sediment and water in glass jars and
agitating them for 30 minutes at 250 rpm on a New Brunswick Scientific G10
gyrotory shaker.  The sediment was allowed to settle out for 18 hours and the
supernatent (elutriate) was siphoned off for the toxicity tests.  Replacement
elutriate was generated by adding fresh water to the same sediment and
repeating the steps just described.  It was necessary to aerate all
elutriates prior to the bioassays, bringing their dissolved oxygen
concentrations to 8.3 - 8.5 mg/L.

     Samples were prepared in duplicate.  Duplicate data were analyzed for
statistically significant differences within treatments.  If no significant
differences in reproduction, survival or dry weight were found within
duplicate treatments, the data were pooled for one-way ANOVA and pairwise
comparisons against other treatments.

     In experiment #1, the Station 4 elutriate caused 100% fathead minnow
larval mortality within 48 hours, while the Station 12 control elutriate
exhibited 100% survival.  The Station 4 elutriate also caused 100% inhibition
of Ceriodaphnia reproduction.

     Because a 1:4 sediment to water ratio at Station 4 caused complete
mortality of the larval fish and complete reproductive inhibition of
Ceriodaphnia. additional ratios of 1:8 and 1:12 were tested for experiments
#2 and #3.  All other procedures were identical to those in experiment #1.
Only the Ceriodaphnia assay was performed.  A dilution effect was apparent
from the experiment #2 data.  The 1:4 elutriate from Station 4 sediment
reduced daphnid survival by 30-40% and reproduction by about 50%, while the
effects of the 1:12 ratio were virtually indistinguishable from those of the
controls.

     Experiment #3, surprisingly, showed only a decrease in larval fish
growth at Station 4 at a 1:4 elutriate, while larvae survival and
Ceriodaphnia survival and reproduction were not significantly (p < 0.05)
affected by any of the elutriate solutions.  The drastic change in toxicity
at this one station over just 4 months was completely unexpected.  It may be
an indication of real temporal variability in sediment toxicity, due to
unknown processes; it may be an indication of extreme spatial sediment
heterogeneity, and thus an artifact of inadequate field sampling; or it may
be an indication of unsuspected problems associated with actual toxicity test
procedures.  Consequently, it is recommended that sediment elutriate toxicity
tests should investigate temporal and small-scale spatial variability.
                                     96

-------
7.2.7  Fathead Minnow Toxicitv Bioassavs

     In addition to the bioassays employing the cladoceran Cerlodaphnia.  a
similar test using fathead minnow larvae (Pimeohales oromelas Rafinesque) was
performed to Investigate potential toxicological  effects of river water on
larval fish growth and survival (Dolan et al_.,  1984).  As in the daphnid
assays, the larval minnows under the test conditions were subject to both
direct and indirect toxicological effects.

     The procedure used was the seven-day renewal  method of Norberg and Mount
(1982).  The test tanks were constructed from plate glass and measured
30.5 cm x 15.2 cm x 10.2 cm deep.  Each tank was subdivided into four
replicate compartments, each 12.7 cm x 7.6 cm x 10.2 cm.  A wire mesh screen
(30.5 cm x 10.2 cm) covered the open ends of the compartments allowing a
one-inch sump area from which water was removed or renewed without disturbing
the larvae.  Two liters of test solution were used in each test tank and
water was renewed daily to maintain acceptable dissolved oxygen (>6 mg/L)
concentrations.  Ten fish were placed into each of the four replicate
compartments with a large bore pipette, resulting in 40 larvae per treatment.
The test was conducted at 25°C with a photoperiod of 16 hours/day.  Fathead
larvae were fed newly hatched (less than 24 hour old) brine shrimp nauplii on
a daily basis at the rate of about 1 ml of dense nauplii per compartment per
day.  At test termination on the seventh day, fry were counted, dried at
105*F in preweighed aluminum boats for at least two hours, and weighed to the
nearest 0.0001 gram.  Four replicates of 10 pre-test larvae each from Day 1
were also collected at random as controls for dry weight (growth) comparisons
with test larvae.  Mean dry weights were calculated by dividing the total
group weight by the number of larvae in the group.

     These assays were performed on river water and effluents collected on
surveys 3, 4, 6, 7, 8, 9, and 10.  The larval fish were obtained from the
U.S. EPA Newtown Fish Toxicology Station in Cincinnati, Ohio.  Mortality and
growth data were generated from tests using both 100% (undiluted) sample
water and a serial dilution (100, 50, 10, 5 and 1%) of sample water.
Dilution and control water was either Lake Huron or offshore Lake Erie water.

     The 1983 mortality assays (surveys 3 and 4), like the Ceriodaohnia
tests, showed increasing mortality at the downstream stations, particularly
within the Monroe Wastewater Treatment Plant (WWTP) effluent plume.  The
elevated toxicity decreased in the lower river as the River Raisin plume was
diluted with Lake Erie water.  In 1984, however, larval survival was
uniformly high throughout the length of the river.  No cause has yet been
determined for this apparent decrease in toxicity between years.

      In 1983 the larval growth assay was performed only during survey 3 and
only on the WWTP effluent (Station 7).  Although there was some variability
in the dilutional series trend,  there was an apparent decrease in growth with
an increasing concentration of effluent.  In 1984, utilizing water from a
series of stations located along the length of the river, there was an
apparent increase in toxicity  and an associated decrease in larval growth at
the downstream stations in 4 of  the 5 surveys.

                                      97

-------
     Multiple regression analysis of larval fathead minnow survival and
growth and dissolved metals concentrations showed a large degree of
variability between surveys and did not reveal  any statistically significant
(p < 0.05) relationships.  This contrasted with Ceriodaphnia data in which
there was a positive correlation between toxicity and dissolved metals.

     Overall these tests indicated an increase in toxicity at the downstream
stations, particularly below the WWTP.  Dilutions within the river and of
river water entering Lake Erie moderated this toxicity.

7.2.8  Fathead Minnow Flow-Through Bioassav

     During survey 1, the Michigan Department of Natural Resources (MDNR)
Surface Water Quality Division conducted two 96-hour, continuous flow-through
bioassays on the chlorinated and dechlorinated effluents of the Monroe
Municipal Wastewater Treatment Plant (WWTP).  The test organisms were adult
fathead minnows (Pimeohales promelas). and mortality was the test endpoint.
Tests were conducted in a MDNR bioassay trailer on the grounds of the WWTP.
A continuous flow diluter produced test water dilutions of 2, 5, 7, 10, 13,
15, 18 and 20% chlorinated effluent, and 12, 25, 50, 66, 75, 88 and 100%
dechlorinated effluent.  Diluent and control water used was River Raisin
water collected approximately 50 feet upstream from the WWTP outfall, eight
feet from the southern shore.  Test solutions were monitored for pH,
conductivity, temperature, dissolved oxygen, alkalinity, hardness and total
residual chlorine.

     The dechlorinated effluent produced no mortalities nor any observable
stress symptoms after 96 hours exposure.  The chlorinated effluent, on the
other hand, was acutely toxic to the fish at concentrations as low as 10%.
The 96-hour LC50 occurred with 13% effluent within 95% confidence limits of ±
1%.  Toxicity was attributed to residual chlorine, which averaged 2.9 ± 1.1
mg/L in the undiluted chlorinated effluent and 0.4 mg/L in the 13%
concentration tanks (i.e., at the LC50).

7.2.9  In-Situ Bioaccumulation of PCBs in Clams. Fathead Minnows and Channel
       Catfish

7.2.9.1  Approach--

     The Michigan Department of Natural Resources annually issues consumption
advisories for carp caught in the lower River Raisin due to their elevated
body burdens of PCBs.  These advisories are based on data derived from the
capture and analysis of free-living resident fish.  While this procedure
provides Information about the general occurrence of contaminants in aquatic
organisms and is appropriate for establishing consumption guidelines, the
unknown and highly variable mobility of the resident fish precludes their use
as indicators of site-specific contaminant availability.  Such site-specific
information, however, is useful for the identification of contaminant "hot
spots", and this in turn is useful to the manager who wishes to evaluate and
prioritize specific locations for possible remedial actions.
                                      98

-------
     An increasingly popular procedure for evaluating the in situ.
site-specific bioavailability of aquatic contaminants involves the  caging and
exposure of organisms at specific locations for time periods sufficient to
approximate, or allow the calculation of, contaminant equilibrium.   A number
of organisms and caging schemes have been successfully used (e.g.,  Kauss and
Hamdy, 1985; Pugsley e_I al., 1985; Rice et a!., 1984), but a review of the
literature points out several limiting considerations:

     - The organisms must survive in the often degraded environments
       of interest.

     - The organisms must contain sufficient tissue mass to permit  the
       quantification of trace contaminant concentrations.

     - The caging scheme must allow the organisms to carry out normal
       physiological activities, particularly feeding and respiration,
       for the duration of the exposure.

     Three organisms were chosen for the Monroe Harbor studies with the above
considerations in mind:  adult bivalve mollusks (Lamosilis radiata  Barnes and
Anodonta grandis Say), adult fathead minnows (Pimeohales Drome1 as Rafinesque).
and yearling channel catfish (Ictalurus punctatus Rafinesque).  All contained
enough tissue (>20 g) to provide an adequate mass for analysis when
composited as described in 7.2.9.2.  Bivalves are filter feeders and
therefore capable of feeding while caged.  Fathead minnows are able to
survive on the organisms and organic matter suspended in the water column and
on periphyton growing on the cages.  Channel catfish are benthic omnivores
and are suitable only in cages resting on the bottom.

     Two bioaccumulation experiments were performed, one in 1983 (Rathbun and
Smith, 1985) and one in 1984 (Rathbun et al., 1985).  The 1983 study
investigated two principal questions:

     - Will clams accumulate total PCBs and the PCB homologs in
       proportion to their concentrations in the surrounding environment?

     - Are there differences in the bioaccumulative capabilities of
       the two clam species  and of the two sexes of L. radiata?

     The 1984 study investigated two additional aspects of bioaccumulation:

     - How do water and sediment compare as sources of bioavailable  PCBs?

     - Is a 35 day time-series collection scheme adequate for the
       observation  (or extrapolation and calculation) of equilibrium
       bi concentration factors (BCFs) for total PCBs and the PCB
       homologs  in clams and fish?
                                      99

-------
7.2.9.2  Methods--

     In bioaccumulation studies, It is essential to obtain experimental
organisms with low initial body burdens of the contaminant(s) of interest.
Commercial fish hatcheries are usually reliable sources of "clean" fish;
nonetheless it is appropriate to conduct preliminary analyses prior to the
start of the experiment when utilizing new or questionable sources.  The
adult fathead minnows (45-65 mm) and the young-of-the-year channel catfish
(110-130 mm) used in the 1984 study were obtained from Kurtz's Fish Hatchery
in Elverson, Pennsylvania and Spring Valley Trout Farm in Dexter, Michigan,
respectively.

     Bivalves may also be purchased from a limited number of commercial
suppliers, but the species available may not be native to the Great Lakes
basin (e.g., Alasmidonta varicosa). the origin and history of the mollusks
are usually unknown, the taxonomic information provided by the supplier is
often questionable, and at prices of up to $1.00 apiece they are rather
expensive.  Consequently, most workers utilize natural populations of native
bivalves.  Clams (6-10 cm in length),  for both the 1983 and 1984 River Raisin
studies were collected by hand from the shallow waters of Station 12 in the
River Raisin above Monroe, on both the upstream and downstream sides of the
Telegraph Road Bridge.  This station is upstream of any known local
industrial outfalls, exhibited diverse aquatic plant and animal communities,
and was utilized as a control station.  In 1983 i. radiata and &. grandis
were collected, while 1984 only the more abundant L- radiata was collected.
These two species are among the most common unionid bivalves in the Great
Lakes basin (Mathiak, 1979).  Clam collecting took approximately 16 manhours
each year.

     It should be noted that clam taxonomy can be quite confusing to the
beginning malacologist.  None of the most readily available keys is without
flaws, but suggested references are Burch (1973), Pennak (1978), and Clarke
(1981).

     Containment and exposure of organisms at a specific site poses several
problems.  The containing device must  limit the free movement of the
organisms for recollection purposes yet allow normal physiological activities
like feeding and respiration; it must  be strong enough to withstand repeated
retrieval and handling in the field yet be porous so as to allow free water
circulation; it must be large enough to reduce abnormal crowding of the
organisms yet small enough to be conveniently handled; and it must be placed
and marked so as to permit easy locating and access by the researchers, yet
not be a hazard to other water users nor attractive to vandals.  With these
considerations in mind, submerged, buoyed cages (Figure 7.10) were chosen as
the exposure scheme for these studies.

     Clam cages in 1983 consisted of 45-cm lengths of 10 cm (ID) black,
Nalgene Type II pipe, with plastic screen (1.3 cm openings) covering the ends
and twelve 2-cm holes drilled along the sides.  Cage construction took
approximately 16 manhours.  Twelve clams were placed in each cage, which were
suspended from buoys 1 m below the water surface at Stations 1, 3 and 4, and


                                    100

-------
                                                                      -\
                                   1983
                          1 arr Case
                          x 60cn-"x  25cn
                                    ? Q1--
Figure 7.10.   Cage Designs for Monroe Harbor Bioaccumulation Studies,
                                   101

-------
anchored to the bedrock bottom at shallow Station 12.  These sites were
chosen to provide a gradient of whole water total PCB concentrations.  Single
samples were collected after 25 days of exposure.  The length of the exposure
period was chosen after a review of the literature (e.g., Hartley and
Johnston, 1983; Kauss el al_., 1981).

     A greater number of organisms was required for the 1984 experiments,
necessitating a change in cage design and a reduction in the number of
sampling sites.  Clam cages were constructed of 1.5-cm mesh galvanized metal
hardware cloth, 60 cm x 60 cm x 26 cm high.  Fish cages were constructed of
0.75-cm mesh hardware cloth, 60 cm x 60 cm x 60 cm high.  Cage construction
took approximately 40 manhours.  At Station 12, a clam and a fish cage were
each anchored to the bottom in shallow (1 m) water.  At Station 4, six clam
and six fish cages were suspended from buoys; three clam and three fish cages
hung at 1 m below the surface, and three clam and three fish cages rested on
the bottom in about 3 m of water.  Each clam cage held 35 clams.  It should
be noted that, judging by the mud adhering to the cages and the clam shells
at Station 4, the cages positioned on the bottom sunk several centimeters
into the sediments.  This apparently permitted limited burrowing by the
clams.  Each surface fish cage held 150 fathead minnows and each bottom fish
cage held 50 catfish.  In addition, one bottom fish cage held 50 fathead
minnows.  Triplicate samples were collected after 1,  2, 4, 9, 18, 25 and 35
days of exposure, based on the recommendations of ASTM (1984).  Exposure
occurred between July 24 and August 27, 1984.

     In 1983, whole water samples for organochlorine analysis were collected
daily during the first week of exposure.  In 1984, whole water was sampled on
Days 1, 2, 4, 9, 18, 25 and 35.  Section 7.1 of this report provides details
on the water sampling procedures.

     Sediment samples were not collected in 1983.  In 1984, a grab sediment
sample was collected with a Ponar grab on the first day of exposure.  Section
3.3 of this report provides details on the sediment sampling procedure.

     Sample collection (biota, water and sediment) took approximately 20
manhours in 1983 and 70 manhours in 1984.

     All biota samples from both the 1983 and 1984 studies were returned to
LLRS and frozen.  All samples were processed with their gut contents intact.
Clams were later shucked by hand and the tissue homogenized in a blender.  In
1983, the easily distinguished sexes of L. radiata were processed separately
in order to examine the influence of sex on accumulation.  In 1984 the sexes
were combined.  Whole fish in 1984 were also homogenized in a blender, and
the sexes were combined.  Biota samples were processed as composite samples
whenever possible.  In 1983, each clam sample was composed of from 1 to 6
clams.  In 1984, each clam sample was composed of 5 clams; each fathead
minnow sample, of from 10 to 30 fish; and each channel catfish sample, of
from 5 to 12 fish.  Single samples were prepared for each station in 1983,
whereas in 1984 samples were collected in triplicate.  Biota sample
processing took approximately 12 hours in 1983 and 100 hours in 1984.
                                     102

-------
     Details of the extraction procedures used for biota samples may be found
in Rathbun and Smith (1985), for water samples in Smith e£ al_. (1985), and
for sediment samples in Filkins et aj.. (1985).  Briefly, the biota samples
were Soxhlet extracted for 48 hours with hexane and dichloromethane.  Water
samples (2-12 L) were solvent extracted with dichloromethane (10 parts water
to 1 part DCM) in 4 L amber glass bottles containing 1  L of sample/bottle.
Sediment samples-were Soxhlet extracted for 48 hours with hexane and acetone.
Sample extractions (biota, water and sediment) took approximately 50 manhours
in 1983 and 12 manweeks in 1984.

     All extracts from 1983 were cleaned of lipids with concentrated
HoSC^ (Rathbun and Smith, 1985).  Because of cleanup problems encountered
with some of the water samples (see Section 7.14), extracts from 1984 were
cleaned on Florisil columns (Smith et il., 1985).  Extract cleanup took
approximately 8 manhours in 1983 and 9 manweeks in 1984.

     Analytical procedures are detailed in Rathbun et aJL, 1985.  Briefly,
extracts were analyzed on a VARIAN Model 3700 gas chromatograph equipped with
a 50-m fused-silica capillary column and a WN1 ECD.  Sample peaks were
quantified by comparison to a standards library of 110 congeners via a
modified version of the multiple regression program COMSTAR (Burkhart and
Weininger, 1987).  Sample analysis took approximately 3 manweeks in 1983 and -
20 manweeks in 1984.

7.2.9.3  Results--

     The cage deployment schemes utilized in these studies appear to have
avoided the potential problems outlined in 7.2.5.2.  Loss of organisms due to
mortality were not recorded in 1983 but was minimal in 1984:  less than 1%
for the clams and channel catfish and about 5% for the fathead minnows.  The
organisms were apparently feeding normally as judged by their lipid contents;
in 1983 there was no obvious trend in clam lipid content, and in 1984 the
clam, surface fathead minnow, and channel catfish lipid contents did not
change significantly (p < 0.05) over the 35 days, while lipids in bottom
fathead minnows  actually increased significantly  (p < 0.05).  Condition
factors (wet weight/total length) and tissue moisture contents, other
measures of organism health, also did not change significantly (p < 0.05)
during the 1984  study.

     The 1983 data indicated that the total PCB bioconcentration factors
(BCF; tissue PCB concentration/water PCB concentration) of the three clam
groups (i.e., A. grandis and the two sexes of L- radiata) at  the three
exposure sites varied by less than one order of magnitude both within and
between stations despite a 43-fold difference in whole water  total PCB
concentration between stations  (Table 7.9). This  suggests that uptake of
total PCBs by these clams was independent of clam species or  sex or whole
water total PCB  concentration.  Also, all three clam groups accumulated PCB
homolog patterns similar  in composition to those  in their respective
environments  (Table 7.9) despite substantial changes in whole water homolog
patterns between the stations (see Section 7.1.4.3 and  Figure 7.1 for further
discussion of these PCB homolog changes).


                                     103

-------

>»
8
i~
u>



f*»
g
h-
<
H-
CO



«••
5
H-
Crt





=55
<-!i
01
a
«> wo>
— a*
l-*i
u. t_
a
*• o>
«i a *
S Jl'| i
41 U
-> t> -1
55?
«
*™
-It
O)
« *- a»
— « je
1^1
u. u
2 a,
«l <0 -X
5-'|f
±fe_,
55?
(0
'•55
-It
o
ca
«l WO)
l-'^l
u. c.
•
w 01
a> an
5Ji!if
ffc^
55?



K
o
-
Kl
in
o

1
CD
•O
O
00
0
•o
in
o
AJ
(M
O
o
0
s
a
r*i
a
0
S
0-
1
1
H-


o> co in C §
o >c> ui co >o ro •-' o o o
«- in r-
ro ^ M >8 8
ooo'O^jr'i'-ooo
»-OO
*™ «— st' CM
**rcoK>mSo
M^jjt;So'plJ*~O00
°i *^ °. ^
OCOt^OK10s!>»C3O
to r>j «-
PU •* s» C> SG
oi^roioinr^^*«^c30
(NJ ^ *"
ru
 CM
O-O(MO-*>*rO-*tOOOIMcS(5
r^omo.«-rorjooo
•- CM IO »-
>* r>-
oom- CM CO
-*O(MinO>*f^N-OO
»- «- •- CM «-
00 GO
oo*»inh.cM^Nt<-o
CM •- CM «- •-
tn
(MOCO -*K(N-»--*
Kt-OCOO-OCNJ^^OO
«- M - OIL i— o >- < _i
XOXO-JOU
104

-------
     Within 2 to 4 days of exposure,  the 1984 data indicated that the clams,
fathead minnows, and channel catfish  accumulated the same PCB homolog pattern
as that found in the surrounding water and sediment, a pattern they all
retained throughout the 35 days of exposure.   Based on observed accumulation
profiles (Figure 7.11), thirty-five days was  not sufficient for the organisms
to attain equilibrium with regard to  total PCBs.  There was no significant
difference (p < 0.05) in total PCB accumulation between animals caged in the
water column compared to those  on the sediments, although the sediments
contained 10,000 times more total PCBs (1.9 mg/kg versus an average of 150
ng/L) than the whole water.  By Day 35, the two fish species had accumulated
significantly (p < 0.05) more total PCBs than the clams, with concentrations
expressed on a wet weight basis.  Correcting  for the lipid content of the
samples (wet weight PCB concentration/1ipid content), however, reduces the
difference between PCB concentrations in clams and fish at Day 35 to
statistical insignificance  (p < 0.05; Figure 7.12).

     It should be noted that clams usually have a lower lipid content than do
most fish.  In the 1984 study, for example, the L- radiata contained an
average of 0.43% total lipid over the 35 days of exposure, while the channel
catfish averaged 4.17% and the fathead minnows, 3.48%.  Tissue lipid content
is an important parameter influencing the bioaccumulation of organochlorine
compounds, and it should be considered when evaluating organisms for use in
experiments such as these.

7.2.9.4  Evaluation--

     The 1983 experiment demonstrated that bivalve mollusks were suitable
organisms for in situ bioaccumulation studies in Monroe Harbor.  Despite
substantial changes in whole water total PCB concentrations and PCB homolog
patterns between stations after 25 days, the total PCB BCFs were similar for
all three clam groups and the homolog patterns accumulated by the clams were
similar to those found in the surrounding water.

     The first finding, that the clam total PCB BCF was independent of the
whole water total PCB concentration,  suggests that it would be theoretically
possible to derive whole water concentrations from known organism tissue
concentrations and an assumed BCF.  This would eliminate the difficulty many
labs encounter when quantifying trace concentrations  in water samples.  While
the 1983 data suggests that this is feasible, much more information on the
variability of bioaccumulation processes is needed before such calculations
could be relied upon.

     The second finding, that the clams accumulated PCB homolog patterns
closely resembling those of the surrounding water, suggests that these
species might be useful for qualitatively identifying sources of PCBs that
exhibit distinctive homolog "fingerprints".  This approach could be made
quantitative by the application of appropriate pattern-recognition
statistics, and could be made more definitive by considering congener data
rather than homologs.
                                     105

-------
oo
CT)
 C
,g

"o
                                                               c

                                                          S  I  I
                                                            >  u  o
                                                                            o


                                                                            <-,
 l/l

 o
 c
 c
                                                                                                      •o
                                                                                                       to
                                                                                                       0)
                                                                                                       (/I

                                                                                                       03

                                                                                                      CJ
 •*-> .C
 .c l/>
 OVr-
 •t- «t-
 tt) •!->
 OJ r—
2 OJ
--- C
    c
CO IQ
O -C
CL. O
                                                                                                       QJ
                                                 106

-------
oo
_g
"5
                                                             I I
                                                             o o
                                                             m ID

                                                                       m  o

                                                                       O  <1
                                     \
                                            \
        O
        ID
O
o
                                                                              . in
                                                                               oo
                                                                                  o
                                                                                 o
                                                                               en

                                                                               10
                                                                               OQ
                                                                               O
                                                                               Q.
                                                                               (N
o
o
o
m
                                                                               O
                                                                               OJ i—
                                                                               S-  (U
                                                                               s-  c
                                                                               O  C
                                                                               o  na
                                                                                  .c
                                                                                I  O
                                                                               C  C
                                                                               O)  A3
                                                                               4->
                                                                               C   •
                                                                               O  (/I
                                                                               O  2
                                                                                   O
                                                                               •o  c
                                                                                                        TJ
                                                                                                     **- (O
                                                                                                     O O)
                                                                                                        -C
                                                                                                     O) 4->
                                                                                                     -at 
-------
     Data from the 1984 study showed that clams caged on the sediments did
not accumulate more total PCBs than did clams caged in the water column,
despite the much higher concentration in the sediment.  This suggests that
the PCBs associated with the bedded sediments were much less bioavaiTable
than those present, in whatever phase, in the water column.  The experimental
designs employed in this study, unfortunately, did not permit an evaluation
of the contributions of the dissolved and suspended particulate phases to PCB
uptake.  The algal bioayailability experiments (Section 3.3.4.3.2), however,
demonstrated that a radiolabelled hexachlorobiphenyl bound to resuspended
sediments was taken up by three species of algae.  Future in situ
bioaccumulation studies should investigate the relative contributions of
these phases to PCB uptake in different organisms.

     Results of the 1984 study also suggest that 2 to 4 days of exposure was
adequate time for the channel catfish, fathead minnows, and clams to
accumulate the PCB homolog pattern of the surrounding water and sediment.
This, along with the 1983 findings, supports the use of caged organisms as
qualitative tracers of PCB sources processing distinctive "fingerprints".
The similarity of the whole water and sediment PCB homolog patterns also
suggests that they share the same PCB source, or that one is the source for
the other.

     While similar homolog patterns were quickly accumulated by all of the
organisms, the 1984 data also indicated that 35 days of exposure was not
sufficient to attain equilibrium with total PCBs.  The data do not seem
appropriate for nonlinear extrapolation to steady state concentrations,
although this is still being evaluated.  Since extending the exposure time
significantly increases problems of changing water temperatures and
vandalism, future experiments that require data for equilibrium
concentrations might try an accelerated bioaccumulation study design (Branson
e_t a!., 1975).  In this design, a short term (4 days) accumulation period is
followed by a depuration phase.  Uptake and depuration rate constants (Ku and
Kd) are calculated, and the steady state BCF is calculated as Ku/Kd.  This
procedure is cheaper than the time series design employed in 1984 and is
especially appropriate for biologically stable contaminants with large BCF
values, like PCBs.

     It should be noted that, as a group, PCBs include congeners of
substantially different partition coefficients and stereochemical structures.
As such, a total PCB value alone is not adequate for evaluating potential
toxldty or other effects.  Data on individual PCB congeners is required for
these purposes.  These two studies employed high-resolution, capillary gas
chromatography of sufficient sensitivity to produce such data, and additional
work with these data will investigate the influence of stereochemistry on the
accumulation of Individual PCB congeners.

     Two Important aspects of bioaccumulation not addressed by these two
experiments are:  contaminant depuration kinetics, which would provide
information on organism detoxification capabilities and ecosystem response to
remedial actions, and; physiological "stress tests", which would allow
partial verification of the assumption that accumulated organism body burdens


                                    108

-------
reflect the ambient contaminant concentrations in seriously degraded
environments.

7.2.9.5  Additional In Situ Bioaccumulation Experiments--

     Prior to choosing native bivalves, channel  catfish,  and fathead minnows
for use in the in situ bioaccumulation experiments just described,  a number
of other organisms were evaluated as potential test subjects.  The  Intent of
these preliminary experiments was only to evaluate caged organism survival
and, therefore, no analytical samples were collected (with one exception
described in the following).

     Mayfly larvae (Stenonema sp. and Isonvchla sp.), snails (Campeloma sp.),
and sphaeriidian clams were collected at Station 12, and crayfish were
purchased from a local bait store.  All organisms were exposed to subsurface
station 4 water in plexiglass grid cages, composed of thirty-six 5 cm x 5 cm
x 5 cm cells with 0.3 cm mesh stainless steel screen on each side (not
illustrated), for periods of 4 to 8 days in the summer of 1983.  The mayfly
larvae survival was poor, around 50%.  Survivorship of the snails and
sphaeriidian clams was much better at 75 to 100%.  These organisms, however,
presented analytical difficulties in that the tissue could not easily be
separated from the shell material.  Crayfish survival was also high, at
90-100%.

     Consequently, crayfish were included in the 1983 bioaccumulation study,
described earlier.  They were exposed in the grid cages for 25 days at
stations 12, 1, 3 and 4.  After collection the samples were homogenized in a
blender, extracted, and analyzed as described in Section 7.2.6.2.  Upon
analysis, however, the extracts were found to contain a series of interfering
compounds that obscured many of the PCB peaks and prevented quantification.
They were tentatively identified as petroleum hydrocarbons, but were not
found in the 1983 or 1984 clam sample extracts.  An additional problem with
these crayfish was that their origin was not known; the bait store bought
them from a number of private collectors and no reasonable amount of
preliminary analytical screening could assure that all of the experimental
organisms were free from contamination.  Lastly, the tissue lipid content of
the crayfish, unlike that of the clams, decreased substantially during the 25
days of exposure.  The control sample contained 1.45% of total lipid while
the caged samples ranged from 1.00 to 0.54%.  Apparently, the crayfish did
not or could not feed adequately while caged in the water column.

     After considering the results of these preliminary experiments, we chose
the large filter-feeding unionid bivalves, adult fathead minnows, and channel
catfish for use 1n the in situ bioaccumulation experiments.

7.2.10  Resident Larval Fish Studies

     Nursery areas and the successful recruitment of larval fish are
essential to the stability of fish populations.  Undoubtedly, estuaries and
river mouths in the Laurentian Great Lakes serve as spawning and nursery
grounds for many species.  The structure, growth, mortality, and pathology of

                                     109

-------
larval fish were investigated in the River Raisin to discern whether
toxicological effects were evident (Fay et al_., 1985).

     Seven sampling stations were used in larval fish studies which were
distributed along the longitudinal axis of the River and extended into
western Lake Erie (Figure 7.13).  Longitudinal-oblique tows were taken using
a 0.75 m diameter, conical oceanographic plankton net with 0.57 mm mesh.  Tow
duration was six minutes behind an outboard motor-powered boat travelling at
4-5 knots.  Flow rates were measured using a center-mounted General Oceanic
Model MKII flowmeter.  Because of the diurnal activity of larval fish,
samples were collected at night, commencing at least 45 minutes after sunset.
Three replicate samples were taken from all stations.  Larvae used for
morphological analyses and pathological examinations were preserved in
Dietrick's fixative.  Samples for growth and mortality studies were preserved
in buffered formaldehyde.

     Larval collections were washed and sorted to remove any extraneous
biological elements or debris.  A Bausch and Lomb stereo dissecting
microscope with a magnification range of 6-100X was used to identify larvae
to the lowest taxonomic category possible.  Taxonomic references such as Auer
(1982) and Synder (1976) were used to identify species of larvae and
determine their developmental stage.

     Descriptive statistics calculated during the study included larval fish
density by site and season, average length by site and season, simple growth
rates, and instantaneous growth and mortality rates.  Simple growth rates
were calculated from the ratio of difference in length to difference in time,
dl/dt.  Instantaneous growth rate coefficients were calculated as follows:


                            G(t - tj
              L(t)  =  L(tQ)       °


where:

      L(t) = length at final time

     L(t0) = initial length

         G * growth rate

         t - final time

        t0 - initial time.


Mortality coefficients were calculated using the equation:


                            -z(t - tj
              N(t)  =  N(tQ)
                                     110

-------
Figure 7.13.   River Raisin Larval  Fish  Sampling
             Stations,  1983-1984.
                      Ill

-------
where:

      N(t) « number of larvae at final time

     N(t0) - number of larvae at initial time

         Z - mortality rate

         t - final time

        t0 * initial time.


Methods used for determining instantaneous growth and mortality rates are
found in Hackney and Webb (1978).

     The gizzard shad, Dorosoma cepedianum. was the dominant larval fish
(<70%) collected during 1983 and 1984 in the River Raisin (Table 7.10).
Other common taxa observed included Notropis antherinoides (emerald shiner),
Cvorinus caroio (common carp), Morone spp. and Aolodinotus arunniens
(freshwater drum).  Some variation was observed in terms of total density and
predominant species composition between the two field years (Table 7.11).
Generally, the twenty most abundant species were collected from all stations.

     Simple growth rates indicated greatest growth in the mid-portion of the
study transect (Stations 2-4) for gizzard shad (Table 7.12) and emerald
shiner.  Instantaneous growth rates indicated greatest growth at upstream
stations with decreasing growth proceeding lakeward (Figure 7.14) for both
species.  Both growth estimation methods indicated decreased growth at
downstream stations.  Similarly, cumulative probabilities of median fish
length showed inhibited growth proceeding from upstream to Lake Erie stations
(Figure 7.15).  Growth rates exhibited an inverse relationship with zinc,
copper, chromium in sediments (Table 7.13).  Zinc and copper toxicity
increased by at least 1 toxic unit as downstream stations were encountered
and larval fish toxicity increased by 0.5 toxic units or greater (Figure
7.16).

     Mortality rates were extremely variable between years, both
quantitatively and spatially (Table 7.14).  Greatest gizzard shad mortality
was observed in downstream stations in 1983 and in upstream stations in
1984.  In general, greater mortality was observed in 1984.  When comparing
instantaneous growth and mortality over the two years, Station 2 exhibited
the lowest growth and mortality in 1983 and the greatest growth and mortality
in 1984.

     Histopathological examination of fish larvae indicated the presence of
lesions in River Raisin populations.  Numerous types of lesions were observed
in gizzard shad (Table 7.15).  Epithelial necrosis was the dominant
lesion-type affecting numerous organs and tissues.  However, larval fish
examined fromt he Lake Erie control station 70 also exhibited comparably high
lesion occurrence.  Poor control site selection or population changes due to


                                     112

-------
£
UJ
   .05
   .04
t  .03
Ui
O
o

UJ
X


g  .01
o
                                    I
                                              I
                                    4        5

                                  STATION  NO.
                                                                70
        Figure 7.14.
                      Instantaneous  Growth Rates of River Raisin
                         Emerald Shiners, 1983.
                                   113

-------
rn





O
<
Z
(/>
GIZZARD
V*

_
co to
CM *- tr> o "
co ^ t£) —

^ V ^ 1>
rV "M *
,-co ^L "fc S
co ^* ^ ^*"~ ^^
^I v* ^ ^TtO Jc ™
co«~ ^^ "^ D
n *- v *>


0
O)

8g
_J
20 50
PROBAB
0
T"

o o o o o
h* 
                                                   o>
O
«£
                                                     o m
                                                     m <
                                                        co
                                                        O
                                                  -Q
                                                   O
                                                   s_
                                                   .
                                                  r^


                                                   OJ


                                                   3
   O
   CO
                              o
                              CM
 (LULU)  H10N3T  NVIQ3W
                    114

-------







E
O
X
e














DI~
-

n

	 D-
	 ,,
— D-
i i i
CO 04 i- O


i

— Q—
Q
_



in
i

^ £i
o
^^
X
in O
T-: H
X
1 •
"* i
CM ^
DC
* 3
CO

in




u> -
• 0

X
^^

i
i
-

^^^^^^J^^^l_

5 -Q
= i .^3.
if Q '
m O
0<
i i i
T- «« 0


I
1
— ru-
o- ,
-O- i
1
1

1
1
in
r

in
*>
tf\ *^~
10 H 1
• S* (T3
r- S -o
S o
O)
««.
CM g
•f»
»^4
in |
co §
C_J
o
in i §
• • -|~
i ; *..
! 3
^
i
in z
* o

^









•
c
to
Of
(1)
>
ex
+•>
u
X
0
(—
JZ
(/I
u.
r—
"3
>
U
                                     UJ
                                          00 C
                                   CO
(SlINO DIXOl) NOIiVyiN30NOO
                     115

-------
TABLE 7.10.  ABUNDANCE OF LARVAL FISH COLLECTED IN THE RIVER RAISIN,  1983


1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42


Gizzard shad
Emerald shiner
Carp
Morone sp.
Freshwater drum
Spottail shiner
Channel catfish
Yellow perch
Lepomis spp.
Unidentified Cyprinid
Al ewi f e
Walleye
White bass
Logperch
Brook silverside
Rock bass
Rainbow smelt
White sucker
Trout-perch
Bluntnose minnow
Pomoxis sp.
Tadpole madtom
White crappie
Sauger
Stonecat madtom
Johnny darter
Unidentified
Unidentified percid
Silverjaw minnow
Quillback carpsucker
Golden shiner
Green sunfish
White perch
Goldfish
Central stoneroller
Lake chubsucker
Creek chub
Unidentified catostomid
Yellow bullhead
Ictalurus sp.
Bluegill
Unidentified Fundulus
TOTAL
TOTAL * LARVAE
COLLECTED. 1983
11,440
919
814
701
512
345
245
215
114
99
63
67
49
44
49
29
25
16
16
14
12
9
8
7
6
4
5
2
3
2
2
2
2
1
1
1
1
1
1
1
1
1
15.849
% OF TOTAL
COLLECTED. 1983
72.1
5.8
5.1
4.4
3.3
2.2
1.5
1.4
0.7
0.6
0.4
0.4
0.3
0.28
0.27
0.18
0.17
0.101
0.100
0.088
0.080
0.057
0.050
0.040
0.040
0.020
0.020
0.020
0.019
0.013
0.013
0.010
0.010
0.006
0.006
0.006
0.006
0.006
0.006
0.006
0.006
0.005
99.8342
                                  116

-------
   TABLE 7.11.  RANKING OF SPECIES ABUNDANCE IN THE RIVER RAISIN, 1982-1984

SPECIES
Gizzard shad
Emerald shiner
Carp
Morone sp.
Freshwater drum
Spottail shiner
Channel catfish
Yellow perch
Lepomis spp.
Cyrpinid (unid)
Alewife
Wai 1 eye
White bass
Logperch
Brook silverside
Rock bass
Rainbow smelt
White sucker
Trout-perch
Bluntnose minnow
Pomoxis sp.
Tadpole madtom
Damaged larvae
Quill back carpsucker
Burbot
Largemouth bass
Lake whitefish
Northern hoasucker
JUDE et al.
1982
1
7
5
3
2
9
11
4
17
12
NF
15
NF
18
NF
NF
8
16
14
23
20
24
6
10
13
19
21
22
OSU
1983
1
2
3
4
5
6
7
8
9
10
n
12
13
14
15
16
17
18
19
20
21
22
NF
30
NF
NF
NF
NF
OSU
1984
1
8
3
4
6
9
5
12
7
28
22
13
3
15
NF
19
10
11
14
29
17
16
NF
NF
NF
NF
31
NF
NF = Not found.
                                    117

-------
TABLE 7.12.
SIMPLE GROWTH RATES OF GIZZARD SHAD IN THE
      RIVER RAISIN, 1983

SPECIES STATION YEAR
Gizzard 1 1983a
shad b
c
d
2 1983a
b
c
3 1983a
b
c
4 1983a
b
5 1983a
b
c
d
e
6 1983a
b
c
d
e
70 1983a
b
c
d
e
INITIAL
DAY
199
202
192
171
160
171
174
160
164
167
160
209
150
160
171
188
209
150
167
174
160
199
164
167
171
195
199
FINAL
DAY
220
244
237
181
209
111
230
195
209
216
223
251
234
227
230
241
216
111
234
230
181
213
216
227
213
220
234
INITIAL
SIZE
(mm)
40.5
31.2
14.7
4.0
3.5
3.5
4.3
3.5
3.9
3.8
3.0
12.3
6.7
6.8
6.0
10.1
13.9
8.0
8.9
9.7
4.0
11.7
7.8
7.5
5.2
11.4
6.8
FINAL
SIZE
(mm)
49.2
48.2
35.3
6.5
41.6
40.4
33.5
38.0
47.4
49.9
45.3
41.7
37.3
31.5
27.2
30.0
16.4
32.9
30.8
28.3
9.7
16.9
24.9
26.0
19.8
20.4
17.8
dl (mm)
dt (day)
0.41
0.40
0.46
0.25
0.78
0.66
0.52
0.98
0.97
0.94 -
0.67
0.70
0.36
0.37
0.36
0.38
0.36
0.32
0.33
0.33
0.27
0.37
0.33
0.31
0.35
0.36
0.32
                         118

-------
       TABLE 7.13.  METAL CONCENTRATIONS AND LARVAL GIZZARD SHAD GROWTH
                             IN RIVER RAISIN,  1983
STATION
MEAN WATER (ppbl
Zn    Cr     Cu
                                         SEDIMENT (Dpm)
Zn
Cr
Cu
GIZZARD SHAD
GROWTH RATE
COEFFICIENT
  (%/DAY)
1
3
4
5
70
1.6
2.7
2.4
1.7
1.0
.21
.20
.12
.90
.12
2.4
1.9
2.4
2.3
2.2
26
107
321
390
356
9
17
125
166
157
8
42
106
230
124
0.051
0.036
0.024
0.019
0.025
          TABLE 7.14.
        MORTALITY RATES OF MONROE HARBOR GIZZARD SHAD,
                  1983 AND 1984
YEAR
                   STATION
                        Z (%/DAY)
1983
1984
                      1
                      2
                      3
                      4
                      5

                      1
                      2
                      3
                      4
                             .018
                             .001
                           -.010
                             .043
                             .044

                             .073
                             .082
                             .049
                             .049
                                     119

-------
             TABLE 7.15.  HISTOPATHOLOGIC LESIONS IN MONROE HARBOR
                         LARVAL GIZZARD SHAD,  1983-1984

% AFFECFTED
TISSUE
Olfactory Organ
Lateral Line
Oropharynx
Esophagus (Ant.)
Gills
Excretory Kidney
Excretory Kidney
Intestine
Gill Parasites
LESIONS STATION 4
Epithelial Necrosis




Tubular Epithelial Necrosis
Hyaline Droplet Degeneration
Epilthelial Mecrosis

95
94
96
92
92
94
32
70
72
STATION 70
91
90
97
91
100
100
34
93
28
migratory behavior may have been factors producing this unexpected outcome.
However, gill parasites were considerably more abundant in River Raisin
gizzard shad compared to those from Lake Erie, possibly suggesting that
populations were distinct and migration between sites was minimal.  Lesions
observed during the study are typically associated with acute exposure to a
wide array of toxicants.

     Extreme variability was exhibited in various phases of the larval fish
study.  Variability in yearly average species composition and density
indicate that sampling times between years must be determined by factors such
as ice-out initiation, water temperature, water level, flow, and photoperiod
and not necessarily by date to obtain more reproducible results.
Undoubtedly, natural variability will be encountered but should be minimized
to the greatest extent possible.

     Optimal larvae size for identification and histopathological examination
appears to be between 20 and 50 mm in length.  In this size range the larvae
are sufficiently developed, in terms of morphological characteristics used in
taxonomy, as well as for use in tissue analyses.

     Results of simple and instantaneous growth rate calculations for gizzard
shad indicated somewhat different spatial patterns.  All estimates indicated
that growth was greater in the upper to mid-stream reaches of the River than
in the downstream and Lake Erie zone.  Generally, simple growth rate
calculations adequately reflect true growth assuming no population migration
and a representative sample yield.  Growth rates appeared to be negatively
associated with Zn, Cr, and Cu in the sediment but not in the water column;
metal toxicity in the water column increased by approximately 0.5 toxic units
proceeding downstream.  Similarly, a negative relationship between growth
rates and PCBs in water and sediment was observed.
                                     120

-------
     Histopathological studies revealed the presence of lesions in River
Raisin larval fish populations.  Epithelial necrosis was the most common
lesion detected, afflicting several  organs and tissue types.  The lesions
observed in gizzard shad were similar to those known to be caused by toxic
exposure.  Larval fish from the control site exhibited comparable or greater
occurrence of lesions than river specimens.  The choice of this control
station might have been poor, since  all fish in the region were afflicted
with a similar incidence of lesions.  Migratory activity may have resulted in
sampling the same population at all  sites.  A control site, considerably more
removed from the test site(s), would reduce the possibility of migratory
activity.

7.2.11  Contaminant Body Burdens of  Resident Fish Populations

     Typically, fish populations are the most important ecosystem component
in terms of commercial and recreational use.  Issues concerning fish
population composition, abundance, and health generally draw the greatest
public attention, particularly in regard to human health implications.
Toxicants which have accumulation potential are routinely sought in fish
tissue, to determine whether a human health risk exists through consumption.
The approach employed in the River Raisin was to analyze fish tissue for body
burden concentrations of PCBs to assess the spatial distribution of
contamination, the species involved, the environmental exposure factors, and
the potential for human health risk.

     During 1983 and 1984, seven species of River Raisin fish were analyzed
for body burden concentrations of PCB.  Samples were collected from the
mid-portion of the River and lakeward.  Sample preparation and extraction
procedures for fish tissue followed  the same methods as those used for other
biological samples (Rathbun and Smith, 1985); for water and sediment methods,
see Sections 7.1 and 7.3 of this report.  Both fillet and whole fish were
processed for analysis.

     Greatest total PCB concentrations were observed in carp; values ranged
from 0.21 - 100.0 mg/kg (Table 7.16).  Twenty-five of 31 carp samples
exceeded the USFDA action level of 2 mg/kg.  Generally, fillets contained
lower PCB concentrations; of the six fish which did not exceed FDA actions,
four were fillet analyses.  Young-of-the-year emerald shiners exhibited
relatively high concentrations, ranging from 0.48 to 3.7 mg/kg; 5 of 7
emerald shiners exceeded the FDA action level.  One of eight small mouth bass
(3.4 mg/kg), three of eleven gizzard shad  (maximum of 2.9 mg/kg) and a single
mirror carp (26.0 mg/kg) also exceeded the action level.  Single samples of
rock bass and largemouth bass exhibited relatively low PCB concentrations.
Based on these data, the Michigan Department of Natural Resources re-issued a
fish consumption advisory for the lower River Raisin.

     These results indicated severe PCB contamination of several River Raisin
fish species.  They also demonstrated the differential bioaccumulation
potential of selected species.  Concentrations in particular species, such as
commom carp, are probably due to direct sediment exposure and a benthic-based
food chain.  Conversely, species with decidedly planktivorous-based food

                                     121

-------
00
c*>
CO
oo
en
CO
CO
ae.
LU
CD


£
r«.


LU

CO
          00
          £•5;
          o —'


             o
             t—t
             o.
             >—I
             —I

             a*
          O.X




          LU O
          CO
          o
          C3 O)


          LU
             X

             CO
             O.



             CO




             LU


             <
             5
             CO
             LU
             a.
             CO
                     to
                           O i—
                                esjr—
                                              csiesi    cncooromm
                                                       CM CO O      CvJ
                            -i— i — r— vo «o co oo 00
                                                                                i— i — o
                                                            i— Ift^-VO    CO i— lf> VO
                      ooooooooooo
                      O.O.O.O.O.O.O.O.O.O.O.
                      EEEESSEEESE
                      ooooooooooo
                     oouuuuuuuoo
                                                                                         CM 10 LO
                     zzzzzzzzz
^•^•OOi—r—OOOOp-r-<\JOO
r—r—r— OCMr— evICAJCMr—i—
r«.r^oocf>r«.oovototor>>rv
ooooooooooo
cocococococo^-^*"***
00 00 00 00 00 00 00 00 00 00 00
cMcv/csjcvjcjcsjr^.f-.rv.r^.r^r^r^.r^r^
CMCVICslCSJCMCMCSJCMCMCMCMCSICMCVJCJ
Tj-^-^l-«a-^-«9-U3l£)lOlOVOVOV£)UJU3
ooooooooooooooo
cocococococo^-^-'a-**-«r«a-«S'^'«tf-
ooQOoooocooocococooooooooooooo
                       1 m i
                        •o
                         fO

                        CO

                        •o


                      >  N
                      i.  IM
                      (O -r-
                     —I U)
                                  1 tn <
                                                        i.

                                                       o
                                                                                                       (O 0)
                                                                                                       fcO r—
                                                                                                          o.
                                                                                                        o
                                                                                                       JQ
                                                                                                       Or-
                                                                                                        II   H
                                                   122

-------
         03
         O — »
         CL. en
            en
            S
            -


            Q
            i— i
            o.
         LU
         Q.
o
o
oo
en
co
co
en
co
co

i
on
            S-
            o
         CO
         tD
CO
o
a.
CO
 vo CM ^- 1— ««• r- o
   • —            r—   CM                 CM
CM in co en ^- m i— o in co r^ CM CM oo cn en
oocMr- r-.cniocnr^.oinr^r-.coco**'^1
                            co en .— co o o in
                                      r— CM
                                                 to
                                                      i— i— o
                                                            a>
                                                            4->
                                                            0
                                                            Q.
VO
CM
CM
CM
                                                                                      CM
ininoom^-cnocno^oi0^voocooooinin*t-ininoo
incoo^-inr— i— incncocnmtocoor^vocnvocMCMCOr^^-or^.
«*-inm^-^-««'<9'^-cococMeMCMincococo^-^'^-«*cococo^-^>
  i co co co c^j co ^o ^^ co f^" ^"^ r^^ ^^ ^o ^o ^^ f^n> ^o ^^ C5 co ^^ c^j ^^ ^o ^^



  •mcniovor—t^vocMcocncocMincooococMcncnr—r^cocnocn
  •omi—cMOOcnocncocMCMCMr—^roocoinincnr—oof^iocooo
  • CM CM i—i—      r—              CM        r—i—   i—            i—
iv.r^r«.tv.i»»minininr-.r».r»»r«.rv.r>.i~.
CMCMCMCMCMr— r~r— r— CMCMCMCMCMCMCM
VOVOiOVOiOCOCOCOCOlOlO^OVOlOVOVO
oooooooooooooooo
^•^•^•^•<»COCOCOCO^'*'«-^'^'^'«9'
cooooooooocooooooocOoooocooococo
                    ^•^•^•^•^-cocococo^-^r^-^-^-^-*
 r».
 CM
 1O
 o
 ^>
 00
                     c
                     0
                     a.
                     i-
                     re
                    o
                                                                   to
                                                                   o

                                                                   J_
                                                                   o
                                                                   i-
                                                                                               00
                                                                                               o>
         (O O)

           "o.


         O CO
         .a
           -M
         0) 0)
         ^— r—
         O i—



         II  II
                                                  123

-------







— *
Q
h-

O
O
00
en
CO
00
en

GO
(—1
LL.
z
HH
OO
i
CtL
>
oc
2
CO
00
O
Q.

•
VO
^
LU
CO
r-








































































CO
£j?
O*-*
a
«— «
a.
«

r-
L*
LU O
_J
00 *""
3-
1— — «.
eg g
LU •— ••
i


1—
3 "en
HM *~ -
LU

X
|,|
hU
Z LU
1—1 —i
a.
o 5


LU
r-


^)
Un»
f~*
H**
oo






00
LU
1—4
O
LU
a.
oo





o oo Is* o ^ vo CM cn en
co

^ LO CM O CM r— VO ^" VO O
O CM CO CM r— CM ««• CM CM CM




„ ___„«««,-, «


•* Oi— O LO O CM LO CM CO
r**^ ^^ ^^ f^ LO co ^^ CM CM oo
f— CMCMCMCMCMCMCMCM >—

^% ^^ CM vo cn LO oo vo CM en
^^ oo r^ oo ^™ ^^ LO co CM ^^
co aors.oocoLocncM^- cn
r~~ co »•— co ^^ ^^ cn f**» LO cn
r— i — CM r— CM i— r— r— i—


X u-3cu.^u-u.s:^ u_





CM CMCMCMCMCMCMCMCM CM
tp if> tp ip tfi tp tp tp t^ tf>
O OOOOOOOO O
00 0000000000000000 00

*r •* * «*• * *r * * ^- *•



(/I 00
r« (3
CO CO

.c x:
^j +^
CO "^ 3
CO 0 0
fO ZZ Z
CO
i— 0)
^ i— cn
o s^
z



^v



^^^
z

cn



r—
00
o
oo

*





&.
nj
a>
•>-
1
4-
O

OI
c

0
•>-

^


CD
VO
CO
0)
o
o.
E
o
X

^f
"\
z



\
z


<
z~

0



o
cn
0
00

f-




!„
0)
c

f
00

•a

fO
i.
0)
E
LU

cn


C\J
o
co
0)
o
o.



*f
~^~
z



-\



<
z"

o



o
cn
o
00

1—



















00
CM

00
o
co
0)
o
a.
§
i

^£
~^




^N^



<
z"

0



o
cn
0
00

««r



















cn
CM

CO
r—
CO
0)
•r—
O
a.
i
i

^^
•^
z



>vv
z


<:
"^x.

0



o
cn
0
co

*•



















„
CM

f»^
i —
CM
9)
•9—
O
O.
o
X

^^
'Vs^
z



\.
z


<
z"

o



o
cn
0
00

LO



















00
CO CO

00 ^^ ^^
LO 00 CO
CM r— CM
0) Q) CO
000
a. a. a.
E E E
O O O
3: z z

^c ^c ^^
>S*K. ^^S* "****•,,
z z z



\\\
z z z


«<
z'z'z'

0 O O



Or- 0
cn oo cn
000
oo oo oo

LOLOr-





















































a>
"o.
E co
re a>
CO r—
o.
>> E
-o re
O CO

^_i
CD 0)
r— P—
O r—
JC. -P-


II N

2 LL.
124

-------
chains, such as largemouth and small mouth bass,  exhibited considerably lower
concentrations.  Most fish collections for these analyses occurred at Station
4 where elevated PCB concentrations in the sediment and water column were
found.  From limited data on gizzard shad, relationships between fish body
burdens and PCB concentrations in water and sediment were exhibited (Table
7.17).  Collections and analysis of fish species with similar body weights
along contaminant gradients should be employed to determine exposure-effects
relationships.

7.2.12  Evaluation of Biological Studies

     lexicological effects were observed in a number of exposure studies
employed in the River Raisin.  Biological toxicity testing is a critical step
in determining the health of ecosystems.  Relatively high concentrations of
organic contaminants and heavy metals cannot be used as the sole criteria for
identifying environmental problems.  If biological effects cannot be
demonstrated, due either to the unavailability of toxicants to biota or to
concentrations which do not induce a response, then it is more difficult to
argue that a problem exists, no matter what toxicant concentrations are.
Conversely, if considerable pollutant concentrations are observed in fish or
drinking water, posing the risk of direct human exposure, a potential health
risk exists and mitigative action should be initiated.


        TABLE 7.17.   PCBs  IN WATER,  SEDIMENT, AND LARVAL GIZZARD SHAD,
                               RIVER  RAISIN,  1983



STATION
1
3
4
5
70
MEAN TOTAL PCB-
GIZZARD SHAD
(ppm)


1.28
0.77
0.4
MEAN TOTAL
PCB-WATER
(DDb)
0.008
0.054
0.229
0.015
0.032
TOTAL PCB-
SEDIMENT
(ppm)
0.008
0.05
2.7
1.9
0.14
     Bioassays employed should include both acute and chronic endpoints and
should address ecosystem structure and function.  Both primary and secondary
ecosystem functions are effective assay endpoints.  Ecosystem function tests
assess the processes of metabolism, physiology, feeding, reproduction,
growth, and mortality.  Factors such as survival, bioaccumulation,
morphological considerations, and body burden, which are affected by the
above processes, should be addressed also.  Additional assays which address
other factors can be included to complete the battery of tests; these would
assess carcinogenicity, mutagenicity, and behavior modification.
                                     125

-------
     In a comprehensive approach to toxiclty testing, a suite of bioassays
must be employed which span several trophic levels.  Representative species
of particular trophic levels may have differing sensitivities to certain
exposure concentrations and toxicants.  Results from the River Raisin
demonstrated this point very well.  Phytoplankton carbon uptake in effluent
waters near the Wastewater Treatment Plant exhibited both inhibition and
stimulation and did not show a clear trend at this site.  Conversely, both
zooplankton grazing and reproduction were inhibited at this location.
Moreover, zooplankton reproduction was suppressed to a greater extent than
zooplankton grazing.  This result indicates that the effluent had an effect
directly on reproduction or reproductive physiology and that decreased
reproduction was not caused by feeding inhibition.

     Bioassay sensitivity, in terms of the function and test organism
considered, must be evaluated.  An exact statistical protocol has not been
established for this procedure.  Preliminary screening can be accomplished by
correlation analysis, but this would necessitate complete coordination of all
assays on the same samples.  Two apparently sensitive assays used in the
River Raisin study were the zooplankton grazing and Ceriodaohnia 7-day
assays.  These assays generally produced consistent results in time although
spatial differences were observed.  Additionally, within any assay,
reproducibility of results were consistent.  These factors in themselves, may
explain why informative results were obtained.  Specific organisms used in
bioassays must be sensitive to environmental conditions but not so sensitive
as to produce sharp threshold responses yielding only live or dead organisms.
In particular, species must be selected to produce dose-response functions
where concentration-effect resolution is obtained, so that the EC10 and EC50
responses can be calculated.  If sharp dose-response curves are obtained,
additional dilution ratios may be required to negate the problem.  Similarly,
if inhibition responses do not increase monotonically with increasing
exposure concentrations, at least to some degree, the resultant dose-response
functions are very difficult to interpret.

     Clams, fathead minnows, and channel catfish all appear to be suitable
test organisms for PCB bioaccumulation studies.  Bioaccumulation of
organochlorine compounds in native species is usually known from surveillance
activities.  Placement of caged specimens, however, indicate site-specific
bioaccumulation potential and can be used for locating sites with the
greatest potential impact.  Toxic effects on resident, larval fish growth
were observed and were demonstrated in several ways.  These studies produced
a fairly consistent spatial pattern and indicated where toxic effects were
most likely occurring.

     Bioassays and biological studies that were conducted, other than those
mentioned above, yielded various degrees of information.  However, the lack
of reproducibility and/or the extreme variability that occurred diminished
the value of findings in some cases.  Variability must be kept to a minimum,
and this would be achieved best through sound experimental design or by the
control of as many variables as possible.
                                     126

-------
     In many cases, variable control can be accomplished through a
combination of field and laboratory methods such that confounding factors are
eliminated and particular effects can be determined reproducibly.  For
example, phytoplankton assays used mixed algal assemblages in this study,
which undoubtedly represented differing species composition,  biomass,
biovolume, and physiological states.  These factors contribute to variability
but could be negated by using monoclonal stock cultures of species
representing different physiological groups (i.e., greens, blue-greens, and
diatoms), held in steady state, where productivity rates, chlorophyll a
concentrations, and biovolume are known.  The confounding influence of
variable nutrient concentrations in the water column can be diminished by
saturating culture media with phosphorus, nitrogen, and silica.  Using a more
controlled assay format, variability will be reduced and reproducibility and
confidence will be enhanced.  Admittedly, controversy exists concerning this
topic but all Areas of Concern usually have eutrophic waters, and generally
phytoplankton are not nutrient-starved.

     It appears that considerable field and laboratory methodology and
procedural development is necessary for many of these studies.  These include
the choice of collection times, gradient sampling, consistent station
sampling, a dilutional series with finer resolution, proper control station  .
selection, and the development and use of statistical techniques for
interpretation.  A critical question for all assays and particularly for
chronic assays is that of sample handling and storage time of test water.
Biochemical changes may occur in sample water during storage as resupply
media.  Specific experimentation geared toward determining storage time
effects and the potential variability introduced to assays should be
initiated.  Most of the tests employed in the River Raisin study are costly,
and are time- and effort-intensive.  The development of rapid, cost-efficient
bioassay methods is much needed.

     When a field plan and  sample design are being developed for toxicity
studies, sampling in a food chain context should be considered.  Body burden
determinations would yield  bioconcentration factors for specific food chains
of interest.  The same determinations would also be used  for fish population
health assessment, consumption advisories, and exposure relationships with
other environmental variables; this could be applied to most trophic levels.
Similarly, the trophic levels used  in food chain studies  should be employed
for bioassay tests.  It would be extremely informative if the toxicity to
indigenous species were known.

     When investigating complex mixtures of water  or sediment, it is very
difficult to determine cause-effect relationships  and discern specific causal
agents.  Statistical methods which may be used for preliminary screening are
correlation and multiple regression analysis, analysis of covariance,
analysis of variance, canonical correlation, and factor analysis.  Results of
these studies also appear to be amenable to analysis using multivariate
techniques such as principal component analysis or correspondence analysis.
These techniques will identify potential exposure-effect  associations.   The
utilization of toxic units, EC50 responses, and exposure  probabilities also
appears to be very informative in determining cause-effect relationships.

                                     127

-------
These methods standardize diverse data sets and indicate an exposure-effects
range which may be expected, aiding in interpretation.

     The Monroe Harbor project provides a foundation for future biological
toxicity studies to determine exposure and effect.  Certainly, numerous
features of this study will be built on for improvements in the future.
Prominent effects observed in this study, as determined from correlation or
association, include the inhibition of phytoplankton production by chromium,
zooplankton feeding inhibition by copper and zinc after correction for
hardness, and zooplankton reproduction inhibition by zinc after corrected for
alkalinity.  The water and sediment of the River Raisin contain PCB
concentrations that are bioavailable for uptake by clams and fish.  Gizzard
shad and emerald shiner growth rates appeared to be negatively associated
with zinc, copper, and chromium found in sediments.  Similarly, PCB
concentrations in gizzard shad exhibited a positive relationship to PCB
concentrations in sediment and water.  Numerous tumors and associated lesions
were observed in native fish and organochlorine concentrations were elevated
to such an extent that the fish should probably not be consumed by humans.


7.3  SEDIMENT CHEMISTRY

     In many of the Great Lakes Areas of Concern it is felt that in-place
pollutants (pollutants previously deposited from the water column to bottom
sediments) are contributing to the degraded conditions that exist.  Before
agencies can establish rational management programs (Remedial Actions Plans)
for Areas of Concern, it is necessary to identify the relative importance of
contaminated sediments for each system of interest.  Because not all systems
behave identically with respect to in-place contamination, a general
methodology must be developed that address questions relative to exposure and
biological effects of in-place pollutants within the context of an
integrated, system-wide exposure-effects assessment.  Presented below is an
approach and methodology, as applied to the Monroe Harbor problem, for
assessing the significance of bottom sediments relative to the transport,
fate, and effects of contaminants in an Area of Concern.

7.3.1  Physical Characterization

7.3.1.1  Approach--

     Physical characterization of bottom sediments is essential in order to
assess the potential for transport of contaminants across the sediment-water
interface and within the sediment profile.  An effort must be made to survey
the bottom sediments of the study system for those parameters which are most
likely to affect solid and dissolved phase contaminant transport.  The
physical characteristics selected for the survey of bottom sediments in
Monroe Harbor were sediment moisture content, loss on ignition (volatile
solids content), sediment density (dried and after ignition), porosity, and
particle size distribution.
                                    128

-------
     The Michigan Department of Natural  Resources (MDNR)  established 22
transects in the study area and collected sediment samples for qualitative
sediment characterization at three points along each  transect indicated as
south (S),  center (C), and north (N).   Transect 42, located in the turning
basin, was sampled at four positions,  south (S),  center south (CS),  center
north (CN), and north (N) (Figure 7.17).   Table 7.18  presents the results of
the MDNR qualitative sediment survey.   The results suggested a transition in
sediment composition from sand and gravel above the turning basin to silt and
detritus from the turning basin downriver to the river mouth.  The results of
this type of qualitative survey are used, in addition to  any other historical
data available, to plan the location of sampling transects in the study area.

     In order to obtain an accurate representation of the spatial
distribution of sediment physical characteristics in  the  study system,
surface sediment samples (0-2 cm) were collected at as many as four stations
along several transects of the study area.  In addition,  vertical profiles
(0-10 cm) were collected at three stations in order to assess contaminant
transport within the sediments.  Two surveys were conducted, one in July of
1983 and another in May of 1984.  A complete list of  the  sediment samples
collected is presented along with their physical  and  chemical characteristics
in Table 7.19, and the locations of the stations can  be found on the study
area map in Figure 7.17.

7.3.1.2  Methods--

     7.3.1.2.1  Field—During the 1983 survey sediment samples were collected
by using either a homemade, hand-driven corer or a gravity coring device
(Wildco Corp., Saginaw, MI), depending on water depth.  Neither of these
devices was satisfactory for sampling in areas of hard sand or gravel bottom,
and thus no samples could be collected throughout most of the upstream
section of the study area.  During the 1984 survey, the addition of a Ponar
dredge (Wildco Corp., Saginaw, MI) to the sampling equipment allowed samples
to be collected in areas of hard sand bottom which could  not be sampled
during the 1983 survey.  This enabled the sampling team to collect samples in
the upper section of the study area, upstream from most of the current point
source inputs to the system.  It was expected that these samples would thus
serve as indicators of "background" levels of contamination from sources
outside the study area.

     Immediately upon collection, core samples were cut into two centimeter
sections and placed in acid-washed polyethylene containers or solvent-washed
glass jars, for metal and organic analyses, respectively.  Samples were
transported in coolers and, upon arrival at the laboratory, were stored in
darkness at 4*C to minimize biological and chemical alteration.

     7.3.1.2.2  Laboratory--The following is a brief  description of the
methods used for physical characterization of the sediment material.

     1.  Moisture Content - After thorough homogenization, an aliquot of each
sample was placed into a tared, aluminum dish and weighed.  These aliquots
were then dried for three hours at 105'C, cooled in a dessicator, and


                                     129

-------
                                                    I/I
                                                    c
                                                    ro
                                                    S_
                                                    to

                                                    to
                                                    c
                                                    o
                                                    (O
                                                    +J
                                                    I/)
                                                   •o
                                                    0)
                                                    o
                                                    JD
                                                    S_
                                                    OJ
                                                    o
                                                    c
                                                    o
                                                    ts
                                                    (J
                                                    o
                                                    o>

                                                    3
130

-------
               TABLE 7.18.  QUALITATIVE SEDIMENT SURVEY - MDNR
SAMPLE
 SITE	DESCRIPTION
  2N     Sandy silt and some gravel.   Brownish.
  2M     Sandy silt.  Brown.
  2S     Sandy silt and fine gravel.   Brown.
  3S     Silt and sand.  Some detritus.  Brown.
  3CS    Sand and fine gravel.  Brown.
  3C     Sand and fine gravel.  Light brown.
  3N     Silty sand.  Gray color.  Some detritus.  Very slightly anaerobic.
  4S     Sandy silt.  Brown.  Some gravel.
  4C     Coarse sand, some gravel.  Brown.
  4N     Silt.  Light brown.  No unusual odor.
  5S     Light brown silt.
  5CS    Silt, some sand.  Brown.  No unusual odor.
  5C     Fine gravel and sand.  Some silt.  Light brown.
  5N     Hard rocky bottom.
  6S     Hard rocky bottom.
  6C     Coarse sand and fine gravel.  No odor.   Brown.
  6N     Coarse sand and fine gravel.  Brown.
  7C     Hard rocky bottom.
  7S     Gray silt.  No odor.  Coarse detrital material.
  7N     Fine gravel/sand.  Some silt.  Light brown.
  8N     Sandy silt.  Gray-brown in color.
  8C     Hard rocky bottom.
  8S     Silt, some sand and gravel.   Brown color.
  9N     Sandy silt.  Gray-brown.
  9C     Fine gravel and sand.  Gray.  Slight sewage odor.
  9S     Silty, sand and gravel.  Black.  Slight oily odor.
 10S     Hard rocky bottom.
 IOC     Fine gravel, sand and silt.   Gray.  Slight sewage odor.
 ION     Silt.  Black in color.  Slight anaerobic odor.
 IIS     Hard rocky bottom.
 11C     Hard rocky bottom.
 UN     Silty sand.  Gray.
 12S     Clay and silt.  Gray.
 12C     Fine gravel and sand.  Some silt.  Gray.
 12N     Silt, some gravel.  Gray color.
 13N     Silt, some clay.  Gray color.  No  unusual odor.
 13C     Fine gravel, sand, silt.  Gray color.
 13S     Silty sand.  No odor evident.  Dark gray.
 42CN    Silt.  Brown-gray.  Some detritus.
                                    131

-------
          TABLE 7.18.  QUALITATIVE SEDIMENT SURVEY  - MDNR  (CONT'D.)


SAMPLE
 SITE	DESCRIPTION	

 42C     SIHy sand.   Grayish.
 42CS    Silt.  Gray-brown color.  Some detritus.
 42S     Silty sand.   Slight oily odor.  Dark gray.   Some detritus.
 43S     Silt.  Dark gray.  Slight oily odor.
 43C     Silty.  Gray.  No unusual odor.  Some detritus.
 43N     Rocky.
 44S     Dark gray silt.   Slight anaerobic, oily.
 44C     Light brown silt.  Faint oil  odor.  Some  detritus.
 44N     Oily odor, hard packed silt.   Some sand/gravel.
 45S     Silty sand.   Dark gray.  Slight anaerobic.
 45C     Silt.  Brown-gray.  A little  detritus.
 45N     Sandy silt.   Slight oil odor.  Gray.
 46N     Black, oily/anaerobic odor.   Silt.  Some  detritus.
 46C     Silt.  Brown-gray.  No unusual odor.
 46S     Black silt.   Oily odor and appearance.
 47S     Silty sand.   Some detritus.   Brown gray.
 47C     Silt.  Brown gray.  No unusual odor.
 47N     Gray.  Oily silt.  Some detritus.  Oily odor.
 48S     Black silt,  oily.  Some detritus.  Oil  odor.
 48C     Brown-gray silt.  No odor.  Slight detritus.
 48N     Sandy silt.   Gray.  Slight oil odor.   Fine  detritus.
 49S     Silty sand.   Gray.  Slight oil odor.   Some  gravel.
 49C     Silty.  Brown-gray.  Some detritus.
 49N     Silty sand.   Gray-brown.  Some detritus.
 BOS     Silt.  Gray-brown.  No unusual odor.   Some  detritus.
 50C     Silt.  Brown-gray.  Some detritus.
 BON     Silt.  Brown-gray.  No unusual odor.   A little detritus.
                                    132

-------
reweighed (Black, 1965 and APHA,  1980).   Analyses performed In duplicate
provided precision estimates for the analysis:   Expected Range of Duplicates
(ERD) - ±0.037(X) where X - mean of duplicate determinations.

     2.  Loss on Ignition - Dried samples were powdered and re-homogenized,
and an aliquot of each was weighed into a pre-ignited and tared porcelain
crucible.  The crucibles were then baked at 550*C for one hour, cooled in a
dessicator and weighed (APHA, 1980).  Precision:  ERD = ±0.084(X).

     3.  Density - Dried samples were powdered and re-homogenized, and an
aliquot of each was weighed into a previously-calibrated pycnometer bottle.
The bottle was then filled with distilled water, and weighed (Kezdi, 1980).
The temperature at which the measurement was taken was recorded to the
nearest 0.05*C.  Density was then calculated using the equation:


              Pst  =  [Ms/(Mpw + Ms) - Mpt]  x  Pw


where:

     MS  = mass of sediment (g)

     Mpw - mass of pycnometer bottle and water (g)

     Mpt = mass of pycnometer bottle and suspension (g)

     Pw  - density of water at specified temperature (g/cnr*)

     Pst = density of sediment (g/cm^)

The density of the mineral fraction of the sediment samples was also
determined; in this case, however, sample residue remaining after ignition at
550*C, rather than after oven-drying at 105*C, was used (Kezdi, 1980).
Precision:  ERD = ±0.034(X) for dry density and ±0.042(X) for  ignited
density.

     4.  Porosity - Porosity was not measured directly on the  sediment
samples, but was estimated using a relationship developed from Black  (1965)
which  incorporates other measured parameters (moisture content and dried
sediment density).  Porosity was calculated from the equation:
                    7W+ (Pw -


where:

     e  = porosity expressed as a fraction of total sample volume

     W  - moisture content  (%)/100


                                     133

-------
     Pw » density of water (g/cnr)

     Pp « density of dried sediment (g/cm3)

     5.  Particle Size Distribution - A micro-sieve set (Fisher Scientific)
with eight interchangeable mesh inserts was used to determine the particle
size distribution of the composite sediment used in adsorption/desorption
experiments.  Inserts had sieve openings of 180, 125, 90, 63, 20, 10, 5, and
1 micron(s).  The microsieve set was modified for wet sieving and a standard
wet sieve procedure, as outlined in Black (1965), was followed.

7.3.1.3  Results and Evaluation--

     Qualitatively, bottom sediment samples collected during 1983 and 1984
indicated that an abrupt change in depositional environment occurs within the
study area at the Monroe Harbor ship turning basin (Transect 13, Figure
7.17).  Bottom sediments upstream from the basin were found to consist mainly
of quartz, sand and gravel with abundant pebbles and shell fragments, while
Monroe Harbor sediments were composed predominantly of silt and fine sand
with some clay and organic detritus.  A qualitative sediment survey performed
in the study area during 1983 by the MDNR noted a similar depositional
pattern.  The pronounced change in stream morphometry and subsequent decrease
in stream velocity and turbulence that occurs between the upstream section of
the river and the Monroe Harbor may account for the deposition of more
fine-grained material in the harbor area.

     Comparison of samples between 1983 and 1984 surveys indicated somewhat
different appearances.  An oily sheen and odor noted in many of the 1983
samples was predominantly absent in the 1984 samples, and a fibrous material
thought to be paper pulp waste material was observed only in the 1984
sediments.

     The results of the quantitative physical characterization of River
Raisin bottom sediments are presented in Table 7.19, along with summary
statistics for both surveys and the composite.  Moisture content was
significantly greater (p < 0.01) in the samples collected during 1983 than in
1984.  This parameter may vary because of its dependence on several factors
including particle size and composition, degree of sediment compaction, and
hydrologic conditions.  The difference between years was significant even
ignoring the sediments collected at Transect 2, which were assumed to
represent "background" conditions.  The reasons for the higher solids in the
1984 samples are not known but may be due to different flow conditions
preceding each survey, causing variation in amounts of less-compacted
sediments transported out of the system.  Moisture content was also greater
in surfidal sediment (0-2 cm) compared to deeper layers (2-12 cm),
indicating active compaction with sediment depth.

     Loss on ignition, or volatile solids content, often is used as an
estimate of sediment organic carbon content.  Occluded water and some
volatile inorganic constituents, however, are also detected during this
procedure and can confound the results.  The composition of the total loss on

                                     134

-------
TABLE 7.19.  PHYSICAL  AND CHEMICAL CHARACTERISTICS OF MONROE HARBOR  SEDIMENT SAMPLES


Staton
Location
1983
6
11
9S
12S
42S
42SC
42NC
42NC
42NC
42NC
42NC
42H
47S
47C
47C
47C
47C
47C
47N
SOS
SOS
SOS
SOS
SOS
50C
SON
1984
2S
2C
2N
13S
13C
13N
42S
42N
47S
47C
47N
47S
47S
47S
SOS
50C
SON
11
43S
43N
Sample Size
Sample Size
Mean, 1983
Mean, 1984
Depth in
SediMnt
Core (cm)

0-3
0-3
0-2
0-2
0-2
0-2
0-2
2-4
4-6
6-8
8-10
0-2
0-2
0-2
2-4
4-6
6-8
8-10
0-2
0-2
2-4
4-6
6-8
8-10
0-2
0-2

0-3
0-3
0-3
0-2
0-2
0-2
0-2
0-2
0-3
0-2
0-2
3-6
6-9
9-12
0-2
0-2
0-2
0-3
0-3
0-3
, 1983
, 1984


Moisture
Content
(X)

71.5
77.3
69.8
37.8
76.6
68.8
75.8
67.8
63.7
60.6
54.6
79.2
63.3
78.1
63.5
61.4
58.2
57.0
79.4
72.6
45.3
47.8
50.6
56.0
73.5
65.0

18.2
35.7
25.4
46.3
29.0
53.0
51.7
66.2
42.9
71.6
66.2
36.6
51.9
48.4
62.6
61.8
46.5
55.2
55.5
62.7
26.0
20.0
64.4
49.4
Loss on
Ignition
(X)

10.2
9.7
11.7
4.3
11.4
10.7
11.6
11.8
11.7
11.0
11.4
13.4
8.3
10.5
10.0
10.6
10.3
9.7
10.7
10.1
4.9
5.8
8.0
9.3
9.7
8.7

1.5
1.4
1.8
7.3
5.5
6.3
12.7
10.0
5.0
9.1
9.7
6.1
11.2
11.3
7.0
7.0
5.8
6.7
9.2
9.3
26.0
20.0
9.8
7.2
Dry
Density
(a/cm3)

2.26
2.01
2.14
2.33
2.15
2.08
2.06
2.24
2.11
2.17
2.17
2.15
2.11
2.28
2.38
2.39
2.26
2.12
2.03
2.03
2.24
2.41
2.38
2.19
2.11
2.21

2.57
2.66
2.65
2.58
2.36
2.34
2.32
2.25
2.42
2.29
2.23
2.25
2.17
2.29
2.31
2.25
2.31
2.25
2.21
2.31
26.0
20.0
2.2
2.4
Ignited
Density
(Q/cm3)
*
2.52
2.41
2.59
2.56
2.61
2.62
2.41
2.61
2.53
2.71
2.43
2.62
2.52
2.56
2.62
2.57
2.62
2.69
2.44
2.46
2.65
2.36
2.66
2.98
2.51
2.55

2.65
2.63
2.64
2.68
2.64
2.63
2.50
2.43
2.65
2.51
2.65
2.63
2.62
2.67
2.73
2.57
2.63
2.66
2.52
2.59
26.0
20.0
2.6
2.6

Porosity
(Fraction)

0.850
0.873
0.832
0.586
0.875
0.821
0.865
0.825
0.788
0.770
0.723
0.891
0.784
0.891
0.805
0.792
0.759
0.737
0.887
0.843
0.650
0.688
0.709
0.736
0.854
0.804

0.363
0.596
0.474
0.690
0.491
0.725
0.713
0.815
0.645
0.852
0.814
0.565
0.701
0.682
0.795
0.785
0.668
0.735
0.734
0.795
26.0
20.0
0.8
0.7
Total
Cu
(mg/ka)

85
124
76
42
99
71
106
109
67
52
51
107
131
176
77
112
117
97
231
113
49
45
86
123
85
90

13
8
10
31
19
35
40
57
91'
76
82
148
845
857
51
56
59
69
60
58
26.0
20.0
97.0
133.3
Total
Cr
(ma/kg)

281
356
274
108
304
269
321
316
218
194
146
352
210
449
214
261
256
236
390
271
135
133
253
324
219
201

23
26
26
98
59
105
114
164
139
183
167
156
853
915
143
149
129
208
162
150
26.0
20.0
257.3
198.3
                                         135

-------
TABLE 7.19.   PHYSICAL AND  CHEMICAL CHARACTERISTICS OF MONROE HARBOR SEDIMENT SAMPLES (CONT'D.)

Depth in
Staton Sediment
Location Core (cm)
1983
6
11
9S
12S
42S
42SC
42NC
42NC
42NC
42NC
42NC
42N
47S
47C
47C
47C
47C
47C
47N
SOS
SOS
SOS
SOS
SOS
50C
SON
1984
2S
2C
2N
13S
13C
13N
42S
42N
47S
47C
47N
47S
47S
47S
SOS
50C
SON
11
43S
43N
Sample Size,
Sample Size,
Mean, 1983
Mean, 1984
0-3
0-3
0-2
0-2
0-2
0-2
0-2
2-4
4-6
6-8
8-10
0-2
0-2
0-2
2-4
4-6
6-8
8-10
0-2
0-2
2-4
4-6
6-8
8-10
0-2
0-2

0-3
0-3
0-3
0-2
0-2
0-2
0-2
0-2
0-3
0-2
0-2
3-6
6-9
9-12
0-2
0-2
0-2
0-3
0-3
0-3
1983
1984


Total
Zn
(ma/kg)
95
158
67
17
78
65
125
133
93
73
55
116
85
148
87
110
104
90
166
94
48
50
88
114
98
72

7
9
9
23
13
16
20
30
42
42
39
73
591
632
27
34
30
62
36
35
26.0
20.0
93.3
88.4
Pore Water
Cu
(ua/L)
-

25
15
41
70
32
8
114
23
37
22
22
251
77
78
1
18
61
12
0
64
0.0
20.0
N/A
48.5
Pore Water
Cr
(uq/L)
-

28
32
56
175
95
37
134
15
58
29
28
264
192
217
36
9
98
34
13
20
0.0
20.0
N/A
78.5
Pore Water
Zn
(ug/L)
-

45
6
13
180
46
6
313
16
31
6
35
148
60
62
1
5
100
5
1
5
0.0
20.0
N/A
54.2
Total Inorganic
Carbon Carbon
(X) (X)
2.98
3.59
3.92
2.62
2.95
4.71
5.18
4.90
4.68
3.49
4.75
4.90
3.84
4.20
4.17
4.02
4.67
4.44
4.73
3.71
3.20
4.92
4.11
4.71
3.67
4.09
.53
.03
.26
.55
.81
.32
2.20
.93
.42
.81
.54
.70
.58
.12
.37
.35
.39
.56
.78
.53
.58
3.78
.53
.51
1.19
1.41

1.97 1.55
2.06 1.06
1.78 1.50
4.49 1.77
6.13 1.11
5.04 1.51
4.98 1.57
5.92 1.5.4
2.97 1.40
5.38 1.55
4.25 1.53
4.55 1.53
5.59 2.54
5.85 1.41
3.54 .32
1.98 .25
3.63 .51
3.71 .36
4.98 .46
4.96 .62
26.0 26.0
20.0 20.0
4.1 1.5
4.2 1.5
Organic
Carbon
(X)
1.45
2.56
2.66
1.07
1.14
3.39
2.98
2.97
3.26
1.68
3.21
3.20
2.26
3.08
2.80
2.67
3.28
2.88
2.95
2.18
1.62
4.14
2.58
3.20
2.48
2.68

0.42
1.00
0.28
2.72
5.02
3.53
3.41
4.38
1.57
3.83
2.72
3.02
3.05
4.43
2.22
0.73
2.12
2.35
3.52
3.34
26.0
20.0
2.6
2.7
                                           136

-------
ignition vis-a-vis these three components was not known, but loss on ignition
values were significantly lower (p < 0.01) for the 1984 samples than
for the 1983 samples.  Organic and inorganic carbon values were not
different, on the average, between the two groups of samples.

     Density measurements of dried and ignited samples varied little between
the two surveys and gave values that would be anticipated for bulk river
sediments and their mineral fraction, respectively.

     Porosity, like moisture content, is an indicator of the degree of
compaction of the sediments, and can be of interest when considering sediment
resuspension and potential for diffusive flux of contaminants into the water
column.  Like the moisture content, porosity of the sediments decreased
significantly in samples collected in 1984 compared to those in 1983.  The
one sediment profile collected in 1984 did not show the typical decrease in
porosity with depth (seen in the 1983 profiles) that is characteristic of
relatively undisturbed sediments.

7.3.2  Metallic Contaminant Characterization

7.3.2.1  Approach--

     The focus of this section is on the contamination of River Raisin bottom
sediments by heavy metals.  In this particular system zinc, copper, and
chromium were the metals of concern; selection of the metals of concern must
be made on a site-specific basis from a preliminary survey or from historical
data.  The chemical parameters analyzed in the River Raisin sediments
included:  total and pore water zinc, copper, and chromium; total organic
carbon; and total inorganic carbon.  These measurements contributed to such
assessments as:  the spatial distribution of metal contamination in the study
system sediments; an estimate of the historical contamination of the system
through sediment profiles; the partitioning of metals in bottom sediments
between solid and aqueous phases; the potential for water column
contamination by in-place metals through particle resuspension and/or aqueous
phase diffusion as an internal source in mass balances; data necessary to
evaluate sediment toxicity measurements.

     Samples for chemical characterization were aliquots of those collected
as described in Sections 7.3.1.1 and 7.3.1.2 above.  The inherent spatial and
temporal variability of bottom sediments in systems such as the River Raisin
makes it imperative to analyze both physical and chemical parameters on
aliquots of the same samples if one hopes to deduce significant
i nterrelationshi ps.

7.3.2.2  Methods--

     7.3.2.2.1  Field—The field procedures for collection of bottom
sediments, presented in Section 7.3.1.2, included taking aliquots for
chemical analyses.  Little additional field time was added to the effort when
sediment aliquots were also used for the chemical analyses.
                                    137

-------
     7.3.2.2.2  Laboratory--The following Is a brief description of the
methods used for the chemical analyses performed on the bottom sediments.

     Total organic carbon and total  inorganic carbon were measured by wet
combustion and gravimetric techniques.  The methods and apparatus used in
these determinations are described in detail in Black (1965).

     Total Copper (Cu), chromium (Cr), and zinc (Zn) concentrations in the
sediment samples were determined by nitric acid digestion followed by flame
atomic absorption spectrophotometry.   The digestion method used was a
modification of EPA Method 3010 (U.S. EPA, 1982).  Hydrochloric acid was
eliminated from the digestion due to interference of the chloride ion with
zinc analysis (Mcllroy, 1984).

     Digested samples were analyzed by flame atomic absorption using a VARIAN
AA-475 Spectrophotometer.  Background correction was not utilized since
previous experience with the analysis of these metals in similar matrices,
along with information available in the literature (Mcllroy, 1984), suggested
that no significant chemical interferences would occur.  Instrument
operational parameters were within the optimum working range recommended by
the manufacturer, and all readings were within the linear range of the
working absorption curve as deduced by analysis of standards.

     Concentrations of copper, chromium and zinc in sediment interstitial
(pore) water were determined by using centrifugation to separate interstitial
water from the sediment.  This was followed by flame or flame!ess atomic
absorption analysis of the supernatant pore water, depending on the detection
limit for the particular metal.  Sediment pore water was extracted by
centrifuging a 1-5 g sample aliquot at 35,000 x gravity for 45 minutes in a
Beckmann Model J2-21 centrifuge.  The temperature was held at 4*C during
centrifugation, to inhibit chemical or biochemical reactions.  Following
centrifugation the supernatant was decanted into 15 ml graduated tubes,
acidified to pH 2 or less with nitric acid  (J.T. Baker Co. "Instra-Analyzed"
grade), and analyzed immediately.

7.3.2.3  Results and Evaluation--

     The chemical characterization results  for River Raisin bottom sediments
are presented, along with physical characterization data, in Table 7.3-2.
One use for these data would be to compare  average values for the study
system with those from other systems or with toxic levels known to produce
environmental impacts.  For example, the average values obtained for total
copper, zinc, and chromium  (112.7, 231.7 and 91.3 ug/g, respectively) are
elevated in comparison with sediment analysis results for southern Lake Huron
(Rygwelski, 1984).  Average values obtained for soluble (pore water) metals
were at less than toxic levels with the exception of copper, which at 48 ug/L
exceeded the chronic LC50 value for bivalves of 17 ug/L (Hart and Fuller,
1974).

     These data also permit a  "mapping" of  the sediment contamination in the
study system.  One can attempt to locate areas of excessive  sediment toxic


                                     138

-------
levels ("hot spots") for the purpose of assessing the potential for in-place
pollutant impacts.  For example, sediment from River Raisin transect 47 (see
Figure 7.17) appeared to contain elevated metals levels relative to other
locations; therefore, the assessment of potential sediment-water transport of
metals from this area is more crucial than from other areas of the system.
Another observation is that, while surface sediment metal levels in transect
42 sediments (turning basin) were somewhat elevated in 1983, they were
measurably lower in the 1984 survey.  A possible explanation for this finding
is that high spring flows resuspended finer turning basin sediments, leaving
larger-sized, less-contaminated sediments behind.

     While correlations between sediment characteristics should be evaluated
with caution due to the circumstantial nature of the evidence they provide,
they may nonetheless give some insight into what sediment parameters most
significantly affect the transport and fate of contaminants in the system of
interest.  Of particular interest, therefore, are the correlations between
total and soluble metal concentrations and other sediment parameters.  A
bivariate correlation analysis revealed that the three total metals (copper,
zinc, and chromium) were very strongly intercorrelated (p < 0.0001) and also
correlated significantly with total inorganic carbon (TIC) (p < 0.01 for Cu;
p < 0.05 for Zn and Cr), stratum depth (p < 0.01 for all three metals), and  .
loss on ignition (p < 0.05 for Cu; p < 0.01 for Zn and Cr).  The strong
intercorrelation observed between total copper, zinc, and chromium suggests
two possibilities:  1) that each of these metals has become associated with a
particular sediment component, which would lead to increased partitioning to
sediments enriched with a greater fraction of that component; and 2) that the
metals had been discharged to the system from a small number of sources,
resulting in their co-occurrence due to rapid association with aquatic
particulate matter upon entering the river.  The lack of a strong correlation
between total metals and other sediment characteristics does not support the
former hypothesis.  It is more likely that surficial distribution of these
metals in the River Raisin-Monroe Harbor is controlled largely by the
transport characteristics of the particles with which they are associated.

     The above-mentioned positive correlations between stratum depth and
total metal concentrations may indicate that inputs of heavy metals to the
River Raisin-Monroe Harbor system were greater in the past than at present,
which is consistent with the industrial history of the area.  However, the
complex nature of sediment transport in rivers as well as the possible
effects of dredging makes it difficult to draw any conclusions from these
correlations.

     In order to gain some insight into metals partitioning in bottom
sediments and to reduce the effect of random variation on the analysis of
correlations, a principal components-factor analysis were performed on the
sediment data.  It was found that four factors were adequate to explain 82.5
percent of the variation among the data.  The four factors were grouped as
fol1ows:

     Factor 1 (35% of common variance) - total metal levels of the sediments,
pore-water Zn concentration, sediment depth, and inorganic carbon content;


                                     139

-------
     Factor 2 (27% of common variance) - physical  properties (moisture,
density, porosity), loss on ignition,  and transect.

     Factor 3 (20% of common variance) - pore water metal concentrations;

     Factor 4 (16% of common variance) - organic carbon content of sediments.

     A similar analysis performed on sediments from the Vesdre River, Belgium
by Houba e_t aJL (1983) obtained comparable results.   In both cases the
results suggest that the metals originate from sources in close proximity and
that relatively rapid partitioning to particulate matter, followed by
sediment transport and deposition, may be primarily responsible for their
distribution within this aquatic system.

7.3.3  Organochlorine Contaminant Characterization

7.3.3.1  Approach--

     Analysis of sediment samples collected on a reconnaissance survey and
review of historic data from the study site indicated that polychlorinated
biphenyls (PCBs) were the major class of organochlorine contaminant present. .
In this report, PCBs will be the only organochlorine contaminant discussed.

     The sediment sampling plan was designed to provide measurements of the
surficial and vertical distribution of PCBs in sediments of the study area.
The PCB sediment data provides necessary data for mass balance measurements,
and for the sediment toxicity and bioaccumulation components of this study.

7.3.3.2  Methods--

     For a more complete description of field and laboratory methods, see
Filkins el a!., 1985.

     7.3.3.2.1  Field—To provide comparability of data, sediment samples
analyzed for PCBs were aliquots of the samples collected for metallic
contaminant and physical characterization.  The procedures for the collection
of sediment samples are presented in Section 7.3.1.3.

     7.3.3.2.2  Laboratory--Upon return to the laboratory, samples were
stored  in the dark at 4'C until extraction.  A sediment  sample, 6-38 grams,
was mixed with anhydrous sodium sulfate and Soxhlet extracted with equal
volumes of n-hexane and acetone for six hours.  Prior to analyzing all the
survey  samples, extraction and analysis of selected Monroe Harbor sediments
were performed to ascertain the length of extraction time required for
optimal recovery.  The extract was concentrated in a Kuderna Danish  apparatus
on a steam bath.  The extract was further concentrated under a stream of dry
nitrogen gas in a graduated tube heated to 30-40'C in a  hot water bath.

     An aliquot of extract was diluted and potentially interfering compounds
removed by exposure to an equal volume of concentrated sulfuric acid.  The
aqueous layer was  removed by freezing in an acetone dry  ice bath and the


                                      140

-------
hexane layer decanted.  Sulfur, which would interfere with the analysis, was
removed by mixing the cleaned extract with nitric acid-washed granular
copper.

     High resolution, fused silica, capillary, gas chromatography was
performed on a VARIAN Model 3700 gas chromatograph equipped with a 63Ni
electron capture detector.  The individual PCB congeners were quantified and
summed to provide PCB homolog and total PCB values.  Further information on
PCB analysis and quantitation can be found in Section 7.1.3.1.

7.3.3.3  Results and Evaluation--

     For detailed results of the PCB characterization of Monroe Harbor
sediment, see Filkins et al_., 1985.

     The use of high resolution, capillary, gas chromatography provides a
more accurate quantitation of total PCB values and homolog values than would
be possible with a conventional packed column approach, since potentially
interfering compounds are more likely to be resolved from the PCBs.

     The sampling design allows one to describe the surficial distribution
(0-2 cm) of PCBs in the sediments for collection made in 1983 and 1984.  As
was described earlier, individual congeners were quantitated and summed to
provide a total PCB value for the surficial sediments.  As an example,
surficial total PCB data from samples collected in 1983 are plotted in Figure
7.18.

     Sediment samples collected at Stations T6/S and T9/S were composed of
sand and gravel with some silt and contained PCB concentrations of 0.2 and
0.092 mg/kg, respectively.

     Station 3A is located just downstream of the Monroe Wastewater Treatment
Plant at the mouth of a canal which, in the past, discharged waste water from
paper processing plants into the river.  PCB concentrations at this site
showed a 4-fold higher concentration of PCBs (0.49 mg/kg) relative to
Stations T9/S and T6/S.

     Station T12/S is located adjacent to a marshy undeveloped area where the
sediment was characterized as clay and silt, gray in color.  The total PCB
concentration was increased by 2.5 times (1.3 mg/kg) over that at Station 3A.

     Station 8A is located at the mouth of Mason Run, where sediment was
characterized as hard and peaty, not typical of the sediments seen on the
River Raisin.  The PCB concentration measured at this station was the lowest
record (0.089 mg/kg) for PCBs in surficial sediments collected in 1983.

     Stations located along Transect T42 in the turning basin had PCB
concentrations ranging from 0.3 to 0.55 mg/kg.  The sediment along this
transect consisted of silt and sand with some detritus, and appeared gray
with a slightly oily odor.
                                    141

-------

                                                00
                                                CTi
                                                l/l
                                                +J
                                                c
                                                O)



                                                (1)



                                                03

                                                O
                                                o
                                                J2
                                                S-
                                                o
                                                 QJ
                                                 O
                                                 C
                                                 o
                                                •o

                                                ca
                                                o
                                                D.
                                                 itl
                                                00

                                                  •
                                                r~-

                                                 01

                                                 3
142

-------
     Station 54, located just downstream from the turning basin, exhibited a
5-fold Increase in PCB concentration compared to that of T42.  Sediments
collected consisted of silt with some sand showing traces of oil and wood
fragments.

     Transect T47 1s located 1-km downstream of the turning basin and
exhibited PCB values ranging from 1.9 to 5.1 mg/kg.  The PCB concentration of
5.1 mg/kg was the highest concentration of PCBs found in surficial sediments
collected during 1983.  The sediment found along Transect 42 was silty
showing some detritus.  At Station T47/N, the sediment looked black and oily,
exhibiting an odor of oil.

     Transect 50 is located near the river mouth and had PCB concentrations
ranging from 0.28 to 0.30 mg/kg.  Sediments were silty, gray-brown in color
and exhibited some detritus but no unusual odor.

     Station 11 located 5-km offshore exhibited a low PCB concentration (0.14
mg/kg).

     Surficial sediments were collected at 18 stations in 1984.  Three
transects and Station 11 were sampled in both 1983 and 1984.  Figure 7.19 is
a plot of PCB concentrations for surficial sediment samples collected in
1984.

     Monroe Harbor transects 42, 47, 50 and Station 11 were sampled in July
1983 and May 1984.  Comparing the plotted total PCB concentrations for
surficial sediments (Figures 7.18 and 7.19), Transect 42 shows an Increase in
PCB concentrations at Stations T42/S and T42/N.  Samples could not be
obtained from the center of the transect in the turning basin, indicating
that some type of scouring event may have occurred between July 1983 and May
1984.  This event may have been associated with the strong currents
encountered during spring runoff.  Transect 47 showed no clear indications of
an increase or decrease in PCB concentrations as a whole.  However, Station
T47/N did exhibit a decrease in PCB concentration (1.4 mg/kg) by a factor of
3.6 from the previous year.  The total PCB concentrations at Transect 50
appear to have remained constant from 1983 to 1984.  At Station 11 in Lake
Erie, a comparison of the samples collected in 1983 and 1984 showed an
increase by a factor of 2.4 in PCB concentrations.  It is not clear whether
these changes seen at certain stations from one year to the next are real, or
are the result of sampling variability.

     Two cores were collected at each station and core intervals were
composited by depth for analysis.  The collection and analysis of individual
cores, 1n sufficient number to statistically describe the variability among
the cores collected at each station, would have allowed a more accurate
comparison of the samples collected from one year to the next.  In addition,
the knowledge of the sampling variability at a site would have permitted the
accuracy of the sample's representation of the sediments at that station to
be evaluated.  Resources were not available for this optimal sampling
approach and, therefore, samples were composited.
                                     143

-------
                                                                                                    00
                                                                                                    CTl
                                                                                                     0)
                                                                                                     O)
                                                                                                    CO
                                                                                                     ITS
                                                                                                    •^
                                                                                                     u
                                                                                                     t.
                                                                                                     O
                                                                                                     CO
                                                                                                     (1)
                                                                                                     O

                                                                                                     C
                                                                                                     O
                                                                                                     O
                                                                                                     O)
                                                                                                     o
                                                                                                     C
                                                                                                     o
                                                                                                    • o

                                                                                                     CO
                                                                                                     o
                                                                                                     Q-


                                                                                                     (O
                                                                                                     4->
                                                                                                     O
                                                                                                      0)
D.UCD       2 O
                                                    144

-------
     Obtaining samples by coring, where possible, rather than by Ponar
dredge, provided Information on the vertical distribution of PCBs within the
sediment.  At Station 4, a 10 cm core cut into 2 cm slices showed little
change in PCB concentration from the sediment surface to a depth of 10 cm
(Figure 7.20).  However, results from Stations T43/S and T47/C showed a sharp
increase in PCB concentrations at the 12-14 cm depth and the 8-10 cm depth,
respectively (Figures 7.21 and 7.22).  To address bioavailability of the
contaminants in the sediments, it is necessary to know to what depth in the
sediment the organisms penetrate and the concentration of contaminants in
that zone.  Unless core slices are analyzed, the detailed vertical
distribution of contaminants can not be determined.  A Ponar grab sampler by
virtue of its method of collection can provide only composite sampling of
contaminants.

     Samples for analysis of individual PCB congeners or homologs may be
treated in the same manner as described above for total PCBs.  However, the
quantitation of homologs and congeners provides an opportunity to evaluate
the composition of the total PCB concentration as reflected in their
respective patterns.  One approach to this pattern recognition is to plot the
homolog concentration as a percentage of the total PCB concentration (Figure
7.23).  Inspection of these plots shows a similar pattern in river sediment
homolog composition, the dominant homologs being the tri- and
tetra-chlorinated biphenyls.  However, results from Station 11A (5-km
offshore) and, to a lesser extent, from Station T50/C indicated an increase
in the percent composition of the hexa- and hepta-chlorinated biphenyls
homologs.

     One explanation for this change in homolog patterns might be the greater
influence of Lake Erie (Station 11) water and sediment on Station T50.  This
influence might be explained by the large volume of cooling water required by
the Edison Power Plant.  During spring runoff, the River Raisin makes up more
than 95% of the cooling water.  However, during low flow, in the summer, the
river makes up less than 5%, the balance of water being from Lake Erie.  Due
to the Edison's cooling water requirements, the River Raisin is diverted at
T48 (Figure 7.17) and discharged south of the river mouth at Plum Creek.  As
a result, sediments at Station T50 might be influenced more by Lake Erie
water (Station 11) than the River Raisin, as indicated by its homolog
patterns and lower PCB concentrations.

     Pattern recognition analysis can also be used to compare contaminant
mixtures in sediments to those in the water column and biota to gain a better
understanding of the movement of contaminant within food chains (Section
7.2).

7.3.4  Adsorption/Desorption Experiments

7.3.4.1  Approach--

     In regard to toxic pollutant transport, one of the most significant
mechanisms is the association (either by adsorption or uptake) of a
contaminant with both nonviable and viable particulate matter, followed by


                                     145

-------
                                                        CO

                                                         T-
                                                            co -o
                                                                01
                                                            ut   C
                                                            ID  O
                                                            c^  »^»
                                                            4-> 4->
                                                            C  fO
                                                            a> 4J
                                                            o co
                                                                                                         CQ
                                                                                                         ,l_>
                                                                                                         a.
                                                                                                          o
                                                                                                          CVJ
                                                                                                          OJ
                                                      146

-------
                                                   o
                                                  .a

                                                   HJ
                                                    **•
                                                   I- 00
                                                   0) Of
                                                      0)
                                                  .C  1.
                                                  •*->  o
                                                   0.0
                                                   *°  .!=

                                                   to  "5
                                                   C  O)
                                                   O  OO
                                                   J- co
                                                  •»-> ^~
                                                   C (—
                                                   0)
                                                   u  c
                                                   c  o
                                                   o •«-
                                                  CJ -M
                                                      
                                                   o
                                                   a>
147

-------
                                                                                                    O
                                                                                                   -C
                                                                                                    ra
                                                                                                    O
Q. O  ffl
                                                                                                    to

                                                                                                    (T3 '*"**
                                                                                                    > <3-
                                                                                                    S_ CO
                                                                                                    CU Ol
                                                                                                        QJ
                                                                                                    J=  S_
                                                                                                    4J  O
                                                                                                    Q.O
                                                                                                    Ol
                                                                                                    O -U

                                                                                                    •P  (V
                                                                                                    t/l T3
                                                                                                    C  (V
                                                                                                    O l/>
                                                                                                    cu
                                                                                                    O  C
                                                                                                    C  O
                                                                                                    O •!-
                                                                                                    ce

                                                                                                    Q.


                                                                                                     tO

                                                                                                     O
                                                                                                     OJ
                                                                                                     i-
                                                                                                     3
                                                                                                     cn
                                                148

-------
   HOMOLOG PATTERNS
   1984 MONROE HARBOR  SEDIMENTS
   0 - 2 cm
                 ,2/N
                 X13/C
h
T» «
>
/
•V
/
/
/
/
X
/
/
/
/
*
k ^, .^
            .'42/N
^    _	47/C
      	50/C
(Fig. 5.3)-
              11
                                  40
                        ?     2O
                       PCB
                                    0123456789
                                  40
                            PCB

                                          2  3  4   5   6   7   8   9
                                  4O
                            %
                            PCB  20

                                   o

                                          23456789
                                  4O
                            I
                            PCB
             ^

             P

          II
                                                «:S
                                          23456789
                        Z
                        PCB
4O



20


 O


40
                                    01  23456789
                            PCB
                                  20 ^

                                    01   23456789
 Figure 7.23.   PCB Homolog Patterns in Surficial  Sediments
           From Selected Monroe Harbor Stations.
                            149

-------
the transport of the Interacting particles.   Association with participate
matter significantly alters the transport regime of a chemical by introducing
additional transport processes, such as deposition and resuspension.
Furthermore, the association of a toxicant with sediment material can
indirectly (by altering its bioavailability) affect the rate and extent of
its biochemical transformations and biotic accumulation, thus potentially
altering its exposure potential and/or toxicity.  Any toxicant
exposure/effects assessment must, therefore, investigate the behavior of the
contaminants of interest with respect to their interactions with aquatic
particulate matter.

     The approach taken in this study was to conduct adsorption and
desorption equilibrium and rate experiments using a well-character!zed bottom
sediment sample, composited from several locations within the Monroe Harbor
turning basin.  Total suspended solids concentrations (200 mg/L), pH (8.3),
and total alkalinity (3 meq/L) typical of the lower River Raisin-Monroe
Harbor system during high flow conditions were duplicated in an experimental
suspension medium.  This synthetic river water also contained an indifferent
electrolyte (NaN03) to maintain a constant ionic strength.  Adsorption/-
desorption experiments were performed in this medium with both heavy metals
(Copper (II), Chromium (III), Zinc (II)) and organics (Hexachlorobiphenyl
(HCBP), and Hexachlorobenzene (HCB)).

     In addition to working with river bottom sediments, sorption/desorption
experiments were performed to determine equilibrium partition coefficients
and the rate and extent of desorption of hexachlorobenzene (HCB) and
hexachlorobiphenyl (HCBP) with several species of phytoplankton common to the
lower River Raisin.  The potential for algae in a system, having much higher
organic content than allochthonous solids of soil origin or resuspended
bottom sediments, to become a major compartment for hydrophobic organic
compounds must be investigated.  Such was the intent of this part of the
experimental effort.

7.3.4.2  Methods--

     Presented in this section is a brief description of the experimental
design employed in the contaminant-particle association studies on the River
Raisin project.  Details of these experiments can be found in the full
project report from Clarkson University (DePinto et aj.., 1986).  The
experimental design for this study represents only an example of what might
be done in this area on a given exposure/effects assessment project.
Depending on project objectives and resources available, the experimental
design for studies of this nature would have to be developed on a
site-specific basis.

     The experimental work on adsorption and desorption of toxics performed
in this study was relatively extensive.

     7.3.4.2.1  Meta1s--It is widely recognized that association of trace
metals with aquatic particulate matter  is greatly influenced by pH and by
concentrations of trace metals, adsorbent (sediment), and competing ligands


                                      150

-------
(Benjamin and Leckie, 1981).  Consequently, these characteristics of the
River Ra1sin-Monroe Harbor system were determined from a review of water
quality data collected on the lower River Raisin (LEWMS, 1979; E. Smith,
personal communication).  From these data a synthetic river water was
prepared for use as an aqueous medium for all experimental work:  pH 8.3, 3
meq total alkalinity/L, and 200 mg total suspended solids/L; NaN03 was
added to yield a "concentration of 0.1 N to maintain a constant ionic strength
among experimental treatments.

     Adsorption of Cu, Zn, and Cr by River Raisin sediments was evaluated by
generation of adsorption isotherms over successively longer adsorption
equilibration periods.  The metal-sediment-water system was considered to be
at equilibrium with respect to metal adsorption when the value of the
partition coefficient (slope of the isotherm) obtained did not increase
significantly with increased equilibration time.  Desorption of the metals
was studied by replacing the aqueous phase from metal-containing
water-sediment mixtures with metal-free aqueous media after the sediments had
reached an adsorption equilibria.  Replicate sets of sediments resuspended in
metal-free media were allowed to desorb for periods of varying length.  This
permitted evaluation of both the rate and extent of desorption from a single
experiment.  A flowchart of the metal adsorption/desorption experiments is
shown in Figure 7.24.

     7.3.4.2.2.  Oraanics--The experimental design for this portion of the
study was developed around the need for further information on the behavior
of toxic organic compounds relative to partitioning to..aquatic sediments and
phytoplankton.  The compounds used in this study were   C labelled
hexachlorobenzene (HCB) and 2,4,5,2',4',5' hexachlorobiphenyl (HCBP), two
organic pollutants considered to be representative of the hydrophobic
contaminants found in the River Raisin sediments and water column.  The
experiments performed could be categorized into four general areas:

     a.  Equilibrium sorption experiments using River Raisin experimental
         sediments;

     b.  Sorption/desorption partitioning to phytoplankton;

     c.  Desorption kinetic studies; and

     d.  Bioavailability of sediment-associated organics to phytoplankton.

     The first general area of experimentation consisted of determining
sorption partition coefficients for HCB and HCBP to River Raisin sediments.
The sorbent used for these experiments was the same composite sediment used
for the metal adsorption/desorption experiments, described previously.
Aqueous chemical conditions (pH, alkalinity, ionic strength) were identical
to those In the metal experiments.

     The next three areas of experimentation involved the use of
phytoplankton as an interacting sorbent.  Representative species from three
major phycological taxa, all typically present in the lower River Raisin,


                                     151

-------
 Metal-Free

 Medium
Experimental
   Metal
                  Sediment

                  Sample
 Stock Slurry

 ~lg TSS/L
 Experimental
  Suspension

-200 moTTSS/L
                     I
                   Replicate

                   Aliquots
                                             1
                          Replicate

                          Aliquots
                                             at
Incubate
Required
Period
                                           HNO3

                                           Digestion
                                             Metal
                                            Analysis
                                             (AAS)
Figure 7.24.   Flowchart of Metal  Adsorption/Desorption  Experiments,
                                  152

-------
were chosen as test species:  a chlorophyte, Scenedesmus auadricauda. a
diatom, Cvclotella sp. (both obtained from the Richard C. Starr collection,
Department of Botany, University of Texas-Austin) and a cyanophyte,
Microcvstis sp. (obtained from Dr. G-Yull Rhee, Division of Labs and
Research, New York State Department of Health, Albany, New York).  Both
continuous flow reactors (chemostats) and batch algal cultures were utilized
to produce a sorbent, depending upon the needs of the experiment.  Continuous
flow reactors were used to produce steady-state phytoplankton cultures for
the evaluation of the effect of species and physiological state on sorptive
and desorptive partitioning of HCB and HCBP.  Batch algal cultures sufficed
for developmental experiments, the determination of desorption rate
constants, solids effects investigations, and bioavailability evaluations.

     Several procedures were developed to accomplish the goals of this
research.  Due to hydrophobicity of the chosen test compounds, extreme care
was required in handling, dissolving, and storing aqueous solutions of these
compounds to minimize losses over time.  Quantification of compound
concentration via liquid scintillation analyses required the development of
techniques to ensure efficient and accurate counting of both particulate and
dissolved fractions.  Numerous preliminary experiments were conducted to
evaluate equilibration times for the test compounds and phytoplankton prior
to the sorption and desorption experiments.  To evaluate the rate and extent"
of desorption, an air purging apparatus and associated experimental
techniques were developed.  A density-dependent separation technique was
adapted to evaluate the extent of bioavailability to phytoplankton of HCBP
associated with River Raisin sediments.  A detailed account of the procedural
development of this will not be discussed in this report; however, it can be
found in Autenrieth (1985).

     Phytoplankton sorption and desorption isotherm experiments were
conducted to determine partition coefficients for HCB and HCBP.  The
experimental variables for these studies included:  sorbate organic
compounds, algal species, physiological state of algae, and concentration of
algal sorbent.  The desorption kinetics experiments employed an air-purging
apparatus for HCB-equilibrated algal suspensions and River Raisin sediments.
The purging apparatus was similar in design to that used by Karickhoff
(1980); it was employed to maintain a favorable desorption gradient, a
requirement for simple batch reactors.

     The purpose of the bioavailability experiments was to determine whether
sediment-bound HCBP could be taken up by algal populations when both solid
phases were incubated in the same medium.  The algal-bioavailability of
particulate-associated organics is considered to be a dynamic competition for
the compound between the algae and the sediments.  Hence, the bioavailability
experiments involved incubation of an initially contaminant-free algal
culture in the presence of a suspension of contaminated sediments for a fixed
period of time.  Following the incubation, determination of the quantity of
contaminant that had been transferred to the algal cells was considered to be
a measure of the algal-availability of that particular compound under the
specific environmental conditions of the incubation.  A radio-labelled
compound was used for these experiments, and the contaminant concentration in


                                     153

-------
each compartment of the batch systems (algae, sediments, and water) was
determined after quantitative separation of algae and sediments by density
gradient centrlfugation.

7.3.4.3  Results and Evaluation

     7j3.4.3.1  Metals--Adsorption equilibria showed a linear dependence of
total particulate metal content to soluble metal concentration.  Partition
coefficients (48-hour equilibration) amounted to -50, 30, and 25 L/g for Cu,
Cr, and Zn, respectively.  The dependence became nonlinear, however, at
adsorption densities greater than approximately 6,000 yg/g for Cr and 9,000
ug/g for both Cu and Zn.  Partition coefficients computed from water column
field data gave a much wider range of values for the same metals compared to
the laboratory observations.  Metals partition coefficients calculated from
total and pore water bottom sediment data were approximately a factor of five
lower than the laboratory equilibrium results.  Several factors contributed
to differences in coefficients between the field and laboratory; a more
thorough discussion of metals partitioning to River Raisin sediments can be
found in Young e_t a].., 1987.

     Desorption of Zn was completely and rapidly reversible (24-48 hours) in .
contrast to that of Cr, which was much slower; Cu desorption was Intermediate
between that of Zn and Cr.  Metal desorption from the experimental sediments
could be described by assuming the metals were retained in reversible
("loosely-bound") and resistant ("tightly-bound") forms.  However, overall
desorption did not obey closely the reversible-resistant behavior described
by Di Toro and Horzempa (1982).  Rather, metal desorption did not reach a
metastable equilibrium even when permitted to desorb for periods as long as
24 days.  The observed desorption, however, was readily described by a simple
mathematical model that treated sorbed metal ions as existing in two forms
which desorbed at different rates ("fast" and "slow"), either simultaneously
or sequentially (Young et at., 1987) with different mobilities.  Actual
mechanisms responsible for the observations could not be tested adequately
with the available data; however, aqueous chemical considerations pointed
toward the usefulness of differentiating adsorbed from precipitated species.

     7.3.4.3.2  Orqanics--Values obtained for River Raisin experimental
sediment-water partition coefficients were 14 L/g for HCB and 40 L/g for
HCBP.  While partition coefficients for HCB in the Great Lakes generally are
not  available, a number of researchers have reported partition coefficients
for  HCBP using Great Lakes sediments as the adsorbent phase.  Values reported
by Horzempa and Di Toro (1983) for partitioning of HCBP to 1100 mg/L of
Saginaw Bay sediments, were 9-14 L/g (9,000-14,000 L/kg).  These sediments
varied somewhat in composition but  in general were similar to Monroe Harbor
sediments, being composed predominately of silt with total organic contents
of 1-5%.  Voice and Weber (1983) also reported partition coefficients for
PCBs on Saginaw Bay sediments, but used the PCB mixture Aroclor 1254 rather
than an individual congener.  Partition coefficients were evaluated for a
variety of suspended solids concentrations and were generally about an order
of magnitude higher than those reported by Horzempa and Di Toro.  The
partition coefficient of 40 L/g obtained for HCBP in this study falls

                                     154

-------
approximately mid-way between the values reported by Horzempa and 01 Toro
(1983) and Voice and Weber (1983).  Our measured values are also within a
factor of two of the predictions of the Karlchoff (1980) model that is an
empirical regression based on the octanol-water partition coefficient of the
sorbate and the organic content of the sediment.

     Sorbents for the study of batch sorption and desorption of HCB and HCBP
as a function of algal species and algal physiological state were produced
using chemostat cultures of Scenedesmus ouadricauda. Microcvstis sp., and
Cvclotella sp.  The sorption (Ks) and desorption (Kj) partition
coefficients for these experiments are presented in Table 7.20.  An example
of the data used to generate the partition coefficients in Table 7.20,
showing the plots from the HCB-Cvclotella systems, is presented in Figure
7.25.

     Inspection of Table 7.20 yields several interesting observations:

     1.  Cvclotella (a centric diatom) had partition coefficients
approximately an order of magnitude larger than did either Scenedesmuj (a
green alga) or Microcvstis (a blue-green alga).

     2.  Partition coefficients for both compounds to Scenedesmus and
Microcvstis were similar to the experimental values determined using the
River Raisin sediments.

     3.  With respect to variation of sorption partitioning with algal
physiological state, only the Cvclotella cultures exhibited a significant
trend with respect to chemostat dilution rate.  There was a significant
decrease in partition coefficients for these cultures as dilution rate
(equivalent to growth rate at steady-state) increased.  This is further
evidence for a cell constituent-related partitioning, the value of which
changes as a function of growth rate for Cvclotella but not significantly for
the other two species.  Perhaps the slower-growing diatoms tend to build up
higher cell levels of certain organic constituents, such as lipids, that are
primary responsible for the increased affinity for these organochlorines.

     4.  The desorption partition coefficients for all three species were
consistently higher than the corresponding sorption values.  These results
indicate a hysteresis in the sorption-desorption process for organic to algae
partitioning that has been observed for the same class of compounds to
aquatic sediments (01 Toro and Horzempa, 1982).

     5.  Finally, with the exception of Microcvstis. the HCBP exhibited
stronger partitioning than the HCB.  Since HCBP is more hydrophobic than HCB,
this is not a surprising result.

     A typical desorption kinetics experimental result is presented in Figure
7.26; which shows the cumulative HCB purged (using the air purging system to
maintain a desorption gradient) from Hicrocvstis as a function of time.
These studies, in general, revealed two desorption components:  a
fast-release component that came off at a first-order rate of approximately


                                     155

-------
  O)
  3
  O

  O
  u
  I
  O
  (A
     O 0.14 day-1
                      SORPTIVE PARTITIONING
                             Cyclotella sp.
                   Diss. Cone. (ug/L)    O 0.50 day-1
                    0.25 day-1
  o>
  3
  U

  O
  O

  •o
  «
  a
  o
to-


70-



60.



X



40



X



20
      O  0.14 day-1
                     DESORPTIVE PARTITIONING
                          Cyclotella sp.
                            02
Diss. Cone, (ug'L)
0.25 day-1
                                     O 0.50 day-1
Figure 7.25.
        Sorptive and Desorptive Partitioning of HCB
           in Cyclctella  Systems.
                             156

-------
                                                  c
                                                  3
                                                  O
                                                  re
                                                  4-1
                                                  o
                                            i—
                                                  
-------
        TABLE 7.20.   SORPTION AND DESORPTION PARTITION COEFFICIENTS FOR
           HEXACHLOROBENZENE  (HCB) AND HEXACHLOROBIPHENYL (HCBP) TO
             PHYTOPLANKTON HARVESTED FROM STEADY-STATE CHEMOSTATS




ORGANISM
Scenedesmus
quadricauda


Microcvstis sp.



Cvclotella sp.




DILUTION
RATE
(DAY-1)
0.08
0.30
0.50
0.24
0.12
0.23
0.44
0.19
0.14
0.26
0.49
0.27



COMPOUND
HCB
HCB
HCB
HCBP
HCB
HCB
HCB
HCBP
HCB
HCB
HCB
HCBP
SORPTION
PARTITION
COEFFICIENT
kd (L/9)
17.9
23.0
6.5
82.0
17.1
22.0
19.0
18.5
232.
167.
118.
448.
DESORPTION
PARTITION
COEFFICIENT
kd (L/g)
32.7
43.2
49.8
245.
114.
93.3
66.7
28.4
256.
197.
174.
"
1-4 day"1; and a slow-release component that was released at a much slower
rate of between 0.01 and 0.05 day'1.   The amount of HCB in each component
comprised about one-half of the initial total amount sorbed to the particles.

     Finally, bioavailability experiments, although preliminary, demonstrated
that a significant amount of sediment-bound HCBP was taken up by a-lgae
maintained in contact with the sediments for up to 24 hours.  This exchange
of material, which has major implications relative to organic contaminant
transport and bioaccumulation, was attributed to the preferential sorption of
nonpolar organics to higher organic content particulate material and, in
particular, to the lipophilicity of HCBP.
                                     158

-------
                                  SECTION 8.0

                        DATA BASE DOCUMENTATION SUMMARY
     Proper data management and quality is an essential component of a
complex integrated environmental study.  If the requirements of the system
are not well thought out and defined, confusion and delays will be
encountered.  Components of the data management process include:

                 1.  Project Leadership
                     Computer Programmers
                     Data Entry Clerks
                     Computer Operation Support

                 2.  Computer Hardware

                 3.  Documentation Procedures

     The initial step in the process is for project management to define the
types and sources of data and how those data are to be processed.  As
specifics become known, computer programmers can develop procedures for
storage and processing of the data.  Often the data are received on forms, as
listings, or in some other non-computerized fashion.  Data entry clerks are,
therefore, required to code and enter the information on the computer.  The
data and programs are then put on an in-house computer or on a remote system
that is accessed via in-house terminals.  These procedures must also be well
documented.  All too often programs and procedures are re-written because
details and knowledge of previous work have been lost because of lack of
documentation.

     There were several sources and types of data used in support of the
Monroe Harbor Project.  For example, physical data were obtained from the
City of Monroe, the Monroe Waste Water Treatment Plant, the U.S. Geological
Survey (USGS), the National Oceanographic and Atmospheric Administration
(NOAA), and through inhouse efforts.  Point source data were received from
the Michigan Department of Natural Resources.  Sediment data were obtained
from Clarkson University.  With such a large quantity of data, it would have
been difficult to distribute, analyze, graph, or process the information
without the aid of a computer.

     Prior to the study, the data base needs for the inhouse water quality
data were defined.  The data base had to be able to handle the final
concentration values for each of the conventional parameters, as well as for

                                     159

-------
total PCBs and homologs.  The total  number of parameters exceeded the
capability of the existing data base.   Special treatments for duplicates and
composites were also given.  For the first three surveys, each composite
sample corresponded to three grab samples.  This relationship was taken into
account in retrieval programs which  included raw data listings or statistical
printouts.  Retrieval specifications were entered interactively.  Midway
through development of the water quality data base,  the need for a
specialized organic analytical data  base became apparent.  The resultant
organics data base allowed the analyst to screen the data and prepare
intermediate data results.  This organics data base  was to be the only
storage location for PCS congener level data and pesticide data.  Organic
data were transferred to the water quality data base after final approval was
received.

     The field data for sediment and biota were entered into separate data
files.  Chemical data for the sediment samples were  eventually added to the
sediment field data.  Since the computer files were  not in a readable format,
programs to print the data were written.  A readable listing of all sediment
data could be generated by running the appropriate program.

     Two report-generating programs  were developed for the biota data.  One  .
listed all field data for the Monroe Harbor study.  The other listed field
data for specific sample identity (ID) codes.  Complete organic data for
biota samples were only available in the organic data base.

     When data from outside agencies were received,  a meeting with data
processing personnel was held.  The data and any required processing were
discussed.  Difficulties and progress were reported  periodically.  When the
project was completed, appropriate documentation forms were filled out and
the project was completed.

     A few improvements can be made to this approach.  First, the data
processing personnel should have knowledge of the sources and types of data
prior to its arrival.  Advanced knowledge of data types, processing and data
integration, combined with good documentation can save time and emotional
stress.  A second problem encountered was the confusion of using several data
bases for inhouse data.  It is much more desirable to have one, all
inclusive, data base.  This approach is more user-friendly.  Users do not
need to know how to run several retrieval programs,  one for each type of
data.  Instead they may familiarize themselves with only one retrieval
program.  A second advantage  is that copying data from one system to another,
such as from the organic data base to the water quality data base, requires
extra time and energy.  Maintaining consistant data sets was a constant
struggle.  Transfer from the organics data base to the water quality data
base was done when a bulk of data became available.   The organics data base
did not have access to the field information, therefore, erroneous sample  ID
codes could be entered  into the system.  These were detected when GIDS, the
water quality data base, was  updated.   In general, problems could be fixed in
one data base, but not  in the other.  Development of an  integrated data
system for the Large Lakes Research Station is currently in progress.  This
will make retrievals easier, more complete, and eliminate the need to access

                                      160

-------
two data bases for one set of related data.  The Integrated system will allow
for data checking so that errors will be minimized.   Having one copy of a set
of data will also eliminate the need for making the  same correction in two
places.  Another benefit is that there will be no delays in making
information available to end users.

     Appendix B contains specifics about data files  and programs for the
Monroe Harbor study.
                                     161

-------
                                  SECTION 9.0

                                  REFERENCES
American Society for Testing and Materials.  1984.  Proposed Standard
     Practices for Conducting Bioconcentration Tests With Fishes and
     Saltwater Bivalve Mollusks, Draft No. 8.  ASTM Committee E-47 on
     Biological Effects and Environmental Fate, Philadelphia, Pennsylvania.
     58 pp.

APHA.  1975.  Standard Methods for the Examination of Water and Waste Water.
     14th Edition.  Amer. Public Health Assoc., Washington, D.C.  1193 p.

APHA.  1980.  Standard Methods for the Examination of Water and Waste Water.
     15th Edition.  Amer. Public Health Assoc., Washington, D.C.

Auer, M.A. (Ed.).  1982.  Identification of Larval Fishes of the Great Lakes
     Basin With Emphasis on the Lake Michigan Drainage.  Great Lakes Fishery
     Commission Special Publication 82-3.  Ann Arbor, Michigan.  744 pp.

Autenreith, R.L.  1985.  Adsorption and desorption of hydrophobic organic
     compounds by algae.  Ph.D. dissertation, in preparation.  Clarkson
     University, Potsdam, New York.

Bedford, K.W. and M. Abdelrhman.  1986.  Improved acoustic and analysis
     methods for parameterizing resuspension in Monroe Harbor/Raisin River,
     Part 1 - Review of near bottom transport processes of importance to Lake
     Erie and the Great Lakes.  Draft report.

Bedford, K.W.  1986.   Improved acoustic and analysis methods for
     parameterizing resuspension in Monroe Harbor/Raisin River, Part 2 -
     Field and analytical procedures for initial estimation of sediment
     resuspension and  transport.  Draft report.

Bedford, K.N., C.M. Libicki, and R. Van Evra.  1986.  Improved acoustic and
     analysis methods  for parameterizing resuspension in Monroe Harbor/Raisin
     River, Part 3 - An acoustic procedure for measuring sediment
     concentration profiles.  Draft report.

Bedford, K.W., 0 Wai,  and C.M. Libicki.  1986.  Improved acoustic and
     analysis methods  for parameterizing resuspension in Monroe Harbor/Raisin
     River, Part 4 - An analysis of the Raisin River/Monroe Harbor.  Draft
     report.


                                      162

-------
Benjamin, M.H. and J.O.  Leckie.  1981.   Conceptual  model  for
     metal-liquid-surface Interactions  during adsorption.  Environ.  Sci.
     Techno!., 15(9):1050-1057.

Black, C.A.  1965.  Methods of Soil  Analysis.  Amer.  Soc. of Agronomy,
     Madison,  Wisconsin.

Branson, D.R., G.E. Blau, H.C. Alexander,  and W.B.  Neely.  1975.
     Bioconcentration of 2,2',4,4'-tetrachlorobiphenyl  in rainbow trout as
     measured by an accelerated test.   Trans. Am.  Fish.  Soc., 104:785-792.

Bridgham, S.,  D.C. McNaught, and C.  Meadows.   1987.   The effects  of
     photosynthesis in two Great Lakes  of complex  effluents from  the River
     Raisin.  ASTM Symposium on Functional Testing for Hazard Evaluation.
     In press.

Burch, J.B.  1973.  Freshwater Unionacean Clams (Mollusca:  Pelecypoda) of
     North America.  Biota of Freshwater Ecosystems Identification Manual  No.
     11.  U.S. Government Printing Office, Washington,  D.C.  176  pp.

Burkhart, L.P. and D. Weininger.  1987.  Anal. Chem., in press.

Cairns, M.A.,  A.V. Nebeker, J.H. Gasktatter,  and D.F. Krawczyk.  1984.
     Evaluating acute toxicity of contaminated Wisconsin sediments using
     liquid- (elutriate)  and solid-phase bioassays with freshwater
     invertebrates.  USEPA, ERL-Corvallis, Oregon.   Draft report.

Clarke, A.H.  1981.  The Freshwater Molluscs  of Canada.   National Museums  of
     Canada, Ottawa, Canada.  446 pp.

Cole, R.A.  1978.  Entrainment at a once-through cooling system on western
     Lake Erie.  Institute of Water Research  and Department of Fisheries and
     Wildlife, Michigan State University.   EPA-600/3-78-070, pp.  1-10.

DePinto, J.V., T.C. Young, R.L. Autenrieth, and T.W.  Kipp.  1986. .
     Interactions of heavy metals and  organics with particulate matter  from
     the lower River Raisin, Michigan.   Final report for Project  No.
     CR810776, Large Lakes Research Station,  ERL-Duluth,  Grosse He,
     Michigan.

Di Toro, D.M.  and L.M. Horzempa.  1982.  Reversible and resistant components
     of PCB adsorption-desorption isotherms.   Environ.  Sci. Technol., 16(9):
     594-602.

Di Toro, D.M., R.P. Winfield, J.P. Connolly,  S.M.  Kharkar, W.A. Wolf, C.J.
     Pederson, J.R. Blasland, and J.O.  Econom.  1985.  Development of toxic
     substances models of fate and ecosystem  effects in Monroe Harbor - River
     Raisin, Michigan.  Draft report.
                                     163

-------
Dolan, D.M., M.I. Gessner, S.  Hendrlcks,  D.A.  Griesmer,  and K.R.  Rygwelskl.
     1985.  Correlations of bioassay results and toxicant concentrations at
     Monroe Harbor, Michigan,  1983-1984.   Draft report.

Environment Canada.  1979.  Analytical  Methods Manual.   Inland Waters
     Directorate, Water Quality Branch, Ottawa, Canada.

Fay, L.A., M. Gessner, P.C. Stromberg,  and J.  Hageman.   1985.  Effects of
     toxic substances on growth, mortality and pathology of larval  fishes in
     the River Raisin, Michigan.  Draft report.

Filkins, J.C., M.D. Mullin, W.L. Richardson, V.E. Smith, J. Rathbun,  S.G.
     Rood, K.R. Rygwelski, and T. Kipp.  1985.  Report  on the distribution
     of polychlorinated biphenyls in sediments of lower River Raisin; Monroe
     Harbor, Michigan, 1983 and 1984.  Draft report.

International Joint Commission.  1983.   1983 report on  Great Lakes water
     quality, Appendix - Great Lakes surveillance.  Great Lakes Water Quality
     Board, Windsor, Ontario,  Canada.  130 pp.

International Joint Commission.  1985.   1985 report on  Great Lakes water
     quality.  Great Lakes Water Quality Board, Windsor, Ontario, Canada.

Griesmer, D.A.  1987.  Personal communication.  Computer Sciences
     Corporation, Grosse lie,  Michigan.

Hackney, P.A. and J.C. Webb.  1978.  A method for determining growth and
     mortality rate of ichthyoplankton.  in - Proc. Fourth Nat. Workshop on
     Entrainment and Impingement, L.D.  Jensen (ed.),  Ecological Analysts,
     Inc., Melville, New York, pp. 115-124.

Hamilton, M.A.  1984.  Statistical analysis of the seven-day Ceriodaphnia
     reticulata reproductivity toxicity test.  Report to the U.S.
     Environmental Protection Agency, ERL-Duluth, Minnesota.

Hart, C.W. and S.L.H. Fuller  (Eds.).  1974.  Pollution  Ecology of Freshwater
     Invertebrates.  Academic Press, pp. 389.

Hartley, D.M. and J.B. Johnston.  1983.  Use of the freshwater clam Corbicula
     manilensis as a monitor  for organochlorine pesticides.  Bull. Environ.
     Contain. Toxicol., 31:33-40.

Hobble, J.E.  1969.  A method for studying heterotrophic bacteria,  in - A
     Manual on Methods for Measuring Primary Production in Aquatic
     Environments.  Blackwell  Scientific Publications,  Oxford, pp. 146-151.

Horzempa, L.M. and D.M. Di Toro.  1983.  The extent of reversibility of
     polychlorinated biphenyl  adsorption.  Water Res.,  17(8):851-859.
                                     164

-------
Houba, D., J. Remade, D. Dubois, and J. Thorez.  1983.  Factors affecting
     the concentrations of cadmium, zinc, copper, and lead In the sediments
     of the Vesdre River.  Water Res., 17(10):1281-1286.

Jones, J.D.  1983.  Personal communication.  University of Michigan, Ann
     Arbor, Michigan.

Jude, D.J., P.J. Mansfield, and M. Perrone, Jr.  1983.  Impingement and
     entrainment of fish and effectiveness of the fish return system at the
     Monroe Power Plant, western Lake Erie, 1982-1983.  University of
     Michigan, Great Lakes Research Division Special  Report 101, Ann Arbor,
     Michigan.

Karickhoff, S.W.  1980.  Sorption kinetics of hydrophobic pollutants in
     natural sediments.  In - Contaminants and Sediments, Vol. 2., R.A. Baker
     (ed.), Ann Arbor Science, Michigan, pp. 193-025.

Kauss, P.B., M. Griffiths, and A. Melkic.  1981.  Use of freshwater clams in
     monitoring trace contaminant sources areas,  in  - Proc.  of Technology
     Transfer Conf. No. 2, pp. 371-388.

Kauss, P.B. and S.Y. Namely.  1985.  Biological  monitoring of organochlorine
     contaminants in the St. Clair and Detroit Rivers using introduced clams,
     Elliptic comolanatus.  J. Great Lakes Res., 11(3):247-263.

Kezdi, A.  1980.  Handbook of Soil Mechanics, Vol. 2:  Soil Testing.
     Elsevier Scientific, New York.

Mathiak, H.A.  1979.  A River Survey of the Unionid Mussels of Wisconsin:
     1973-1977.  Sandshell Press, Horicon, Wisconsin.  75 pp.

Mcllroy, L.M.  1984.  Partitioning of copper and zinc to Flint River
     suspended solids.  Master's thesis, Clarkson University, Potsdam, New
     York.

McNaught, D.C.  1975.  Selective feeding by Diaptomus and Daohnia.  Verh.
     Internet. Verein. Limnol., 19:2935-2942.

McNaught, D.C.  1982.  Short cycling of contaminants  by zooplankton and their
     impact on Great Lakes ecosystems.  J. Great Lakes Res.,  8(2):360-366.

McNaught, D.C., S. Bridgham, and C. Meadows.  1984.   Effects of contaminants
     in Raisin River on ecosystem functions of the bacteria,  phytoplankton
     and zooplankton.  Draft report.

McNaught, D.C., S. Bridgham, and C. Meadows.  1987.   The effects upon
     zooplankton grazing in Lake Erie of complex effluents from the Raisin
     River.  ASTM Symposium on Functional Testing for Hazard Evaluation.  In
     press.
                                    165

-------
Michigan Department of Natural  Resources.   1979.   River quality in the River
     Raisin basin.  Water Quality Division,  Biology Section,  Lansing,
     Michigan, pp. 3-4.

Michigan Department of Natural  Resources.   1985.   Unpublished data on  Port of
     Monroe Landfill.  Water Quality Division,  Lansing, Michigan.

Michigan Department of Natural  Resources.   1987.   Michigan sites of
     environmental contamination priority lists - Act 307.  Lansing,
     Michigan, pp. 213.

Monroe County Drain Commission.  1984.   Environmental assessment - Monroe
     metropolitan area.  pp. 1-6.

Mount, D.I. and T.J. Norberg.  1984.  A seven-day life cycle cladoceran
     toxicity test.  Society of Environmental Toxicology and Chemistry
     (SETAC), Vol. 3, pp. 425-434.

Mozola, Andrew J.  1970.  Geology for environmental planning in Monroe
     County, Michigan - report investigation 13.   Geological  Survey Division,
     Michigan Department of Natural Resources,  Lansing, Michigan,  pp.  1-8.

Norberg, T.J. and D.I. Mount.  1982.  A seven-day early life stage growth
     test using the fathead minnow (Pimeohales oromelas).  USEPA,  ERL-Duluth,
     Minnesota, unpublished manuscript.

Norberg, T.J. and D.I. Mount.  1983.  A seven-day early life stage growth
     test using the fathead minnow (Piroephales promelas).  USEPA,  ERL-Duluth,
     Minnesota, unpublished manuscript.

O'Connor, D.J. and J.P. Connolly.  1980.  The effect of concentration of
     adsorbing solids on the partition coefficients.  Water Res.,
     14:1517-1523.

Pennak, R.W.  1978.  Freshwater Invertebrates of the United States, 2nd
     Edition.  John Wiley & Sons, New York.  803 pp.

Perkin-Elmer Corporation.  1977.  Analytical methods using the HGA graphite
     furnace.  Perkin-Elmer Corporation, Norwalk, Connecticut.

Pritchard, D.W.   1969.  Dispersion and flushing of pollutants in estuaries.
     0. of Hydraulics Division, ASCE, HY1, pp. 115-124.

Pugsley, C.W., P.D.N. Hebert, G.W. Wood, G. Brotea,  and T.W. Obal.  1985.
     Distribution of contaminants  in clams and sediments  from the Huron-Erie
     corridor, I  -  PCBs and octachlorostyrene.  J. Great  Lakes Res., 11(3):
     275-289.
                                      166

-------
Rathbun, J.E., M.L. Gessner, V.E. Smith, D.M. Lemon,  D.J. Brokaw, M.A. Hoeft,
     W.L. Richardson, and K.R. Rygwelski.  1985.   Bioaccumulation of total
     PCBs and PCB homologs In caged clams,  channel  catfish and fathead
     minnows in the Monroe Harbor - River Raisin,  Michigan (1984).  Draft
     report.

Rathbun, J.E. and V.E. Smith.  1985.  Bioaccumulation of total PCBs and PCB
     homologs in clams in the River Raisin, Michigan  (1983).   Draft report.

Rice, C.P., D.S. White, M.S. Simmons, and R. Rossman.  1984.   Assessment of
     the effectiveness of the cleanup of the PCBs  from the south branch of
     the Shiawassee River:  field results.   Final  report to the Michigan
     Department of Natural Resources, Water Quality Division.  89 pp.

Robbins, J.  1980.  Sediments of southern Lake Huron; elemental composition
     and accumulation rates.  USEPA, Ecological Research Series
     #EPA-600/3-80-080.

Rygwelski, K.R.  1984.  Partitioning of toxic trace metals between solids and
     liquid phases in the Great Lakes,  in - Toxic Contaminants in the Great
     Lakes, J.O. Nriagu and M.S. Simmons (eds.)» John Wiley and Sons, Inc.,
     New York, pp. 321-333.

Smith, V.E., K.W. Lee, J.C. Filkins, K.W. Hartwell, K.R. Rygwelski, and J.M.
     Townsend.  1977.  Survey of chemical factors  in  Saginaw Bay (Lake
     Huron).  USEPA, Ecological Research Series #EPA-600/3-77-125.  pp. 143.

Smith, V.E.  1983.  Personal communication.  Cranbrook Institute of Science,
     Large Lakes Research Station, Grosse He, Michigan.

Smith, V.E., J.E. Rathbun, S.G. Rood, K.R.  Rygwelski, W.L. Richardson, and
     D.M. Dolan.  1985.  Distribution of contaminants in waters of Monroe
     Harbor (River Raisin), Michigan and adjacent  Lake Erie.   Final Project
     Report, Grant 810232, USEPA, Large Lakes Research Station, ERL-Duluth,
     Grosse lie, Michigan.  153 pp.

Snyder, D.E.  1976.  Terminologies for intervals of larval fish development.
     In - Great Lakes Fish Egg and Larvae Identification, J.  Borrmann (ed.),
     U.S. Department of the Interior, Fish and Wildlife Service, Washington,
     D.C., FWS/OBS-76/23, pp. 41-60.
Technicon Industrial Systems.  1971.
     Methods, Tarrytown, New York.

Technicon Industrial Systems.  1972.
     Methods, Tarrytown, New York.
Technicon Autoanalyzer II, Industrial
Technicon Autoanalyzer II, Industrial
U.S. Army Corps of Engineers.  1967.  Water quality data for the Lake Erie
     basin.  Lake Erie Wastewater Management Study, Buffalo, New York, pp.
     27-28.
                                     167

-------
U.S. Environmental Protection Agency.   1971.   Methods for chemical analysis
     of water and wastes.  USEPA,  Ecological  Research Series 116020-07/71.

U.S. Environmental Protection Agency.   1975.   Monroe, Michigan.  Report on
     the degree of pollution of bottom sediments.   1975 Harbor Sediment
     Sampling Program, April 9, 1975.   U.S.  Environmental Protection Agency,
     Region 5, Great Lakes Surveillance Branch,  Chicago, Illinois.  9 p.

U.S. Environmental Protection Agency.   1982.   Test methods for the evaluation
     of solid wastes.  Office of Solid Waste and Emergency Response
     Publication 1846, 2nd Edition, Method 3010, Washington, D.C.

U.S. Environmental Protection Agency.   1984.   Technical guidance manual for
     performing waste load allocations.  Office of Water Regulations and
     Standards Monitoring and Data Support Division, Water Quality Analysis
     Branch, USEPA, Washington, D.C.

U.S. Environmental Protection Agency.   1985.   Technical support document for
     water quality-based toxics control.  Office of Water Enforcement and
     Permits and Office of Water Regulations and Standards, USEPA, Ecological
     Research Series #EPA-440/4-85-032, Washington, D.C.  74 pp.

Veith, G.D. and D.W. Kuehl.  1980.  Unpublished data.  EPA/ERL-Duluth,
     Minnesota.

Voice, T.C. and W.J. Weber.  1983.  Sorption of hydrophic compounds by
     sediments, soils, and suspended solids - I and II.  Water Res., 17(10):
     1433-1451.

Vollenweider, R.A. (Ed.).  1969.  A Manual on Methods for Measuring Primary
     Production in Aquatic Environments.  Blackwell Scientific Publications,
     Oxford.  213 pp.

Young, T.C., J.V. DePinto, and T.W. Kipp.  1987.  Adsorption and desorption
     of Zn, Cu, and Cr by sediments from the Raisin River (Michigan).   J.
     Great Lakes Res., in press.

Zar, J.H.  1974.  Biostatistical Analysis.  Prentice-Hall,  Inc., Englewood
     Cliffs, New Jersey, pp. 109-113.

Ziegler, C.K. and W. Lick.  1986.  A numerical model of the resuspension,
     deposition, and transport of fine-grained sediments  in shallow water.
     Draft report.
                                      168

-------
                                  APPENDIX A

                         PCB MASS BALANCE MODELS - 1983
     No detailed project report on PCB models In Monroe Harbor exists as a
source for more Information on the topic.  The Interested reader may,
therefore, find the following additional  information on PCB modeling helpful.

     Model schematics based on Pritchard's general  estuarine model  are
presented in Figure A-l.  The two models  were applied to an upper reach of
the study site that included the area between Station 1 and Station 4.  The
first and second schematics represent the July and  fall surveys, 1983,
respectively.

     Definitions and terms used in the schematics follows:

     Subscript "L" - lower layer

     Subscript "U" - upper layer

             cu(P) * Concentration of pollutant in  point source discharge
                     (mass/volume)

        cu(n-l  n) * Concentration of pollutant in  upper water layer at the
              '      boundary between segments n-1  and n (mass/volume)

                En » Vertical exchange coefficient  between upper and lower
                     water layers of segment n ((length)vtime)

        Qii/n i  n\ * Net fl°w rate in upper water layer from segment n-1 to
          1   '  ;   n (volume/time)

             Qy(n) " Net upward or downward vertical flow between water
                     layers (volume/time)

             QU(P) - Flow rate of point source discharge to upper water
               v     layer (volume/time)

     Specific solutions to Pritchard's model were determined for the July,
September, and October surveys (1983). A mass balance equation was written
for each layer and for each given model condition in Figure A-l.  These
equations were solved simultaneously.  The resulting solutions were then
simplified by making appropriate assumptions.


                                     169

-------
SCHEMATIC A-l-A


Survey 1, July 1983  (upper  reach  from  Station  1 to 4)
         n-1
Station 1
CU(P)
Qu(P)
  \
Station 4      n+1
            / / / / / / x / '/

                        /
                        }
«"("•!. n)
  '   L_£)      A      -"u(n



               QV(n)
                                                         nufn»
                                            'Cu(n)       Qu(n,
                                                  J\
                                            *CL(n)
                     i. _
                                      L(n+l, n)

                                          1' ni
                              7/X / 7 / /  / r s ' r / S '/ S f S f S /r s 7
SCHEMATIC A-l-B


Survey 2 and 3, September 1983 and October 1983 (upper reach  from
Station 1 to 4)

                              CU(P)

                              Qu(P)
          n-1        Station 1   \       n               Stat^ion__4_  _ ntl
1
^u(n

^///x/,^/.
>*
/
-l.n) ^
-1, n)
? I
i
QV(n)

E
*Cu(n) >

\
*CL(n)
f / f s s f r /
n) Cu(n+l.
5u(n.+l,
. . . . i . .
1
L ( n , n
QL(n, n
s / / ' S / f s
n)
n)

fl)
*1)
x ^..^^^
      Figure A-l.  Pritchard's Models for River Raisin/Monroe Harbor.
                                  170

-------
     General assumptions used in all  of Pritchard's models are:
     - Steady-state condition (no changes in the system with respect to time)
     - Conservative constituent
     - Concentration of constituent is uniform within each layer of each
       segment
     - The constituent concentration  at the boundary between segments or
       layers is equal to the average concentrations in the two adjacent
       segments or layers.
JULY 1983 - Upper Reach (Station 1 to Station 4)
--Given the July 1983 flow conditions (Figure A-l-A),  the following mass
  balance equations can be written for any conservative substance:
Upper Layer Balance:
Wu(n-l,  n)l  Cu(n-l,  n)  +
           uation
Lower Layer Balance:
                   L(n+l, n) + ^ 
                    L(n)
Adding A-l and A-2 yields:
              u(n-l, n)
                                             u(n,
           [Qu(P)] Cu(P) +QL(n+l, n) CL(n+l, n)
                                     171
                                                                Equat1on A'3

-------
Conservation of hydraulic continuity requires that,


     QL(n+l, n) - Qu(n,  n+l) ' Qu(n-l,  n)  ' Qu(P)                E1uat1on A'4
Therefore, Equation A-3 can be rewritten


     [Qu(n-l, n)] Cu(n-l, n) ' [Qu(n,  n+l)] Cu(n,  n+l)          Equation A-5


         + Qu(P) Cu(P) + [Qu(n, n+l) " Qu(n-l, n)


         - Qu(P)] CL(n+l, n) =0
Qu(n, n+i) can now be calculated using any conservative tracer if the other
varia'bles are known.  Using Equation A-4, Q|_(n+l, n) can be calculated.  Now
that all of the flows are known in Equation A-3, the mass balance using this
simple input-output model can be computed for any constituent.  Positive
solutions suggest the modeled system is a sink.  Negative solutions suggest
the system is a source for the constituent.
SEPTEMBER AND OCTOBER 1983 - Upper Reach (Station 1 to Station 4)

--Given the September and October 1983 flow conditions (Figure A-l-B), the
  following mass balance equations can be written for any conservative
  substance:


Upper Layer Balance:


     [Qu(n-l, n)] Cu(n-l, n) + £Qu(n+l, n)l Cu(n+l, n)


                   Cu(P) + ^ 
         - QV(n) Cu(n)
                                     172

-------
Lower Layer Balance:
     [QV(n) Cu(n) + [En3 (Cu(n) " CL(n))
           [Q
       L(n,
                        L(n,
                                                          Equation A-7
Adding A-5 and A-6 yields:
     [Qu(n-l, n)] Cu(n-1, n) + [Qu(n+l, n)1 Cu(n+l, n)
   + tQu(P)] Cu(P) - [QL(n, n+l)lCL(n,
                                                   =0
                                                          Equation A-8
Conservation of hydraulic continuity requires that,
L(n,
                  Qu(n-l, n) + Qu(P) + Qu(n+l, n)
                                                          Equation A-9
Therefore, Equation A-8 can be rewritten
     tQu(n-l, n)1 Cu(n-l, n) + [Qu(n+l, n)1 Cu(n+l, n)


         * [Qu(P)] Cu(P) " [Qu(n-l, n) + Qu(P)
           Qu(n+l, n)] CL(n,
                                                          Equation A-10
Qu(n, n+1) can now &e calculated using any conservative tracer if the other
varia'bles are known.  Using Equation A-9, QL(n, n+1) can be calculated.  Now
that all of the flows are known in Equation A-8, the mass balance using this
simple input-output model can be computed for any constituent.  Positive
solutions suggest the modeled system is a sink for the constituent.  Negative
solutions suggest the system is a source for the constituent.
                                      173

-------
                                  APPENDIX B

                            DATA BASE DOCUMENTATION



B.I  PHYSICAL DATA

B.1.1  Flow Data

B.I.1.1  Flow and Water Chemistry Data From the Second Low Head Dam, Monroe--

     The City of Monroe routinely monitors river level and chemistry data at
the second low head dam.  Copies of the data for the period February 1983
through October 1984 were received through the Monroe Waste Water Treatment
Plant.  The sheets were coded and put on the MicroVAX II; data set -
[MONROE.FLOWJMHTEST.DAT.  A look-up table needed to convert river level to
flow is in file [MONROE.FLOW]MHFLOW.TAB.  A printout of the river flow and
chemistry data can be obtained by running program MHFLOW, from the Monroe
account, FLOW, subdirectory.

B.I.1.2  Waste Water Treatment Plant, WWTP, Instantaneous Secondary Flow--

     Even-hour secondary flow data for September 1983 through December 1983,
were received from the WWTP.  It was coded and put into computer file
[MONROE.FLOW]EVNHRFLO.DAT.

B.I.1.3  USGS Daily Flow Data--

     There is a USGS gage, 04176500, at Ida-Maybee Road.  Flow data from this
gage can be retrieved from the STORET FLOW file.  STORET is the EPA water
quality data base.

B.I.2  River Current Data

     River current data were collected by Wapora, Inc., a consulting firm,
during Surveys II and III.  The data were keypunched and entered into
computer files [MONROE.FLOW]VELCTY.MHZ and [MONROE.FLOW]VELCTY.MH3.

B.I.3  Wind Data

     Wind data reports were obtained from the Fermi II  (58944) and Detroit
Edison (58953) power plants for the time period April 1983 through December
1984.  The wind direction and speed were put into computer files
                                     174

-------
[MONROE.WIND3MHWINDD.DAT and [MONROE.WINDJMHWINDV.DAT.  The contents of
either file can be listed by running program WNDRPT.

B.I.4  Lake Level Data

     NOAA provided a tape containing lake level data for three stations,
Monroe (3087), Fermi (3090), and Toledo (3085), for February 1983 through
October 1984.  The data were transferred to file [MONROE.LAKE]NOAAIN.DAT.
PLTSTG will plot one month of data for one station.

B.I.5  Hvdrolab® Profile Data

     Hydrolab* data were collected during Survey I and for the day prior to
Survey IV by in-house staff.  Wapora, Inc. collected Hydrolab® data during
Surveys II and III.  This Hydrolab® data consisted of temperature, pH,
dissolved oxygen, and conductivity.  These data were collected from stations
located on transects across the river.  Three positions on a transect were
defined as the north, central, and south channels.

     The dates corresponding to surveys are:

            830712-830716 - Survey I
            830913-830917 - Survey II
            831025-831028 - Survey III
            840403 - Day prior to Survey IV.

     All  of the data are in file [MONROE.TRAN]MHTRAN.SRT.  In order to
generate plots of parametric data for a given channel  position on any
transect along a given river reach, the data set containing all of the data
were copied and manipulated to make a working file in a direct-access format.
First, DELZER copied data from MHTRAN.SRT that were collected from a
"channel" position on a transect.  The results were put into
[MONROE.TRANJMHTRAN.USE.  Next, MHDIR converted the data to direct access
file [MONROE.TRAN1MHTRAN.DAT.  Program POINT accessed the file and set
pointers that facilitated quick access.   Finally,  program TRANP produced
channel plots.

B.I.6  Transect Bottom Depth Data

     The depth of the river was measured along transects on August 7, 1984.
The data were coded and put into file [MONROE.TRAN]TRANP.M84.

B.I.7  Precipitation Data

     The WWTP supplied precipitation data for January 1983 - October 1984.
The information was keypunched and put into file [MONROE.WIND]MHPRECIP.DAT.
                                    175

-------
B.2  WATER QUALITY DATA

B.2.1  Grosse He Data System (GIDS1

     Water quality data are stored In the Grosse He Data System,  GIDS,
located on the MicroVAX II, account GIDS.  Monroe Harbor data are  divided
into four subsystems.

B.2.1.1  Monroe Section--

     The Monroe section contains water quality data for Surveys I-X.  It also
contains data collected and analyzed during non-survey periods. Table B.I
provides a list of parameters for which data were collected.  Table B.2 lists
the stations sampled by survey.

B.2.1.2  MHSPEC Section--

     This includes special study data such as turning basin samples and WWTP
effluent dilutions.  The following is a list of samples in the MHSPEC
subsection.
            Sample ID

            MH0001WTOOE1
            MH4006WT04F1
            MH3513WT07NA
            MH3513WT07NB
            MH3513WT07NC
            MH3513WT07ND
            MH3517WT07NA
            MH3517WT07NB
            MH3517WT07NC
            MH3517WT07ND
            MH0106WT07E1
            MH0107WT07E1
            MH0108WT07E1
            MH0109WT07E1
            MH0110WT07E1
            MH0111WT07E1
            MH0112WT07E1
            MH0113WT07E1
            MH0114WT07E1
            MH0115WT07E1
            MH0116WT07E1
            MH0117WT11E1
            MH0118WT11E1
            MH0119WT11E1
            MH0120WT25E1
            MH0121WT25E1
            MH1001WTTBN1
            MH2033WTTBN1
840404
840404
831026
831026
831026
831026
831028
831028
831028
831028
840223
840223
840223
840223
840223
840223
840223
840223
840223
840223
840223
840223
840223
840223
840223
840223
830716
830914
Day 1
Day
                         With  unfiltered
                         Lake  Erie
   Description

 Spring water  in  glass  bottle
 Fathead minnow waste water
'50% WWTP  EFFL H
 10%
 5%
.1%
 50%
 10%
 5%
.IX
 100%
 50%
 10%            I  With  20m Lake
 5%             fErie  filtrate
 2%
 1%
 50%
 10%
 5%
 2%
 1%
 20y filtered  lake control
 20p filtered  lake control
 Unfiltered lake  control
 Lake Erie intake control
 Lake Erie intake control
 Turning  Basin
 Turning  Basin
                       l»With unfiltered
                         Lake Erie
                                     176

-------
TABLE B.I.  MONROE HARBOR PARAMETERS AND CODES
CODE
10
299
78
400
95
410
50060
608
613
900
940
530
520
1030
1034
1040
1042
1090
1092
110000
110001
110002
110003
110004
110005
110006
110007
110008
110009
110010
210000
210001
210002
210003
210004
210005
210006
210007
210008
210009
210010
DESCRIPTION
Temperature °C
Dissolved Oxygen mg/L
Secchi Depth m
pH S.U.
Specific Conductance yS/cm
Alkalinity mg/L (as CaC03)
Total Residual Chlorine mg/L
Dissolved Ammonia mg/L
Nitrite mg/L
Hardness mg/L (as CaC03)
Chloride mg/L
Suspended Solids mg/L
Volatile Suspended Solids mg/L
Dissolved Cr yg/L
Total Cr ug/L
Dissolved Cu ug/L
Total Cu yg/L
Dissolved Zn yg/L
Total Zn yg/L
Total PCB ng/L
Total PCB Homolog 01 ng/L
Total PCB Homolog 02 ng/L
Total PCB Homolog 03 ng/L
Total PCB Homolog 04 ng/L
Total PCB Homolog 05 ng/L
Total PCB Homolog 06 ng/L
Total PCB Homolog 07 ng/L
Total PCB Homolog 08 ng/L
Total PCB Homolog 09 ng/L
Total PCB Homolog 10 ng/L
Dissolved PCB ng/L
Dissolved PCB Homolog 01 ng/L
Dissolved PCB Homolog 02 ng/L
Dissolved PCB Homolog 03 ng/L
Dissolved PCB Homolog 04 ng/L
Dissolved PCB Homolog 05 ng/L
Dissolved PCB Homolog 06 ng/L
Dissolved PCB Homolog 07 ng/L
Dissolved PCB Homolog 08 ng/L
Dissolved PCB Homolog 09 ng/L
Dissolved PCB Homolog 10 ng/L
                     177

-------
TABLE B.I.  MONROE HARBOR PARAMETERS AND CODES (CONT'D.)
   CODE                       DESCRIPTION
   310000                Particulate  PCB   ng/L
   310001                Particulate  PCB  Homolog  01   ng/L
   310002                Particulate  PCB  Homolog  02   ng/L
   310003                Particulate  PCB  Homolog  03   ng/L
   310004                Particulate  PCB  Homolog  04   ng/L
   310005                Particulate  PCB  Homolog  05   ng/L
   310006                Particulate  PCB  Homolog  06   ng/L
   310007                Particulate  PCB  Homolog  07   ng/L
   310008                Particulate  PCB  Homolog  08   ng/L
   310009                Particulate  PCB  Homolog  09   ng/L
   310010                Particulate  PCB  Homolog  10   ng/L
                          178

-------
TABLE B.2.  MONROE HARBOR STATIONS SAMPLED BY SURVEY

STATION
01
02
03
04
05
06
07
08
09
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
28
29
30
31
32
33
34
35
36
38
39
40
41
42
43
44
45
46
47
YEAR
MONTH/- 7/12
DAY 7/17
SURVEY 1
X
X
X
X
X
X
X
X
X
X
X


































1983
9/13
9/18
2
X

X
X
X
X
X
X
X
X
X



































10/25
10/28
3
X

X
X
X
X
X
X
X
X
X

X
X
X
X
X
X




























4/4
4
X


X


X
X


X







X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X








5/9
5
X


X


X
X


X













X
X

X








X
X
X
X
X
X
X
X
X
1984
5/30 6/12 7/10 9/25 8/1
6 7 8 9 10
X X X X X


X X X X X




















X X
X X X X X

X X XXX








X X X X X
X X X X







                        179

-------
     Parameters analyzed are those  from the  Monroe  section  less  the  organic
information.
                    Parameter
            Code

            10  "
            299
            78
            400
            95
            410
            50060
            608
            613
            900
            940
            530
            520
            1030
            1034
            1040
            1042
            1090
            1092
B.2.1.3  MHTCHM Section--
  Description

Temperature  °C
Dissolved Oxygen  mg/L
Secchi Depth  m
pH  S.U.
Specific Conductance  uS/cm
Alkalinity  mg/L (as CaC03)
Total Residual Chlorine  mg/L
Dissolved Ammonia  mg/L
Nitrite  mg/L
Hardness  mg/L (as CaC03)
Chloride  mg/L
Suspended Solids  mg/L
Volatile Suspended Solids  mg/L
Dissolved Cr  yg/L
Total Cr  ug/L
Dissolved Cu  yg/L
Total Cu  ug/L
Dissolved In  pg/L
Total Zn  pg/L
     Prior to using stored water for toxicity tests,  chemical  measurements
were taken.  The parameters measured were:
            Code

            95
            299
            410
            400
            530
                   Parameter
  Description

Specific Conductivity  jjS/cm
Dissolved Oxygen  mg/L
Alkalinity  mg/L (as CaC03)
P.M.  S.U.
Suspended Solids  mg/L
B.2.1.4  SEDCHM Section--

     Sediment toxicity studies were performed during Surveys IX, XI, and XII,
This subsection contains chemistry data for these samples.

B.2.2  Oroanics Analytical Data Base

     An analytical data base was developed for the organic data located on
the MicroVAX II, account MHOPAK.  Each sample was assigned a process file
name when it was analyzed by gas chromatography.  To retrieve information

                                     180

-------
from the data base, a file is created containing process file name and,
optionally, a line number.  The user can retrieve a list of samples and
process file names by running program DMPLAB.  Program MHOPAK will retrieve
data for the process files in the user created file.  Raw data or statistics
can be reported.  All organic contaminants data for Monroe Harbor are in this
data base.

B.2.3  Point Source Data

     Data from point sources discharging to the River Raisin in the study
area were obtained from the Michigan Department of Natural Resources.  Daily
data for January 1983 through October 1984 were entered in STORE! for the
Monroe WWTP, the Ford Motor Company and Union Camp facilities.  The STORET
retrieval which lists all of the data is in a data file on the NCC-IBM
computer DVGF528.MONROE.STORET.RETJCL.


B.3  BIOTA DATA

B.3.1  In-House Ceriodaphnia and Fathead Minnow Data

     Toxicity data for inhouse Ceriodaphnia bioassays were entered into
temporary data sets on the MicroVAX II.  A bootstrapping program,
[BSVAR.MONROE]BSVAR, was run to calculate percent survival using two methods.
Results from the printout were coded and combined with manually calculated
fathead minnow results.  The results were put into two files on NCC-IBM and
can be processed using SAS:

     1.  The data set DVGF528.SAS.MHCERIO.TOXDATA contains water toxicity
         test results.

     2.  The data set DVGF528.SAS.SEDIMENT.TOXDATA contains sediment
         toxicity test results.

     In this form, the data sets are now ready for SAS analysis.  The data
sets DVGF528.SASJCL.TOXDATA and DVGF528.SASJCL.SEDTOX contain the SAS
statements needed to read toxicity test results into SAS data sets.

     The corresponding chemistry data for water (DVGF528.SAS.GIDSDATA) and
sediment (DVGF528.SEDTOX.DATA) can be read into SAS data sets by using the
files DVGF528.SAS.MHMERGE, and DVGF528.SASJCL.METTOX, respectively.

B.3.2  Ohio State University Larval Fish Data

     Ohio State University larval fish data were received on a data tape.
Two files were transferred to the EPA National Computer Center, NCC-IBM:
KPMF528.0SU.LARVAL.FISHI and KPMF528.0SU.LARVAL.FISH2.
                                     181

-------
B.3.3  University of Minnesota Biota Data

B.3.3.1  CeHodaohnia Bioassays--

     Because of the quantity of data to be processed, a preprocessor,
[BSVAR.MONROE]PBSVAR, was written to prepare data for the bootstrapping
program.  Data were keypunched directly from the log sheets.  BSVAR was
modified to calculate an R statistic and to append the results to the end of
file [BSVAR.MONROE]CERIO.SAV.  The new version of the program was named
RBSVAR.  Once all of the data were processed, the output file CERIO.SAV, was
transferred for use on SAS, Filename - DVGF528.MCCERIO.TOXDATA.

B.3.3.2  Phytoplankton Productivity Bioassays--

     Data were keypunched and put in a file for use with SAS, Filename -
DVGF528.MCNAUGHT.GRAZE.

B.3.3.3  Zooplankton Grazing Bioassays--

     Data were put on NCC-IBM for use with SAS, Filename -
DVGF528.MCNAUGHT.GRAZE.

B.3.3.4  Bacterial Productivity Bioassays--

     Data were put on NCC-IBM for use with SAS, Filename -
DVGF528.SAS.BACTERIA which can be read using DVGF528.SASJCL.BACTERIA.

B.3.4  Other Inhouse Biological Data

     All field data were put into data file  [MONROE.BIO]BIOFLD.DAT.  A
formatted report of all  data can be generated by running BIOFLD.  Program
BIODAT will print a report for specific sample ID codes that were input to a
file prior to program execution.  Organic data for fish and clam samples can
be obtained through the organic data base.


B.4  SEDIMENT DATA

     The physical and chemical data for Monroe Harbor sediment samples were
combined with total PCB and homolog data.  These results are in file
[MONROE.SED]COMBNE.OUT.   A formatted listing of the data can be made by
running program SEDRPT.   Pesticide and congener level PCB data can be
retrieved fro* the organic data base.


B.5  OTHER DATA

     Heidelberg College did a general pesticide/herbicide analysis of water
samples from Stations 1  and 4.  These pesticide results from 1984 are in file
[MONORE.MISCJMHPEST.DAT.
                                     182

-------