United States
            Environmental Protection
            Agency
Office of Research and
Development
Washington DC 20460
EPA/600/R-93/183
September 1993
oEPA      Guidance Manual

            Bedded Sediment
            Bioaccumulation Tests

-------
                                              EPA/600/R-93/183
                                              September 1993
  GUIDANCE  MANUAL:  BEDDED SEDIMENT BIOACCUMULATION TESTS
      Henry Lee II1,  Bruce L. Boese1, Judy  Pelletier2,

 Martha  Winsor2,  David T. Specht1, and Robert  C.  Randall1
 Bioaccumulation Team
 Pacific Ecosystem Branch
 Environmental Research Laboratory -  Narragansett
 U.S. Environmental  Protection Agency
 Hatfield Marine Science Center
 Newport, Oregon 97365

2AScI Corporation
 Hatfield Marine Science Center
 Newport, Oregon  97365
                       September,  1989

                ERL-N Contribution No. Nlll
                                           OS Printed on Recycled Paper

-------
                             PREFACE
     The recommendations in this document are based on our
experience and best scientific judgment.  It is our hope that the
procedures suggested here will aid scientists in conducting
"routine" and experimental sediment bioaccumulation tests, as
well as aid regulators in determining when bioaccumulation tests
are needed, in evaluating the QA/QC procedures of such tests, and
in interpreting the results.  The recommendations made here,
however, do not constitute an official policy or standard
procedure by the U.S. Environmental Protection Agency.  Mention
of trade names or commercial products does not constitute
endorsement or recommendation of use by the Environmental
Protection Agency.

-------
         1993  UPDATE TO "BIOACCUMULATION GUIDANCE MANUAL"

     In the four years since we  wrote this document, the "Guidance
Manual"  has  become  the  standard for assessing  dredge materials
proposed for ocean disposal  (U.S.  EPA/ACE, 1991), dredge materials
proposed  for inland disposal  (U.S.  EPA/ACE,  1993  draft),  and
several  scientific  studies.   Fortunately,  the procedures in this
document appear  to have stood the test  of  time  relatively well.
Nonetheless, there are several modifications.  These modifications,
as  well  as  procedures  for  freshwater  organisms,  are  being
incorporated into an ASTM guidance document that should be out in
1994 or 1995.
I. VERSIONS OF GUIDANCE MANUAL

     The 1989 version of the Guidance Manual was not published and
was distributed only as  photocopies.   In the  1989 version, Table
IV-3A in the Appendix IV was in error.   In  1991, the Army Corps of
Engineers agreed to photocopy and distribute the Manual in support
of the new  "Green Book" (U.S. EPA/ACE,  1991).   This  version was
single spaced  and contained  a  corrected Table IV-3A.   The 1993
version contains  the corrected  Table IV-3A,  corrects  an error in
Figure IV-1, and makes a few editorial corrections.


II. NON-INDIGENOUS TEST SPECIES

     Over the last few years,  there has been a  growing awareness of
the ecological and  economic damage  caused  by introduced species.
Because  both  east  and  west  coast species  are  often used  in
bioaccumulation tests,  there is a  real  potential of  introducing
bioaccumulation test species or associated fauna and flora  (e.g.,
pathogens, algae used in transporting the worms) .   Any user of this
document needs  to understand that  it  is their responsibility to
assure  that non-indigenous  species are not  released  into  the
environment.

     The general procedure  to contain non-indigenous species is to
collect  and  then poison  all  water,   sediment,  organisms,  and
associated  packing  materials   (e.g.,   algae,  sediment)  before
disposal.   Chlorine bleach  can  be used  as   a  poison.   Double
containment system is used  to keep any spillage from going down the
drain.  Guidance on procedures used  in  toxicity tests can be found
in Appendix B of  DeWitt  et al.  (1992).   Permits  to import, hold,
and use non-indigenous  species  may  be  required by state Fish and
Game departments or other state agencies.  Flow-through tests can
generate large quantities  of water, and the  researcher needs to
plan on having sufficient storage facilities.


III.  SEDIMENT RENEWAL

     We no longer  recommend the addition of supplemental sediment


                               i i i

-------
during 28-day exposures  (Chapter X) .  In tests longer than 28 days,
sediment  addition  is  still   an   option  to  maintain  pollutant
concentrations and food  quality/quantity as  is periodic renewal of
all the test sediment.

IV.  DURATION OF TEST

     Our recent work on  field  sediments  contaminated with DDT, DDT
metabolites,  and   dieldrin   indicates   that  28   days   may  be
insufficient to obtain  80% of  steady-state  tissue residues for a
number of higher Kow compounds  (see Boese and Lee,  1992 for summary
of unpublished  data  and  an  expanded review of  the percent  of
steady-state obtained in  28  days).   These  results  emphasize the
importance of conducting tests of at least 28 days duration.

     In tests where  it  is critical to measure the residue within
_>80% of steady-state, it may  be  necessary to conduct  longer term
tests or at  least to  expose a few individuals for a longer duration
to test  whether steady-state was achieved.  In  screening tests
where is important not to underestimate  the  residues, the residues
obtained from 28-day tests could be multiplied by a "steady-state
correction factor".   The  "steady-state  correction factor" is the
reciprocal of the decimal fraction of the amount  of steady-state
tissue residue obtained  after  28  days.   For  example, if the tissue
residue after 28 days is 0.33  of  the residue at steady state, the
correction  factor  would be  I/.33  = 3.    These  correction values
would be  obtained from previously conducted lab  studies  (e.g.,
Table IV-1 of Guidance Manual; Boese and Lee, 1992).


V.  GUT PURGING

     As far as we are aware,  the  effects  of gut  purging have not
yet been assessed quantitatively.  Therefore, the recommendations
in Chapter X and Appendix X-l remain the best available guidance.


VI. TERMINOLOGY AND COMPARISON OF BIOACCUMULATION MODELS

     Unfortunately, the  terminology used in  bioaccumulation  is not
standardized and has continued to  evolve, or at least change, since
the  Guidance  Manual  was  written.    The  "thermodynamic-based
bioaccumulation" model referred to in Appendix 1-1 of the Guidance
Manual   is   better   termed   the   "equilibrium   partitioning
bioaccumulation"  (EqP)  bioaccumulation model.    Since the term
"accumulation factor" (AF)  of the EqP bioaccumulation model first
appeared in the peer-reviewed  literature,  the term "biota sediment
accumulation factor"  (BSAF)  has  been used  by some authors.   It
appears that the two terms are interchangeable,  though the reader
is cautioned to check the units.  Though it was not emphasized in
Appendix 1-1 of the Guidance Manual, AFs have units of gC/gL (see
Lee,  1992) and it  is critical  that both the tissue pollutant and
lipid concentrations be  measured in or converted to the same units
                                IV

-------
 (i.e., both wet or both dry) .  Additionally,  the sediment pollutant
concentrations and the total organic carbon  (TOO concentration of
the sediment need to be in consistent units.

     The  guidance  manual  for  open water disposal  of  dredge
materials    (U.S.    EPA/ACE,    1991)    uses   the   "theoretical
bioaccumulation potential"  (TBP).   The  TBP  is  the application of
the EqP  model using  an  accumulation factor of 4.0.   Again the
reader is cautioned  to use consistent units for lipid and tissue
residue and for sediment and lipid concentrations.

     We  now recommend the  use of  "ks"  rather  than  "k^"  for the
first-order  uptake   rate   coefficient   from  sediment  to  help
distinguish it from uptake from water (see Lee,  1991).  It is also
recommended to consider ks a coefficient rather than  a  constant.

     More thorough comparisons  of bioaccumulation models than given
in Appendix 1-1 can  be found in Boese and Lee   (1992),  Landrum et
al. (1992), and Lee  (1992).
VII. FIELD VALIDATION

     The  bioaccumulation  test  methodology  is  presently  being
evaluated by  comparing residues in  field-captured Macoma nasuta
with residues  in  laboratory-exposed  M.  nasuta and Nereis virens.
The initial data for sum DDT with M. nasuta indicates that  only 34%
of the steady-state residue was obtained within 28  days at the one
station where  a long-term  (90  day)  test was conducted.    However,
when the 28-day residues from other stations were  multiplied by the
steady-state correction factor,  the lab residues were within 3-fold
of residues  in field-captured M. nasuta  in 7 of  8  cases.   This
comparison  included 6   sites that  varied by  almost  3 orders-of-
magnitude in sum DDT sediment concentrations,  indicating  that the
procedures work over a  wide range of sediment  contamination.


VIII. REFERENCES

Boese,  B. and  H.  Lee II.   1992.   Synthesis of Methods to Predict
Bioaccumulation   of  Sediment   Pollutants.     U.S.  EPA  Report.
ERL-Narragansett No. N232.   87 pp.

DeWitt, T.,  M. Redmond, J.  Sewall,  and R.  Swartz.   1992.  Develop-
ment of a  Chronic  Sediment  Toxicity  Test  for  Marine  Benthic
Amphipods.   Chesapeake  Bay Program/TRS 89/93.

Landrum, P., H. Lee II, M.  Lydy.  1992.  Toxicokinetics in aquatic
systems: Model comparisons  and  use  in hazard assessment.   Environ.
Toxicol. Chem. 11:1709-1725.

-------
Lee II, H.   1991.   A  clam's  eye view of  the  bioavailability of
sediment-associated  pollutants.     in   Organic  Substances  and
Sediments in Water.  Vol. Ill: Biological.  R.  Baker (ed.), Lewis
Publ.,  N.Y.,  pp 73-93.

Lee  II,   H.   1992.     Models,  muddles,   and  mud:   Predicting
bioaccumulation of  sediment-associated  pollutants.   in Sediment
Toxicity Assessment.  G.  Allen Burton (ed.), Lewis  Publ., N.Y.  pp
267-291.

U.S. EPA/ACE.  1991.    Evaluation of  Dredged  Material  for Ocean
Disposal  (Testing  Manual).    U.S.   EPA  Report  No.  503/8/91/001,
Office of Marine and Estuarine Protection, Washington,  DC.

U.S. EPA/ACE.  1993.  Evaluation of Dredged Material Proposed for
Discharge  in Inland and Near  Coastal  Waters  -  Testing  Manual
(Draft).

-------
                        ACKNOWLEDGEMENTS
     We greatly appreciate the time and effort Jack H. Gentile,
Peter F. Landrum, Norman Rubinstein, and Richard K. Peddicord
spent in reviewing a document of this magnitude.  We also
appreciate the comments of James Heltshe and Donald J. Reish on
specific chapters.  All reviewers offered important comments and
suggestions,  and the document benefited from their insights.  We
also appreciate Richard Latimer's initial prodding for us to
initiate this task.  George Ditsworth, and Donald Schults allowed
us to use their unpublished data.  Karl Rukavina, Richard Lapan,
and Tricia Lawson offered numerous editorial comments and helped
assure there were no typoos left in the document.  The data on
the sediment pollutant concentrations in Yaquina Bay were
originally collected from research partially funded by the Puget
Sound Program of Region X of the U. S. Environmental Protection
Agency.  Environmental Research Laboratory-Narragansett
Contribution No. Nlll.
                             VI 1

-------
                        TABLE OF CONTENTS


PREFACE 	  i i

1993 UPDATE 	i i i

ACKNOWLEDGEMENTS 	vii

LIST OF TABLES 	xi i

LIST OF FIGURES 	Xl'ii

ABSTRACT 	  xi V

CHAPTER I:  INTRODUCTION 	   1

CHAPTER II: CONTROL VERSUS REFERENCE SEDIMENT  	   6
   A. DEFINITION OF CONTROL AND REFERENCE SEDIMENT  	   6
   B. CRITERIA FOR CONTROL AND REFERENCE SEDIMENTS  	  10
   C. STANDARD REFERENCE SEDIMENT  	  13

CHAPTER III. PRINCIPLES OF EXPERIMENTAL DESIGN 	  16
   A. OBJECTIVES AND DEFINITIONS 	  16
   B. HYPOTHESES TESTING 	  18
   C. REPLICATION 	  20
   D. RANDOMIZATION 	  27
   E. PSEUDOREPLICATION 	  27
   F. AVOIDING OR REDUCING PSEUDOREPLICATION  	  30
   G. COMPOSITING SAMPLES 	  33

CHAPTER IV: TEST DURATION AND SAMPLING SCHEDULES  	  36
   A. STANDARD 28-DAY BIOACCUMULATION TEST  	  36
   B. LONG-TERM UPTAKE TESTS 	  41
   C. ESTIMATING STEADY-STATE TISSUE RESIDUES  FROM  UPTAKE
      AND DEPURATION RATES 	  44

CHAPTER V:  SEDIMENT COLLECTION, HOMOGENIZATION, MANIPULATION,
           AND STORAGE 	  48
   A. SEDIMENT COLLECTION AND TRANSPORT 	  48
   B. SEDIMENT SPIKING AND MANIPULATION 	  54
   C. LABORATORY SEDIMENT STORAGE  	  54
   D. SEDIMENT PREPARATION AND HOMOGENIZATION  	  55

CHAPTER VI: SEDIMENT CHARACTERIZATION 	  58
   A. GRAIN SIZE 	  58
   B. TOTAL SOLIDS CONTENT 	  60
   C. ORGANIC CARBON 	  61
   D. ADDITIONAL SEDIMENT CHARACTERISTICS 	  63
   E. INTERSTITIAL WATER	  65
                               IX

-------
CHAPTER VII:  ORGANISM SELECTION  	  69
   A. SELECTION CRITERIA 	  69
   B. RECOMMENDED SPECIES 	  70
   C. NUMBER OF SPECIES TESTED AND MULTIPLE
      SPECIES TESTS 	  72

CHAPTER VIII: ORGANISM COLLECTION, MAINTENANCE, TRANSPORT,
              AND ACCEPTABILITY  	  74
   A. COLLECTION AND TRANSPORT 	  74
   B. CULTURING AND PURCHASING TEST ORGANISMS  	  77
   C. PRE-EXPERIMENTAL MAINTENANCE 	  78
   D. ORGANISM ACCEPTABILITY AND BACKGROUND
      CONTAMINANT LEVELS 	  82

CHAPTER IX: SEDIMENT EXPOSURE SYSTEMS  	  85
     A. SYSTEM REQUIREMENTS 	  85
     B. EXPOSURE SYSTEM DESIGN 	  91
     C. MULTIPLE SPECIES EXPOSURE CHAMBERS  	  95

CHAPTER X: EXPERIMENTAL INITIATION, MAINTENANCE,
           AND SAMPLING 	  97
   A. EXPERIMENTAL INITIATION AND MAINTENANCE  	  97
   B. SCHEDULE FOR ABIOTIC AND BIOTIC  POLLUTANT SAMPLES  ....  102
   C. METHODS OF BIOTIC SAMPLING  	  105
   D. GUT PURGING 	  106
   E. ACCEPTABLE LEVELS OF MORTALITY  	  Ill
   F. CHAIN OF CUSTODY	  112

CHAPTER XI: POLLUTANT AND LIPID ANALYSIS  	  114
   A. POLLUTANT ANALYSIS 	  114
   B. LIPID ANALYSIS 	  118
   C. SAMPLE STORAGE 	  121
   D. REPORTING OF RESULTS 	  122

CHAPTER XII:  STATISTICAL ANALYSES  	  123
   A. TESTS FOR NORMALITY AND HOMOGENEITY OF VARIANCES  	  124
   B. PAIRWISE COMPARISONS 	  126
   C. MULTIPLE COMPARISONS 	  131
   D. INTERPRETATION OF COMPARISONS OF TISSUE  RESIDUES  	  133
   E. ADDITIONAL ANALYSES 	  134

CHAPTER XIII: REGULATORY STRATEGIES FOR USE OF
              BIOACCUMULATION DATA 	  137
   A. NO FURTHER DEGRADATION 	  138
   B. TISSUE RESIDUE EFFECTS 	  139
   C. WATER QUALITY CRITERION TISSUE LEVEL APPROACH  	  140
   D. FDA ACTION LIMITS	  142
   E. HUMAN HEALTH RISK ASSOCIATED WITH SHELLFISH  	  143
   F. TROPHIC TRANSFER OF POLLUTANTS INTO PELAGIC
      FOOD WEBS 	„.......„„„...„	  144

-------
                            APPENDICES
                (Numbered sequentially by Chapter)

APPENDIX 1-1: ADDITIONAL METHODS TO  PREDICT  BIOACCUMULATION..147
   A. FIELD COLLECTION  	  147
   B. BIOACCUMULATION FACTORS AND ACCUMULATION FACTORS  	  149
   C. TOXICOKINETIC BIOACCUMULATION  MODELS  	  153

APPENDIX 1-2: SAFETY AND WASTE DISPOSAL  	  157
   A. PERSONNEL SAFETY  	  157
   B. HAZARDOUS WASTE DISPOSAL 	  157

APPENDIX III-l: DETERMINING NUMBER OF REPLICATES  	  159

APPENDIX IV-1: ADEQUACY OF 10-DAY AND 28-DAY EXPOSURES  	  164

APPENDIX IV-2: ALTERNATE TEST DESIGNS  	  170
   A. SHORT-TERM TESTS  	  170
   B. ESTIMATING STEADY-STATE FROM UPTAKE RATES  	  170
   C. GROWTH DILUTION 	  171

APPENDIX IV-3: CALCULATION OF TIME TO STEADY-STATE  	  174

APPENDIX V-l: TECHNIQUES FOR SEDIMENT MANIPULATION  	  178
   A. SEDIMENT SPIKING  	  178
   B. INCREASING SEDIMENT TOC 	  182
   C. DECREASING SEDIMENT POLLUTANT  AND/OR
      TOC CONCENTRATIONS 	  185

APPENDIX VII-1: SELECTION CRITERIA FOR TEST  SPECIES  	  187
   A. INDIGENOUS VERSUS SURROGATE SPECIES 	  187
   B. REQUIRED CRITERIA  	  188
   C. DESIRABLE CHARACTERISTICS	  190
   D. RECOMMENDED AND SECONDARY SPECIES  	  196

APPENDIX VIII-1: SOURCES FOR TEST ORGANISMS  	  198

APPENDIX IX-1: SPECIAL PURPOSE EXPOSURE  CHAMBERS  	  199
   A. CLAMBOX 	  199
   B. WORMTUBES 	  199
   C. SEDIMENT RESUSPENSION SYSTEMS  	  201

APPENDIX X-l: ADDITIONAL TECHNIQUES  FOR  CORRECTING FOR
              GUT SEDIMENT 	  204
   A. MODIFICATIONS TO 24-HOUR PURGE AND DISSECTION  	  204
   B. CALCULATING POLLUTANT MASS OF  GUT  SEDIMENT  	  205
   C. USE OF CONSERVATIVE TRACE ELEMENTS 	  206

GLOSSARY: 	  208

BIBLIOGRAPHY  	  213

-------
                         LIST OF TABLES
          (Numbered sequentially by Chapter or Appendix)

TABLE II-1: REPRESENTATIVE CONTROL SEDIMENT
            CONCENTRATIONS 	  11

TABLE III-l: RANGES OF COEFFICIENTS OF VARIATIONS  (CV) FOR
             VARIOUS ORGANISMS AND POLLUTANTS  	  24

TABLE IV-1: PERCENT OF STEADY-STATE TISSUE RESIDUE
            OBTAINED AFTER 10-DAY AND 28-DAY EXPOSURES  	  38

TABLE VI-1: WENTWORTH GRADE CLASSIFICATION OF  SEDIMENTS  ....  59

TABLE VI-2: RATIOS FOR CONVERTING LOSS ON IGNITION  (LOI)
            TO TOTAL ORGANIC CARBON  (TOC)  	  64

TABLE VII-1: PERTINENT CHARACTERISTICS OF TEST SPECIES  	  71

TABLE VIII-1 REPRESENTATIVE CONTROL ORGANISM TISSUE
             RESIDUES 	  84

TABLE X-l: ERRORS ASSOCIATED WITH GUT SEDIMENT/PURGING  	  107

TABLE X-2: DEPURATION LOSS OF POLLUTANTS DURING  24 AND
           72 HOUR GUT PURGES 	  110

TABLE XI-1: U.S. EPA CONTRACT LABORATORY PROGRAM
            QUANTITATION LIMITS FOR WATER AND  SEDIMENT WITH
            ESTIMATES FOR TISSUE MATRICES  	  116

TABLE XI-2:  PSDDA LOW LIMITS OF DETECTION FOR WATER,
             SEDIMENT AND TISSUE MATRICES  	  117

TABLE XII-1: SUMMARY OF STATISTICAL ANALYSES  	  129

TABLE XII-2: EXAMPLES OF ANALYSES AND INTERPRETATION
             OF RESULTS 	  135
                       TABLES IN APPENDICES

TABLE IV-1A: INFORMATION GAINED AND REQUIREMENTS  OF  DIFFERENT
             APPROACHES TO ESTIMATING BENTHIC  TISSUE
             RESIDUES  	  167

TABLE IV-3A: ESTIMATED TIME  TO OBTAIN 95%  OF
             STEADY-STATE TISSUE RESIDUE  	  176
                               XI 1

-------
                         LIST OF FIGURES
          (Numbered  sequentially by Chapter or Appendix)


FIGURE 1-1: STEPS IN CONDUCTING SEDIMENT BIOACCUMULATION
            TESTS 	   4

FIGURE II-1: OVERLAP OF CONFIDENCE INTERVALS  IN ORGANISMS
             EXPOSED TO ACCEPTABLE AND UNACCEPTABLE
             REFERENCE SEDIMENTS 	   9

FIGURE III-l: COEFFICIENT OF VARIATION VS. SAMPLE SIZE FOR
              VARIOUS MINIMUM DETECTABLE DIFFERENCES  	  23

FIGURE III-2: MINIMUM DIFFERENCE DETECTABLE BETWEEN
              TREATMENTS FOR A SPECIFIC COEFFICIENT OF
              VARIATION AND SAMPLE SIZE 	  26

FIGURE III-3: RANDOM AND PSEUDORANDOM REPLICATION SCHEMES  ..  29

FIGURE IV-1: TYPICAL UPTAKE-DEPURATION CURVE  	  46

FIGURE IX-1: REPRESENTATIVE SEDIMENT EXPOSURE SYSTEM  	  93

FIGURE XII-1: SAMPLING SCHEMES FOR COMPARISON-WISE VS.
              EXPERIMENT-WISE ERROR RATES  	  132

FIGURE XIII-1: POSSIBLE REGULATORY STRATEGY FOR HUMAN
               HEALTH CRITERIA IN ASSESSING SEDIMENT
               CONTAMINATION 	  146


                      FIGURES IN APPENDICES

FIGURE V-1A: EFFECT OF SEDIMENT SELECTION ON  INGESTED
             DOSE IN NATURAL AND ORGANICALLY  ENRICHED
             SEDIMENT 	  183

FIGURE IX-1A: CLAMBOX EXPOSURE CHAMBER 	  200

FIGURE IX-IB: WORMTUBE EXPOSURE SYSTEMS 	  203
                              XI

-------
                            ABSTRACT

                        GUIDANCE MANUAL:
              BEDDED SEDIMENT BIOACCUMULATION TESTS

          Henry Lee II, Bruce L. Boese, Judy Pelletier,
      Martha Winsor, David T.  Specht,  and Robert C. Randall


     Bioaccumulation tests with bedded sediments are the most
direct method of deriving tissue residue data required for
evaluation of dredge materials and for quantitative ecological
and human risk assessments.  Bioaccumulation tests are also an
important experimental tool for identifying the factors
regulating the bioavailability of sediment-associated pollutants
and to test various Sediment Quality Criteria approaches.
However, the procedures for conducting such tests have not been
standardized, making it difficult to compare studies.  This
manual gives detailed guidance on how to conduct "routine" bedded
sediment bioaccumulation tests with marine or estuarine deposit-
feeding organisms.  All phases of the process are covered, from
formation of the experimental design,  through the actual
exposures, to statistical analysis and interpretation of the
results.

     Because the interpretation of tissue residue data is often
relative to  "control" and "reference" sites, the acceptability of
such sites is considered.  The importance of an appropriate
experimental design, including sufficient statistical power and
replication, is stressed.  Based on recommendations for
statistical power and the use of one-tailed tests, a minimum of
eight replicates is recommended.  Methods to avoid or reduce
"pseudoreplication, a common statistical problem in toxicity
tests, are also discussed.

     For the data generated from the bioaccumulation tests to be
useful in quantitative risk assessments, tissue residues should
be within 80% of the steady-state tissue residue.  Based on a
review of both 10-day and 28-day tests, we conclude that the 10-
day test will not meet this criterion for many environmentally
important pollutants, such as PCBs and DDT.  Therefore, we
recommend a  28-day  exposure as the routine duration for
bioaccumulation tests.  Techniques for conducting long-term
exposures  (>28 days) and kinetic approaches based on uptake and
depuration rates are also presented for cases when more accurate
estimates of steady-state tissue residues are required.

     Sediment collection and preparation, including spiking
techniques,  are discussed as are techniques for collecting and
maintaining  test species in the laboratory.  Based on a number of
criteria, including a required criterion for sediment-ingestion,
five species are recommended as suitable for routine testing.
                               xiv

-------
Another eight species are identified as potential "secondary"
species.  The water quality and sediment requirements for
exposure chambers are discussed, and in most cases, these
requirements can be achieved with relatively simple static or
flow-through systems.  Specific sampling schedules and techniques
are given for the routine 28-day exposures.  To allow comparisons
among studies, we recommend the Bligh-Dyer method as the standard
lipid technique, or, if another lipid method is used, to
intercompare with Bligh-Dyer.

     The statistical analysis of the data is discussed,  and the
use of one-tailed tests is recommended when comparing a test
tissue residue(s) to reference or control tissue residue(s), as
would commonly be the case when testing for "no further
degradation".  Besides the "no further degradation approach",
other regulatory strategies for using tissue residue data are
presented.
                              xv

-------
                     CHAPTER  I.   INTRODUCTION





     Sediments are the ultimate sink for many of the pollutants



entering the marine/estuarine environment through industrial and



municipal discharges, dredge materials, and non-point runoffs.



Sediments are an especially important repository for compounds



that sorb strongly to particles, such as organic pollutants with



high octanol-water partitioning coefficients (Kow)   (e.g., PCBs,



DDT) and many of the heavy metals.  As a general rule, these are



the same compounds that bioaccumulate to high levels.



Accumulation of pollutants in bedded sediments  (i.e., deposited



rather than suspended sediments) reduces their direct



bioavailability to pelagic organisms but increases their exposure



to benthic organisms.



     Bioaccumulation of these sediment-associated pollutants by



benthic organisms can result in a number of ecological and human



health impacts.  Bioaccumulation of pollutants can result in



acute and chronic effects in individual benthic organisms, which



ultimately can translate into alterations in benthic community



structure and function.  Because the benthos are the primary food



for demersal predators, predation on contaminated benthos is an



important pollutant uptake route for many ecologically and



economically important fishes and invertebrates.  Once introduced



into the pelagic foodweb, trophic transfer can spread the



pollutants to higher trophic levels, including sea birds, marine



mammals, and human consumers.  If tissue concentrations in

-------
shellfish or fishes exceed state or Federal regulatory criteria,



such as the FDA Action Levels, areas may be closed to commercial



and recreational fishing with the resultant economic loss.



     Given the importance of bioaccumulation by benthic



organisms, regulatory agencies need scientifically credible,



cost-effective methods to measure benthic tissue residues



resulting from exposure to existing sediments, as well as methods



to predict tissue residues resulting from projected levels of



sediment contamination.  Additionally, development of



standardized protocols to assess the bioavailability of sediment -



associated pollutants are required to assist in the development



of Sediment Quality Criteria.  This need to assess or predict



tissue residues in benthic organisms has long been recognized,



and a variety of field, laboratory, and theoretical approaches



have been developed, as discussed in Appendix 1-1.



     Of these approaches,  the laboratory test offers great



promise both in assessing existing field sediments or dredge



materials and as an experimental technique to gain insights into



the factors regulating bioavailability.  Unfortunately,  the



techniques for conducting sediment bioaccumulation bioassays have



varied considerably depending on the specific regulatory or



experimental goals,  making it difficult to compare results.



Probably the most frequently used procedure has been the 10-day



test to assess "bioaccumulation potential" of dredge materials



(U.S.EPA/U.S.ACE,  1977),  though recent evidence indicates the 10-



day exposure is not adequate (see Chapter IV and Appendix IV-1).

-------
     The purpose of this manual is to develop a standardized



approach to conducting sediment bioaccumulation tests with



sediment-ingesting organisms exposed to bedded marine/estuarine



sediment.  The manual covers all aspects of the bioassay



procedure, from experimental design to interpretation of the



results  (see Figure 1-1), though detailed guidance in certain



areas  (e.g., analytical methods) is beyond the scope of this



manual.  These guidelines are designed for the "routine" testing



of sediments and are not tailored toward any specific regulation



or geographical location.  The data obtained from the recommended



28-day test should, in most cases, generate the type of



information required for quantitative ecological and human health



risk assessments.  Users of this manual must recognize, however,



that the proposed standard 28-day test may underestimate actual



steady-state field tissue residues under certain conditions.



These possible sources of underestimation and alternate test



methods are discussed.



     The procedures presented here are based on our own



experiences, results from numerous published sediment



bioaccumulation tests, and to the extent appropriate, on the



standard bioconcentration tests for water uptake (ASTM, 1984),



the draft ASTM guidelines for sediment toxicity (ASTM, 1988a) and



sediment storage, characterization, and manipulation  (ASTM,



1988b), and the draft Ecological Evaluation of Proposed Discharge



of Dredged Material into Ocean Waters  (U.S. EPA/U.S. ACE, 1988)



in preparation by Battelle.  The reader is cautioned that the

-------
                   FIGURE  1-1
 Steps in Conducting Sediment Bioaccumulation Tests
         Define Control and Reference Sediment
        	Chap. II	
     Formulate hypotheses and experimental design
                       Chap. Ill
          Set duration and sampling schedule
                       Chap. IV
        i
  Collect Sediment
      Chap. V
Choose test species
     Chap. VII
        1
Characterize sediment
      Chap. VI
       1
Collect/maintain test
 species — Chap. VIII
                            r
       I
                                   Construct exposure
                                   system — Chap. IX
           Initiate experiment — Chap. X
           Pollutant analysis  — Chap. XI
           Statistical analysis - Chap. XII
             Interpretation - Chap. XIII

-------
final versions of these draft documents may differ in their



recommendations, and the final documents should be consulted.



     The reader is also cautioned that the area of sediment



bioavailability/bioassays is highly dynamic and there are diverse



opinions among these various sources.  When there is no



consensus, our rationale for recommending a technique is



discussed in the text or one of the appendices.  A manual such as



this must be considered a "living" document, as modifications



will inevitably be needed both as a result of future research and



as we gain a better understanding of the information required for



quantitative risk assessments.



     Though the methods in the manual have been directed toward



"routine" 28-day tests with marine/estuarine sediment, many of



the laboratory procedures are applicable to experimental



exposures of different durations and to brackish/freshwater



sediments.  The standard 28-day test recommended here is not



exclusive of other methods to assess or predict bioaccumulation



(see Appendix 1-1).  The specific technique(s)  used will depend



upon the accuracy and precision required and the specific goals.



In many cases, the various approaches complement each other and



could be used sequentially or concurrently.



     When working with contaminated field sediments or



experimentally spiked sediments, adherence to safe laboratory



practices is critical at all stages, as is disposal of all wastes



in an environmentally proper and legal manner.   The specifics of



laboratory safety and waste disposal are beyond the scope of this



manual, but some guidelines are given in Appendix 1-2.

-------
          CHAPTER II: CONTROL VERSUS REFERENCE SEDIMENT






     Tissue residues occurring in organisms exposed to the test



sediment(s) are statistically compared to those occurring in



organisms exposed either to "control" or "reference" sediments.



Thus, the difference between control and reference sediments is



critical to the interpretation of the results.  Before initiating



a test, determine if a particular sediment constitutes an



acceptable control and whether to use both a control and a



reference sediment or just a control sediment.






A. DEFINITIONS OF CONTROL AND REFERENCE SEDIMENTS



        A "control" sediment is a pristine sediment or, more



practically,  a sediment with very low levels of pollutants.



Essentially any contaminants in control sediments originate from



the global spread of pollutants and do not reflect any



substantial input from local point or non-point sources.  The



comparison of the test sediment versus the control is a measure



(within the statistical limits of the test)  of any



bioaccumulation from the test sediment beyond the inevitable



global background contamination.   The use of control sediment



also provides information on any contamination from the seawater



or the exposure system.  To the extent possible,  grain size, TOC,



and other key physical characteristics of the control sediment



should closely resemble those of the test sediment.

-------
        In comparison, a "reference" site may contain low to



moderate levels of pollutants.  There are two slightly different



ways in which to use a reference sediment.  In the first case,



the reference sediment is used as an indicator of the localized



sediment conditions exclusive of the specific pollutant input



being studied.  Such sediment would be collected near the site of



concern, and would represent the background conditions resulting



from any localized pollutant inputs as well as the global input.



This is the manner in which reference sediment is used in the



dredge material evaluations (U.S. EPA/U.S. ACE, 1988).  This



document states that reference sediment should be collected "near



the disposal site but should not have been influenced by previous



disposal of dredged materials."  As the purpose is to compare the



reference sediments to the dredge material (i.e., test.sediment),



the reference sediments should be similar to the dredge material



in grain size, TOC, and other physical-chemical characteristics.



     A second, though less common, use of reference sediments is



as a measure of the tissue residues at a particular site before a



specific pollutant input.  This differs from the previous



application in that the reference sediment is used as a measure



of background conditions at the specific site rather than as an



indicator of background conditions at another location.   For



example, if a sewage outfall pipe was being relocated in an



urbanized near-shore environment, tissue residues in organisms



exposed to sediment collected at the present discharge site (test



sediment) would be compared to those in organisms exposed to

-------
sediment from the new site (reference sediment).   The difference



in the tissue residues is an estimate of how much the relocation



of the sewage discharge will increase the tissue residues at the



new site.  As the purpose of this comparison is to predict what



will happen at a specific location, it may be impossible to



closely match the physical-chemical characteristics of the



reference and test sediments.



     Understanding the type of information generated by reference



sediments is critical for correct interpretation of the tests.



Reference sediments usually contain measurable concentrations of



a number of pollutants.  Uncontaminated organisms exposed to



reference sediments will eventually bioaccumulate pollutants to



the level of the organisms living at the reference site.



Comparing tissue residues from test sediments to those from a



reference sediment determines whether the test sediment results



in a significant incremental increase in tissue residues, not



whether there is any bioaccumulation (i.e., bioaccumulation



potential).



     Use of a reference site is appropriate when a "no further



degradation" approach is used to determine the suitability of an



industrial or municipal discharge or a disposal operation (see



Chapter XIII).  Even in cases when the use of a reference



sediment is appropriate, the sediment should not contain more



than low to, at most, moderate levels of pollutants.  If the



reference sediment contains too high a level of a contaminant,



the tissue residues in organisms exposed to reference sediment

-------
                             FIGURE II-1
             Overlap of Confidence Intervals in Organisms
     Exposed to Acceptable  and Unacceptable Reference Sediments
Unacceptable
tissue residue
   level
    Tissue
   residue
             Acceptable  Reference
                           Test
Reference
                   Unacceptable References
                                   Test
                 Control
                                            Control
                                                       Reference (b)
Reference (a)

-------
may not differ significantly from those in the test sediment even



though the organisms exposed to the test sediments accumulated an



unacceptable tissue residue (see Figure II-1).  The situation in



Figure II-1 is an extreme example, but it does illustrate that



the results of a comparison of reference and test sediments



depends upon the absolute level and the variation in the



pollutant concentrations at the reference site.






B.  CRITERIA FOR CONTROL AND REFERENCE SEDIMENTS



     There are no simple criteria available to judge the



acceptability of a sediment as a control or reference sediment.



Ideally,  the concentration of every anthropogenic pollutant



(e.g., PCBs, DDT) in a control sediment should be significantly



indistinguishable from zero, and the concentrations of naturally



occurring compounds (e.g., metals) should be within natural



levels.  In practice,  it will often be difficult to meet these



criteria.  One alternative is to use the pollutant levels in



Table II-1, which gives a range of pollutant concentrations in



control sites on the West coast, as a guide.  Sediment with



pollutant concentrations falling within the ranges in Table II-1



should represent adequate control values for the compounds



measured.  Alternatively, the concentrations at a putative



control site can be compared to the sediment concentrations



(normalized by the silt-clay fraction) given in NOAA (1988) .



This document presents raw data for both organics and metals for



about 200 near-coastal sites throughout the United States with



the concentrations for the highest and lowest 10 stations in






                                 10

-------
   TABLE II-l:  Representative Control Sediment Concentrations
COMPOUND
S. CALIF.1
BaPa
BFb
DDT
NAPHC
PAHd
PCB
Ag
As
Cd
Cr
Cu
Hg
Ni
Pb
Zn
(15-150)*
(<5.0-18)*
0.06-2.0
3-15
0.001-2
6.5-40
2.8-30
<1.0
<20.0
<10.0
<70.0
LOCATION
PUGET SOUND2 YAQUINA BAY, OR 3
7-30
7-80
0.03-0.6
3-30e
2-60
<0. 02-1.0
1.2
3-15
3.1-18.3
20.9
10-50
0.02-0.12
13.0
8.0
~
29-66
26.2
<0.01
<0.01f
<0.01
<0.01
0.559
0.47
19.3
6.3
14.5
5.5
26.3
Organics are in ppb dry wt.  Metals are in ppm dry wt.

  Not Considered Control Values

1 = Southern California   (Bascom, 1984; Brown et al.,
    Thompson et al.; 1984)
2 = Puget Sound, Washington   (Konasewich et al., 1982)
3 = Yaquina Bay, Newport, OR   (unpublished data)

f^Benzo (a) pyrene
"Benzofi,b,k)fluoranthene
^Napthalene
dPolyaromatic Hydrocarbons
®(Brown et al.,  1984)
£(Schults, D.W., unpublished data.
  U.S.EPA, Mar.  Sci. Ctr, Newport, OR)
g(Swartz et al., 1984)
1984;
                                 11

-------
tabular form.  Sediment concentrations falling within or near the



lowest 10 stations would be acceptable as controls.



Concentrations substantially above those in Table II-1 or the



normalized values for the lowest 10 stations in NOAA (1988)



should not be considered controls, with the possible exception of



sediments containing natural high levels of certain metals.



     Because the acceptability of a reference sediment depends in



part on the local background pollutant levels and how the



reference sediment will be used, no specific criteria will be



suggested here.  However, the appropriateness of the proposed



reference site should be examined carefully if the silt-clay



normalized concentrations fall in the upper half of the



concentrations presented in NOAA  (1988).



     Because the regulatory interpretation of bioaccumulation



tests is often based on a comparative approach, having



scientifically defensible definitions and criteria for control



and reference sediments are critical.  The suggestions presented



here represent a preliminary attempt at such criteria.  To



develop more rigorous criteria, a regional statistical analysis



of the existing sediment data and any ecological effects data is



required.  Defining control and reference concentrations by their



frequency of occurrence and by their correlation with adverse



biological effects would then be possible.



     It is important to emphasize that the comparison of a test



sediment versus a reference sediment tests for an incremental
                                 12

-------
increase in bioaccumulation and not whether any bioaccumulation



would result.  Therefore, a control sediment treatment should



always be used in addition to any reference sediments.






C. STANDARD REFERENCE SEDIMENTS



     Variation in organism behavior and physiology can



substantially affect pollutant uptake.  For example, uptake in a



test species could vary seasonally in response to changes in



lipid content or temperature, or vary non-seasonally in response



to organism health or site of collection.  The extent of this



variation should be assessed, especially if results will be



compared from tests conducted at different seasons or from tests



using organisms collected at different sites.



     Organism variation can be assessed by using a "standard



reference sediment," a well characterized sediment containing a



known and constant pollutant concentration.  This standard



reference sediment treatment is a  positive control and would be



conducted in addition to the normal control,  which is a negative



control.  Differences among studies in tissue residues in the



standard reference sediment would measure the inherent variation



associated with a test species.  Use of a standard reference



sediment would also help in standardizing results from different



laboratories and/or different species.




     Although positive controls have been suggested for sediment



toxicity tests (e.g.,  Johns et al.,  1989),  they have not been



adequately considered for use in sediment bioaccumulation tests.



Part of the problem is the lack of  any standardized sediment






                                 13

-------
suitable for bioaccumulation tests.  In the absence of a national



standard sediment, an interim solution is for each laboratory to



make its own standardized sediment.



     Because of potential spatial and temporal variations in



pollutant concentrations in field sediments, the use of a



laboratory spiked sediment is recommended as the standard.



Spiking methods are discussed in Appendix V-l.  The sediment used



for the reference should be collected at the site where the test



organisms are collected or, if that is impractical, the physical



characteristics (e.g., grain size, TOO  should closely match



those at the collection site.  The unspiked sediment can be



stored for long periods either by freezing or by drying.



Although both of these processes can affect the physical



integrity of a sediment, the purpose of the standard reference is



to provide a constant exposure regime, not necessarily a natural



one.  Before either of these storage techniques are used,



survival and/or behavioral bioassays should be conducted on



previously frozen or dried uncontaminated sediment to assure that



the technique does not adversely affect the test species.  The



sediment would be spiked just before its use.



     Ideally, the standard reference sediment would be spiked



with a suite of compounds ranging in chemical properties.



Alternatively, a single neutral organic and/or a single metal



could be chosen as a representative compound(s).   A specific PCB



congener, not an Aroclor, is a good candidate for the organics



because of the wealth of information on PCBs, their high





                                 14

-------
bioaccumulation potential, and their resistance to metabolism.



We suggest using 2,2',4,4',5,5' hexachlorobiphenyl (IUPAC #153),



which is the most frequently occurring PCB congener in



environmental samples (McFarland and Clarke, 1989) and is readily



bioaccumulated by marine worms and clams  (Rubinstein et al.,



1987; McElroy and Means, 1988; Lake et al.,  in review;



unpublished data).   DDT is another possibility as an organic



reference toxicant.  The use of radiolabeled PCB or DDT is



acceptable and would reduce the analytical load, though waste



disposal of a mixed waste could be a problem.  Cadmium is



suggested as a general reference metal.  Bioaccumulation of



sediment-associated cadmium has been studied in a number of



organisms (e.g., Ahsanullah et al.,  1984) and has been suggested



as the reference toxicant for Neanthes growth tests (Johns et



al., 1989)
                                 15

-------
          CHAPTER  III.  PRINCIPLES  OF EXPERIMENTAL DESIGN






     Care in experimental design is necessary to ensure the data



generated are relevant to the problem under investigation, as



well as to maximize the information gained relative to the cost.



This chapter will discuss some basic principles of experimental



design and provide examples as they relate to bioaccumulation



tests.  The following chapter presents specifics of the



experimental design such as test duration and sampling schedule.






A.   OBJECTIVES AND DEFINITIONS



     The objectives of these experiments are to quantify the



amount of pollutants bioaccumulated by organisms exposed to



sediments or dredge materials and to test whether this



accumulation is statistically greater than that occurring in a



control or reference sediment.  Each experiment consists of at



least two treatments; the control and one or more test



treatment(s).  The test treatment(s) consist(s) of the



contaminated or potentially contaminated sediment(s) or dredge



material(s).  A control sediment is always required to ensure



there is no contamination from the experimental set-up, but some



designs will also require a reference sediment  (discussed in



Chapter II).   Uptake from the control sediment or from the



reference sediment  (when appropriate) is used to provide baseline



values to compare with uptake from the test sediment.  Thus, the



reference sediment functions as the  "control" treatment during



comparisons with  test sediment but also functions as a test
                                 16

-------
treatment during comparisons with the control sediment.  Since



the statistical term "control" treatment could be confused with



the control sediment, in this chapter we will use the combined



descriptor control/reference when referring to the sediment used



as the "control" treatment.



     The organism(s)  to which a single application of treatment



is applied is the experimental unit.  This will be either a



single organism or group of organisms (i.e., composite, see



Section G) placed in an aliquot of a particular type of sediment



in an exposure chamber.  The specific type of sediment



constitutes the treatment, for example,  the dredge material is



the test treatment.  If a clam is placed in a beaker containing



sediment, the clam is the experimental unit and the beaker is the



exposure chamber.  If several worms have to be composited to



supply sufficient biomass for chemical analysis, the group of



worms would constitute the experimental unit and the beaker or



aquarium containing them would constitute the exposure chamber.



If an aquarium is physically subdivided, such as containing



several beakers each with an aliquot of sediment, then the



organism(s) placed in each beaker is the experimental unit.  The



important concept is that the treatment (sediment) is applied to



the experimental unit as a discrete unit.



     Experimental units must be independent and not differ



systematically.  This chapter will discuss the procedures



required to assure independence and randomization of the
                                 17

-------
experimental units, as well as the importance of replicating the



experimental units to assure a sufficiently powerful statistical



test.






B. HYPOTHESIS TESTING



     Statistical testing requires the establishment of the null



(Ho) and alternative  (Ha) hypotheses prior to conducting an



experiment.  In most cases, the tests for the bioaccumulation



bioassays will be one-tailed rather than two-tailed.  One-tailed



tests are used because the purpose of the experiment is to



determine whether uptake from test sediment is significantly



higher than uptake from control/reference sediment.  If uptake is



lower in the test sediment than in the control/reference



sediment(s), presumably no testing is required (see Chapter XII).



For these experiments, a one-tailed test will be performed where



Ho is that the mean tissue residue of organisms in a test



treatment is equal to the mean tissue residue of organisms in the



control treatment.  Ha is that the mean tissue residue of



organisms in a test treatment is greater than the mean tissue



residue of organisms in the control treatment.  Each test



treatment is compared to the control treatment separately.



     Levels of statistical significance are stated by setting



values for Type I and II errors.  A Type I error occurs when Ho



is rejected falsely (i.e., Ho is correct but rejected) and the



probability of a Type I error is usually termed "alpha" and given



a significance level of 0.05.  In other words, if the tissue



residues are equal (Ho is true) and the experiment were to be






                                 18

-------
repeated many times, an incorrect conclusion  (i.e., Ho rejected



with the conclusion that tissue residues are not equal) would



occur 5% of the time.  Type I error can be considered the



"discharger's " risk as it is the probability of incorrectly



ascribing bioaccumulation to a sediment or dredge material.



     A Type II error occurs when Ho is falsely accepted  (i.e., Ha



is true) and is termed "beta".  The converse of a Type II error



(1-beta) is the statistical "power" of the test, which is the



probability of correctly rejecting Ho  (i.e., Ha is correct).



We recommend a value of 0.05 for beta  (power = 0.95) as the



standard for the bioaccumulation tests.  This means that if there



were a true difference between test and control/reference tissue



residues and the experiment were to be repeated many times, an



incorrect conclusion  (i.e., tissue residues equal when actually



there is bioaccumulation) would occur 5% of the time due to



chance.  Type II error can be considered tne  "environmental" risk



as it is the probability of incorrectly concluding that a



sediment or dredge material will not result in bioaccumulation.



Using a one-tailed test, as recommended here, instead of a two-



tailed test increases the power of the test given a set Type I



error and reduces the number of replicates required.



     By using the same value  (0.05) for both Type I and Type II



errors, an equal probability of error is assigned to both the



"discharger" risk and the "environmental" risk.  An implicit



assumption of assigning equal risks is that the "cost" of making



either type of error is equal.  It is difficult, and often






                                 19

-------
subjective, to compare the monetary costs of pollution



treatment/dredging to the environmental/human health costs of



degradation of marine ecosystems.  Therefore, in the absence of



other data, we believe the use of equal risk is the most



defensible procedure.  In cases where it can be demonstrated that



the "dischargers" risk is substantially greater than the



"environmental" risk, the beta could be increased to 0.20, which



would reduce the probability  (power) of correctly detecting



bioaccumulation to 80%.



     Each pollutant must be considered and tested separately.



Different pollutant tissue residues determined from the same



experimental units are not independent and so can not be compared



using the standard statistical tests.  The appropriate



statistical procedure for comparisons between different



pollutants  (e.g., comparisons of PCB congeners from the same



tissue samples) is repeated measures ANOVA (see Chapter XII).






C. REPLICATION



     An important principle in experimental design is the



replication of experimental units.  Replication is the assignment



of a treatment to more than one experimental unit, which in the



bioaccumulation experiment is the organism (or composite of



organisms) to which a single  treatment  (e.g., test or



control/reference sediment) is applied.   The variation among



replicates is a measure of the within-treatment variation which



includes random variation among individuals as well as sampling
                                 20

-------
and analytical errors.  This variation provides an estimate of



within-treatment error used for assessing the significance of



observed differences between treatments.  In experiments without



replication, inferential statistical testing is not possible.



     A minimum number of replicates is needed for sufficient



statistical power to determine whether the tissue residues of the



test organisms are greater than those of the control.   The number



of replicates required can be calculated from an estimate of the



variance or coefficient of variation of tissue residue values and



a predefined minimum detectable difference between the two means.



The minimum detectable difference is the smallest absolute



difference between two means that is statistically



distinguishable.  For example, if the tissue residues in the



organisms exposed to the test sediment must be at least twice as



great as those in the control sediment to be statistically



distinguishable, the minimum detectable difference of the means



is 2.0.  Besides the absolute difference between two means, the



minimum detectable difference can be expressed as a proportion of



the mean or as a proportion of the variance (see Appendix III-l).



     The smaller the minimum detectable difference, the greater



the number of replicates required for a given significance level



and power.  Although there is no consensus on what constitutes an



acceptable minimum difference, we suggest the bioaccumulation



experiment be designed to detect a 2-fold difference between



tissue residues in the test and control sediments or the test and



reference sediments.  In most cases, a 2-fold difference should
                                21

-------
provide a sufficiently precise result to address the ecological



and human health concerns.  A smaller minimum detectable



difference would be required if the 2-fold range around the



control/reference sediment overlapped a human health or



environmental criterion  (e.g., FDA action limit), though this



should be rare if the control/reference sediments are chosen



correctly.




     In some cases,  it may not be necessary to distinguish a 2-



fold difference in tissue residues among the treatments, such as



when there are large differences between the treatments.  For



example,  if the test sediment is suspected of being relatively



contaminated (e.g.  pollutant concentrations greater than 50 times



control values),  then a 5-fold minimum detectable difference



would be sufficient to find significant differences between the



treatments.  In other cases, it may not be possible to achieve



the 2-fold minimum detectable difference with a reasonable



experimental design, such as with compounds with high analytical



variation.  When it is impractical to achieve the 2-fold minimum



detectable difference, we recommend using sufficient replicates



to distinguish a 5-fold difference in tissue residues between



treatments, with the caveat that the 5-fold range around the



control/reference does not overlap a sediment or tissue residue



criterion or end-point.



     Appendix III-l provides instructions on computing the number



of replicates (n).   Figure III-l may be used to determine n from



a coefficient of variation and a minimum detectable difference
                                 22

-------
to
U)
        400-
o 300
i—
•<

-c
        200
      o 100
      o
                                  FIGURE III-1
                     STATISTICAL SIGNIFICANCE = 0.05
                         POWER OF TEST =0.95
                           ONE-TAILED TEST
                                                             I
                                                            10
                                                            I
                                                           11
12    13
                                                                                6=5
                                                                          6=4
                                                                                6=3
                                                                          6=2
                                                                          6=1
i
H
                                          SAMPLE SIZE
                      COEFFICIENT  OF  VARIATION  VS SAMPLE  SIZE FOR
                       VARIOUS MINIMUM DETECTABLE DIFFERENCES (5)
                             EXPRESSED  AS A PROPORTION OF THE MEAN

-------
TABLE III-l: Ranges of Coefficient of Variations (CV) for Tissue
             Residues Reported for Benthic Organisms
POLLUTANT
ORGANISM
CV
REFERENCE
Cadmium Modiolus demissus
Mytilus edulis
Mya arenaria
Mulinia lateralis
Callianassa australiensis
Mercury Modiolus demissus
Mytilus edulis
Copper Neanthes arenaceodentata
Zinc Nereis divers icolor
Octolasion tyrtaeum
Corbicula fluminea
Kepone Crassostrea virainica
PCB Octolasion tyrtaeum
Corbicula fluminea
Nereis virens
Uca spp.
HCB Ma coma nasuta
BaP Amphipods
Ma coma inquinata
Abarenicola pacif ica
Napthalene Ma coma inquinata
Phenanthrene Ma coma inquinata
Abarenicola pacif ica
Chrysene Ma coma inquinata
Abarenicola pacif ica
*Samples were composited resulting in (usually)
1. Breteler and Saksa, 1985
2. Jackim et al . , 1977
3. Ahsanullah et al . , 1984
4. Pesch and Morgan, 1978
5. Renfro and Benayoun, 1975
6. Mac et al. , 1984
7. Morales-Alamo and Haven, 1983
8. McFarland et al . , 1985
9. Rubinstein et al . , 1984
10. U.S. EPA, 1986b
11. unpublished data, 1989
12. Reichert et al . , 1985
13. Augenfeld et al . , 1982
14. Roesijadi et al . , 1978
4-54%
4-61%
18-22%
35-49%
5-67%
5-34%
5-53%*
8-60%
42%
12-30%
7-8%
8-80%
2-23%
10-74%
5-40%
31-75%
23-33%
4-22%
4-36%
9-24%
50-100%
17-56%
10-31%
11-46%
2-46%
lower CV's














1
1
2
2
3
1
1
4
5
6
6
7
6
6,8
9
10
11
12
13
13
14
13
13
13
13















                                 24

-------
expressed as a proportion of the mean using a significance level



of 0.05 and a power of 0.95 (beta value of 0.05) for one-tailed



tests.  If no other information is available, the coefficient of



variations for tissue residues in various benthic species given



in Table III-l can used as guides for Figure III-l.  The values



in Table III-l should be used as lower estimates as many were



derived from composites (which will produce lower CVs than will



individual samples),  used radiolabeled compounds, and are the



results from successful, published experiments.



     ASTM (1984) recommends at least four replicates to determine



bioconcentration factors.   Because of the likelihood of a greater



variation in sediment exposures compared to water exposures, we



recommend a minimum of eight replicates as the  "default" number



of replicates to provide a statistical power of 95%.  Figure III-



2 may be used to determine if eight replicates are adequate for a



specified coefficient of variation and minimum detectable



difference expressed as a proportion of the mean.  In some cases,



when variability is low or less power is required, as few as five



replicates can be used, though five should be an absolute



minimum.  In this discussion,  the number of replicates refers to



the number analyzed for tissue residues and not the number



exposed.  It is prudent to include an extra replicate or two for



each treatment in case of mortality or the loss of samples during



chemical analysis.
                                 25

-------
to
         400 H
          300-
                                   FIGURE  III-2
STATISTICAL SIGNIFICANCE = 0.05

    POWER OF TEST =0.95

      ONE-TAILED TEST
                                   MINIMUM DETECTABLE DIFFERENCE
                  THE MINIMUM DIFFERENCE  DETECTABLE BETWEEN TREATMENTS
                         (EXPRESSED AS A PROPORTION  OF  THE  MEAN)
               FOR  A SPECIFIC COEFFICIENT OF VARIATION  AND  SAMPLE SIZE  (N)

-------
D. RANDOMIZATION



     Randomization is the unbiased assignment of treatments to



the experimental units  (i.e., organisms or composites of



organisms) ensuring that no treatment is favored and that



observations are independent.  This is necessary for valid



statistical testing.  Randomization is often performed by using



tables of random numbers.  For these experiments, it is important



to randomly assign the organisms to the control and test



treatments, to randomize the allocation of sediment (e.g., not



take all the sediment in the top of a jar for the control and the



bottom for spiking), and to randomize the location of exposure



units.  For example, a bias in the results may occur if



assignments are not randomized and all the largest animals are



placed in the same treatment.






E. PSEUDOREPLICATION



     The appropriate assignment of treatments to experimental



units is critical to avoid a common error in design and analysis



recently termed "pseudoreplication" (Hurlbert, 1984).



Pseudoreplication occurs when inferential statistics are used to



test for treatment effects even though the treatments are not



replicated or the replicates are not statistically independent



(Hurlbert, 1984).



1. Lack of Replication




     The simplest form of pseudoreplication is treating



subsamples as true replicates of the experimental unit (Figure



III-3a).  For example,  two aquaria are prepared,  one with






                                27

-------
control sediment, the other with  test  sediment, and  five



organisms are placed in each aquarium.  Even  if each organism  is



analyzed individually, the five organisms are not  true replicates



because the treatment  (i.e., sediment  type) is applied to the



aquarium as a whole and not to each individual organism



separately.  In this case, the experimental unit is  the five



organisms and each organism is a  subsample, therefore, there is



no replication of experimental units in this  particular design.



2. Segregation




     A less obvious form of pseudoreplication is the physical



segregation of replicates by treatment, potentially  resulting  in



a systematic error (bias)  and lack of  independence.   For



example, all the control experimental  units are placed in one



area of a room and all the test experimental  units are in another



(Figure III-3b).  Spatial effects (e.g., different lighting,



temperature) could bias the results for one set of treatments



making it impossible to distinguish true effects of the treatment



from the effects due to the physical layout of the experiment.



Random physical intermixing of the experimental units is



necessary to avoid this type of pseudoreplication.



     A more common form of segregating replicates is the use of



separate aquaria for each treatment.   For example,  segregation



would occur if all the control experimental chambers (e.g.,



beakers) are placed in one aquarium and all the test experimental



chambers in another aquarium (Figure III-3c).   Any effects due  to
                                 28

-------
               FIGURE III-3
       Random and Pseudorandom
          Replication Schemes
a. no replication

b. segregation
C. segregation
                  aquarium 1
d. randomized
   with interdependent    \	1—
   replication      g|  j—j |—j
                     aquarium 2
                        cm
e. completely
  randomized

f. randomized
  block
           cm
      cm
    a-d
    e-f
aquarium 1   aquarium 2   aquarium 3

control experimental unit

test experimental unit

pseudoreplication

strict replication
                    29

-------
temperatures or different lighting conditions could bias the



results for one of the treatments.  Replicate aquaria are



necessary in this case.  Section III-F gives suggestions on



addressing this type of pseudoreplication.



3. Randomization With Interdependent Replicates



     Randomized spatial interspersion does not necessarily



preclude pseudoreplication.  If the replicates are physically



interdependent, spurious effects can bias one treatment over



another.  This can occur if all the aquaria replicates of the



control are serviced by the same water supply system while all



the treatment aquaria replicates are serviced by another water



supply system  (Figure III-3d).   Any differences between supply



systems may potentially bias one set of aquaria over another.



Thus, even if the aquaria replicates are physically interspersed,



the replicates are not independent.  To avoid pseudoreplication,



each experimental unit should have its own water or air supply,



all branching off a common supply and there should be no flow of



water from one exposure system to another.





F. AVOIDING OR REDUCING PSEUDOREPLICATION



1. Avoiding Pseudoreplication



     Pseudoreplication can be avoided by properly identifying the



experimental unit, providing replicate experimental units for



each treatment, and applying the treatments to each experimental



unit in a manner that includes interspersion and independence.



The simplest design that avoids pseudoreplication is the



completely randomized design (Figure III-3e).  In this design,





                                 30

-------
treatments are randomly assigned to the experimental units



independent of location and each experimental unit is maintained



in a separate exposure chamber with a separate water and air



supply.



     A randomized block design is also appropriate.  A block is a



set of relatively homogeneous units to which treatments are to be



applied, such as all the beakers within an aquarium.  In the



randomized block design, all the treatments are randomly assigned



to each block, and there are multiple blocks.  For example, if



there are two treatments and eight beakers per aquarium, each



aquarium is randomly assigned four beakers with control sediment



and four beakers with test sediment  (for another example see



Figure III-3f).  One drawback of this design, however, is that



since both test and control organisms are in one aquarium



(block), the potential exists for contamination of controls by



test sediment.  This is especially likely with organisms that



eject sediment into the water, such as Macoma during the



production of pseudofeces.  If this design is used, the aquaria



and/or control exposure chambers need to be monitored to assure



that cross-contamination does not occur.



2. Reducing Pseudoreplication Effects



     Totally  avoiding pseudoreplication may be difficult or



impossible given resource constraints.  For example, one common



experimental  design segregates the experimental treatments in



separate aquaria.  In this case, the beakers containing the test



sediment are  placed in separate aquaria from beakers containing





                                 31

-------
 the  control  sediment  (see  Figure  III-3C).   Such  a  design avoids



 the  problem  of  cross-contamination between  the test  and  control



 sediment and does not  require a separate aquarium  for  each



 beaker.  However, because  the beakers are segregated by  treatment



 type, their  distribution is not random.  In such cases,  the



 experimental unit may  be defined  as the replicated unit



 (organism(s) in the beaker with each beaker as a replicate), but



 with the stated assumption that there is no effect due to the



 physical segregation  (aquaria effect in this example).




     With this design, we  recommend using replicate  aquaria for



 each treatment type to enable comparison of  results  between



 aquaria within a given treatment  using a nested  ANOVA.   If



 aquaria effects are apparent, the data from one  or more  aquaria



 may be considered invalid, or the differences due  to the aquarium



 effect may be deemed trivial compared to the treatment effects.



 For example, if there  is a significant difference  among  test



 aquaria results, but that difference is much less  than the



 difference between test and control aquaria, the aquaria effects



may be considered unimportant to  the results of  the experiment.



 However,  moderate to large differences between aquaria of the



 same treatment would suggest a local contamination problem or



 other type of bias and the experiment should be  repeated.  If no



 significant aquaria effects are detected,  the organism(s) within



 each beaker are properly considered the experimental unit and



each beaker a replicate.   The analysis (see Chapter XII)  is then



performed as if the beakers were not segregated  into aquaria.






                                32

-------
G. COMPOSITING SAMPLES



     Compositing consists of combining samples  (e.g., organisms,



sediment) and chemically analyzing the mix rather than the



individual samples.  The chemical analysis of the mix provides an



estimate of the average concentration of the individual samples



making up the composite.  Compositing will be used in



bioaccumulation experiments primarily when the biomass of an



individual organism is insufficient for chemical analysis.



Several individuals can be composited into a single experimental



unit with sufficient biomass and the analysis performed on the



composite.  Compositing is also used when the cost of analysis is



high.  For example, sediment pollutant analysis can be based on



sediment composited from several exposure chambers of the same



treatment to reduce the analytical sample load.  Replicate



sediment or tissue composites  (i.e., experimental units) are



required if statistical testing is planned.



     For the tissue composite to be unbiased, the individuals



must be randomly assigned to the various treatments.  Each



organism or sediment sample added to the composite must be of



equal size (i.e., wet weight) and the composite must be



completely homogenized before taking a sample.  If compositing is



performed in this manner, the value obtained from the analysis of



the composite is the same as an average obtained from analyzing



each individual sample  (within any sampling and analytical



errors) .   If replicate composites are made, the variance of the



replicates will be less than the variance of the individual






                                 33

-------
samples, providing a more precise estimate of the mean value.



This increases the power of a test between means of composites



over a test between means of individuals or samples for a given



number of samples analyzed.




     However, if composites are made of individuals or samples



varying in size  (e.g., varying weights of sediment samples) or



quality (e.g., disproportionate number of gravid females in one



composite),  the value of the composite and the mean of the



individual organisms or sediment samples are no longer



equivalent.   The variance of the replicate composites will



increase,  decreasing the power of a test between means.  In



extreme cases, the variance of the composites can exceed the



population variance (Tetra Tech, 1986a).   Therefore, it is



important to keep the individuals or sediment samples comprising



the composite equivalent in size and quality.  If sample sizes



vary,  consult the tables in Schaeffer and Janardan  (1978)  to



determine if replicate composite variances will be higher than



individual sample variances, which would make compositing



inappropriate.



     It is not advisable to composite samples if an estimate of



the population variance is required or if information about the



range in values obtained for individuals is needed.  For example,



tissue samples should not be composited if it is important to



know the percent of individuals exceeding the FDA Action Limits.



Compositing also requires more individuals (assuming individuals
                                 34

-------
can be analyzed)  so it is not advised when space or cost keeps



the number of individuals at a minimum.  When there is extra



sediment or tissue, archive individual samples in case a measure



of the population variance or the concentration in a particular



exposure chamber is desired latter.
                                35

-------
         CHAPTER IV:  TEST DURATION AND SAMPLING SCHEDULES






     Besides the statistical issues addressed in the previous



chapter, an environmentally-relevant bioaccumulation test



requires an appropriate exposure duration and sampling schedule



for organism, sediment, and water samples.  This chapter



discusses these topics for the standard 28-day and long-term



bioaccumulation tests, as well as a kinetic approach.  Additional



discussion of the 10-day versus 28-day tests are give in Appendix



IV-1.  Alternative test designs, which may be applicable under



special circumstances or for research purposes, are presented in



Appendix IV-2.






A. STANDARD 28-DAY BIOACCUMULATION TEST



1. Steady-State and Duration



     Ideally, the duration of a bioaccumulation test should be



sufficient for the organisms to reach steady-state tissue



residues, where steady-state is operationally defined as the lack



of any significant difference  (ANOVA, alpha = 0.05) among tissue



residues taken at three consecutive sampling intervals (ASTM,



1984).  The time to  reach or approach steady-state varies



drastically among different compounds, but in general, the tests



should be designed to generate environmentally-relevant data on



high Kow organics (e.g., PCBs, DDT) and heavy metals.  Therefore,



we recommend a 28-day exposure as the standard duration.   As



discussed in Appendix IV-1, a 28-day exposure will result in



tissue residues within 80% of the steady-state tissue residues in
                                 36

-------
most cases.  When steady-state is not approached within 28 days,



tissue residues of organics usually appear to be within 2-4 fold



of steady-state concentrations (Table IV-1),  which is considered



acceptable for the ASTM bioconcentration test (ASTM, 1984).



Although a 28-day sediment exposure will not assure attainment of



steady-state for all environmentally important compounds,  it will



improve the predictive power of the test compared to the commonly



used 10-day exposure.  For cases where more accurate estimates of



the steady-state tissue residues are needed,  a long-term



bioaccumulation test (Section B)  or a kinetic uptake approach



(Section C) is recommended.



2. Biotic Sampling Schedule



    Biological samples are used to determine the amount of



pollutants accumulated from the test sediment and to



statistically compare these values to the amount of pollutants



accumulated from control and reference sediments.  To set  the



baseline conditions for these comparisons,  bioassay organisms



should be analyzed for pollutant and lipid content immediately



before initiation of the experiment (tr, samples) .  As discussed



in the previous chapter (Chapter III),  eight replicates are



assumed as the number required to achieve sufficient statistical



power.  Therefore,  eight replicate organisms or composites (i.e.,



experimental units)  should be analyzed at tQ.  The organisms



sampled at tQ should be chosen randomly from the same set of



organisms used in the various sediment treatments.  If



compositing of individuals is necessary to obtain sufficient
                                 37

-------
TABLE IV-1: Percent of Steady-State Tissue Residue of Neutral
            Organics Obtained After 10 and 28 Day Exposures to
            Bedded Sediment.


Compound
% of Steady-State
Tissue Residue
10 -DAY 28 -DAY Species

Estimated
by Ref .
Phenanthrene       100   100
Benzo(a)pyrene      96   100
Benzo(a)pyrene      96   100
Phenanthrene        94   100
Hexachlorobiphenyl  88   100
Phenanthrene        67    95
Aroclor 1260        53   100
Benzo(a)pyrene      43    75
Chrysene            43    87
Hexachlorobenzene   35    70
Benzo(a)pyrene      32    66
Aroclor 1242        29    82
Aroclor 1260        27   100
Aroclor 1254        27   100
Aroclor 1242        18    87
Aroclor 1254        12    82
Aroclor 1254         9    25
Macoma inquinata     G
Hexagenia limbata    K
Mysis relicta        K
Mysis relicta        K
Hexagenia limbata    K
Pontoporeia hoyi     K
Macoma balthica      G
Macoma inquinata     K
Macoma inquinata     G
Macoma nasuta        K
Pontoporeia hoyi     K
Cerastodema edule    G
Cerastodema edule    G
Cerastodema edule    G
Nereis virens        G
Macoma balthica      G
Nereis virens        K
1
3
3
3
3
3
4
1
1
2
3
4
4
4
4
4
5
K = Steady-state tissue residue estimated from kinetic uptake
    model.

G = Steady-state tissue residues estimated from graphs of tissue
    residues versus time.  Often, the 10-day and 28-day values
    had to be interpolated.  "Steady-state" was defined as the
    time period with the maximum tissue residue.

SOURCES:
1 = Augenfeld et al.,  1982
2 = Boese et al., in press and unpublished data
3 = Landrum and Poore,  1988
4 = Langston, 1978
5 = McLeese et al., 1980
                                 38

-------
biomass, the same compositing scheme should be used for all



sampling periods throughout the experiment.  At the end of the



28-day test period  (t28),  eight replicate organisms or composites



should be taken from each of the treatments and analyzed for



pollutants and lipids.  For comparing test and control sediments,



the simplest design results in 24 tissue samples  (8 controls at



tQ, 8 controls at t28, and 8 test at t28) .



     We recommend including an extra one or two replicates in



each treatment in case a sample is lost.  Additionally, several



extra individuals or composites should be taken at the initiation



of the experiment.  These extra samples should be frozen until



the tissue residue data has been analyzed and interpreted.  The



method of physically sampling the organisms is discussed in



Chapter X.



     Time-series samples may be taken during the 28-day exposure



to document uptake kinetics.  This type of information can be



very helpful even if it is necessary to limit the analytical load



by taking only a single sample or, preferably, a single composite



at each sampling period.  However, if the data will be



statistically compared to determine if steady-state has been



attained,  replicates are required at each sampling period.  The



sampling interval for these samples should approach a geometric



progression with sampling periods of no greater than one week



(e.g.,  day 0, 2,  4,  7, 14,  21,  and 28).  A sample at 10 days is



recommended if there are previous 10-day exposure data.  Begin



the series on Monday to avoid weekend sampling.
                                 39

-------
3. Abiotic Sampling:  Frequency and  Replicates



     The physical and chemical properties  of each  test,  control,



and reference sediment need to be characterized  immediately after



collection.  This includes, at a minimum,  grain  size




distribution, moisture content, pollutant  concentrations, and TOG



(or LOI)  (see Chapter VI).  Depending upon the length of storage,



it may be necessary to remeasure these physical  and chemical



parameters, with the  possible exception of grain size



distribution, immediately prior to  the start of  the



bioaccumulation test  (i.e., tQ).  If these tQ samples will be



statistically compared to samples taken at the end of the test



period  (t28), eight replicate samples are  required.



     At the end of the bioaccumulation test (t28), take sediment



samples from each exposure chambers for measurement of pollutant



concentrations,  TOG, and moisture contents.  It  is usually not



necessary to remeasure grain size.   Preferably,   these analyses



should be conducted on the sediment from each beaker or aquarium



(i.e.,  experimental unit).  Measurements on individual



experimental  units may help explain any unexpected variation



among the replicates.   If eight replicates are used per



treatment,  this would result in a total of 24 sediment samples (8



controls at tQ,  8 controls at t28,  and 8 test samples at t28).



     If this  is too large an analytical load,  an alternative is



to analyze a  composite sample from each treatment composed of



equal aliquots of sediment from each beaker or aquarium within



the treatment.   Additionally,  a sediment sample  from each beaker
                                40

-------
or aquarium should be taken and archived.  If the tissue residue



data are more variable than expected or if there are "unusual"



data points, these individual sediment samples should be



analyzed.  Additionally, individual sediment samples should be



analyzed if the differences in pollutant concentrations in the tQ



and t2g sediment samples are greater than would be expected from



analytical variation alone.  It would then be possible to



determine if significant changes in pollutant concentration or



TOG had occurred during the course of the experiment.






B. LONG-TERM UPTAKE TESTS



1. Criteria and Limitations



    In some cases, body burdens will not approach within 80-90%



of the steady-state body burdens in a 28-day test (see Table IV-1



and Appendix IV-1).  Organic compounds exhibiting these kinetics



will likely have a log Row > 5, be metabolically refractory



(e.g., highly chlorinated PCBs, dioxins), and exhibit low



depuration rates. Many of these same organic compounds biomagnify



in aquatic food webs and pose a human health risk.  Additionally,



tissue residues of several heavy metals may gradually increase



over time so that 28 days is inadequate to approach steady-state.



Depending on the goals of the study, it may be necessary to



conduct an exposure longer than 28 days (or a kinetic study as



discussed in Section C)  to obtain a sufficiently accurate



estimate of steady-state tissue residues of these compounds.



Although these longer term studies generate more accurate data



for these compounds,  they require greater resources, increase the






                                 41

-------
analytical load,  and increase the likelihood of problems



involving the maintenance of the organisms and temporal changes



in sediment pollutant concentrations.



2. Biotic Sampling



     In the long-term studies,  the exposure should continue until



steady-state body burdens are attained.  As mentioned, steady-



state is documented by the lack of any statistical difference in



tissue residues in three consecutive sampling intervals (ASTM,



1984).  ASTM (1984) recommends a minimum of five sampling periods



(plus t0) when conducting water exposures to generate



bioconcentration factors (BCFs).  For bioconcentration tests,



ASTM (1984) recommends sampling in a geometric progression with



sampling times reasonably close to S/16, S/8, S/4, S/2, and S,



where S is the time to steady-state.  This sampling design



presupposes a fairly accurate estimate of time to steady-state,



which is often not the case with sediment exposures.



     To document steady-state from sediment exposures, we



recommend placing a greater number of samples at and beyond the



predicted time to steady-state.  With a pollutant expected to



reach steady-state within 28-50 days, samples should be taken at



days 0, 7, 14, 21, 28, 42, 56, and 70.  If the time to steady-



state is much greater than 42 days,  then additional sampling



periods at 2 week intervals should be added  (e.g., day 84).



Slight deviations from this schedule (e.g., day 45 versus day 42)



are  not critical, though for comparative purposes, samples should



be taken at t2g.  An estimate of time to steady-state may be






                                 42

-------
obtained from the literature or approximated from structure-



activity relationships (Appendix IV-3),  though these values



should considered the minimum times to steady-state.



     Compared to the ASTM bioconcentration sampling schedule,



this schedule increases the likelihood of statistically



documenting that steady-state has been obtained though it does



not document the initial uptake phase as well.  If accurate



estimates of the first-order uptake coefficient (kl, see Section



C) are required, add sampling periods during the initial uptake



phase (e.g., days 0, 2, 4, 7, 10, 14, 17, etc.).



     One problem with longer exposures is the greater probability



of the test organisms reproducing.  Spawning can drastically



affect lipid content and possibly pollutant concentrations



(Niimi,  1983).   Additionally, because many species die after



spawning, it is prudent to add extra replicates.  Increasing the



total number of replicates by an additional 10-20% should suffice



in most cases.   If not needed, archive these extra individuals at



the end of the test as replacement samples in case of analytical



failures or analyze them to increase the statistical power of the



final sampling period.



3. Abiotic Samples



     The bioavailable fraction of the pollutants as well as the



nutritional quality of the sediment are more prone to depletion



in these extended tests than in the 28-day exposures.  To



statistically document whether such depletions have occurred, at



least eight replicate sediment samples are required for physical





                                 43

-------
 and  chemical analysis  from each  sediment  type  at  the  beginning

 and  the  end of  the  exposure.  Additionally,  we recommend

 archiving  sediment  samples from  every biological  sampling period.

     To  minimize the depletion of  sediment pollutants or

 nutrients, completely  replace the  sediment with stored  sediment

 or freshly spiked sediment on a  regular basis  (e.g., monthly).

 Sediment must be renewed carefully to avoid  damaging the test

 organisms, especially  polychaete worms.  Another  way to minimize

 depletion  of pollutants is to add  fresh sediment  periodically

 (see Chapter X).  Over a long experiment, however, the  exposure

 container  may be entirely  filled,  necessitating the replacement

 of the sediment anyway.  Replenishment sediment should  be sampled

 and analyzed for the recommended parameters.   Do  not feed the

 organisms  a supplemental food (e.g., fish flakes)  as this will

 reduce exposure to  ingested sediment and may result in  an

 underestimation of  sediment bioavailability and steady-state

 tissue residues.


 C. ESTIMATING STEADY-STATE TISSUE RESIDUES FROM UPTAKE AND
   DEPURATION RATES

     Several methods have been published which can be used to

predict steady-state pollutant levels from uptake and depuration

kinetics (Spacie and Hamelink,  1982; Davies and Dobbs, 1984).

All of these methods were derived from fish exposures and most

use a linear uptake, first-order depuration model  which may be
                                 44

-------
modified for uptake of pollutants from sediment:



          Ct(t) = kl*Cs/k2*(l-e~k2*t)              (1)



Where:    Ct = pollutant concentration in tissue at time t



          Cs = pollutant concentration in sediment.



          kl = uptake constant  (day"1 )



          k2 = depuration constant  (day"1)



          t = time  (days)



     As time approaches infinity, the maximum or equilibrium



pollutant concentration within the organism  (Ctmax) becomes:



          Ctmax = Cs*kl/k2                         (2)



     Correspondingly, the bioaccumulation factor (BAF) for a



compound may be estimated from:



                    BAF = kl/k2                    (3)



     The kinetic approach requires an estimate of the uptake rate



constant (kl)  and the depuration rate constant  (k2), which are



determined from the changes in tissue residues during the uptake



phase and depuration phase, respectively.  The uptake experiment



should be short enough that an estimate of kl is made during the



linear portion of the uptake phase  (Figure IV-1) to avoid an



unrealistically low uptake rate due to depuration.   The



depuration phase should be of sufficient duration to smooth out



any loss from a rapidly depurated compartment (Figure IV-1) .



Unless there is reason to suspect that the route of exposure will



affect the depuration rate, it is acceptable to use a k2 derived



from a water exposure.  The durations of the uptake and



depuration experiments will vary with animals species,
                                 45

-------
FIG.  IV-1:  Idealized  Uptake -  Elimination  Curve
                                          "k   M
Ct(t)  =
         Uptake Phase
                  k2/k   *
                                   (  1-e
                      Cs = k  * Ct
                      Steady-state
               Linear uptake phase
        * Cs » k
                            Ct
                           Elimination  Phase
                                       Fast component
                             Slow component
   (Contaminated  Sediment) '(Clean  Sediment)
                    TIME
    tissue concentration (ug/g)
    uptake constant  (1/time)
    sediment concentration (ug/g)
    elimination  constant (1/time)
t  = time
       Ct
       k-,
       Cs
         2

-------
compound, pollutant concentration, analytical detection limits,



and test sediment.  As a result, no specific guidelines will be



presented here.  For a discussion of this method for



bioconcentration studies in fish, see Davies and Dobbs (1984),



Spacie and Hamelink (1982), and the ASTM standard practice for



conducting bioconcentration tests (ASTM, 1984).   Effects of



growth on the estimation of kl and k2 and how to correct for



growth dilution effects on the estimate of steady-state tissue



residues are discussed in Appendix IV-2.
                                 47

-------
         CHAPTER V. SEDIMENT COLLECTION, HOMOGENIZATION,
                    MANIPULATION, AND STORAGE


     Bioaccumulation tests use sediments collected in the field

and brought back to the laboratory or manipulated experimentally

in the laboratory.  In both cases, the handling can result in

loss of fine sediments, interstitial water, and water soluble

compounds; oxidation of compounds; or contamination by metals and

organics.  This disruption can change physicochemical properties

such as grain size distributions, pollutant concentrations,

sorption equilibria, speciation, and complexation, thereby

affecting pollutant bioavailability (Plumb, 1981; Jennett et al. ,

1980; Holme and Mclntyre, 1984).  Although some changes are

unavoidable, they can be minimized with appropriate techniques.

In this chapter, we provide guidelines on handling sediments

during and after collection.  We cannot, however, formulate a

standard operating procedure applicable in all cases because the

techniques used depend on the goal of the experiment and the

pollutants of concern.  In particular, techniques optimally

suited to study metals may not be not be suitable for organic

compounds (Plumb,  1981; ASTM,  1988b).


A. SEDIMENT COLLECTION AND TRANSPORT

1. Depth of Collection and Sediment Collection Techniques

     The depths from which sediments are collected can affect

bioaccumulation test results;  therefore, a consistent depth

should be used in all collections.  We recommend sampling the

upper 2-3 cm layer, a depth commonly used for toxicity and


                                 48

-------
bioaccumulation  tests  and  sediment  chemistry  (e.g.,  Jenne  et  al.,



1980;  Plumb,  1981; Swartz  et al., 1985a,  1986; NOAA,  1988;



Ferraro et al.,  1990).  Advantages  in  sampling the upper layer



include that  the sediment  is more recently deposited, more



consistent in  pollutant concentrations,  in contact with the



overlying water,  and the most biologically active zone.  However,



tests  on dredge  materials  may require  that representative  samples



be collected  from deeper layers, up to several meters deep, in



areas  intended for dredging.



       To collect intertidal sediment samples by hand, use



shovels, scoops,  spatulas, or coring tubes.  Hand skim or  core



with one of the  above mentioned tools  the upper 2 cm sediment



layer.  To maintain the sample layers  intact, deposit the



sediment sample  into an appropriate container or, if a corer  is



used,  plug the top and bottom of the tube.  Core samples may be



sectioned later  at specific depth-intervals for analytical and



bioaccumulation  tests  (Plumb, 1981; Holme and Mclntyre, 1984;



NOAA,  1988) .



     Box corers  and benthic grabs are  commonly used  to collect



sediments in subtidal waters.  Sampler choice will vary according



to firmness of substrate,  volume of sediment needed, and type of



ship available.  Box corers are the preferred collection device



because they disturb sediment layers the least and retain fine



particles.   A Smith-Mclntyre or modified Van Veen grab, though



more disruptive  to sediment layers  than a box corer, is



acceptable.  Compared to the box corer, these grabs operate in





                                 49

-------
sandier bottoms, are easier to handle, require fewer personnel,



and operate in heavier seas (Plumb, 1981; Holme and Mclntyre,



1984; Tetra Tech, 1986b; NOAA, 1988) .   Scrape surficial sediment



from the grab or box corer samples and immediately store in



appropriate containers.  Consider flocculent material part of the



sample (Lauenstein and Young,  1986) .



     If depth profiles are of interest, the original sediment



layering must be preserved.  Take core samples from the center of



the grab sample once on shipboard and section them vertically at



specific depth intervals (Plumb,  1981).  To minimize oxidation



and changes in other chemical properties, place plastic or



Teflon  bags or containers of appropriate composition and



diameter over the ends of core tubes and extrude samples to



specified depths.



     Construct all collecting equipment with appropriate



materials and clean equipment to reduce the possibility of



contamination.  When organic pollutants are the primary concern,



avoid contaminating sediment with various plastics, especially if



phthalate esters, used in flexible plastics, will be quantified.



When metals are the primary concern, avoid contaminating



sediments with any metal, including stainless steel.  Subsampling



from the center of grabs as well as coating, covering, or lining



equipment with silicone rubber, TeflonR, plastic, polypropylene,



or polyethylene will eliminate direct contact of samples with the



equipment.  For instance, TeflonR bags can be used to cover
                                 50

-------
stainless steel scoops (Lauenstein and Young, 1986).   When both



metal and organic contaminants are the interest, use glass when



possible.



     To remove organics/metal contamination, initially wash the



equipment with a non-phosphate detergent, then consecutively



rinse with distilled water, a water-miscible organic solvent, 5-



10% hydrochloric or nitric acid, and finally deionized-distilled



water  (Plumb, 1981; Lauenstein and Young, 1986; U.S. EPA/U.S.



ACE, 1988).  Bake  (>350°C) all glass equipment  (e.g., jars,



corers, trays) before use.  Glassware to be used in metal



analyses should be stored wrapped in TeflonR sheets or plastic



wrap, whereas glassware to used in organic analyses should be



stored wrapped in TeflonR or aluminum foil.  Wash new plastic



containers and equipment as described and then leach them with



distilled water before use.  Rinsing grabs or corers with



seawater between stations should suffice in most studies, though



it may be necessary to use a brush or a detergent to remove



highly cohesive sediments.  When it is critical to remove all



contaminants, Lauenstein and Young  (1986) recommended rinsing



grabs or corers with methylene chloride, followed by a seawater



rinse.  However, methylene chloride generates a hazardous



substance, and its use on a ship would have to be carefully



controlled so as not to endanger the workers or release a



hazardous waste into the environment.  Safer alternatives are



methanol and ethanol alcohol.
                                 51

-------
     Specifics of the field sampling design, such as the number



of sites and the number of samples per site, depend upon the



goals of the study and the type of spatial resolution required.



Guidance for designing field sampling programs can be found in



Green (1979), Elliott (1983), and Lauenstein and Young  (1986).



2. Field Measurements



     Upon collection, immediately determine sediment temperature



and salinity by inserting a thermometer and an electrode 1 cm



into the center of each sample  (Plumb, 1981; Swartz et al.,



1986).   When metals are of interest, immediately measure pH and



Eh, both of which require undisturbed samples  (Holme and



Mclntyre, 1984).  Important information recorded with each sample



should include the site (name, with latitude and longitude to



tenths of a minute),  replicate number, depth, sampler



description, numbers and kinds of subsamples, sediment



characteristics, temperature, salinity, pH or Eh if measured,



odor-color, penetration depth, sieve size, vessel size,  date and



time, weather conditions,  names of chief scientist and team



members,  and comments (Lauenstein and Young, 1986).



3. Field Storage and Sediment Transport



     Physical,  chemical,  and biological changes in sediment



samples can occur rapidly, resulting in changes in sediment



quality and/or bioavailability during the transport of sediment.



Temperature, pH, and dissolved oxygen are often the rate



controlling factors for these changes (Jennett et al.,  1980).   To



diminish these effects,  store the sediment sample in a bag or jar
                                 52

-------
immediately after collection.  TeflonR containers or brown



borosilicate glass jars with Teflon -lined lids are recommended



for both metal and organic samples, but regular glass jars with



TeflonR-lined lids are acceptable  (Lauenstein and Young, 1986).



Containers need to be completely cleaned and stored in a covered



container to avoid contamination.  Cleaning protocols used for



the sampling equipment also apply to storage containers.



     Fill jars and bags completely with sediment to eliminate



airspace and retard oxidation of metals, but retain as much of



the interstitial water as possible (Lauenstein and Young, 1986;



U.S. EPA/U.S. ACE, 1988).  Refrigerate sample containers in



insulated cartons or ice chests immediately after collection.  To



maintain a temperature near 4°C, provide containers with



prefrozen, jelled refrigerant packs (e.g. Blue IceR)  or ice.



Make sure that samples are protected from the refrigerant to



avoid cross contamination and freezing of the sample.



     Shipping containers must be durable and leak-proof or lined



with two heavy duty plastic bags.  Add adequate absorbent



material to soak up any leaked liquid.  Pack samples tightly,



using dividers between glass containers and fill all empty spaces



with packing material.  Mark containers with "This End Up" and



"Fragile" labels.  Ship samples by "overnight" or "24 Hour"



carrier to the laboratory immediately after completion of



sampling to protect sample quality.  Refrigerate samples at 4C°



upon arrival.  Guidance for shipping hazardous materials can be



found in CFR 49, Parts 100-177.
                                 53

-------
B. SEDIMENT SPIKING AND MANIPULATION



     Besides using field collected sediments, researchers can



experimentally manipulate sediments to test pollutant



bioavailability under defined conditions.  Addition or spiking of



pollutants to sediments is the most frequent type of



manipulation.  Other types of manipulations include the addition



of inert substances for producing a less polluted sediment and



the alteration of sediment organic content or particle size.



Sediment spiking and manipulation techniques have not been



standardized, and until standard methods are developed or the



techniques are intercalibrated,  exercise caution when comparing



results from different techniques.  Because manipulations can



alter properties of sediments, prepare and manipulate control



sediments in the same manner as test sediments.  Several sediment



manipulation techniques are outlined in Appendix V-l.





C. LABORATORY SEDIMENT STORAGE



     Keep the time between the collection and/or spiking of a



sediment and its use in bioassays to a minimum.  If there is a



delay of more than about 24 hours, store the sediment in air-



tight containers at 4°C in the dark.   Because freezing may



physically disrupt sediment,  sediments for biological testing



should not be frozen or freeze-dried (Tetra Tech, I986b),  with



the possible exception of sediment stored for use as a "standard



reference sediment" (Chapter II).   If metals are the major source



of contamination in a field sediment,  the sediment should be
                                 54

-------
stored in the absence of air to minimize oxidation of reduced



forms.  Nitrogen can be used to fill the headspace in the



container.  Air-tight glass containers are recommended for



sediments polluted with either metals or organics (Plumb, 1981;



Lauenstein and Young, 1986).  High density polyethylene and



TeflonR containers are also acceptable.  Remove large organisms



and extraneous material, such as bivalves or twigs,  from the



sediment before storing.



     There is no consensus on maximum storage time other than



that it should be kept to a minimum.  This lack of consensus



reflects the use of different end-points, sediment types, and



storage procedures.  Little information exists storage effects,



though pollutants that are volatile, biodegrade rapidly, or



undergo rapid oxidation-reduction reactions should be the most



prone to changes in concentrations and/or bioavailability.  Given



the present state-of-the-art, the maximum time recommended for



storage of dredge materials used for biological testing is 6



weeks at 4°C  (U.S. EPA/U.S. ACE, 1988), which seems reasonable.



The storage of samples for analytical and physical analysis is



discussed in Chapter XI.





D. SEDIMENT PREPARATION AND HOMOGENIZATION



     Before using a field sediment, remove any extraneous



materials (e.g., macroalgae, twigs, garbage, and rocks) and large



organisms (e.g., bivalves).  Disturb the sediment as little as



possible during this process.  The simplest technique is to
                                 55

-------
gently spread the material out in a glass pan and remove any



large objects with forceps.  However, if metals are of primary



importance, keep contact with air to a minimum and use plastic



tools.



     If a field sediment is going to be experimentally



manipulated (e.g., spiked with a pollutant), sieve the sediment



through a 1-2 mm mesh sieve to remove the extraneous materials.



Using as small a volume of water as possible, sieve the sediment



over a large container  (e.g., garbage pail).  After letting the



suspended fines settle for 6 to 24 hours, carefully siphon off or



decant the overlying water and mix the settled fine particles



back into the sediment.



     After settling or storing sediments, mix them well



immediately before taking aliquots for chemical analysis,



spiking, or using in bioaccumulation tests.  This will assure



homogeneity as well as mix any separated interstitial water back



into the sediment.  Stir with a spoon or rod made of an



appropriate material.  If grab samples were divided into several



containers, mix the respective sediment samples together before



sampling or using in biological tests.  Large masses of sediment



can be manually mixed in an appropriately cleaned glass tray or



plastic tub, or placed in jars and rotated on a rolling mill.



Homogenize control sediments in the same manner as test



sediments.



     Visually inspect the sediment to judge the extent of



homogeneity.  Excess water on the surface of the sediment can
                                 56

-------
indicate separation of solid and liquid components.  If a



quantitative measure of homogeneity is required, take replicate



(3-8)  subsamples from the sediment batch and analyze for TOC,



chemical concentrations, or the percent fines.



     Jenne et al.  (1980) cautioned against prolonged stirring



which can abrade floes and change the sediment's physicochemical



properties, such as dissolved organic matter  (DOM).  However, all



changes to the sediment are probably impossible to avoid.  Recent



results suggest that even stirring can increase interstitial



water DOM concentrations (DeWitt, T. pers. comm. OSU, Mar. Sci.



Ctr.,  Newport, OR).
                                 57

-------
              CHAPTER VI. SEDIMENT CHARACTERIZATION






     Bioavailability of sediment-associated pollutants depends on



the physical and chemical composition of the sediment, which is



often specific to a particular site and may change due to



experimental manipulations (Gambrell et al.,  1980; Jennett et



al.,  1980).  Therefore, always measure the parameters potentially



affecting bioavailability: pollutant concentrations, grain size



distributions, organic carbon, and total solids contents  (Plumb,



1981; NOAA, 1988).   A number of other sediment and interstitial



water measurements such as pH, Eh, cation exchange, and acid



volatile sulfides will aid interpretation of the results,



especially those of metals.






 A.  GRAIN SIZE ANALYSIS



     Grain size analysis is the measure of the frequency and



distribution of the disaggregated mineral particles comprising



the sediment.  Distributions are commonly reported on the



Wentworth scale, which classifies particles as coarse sand,



medium sand, fine sand, very fine sand, silt, and clay (Folk,



1980, Holme and Mclntyre, 1984) (Table VI-1).  Particle sizes are



either expressed in millimeters or on a phi scale, where phi =



-Iog2 particle diameter in millimeters.  Quantification of the



fine fraction (silt-clays <0.0625 mm or > 4 phi) is important



because pollutants predominantly associate with, and many



deposit-feeders ingest, this fraction.
                                 58

-------
TABLE VI-1: Wentworth Grade Classification of Sediment
                                 GRADE  LIMITS
    NAME
                            mm               Phi
     Coarse  sand       1.000  -  0.500         +1.0

     Medium  sand       0.500  -  0.250         +2.0

     Fine  sand          0.250  -  0.125         +3.0

     Very  fine  sand    0.125  -  0.062         +4.0

     Silt               0.062  -  0.004         +8.0

     Clay               <  0.004               >8 . 0
                           59

-------
     The sand fraction of a sediment is determined by first



drying the sediment  (air dried or at <105°C) and then removing



the organic matter with 6% hydrogen peroxide  (Buchanan, 1984) .  A



known mass of dry sediment is then sieved through a standard



series of sieves graded on the Wentworth scale  (e.g., <.500,



<.250, <.125, <.0625 mm, and pan) or on the phi scale  (e.g., at



1/2 phi intervals) (Buchanan, 1984; Holme and Mclntyre, 1984).



Manually or mechanically shake about 25 grams of dry sediment for



about 20 minutes and then weigh the contents of each sieve.  To



further characterize the silt-clay fraction  (i.e, the sediment in



the bottom pan), an hydrometer and pipette analysis is used



(Jenne et al.,  1980;  Buchanan,  1984; Holme and Mclntyre, 1984).



Suspend the pan fraction in a graduated cylinder of



deionized/distilled water, take periodic water samples from a



known depth, and weigh the resultant suspended particles.  The



weight of the silt-clay fractions can then be calculated from the



estimated settling velocities of particles of different sizes.





B. TOTAL SOLIDS CONTENT



     Sediments are composed of both solids and interstitial water



(IW),  with the relative proportion of the two phases varying due



to physical factors  (e.g., percent sand) and biological factors



(e.g., intensity of bioturbation).   The total solids content, the



percent of wet sediment comprised of particles,  is used to



convert sediment pollutant concentrations from wet to dry weights



and to record changes in the ratio of water to sediment which can



cause desorption of pollutants from sediment (Plumb,  1981).






                                 60

-------
Measure total solids content by drying a wet sediment  sample  at



105°C to a constant dry weight.  The percent solids  content is



the dry weight divided by the initial wet weight multiplied by



100  (APHA, 1985) .  The percent moisture content of a sediment is



100% less the percent solids.






C. ORGANIC CARBON




     Organic carbon content is one of the primary factors



regulating sediment bioavailability  (e.g., Rubinstein  et al.,



1983); therefore, measuring organic content on every sediment



tested is critical.  One direct measure of organic matter  is



total organic carbon (TOO, a measure of the total amount  of



oxidizable organic carbon.  In comparison to TOG, total carbon



(TC) is the measure of both the volatile organic and inorganic



nonvolatile materials,  such as carbonates and bicarbonates  (U.S.



EPA/U.S. ACE, 1988).  Because of its relation to the binding  of



organic pollutants, TOC is the more biologically relevant



measure.



     Both TOC and TC can be determined by wet or dry (combustive)



oxidation techniques (Plumb, 1981; Holme and Mclntyre,  1984).



Oxidation of an untreated (no acidification) sediment measures



total carbon; whereas,  oxidation of an acid treated  sediment



measures TOC (Plumb, 1981).  The common wet methods  use a  chromic



acid oxidation technique developed by Walkley and Black (1934;



also see Holme and Mclntyre, 1984) and el Wakell and Riley



(1956).   Buchanan and Longbottom  (1970)  describe a technique  to
                                 61

-------
use on sediments containing coal.  The differential combustion
method, used in commercially available carbon analyzers, oxidizes
the carbon with heat.  The amount of C00 generated at different
                                       £t
temperatures is a measure of the total and organic carbon.
Carbon analyzers are simpler to use than wet techniques and are
recommended.  However, exercise care in using or purchasing
carbon analyzers because instruments designed for measuring TOC
in seawater may be too sensitive for sediments, requiring
extensive dilution of the samples.
     A commonly used method to estimate organic matter is loss on
ignition (LOI)  or total volatile solids (TVS),  which is the
percent loss of weight after combustion.  The organic matter is
combusted by heating dry sediment at 550°C for 1-4 hours.  LOI is
calculated as the difference between the dried and combusted
sample weights divided by the dry weight,  multiplied by 100 to
convert the number to a percent  (Dean, 1974; Byers et al.,  1978;
Plumb, 1981; APHA,  1985).   Structurally bound water and
carbonates, found particularly in clays and in calcareous sands,
may be lost along with volatile solids during combustion,
distorting actual organic carbon values (Dean,  1974).  Because
LOI includes these non-organic constituents, it is not a direct
measure of sediment organic carbon content.
     When determining LOI,  spread a small aliquot (a few grams)
of sediment out thinly in a porcelain ignition dish at least 75
mm in diameter.  This exposes the sediment to an adequate oxygen
supply for oxidation of the organics.  Clumps of sediment and/or
                                 62

-------
large amounts of sediment have a tendency to block oxygen



transfer to the center of the clump, thereby yielding



significantly lower and/or inconsistent LOI results.  We analyze



2g samples of sediments and ignite the dry total solids residue



for 1.5 hours at 550°C.



     Of the two methods, LOI is simpler and cheaper than TOG.



However, most research and regulatory applications require a more



accurate measure than LOI (see Dean, 1974; Byers et al., 1978;



Mook and Hoskin, 1982).  Because TOC is one of the parameters



used to calculate Accumulation Factors (see Appendix 1-1), we



recommend using LOI as a rapid survey method, but performing TOC



analyses on all sediments used in bioaccumulation tests.  If



analyzing TOC on all sediments is impossible, determine the



conversion between LOI and TOC on a few samples or use .the values



in Table VI-2.  However, use caution when using these conversion



factors because the relationship between LOI and TOC can vary



several fold among sediments (Dean, 1974;  Byers et al., 1978;



Ditsworth, G., pers. comm.,  U.S. EPA, Mar. Sci. Ctr., Newport,



OR) .





D. ADDITIONAL SEDIMENT CHARACTERISTICS



     Measure salinity in the overlying and interstitial water



during the initial characterization of sediments in the field.  A



refractometer measures salinity with sufficient accuracy in most



cases and requires only a few drops of interstitial water.  In



studies of metal bioavailability, measure pH and Eh, both of
                                 63

-------
TABLE VI-2:  Ratios For Converting Loss on Ignition  (LOI) to Total
               Organic Carbon  (TOG) For Various Sediment Types.
Particle size
%>63um %<63um
36
med.
63.5
34.1
50.2
64
silt
36.0
65.9
49.8
TOG
1
10
1.21
2.29
2.10
LOI
2.
22
4.
7.
6.
13
.9
30
79
53
TOC/LOI
0
0
0
0
0
.47
.44
.28
.29
.30
Reference
Dean, 1974
Byers et al . , 1978
unpubl . data
unpubl . data
Ditsworth , pers . comm.
(Mean of 12 sediments)
19-81   19-81   0.53-3.50 2.43-10.10  .21-.37  Ditsworth1, pers. comm.
                                               (Range of 12 sediments)
1 G.  Ditworth,  U.S.EPA, Mar. Sci. Ctr.,  Newport, OR
                                 64

-------
which can affect pollutant solubility and mobilization  (Engler,



1980).  Measure these parameters during  field  collection,



repeating the measurements after handling and  storage during



which values can change rapidly.  Use a  meter  or probe, and a



platinum electrode, to measure pH and Eh, respectively.   Insert



the probes 1-3 cm below the sediment surface and allow  to



stabilize  (Plumb, 1981).  Discussion of  pH and Eh can be  found in



Fenchel  (1969) and Pearson and Stanley  (1979).



     Additional characteristics potentially affecting metal



availability are amorphic oxides of iron and manganese, iron



sulfides, cation exchange sites, and selectively extractable



fractions using extractants such as hydrogen peroxide,  acetic



acid, and organic chelates (Luoma and Jenne, 1976; Jennett et



al., 1980; Plumb, 1981).  Consider measuring these parameters in



studies of metal bioavailability.  Recent work suggests that acid



volatile sulfides (AVS) may be the primary factor regulating the



bioavailable fraction of many metals (DiToro et al., in review).



If further investigations support the initial  findings, it may be



possible to use AVS as a "normalizer" for sediment metals much as



TOC is used for neutral organics.  Morse et al. (1987) gives a



method to measure AVS.





E. INTERSTITIAL WATER



     Interstitial water is an integral part of bedded sediments



and pollutants associated with the interstitial water may play a



major role in the uptake and toxicity of a number of compounds



(Adams et al.,  1980;  Swartz et al.,  1988 ;  Landrum 1989).





                                65

-------
Because of the potential role of pore water in controlling uptake



and toxicity, its characterization can be very informative.



"Dissolved" pollutant concentrations, complexed pollutant



concentrations, and dissolved organic matter  (DOM) are the most



important parameters to measure.



     A number of techniques have been used to collect



interstitial water for pollutant or nutrient analysis, including



squeezing of sediment cores (Presley et al., 1967; Robbins,



1977), centrifugation (Edmunds and Bath, 1976; Plumb, 1981;



Landrum et al., 1984), suctioning (Plumb,1981; Knezovitch and



Harrison, 1987),  equilibration with dialysis membranes (Mayer,



1976)  or porous TeflonR cups  (Zimmermann et al.,  1978), and



displacement with other liquids  (Bately and Giles, 1980).



Manheim  (1974)  published a comparative study and Bately and Giles



(1980) reviewed several methods.



     Centrifugation is a straight-forward technique suitable for



the routine collection of small to moderate amounts of



interstitial water for pollutant analysis.  There is no standard



procedure, but centrifuging the sediment sample at 7000 - 9000



rpm for 5-10 minutes should suffice in most cases  (Bately and



Giles, 1980; Plumb,  1981).   Higher rpms are required if any



suspended particles remain, because particulate matter in the



supernatant will result in erroneously high "dissolved" pollutant



concentrations.  If the speciation of metals will be examined,



the centrifuge tubes should be oxygen-free.   Following



centrifugation, the supernatant is often vacuum filtered to






                                 66

-------
remove particulate matter.  There is no standard pore size, but a



0.45 urn NucleporeR membrane filter is frequently used (U.S.EPA,



1979; Plumb, 1981; Landrum et al.,  1984).  The filters should be



precleaned in 10% nitric acid for 24 hours and then soaked in



deionized water for another 24 hours (Lapan, R. pers. comm.



U.S.EPA, ERL-N, Narragansett, RI)  A potential source of error is



sorption of dissolved pollutants onto the filter, resulting in an



underestimation of dissolved pollutant concentrations.  Measure



the sorption of the pollutants onto the filter by determining the



loss of pollutants from standard solutions passed through the



filter.



     Pollutants in the interstitial water may either be truly



dissolved or complexed with DOM  (Engler, 1980) .  In general, the



complexed pollutants have a lower bioavailability than the free



forms  (Luoma and Bryan, 1978; Jenne et al. 1980)  .  As DOM levels



may change due to handling (DeWitt, T., pers. comm., OSU, Mar.



Sci. Ctr., Newport, OR) or biotic effects (e.g.,  excretion of



organics), the bioavailability of interstitial water pollutants



may change over the course of an experiment or vary between field



and laboratory sediments.  The concentration of DOM can be



determined either by using wet oxidation methods (Walkley and



Black, 1934; Holme and Mclntyre, 1984)  or a carbon analyzer.  As



mentioned, carbon analyzers that measure the relatively low



concentrations of DOM in interstitial water with good accuracy



are not as efficient at measuring the high concentrations of TOC



in sediment.
                                 67

-------
        Several techniques have been developed  to measure  the



concentrations of the dissolved and complexed pollutants in



interstitial water.  A relatively simple technique has been



developed by Landrum et al.  (1984) to separate  and quantify



total, freely-dissolved, and colloidally-bound  non-polar organic



pollutants in interstitial water.  This technique uses a Sep Pak



C-18R cartridge to separate the bound and dissolved fractions.



The bound fraction of the pollutant will pass through, while the



dissolved fraction will be retained by the column.



     Besides direct measurement, equilibrium interstitial water



pollutant concentrations can be calculated from the TOG, bulk



sediment concentration, and Koc for a compound  (Karickhoff,



1984).  Because all the aforementioned collection techniques may



alter interstitial water characteristics through adsorption/



desorption of pollutants or suspension of particulate matter



(Word et al., 1987), the calculated value for interstitial



equilibrium concentrations may be more accurate for very high Kow



compounds (Karickhoff,  S.,  pers. comm.,  U.S. EPA,  ERL-A, Athens,



GA).   Calculated concentration values will not monitor temporal



changes in interstitial water but will serve as check on directly



measured concentrations.   Measured concentrations substantially



higher than calculated could indicate suspended particles;



whereas,  measured concentrations lower than expected could



indicate either that equilibrium had not been attained or that



the pollutant had adsorbed to the filter.
                                 68

-------
                 CHAPTER VII. ORGANISM SELECTION






A. SELECTION CRITERIA



     The choice of the test species can have a major influence on



the success, ecological significance, and interpretability of a



bioaccumulation test.  Given the potential range in environmental



characteristics, no one species is best suited for all



conditions.  There are, however, two characteristics required of



any bioaccumulation test species, as well as a number of other



desirable characteristics.  These characteristics are summarized



in this chapter and Table VII-1 and discussed in more detail in



Appendix VII-1.



     The first required criterion is that the test species ingest



sediment.   This requirement is critical because recent work has



demonstrated that ingested sediment is the major uptake route for



higher Row compounds for some species  (Landrum, 1989; Boese et



al.,  in press).  Many benthic invertebrates are flexible in their



feeding mode, and this requirement does not preclude the use of



facultative filter-feeders (e.g., Macoma) as long as the only



exposure route during the experiment is from bedded sediment



(i.e, no resuspended particles or phytoplankton).  The second



required attribute is that the test species be sufficiently



pollutant resistant to survive the duration of the exposure with



a minimum level of mortality.  This requirement precludes the



species routinely used in sediment toxicity testing  (e.g.,



Rheooxvnius).
                                 69

-------
     Besides the required criteria,  there are a number of



desirable  characteristics -- ease of collection,  year-round



availability, ability to culture the species, adaptability to



laboratory conditions,  suitable size, tolerance to a wide range



of sediment types and salinities,  suitability for sublethal or



toxicokinetic tests, ecological or economic importance, having a



high bioaccumulation potential, compatibility with other species,



and a low capability of metabolizing PAHs and other contaminants.



The importance of these various criteria depends upon the



specific goals of the research and the sediment tested.  However,



using an organism large enough to supply sufficient biomass for



chemical analysis is important in nearly all cases.  Ideally, the



test species should be large enough to allow chemical analysis on



individuals.  Even when individuals are composited, compositing a



smaller number of larger organisms is easier than dozens or



hundreds of smaller specimens.





B. RECOMMENDED SPECIES:



     An evaluation of the suitability of potential test species



is summarized in Table VII-1.  This evaluation is not based on



extensive comparative studies and should be considered a guide



rather than a definitive characterization of the species.  Based



on this analysis, we identified five recommended bioaccumulation



test species and another eight "secondary" taxa.  The recommended



species meet all or nearly all of the desired criteria and are



well established as bioaccumulation test species.  The recommended
                                 70

-------
                                 TABLE VII-1
           PERTINENT CHARACTERISITCS OF TEST SPECIES
SPECIES
Abarenicola spp.
Arenicola spp.
Callianassa spp.
Capitella spp.
*
Macoma balthica
*
Macoma nasuta
Nephtys incisa
*
Neanthes arenaceodentata
Nereis virens
Nereis diversicolor
Nucula spp.
Palaemonetes pugio
Yoldia limatula *
Feeding
Type
Fun
Fun
SSDF
SDF
SDF
SDF
SSDF
SDF/0
SDF/0
SDF/0
SSDF
SDF
SSDF
Biomass o Pollution Culture Commercial Bio.
S /«"> Tolerance Potential Availability lnfo
4-+ >is 4- - -
4-4- >is 4- -f
4-4- >i° -? - 4-
- >io ++ + +
4- >io 4- ~~ "~
+ + >io + - ~
+ >25 4. ~~ —
+ ? >28 4- 4- 4- +
4-4- >io ++ - +
4- + >10 4-4- - +
+ ? -4- — —
4-? >io - ? + 4-
4- >25 ? + - -
i
~r

4-4-
44-
4 +
4-
4-4
4-4-
44-
+
4-4-
4-
Fun = Funnel feeder
0 = Omnivore
SDF  = Surface Deposit Feeder
SSDF = Subsurface Deposit Feeder
Recommended test species
+ =  good, sufficient
++ =  very good
- =  poor, insufficient

Bio Info. = information on bioaccumulation toxicity

-------
species are  the polychaetes Nereis diversicolor. Neanthes  (Nereis)



virens, and  the bivalves Macoma nasuta. Macoma balthica. and Yoldia



limatula.  These species have been used in a  substantial number  of



experimental bioaccumulation studies and  in regulatory monitoring.



Within their tolerance levels, these species  should serve as



suitable test species, and we recommend using at least one of these



species in all tests, at least until the  suitability of other



species has been demonstrated locally.




     The secondary bioaccumulation species meet the required



characteristics but are to some extent deficient in one or more



of the important desired characteristics  and/or there is



insufficient information to make a final  evaluation.  However,



some of these secondary taxa offer potential  advantages such as



large size (arenicolid worms), additional phylogenetic groups



(i.e., crustaceans), adaptability to culturing (e.g.,  Neanthes



arenaceodentata).  and high pollution tolerance (Capitella spp.).



The importance of these various advantages  depend upon the site



specific situation  (e.g.,  level of toxicity of sediment).





C. NUMBER OF SPECIES TESTED AND MULTIPLE SPECIES TESTS



     Species as well as larger phylogenetic groups vary in their



tendencies to bioaccumulate pollutants both in response to their



modes of exposure and to their metabolic characteristics.  The



extent of these interspecific variations are not well  understood,



and both the magnitude and direction of species differences can



vary with pollutant (e.g.,  metals vs.  organics)  and perhaps with
                                 72

-------
sediment type.  Thus, utilization of two or more species from



different major taxa increases the probability of accurately



assessing the maximum field tissue residues.



     The actual number of species and taxa used depends upon the



goals and scale of the project and the range of pollutants in the



sediment.  In general, use of a single species should be adequate



for a general survey of an area or assessing a small discharge or



volume of dredge material.   In interpreting the data from a



single species test, however, it should be recognized that no one



species is likely to maximize uptake from all pollutants.  Two



species are recommended when assessing a moderate to large sized



discharge or dredging operation.  The species should be of



different major taxa, and a polychaete and a bivalve are



recommended.  It is especially important to include a bivalve if



PAH contamination is of concern, as bivalves have a reduced



capability to metabolize PAHs compared to amphipods or



polychaetes  (Varanasi, et al. 1985).  The addition of an



arthropod species or additional polychaete and/or bivalve species



may be justified when assessing a large discharge or dredging



operation, especially if there is a wide range of pollutants.
                                 73

-------
    CHAPTER VIII:  ORGANISM COLLECTION,  MAINTENANCE,  TRANSPORT,
                        AND ACCEPTABILITY


     To assure unbiased test results, bioassay organisms must be

in good health and have minimal background pollutant

contamination.  Reasonable efforts must be taken to minimize

stress during collection and transport to holding facilities.

Holding facilities must provide high quality water and conditions

suitable for the maintenance of the test species.  This chapter

describes techniques and facilities which will meet these general

requirements.


A. ORGANISM COLLECTION AND TRANSPORT

1. Field Collection of Test Organisms

     The logistics of collecting intertidal species is usually

much simpler than those of collecting subtidal species, and

intertidal collection is recommended when possible.  Infaunal

organisms can be collected by turning the sediment over with a

shovel and picking out larger species  (e.g.,  clams) or by gently

sieving the sediment in the field.  For most of the

bioaccumulation test species,  a sieve size of 4-6 mm will collect

adequate numbers while minimizing damage and sorting time.

Collection equipment should not have been used in contaminated

sites or should have been adequately cleaned.

     Subtidal organisms can be collected by grabs, dredges,  or

suction samplers (see Holme and Mclntyre, 1984) .   Dredges sample

a larger area than grabs and are usually more proficient at

collecting shallow-buried organisms, though there is a greater


                                 74

-------
possibility of damaging specimens.  Grabs are recommended for



collecting more deeply buried species.  Suction lifts are also



useful for collecting larger, deeply buried bivalves, though they



require the use of SCUBA divers, and a greater likelihood of



damage exists.  Electro-shocking, chemical poisons, and other



harsh methods of collection are not acceptable.



     Remove organisms from the collection device as soon as



possible and submerge them in ambient seawater or sediment



contained in ice chests or uncontaminated plastic buckets.  Avoid



overcrowding animals in collection containers.  Discard organisms



with signs of disease or obvious defects  (e.g., bivalves with



cracked shells).



     State or local authorities may require collection permits or



ban collection  from specified areas.  Collection of regulated



species  (e.g.,  bay clams) may require a local license, be limited



to a season, and preclude certain collection techniques.



Additional permits or precautions may be  required when importing



non-indigenous  species.  Check with State authorities about the



local regulations before collecting or importing specimens.



2. Organism Transport



     For organism maintenance, ASTM  (1984) recommends not more



than a 3°C change in water temperature within a 12 hour period



and an oxygen concentration  of between 60 and 100% of saturation.



If the time between collection and return to laboratory is short



 (less than 1-2  hours) and ambient temperature is not extreme,



simple precautions should meet  these  requirements.   If possible,





                                 75

-------
collection buckets or  ice-chests  should be kept  out of direct



sunlight and must not  be left in  closed vehicles.  The water in  the



containers should be changed periodically while  collecting and



immediately before returning.  If the time before returning to the



laboratory is several  hours and/or air temperature is high, use  a



portable aerator to maintain oxygen levels.



     Successful long-distance transport of organisms, whether in



a vehicle or through the mail, requires packaging that retains



moisture and maintains an adequate supply of oxygen.  This can be



accomplished by placing animals in a minimum amount of water (a



few milliliters) in a  sealed container filled with air (e.g.,



Whirl-PakR bag).  Alternatively, test animals may be placed



between wet nylon or seagrass (e. g., Zostera) and surrounded by



layers of wet paper towels,  all contained in polyethylene



ZiplockR  bags  (Robinson, A., pers. comm. AScI, Mar. Sci. Ctr.,



Newport, OR; Gulf Specimen Co.,  pers. comm.,  Panacea,  FL).   Wet



sediments may also be used to retain moisture.  These sediments



should have a low organic content (e.g.,  ashed sediment,  beach



sand) as they are not as likely as natural sediment to turn



anoxic.  Regardless of the moisture retaining agent,  the



container should have a large air space to maintain aerobic



conditions.   Air trapped within a plastic bag has the added



advantage in preventing animals from being crushed.



     Containers  with organisms should be placed in ice chests or



insulated shipping containers,  with packets of jelled (e. g.,



Blue IceR)  refrigerant placed at or taped to the inside of  the  top






                                 76

-------
of the container.  Jelled refrigerants are preferred to ice to



avoid melt water and a layer of insulating material should be



placed between the refrigerant and animals.  Add sufficient



refrigerants to maintain the water temperature in the containers at



or a few degrees below the water temperature at the collection site



but not so much as to cold shock the organisms.  Insulating



material should fill all extra space in the shipping container,



protecting and securing the bottles and bags in the carton.  Pack



shipment containers to obtain a low center of gravity, and label



plainly to keep package upright.  Every effort should be made to



provide overnight or 24-hour delivery.  If the organisms are



transported by vehicle, periodically monitor the temperature and



drain any melt water and replace the ice as required.





B. CULTURING AND PURCHASING TEST ORGANISMS



     A successful culture of an appropriate test species has the



advantages of providing a ready supply of specimens with a known



history.  However, culturing of marine/estuarine organisms is not



a task to be undertaken lightly, and is usually justified only if



regular tests are planned.  Although a few sediment ingesting



polychaetes can be cultured with relatively simple equipment (see



Appendix VII-1), the majority of recommended test species are not



routinely cultured.



     Some test organisms can be purchased from biological supply



houses, local collectors, colleges, or bait shops.  Purchasing



specimens can be cost-effective if the laboratory is not well
                                 77

-------
equipped for collection  (e.g., does not have a boat) or if tests



will be conducted only occasionally.  There are several companies



that specialize in supplying bioassay organisms (see Appendix



VIII-1),  although most do not presently supply appropriate



benthic bioaccumulation organisms on a routine basis.  Check with



a supplier even if bioaccumulation test species are not currently



carried as availability of particular species may change or the



supplier may be able to fill special orders.



     Maintain purchased organisms in the laboratory for at least



one week to acclimate them to local conditions and to monitor



their health.  Before beginning bioaccumulation tests, analyze



the purchased organisms for background pollutant levels to



determine if they meet the criteria for control organisms  (see



Table VIII-1).






C. PRE-EXPERIMENTAL MAINTENANCE



     In general, the guidelines presented here are based on our



experience with Macoma nasuta or modified for deposit-feeders



from ASTM Standard Procedures (ASTM 1980, 1984) for maintaining



filter-feeding bivalves for bioconcentration tests.  Most of the



bioaccumulation test species are adaptable to laboratory



conditions,  so elaborate procedures are not usually required for



the maintenance of adults.  Additional information on the



maintenance of benthic invertebrates can be found in King and



Spotte (1974),  Dean and Mazurkiewicz (1975), Kinne  (1976-1977), -



and National Research Council Committee on Marine Invertebrates



(1981).






                                 78

-------
     When maintaining a non-indigenous species,  permits may be



required from state or local authorities.   This may require fail-



safe precautions against accidental release of such organisms



into the local environment  (i.e., double containment, diked



seawater drains, siphon breaks, etc.).  Equipment, water, wastes,



and dead animals may require sterilization before disposal.



1. Water Quality



     Constant water quality should be maintained in holding



aquaria, keeping dissolved oxygen between 60-100% saturation and



un-ionized ammonia concentrations <20 ug/L (ASTM, 1984).  Flow-



through seawater with a minimum flow-rate of 1 L/hr per gram of



wet tissue is recommended as a means of maintaining water



quality.  Filtering incoming seawater is generally inadvisable



with benthic invertebrates as the settling particles supply a



natural source of food.  If flowing seawater is unavailable,



organisms can be maintained in static systems using collected



seawater.  Store replacement seawater in covered containers in



the dark at 4°C to keep salinity and water quality constant.  If



collection of natural seawater is impractical, artificial



seawater may be used, though it should be demonstrated that the



growth and behavior of the  test species is not altered by using



artificial salts.  Prepare  artificial seawater with deionized



water or with distilled and charcoal-filtered water.  Static



systems should always be aerated, and preferably equipped with a



recirculating aquarium filter with replaceable activated



charcoal.





                                 79

-------
     Regardless of whether flow-through or static systems are



used, the seawater should be analyzed for background levels of



pollutants, especially if it is collected from an urbanized area.



If a pollutant is detected in the water, its potential uptake can



be estimated by multiplying the water concentration by the



bioconcentration factor  (BCF) for that compound.  If the



calculated tissue residue is greater than that acceptable for a



control organism (Table VIII-1), a different water supply is



required.  BCF values and methods to estimate BCFs can be found



in Bysshe et al. (1982).




     Temperature should not vary more than 3°C in a 12 hour



period and salinity should not vary by more than 2 g/kg or 20% of



the average, whichever is larger (ASTM,  1984).  In flow-through



systems, a storage tank within the laboratory will help



ameliorate natural fluctuations in temperature.  In estuarine



areas,  a storage tank may be necessary to supply high salinity



water during low salinity periods.



2.  Sediment Quality



     Maintain animals in a sufficient amount of clean sediment to



allow them to burrow naturally, which in nearly all cases will be



at  least 3 cm.  This sediment must  be analyzed for pollutant



concentrations,  which should not exceed the level acceptable for



a control sediment (Table II-1).   Periodically add fresh sediment



of  the same type to maintain an adequate food supply (i.e.,



detritus and associated microbes).   One  to three times a week,



add about two millimeters of fresh  control sediment to the
                                 80

-------
sediment surface.  This sediment replenishment should be



sufficient if the organisms are not over-crowded.  If the



sediments become heavily pelletized with fecal material, remove



the organisms and replace the sediment.  The addition of other



types of food are not recommended except in special cases of



long-term maintenance.  These foods include detritus or decaying



seaweeds, cultured marine phytoplankton and zooplankton,



microencapsulated diets, formulated feeds such as fish flakes



(e.g., TetraMinR), or small bits of clam or other tissues for



omnivores (Lee and Muller, 1972).  Check the background pollutant



levels of all foods.



3. Organism Health And Acclimation



     Field collected organisms should be held in the laboratory



for at least four days before commencing a bioassay,  and



purchased organisms held for at least a week.  Discard any



organism if injured or behaving abnormally.  In general, animals



should not be held longer than two weeks before testing.  If



longer maintenance periods are needed, the investigators should



have experience with the species and should monitor for any signs



of stress (e.g., reduced sediment processing rate, unusual tube



construction).   A flow-through system is strongly advised if



long-term maintenance is planned.



     To avoid the spread of diseases,  organisms collected more



than a week apart should be maintained in separate aquaria,  each



with an independent water supply.  The organisms should be



checked every day or so, and any diseased,  dying, and dead





                                 81

-------
organisms promptly removed.  Black spots on the surface of the



sediment can mark the location of dead organisms.  Should a



question arise concerning the health of the animals,  a behavioral



test such as time to rebury or analysis of lipid content is



recommended.



     If holding and experimental conditions are different,



gradually acclimate the test organisms to experimental



conditions.  This transition may be accomplished using serial



water dilutions until the proper temperature, salinity, and pH



are reached.  Acclimation for temperature should proceed no



faster than 3°C in 72 hr (ASTM, 1984).  Maintain animals at the



test temperature and salinity for at least two days before



commencement of an experiment.  No more than 3% mortality is



permitted within 48 hr before the test  (ASTM, 1984).





D. ORGANISM ACCEPTABILITY AND BACKGROUND CONTAMINANT LEVELS



     Specimens selected for a test must tolerate the physical-



chemical conditions  (e.g.,  TOC content, interstitial salinity) of



the test substrate and be free of disease or stress from capture



or handling.  All specimens should be collected from the same



site, and preferably at the same time.  All organisms used in a



given test should be as uniform in age and size as possible, and



bivalves should be of the same year class.  For bioconcentration



tests, ASTM  (1984) stipulates that the length  (umbo to distal



valve) of the largest clam should be no greater than 1.5 times



larger than the smallest clam.  For Macoma nasuta. this would be
                                 82

-------
equivalent to a 3.7-fold difference in wet flesh weight.  It is



important to correctly identify the test species, and voucher



specimens should be kept from each collection.



     High pollutant background levels in the test specimens may



confound the results, making it difficult to detect differences



between treatments.  Therefore, tissue residues in the test



organisms should be no greater than those expected in organisms



living in control sediment (see Chapter II).  Approximate



acceptable background tissue concentrations for test species are



given in Table VIII-1.  These values are from organisms collected



from sites which appeared to meet the criteria for a control



site.  For compounds not listed in Table VIII-1, the ASTM (1984)



criterion of the background tissue residue not exceeding 10% of



the expected steady-state can be applied.  First-order estimates



of steady-state tissue residues can be obtained from data on



other species or from the thermodynamic-based bioaccumulation



model for neutral organics (see Appendix 1-1).
                                 83

-------
 TABLE VIII-1:   Representative Control Organism Tissue Residues
COMPOUND
Organics
(ppb wet wt.)
             Various East
             Coast Sitesa
Puget Sound*3   Yaquina Bay, ORC
CB
B(ibk)F
BaP
DDT
HCB
Naph
PAH
PCB
pesticides
<1.0-70

0.3-6.0*
<0.08-3.8
0.02-0.17
<1. 0-9.1
0.02-7.2
10-70
<0.03-0.6

<10
2.3-<10*
<1.0-<5.0
<130
<0.05
<2-17*
<2.0-10



1.9
3.9





Metals
(ppm wet wt.)

Ag
As
Cd
Cr
Cu
Hg
Ni
Pb
Zn
             Various East
             Coast Sitesa

              0.2-2.6
              1.5-3.9
            <0.06-4.0
             0.26-2.5
              0.1-7.2
            <0.05-1.2
             <0.4-7.0
             <0.6-2.6
              2.4-30
Puget Sound"   Yaquina Bay, ORC
                  <0.005
   1.0
                  <2.0
a
b
c
*
Tetra Tech, 1985a
Konasewich et al.,  1982
unpublished data
Tetra Tech, 1985a and Konasewich et al.,
             1982
CB - Chlorinated benzenes
B(ibk)F  - Benzo(i,b,k)fluoranthene
BaP  - Benzo(a)pyrene
HCB  - Hexachlorobenzene
Naph - Naphthalene
PAH  - Polyaromatic hydrocarbons
PCB  - Polychlorinated biphenyls
                                 84

-------
              CHAPTER IX. SEDIMENT EXPOSURE SYSTEMS





     In designing the exposure systems, considerations for



maintaining an adequate environment for the test organisms must



be combined with considerations for pollutant behavior, costs of



construction and maintenance, and ease of operation.  The



recommendations made here are based on the assumption that each



"exposure chamber" will hold a single experimental unit (e.g.,



individual organism or composite of a single species in a



beaker),  but that these may be placed into a larger aquarium or



tank to maintain water quality.  The "exposure system" is



composed of the replicate exposure chambers, any aquaria or tanks



which hold the exposure chambers, the water delivery system, and



any pollution abatement system.  The recommendations are also



based on the standard 28-day exposure duration.  Discussion of



specialized exposure chambers is limited to Appendix IX-1.






A. SYSTEM REQUIREMENTS



     In designing the exposure system, two critical



considerations are the amount of sediment per individual and the



volume and turnover rate of water.  Adequate amounts of sediment



and overlying water are required to assure that supplies of food



and pollutants are not substantially depleted, water column



oxygen concentration is not depressed, metabolites are diluted,



and the organism's feeding behavior is not impaired.  The ASTM



criteria for bioconcentration  (ASTM, 1984) and toxicity tests



(ASTM, 1980) offer general guidance to the design of a sediment
                                 85

-------
exposure system, but several of their specific recommendations



had to be modified because of the different requirements for



deposit-feeders compared to fish and filter-feeders.



Additionally, the design criteria for maintaining a constant



pollutant concentration in the water are inapplicable as the



sediment instead of water is the exposure mode in these tests.



1. Sediment Requirements



     Sediment serves as the habitat, source of food, and source



of pollutants for the test organisms.  If insufficient sediment



is added, organisms may reingest the same particles.



Alternatively, if the fecal pellets are resistant to breakdown,



there may be a reduction in the appropriately sized particles,



especially with the more selective deposit-feeders.  Both of



these processes could reduce the mass of pollutant available to



the test organisms.  Although both reingestion and pelletization



of sediments occurs in field (see Lee and Swartz, 1980), the



rates may be exaggerated in laboratory systems.  For this reason,



it is critical to supply a sufficient sediment mass during the



entire course of the laboratory exposure.



     Assuming periodic sediment additions to the exposure



chambers (see Chapter X),  we recommend initially adding at least



50 grams wet sediment for each 1 gram wet flesh tissue  (excluding



shell) for surface deposit-feeding bivalves.  For funnel-feeders



such as arenicolid worms,  at least 200 grams of wet sediment to



each 1 gram of wet flesh tissue may be required for construction



of a normal feeding burrow.  Besides the mass of sediment, the
                                 86

-------
sediment must be deep enough to allow normal burying and feeding.



The initial depth for Macoma should be at least 2 cm and



preferably 3-5 cm, whereas a large lugworm may require 5-10 cm of



sediment.  While these amounts should bracket the needs of most



organisms, the initial amount of sediment added should equal or



exceed consumption requirements for the 28-day exposure.  For



example, if a particular species processes 2 grams of sediment



per gram of tissue per day, then at least 56 grams per gram of



tissue should be added initially.  A more accurate estimate of



sediment requirements for selective deposit-feeders can be



generated by using the processing rates of the ingested size



fraction, though this information will .not usually be available.



     If periodic sediment additions are not made, then the



initial amount added should exceed the total amount processed



over the duration of the experiment by at least 2-fold, and



preferably 5-fold.  Thus, for the organism with a 2g/g-tissue/day



sediment processing rate, about 250-300 grams of sediment should



be added per gram of tissue.  Compilations of sediment processing



rates  (e.g.. Lee and Swartz, 1980) can be used to estimate these



requirements.  It must be recognized, however, that in a



laboratory, an organism may deplete the food or pollutants within



its specific feeding zone regardless of the amount of sediment



added.  This is especially likely with surface deposit-feeders.



For this reason, we strongly recommend periodic additions of the



treatment sediment, as discussed in Chapter X.
                                 87

-------
2. Water Quality



     The water quality requirements for conducting the tests are



similar to those for pre-experimental maintenance  (Chapter VIII).



That is, dissolved oxygen should be between 60 and 100% of



saturation and the un-ionized ammonia concentration should not



exceed 20 ug/L (ASTM, 1984).  These criteria can be met either by



using a flow-through system or a static batch-replacement mode.



     For a flow-through system, ASTM  (1984) recommends not more



than one filter-feeding bivalve (40-60 mm from umbo to edge of



distal valve) per liter per hour.   This would be equivalent to



about a minimum flow of one L/hr/g wet tissue for an oyster.



However, this requirement is based on feeding, and does not



account for sediment oxygen demand.  In addition to the flow rate



per gram tissue,  flow-through systems should be designed to



achieve at least six turnovers a day.



     In static systems,  the water volume to loading ratio needs



to be sufficient to allow the maintenance of oxygen levels .>60%



of saturation.  A gentle aeration helps maintain the oxygen level



as does changing the water two or three times a week.   As an



example, 10 Macoma nasuta (mean wet flesh weight of about 1.3 g),



each in a 100 ml beaker with an initial 50 grams of sediment,



have been successfully maintained in a 10 L aquarium with 8 L of



filtered seawater (Ferraro et al.,  1990).   The aquarium was



gently aerated and the water changed three times a week.



     In determining the oxygen demand for the system,  it is



important to take into account the total sediment oxygen demand.
                                 88

-------
In most cases, sediment microbial demand will be several fold

greater than the oxygen utilization by the test species.

Therefore, calculations based on sediment-free exposure systems


will underestimate the actual oxygen requirements.  The total
                                                             Q
oxygen demand of sediments ranges from <1 to over 100 ml 02/mz/hr


(e.g., Hargrave, 1969; Smith et al., 1973; Smith, 1978; Davis and


Lee, 1983).  In general, total oxygen demand will increase with


temperature and in organically enriched sediments, and the water

flow or volume should be increased accordingly.


     Aeration will help ensure a proper oxygen concentration is


maintained, and is required in a static system.  The air should


be filtered and free of oil and moisture.  The volume should be

sufficient to turnover the water but not enough to resuspend the


sediment.  This can be achieved with a supply of approximately


0.1 SCFH  (standard cubic foot per hour) per 10 liter aquarium via


an air stone or pipette.  Position the air stone or pipette


outside of beakers maintained in aquaria, or sufficiently far


above the surface to avoid resuspension in individual beakers or


aquaria.  Check the air-stone or bubbler frequently and remove

any salt crystals forming at the orifices.  If air is provided

from a compressed air tank, specify that the composition include


about 0.3-1.0% C02 to help control pH.  If not specified, no C02

will be present.


     Seawater is well buffered, but in static systems metabolites


and waste materials (i.e., ammonia)  can build up, lowering pH.


Maintain pH between about 6.5 and 8.0  (ASTM, 1984).  As mentioned,
                                 89

-------
aeration will help maintain pH as will periodic replacement of



water.  In static systems, the addition of clean,  crushed oyster



shell to the bottom of the aquarium can provide a good buffering



system.



     ASTM (1984) recommends that overlying water salinity should



vary less than 2g/kg or 20% of the average, whichever is higher.



In areas where salinity varies (as in water drawn from estuaries



with a seasonally high riverine contribution),  store a quantity



of high salinity water sufficient for the expected period of low



salinity, or preferably, to maintain salinity over the duration



of the exposure.



     Because phytoplankton and suspended material are a sink for



pollutants and a food for facultative filter-feeders, it is



important to filter the water to remove suspended particulates



(>5 um) during the test.  Filtration can be accomplished with in-



line cartridge filters  (commercially available with 2.5-5.0 um



pores) or in batch mode.  The ASTM (1984) precautions concerning



the adequate concentration of phytoplankton necessary as a food



source are not relevant for deposit-feeders.



3. Temperature and Light



     The tests should be conducted as close as possible to one of



the seven temperatures recommended by ASTM (1984)  - 7, 12, 17,



22, 27, and 32°C.  A temperature corresponding to the average



spring-summer temperature of the study site would simulate the



biologically most active season.  Most commonly, this will be



12°C in the Pacific Northwest, 17°C in mid-latitudes, and 22 or
                                 90

-------
27°C on the Gulf Coast.  With flow-through systems, it can be



difficult to exactly adjust the temperature, though a large head



tank in a controlled-temperature room will temper the water.  The



maximum difference between the minimum and maximum temperatures



must not exceed 10°C (ASTM, 1984).



     Light should be provided by means of cool-white fluorescent



lights at an intensity of about 400 foot-candles.  Other sources



(incandescent, fluorescent/incandescent, photosynthetically



active radiation augmented) may be required for special purposes.



Ultraviolet radiation, especially UV-B, is generally missing from



these artificially supplied spectra.  Although UV-B radiation can



enhance the toxicity of certain pollutants  (phototoxicity), this



probably will not greatly affect organisms buried in sediment.



ASTM  (1984) recommends a 16h day, 8h night as a convenient



light/dark cycle.  Schedules of 12/12 or 14/10 hrs day/night are



also acceptable, and may be useful in delaying maturation and



spawning of some species.  We have routinely used a 12/12



schedule.  These various day/night cycles can be controlled by



use of timing devices  in the light circuits.





B. EXPOSURE SYSTEM DESIGN



1. Materials Compatibility



     Materials used in the exposure system should not induce any



reaction by the organisms or affect pollutant concentration or



bioavailability.  Borosilicate glass  (Pyrex , Kimax , or



equivalent) and soft glass  (soda-lime, window) have proved



generally non-reactive to metals and organics, and are the





                                 91

-------
preferred materials where their fragility is not a major



limitation.  Most rigid plastics  (polyolefins, engineering resins



and fluoropolymers) are acceptable after conditioning, such as



soaking in deionized water for several days.  Some plastics,



generally flexible types that contain mobile plasticizers



(phthalate esters), need to be tested for toxicity and pollutant



compatibility.  These flexible plastics should not be used if the



uptake of phthalate esters will be studied.  Because the alloy



components of many stainless steels may react with saltwater,



stainless steel should not be used in direct contact with



seawater.  Choose another material if pollutant sorption to



internal surfaces of containers is a problem.



     Any sealant used to construct chambers must be non-toxic,



such as Dow-CorningR #8641 clear,  non-toxic silicone-rubber  (i.e,



meets FDA Regulation 21 CFR 177.2600).  Such materials are



usually specified for aquarium use and do not contain fungicides



(e.g., arsenic compounds).  Exposed sealant at joints should be



minimal.  Place sealant used for mechanical reinforcement on the



outside of the joint.  Plastics and sealants must be chosen



carefully, as both may sorb pollutants.  Product literature on



the material is helpful in determining the compatibility of a



particular plastic to a pollutant.



2. Exposure Chamber



     The actual exposure chamber can consist of glass boxes,



beakers, aquaria, or other containers of appropriate material



(see Figure IX-1).  For most species, beakers are an inexpensive






                                92

-------
              FIGURE IX-1



Representative Sediment Exposure System
                    93

-------
exposure chamber.  The diameter of the beaker and the sediment



depth in the beaker should be sufficient to allow the organism to



bury and/or construct normal tubes.  With Macoma nasuta. about 50



mm diameter beakers (Corning #1000, 100 ml) with about 50 g of



sediment were sufficient for 25-32 mm clams (anterior to



posterior measurement) to bury normally.  With 35-48 mm clams, 68



mm diameter beakers (Corning #1000, 250 ml) with about 100 g of



sediment were required.  In both cases, the initial depth of the



sediment was about 3-5 cm depending on sediment type.  If the



beakers are placed into an aquarium, the beaker height should be



several centimeters less than the water height to allow for



circulation into the beakers.  For funnel-feeding arenicolid



worms, a long narrow glass box (about l"w,  7"1, 4.5"h) is a more



appropriate shape  (see Figure IX-IB).  The opening of the



exposure chamber should allow the periodic addition of feeding



sediment.



3. Static Exposure System



     The simplest static systems are individual aquaria or



beakers filled with water as commonly used in sediment toxicity



tests (e.g., Swartz et al.,  1985a).  A more common design for



bioaccumulation tests are sets of beakers submerged in aquaria



(e.g., Ferraro et al., 1990).   The beakers or aquaria should be



covered to reduce evaporation and gently aerated to maintain



dissolved oxygen levels.  Drain tanks about once every two days



by siphoning water from the aquaria, but not the individual



beakers.  Retain the waste water in a container for treatment if





                                 94

-------
it is suspected of containing unacceptable pollutant levels for



disposal down the sanitary sewer.  Gently refill the aquarium



with water of the correct salinity and temperature, and restart



the aeration.



4. Flow-Through Exposure Systems



     For flow-through systems, chambers may be sets of beakers



maintained in aquaria (Figure IX-1) or entire aquaria.  Flow-



through systems have the advantages of removing waste products



and maintaining oxygen.   Though desirable, a flow-through design



is not normally required for a successful bedded sediment test.



To avoid cross contamination, water flowing through one container



must not flow into another container.  Water exiting the systems



should be passed through a charcoal filter if substantial



desorption of pollutants from the sediment is anticipated.



Similarly, resuspended sediment should be trapped and retained as



waste.  Examples of conducting flow-through tests can be found in



U.S. EPA  (1978), Rubinstein et al.  (1980), and Rubinstein et al.



(1987).





C. MULTIPLE SPECIES EXPOSURE CHAMBERS



     If several species are being tested, it is possible to place



multiple species within each exposure chamber.  The advantages of



multiple species per container include reduced space requirements



and a lower cost because of the reduction in the number of



chambers constructed and maintained.  The greatest disadvantage



is the potential for negative interactions among the species,
                                 95

-------
such as competition or predation, which could have unknown and



variable effects on uptake.  For example, uptake could be reduced



in Macoma if siphon clipping by epibenthic shrimp or nereid worms



reduced the clam's feeding rate.  Another disadvantage is that



the accidental loss of a chamber reduces the number of replicates



for each of the species.



     If multiple species are placed within exposure chambers, the



amount of sediment initially added should at least equal the sum



of the amount required for each individual species.  Most of the



potential interactions are density dependent, so increasing the



area of the chambers  (while maintaining a sufficient sediment



depth) should reduce the intensity of any negative interactions.



An alternate design is the physical partitioning of the aquarium



with screens to separate the species  (e.g., Rubinstein et al.,



1987).



     Regardless of the specific design, the same numerical ratio



of one species to another must be maintained in replicate



chambers.  It should also be noted that a paired-comparison



approach should be used when statistically comparing the tissue



residues of two species kept in the same chambers.
                                 96

-------
   CHAPTER X:  EXPERIMENTAL  INITIATION, MAINTENANCE AND  SAMPLING





A. EXPERIMENTAL INITIATION AND MAINTENANCE



1. Pre-Experimental Preparations



     Coordinate the collection and acclimation of the bioassay



organisms with the collection of the sediments so the experiment



may begin with a minimum of delay.  The glassware, water delivery



system,  and any stored water should be ready, as well as sampling



containers, labels and related paraphernalia.  Beakers and other



containers should be pre-labeled.  A detailed work schedule,



showing daily tasks and persons responsible for accomplishing



them, should be prepared before the sediment arrives.  A



prearranged numbering scheme should be agreed upon with the



analytical chemists.  It is critical to keep the analytical



chemists well informed of the sampling schedule so they can



prepare for the sample load.  Arrange with maintenance personnel



to look for power failures, pump leaks,  breakage of aquaria,



inadvertent switching on of lights at night, and other accidents.



Provide telephone numbers for key personnel responsible for



maintenance of the experiment in a prominent location  (e.g., on



the door of the laboratory).  Any safety warnings should also be



posted at entry points.



2. Experiment Initiation



     Weigh all individual organisms or composites of organisms,



while taking care to minimizing exposure of soft-bodied organisms



to the air.  To avoid temperature shock, maintain the air



temperature of the room at the experimental water temperature.





                                 97

-------
All bivalves should be measured  (anterior to  posterior valve),



weighed, and individually marked with a random number.  When



possible, the number on the clam should be the same as the number



on their exposure chamber.  Clams can be marked with a laboratory



marking pen (e.g., SharpieR pen) by first scrubbing the shell



with a KimwipeR or other soft paper towel, blotting the shell,



and then allowing them to dry (about 15 minutes at 12°C).   Mark



the same valve (i.e, right or left)  in all clams.  Discard any



organisms not meeting the criteria for size or condition.



Maintain some extra individuals for potential replacements within



the first 24 hours.  Also, randomly choose some specimens for



wet-to-dry weight conversions and for long-term storage for



potential lipid analysis with a different technique (see Chapter



XI) .



     Distribute measured aliquots of homogeneously mixed sediment



to each exposure chamber.  Weighing the sediment aliquot is



preferable,  but sediment volume can be used to estimate mass for



a particular sediment type.  During the process of measuring out



aliquots of sediment, periodically re-stir the source to avoid



separation of the fines and interstitial water.  If beakers are



used as the exposure chamber, gently tap the beaker to



consolidate the sediment and eliminate air bubbles.  To avoid the



loss of surficial fines when filling the beakers, place a plastic



film over the sediment surface,  slowly fill the beaker with



water, and then withdraw the film using forceps.  Carefully place



the water-filled beakers into filled aquaria and allow any



suspended fines to settle.





                                 98

-------
     If aquaria or other large containers are used as the



exposure chambers, stir the sediment after adding the appropriate



amount to mix sediment and remove any bubbles.  As with the



beakers, a plastic film should be placed over the sediment



surface when filling the aquarium with water.  Position any



aerating device so that the induced turbulence does not resuspend



sediment.



     Add the organisms after allowing the sediments to



consolidate and any suspended particles to settle, which will



normally take from 15 minutes to a day.  Place animals on the



surface of the sediment and allow them to bury.  To facilitate



burial, place Macoma nasuta left valve down  ("bent-nose" up) on



the sediment.  Mobile organisms, such as the polychaetes, should



be observed for a sufficient period to assure that they.bury in



the correct chamber and do not swim into another chamber.  For



mobile worms, it may be necessary to place a screen on the tops



of beakers to keep them from swimming out.



2. Experiment Maintenance



     Replace any animals whose behavior is abnormal (failure to



bury in the sediment, etc.) within the first 24 hours.  Observe



the chambers daily and note any signs of abnormal activity  (e.g.,



reduced production of fecal pellets, avoidance of the sediment).



Remove any beakers with dead organisms.  It is especially



important to monitor for dead organisms in a static system.



Record temperature, salinity, and other water quality parameters



on a periodic basis  (see Chapter IX).  Replenish water in static
                                 99

-------
experiments according  to a preplanned  schedule  (Chapter  IX),  and
dispose of drained water in accordance with  applicable rules  for
hazardous waste.
3. Sediment Renewal
     We strongly recommend periodic additions of small amounts of
the appropriate sediment type to each  exposure chamber.  Because
the bioavailable fraction may only be  a small portion of the
total sediment-associated contaminant  (see Landrum, 1989),  it is
possible for organisms to deplete the  available fraction,
especially in organisms which ingest sediment from a restricted
feeding zone (e.g., surface deposit-feeding bivalves).  For
example, depletion of the bioavailable fraction may be the reason
that tissue residues of 35 of 37 compounds declined between day
39 and day 79 in Oliver's (1987) study of uptake by oligochaetes.
Also, without organic input from settling phytoplankton and with
low light levels inhibiting benthic microalgae,  it is possible
for the nutrient quality of the sediment to decline over the
course of an experiment.  Periodic sediment renewal should reduce
these potential laboratory artifacts and help maintain a more
constant pollutant concentration and food supply.   The periodic
addition of sediment results in a pulsed-renewal exposure.
Without the addition of new sediment,  the exposure is a single-
dose exposure.
     The daily amount of sediment added should equal or exceed
the daily sediment processing rate of the organism.   Sediment
ingesting clams such as Macoma require about 1 gram wet sediment
                                 100

-------
per gram of wet tissue mass per day while arenicolid worms  (2-6 g



wet weight) require about 10 g of sediment per day.  It is



sufficient to add the sediment two or three times a week  (e.g.,



about 3.5 g twice a week for a 1 gram Macoma).  Previously, we



had frozen the sediment immediately before its addition to  reduce



loss of fines  (Specht and Lee, 1989; Boese et al., in press);



however, given the unknown effects of even short-term freezing,



we recommend adding the sediment using a cut-off plastic syringe.



We have successfully used a 3 cc syringe with both a fine sand



and a silt-clay.  The volume of sediment in the syringe is  a



simple way to estimate mass of sediment, though the volume  to



weight ratio has to be determined for each sediment.



     For long-term exposures  (>28 days), we recommend



periodically replacing all the sediment in the chambers.



Replacement of sediment reduces the possibility of depletion of



the bioavailable fraction of the pollutants and/or food, and



avoids excessive pelletization of the sediment.  Additionally,



the periodic addition of surface sediments will overfill most



chambers within a few weeks, requiring a complete replacement of



the sediment.  Replacement on a 28-day schedule should suffice,



and coordinates with the long-term sampling schedule (see Chapter



IV).   If a field sediment is being tested, all the sediment



should be collected at the same time and the renewal sediment



stored until needed.  If a spiked sediment is being tested, it



may be preferable to spike new sediment for replacement.
                                 101

-------
     Do not feed the test organisms a supplemental source of food



in either 28-day or long-term experiments.  By ingesting the



added food, the organisms are presumably ingesting less sediment,



which could result in an underestimation of the bioavailability



of the sediment-associated pollutants.  The addition of food is



not required as shown by long-term maintenance (>28 days) of



deposit-feeding bivalves (e.g., Specht and Lee, 1989),



polychaetes (e.g., McElroy and Means, 1988), and crustaceans



(e.g., Landrum, 1989)  without supplementing the sediment with an



artificial food source.






B. SCHEDULE FOR ABIOTIC AND BIOTIC POLLUTANT SAMPLES



     Samples of sediment, water, and biota should be taken for



pollutant analysis before,  during, and after testing.  Sampling



techniques and apparatus may vary with the nature of the



sediment, species of test organism, and compound(s) of interest.



As the manner in which samples are taken may affect the analysis,



consistency in sampling for any given parameter is essential.



1. Overlying Water



     Although no pollutants are intentionally added to overlying



water in sediment bioaccumulation tests, contaminants may be



introduced from the water supply system, leached from the



sediment, or present on resuspended particulates.   The activities



of some species (e.g.,  Yoldia)  can resuspend considerable amounts



of fine-grain material directly into the water column.  With a



randomized block design  (Figure III-3f), bioturbation may lead to



cross-contamination between treatments.   This potential uptake






                                 102

-------
from the water needs to be quantified to differentiate it from



uptake from the bedded sediment and to check for possible cross-



contamination among treatments.



     At a minimum, overlying water should be sampled for



pollutants from each treatment at the beginning, middle, and end



of the test period (i.e., TQ, T14, and T2Q).  A sample from each



aquarium should be analyzed if statistical comparisons are



planned, though in many cases it would be acceptable to composite



water samples from aquaria of the same treatment.  If samples are



composited, individual samples from each aquarium should be



archived in case a more detailed analysis is required.  Samples



should also be taken during periods of high turbidity or other



unusual water quality.



        Overlying water should be sampled at mid-depth from each



exposure unit.  When experimental units share the same overlying



water (e.g., test beakers within the same aquarium), overlying



water should be sampled from mid-depth of the entire container.



Care should be taken to avoid disturbing the flocculent material



at the sediment-water interface.  Sampling apparatus  (pipettes,



sample vials)  should be made of materials that do not appreciably



absorb or leach pollutants.  To guard against cross-



contamination, rinse off the sampling apparatus after each use.



Sample volumes will depend upon the analytical technique used,



but may range from about 1 to 100 ml.
                                 103

-------
2.  Sediment and Interstitial Water



     Sample all test, control, and reference sediments before the



addition of organisms (tQ sample) and at the end of the exposure



 (typically t28).   These sediment samples should be analyzed for



pollutant concentrations, TOG, and moisture content.  In most



cases, it is adequate to conduct the grain size analysis only on



the initial sample.



     One procedure for sampling sediment for organic pollutants



from exposure chambers is as follows:



   1. Remove overlying water from the exposure chamber by



   siphoning or decanting, taking care not to disturb the surface



   floe.  Depending on the procedure, interstitial water samples



   may be taken at this stage.



   2. Remove the test organism(s) from the sediment.  Larger



   bivalves can be directly removed with forceps.  Spread the



   sediment out in a tray to remove small bivalves and



   polychaetes.  Do not use any water to remove the sediment from



   the exposure chambers.



   3. Homogenize the test sediment from each exposure chamber by



   stirring with a TeflonR coated spoon, glass rod, or other



   inert utensil.  Take a sediment sample from each exposure



   chambers, place in a labeled sample vial and freeze.  These



   individual samples will either be analyzed or archived if



   composites are analyzed.



   4. If composites are going to be taken, the compositing



   strategy will depend upon how the exposure chambers were
                                 104

-------
   allocated among aquaria.  If only one treatment type is placed



   in each aquarium, composite all the beakers within an



   aquarium.  If the exposure chambers are allocated randomly



   among aquaria, combine all the sediment from each treatment



   (i.e., sediment type) regardless of aquarium.  In both cases,



   homogenize the sediment, take replicate samples from each



   composite, and freeze until analyzed.



     Extra sediment samples should be taken from individual



exposure chambers (and from any composites) and frozen in case



there is an analytical failure or greater statistical power is



required.



     Because this procedure exposes the sediment to the air,



reduced metal forms will be oxidized.  If metal speciation will



be studied, the procedure should be modified, especially steps 2



and 3, to minimize the sediment's exposure to air.  One



possibility is to take small sediment cores from the exposure



chambers.  Regardless of sampling scheme, interstitial water



should be collected at the same time as the sediment samples.



Interstitial water may be collected by a variety of methods



including centrifugation, sediment squeezing, and dialysis



membranes  (see Chapter VI).





C. METHODS OF BIOTIC SAMPLING



     Test organisms need to be carefully removed from the



sediment, as described above, and all adhering particles removed.



A gentle rinse with clean seawater will help remove particles



from polychaetes.  In general, organisms should be placed in





                                 105

-------
control sediment to purge their gut contents for 24 hours before



chemical analysis  (see Section D).  At the end of the purging



duration, collect the organisms by gently spreading the sediment



out in a tray and removing the organisms using forceps or gently



sieving the sediment.




     After collection, rinse the organisms with clean seawater,



blot them dry, and then weigh them.  Measure the shell length of



bivalves.  Organisms should be analyzed immediately or frozen in



baked-out aluminum foil or glass vials.  The entire soft-tissue



of each individual or composite of individuals from an



experimental unit should be prepared for analysis.  In many



cases, the tissue from each experimental unit will first be



homogenized and then subsamples taken for organic, metal,  and



lipid analyses, and archiving.  The type of homogenization



technique will depend upon size and tissue consistency of the



organism, the pollutant of interest,  and the analytical



procedures used for pollutant analysis.





D. GUT PURGING



     When a whole-body tissue analysis is conducted on a deposit-



feeder,  any pollutants associated with the mineral particles and



detritus in the gut are included.  Depending on the mass of



sediment and the associated pollutant concentration,  the gut



sediment can measurably increase the apparent whole-body tissue



residue.  Allowing the organism to purge its gut  (i.e., defecate)



in uncontaminated sediment reduces or eliminates this positive bias
                                 106

-------
TABLE X-l: Errors Associated With Gut Sediment/Purging

 I.  Gut Sediment Introduces Greatest Error:

    1. In organisms that selectively ingest high organic particles.

    2. In organisms with a large gut capacity.

    3. During early stages of uptake when tissue residues are low.

    4. For compounds not extensively bioaccumulated, especially
       high Koc compounds with steric hindrance to uptake.


  II. Purging Introduces Greatest Error:

    1. For rapidly depurated/metabolized compounds.

    2. In organisms which do not clear gut in water.
                                 107

-------
However, pollutants will depurate or be metabolized during
purging, resulting in an underestimation of the bioaccumulation.
The type and extent of the error will depend upon many factors,
including the feeding behavior of the organism and the nature of
the pollutant.  Factors influencing the errors associated with
gut sediment or purging are summarized in Table X-l.
1. Standard 24-Hour Purge
     Organic compounds with high Kow values (e.g., PCBs, DDT,
BaP)  are usually the greatest environmental concern in terms of
bioaccumulation.  Most of these compounds are slowly depurated,
so a relatively small amount should be lost during purging.
Therefore, we recommend a 24-hour gut purging as the standard
procedure for sediments known or suspected to contain more than
trace amounts of these pollutants.  A 24-hour depuration period
is sufficient for organisms to defecate the majority of their gut
contents without introducing substantial errors from pollutant
depuration or metabolism.
     Many deposit-feeders require the ingestion of sediment to
completely void their gut contents, so organisms should be placed
in control sediment to assure complete purging.  Reference
sediment should not be used as the purging sediment.  Maintain
environmental conditions  (e.g., temperature, salinity) as during
the exposure phase.  The organisms in the control and reference
sediment(s) should undergo the same purging treatment as
individuals exposed to the test sediment.  Organisms from
different treatments should be kept in separate containers to
                                 108

-------
avoid any possibility of cross-contamination.  Observations



should be made on whether feces were produced during the purging



period and on the general health of the organisms.



     There are several other techniques or modifications to the



standard 24-hour purge which may be considered in specific cases.



These methods are discussed in Appendix X-l.



2. When Not To Purge



     There are certain situations when gut purging may introduce



a greater error than leaving the gut sediment.  In the first



situation, the primary focus of the study is comparing laboratory



and field studies.  In most cases, it is impractical to purge



field collected organisms.  Therefore, to assure that the



laboratory and field results are directly comparable, laboratory



organisms should not be purged.   In the second case, the primary



focus is to determine the trophic transport of pollutants.  As



deposit-feeders extract sediment-associated pollutants in their



guts (e.g.,  Lee et al.,  in press), it is likely that predators



would also extract a certain percentage of the pollutants from



their prey's gut sediment.  In the final case, the primary focus



is on lower molecular weight PAHs.  These compounds can be



depurated and metabolized rapidly (see Table X-2)  so that a 24



hour purge can result in a greater error than leaving the gut



sediment.
                                109

-------
TABLE X-2: Depuration Loss Of Pollutants During 24 and 72 Hour
           Gut Purges
  COMPOUND
ORGANISM
% LOST (HRS)
REF.
                                 24
                          72
  PCB   Crangon septemspinosa

  HCB   Macoma nasuta

  BaP   Pontoporeia hoyi


  Phe   Pontoporeia hovi
                   3

                   4
          8   McLeese et al.,  1980

          12   unpublished data
                  4-28   11-64  Landrum & Poore,
                                 1988

                  11-54   28-90  Landrum & Poore,
                                 1988
PCB = Aroclor 1254
HCB = Hexachlorobenzene
BaP = Benzo(a)pyrene
Phe = Phenanthrene
                                 110

-------
E. ACCEPTABLE LEVELS OF MORTALITY



     According to ASTM  (1984) guidelines for bioconcentration



tests, a test is unacceptable if "more than 10% of the organisms



in any treatment died or showed signs of disease, stress, or



other adverse effects."  This criterion is applicable to studies



of spiked sediments in which it is possible to adjust pollutant



concentrations.  Repeat any 28-day spiking experiment at a lower



pollutant concentration if 10% or more of the organisms in any



treatment die or show overt signs of stress.  Signs of stress



include avoidance of the sediment, non-burial, casting off of



siphons, abnormal tube construction, and reduced ventilation or



sediment processing rates.



     In contrast to most experimental studies of bioavailability,



many of the field sediments or dredge materials of environmental



concern will have moderate to high toxicity.  With these



sediments, it may be impossible or difficult to meet the 10%



mortality criterion.  However, this may not represent a serious



problem as the purpose of evaluating these sediments is to



determine the extent of bioaccumulation which will result from a



particular sediment.  Presumably, the mortality in the laboratory



would mimic the response in the field and so represent the actual



effect of the sediment.



     Because of the different purposes of tests conducted on



field versus spiked sediments, we suggest not automatically



rejecting bioassays with greater than 10% mortality in the test
                                 111

-------
sediment.  The determining factor, in deciding whether to accept a



test treatment with high mortality is whether there are adequate



replicates to obtain sufficient statistical power (see Chapter



III).   If the statistical power is insufficient, the experiment



should be repeated.  Also, mortality or stress in greater than



10% in the control or reference sediment would indicate initially



stressed organisms, contamination of the system, or unacceptable



control or reference sediment.  In such cases, the cause of the



problem should be determined and the experiment repeated.



Consider using a more pollutant-resistant species or diluting the



sediment to reduce toxicity (see Appendix V-l) in any future



tests if the mortality in the test sediment exceeds 25%.



     In some regards, high mortality in field sediments is a moot



problem because any sediment sufficiently toxic to kill a



substantial proportion of the recommended test species presumably



would be unacceptable based on toxicity.  However, even in cases



where a sediment is rejected on the basis of toxicity, a



bioaccumulation test conducted on the diluted sediment may help



identify the compounds responsible for the toxicity.





F. CHAIN OF CUSTODY



     In the event that litigation is expected, it is imperative



to follow proper sample chain-of-custody procedures so that the



results are acceptable for court.  We recommend following the



chain-of-custody procedures published by the National Enforcement
                                 112

-------
Investigations Center  (U.S. EPA, 1986a).  Other sources include



"Quality Assurance and Quality Control  (QA/QC) for 301(h)



monitoring programs: Guidance on Field and Laboratory Methods"



(Tetra Tech, 1986b) and U.S. EPA Contract Laboratory Program



(U.S. EPA, 1988, 1989).
                                 113

-------
             CHAPTER  XI:  POLLUTANT  AND  LIPID  ANALYSIS






A.  POLLUTANT ANALYSIS



    The specifics of the techniques used to analyze sediment,



water, and tissues for pollutants is a complex subject beyond the



scope of this manual.  Discussions of analytical techniques can



be found in Tetra Tech   (1985b, 1986f,g,h) and U.S. EPA  (1988,



1989).  It is possible, however, to offer several guidelines.



First, analytical techniques are media dependent.  Thus, time



should be allocated for modifying the procedures for the various



media and any special conditions (e.g., high TOC sediment, low



tissue biomass).   Second, a harsh extraction technique should not



be used when analyzing sediments for metals since such a



technique can extract biologically unavailable metals from the



mineral matrix.   A discussion of various metal extraction



techniques is found in Luoma and Bryan (1978) and Waldichuk



(1985).  Third,  to the extent possible, the PCB analysis should



be at the level of identifying and reporting specific congeners



rather than Aroclor equivalents.  In particular, the more toxic



planar congeners need to be identified.  A thorough review of PCB



congeners, including which to analyze,  can be found in McFarland



and Clarke (1989).



     The required or desired detection limits will have a major



effect on the choice of analytical techniques and on the ability



to interpret the data.  In some cases,  the detection limits and



analytical procedures will be specified by the pertinent



regulation while in other cases the decision will be determined






                                 114

-------
by the researcher.  If no detection limits are specified, we



recommend that, at a minimum, the analytical techniques meet the



requirements of the U.S. Environmental Protection Agency's



Contract Laboratory Program requirements  (U.S. EPA, 1988, 1989).



The quantification limits from these documents are summarized in



Table XI-1.  These protocols cover both metals and organics in



water and sediment.  Although tissues are not covered by these



protocols, it should be possible to obtain the same



quantification limits as with the sediments.



     Control samples or samples from relatively clean areas



contain low concentrations of pollutants, and may require lower



detection limits to achieve satisfactory results.  The methods



developed for measuring pollutants in samples collected from the



PSDDA control sites in Puget Sound  (U.S. ACE, 1988) are



suggested in such cases.  The PSDDA values include tissues as



well as water and sediment, and are summarized in Table XI-2.



     A complete quality assurance/quality control plan is a



central part of any analytical procedure.  Information on



analytical QA/QC procedures are available from several sources



(U.S. EPA, 1988, 1989; U.S. ACE, 1988).  An important part of any



QA/QC program is the use of reference samples and standards.



Reference samples and standards are available from the U.S. EPA



in Cincinnati, OH; Las Vegas, NV; and Research Triangle Park, NC,



as well as the National Institute of Standards and Technology



(Office of Standard Reference Materials, Room B311, Chemistry



Building, NIST, Gaithersburg, MD 20899).





                                 115

-------
TABLE XI-1:  U.S. EPA Contract Laboratory Program Quantitation
Limits for Water and Sediment With Estimates for Tissue Matrices
Organics
   Water (a)   Sediment (b)    Tissue (c)
Volatiles
Semivolatiles
Pesticides/PCB's
    5-10
    10-50
    0.05-1
0.5-10
330-1600
8-160
0.5-10
330-1600
8-160
For individual pollutants - refer to CLP Statement of Work
Metals
    Water (a)
Antimony
Arsenic
Cadmium
Copper
Lead
Mercury
Nickel
Silver
Zinc
Metals not listed
    20-300
    5-100
    0.5-10
    5-100
    5-100
    0.2-20
    5-100
    1-25
    0.2-4
refer to CLP Statement of Work
a = ug/L
b = ug/kg wet weight
c = ug/kg wet weight basis.  These values were estimated from the
    sediment values on the premise that tissue and sediment
    pollutant concentrations are of a similar magnitude  and are
    analyzed by similar techniques.
                                 116

-------
TABLE XI-2.
Organics
PSDDA Low Limits of Detection for Water, Sediment
  and Tissue Matrices
                     Sediment (a)
Tissue (b)
Volatiles
Semivolatiles
Pesticides/PCB
Metals
Antimony
Arsenic
Cadmium
Copper
Lead
Mercury
Nickel
Silver
Zinc
a = ug/kg dry
b = ug/kg wet
c = ug/L (ppb)
d = mg/kg dry
e = mg/kg wet


•s
Water (c)
3
1
0.1
1
1
0.2
1
0.2
1
weight (ppb)
weight (ppb)
weight (ppm)
weight (ppm)
10-20
1-50
0.1-15
Sediment (d)
0.1
0.1
0.1
0.1
0.1
0.01
0.1
0.1
0.2

5-10
10-20
0.1-20
Tissue (e)
0.02
0.02
0.01
0.01
0.03
0.01
0.02
0.01
0.20

                                 117

-------
B.  LIPID ANALYSIS




     A number of studies have demonstrated that lipids are the



major storage site for organic pollutants in a variety of



organisms  (Roberts et al.,  1977; Oliver and Niimi, 1983; de Boer,



1988).  Because of the importance of lipids, bioaccumulation



programs have recently attempted to normalize tissue pollutant



concentrations to the tissue lipid concentration.  For example,



lipid concentration is one of the factors required in deriving



the Accumulation Factor (AF)  (see Appendix 1-1).  The approach,



however, has experienced difficulties because of the differences



in the lipid concentrations reported from the wealth of different



lipid methods used (see Kates,  1986 for discussion of lipid



methodology).   Work in this laboratory has shown that differences



in lipid technique can result in 3-fold differences in lipid



concentrations.  These differences in lipid concentrations



directly translate into a similar variation in the lipid



normalized pollutant concentrations or Accumulation Factors.



    To allow lipid normalized tissue residues or AFs to be



compared,  it is necessary to either promulgate a standard lipid



technique or to intercalibrate the various techniques.



Standardization on a single method is difficult because the lipid



methodology is often intimately tied in with the extraction



procedure for pollutant analysis.  Instead,  we recommend that one



lipid technique be chosen as  an "intercalibration standard".



Then, regardless of what method is used,  the results would be



reported in equivalent units  of the standard.
                                 118

-------
     The difficulty with this approach is deciding upon which



technique to use as the standard.  We are presently investigating



the advantages and disadvantages of different techniques, but



have not yet reached a consensus.  As in interim solution, we



suggest the Bligh-Dyer lipid method  (Bligh and Dyer, 1959) as a



temporary "intercalibration standard".  Folch et al. (1957)



developed a total lipid method that extracts the neutral and



polar lipids   (i.e. total lipids) from biological samples using



chloroform and methanol as the solvent system.  Bligh and Dyer



(1959) improved upon the method by providing a cleanup for the



extracted lipid residues.



     The potential advantages of Bligh-Dyer include its ability



to extract neutral lipids not extracted by many other solvent



systems and the use of Bligh-Dyer (or the same solvent system) in



numerous biological and toxicological studies (e.g., Roberts et



al., 1977; Oliver and Nimi, 1983; de Boer, 1988; Landrum, 1989).



Because the technique is independent of any particular analytical



extraction procedure, it will not change when the extraction



technique is modified or changed.  Additionally, the method can



be modified for small tissue sample sizes as long as the solvent



ratios are maintained.  We have successfully used a modified



technique with tissue samples as small as 1 g wet tissue and



micromethods using chloroform - methanol requiring only milligram



amounts of tissue were developed by Herbes and Allen (1983) and



Gardner et al.  (1985).
                                 119

-------
     One potential disadvantage of the Bligh-Dyer is that by



extracting many of the lipids not extracted by other techniques,



it may be extracting lipids that are not important to the storage



of neutral organic pollutants.  Seasonal changes or interspecific



variations in these non-active lipid fraction could obscure the



relationship between lipid content and pollutant accumulation.



However, the standard Bligh-Dyer technique may not solublize



triglycerides (Gagney, P., pers. comm.,  U.S. EPA, ERL-A, Athens,



GA),  thereby underestimating the lipid pool important to



pollutant storage.  Other drawbacks are the use of chloroform,



which is a carcinogen, and the need to conduct an additional



analysis instead of measuring the lipids as part of the normal



organic extraction procedure  (e.g., Rubinstein et al.,  1987; Lake



et al., in review).  However, the alternative lipid methods all



have similar limitations, and we believe that Bligh-Dyer is the



best interim calibration method.



     If the Bligh-Dyer method is not used as the primary lipid



method, compare the chosen lipid method with Bligh-Dyer for each



tissue type.  The chosen lipid method could then be converted to



"Bligh-Dyer" equivalents and the lipid normalized tissue residues



could then be reported in "Bligh-Dyer equivalents".  Because of



the interim nature of this suggestion, we also suggest that extra



tissue of each species be frozen for future lipid analysis in the



event that a different technique proves more advantageous.
                                 120

-------
C. SAMPLE STORAGE



     For organics, the U.S. EPA Contract Laboratory Program  (U.S.



EPA, 1988) requires that the samples be protected from light and



refrigerated at 4°C (± 2°C.) from the time of receipt until they



are extracted and analyzed.  Water samples shall be extracted



within 5 days of the receipt of the sample.  Sediment samples



shall be extracted within 10 days of the receipt of the sample



and if continuous extraction procedures are employed, extraction



of water samples shall be started within 5 days of the receipt of



the sample.



     For inorganics, the U.S. EPA Contract Laboratory Program



(U.S. EPA, 1989) requires that soil and sediment samples be



maintained at 4°C.  (± 2°C.) until analyzed.  Samples for mercury



shall be analyzed within 26 days of the receipt of the sample.



Samples for metals shall be analyzed within 180 days of the



receipt of the sample.



     At times, other program priorities (e.g., analysis of



archived samples) do not allow one to abide by the requirements



set by the Contract Laboratory Program.  In those cases, it is



suggested that the samples either be frozen (-20°C) in air tight



containers or dried depending on the type of sample and the



analyses required.  Purging the container with nitrogen prior to



sealing will delay the degradation of some pollutants as well as



lipids.  Sample containers should be as full as practical to



prevent loss of moisture from the sample.   Sediment samples so
                                 121

-------
preserved are stable for at least 6 months, if not longer (Tetra



Tech, 1986d).  Tissue and water samples are expected to be at



least as stable as sediments.





D. REPORTING OF RESULTS



     Investigators have reported results on either a dry or wet



basis,  usually without a conversion factor between the two and



sometimes without any indication of which was used.  This makes



it difficult, or impossible, to compare results from different



studies.  In general, a dry-weight basis is preferred for both



sediment and tissue pollutant concentrations.  However, certain



analytical techniques use wet tissue or wet sediment,



necessitating the calculation of wet-weight concentrations.   To



allow comparisons among studies, the wet-to-dry weight ratios



should be reported for each tissue and sediment type.  As



mentioned  above, lipid values should be reported in "Bligh-Dyer



equivalents" along with any conversion factor(s) between lipid



methods.
                                 122

-------
                CHAPTER XII: STATISTICAL ANALYSES





     The main objective of statistical testing is to determine



whether the mean tissue residues in animals exposed to the test



sediment are significantly greater from those in the control



and/or reference sediments, or greater than a specified criterion



value such as an FDA Action Limit.  Additional statistical tests



comparing the means of other tissue residues  (e.g., control vs



reference) or sediment characteristics will also be conducted,



but the same principles and methods apply.  A summary of the



standard statistical tests and their interpretation are



summarized in Table XII-1 and Table XII-2.



     To perform statistical testing, replicate samples must have



been taken to provide an estimate of variability.  Non-replicated



samples  (i.e. concentration from a single composite sample)



cannot be compared using these methods.  In these tests, the



concentration of each chemical in a tissue or sediment sample is



considered statistically independent and is compared separately.



Comparisons of tissue residues of different chemicals within the



same organisms requires the use of "repeated measures"  (Section E)



     Standard deviations  (SD or s) or standard errors  (SE) and



number of the replicates  (n) should always be reported in



addition to mean values.  When composited values are used, report



the number of organisms per composite  (if the composite comprises



the experimental unit) or the number of experimental units per



composite, as well as the number of replicate composites sampled.
                                 123

-------
      Prior  to  conducting any  statistical  analyses,  it  is



necessary to decide whether the  comparisons between means  are  to



be multiple or pairwise.  Pairwise  comparisons  include



comparisons of a test and control/reference mean  for tissue



concentrations, sediment characteristics, etc.  Pairwise



comparisons also include the  comparison of the  control with the



reference mean and comparisons of a mean  and a  specified



criterion value such as comparison of a test tissue residue with



an FDA action  limit.  Multiple comparisons involve  comparisons of



more  than two  means simultaneously.  Multiple comparisons  are



used  in cases  such as determining whether three or more test



tissue concentration means are equal or whether all the TOC



values for  the sediments (test(s), control and reference)  are



equal.




     After  the applicable comparisons are determined,  the  data



need to be  tested for normality to determine whether parametric



statistics  are appropriate and whether the variances of the means



to be compared are homogeneous.  If normality and homogeneity of



variances are established,  t-tests can be performed in the case



of pairwise comparisons or ANOVA in the case of multiple



comparisons.






A. TESTS FOR NORMALITY AND HOMOGENEITY OF VARIANCES



     Before conducting parametric statistics,  the data need to be



checked for both normality and homogeneity of variances.   The



data for each chemical or sediment parameter are tested



separately.   Commonly used tests for testing normality are the






                                 124

-------
Kolmogorov-Smirnov one-sample test and the chi-square test (Sokal



and Rohlf,  1981).   However, these tests are not very powerful,



especially if sample sizes are small (such as 8 replicates).



More powerful, but less common, tests of normality such as



Shapiro-Wilk and K2 tests  (D'Agostino and Stephens, 1986) can be



used for small sample sizes.



     If the data are not normally distributed, the data can often



can be transformed to achieve normality.  The logarithmic and



arcsine are two commonly used transformations for concentrations.



It may be necessary to apply different transformations to



different chemical or sediment parameters.  See Sokal and Rohlf



(1981) for a more extensive discussion on transformations.  If



normality cannot be established, nonparametric tests for



comparisons of two means, such as the Mann-Whitney test and the



Tukey's Quick test should be used.  These non-parametric tests



are usually not as powerful as the more common parametric tests,



such as the t-test or Analysis of Variance (ANOVA).  See Daniel



(1978) for a discussion of non-parametric statistics.



     The variances of the samples to be compared should be tested



for homogeneity.  This is performed using an F-test when



comparing two variances or Bartlett's test when comparing



more than two variances.  If the variances are considered



homogeneous, then a t-test or ANOVA is appropriate.  If the



variances are heterogeneous, the data can be transformed in an



attempt to achieve homogeneity.  Under conditions of variance



heterogeneity, a modified t-test for comparisons of two means or





                                 125

-------
approximate tests for multiple comparisons can be performed.  See



Sokal and Rohlf (1981) for a more extensive discussion on



appropriate tests when different treatments have unequal



variances.





B. PAIRWISE COMPARISONS



     Pairwise comparisons are performed using Student's t-test,



using a pooled variance estimate when variances are homogeneous.



Under conditions of variance heterogeneity, a modified t-test can



be used (see Sokol and Rohlf, 1981).   Prior to analysis, it must



be established whether the t-test performed will be a one-tailed



or two-tailed test and whether the Type I error rate should be a



comparison-wise or experiment-wise error rate.  These



considerations are discussed below.



1. One-Tailed versus Two-Tailed Tests



     In formulating a statistical hypothesis, the alternative



hypothesis can be one-sided  (one-tailed test) or two-sided  (two-



tailed test).   The null hypothesis (Ho) is always whether two



values are equal.   A one-sided alternative hypothesis (Ha) is



that there is a specified relationship between the two values



(e.g., one value is greater than the other) versus a two-sided



alternative hypothesis (Ha) which is that the two values are



simply different.   A one-tailed test is used when there is an a



priori reason to test for a specific relationship between two



means such as the alternative hypothesis that the test tissue



residue is greater than the control tissue residue.  In contrast,
                                 126

-------
the two-tailed test is used when the direction of the difference



is not important or cannot be assumed prior to testing.  An



example of an alternative two-sided hypothesis is that the



reference sediment TOC is simply different from the control



sediment TOC.



     Because control tissue residues and sediment pollutant



concentrations are presumed lower than reference values which are



presumed lower than test values, we recommend conducting one-



tailed tests in most cases.  For the same number of replicates,



one-tailed tests are more likely to detect statistically



significant differences between treatments (i.e., have a greater



power).  This is a critical consideration when dealing with a



small number of replicates (such as 8 per treatment).  The other



alternative to increasing statistical power is to increase the



number of replicates, which increases the cost of the bioassay.



     The use of one-tailed tests deviates from the usual



experimental procedure, but is justified where a regulatory



action would be taken only if the tissue residues in organisms



exposed to a test sediment were greater than those in a control



or reference sediment.  For example, a dredge material might be



denied disposal in open water if the tissue residues in the test



sediment  (i.e., dredge material) were significantly greater than



those in the reference sediment.  However, the same regulatory



decision  (i.e., allow disposal) would be reached whether the



tissue residues in the test sediment were equal to or less than



those in the reference sediment.  The same reasoning would apply



when comparing a tissue residue to a tissue criterion.





                                 127

-------
     There  are  cases when a  one-tailed  test  is  inappropriate.



When no  a priori assumption  can be made as to which  treatment  is



higher than the other, a two-tailed  test should be used.   For



example, when comparing TOCs of the  test and reference sediments,



a two-tailed test should be used.  A two-tailed test should also



be used  when one regulatory action will be taken when the  two



treatments  are  equal and another when they are  not equal,



regardless  of which one was larger or smaller.  This would be



unusual  for tissue residues, but would  apply to other benthic



parameters.  For example, a two-tailed  test  should be used when



comparing benthic biomass at a control  and test site because both



enhanced and reduced biomass are indicators  of  organic enrichment



(Pearson and Rosenberg, 1978), so the regulatory question  is



whether  there is any difference between the  two sites.  A  two-



tailed test  should also be used when comparing  tissue residues



among different species exposed to the  same  sediment and when



comparing BAFs or AFs  (see Appendix 1-1).



     The appropriate one-tailed and two-tailed tests for the



bioaccumulation test are summarized in Table XII-1 and



Table XII-2.




2. Comparison-wise versus Experiment-wise Error Rates



     The Type I error rate used in the tests will be chosen



either as a  comparison-wise or experiment-wise error rate



depending on whether one decision is made  for each pairwise



comparison or from a set of pairwise comparisons.   For cases



where test sediments are chosen in a stratified manner or along a






                                128

-------
             TABLE XII-1: Summary of Statistical Analyses
Hypothesis
PAIRWISE COMPARISONS

     Test (s)*
                         Comments
Normality
Equality of Variances
Equality of Means
Equality of Means
Equality of a Mean
  and a constant
Equality of Means
  Chi-square or
Kolmogorov-Smirnov

     F-test
     t-test
  modified t-test
     t-test
nonparametric tests
                         Try transformations
                         if not normal

                         Try transformations
                         if not equal

                         One-tailed with a
                         priori knowledge
                         otherwise two-tailed

                         If variances are not
                         equal

                         One-tailed with a
                         priori knowledge
                         otherwise two-tailed

                         If normality is not
                         established
Hypothesis
MULTIPLE COMPARISONS

     Test(s)*
                         Comments
Normality


Equality of Variances    Bartlett's test
  Chi-square or
Kolmogorov-Smirnov
Equality of Means
Equality of Means
     ANOVA
nonparametric tests
                         Try transformations
                         if not normal

                         Try transformations
                         if not equal

                         If normality is
                         established

                         If normality is not
                         established
  Often more than one test can be used for the same hypothesis.
Each test will have different assumptions.  Chose the test with the
assumptions most closely matching your specific conditions and
requirements.
                                 129

-------
gradient  (see examples a and b in Figure XII-1) and any decisions



will be made on a case by case basis, a comparison-wise Type I



error rate of 0.05 should be used for each comparison.  For



example, a comparison-wise error would be used in deciding which



specific stations along a gradient were acceptable or not



acceptable.



     If the test sediments are selected from a supposedly



homogeneous source {e.g., multiple sediment samples from a dredge



barge, see example c in Figure XII-1) and the decision to accept



or reject the sediment will be made from the results of several



pairwise comparisons, then an experiment-wise error rate of 0.05



should be used.  In this example, a regulatory decision will



depend on the results from all the comparisons of the test



treatments to determine if the sediment in the barge is too



contaminated for disposal.  Each individual comparison is



performed at a lower error rate such that the probability of



making a Type I error in the entire series of comparisons is not



greater than 0.05.  This results in a more conservative test when



comparing any particular sample to the control/reference.  Thus,



a single sediment sample from the barge that would have been



rejected at the 0.05 level may not be rejected at the lower



experiment-wise error rate,  though the probability of rejecting



Ho for the entire set of samples is still 0.05.  Use of



experiment-wise error rates adjusts for the possibility of random



differences when multiple samples are taken from a homogeneous



source  (e.g., if 100 samples were taken, a certain percentage






                                 130

-------
would be greater than the control/reference because of random

fluctuations).   The error rate used in each comparison is a

function of the number of comparisons to be used in the decision

"experiment" and can be computed using the method of Dunn-Sidak

(Sokol and Rohlf, 1981) as:

          alpha' = 1-(l-alpha)1/*                 (1)

where:
          alpha  = Type I error rate used for each

                   pairwise comparison

          alpha  = experiment-wise Type I error rate  (0.05)

          k      = number of comparisons

When an experiment-wise error is used, the power to detect real

differences between any two means decreases as a function of k,

the number of comparisons.


C. MULTIPLE COMPARISONS

     For comparisons involving several means, as in the case of

comparing TOG values among all sediment types, an ANOVA is first

performed to establish whether any of the means are different.

The ANOVA also provides a "best" estimate of the variance

(within-treatment error).  If there are significant differences,

a series of t-tests can be performed for any planned  (a priori)

comparisons  (such as between test and control/reference) to

distinguish which means are different.  For unplanned  (a

posteriori) comparisons, such as between two reference tissue
                                 131

-------
                 FIGURE XII-1
            Sampling Schemes
       for Comparison—wise (a. and b.)
      vs Experiment—wise (c.) Error rates
    Stratified selection of test sediments
   Any
  Harbor,
   USA
       X ~~ test site
    Selection of test sediments
    along  a gradient

   Point source      pollutant gradient
               X  X  X  X   X   X
c.
   X ~ test site

Selection  of test sediments from a
presumably homogenous source
                                  dredge barge
       X ~ test site
                       132

-------
residues, tests such as the T-Method or Tukey Kramer procedure



(Dunnett's test) are more appropriate.  See Sokal and Rohlf



(1981) for unplanned multiple comparison tests to determine which



is most suited for each case.



     It is important to note that an ANOVA is inherently for two-



tailed comparisons.  Therefore if the comparisons can be broken



down into a series of one-tailed pairwise comparisons, it is



preferable to perform the analysis in this manner because of the



increase in power.  However, if the series of comparisons are



two-tailed, an ANOVA can be performed first to determine whether



any additional comparisons should be made.





D. INTERPRETATION OF COMPARISONS OF TISSUE RESIDUES



     If the control mean tissue residues at day 28 are not



significantly greater than the day 0 tissue residues, it can be



concluded that there is no significant contamination from the



exposure system or from the control sediment.  If there is



significant uptake, the exposure system and/or control sediment



should be reevaluated as to suitability.  Even if there is a



significant uptake in the controls, it is still possible to



compare the controls and treatments as long as the pollutant



concentrations in the test tissue residues are substantially



higher.  However,  if control values are high, the data should be



discarded and the experiment conducted again after determining



the source of contamination.
                                 133

-------
     Comparisons between the 28-day control (and/or reference)



tissue residues and 28-day test tissue residues determines



whether there is statistically significant bioaccumulation due to



exposure to test sediment.  Comparisons between control and



reference tissue residues at day 28 determines whether there is



statistically significant bioaccumulation due to exposure to the



reference sediment.  When test tissue residues are compared with



a one-tailed test with a set criterion value  (e.g., FDA Action



Limit),  if no significant difference is detected, the residues



must be considered equivalent to the value even though



numerically the mean tissue residue may be lower.



     The statistical interpretation of these and other tests are



summarized in Table XII-2.





E. ADDITIONAL ANALYSES



1. Testing BAFs and AFs



     Statistical comparisons between ratios such as BAFs or AFs



are difficult due to computation of error terms.  Since all



variables used to compute BAFs and AFs have errors associated



with them, it is necessary to estimate the variance as a function



of these errors.  This can be accomplished using approximation



techniques such as the propagation of error (Beers, 1957) or a



Taylor series expansion method (Mood et al.,  1974).  BAFs and AFs



can then be compared using these estimates for the variance. See



Ferraro et al.  (1990) for an example of this  approach.
                                 134

-------
    TABLE XII-2: EXAMPLES OF ANALYSES AND INTERPRETATION OF RESULTS
HYPOTHESIS
                                TEST
                INTERPRETATION OF
           REJECTION OF NULL HYPOTHESIS
PHYSICAL PARAMETERS
Ho: TOCc=TOCi
Ha:
                             two-tailed  TOC not equal between
                               t-test    control and test sediment i
Ho: TOC =TOCr
Ha: TOCcVTOCr

Ho : TOC =TOCr=TOC1
Ha :
                     . =TOC
                          n
two-tailed  TOC not equal between
  t-test    control and reference

   ANOVA    TOC of one or more
            sediment differs
ADEQUACY OF CONTROL
Ho: Ctc=Ctu
Ha: Ctc>Ctu

TREATMENT DIFFERENCES
Ho: Cti=Ctc
Ha: Cti>Ctc

Ho: Cti=Ctr
Ha: Cti>Ctr

Ho: Ctr=Ct
Ha: Ctr>Ctc

Ho : Ct =Ct =Ct., = . . . =Ctn
Ha : Ct
                             one-tailed
                               t-test
                             one-tailed
                               t-test

                             one-tailed
                               t-test

                             one-tailed
                               t-test

                                ANOVA
            Exposure system
            contaminated
            Sig. uptake from test
            sediment i above control

            Sig. uptake from test
            sediment i aboye reference

            Sig. uptake from reference
            sediment above control

            Uptake from one or more
            sediment differs
Ho :
Ha:
                  _
   ANOVA    Uptake from one or more
            test sediment differs
LONG TERM EXPOSURES
Ho: Ct(j)i=Ct(j+l)i=Ct(j+3)i
Ha: Ct()
                                ANOVA    Ct.^ has not reached
                                         steady-state
Ho = null hypothesis
Ha = alternative hypothesis
Ct = concentration of pollutant in tissue at day 28
Subscripts:  c = control organisms or sediment
             i = l,2,..,n  test organisms or sediment
             j = last sampling period
             n = total number of test treatments
             r = reference organisms or sediment
             u = unexposed organisms
                                 135

-------
2. Comparing Tissue Residues of Different Compounds



     In some cases, it is of interest to compare the tissue



residues of different compounds.  For example, Rubinstein et al.



(1987)  compared the uptake of thirteen different PCB congeners to



test for differences in bioavailability.  Because the values for



the different compounds are derived from the same tissue samples,



they are not independent and tend to be correlated, so standard



t-tests and ANOVAs are inappropriate.  Rather, a repeated



measures technique (repeated testing of the same individual)



should be used where the individual  (experimental unit) is



considered as a random factor and the different compounds as a



second factor.   See Rubinstein et al. (1987)  and Lake et al.  (in



review) for an example of the application of repeated measures to



bioaccumulation data.



3. Analyses for Alternative Test Designs



     Long-term exposures require a test to show that steady-state



has been reached.  An ANOVA should be performed on the last three



sample sets.  ASTM (1984) requires that there be no significant



difference  (p>.05) between the means of these sample sets.  If



apparent steady-state is reached, the mean of the samples taken



during apparent steady-state should be used for the steady-state



concentration value.   For steady-state estimates based on uptake



and depuration tests, see Davies and Dobbs (1984)  or Spacie and



Hamelink (1982) for details on the nonlinear parameter estimation



methods required to estimate these rate constants and steady-



state concentrations.
                                 136

-------
           CHAPTER XIII: REGULATORY STRATEGIES FOR USE
                     OF BIOACCUMULATION DATA


     Bioaccumulation is the link between exposure and effects,

and thus can generate important insights into ecological effects,

human health risks, and the routes and extent of pollutant

exposure.  However, its use in a regulatory context is not as

straightforward as acute toxicity data.  Death is unequivocal and

undeniably bad for an individual organism.  But what of a PCB

tissue residue of 1.5 ppm, or one of 50 ppb?  The answer to

questions such as this are often not clear.  Yet, many of today's

environmental problems are due to the accumulation and trophic

transport of sublethal concentration of pollutants rather than

major die-offs.  Bioaccumulation and consumption of contaminated

seafood is certainly one of the primary environmental concerns of

the public.

     In the previous chapter (Chapter XII), the statistical

procedures to test for increases in tissue residues were

discussed.  It is important to note that statistical differences

in themselves do not necessarily indicate an environmental or

human health problem.  Conversely, the lack of statistically

greater tissue residues in test sediment compared to a proper

control would be strong evidence that the test sediment would not

result in an environmental or human health problem for the

pollutants tested.  The lack of a statistical difference between

a test and reference sediment would indicate that the

environmental problems resulting from the test sediment would be


                                137

-------
no worse than those from the existing reference sediment.



Whether this is environmentally acceptable depends on the present



environmental conditions at the reference site.



     In this Chapter, we summarize some of the approaches to



interpreting the ecological and human health ramifications of a



significant increase in tissue residues in a test sediment or



dredge material.  Some of the approaches are well established,



while other are still in the conceptualization stage.  We only



consider bioaccumulation data in this Chapter, whereas an



environmental assessment would normally include sediment toxicity



testing, benthic community analysis, effluent testing, or a



number of other approaches.  These various approaches generate



different types of information which complement each other.





A. NO FURTHER DEGRADATION



1. Approach



     Compare tissue residues in organisms exposed to a test



sediment to those exposed to an appropriate reference sediment.



If the test tissue residues are not greater than those in the



reference sediment,  it is concluded that the test sediment would



not result in a degradation of existing environmental conditions.



2. Advantages and Applicability



     The approach is straightforward and does not require any



data other than the bioaccumulation tests.  Comparison of



existing or predicted pollution effects to present ecological
                                 138

-------
conditions is an established regulatory approach (e.g., 301(h)).



This approach should be applicable in any area in which an



appropriate reference sediment can be found.



3. Disadvantages



     One potential problem with any approach that uses a field



comparison is the choice of an appropriate reference site.  As



discussed in Chapter II, it is possible to get no statistical



difference between test and reference tissue residues, but still



result in unacceptable degradation.  Without national guidelines,



there may be large differences in the "environmental quality" of



reference sites, and hence, in allowable tissue residues.





B: TISSUE RESIDUE EFFECTS



1. Approach



     Relate the tissue residues to specific physiological or



biochemical effects on the organism.



2. Advantages and Applicability



     The advantage of this approach is that it relates the tissue



residues to an effect on an organism.  If a biochemical end-point



or "biomarker" is used, the approach can generate insights into



the mechanisms of stress.  If such relationships can be



developed, it would be a relatively simple matter to assess the



environmental quality from tissue residues.  The approach is



applicable to all organisms, and depending on the end-points



used, can be adapted for different sensitivities (e.g., early



warning vs. unacceptable impacts).
                                 139

-------
3. Disadvantages



     Because of the various mechanisms by which different



pollutants can affect organisms, and because the actual



mechanisms of toxicity may vary among taxa, there is often a weak



correlation among tissue residues and effects.  The relationship



may also be weak if the stress is due to a small fraction of the



tissue residue (e.g., pollutant affect on nerves).  Presently,



most of the techniques are not simple and often require expensive



equipment.





C. WATER QUALITY CRITERION TISSUE LEVEL APPROACH



1. Approach



     Tissue residues are compared to those residues that would



occur at a water exposure to the Water Quality Criterion (WQC)



concentration.  The tissue concentration which would result from



exposure to pollutants at the WQC are calculated by multiplying



the BCF for a compound by the WQC.  This tissue concentration is



the "Water Quality Criterion Tissue Level" (WQCTL).  Tissue



residues higher than the WQCTL indicate that the integrated



exposure through all routes was greater than allowed under WQC,



and thus is unacceptable.  This approach is similar to the



Equilibrium Partitioning approach to deriving Sediment Quality



Criteria in its use of the WQC as the end-points, except that



tissue residues are used as the measure of exposure rather than



interstitial water concentrations.
                                 140

-------
2. Advantages and Applicability



     As with the Equilibrium Partitioning Approach, the WQCTL



approach is relatively simple, draws upon extensive previous



toxicological work, and is based on a well established regulatory



approach (WQC).   Furthermore, using tissue residues as a measure



of the bioavailable pollutant concentration avoids the numerous



factors which can affect the concentrations and bioavailability



of interstitial water pollutants (e.g., complexation with DOM,



the "solids effects", sampling artifacts, etc.).  Also, this



approach does not assume that interstitial water is the only



uptake route or that an organism's tissue residues can not exceed



a thermodynamic maximum.  As both of these assumptions appear to



be incorrect for some species (e.g., selective deposit-feeders)



with some compounds  (e.g., HCB,  BaP),  relaxing these assumptions



increases the applicability of using WQC as the basis for



regulating sediments.  This approach could be used with any



compound that had a chronic WQC value.



3. Disadvantages



     This is a new proposal, and the idea has yet to be



evaluated.   As with the Equilibrium Partitioning approach, it



assumes that WQC values are applicable to benthic organisms.  The



consistency of BCFs among water column organisms and BAFs or AFs



among benthic species needs to be evaluated.  With benthic



species, the actual exposure may depend upon feeding type (Lee et



al.,  in press)  or source of ventilated water (Winsor et al., in



press).  Therefore, the exposure calculated using one species may
                                 141

-------
not be applicable to a different species.  Rapidly metabolized



compounds may give variable results.  Finally, to the extent that



the WQC was based on tissue residues, the approach may be



circular.





D. FDA ACTION LIMITS



1. Approach



     Compare the tissue residues in the benthic test species with



FDA Action Limits.  Tissue residues exceeding an FDA Action Limit



are considered unacceptable.



2. Advantages and Applicability



     The FDA Action Limits are a well established regulatory



criteria.  The FDA Action Limits are used as end-points in the



evaluation of dredge materials (U.S. EPA/U.S. ACE, 1988).



3. Disadvantages



     There are only a few FDA action limits for seafood.  There



is concern that the criteria are not sufficiently protective of



human health, especially with high seafood consumption rates.



The Action Limits do not consider ecological impacts.  Because of



the human health and ecological limitations, FDA Action Limits



can be considered "one sided" criteria, where exceeding the



limits is unacceptable but failure to exceed the limits is not



strong evidence for acceptability of a sediment.
                                 142

-------
E. HUMAN HEALTH RISK ASSOCIATED WITH SHELLFISH
1. Approach
      Calculate the excess human cancer risk from the consumption
of contaminated shellfish.  The calculation of cancer risk
requires an estimate of the tissue residues, a cancer potency
value, and a lifetime consumption rate of the shellfish.  Non-
cancer human health risks could also be calculated, though in
most cases, cancer will be the greater health risk.  This
approach differs from using FDA Action Limits in that the risk is
based on an estimate of actual human exposure.  Therefore, the
actual allowable concentration would depend upon the rates and
patterns of the consumption of shellfish in an area.
2. Advantages and Applicability
     Determining the excess cancer risk associated with'
consumption of seafood is of obvious and direct concern to the
public.  Such an approach would be applicable in all areas where
shellfish are harvested recreationally or commercially.  The
human cancer model is an established regulatory method.  Cancer
potency values are available for many of the environmentally
relevant pollutants.  For some compounds  (e.g., PCBs, BaP), human
health risk can generate a lower acceptable sediment
concentration than sediment toxicity  (Lee and Randall,  1988), so
for these compounds, this approach is protective of the
environment and human health.
                                 143

-------
3. Disadvantages.



     The local per capita consumption rates of shellfish is



poorly known.  Policy decisions must be made about what



constitutes an acceptable human health risk (e.g., 10"5 versus



10"6 excess cancer risks) and whether to regulate on an average



consumption or the maximum consumption rates (e.g., subsistence



fishing by certain ethnic groups).   Use of default seafood



consumption rates for a site can generate unrealistically low



safe sediment concentrations in areas with little or no



harvesting of shellfish.





F. TROPHIC TRANSFER OF POLLUTANTS INTO PELAGIC FOOD WEBS



1. Approach



     Predict the movement of pollutants from the benthos into



their predators and through the food web, up to and including



human consumers.  The acceptability of a benthic tissue



concentration would be determined from the human health effects



associated with consumption of contaminated fishes (or ecological



effects, if end-points were available).  The same human cancer



risk models would be used as used in calculating the risk



associated with shellfish.  The use of human health risk models



with seafood is discussed in Tetra Tech  (1986e).  This Trophic



Transfer approach is a generalized form of the Shellfish approach



and the two would normally be combined (see Figure XIII-1).



2. Advantages and Applicability



     Trophic transfer of contaminants from the benthos to their



predators is one of the major mechanisms by which certain





                                 144

-------
pollutants are introduced into pelagic food webs.  This approach



addresses the far-field effects of a contaminant as well as the



near-field effects.  The approach would be applicable in any area



where fish are harvested recreationally or commercially.  Models



predicting the average increases in various pollutants per



trophic level have been developed (Young et al., 1987; Young,



1988) .



3. Disadvantages



     One limitation of this approach is the difficulty in



obtaining the data required to quantify the transfer of



pollutants from the benthos to their predators and through the



food web.  The other major limitation is the uncertainty in



assigning human seafood consumption rates.
                                 145

-------
            FIGURE XIII-1
 POSSIBLE REGULATORY STRATEGY FOR
HUMAN HEALTH CRITERIA IN ASSESSING
       SEDIMENT CONTAMINATION
     measure sediment parameters
     predict benthic tissue residues
 fail
          shellfish harvested?
      no
              I  yes
         assess human health
                  I  pass
        fisheries feeding site
  no
                yes
    predict extent of trophic transfer
   I assess human health

fail           I
                     pass
         assess benthic effects
                             I  pass
                            approve
                   146

-------
    APPENDIX 1-1: ADDITIONAL METHODS TO PREDICT BIOACCUMULATION




A. FIELD COLLECTION


     The most direct method of assessing tissue residues in


existing sediments is to measure residues in field collected


organisms.  The field approach is appealing because it avoids any


laboratory artifacts, as well as the time, expense and facilities


required for laboratory tests.  However, use of field collected


organisms has several limitations as a routine method.


     The greatest problem is collecting sufficient tissue biomass
   ^.

of an appropriate species for chemical analysis.  This problem is


especially acute at the most contaminated sites because smaller


species tend to dominate stressed communities and during the


early stages of recolonization (Pearson and Rosenberg, 1978;


Rhoads et al., 1978).  In addition, benthic densities are reduced


under severe stress  (Pearson and Rosenberg, 1978).  Even when


sufficient biomass of a particular species can be collected at a


given station, it will often be impossible to collect the same


species from other stations located along a pollution gradient,


seasonally within a single station, or at an estuarine dredge


site and an open ocean disposal site.


     One possible approach to collecting sufficient biomass is to


composite the various species collected from each site.  Although


mixing species will increase biomass, tissue composites taken


from different stations or seasons are likely to be composed of


substantially different proportions of species and numbers of


individuals.  These compounding factors will make it unclear




                                 147

-------
whether patterns in tissue residues are due to differences in the



sites or interspecific differences in bioaccumulation.  For



example, amphipods have a much greater ability to metabolize PAHs



compared to bivalves  (Varanasi et al.,  1985).  Therefore, a



difference in PAH tissue residues among sites could reflect a



difference in the proportion of amphipods and bivalves rather



than a difference in the bioavailability of PAHs.



     Another problem is that the exposure history of field



collected specimens is usually unknown.  Many benthic species,



especially amphipods and some polychaetes, are mobile during a



portion of their life-history (e.g., Williams and Porter, 1971;



Williams and Bynum, 1972; DeWitt, 1988) and may have migrated



into a site recently.  Although pollutant concentrations in



sediments are usually considered  relatively constant,



resuspension events can obscure sediment-bioaccumulation



relationships.  For example, deposition of resuspended



contaminated sediments in an uncontaminated site would form a



surface veneer available to surface-deposit feeders or filter-



feeders.  In this were the case, a bulk sediment analysis would



underestimate the actual exposure.  Also, field organisms are



potentially exposed to contaminated phytoplankton and to



pollutants dissolved in the overlying water.  If these water



column routes are important, then relating tissue residues to the



field sediment would generate incorrect conclusions regarding



sediment bioavailability.
                                 148

-------
     With these limitations, field collections are not as well



suited as laboratory experiments for the routine prediction of



the tissue residues resulting from dredge materials and pollutant



discharges, or for between-site comparisons of sediment



bioavailability.  Field collections are, however, a powerful



regulatory tool if used in the context of periodic monitoring of



existing sites.  In comparing changes at the same stations over



time, problems with the comparison of different species are



reduced, though there may still be problems with collecting



sufficient biomass.  Field collections also complement the



laboratory studies as a quality assurance check and by providing



data on commercially important species difficult to maintain in



the laboratory  (e.g., lobster).  In some cases, both laboratory



and field assessments of tissue residues are justified by the



size of a discharge or dredging operation or by the high



concentration of pollutants.  Guidelines on sampling designs for



field surveys can be found in Green (1979), Elliott (1983), and



NOAA (1988), while Holme and Mclntyre (1984) contains information



on the sampling techniques.





B. BIOACCUMULATION FACTORS AND ACCUMULATION FACTORS



     Several approaches have been developed to predict benthic



tissue residues directly from sediment concentrations thereby



obviating the need for field collections or bioassays.  The
                                 149

-------
simplest of these approaches is the Bioaccumulation Factor (BAF)



which is:



            BAF = Ct/Cs                     (1)



  Where:



     Ct = tissue concentration (ug/g dry wt)



     Cs = sediment concentration (ug/g dry wt)



     BAFs are empirically derived either from laboratory



bioassays or field collected organisms.  Assuming that BAFs were



constant among species and sediments, multiplying the BAF of a



compound times the sediment concentration would predict the



steady-state tissue residue.  BAFs are analogous to the



Bioconcentration Factors (BCF) which are used to predict tissue



residues from water concentrations:



           BCF = Ct/Cw                      (2)



  Where:



     Cw = concentration in water (ug/g)



     Although the formulas are analogous, BCFs are often



calculated using wet tissue concentrations.



     Sediment characteristics, such as TOG, have a major



influence on the bioavailability of sediment-associated



pollutants and increase the among-site variation in BAFs.  BAF



variability is reduced by normalizing the sediment concentrations



to the TOC content  (Rubinstein et al., 1983).  Normalizing tissue



residues to tissue lipid concentrations reduces variability in



pollutant concentrations among individuals of the same species



and between species (e.g., Veith, 1975; Clayton, et al., 1977).
                                 150

-------
These normalizations are combined in a simple thermodynamic-based

bioaccumulation model for pollutant uptake from sediment  (Lake et

al., 1987; Rubinstein et al., 1987).  The fundamental assumptions

of this thermodynamic model are that the tissue concentration is

controlled by the pollutant's physical partitioning between

sediment carbon and tissue lipids and that the organism and the

environment are at thermodynamic equilibrium.  The method assumes

that lipids in different organisms and TOC in different sediments

partition pollutants in similar manners.  The key value in the

model is the Accumulation Factor  (AF), which when multiplied by

the TOC normalized sediment pollutant concentration predicts the

lipid normalized tissue residue.   (Note: some previous studies

such as Lake et al., (1987) and McElroy and Means  (1988) reported

Preference Factors which are the inverse of the Accumulation

Factor).

     In its simplest form, the model is:

         Ct/L = AF*(Cs/TOC)                  (3)

                or

         AF = (Ct/L)/(Cs/TOC)                (4)

  Where:

     L   = concentration of lipid in organism  (g/g dry wt).
           (decimal fraction)

     TOC = total organic carbon in sediment  (g/g dry wt.)
           (decimal fraction)

     In theory,  AFs should not vary with sediment type or among

species.  Based on the relationship between Koc and lipid
                                 151

-------
normalized BCFs, the maximum AF for neutral organic compounds has
been calculated at about 1.7 (McFarland and Clarke, 1986).
Measured AFs would be lower than this maximum if metabolism of
the compound by the organism is rapid or the organism fails to
reach steady-state body burdens due to limited exposure
durations.  Measured AFs could exceed the calculated
thermodynamic maximum if there is active uptake of the pollutant
in the gut or if there is an increase in the pollutant's gut
fugacity, driving the pollutant from the gut into the body.  The
pollutant fugacity in the gut could increase as the volume of
food decreases during digestion or as a result of the reduction
in the lipids (Gobas et al., 1988).
     Laboratory and field validation of the thermodynamic
partitioning model suggests that for a large number of organic
pollutants, AF values do not exceed the maximum value (Ferraro et
al., 1990).  However, AFs for some highly lipophilic PCB
congeners can exceed the theoretical maximum of 1.7 by as much as
an order-of-magnitude  (Rubinstein, et al., 1987).  Sediments with
the lowest TOCs tend to have the highest AF values (Rubinstein,
et al.,  1987; McElroy and Means, 1988; Ferraro et al., 1990; Lake
et al.,  in review), which is not explained by the present model.
     AFs are also dependent upon the accuracy of the lipid
measurement, and total lipids can vary several fold based on the
extraction technique used.  As discussed in Chapter XI,  we
recommend the Bligh-Dyer lipid method as interim standard for AF
                                 152

-------
determinations.  If another lipid extraction technique is used, a



conversion factor should be provided to allow the conversion of



the lipid values to chloroform-methanol extraction values.



     Although laboratory and field evaluations of the AFs have



shown that they are not statistically constant in all cases



(Rubinstein et al., 1987; McElroy and Means, 1988; Ferraro et



al., 1990; Lake et al., in review), AFs are less variable in



predicting sediment uptake than BAFs (Rubinstein et al., 1987;



Ferraro et al., 1990; Lake et al., in review).  Because of their



minimal data requirements, AFs have great potential as a cost-



effective, first-order estimate of tissue residues.  The



predicted tissue residues can then be used for determining



whether bioaccumulation tests or field surveys are needed.



     For these reasons, the data required to calculate AFs should



be collected and reported in all laboratory tests and field



collections.  Development of an AF database would be extremely



useful in determining the limits of applicability of this



approach, as well as generating the values for specific



chemicals.  After a minimum database has been collected on a



compound, the AFs could be used in deriving a Sediment Quality



Criterion by taking the upper 95% percentile value.





C. TOXICOKINETIC BIOACCUMULATION MODELS:



     Toxicokinetic bioaccumulation models are an alternative to



thermodynamic-based partitioning approaches.  Toxicokinetic



models assume pollutant uptake is a function of the feeding



behaviors and physiological characteristics of the organism.





                                153

-------
Most of these toxicokinetic models (e.g., Norstrom et al., 1976)

assume that the tissue residue can be predicted as the sum of the

uptake from each individual phase (e.g. interstitial water,

ingested sediment)  minus any loss due to depuration or

metabolism.

     In its simplest form, uptake from all phases may be

expressed as:

     dCt/dt =  (Fx*CPx*EPx) - L              (5)

  Where:

     dCt/dt = change in tissue residue with time.

     Fx     = flux of phase x through organism.

     CPx    = concentration of pollutant in phase x.

     EPx    = fraction of pollutant extracted from phase x by the
              organism.

     L      = summation of loss of pollutant through metabolism
              and depuration.

     x      = phase  (W = water, F = food, S = sediment)


     As an example, the uptake from water would be the product of

the amount of water ventilated across the gills  (FW),  the

pollutant concentration in the water (CPW), and the efficiency

with which the pollutant is extracted from the water  (EPW).  As

opposed to the thermodynamic model,  the toxicokinetic model

assumes the uptake from each route is independent and additive,

so an organism exposed to two uptake phases  (e.g., interstitial

water and sediment) would have a higher steady-state tissue

residue than an organism exposed to one phase.  Toxicokinetic
                                 154

-------
models usually assume that uptake efficiency values do not change



as body burdens approach steady-state and that loss (L) can be



modeled as a first-order process.



     These models have been used to successfully predict PCS,



methylmercury, and kepone levels in marine and freshwater fish



 (Norstrom et al., 1976; Jensen et al., 1982; Thomann and



Connolly, 1984).  This approach has only recently been applied to



benthic species, and has been used to model the uptake of



hexachlorobenzene by a marine clam (Boese et al., 1988; Boese et



al., in press; Lee et al., in press;  Winsor et al., in press).  A



slightly different toxicokinetic model has been used to predict



the uptake of various PAHs by freshwater amphipods  (Landrum,



1988, 1989; Landrum and Poore, 1988).  Landrum used this model to



determine the relative importance of interstitial water versus



ingested particulates as an uptake route for these PAHs.



     In contrast to thermodynamic approaches, toxicokinetic



models can predict tissue residues under non-equilibrium



conditions and can account for differences in organism feeding or



ventilatory behaviors due to toxic or natural effects  (e.g.,



growth related changes).  The models can also predict the time



course of uptake and depuration, which can be important in



certain regulatory contexts.  However, the approach requires



relatively sophisticated laboratory experiments to measure the



input parameters.  Because of the extensive data needs and the



ongoing process of developing the laboratory methods,  this
                                 155

-------
approach is not presently suited for the routine prediction of



bioaccumulation.  The toxicokinetic models are appropriate when



detailed analysis of sediment or biological effects on



bioaccumulation are required and as a method to test the



assumptions of various Sediment Quality Criteria approaches.
                                 156

-------
           APPENDIX  1-2.  SAFETY AND WASTE DISPOSAL





A. PERSONNEL SAFETY.




     Personnel involved in any  facet of bioaccumulation  testing,



whether sampling in  the field or performing tests  in  the



laboratory, need to  be protected from exposure to  toxic



chemicals.  Exposure to pathogens must also be considered,



especially when working with sediment collected near  sewage



discharges.  The manner in which personnel will be protected  from



these toxics and pathogens must be determined prior to the start



of any work, keeping in mind that exposure can occur  from



breathing vapors, from physical contact with the skin, or



ingestion of the polluted materials and/or chemicals.  How one  is



protected depends on the type of materials and/or  toxics involved



and is beyond the scope of the manual.  Consult the following



references to determine adequate safety approaches: Sax  (1984),



U.S.EPA (1987b, 1988), ACGIH (1987), and U.S. Coast Guard  (1986).



IRIS (Integrated Risk Information System) is available to local,



state,  and federal public health officials through the Public



Health Network (PHN) of the Public Health Foundation  at  (202)



898-5600 or through  Dialcom, Inc., at (202) 488-0550.





B. HAZARDOUS WASTE DISPOSAL.



     Hazardous waste disposal is a serious problem that  must  be



dealt with properly.  Improper shipping or disposal of toxic




materials may result in environmental damage and/or serious legal



consequences.   The Federal Government has published regulations
                                 157

-------
for the management of hazardous waste and has given the states



the option of either adopting those regulations or developing



their own.  If states develop their own regulations, and about



half of them have, they are required to be at least as stringent



as the Federal Regulations.  As a handler of hazardous materials,



it is your responsibility to know the pertinent regulations



applicable in the state in which you are operating and to comply



with them.  Refer to The Bureau of National Affairs, Inc.,  (1986)



for the citations of the Federal requirements.
                                 158

-------
        APPENDIX  III-l:  DETERMINING NUMBER OF  REPLICATES





     Adequate replication is essential for determining



statistically significant differences between treatments with



sufficient power.   If there is a question that the eight



replicates recommended (Chapter III)  will not provide sufficient



statistical power, then the techniques in this Appendix can be



used to determine the appropriate number.  Determining the



appropriate number of replicates requires estimates of the



variability of each treatment and the minimum detectable



difference.  The minimum detectable difference is the smallest



difference between two means, or between a mean and a constant



value, that needs to be statistically distinguishable.  The



variability is a measure of the within-treatment variation and is



expressed as a standard deviation (SD or s) or coefficient of



variation  (CV) and can be obtained from previous experiments or



the literature.  This information is needed because treatments



with high variation will require more replication to distinguish



differences between treatments than less variable ones.  See



Table III-l for a listing of coefficient of variations for tissue



residues reported for a variety of pollutants.



     The number of replicates required is related to the minimum



detectable difference, and detecting a 2-fold increase in tissue



concentrations requires many more replicates than detecting a



100-fold increase.  There are no standards for an acceptable
                                 159

-------
minimum detectable difference, but we recommend that there should
be sufficient replication to detect 2-fold to 5-fold differences
in tissue concentrations between two treatments (Chapter III).
     Additionally, error rates for Type I and Type II errors must
be chosen.  A Type I error  (alpha) is the probability of
rejecting the null hypothesis when there is no true difference
between treatment means and is usually given a value of 0.05.  A
Type II error (beta) is the probability of accepting the null
hypothesis when there is a true difference between treatment
means.  As discussed in Chapter III, we recommend a beta of 0.05.
This is equivalent to a power of 0.95, where the power of a test
is the probability of correctly rejecting the null hypothesis.
     One equation that can be used to estimate the number of
replicates (n) required to detect a minimum detectable difference
between two means (adapted from Sokal and Rohlf 1981) is:
     n > 2*(s/d)2 *  v)2         (1)
For the comparison of one mean and a constant (e.g., FDA Action
Limit) the formula becomes:
     n >  (s/d)2 * (talphafV + t2beta/v)2           (2)
where:
     n = sample size for each treatment
     s = standard deviation (often a pooled value of the two
          sample variances)
     d = the minimum detectable difference
     v = the number of degrees of freedom  [v = 2*(n-l) for the
         comparison of two means,  v =  (n-1) for the comparison
         of a mean and a constant]

                                 160

-------
     alpha = Experiment-wise or comparison-wise Type I error
             (see Chapter XII).   If a two-tailed test is performed,
             each tail will consist of alpha/2.  If a one-tailed
             test is performed,  the single tail is alpha.
     beta = Type II error  (or 1 - power of test).
     ^"alpha v = critical value for alpha of Student's t-
            distribution with v degrees of freedom.  (Use a
            two-tailed t-table for a two-tailed test and a
            one-tailed table for a one-tailed test.)
     fc2beta v = critical value for 2*beta of Student's t-
            distribution with v degrees of freedom.  (Use a
            two-tailed table.  If a one-tailed table is used, the
            critical value is beta.  The critical value is the
            same whether the test is one- or two-tailed.)
     An iterative approach is used to calculate n since talpha v
and t2beta v are dependent on n through v.  The values for
talpha,v t2beta,v alPha» beta, and v are either set by the
investigator or found in tables.  Therefore, only the standard
deviation and the minimum detectable difference must be
estimated.  Although a minimum detectable difference (d) of 2 is
recommended (see Chapter II), an estimate of the standard
deviation will not be available in many cases.  However, the
ratio of the two (s/d) can be described in several ways,
providing different approaches to estimating these parameters.
Three methods of estimating s/d and their advantages and
disadvantages are:
                                 161

-------
Method tti.



           (s/d) =  [s/(Ul-u2)]                      (3)



Where:



          u-^-Uj = difference between mean U]_ and mean u2,



                   or mean u-^ and a constant.



     Advantages: There may be cases when an absolute difference



between two numbers is of interest such as in a comparison  of a



measured tissue residue and a regulatory action limit.



     Disadvantages: Requires an estimate of the standard



deviation of the sample, a value often difficult to obtain.





Method #2.



           (s/d) =  [(CY/lOO/n^]                    (4)



Where:



          CV = Coefficient of Variation  (expressed as a percent)



          m.^ = a multiplicative factor of u-^ that is the



               minimum detectable difference between mean u^ and



               mean u2  (or criterion value)  (e.g., if m-^ =  5,



               the minimum detectable difference between u-^



               and u2 will be five times the value of u^).



     Advantages: The CV is often easier to estimate than the



standard deviation.  The CV's in Table III-l can be used as



estimates if no other information is available, though it would



be prudent to consider these values as the minimum estimates of



variation.
                                 162

-------
     Disadvantages: The value for m-^ will change whether



comparisons are between control and test values or test and a.



criterion value.  Control values  (tissue residues) will tend to



be low in comparison to the test values  (tissue residues) while



test values may be large and close to a criterion value  (e.g.,



FDA action limits).



Method #3.



          (s/d) = [s/(m2*s)] =  [l/m2]              (5)



Where:



          m2 = a multiplicative factor of s. For example, if



               m2= 2, the minimum detectable difference is 2



               standard deviations  (i.e, u2 will have to be 2



               standard deviations from u1 to be able to detect



               a difference).



     Advantages: No estimates are required of the standard



deviation or CV.



     Disadvantages: The value of m2 may have to vary whether



comparisons are between control and test values or test and



action limits.



     If a comparison between more than two means is anticipated



(as in the determination of steady-state conditions), see Sokal



and Rohlf (1981) for a modification of this approach or Tetra



Tech  (1986a) for tables of estimates.
                                 163

-------
     APPENDIX  IV-1: ADEQUACY OF  10-DAY AND  28-DAY EXPOSURES





     Ideally, organisms should be exposed to test sediments for a



period sufficient to attain steady-state tissue residues.



However, cost considerations often prove prohibitive to



conducting tests long enough to document that steady-state has



been attained.   As a result, bioaccumulation tests historically



have been conducted for a preset duration.  Choosing a single



time period is complicated by the multitude of organic pollutants



and metals found in most field sediments or dredge materials,



each with differing uptake kinetics.  To date, a ten-day exposure



to assess "bioaccumulation potential" has been the most commonly



used time period for the testing of marine sediments (primarily



dredge materials)  (U.S. EPA/U.S. ACE, 1977).  Bioaccumulation



potential is the potential for any uptake of a pollutant by



organisms exposed to a sediment,  and the basic premise was that



if there was going to be bioaccumulation it should be possible to



detect it within 10 days.  Thus,  the original intent of the 10-



day test was as qualitative rather than a quantitative measure.



Since 1977, however,  data from 10-day tests have frequently been



extended beyond its original intent and used as a quantitative



result.



     Because of the wide-spread use of the 10-day exposures, it



is worth assessing its utility both as qualitative measure of



bioaccumulation potential and as a quantitative method to



generate data for ecological and human health risk assessments.



The percent of steady-state tissue residue obtained after 10 days





                                 164

-------
for several organic pollutants was used as a simple measure of



accuracy  (Table IV-l).  To adequately assess bioaccumulation



potential, the 10-day exposure should result in a sufficient



percentage of the steady-state tissue residues to identify which



sediments could be an environmental problem.  Also, the



percentage of the steady-state tissue residue obtained should be



relatively consistent for the same pollutant in different



species.  That is, the 10-day exposure should give a strong and



consistent "signal".  In the quantitative risk assessments, the



benthic tissue residues will be used to predict the amount of



pollutants transported from the sediment to higher trophic



levels, including man.  A large error at the base of the food-web



will result in errors throughout the analysis, especially as some



of the errors may be multiplicative.  As a preliminary measure,



we suggest that for data to be acceptable for quantitative risk



assessment, the resulting tissue residues should be within 80% of



the steady-state tissue concentrations.  An accuracy of 80% for



each trophic step results in the prediction of tissue residues



being within two-fold of the actual residues for a three step



chain  (i.e.,  sediment to benthos to demersal predator to higher



predator or man; or 0.8*0.8*0.8 = 0.51)



     In these studies,  only 29% of the organisms approached



within 80% of the steady-state level in ten days (Table IV-1).



Ten-day tissue residues averaged 51% of the estimated steady-



state value,  and this average included some rapidly accumulated



PAHs.   Tissue residues of PCBs achieved after 10 days averaged





                                 165

-------
only about 25% of the steady-state values,  and ranged from 100%



to a low of 9%.  Other environmentally important compounds with



high Kow compounds,  such as DDT,  dioxins,  and BaP,  are expected



to be similar to the PCBs.



     Ten days is also likely to generate a relatively low



percentage of the steady-state tissue residues for metals.  For



example, mercury levels in fish may not attain steady-state



during the lifetime of the organism (River et al.,  1972; Cross et



al.,  1973), and the minimum time for lead to attain steady-state



in Mytilus edulis was greater than 230 days  (Schulz-Baldes,



1974).



     Based on this preliminary review, we reach several



conclusions.  First, a 10-day exposure generates a low percentage



of the steady-state tissue residues for PCBs and presumably other



high Kow organics and some heavy metals.  These compounds are the



most likely to represent an ecological and human health risk



through bioaccumulation and biomagnification.  Second, the



percentage of the steady-state tissue residue obtained varies



several-fold even within a single compound.  Third, the amount



accumulated within ten days is such a small percentage of the



steady-state concentration that it may be below detection limits



of standard analytical methods or may not be significantly



different than control values.  Thus, the 10-day exposure can



result in false negatives concerning the bioaccumulation



potential of a sediment.  Fourth, the percentage of the steady-



state tissue residues accumulated over 10 days is inadequate for





                                 166

-------
TABLE IV-1A: Information Gained and Requirements of Different
             Approaches to Estimating Benthic Tissue Residues
METHOD
BIOACCUM.
POTENTIAL
FALSE NEG.
BIOACCUM.
POTENTIAL
 ESTIMATES
EQUILIBRIUM
  RESIDUE
 ADDITIONAL
REQUIREMENTS
Accumulation     Yes
Factors

10-Day Test      Yes
28-Day Test      Yes
Kinetic Models   Yes
              No
              Yes
              No
              No
              Yes?
               No
             Approx.
             to Yes

              Yes
Long-Term
Exposures
 Yes
  No
  Yes
Sed Cone.,
TOG, Lipids

10 Days Lab
Time, Tissue
Cone.

18 Days Addi-
tional Lab Time

Additional
Tissue Cone.,
Additional Lab
Time?, Develop-
ment of Techni-
ques

23-70 Days
Additional Lab
Time, Additional
Tissue Cone.
Bioaccum. Potential = Qualitative ability to detect uptake.

False Negative Bioaccum. Potential = Amount accumulated is so low that
  it is incorrectly concluded that no uptake will occur.

Estimates Equilibrium Residue = Tissue residue data sufficiently
  accurate for use in quantitative risk assessments.


Exper. Techniques = Resources devoted to determining the correct uptake
       and depuration periods for specific compounds and organisms
Lab Time = Laboratory time required for biological exposure
Lipids = Tissue samples analyzed for lipid content
Sed. Cone. =  Sediment samples analyzed for pollutants
Tissue Cone. = Tissue samples analyzed for pollutants
TOC = Sediment samples analyzed for TOC
                                 167

-------
a quantitative risk assessment.  Lastly, the  10-day exposure does

not generate any additional insights into bioaccumulation

potential that are not generated by use of the Accumulation

Factors  (see summary in Table IV-1A)

     For these reasons, we conclude that an exposure duration

longer than 10 days is required.  Based on the use of 28 day

exposures in the bioconcentration tests (ASTM, 1984), we

recommend a 28-day exposure as a practical compromise between

cost, data accuracy, and data utility.  When  28-day organic

pollutant levels were compared to observed or estimated steady-
                                                              t
state levels (Table IV-l), steady-state tissue residues were

approached (i.e.^80% of steady- state) in 76% of the tests, and

the mean steady-state pollutant tissue level  increased to 86% of

the steady-state maximum.  An average of 82%  of the PCB steady-

state tissue residues was obtained after 28 days, though in one

instance the value was only 25% of the steady-state residue.

This level of accuracy should be sufficient in nearly all cases

to test for bioaccumulation potential with a  reasonable level of

statistical certainty.  In most cases, the data should be

sufficiently accurate for quantitative risk analysis.  In cases

when more accurate estimates are required,  either a long-term

exposure or a kinetic approach can be used (Chapter IV).

     Besides underestimating tissue residues because of

insufficient duration, single point tests can underestimate

maximum tissue residues when a compound reaches a maximum value

before the sampling period and then declines.   For example,
                                 168

-------
phenanthrene approaches its maximum tissue residue in freshwater



amphipods after about 10 days and then declines (Landrum, 1989).



In this case, a 28-day test would generate a lower value than a



10-day test.  Presumably,  the decline is the result of an



increase in the metabolic degradation rate of the pollutant, and



should be most common with the lower molecular weight PAHs.



Because the ability to degrade PAHs varies among taxa (Varanasi



et al., 1985), a decline in tissue residues should be most



pronounced with amphipods and less so with bivalves.  If low



molecular weight PAHs or other rapidly metabolized compounds are



of interest, time series samples should be taken before day 28



(see Chapter IV).
                                 169

-------
             APPENDIX IV-2: ALTERNATIVE TEST DESIGNS






A. SHORT-TERM TEST



     Some compounds (e.g.,  volatiles)  may attain steady-state in



less than 28 days (see Table IV-1),  so that a 28-day exposure may



not be necessary.  Generally,  10-day tests should be acceptable



with organic compounds which have log Row's <3 that have been



spiked into sediments.  Even with these compounds, a 10-day test



should only be used after it has been documented to approach



steady-state in phylogenetically similar species in less than ten



days, or that the depuration rate (k2) in phylogenetically



similar species is >0.5/day.  When determining the



bioaccumulation of pollutants from field sediments, however, a



28-day test should be used because nearly all field sediments



contain some pollutants with slow uptake kinetics.  Biotic and



abiotic samples should be taken at day 0 and day 10 following the



same protocol as used for the 28-day tests.  If time-series



biotic samples are desired, sample on days 0, l, 3, 5, 7, 10.





B.  ESTIMATING STEADY-STATE FROM UPTAKE RATES



     In theory, it is possible to estimate both kl and k2 from



the uptake phase alone if the experiment continues past the point



when the tissue residues begin to "bend over", indicating that



depuration is sufficient to slow net uptake  (Figure IV-1).  This



approach obviates the need to run a separate depuration



experiment, as is required in kinetic approach discussed in



Section C of Chapter IV.  However, since both kl and k2 are






                                 170

-------
estimated from the fitting of mathematical models, this method is



less reliable than the kinetic approach which uses independent



measures of kl and k2.  Nonetheless, this approach has utility



when time or analytical support is limited, or if a long-term,



time-series uptake test is terminated before steady-state is



attained.  In this design, the sampling schedule should follow



closely that of the uptake phase of the kinetic approach using



both uptake and depuration rates (Chapter V).  Refer to Branson



et al.  (1975) and Foster et al. (1987) for the specifics of



estimating kl and k2.





C. GROWTH DILUTION



     If test organisms grow during an experiment, growth



dilution, the dilution of pollutant concentrations in the tissues



by the increase in tissue mass, will occur.  Taking an extreme



example, if an organism doubled its weight during a depuration



study, it would appear that half of pollutants had been depurated



even if none of the pollutants were excreted from the organism.



Without correction for growth, the depuration rate  (k2)



calculated from this experiment would be incorrect for an



organism growing at a different rate.  Many experiments have not



taken growth dilution into account, which may contribute to the



variation among measured depuration rates  (see Niimi et al.,



1981).
                                 171

-------
     In 28-day experiments, growth dilution is not a problem if



growth is relatively slow and kinetic rate constants are not



derived from the data.  However, for the kinetic approach, growth



dilution can cause errors in estimating uptake and depuration



parameters,  resulting in errors in predicting steady-state



concentrations and time to steady-state.



     If substantial growth occurs during experiments to determine



the rate constants, uptake rate constants will be underestimated



and depuration rate constants will be overestimated.  If these



erroneous constants are used in the kinetic model (Equation l of



Chapter IV)  under conditions of no growth, both steady-state



tissue concentrations and time to steady-state will be



underestimated.  Conversely, an error occurs when correct (i.e.,



derived under no growth) uptake and depuration rate constants are



used in this kinetic model when the organisms are growing.  In



this case, both the steady-state concentrations and time to



steady-state will be overestimated because the model does not



compensate for growth dilution.



     If possible, experiments should be conducted with organisms



that grow very slowly or under environmental conditions that keep



growth at a minimum (such as low temperatures).  If growth can



not be avoided, then growth dilution must be taken into



consideration if a kinetic approach is used.  Assuming that



growth dilution is a first-order process and that growth occurs
                                 172

-------
at a constant rate, the kinetic model  (Equation 1 of Chapter IV)



becomes:



    Ct(t) = kl*Cs/(k2+k3)*[l-e"(k2+k3)*t]    (1)



where:



          Ct = concentration in the organism at time t



          Cs = concentration in the sediment



          kl = the uptake rate constant  [days"1]



          k2 = the depuration rate constant  [days' ]



          k3 = the growth rate constant   [days'1]



          t = time [days]



       The growth rate constant (k3) can be measured from the



change in weight during the exposure experiment or during a



separate growth experiment under similar environmental



conditions.  Equation l assumes that the kl and k2 values are



true uptake and depuration constants measured under conditions of



no growth or, if growth occurs, then growth dilution was taken



into account.  If the depuration rate is measured while organisms



are growing, the rate measured will actually be a function of



growth and depuration and can be modeled as k2+k3.



     Under conditions of growth, and using an estimated growth



constant (k3),  the maximum tissue residues becomes




     ctmax = kl*Cs/(k2+k3)               (2)
                                 173

-------
       APPENDIX  IV-3:  CALCULATION OF  TIME  TO STEADY-STATE





     Having an estimate of the time to reach steady-state tissue



residues is very helpful in designing long-term studies and



assessing the adequacy of a 28-day test.  If no estimate for a



pollutant in phylogenetically similar organisms is available, the



time required to approach steady-state can be estimated from a



linear uptake, first-order depuration model  (see Chapter IV,



Section C).  This model is an approximation for benthic



invertebrates as it was developed for fish exposed to dissolved



organic contaminants.



     Uptake of organic pollutants from water  (dissolved phase)



has been modeled in fish species using a linear uptake, first-



order depuration model  (Spacie and Hamelink, 1982):



     Ct(t)  = kl*Cw/k2*(l-e"k2*t)                          (1)



  Where:



     Ct = pollutant concentration in tissue at time t



     Cw = dissolved pollutant concentration in water.



     kl = uptake rate constant.   [days'1]



     k2 = depuration rate constant.   [days'1]



     t = time  [days].



     This model predicts that equilibrium would be reached only



as time becomes infinite.  Therefore, for practical reasons,



apparent steady-state is defined here as 95% of the equilibrium
                                 174

-------
tissue residue.  The time to reach steady-state can be estimated



by:



  S = In[l/(1.00-0.95)]/k2 = 3.0/k2                (2)



  Where:



     S = time to apparent steady-state  (days)



     Thus, the key information is the depuration rate of the



compound of interest in the test species or phylogenetically



related species.  Unfortunately, little of this data has been



generated for benthic invertebrates.  When no depuration rates



are available, the depuration rate constant for organic compounds



can then be estimated from the relationship between Row and k2



for fish species  (Spacie and Hamelink, 1982):





   k2 = antilog[1.47-0.414*log(Row)].               (3)



     The relationship between S and k2  (using Equation 2) and



between k2 and Row (using Equation 3) is summarized in Table IV-



3A.  This table may be used to make a rough estimate of the



exposure time to reach steady-state tissue residues if a



depuration rate constant for the compound of interest from a



phylogenetically similar species is available.  If no depuration



rate is available, then the table may be used for estimating the



S of organic compounds from the Row value.  However,  as this data



was developed from fish bioconcentration data, its applicability



to the kinetics of uptake from sediment-associated pollutants is



unknown and the estimated S values should be considered as



minimum time periods.  Also, Equation 2 does not account for
                                 175

-------
           Table IV-3A:  Estimated Time to Obtain 95% of
                   Steady-State Tissue Residue

Estimated time  (days) to reach 95% of pollutant steady-state

tissue residue  (S) and depuration rate constants  (k2) calculated

from octanol-water partition coefficients using a linear uptake,

first-order depuration model (Spacie and Hamelink, 1982).  k2

values are the amount depurated  (decimal fraction of tissue

residue lost per day).  Note that the calculated k2 values for

log Row values <3 are expressed in hours and are based on

extrapolation from the relationship between k2 and Kow, as

described by Spacie and Hamelink  (1982).  As a result, caution

should be excercised in using these particular k2 values.




Log Kow                    k2                  S  (days)


   1                     °-48!                    °-3
   2                     0.17                     0.7
   3                     0.07                     1.8
   4                     0.65                     4.6
   5                     0.25                    12
   6                     0.097                   31
   7                     0.037                   80
   8                     0.014                  208
   9                     0.006                  540


     * k2 values expressed in hours, all other k2 values
     expressed in days.
                                 176

-------
growth dilution  (Appendix  IV-2).  To  correct  for growth  dilution,



Equation 3 becomes



     S = In[l/(1.00-0.95)]/(k2+k3) =  3.0/(k2+k3)   (4)



Where:



     k3 = growth rate constant  [days"1]





     Using a linear uptake, first-order depuration model to



estimate exposure time to  reach steady-state  body burden for



metals is problematical for a number  of reasons.  The kinetics of



uptake may be dependent upon a small  fraction of the total



sediment metal load that is bioavailable  (Luoma and Bryan, 1982).



Depuration rates may be more difficult to determine, as  metals



bound to proteins may have very low exchange  rates  (Bryan, 1976).



High exposure concentrations of some  metals can lead to  the



induction of metal binding proteins,  like metallothionein, which



detoxify metals.  These metal-protein complexes within the



organism have extremely low exchange  rates with the environment



(Bryan, 1976). Thus the induction of  metal binding proteins may



result in decreased depuration rate constants in organisms



exposed to the most polluted sediments.  Additionally, structure-



activity relationships that exist for organic pollutants (e.g.,



relationship between Row and BCFs) are not well developed for



metals.
                                 177

-------
       APPENDIX V-l:  TECHNIQUES  FOR SEDIMENT MANIPULATION





     In this appendix we summarize techniques to experimentally



manipulate sediment for bioaccumulation tests.   This is an area



of much interest,  and these guidelines may need revision in the



future based on the ongoing research.





A. SEDIMENT SPIKING



1. Methods Used To Add Pollutants To Sediment



     A variety of methods have been used to dose or "spike"



sediments with pollutants.  Toxicants can be added to overlying



water and allowed to partition with the sediment (Breteler and



Saksa, 1985; Pritchard et al.,  1986), added to dry sediment and



mixed by stirring or agitation (Adams et al., 1985; Foster et



al., 1987), added to wet sediment and mixed by stirring or



agitation  (Stein et al.,  1987), added with a solvent carrier, or



evaporated on the sides of jars and the sediment mixed by rolling



in the jars (Swartz et al.,  1986; McElroy and Means 1988; Boese



et al., in press).  Sediments are also spiked by suspending them



in aqueous solutions of the pollutants that contain carrier



solvents (McLeese et al., 1980).   Alternatively,  carrier-free



aqueous solutions of pollutants can be prepared using generator



columns (Veith et al., 1975),  which then can be used to spike the



suspended sediment.



    Solvents are often used as carriers to add hydrophobic



pollutants to water and sediment.  However, carriers can alter



pollutant bioavailability (Dalela et al., 1979; Hughes et al.,





                                 178

-------
1983) as well as pollutant partitioning (Nkedl-Kizza et al.,



1985) .   Although the exact effects of solvents are currently



being debated, there is sufficient evidence to recommend that



carrier solvents not be added directly to the sediment whenever



possible.  If a carrier solvent must be used, the amount should



be minimized.



     The effects of these various techniques on bioaccumulation



have not been tested, and recent work indicates that DOC may



increase proportionally with the magnitude of disturbance



(mixing), continue to increase after the disturbance, and fail to



return to previous DOC concentrations after 10 weeks (DeWitt, T.,



pers. comm.,  OSU, Mar. Sci. Ctr., Newport, OR).  However, tests



have also shown that bioaccumulation of a PCB was not affected by



a long rolling time  (McElroy and Means, 1988).  Until standard



methods are developed, it is recommended that appropriate caution



be exercised in comparing spiked and field results and when



comparing results from sediments spiked by different techniques.



2. Spiking Methodology



     The following is a summary of the method we have used to



spike two sediment types, a fine-grained sand and a silt, with



hydrophobic pollutants.  These techniques should work well with



similar sediments, but those with drastically different



properties, such as cohesive muds, may require modification.  The



advantages of this method are that it avoids the addition of



solvents to the sediments and assures good physical mixing.
                                 179

-------
     Dissolve the pollutant with an appropriate solvent and place



an aliquot of the mixture into glass jars.  Roll jars in a fume



hood while evaporating the solvent with a gentle stream of



nitrogen or purified air.  After the solvent has evaporated, add



wet sediment with a sufficient water content to allow adequate



movement of the sediments.   We have spiked 2 kg of wet sandy



sediment in 1 gallon jars, and up to 13 kg of wet sandy sediment



in 3 gallon jars.  Leave adequate space in the jar for the



sediment to roll.  Cap jars and roll the sediment slurry



continuously at approximately 12 rpm.  The rolling apparatus can



either be home-built or a commercial ball mill.  Place a catch



basin under the rolling apparatus to contain any contaminated



sediment in case of leakage or breakage.  The sediment can be



rolled at room temperature or 4°C, but the cooler temperature



will slow microbial degradation of carbon and organic pollutants.



At the end of the rolling period, place the jar in a fume hood



and stir the sediment with a large TeflonR-coated stainless steel



or polypropylene spoon.



     Appropriate solids to water ratios are essential when



spiking sediment in order to assure pollutant homogeneity. Sandy



sediments with water content of about 30% and fine grained



sediments with water contents of 60% have been mixed



successfully.  However, sediments higher in organic carbon or



clays may require higher water contents for successful mixing.



Low water content is indicated by "balling up" of the sediment



and/or inadequate movement of the sediment in the rolling jar.
                                 180

-------
The sediment needs to slide on the sides of the jar, if only



slightly, while it is being rolled.  If additional water is



added, the excess can be removed by pipetting or siphoning after



the sediment has settled.  Overnight settling is usually



sufficient, but longer periods may be required for particularly



fine sediments.  Sediment can also be centrifuged to remove the



overlying water, though this may result in a sediment with too



low a water content.  If that occurs, water may be added back to



the solids by stirring the sediment or sieving it through a 1 mm



(or appropriate size) screen with a portion of the decanted



water.



    The mixing times required to assure a homogeneous dispersion



of the spiked pollutant may vary, but sediments spiked with



hydrophobic pollutants (e.g., HCB, PCBs)  are physically



homogeneous within one to three days.  Although the pollutant



concentration is homogeneous in a few days, we normally roll the



sediments for ten days to allow the time for the pollutant to



partition among the various sediment phases.  At the end of ten



days, at least 3-8 replicate samples of the spiked sediment



should be analyzed to confirm pollutant homogeneity.  If the



sample is not homogeneous,  the mixing should continue until



homogeneity is achieved.



     Sediment can be spiked with a water soluble (hydrophilic)



compound by first dissolving the pollutant in water of



appropriate salinity and then following the previously described



steps of adding and rolling the sediment slurry.   This procedure
                                 181

-------
differs only in that the water (solvent)  is not evaporated prior



to adding sediment.  Hydrophilic pollutants may not require a 10



day rolling period to achieve partitioning among the water and



the solid phase.





B. INCREASING SEDIMENT TOC.



     The organic content of a sediment can be increased to



determine effects of organic enrichment on the benthos or effects



of organic matter on pollutant bioavailability.  The organic



content can be increased by adding sewage sludge, manure, humics,



natural detritus or other organic-rich materials.  Such



experiments can generate important insights,  but the



investigator must be aware of several potential confounding



factors.  First, though sediments may be equal in TOC



concentrations, not all organic matter reacts in the same manner.



A highly labile organic source (e.g., dried baby food) is more



likely to drive a sediment to anoxia than a refractory carbon



(e.g., natural detritus).  Second, addition of organic matter



will probably change physicochemical properties such as grain



size, Eh, sorption capacity, and texture of the sediment.



     Lastly, the effect of the organic matter addition on



pollutant bioavailability can vary with the feeding behavior of



the test species (Lee et al., in press).   Figure V-1A illustrates



the effect selective feeding can have on ingested pollutant dose



in natural sediment and in organically enriched sediment.  In



this case, the organism selects particles similar in size and
                                 182

-------
Legend to Figure V-1A: Effect of Sediment Selection on Ingested

Pollutant Dose in Natural and Organically Enriched Sediments:

Figure V-1A illustrates a surface deposit-feeding clam, Macoma
nasuta. in a clambox  (see Appendix IX-1).  The clambox separates
the inhalant (left side) and exhalant (right side) siphons,
allowing the collection of fecal pellets for a determination of
the size and pollutant concentration of ingested particles. In
the upper illustration, the clam is exposed to a natural sandy
sediment, but selectively ingests the finer, high TOC particles.
In the lower illustration, the clam is exposed to a sediment
enriched with finer, high TOC particles.  The clam ingests fine
particles with a similar size and TOC content as in the sandy
sediment.  Because the clam ingests particles with a similar TOC
content, and hence concentration of organic pollutants, the
clam's ingested pollutant dose is similar in the two sediments
even though the bulk sediment pollutant concentrations are
dissimilar.
                                 183

-------
                     FIGURE V-1A
 Effect  of Sediment  Selection on Ingested Dose

in Natural and Organically Enriched Sediments
           Inhalant
             siphon
         «   «
                                    Exhalant
                                      r siphon
                  t     CLAMBOX


                  Natural  sediment
                                                  Fecal
                                                  " pellets
                                Dental dam
             Organically Enriched Sediment
           SJLND GRAINS, LOW TOG, LOW POLLUTANT CONC.

           ORGANIC PARTICLE, HIGH TOG, HIGH POLLUTANT CONC.
                         184

-------
organic content in the two sediment, resulting a smaller



difference in the ingested pollutant dose than would be predicted



from bulk sediment analysis.



     Probably the most "natural" way of augmenting sediment TOC



is either by enriching a sediment with the fine fraction



collected from the same sediment type  (Swartz et al., 1985a) or



by using the fine fraction directly  (DeWitt et al., 1988; Boese



et al., in press).  These methods use naturally occurring organic



matter, though grain size distributions of the test sediment is



substantially altered.  The fine fraction can be obtained by wet



sieving the sediment through a series of sieves.  After sieving,



the fine fraction can be dewatered by allowing the solids to



settle for 16-24 hours before siphoning off excess water.  Excess



water can also be removed by centrifugation and decanting.



Similarly, other high organic materials such as sewage sludge



solids can be dewatered by centrifugation and then added to the



test sediment (Swartz et al.,  1984).





 C. DECREASING SEDIMENT POLLUTANT AND/OR TOC CONCENTRATIONS



     Bulk sediment pollutant concentration and organic content



can be decreased by adding control sediment,  clean sand, ashed



portions of the same sediment,  and other inert inorganic



materials, or by chemical extraction techniques.  Such a



manipulation can be used to reduce the toxicity of a field



sediment (Swartz et al.,  1989)  so that its pollutant



bioavailability or LC50 can be determined.  These techniques,



however,  can unexpectedly change the pollutant sorption capacity





                                 185

-------
and the bioavailability of both metals and organics.  Also,



depending on the feeding behavior of the test species, addition



of particles which are not ingested may have little affect on the



ingested pollutant dose (see Fig. V-lA).



     The least disruptive method of diluting sediments is not



clear.  Using a control sediment with a grain size similar to



that of the test sediment will help maintain the physical



characteristics of the sediment.  However, by adding many



partitioning sites, the addition of a control sediment may



totally disrupt the distribution of the contaminant.  Instead,



the addition of clean sand, with few binding sites, may be a



better alternative (Landrum, P., pers. comm., NOAA, Great Lakes



Enviorn. Res. Lab., Ann Arbor, MI).  The advantages and



disadvantages of the various approaches requires further



investigation.
                                 186

-------
       APPENDIX VII-1: SELECTION CRITERIA FOR TEST SPECIES





A. INDIGENOUS VERSUS SURROGATE SPECIES



     The first decision in choosing test species is whether to



use a representative indigenous species or a surrogate species.



The supposed advantage of indigenous organisms is that they are



the same species which will be impacted by the dredge material or



discharge.  However, benthic communities can undergo drastic



fluctuations in species composition in response to natural (e.g.,



Frankenberg and Leiper, 1977) and pollution events (Pearson and



Rosenberg, 1978) and during recolonization (Rhoads et al., 1978).



Because of this variation, the indigenous species chosen for



laboratory testing may not be closely related phylogentically or



ecologically to the species at the impacted site.



     Many of the common indigenous species do not meet the



criteria for use as a bioaccumulation test species, negating any



advantage of using a native species.  Even when an indigenous



species is acceptable, established surrogate test species offer



several advantages.  There is considerable information on the



maintenance and biology of the recommended test species.



Additionally, as more tests are conducted on these species, a



database will develop allowing the comparison of bioaccumulation



under different environmental conditions.  As ASTM (1984) has



pointed out, it is more advantageous to gather detailed



information on a few species rather than a smattering of



information on a large number of species.
                                 187

-------
     Our recommendation is to use surrogate species for routine



monitoring of sediments and discharges.  If there are local



species which appear to meet the various criteria discussed



below, they can be tested along with the recommended



bioaccumulation species.  If the local species prove acceptable



and the results intercalibrate with the results from the standard



species, the local species could be substituted for the standard



species in future tests.  Local species that do not meet the



criteria but are of special concern (e.g., lobster) can be tested



in addition to the surrogate species but should not be



substituted for them.





B. REQUIRED CRITERIA:



     As ingested sediment can be a major uptake route for higher



Kow compounds (Landrum, 1989; Boese et al., in press),  test



species must ingest sediment.  Using a filter-feeder,  in which



the only route for bedded-sediment exposure is from interstitial



water, may underestimate the total uptake from sediments.  Many



benthic invertebrates are flexible in their feeding mode, and



this requirement does not preclude the use of facultative filter-



feeders (e.g., Macoma spp)  as long as the only route of uptake



during the exposure is from bedded sediment (i.e,  no resuspended



particles)  and as long as the bedded sediment supplies adequate



nutrition.



     The requirement for a sediment ingesting species excludes



obligate filter-feeders, such as Mercenaria.  Mya.  oysters (e.g.,



Crassostrea) and mussels (Mytilus spp), as well as obligate





                                 188

-------
predators such as Glycera spp.  If there is concern about  the



human health consequences of the tissue residues  in edible



filter-feeding bivalves, they should be tested  in addition to but



not substituted for the standard test species.  To accurately



predict the tissue residues in filter-feeders,  a  resuspension



exposure system is required  (see Appendix IX-1).



     Test species must be sufficiently pollutant  resistant to



survive the duration of the exposure with a minimum level  of



mortality.  For bioconcentration tests, ASTM  (1984) states that a



test is unacceptable if more than 10% of the organisms  "died or



showed signs of disease, stress, or other adverse effects."  This



requirement is based on deriving BCFs for single  compounds in



which it is possible to control the pollutant concentration.



This requirement appears too strict for testing of environmental



sediments in which it is difficult to meaningfully manipulate



toxicity.  Nonetheless, if a pollution-sensitive  species is



sufficiently stressed to inhibit normal feeding,  the resulting



tissue residues may underestimate the amount bioaccumulated by a



more hardy species.  Additionally, excessive mortality  can create



problems in the statistical analysis of the data.



     Environmentally collected sediments display  a wide range of



toxicities.   If sediment is extremely toxic, then it may be



necessary to use a highly pollution-tolerant species (e.g.



Capitella spp.).  As a rule, however, these pollution-tolerant



species are small.  Most field collected sediments will be



moderately toxicity,  allowing the use of the moderately





                                189

-------
pollution-tolerant species though not the sensitive species



commonly used in sediment toxicity tests (e.g., phoxocephalid



amphipods).   A general indication of the relative hardiness of



various species can be obtained from the lists of opportunistic



species listed in Pearson and Rosenberg (1978), the species used



to calculate the "Infaunal Trophic Index" (Word, 1978),  as well



as from multispecies bioassays (Swartz et al., 1979)





C. DESIRABLE CHARACTERISTICS:





     Besides the required criteria, there are a number of



desirable characteristics which would make conducting the tests



easier, interpreting the results more straightforward, or allow



the results to be applied to a wider range of habitats.



     On important characteristic, especially  if repeated tests



are planned, is the ease of obtaining the test species in



sufficient numbers at the correct season.  The ease of collecting



specimens is related to a species' abundance,  habitat (intertidal



vs subtidal vs offshore), robustness to collection techniques,



depth in the sediment, and seasonality.  It is important not to



underestimate the time required to collect sufficient numbers of



healthy individuals.  In general, it is prudent to collect twice



the number required, especially polychaetes which are prone to



breakage.  Information on collecting and transporting specimens



is given in Chapter VIII.  As an alternative,  test organisms may



be purchased from biological supply houses or local collectors



(See Appendix VIII).   Local bait suppliers may sell species such
                                 190

-------
as Nereis and Callianassa.



     If a large number of bioaccumulation tests will be conducted



over an extended time period, culturing of test organisms may be



cost effective.  Culturing will provide a ready supply of



organisms of known history, but maintaining a culture to supply



sufficient biomass for bioaccumulation tests will be time



consuming.  A few sediment ingesting polychaetes  (e.g., Capitella



capitata and Neanthes arenaceodentata) can be cultured with



relatively simple equipment  (Reish and Richards, 1966; Reish,



1974, 1985; Dean and Mazurkiewicz, 1975), as can Palaemonetes



(Tyler-Schroeder, 1976a,b).  Although these organisms are



generally suitable test species, most of the species are small,



making it difficult to obtain sufficient biomass.  Culture of



bivalves, larger polychaetes, and most crustaceans is impractical



except for experimental studies.



     Regardless of how test species are obtained, they should be



amenable to laboratory conditions and not require elaborate



holding facilities.  Fortunately, most pollutant-resistant



species are relatively hardy and adaptable to laboratory



conditions.  Most of the bioaccumulation test species listed in



Table VII-1 are reasonably easy to maintain and do not require



flowing seawater.



     Whether field-collected or laboratory-cultured specimens are



used, gravid individuals or individuals which are likely to



become gravid during a test should be avoided if possible.   The



reduction in tissue lipids that often occurs with spawning
                                 191

-------
(Gabbot, 1976: Davis and Wilson, 1983) can result in a



corresponding reduction in the associated pollutants.  Spawning



may also result in unacceptable mortalities.  Certain species,



such as Macoma nasuta in Oregon, have a reasonably well defined



spawning cycle and size at reproductive maturity, making it



possible to minimize the collection of reproductive individuals.



Other species, such an Neanthes virens. change appearance when



reproductively mature.  In extended tests, it may be impossible



to completely avoid gravid individuals, though the occurrence of



the reproductive state should be noted.



     A very important characteristic is organism size.  Test



species need to be small enough to be easily maintained, yet



large enough to supply sufficient biomass for chemical analysis.



The amount of biomass required depends upon the analytical



procedures used as well as the types of analyses required  (e.g.,



metals, organics, lipids).  At least 1 gram of wet tissue is



required in nearly all cases, and commonly up to 5 grams tissue



will be required.  Ideally, the species should be large enough to



allow chemical analysis on individuals.  Chemical analyses on



individuals are possible using the more "sophisticated"



analytical procedures with the larger benthic organisms, such as



Macoma nasuta and lug worms.  Depending on the techniques, it may



be impossible to conduct both metals and organic analyses on an



individual, even when using large species, thereby necessitating



twice as many exposure chambers if both types of pollutants are



required.
                                 192

-------
     An alternative approach to obtaining sufficient biomass is



to composite individuals  (see Chapter III).  Even when



compositing individuals,  the size of the individual is an



important consideration.  It is simpler to handle and count a few



larger individuals  (e.g., Nereis) than dozens or even hundreds of



smaller specimens  (e.g.,  Capitella).



     Species suitable for measures of sublethal stress or



experimental manipulations offer the advantage that toxicokinetic



or toxicological data can be collected concurrently with the



bioaccumulation data.  Growth is the simplest measure of



sublethal stress.  Measuring changes in wet weight is possible



with both polychaetes and bivalves, but wet weight measurements



are prone to error.  Johns et al.  (1989) gives techniques for



conducting a growth bioassay with Neanthes arenaceodentata. some



of which could be adapted to measurement of growth during



bioaccumulation tests.  With bivalves, growth can also be



measured as changes in shell length.  Although shell length has



the limitation of only showing positive growth, the comparison of



shell growth in treatments to the controls is a simple sublethal



measure.   Scope-for-growth,  a measure of the amount of energy



available for growth, has been used frequently as a measure of



stress in mussels  (Bayne  et al.,  1981).   Scope-for-growth



measurements have been conducted on infaunal polychaetes (Johns,



et al.,  1985)  and presumably could be adapted to other bivalves



and larger crustaceans,  though the techniques are not simple.



The arenicolid worms and  the bivalves, particularly Macoma
                                 193

-------
nasuta. lend themselves to exposure chambers in which it is



possible to collect ventilated water and processed sediment



(i.e., feces) for bioenergetic measurements and for



determinations of the pollutant uptake rates (e.g., Pelletier et



al., 1988a,b; Specht and Lee, 1989).  Construction of these



exposure chambers is discussed in Appendix IX-1.



     The more tolerant a species is to sediment, temperature, and



salinity variations, the more types of sediments in which it can



be used.  Using a few widely adaptable species allows a direct



comparison of sediment bioavailability from a variety of



environments or biogeographic regions.  Also, collecting and



maintaining a few widely adaptable species is simpler than



developing techniques for a larger number of less adaptable



species.   The approximate salinity and temperate ranges of



potential bioaccumulation species are given in Table VII-1.



These ranges are estimates of the ranges in which the organisms



could be used in bioaccumulation test and are not the



physiological limits.  For many of the species, the ranges are



based on the general literature and discussions with other



researchers rather than extensive experimentation.  A preliminary



survival test would be advisable before initiating a large



bioaccumulation test using species near the limits of the ranges



given in Table VII-1.



     Because the goal of bioaccumulation tests is to estimate the



maximum likely tissue residues, it is important to chose species
                                 194

-------
which have a high bioaccumulation potential.  Unfortunately,



insufficient numbers of multi-species tests have been conducted



to adequately compare the bioaccumulation potential of a range of



species over a range of compounds.  In general, tissue residues



will be higher in species with a higher lipid contents, which can



vary as much as 10-fold among species (e.g., Rubinstein et al.,



1987).   If PAHs are of concern, at least one test species should



be a bivalve as they have a lower ability to metabolize PAHs than



either polychaetes or crustaceans (Varanasi et al., 1985).



     Infaunal species are preferable over epibenthic deposit-



feeders because the latter are only intermittently exposed to



interstitial water.  Interstitial water is thought to be the



major uptake route for compounds with a Kow less than about 5



(Adams, 1987) and possibly for metals as well  (Assanullah et al.,



1984),  so the potential uptake of these compounds could be



underestimated.  This criterion limits the use of Palaemonetes.



the only well established crustacean on the list of



bioaccumulation species  (Table VII-1).  Infaunal crustaceans



(e.g.,  Callianassa) have not been used extensively as



bioaccumulation species, and appear to be more difficult to



maintain in the laboratory than Palaemonetes.



     Compatibility with other species or with the same species is



important if multiple species or multiple individuals of the same



species are exposed in the same chamber.  Several of the nereid



worms are aggressive to members of the same sex  (Reish and Alosi,



1968; Johns et al., 1989).  Some nereids also prey on smaller



species and Palaemonetes may crop the siphons of bivalves.





                                 195

-------
D. RECOMMENDED AND SECONDARY SPECIES:



     Based on these various criteria, we have identified  five



recommended bioaccumulation test species and another eight



"secondary" test species  (Table VII-l). The "recommended" species



meet all or nearly all of the desired criteria and are well



established bioaccumulation test species in both regulatory and



experimental studies.  The recommended species are the



polychaetes Nereis diversicolor and Neanthes virens.  and the



bivalves Macoma nasuta. Macoma balthica. and Yoldia limatula.



Within their tolerance levels, these species should serve as



suitable test species, and we recommend using at least one of



these species in all tests, at least until the suitability of



other species has been demonstrated locally.



     The secondary bioaccumulation species meet the required



characteristics but either are deficient in one or more of the



important desired characteristics and/or there is insufficient



information to make a final evaluation.   Some of the secondary



species offer potential advantages such as including additional



phylogenetic groups  (i.e., crustaceans), adaptability to



culturing (e.g. Neanthes arenaceodentata).  and high pollution



tolerance (Capitella spp).  The importance of these various



characteristics will depend upon the site specific situation



(e.g.,  level of toxicity of sediment).  We recognize that the



list of secondary bioaccumulation species is not exhaustive,  and



there may be other suitable test species.
                                 196

-------
     This list of recommended and secondary test species includes



mostly estuarine species, reflecting both the hardiness and



relative ease of collecting estuarine species.  As many of these



species extend into shallow marine habitats, they should serve as



suitable surrogates species for near-coastal environments.  There



are no open ocean or brackish water  (<10 ppt salinity) species



included, and the species listed generally do not extend into



these habitats.  However, there does not seem to be a priori



reason why estuarine species should underestimate the tissue



residues of off-shore or brackish water species, though this



assumption should be tested.  Discussion of fishes or



megainvertebrates is beyond the scope of this work, but a list of



recommended species for field monitoring of sewage discharges can



be found in Tetra Tech (1985a).
                                 197

-------
           APPENDIX VIII-1: SOURCES FOR TEST ORGANISMS
                     BIOLOGICAL  SUPPLY HOUSES
     COMPANY
  PERTINENT SPECIES
Gulf Specimen Company
PO Box 237
Panacea, Florida 32346
904-984-5297

Pacific Biological Supply
P.O. 536
Venice, CA 90291
213-822-5757

Sea Life Supply
740 Tioga Ave.
Sand City, CA 93955
408-394-0828
Wood Hole Marine Biological Laboratory
Marine Resources Dept.
Woods Hole, MA 02543
508-548-3705
Nereis. Palaemonetes
Uca. Callinectes
Palaemonetes.  Capitella.
Uca. Callianassa
                          OTHER SOURCES
Don Reish
Dept of Biology
California State U.
Long Beach, CA 90840
213-985-4845

Maine Bait Co.
Newcastle, Maine 04553
207-563-3000

Local Bait suppliers
University Biology or
Marine Sciences Depts.
Neanthes arenaceodentata
Capitella sp.
Nereis virens
nereid worms, Callianassa.
clams,

wide variety
Note: Inquire as to the availability of specific species well
ahead of the time the organisms are required.
                                 198

-------
     APPENDIX IX-1: SPECIAL PURPOSE EXPOSURE CHAMBERS





A. CLAMBOX



     This exposure chamber is designed to separate the inhalant



and exhalant siphons of sediment-ingesting clams having



independent siphons (see Figure IX-1A).  The technique is



applicable for Macoma spp. and other tellinids, though in most



bivalves the two siphons are fused together to form the "neck".



The apparatus allows the isolation and collection of the feces



from the parent sediment and ventilated  (pumped) water from the



input supply.  This allows a direct measure of short- and long-



term ventilation and sediment processing rates  (the Fx terms of



Equation 5, Appendix 1-1)  (Specht and Lee, 1989).  By analyzing



the pollutant content in the feces or the ventilated water, the



amount of pollutant extracted by the clam  (the EPx term' of



Equation 5, Appendix 1-1) can be estimated.  The chamber has been



used to determine the efficiency of uptake of dissolved



hexachlorobenzene  (HCB) by the gills  (Boese et al., 1988), the



efficiency of HCB uptake through the gut from ingested sediment



(Lee et al., in press), uptake from ventilated interstitial water



(Winsor, et al., in press), and the passive sorption of HCB to



the soft-tissues (Lee et al., 1988).





B. WORMTUBES



     These exposure chambers are tubes open on each end, which



simulate the burrow of sediment-ingesting polychaete worms such



as Abarenicola pacifica and Arenicola marina.  The worms pump





                                 199

-------
                               FIGURE DC-1A
                       Clambox Exposure Chamber
                                                  Dental dam membrane
o
o
                           Inhalant siphon
                           Exhalant siphon
                        Contaminated sedlmenft
                            Fecal pellets
            100 ml Beaker
Overflow
standpipe

-------
water and sediment in one direction through the tubes  (Figure IX-



1B).   As with the clamboxes, the feces can be collected and



separated from the parent sediment, allowing the measurement the



sediment processing rate and the collection of the feces for



chemical analysis.  These systems have been used to study the



effects of crude oil on sediment processing rates  (Augenfeld,



1980) and on the uptake rate of cadmium as a function of the



addition of sewage carbon to sediment  (Pelletier et al., 1988b).



Some versions also allow the simultaneous measurement of



ventilation rate and oxygen consumption (Kristensen, E., 1981;



Pelletier et al., 1988a).





C. SEDIMENT RESUSPENSION SYSTEMS.



     This flow-through device automatically maintains a constant



suspended sediment load in the water column, using an electro-



optical feedback mechanism  (U.S. EPA, 1978) which employs an



airlift dosing system, a transmissometer to measure particle



concentration, and a microcomputer which calculates the dose



required to achieve a programmed turbidity  (Sinnett and Davis,



1983; Lake et al., 1985; Pruell et al., 1986).  This system has



been used in several studies on the uptake and effects of



pollutants from resuspended sediments using the mussel, Mytilus



edulis. and the infaunal polychaete Nephtys incisa (Lake et al.,



1985; Nelson et al., 1987; Yevich et al.,  1987).  Other systems
                                 201

-------
for maintaining suspended sediments are given in Rubinstein et



al. (1980) and Peddicord  (1980).  These chambers should be used



when there is concern about bioaccumulation in obligate benthic



filter-feeders (e.g., Mercenaria. Mya. Mytilus).  or facultative



filter-feeders (e.g., Macoma) via resuspended sediments.  This



mode of exposure is important in areas where current or wave



action periodically resuspend sediments and in areas with a



flocculent surface layer.
                                 202

-------
                 FIGURE IX-IB
          Wormtube Exposure Systems
      w
      a

      b
      incurrent

a.
b.
c.
Worm in sediment, 1 L glass box
Worm in glass tube, 30 L aquarium
Expanded view of ventilated water collection
and monitoring device
                      203

-------
       APPENDIX X-l:  ADDITIONAL TECHNIQUES FOR CORRECTING
                         FOR GUT SEDIMENT
A. MODIFICATIONS TO 24-HOUR PURGE AND DISSECTION

     There are a number of other techniques or modifications to

the standard 24-hour purge in control sediment  (Chapter X) which

should be considered in specific cases.  When it is unclear

whether a species is voiding all its gut contents within 24

hours, a marker "sediment" can be added to the control sediment

during the purging.  Marker sediments are inert particles of a

contrasting color or phosphorescence under UV radiation added to

the control sediment.  Observation of feces composed of these

marker sediments is indicator that the gut has been voided.

Techniques for marking sediments for use as tracers are given in

Ingle  (1966).  In cases when it is critical not to have any

sediment in the gut, such as in certain studies of metals, it may

be necessary to purge the organisms in clean water without

sediment.  Before using this approach, it is necessary to

determine whether the test species will satisfactorily void its

gut in the absence of sediment.

     Another approach is to remove the gut sediment by

dissection.  Dissection avoids the problems with the loss of

tissue pollutants during the purging, but is limited to the

larger test species  (e.g., Abarenicola).  Care has to be taken to

minimize loss of body fluids and to avoid contamination,

especially with the metals.  General instructions for minimizing

contamination are available in Lauenstein and Young  (1986).


                                 204

-------
B. CALCULATING POLLUTANT MASS OF GUT SEDIMENT

     It is possible to calculate the mass of pollutant associated

with the gut sediment if both the mass and the pollutant

concentration of gut sediment can be estimated.  For selective

deposit-feeders, the pollutant concentration of the ingested

sediment may be several fold greater than the concentration of

the bulk sediment (see Lee et al., in press), so the bulk

sediment concentration should not be used as an estimate of the

gut sediment.  Instead, the gut concentrations can be estimated

either from the pollutant concentrations of the ingested sediment

or the feces.  Using the fecal pellet concentrations as the input

parameter, the whole body tissue residue (Ctw, including both the

tissue and gut sediment pollutants)  can be expressed as:
     Ctw =
(Mg*CPSf)  +  (Mt*Ct)

   Mg  + Mt
     Expressed on a tissue residue only basis (i.e., no gut

sediment),  the formula becomes:

            Ctw*(Ms + Mt)  - (CPSf*M )
     ct = 	         (2)

                      Mt

 Where:

     Ctw =  whole body tissue concentration (tissue and
               gut sediment)  (ug/g)

     Mg = mass of gut sediment (g)

     CPSf = pollutant cone, in feces (ug/g)

     Mt = mass of tissue (g)

     Ct = tissue concentration without gut sediment  (ug/g)
                                205

-------
     If the ingested pollutant concentration (CPSi)  is used, the



formula is the same except that CPSi is substituted for CPSf.



Use of fecal pellet pollutant concentration underestimates the



average gut pollutant content because some of the pollutants are



extracted from the sediment before defecation.   Conversely,



ingested sediment overestimates the average gut pollutant content



because some of the pollutants have been extracted.   These errors



are not expected to be large, but both methods could be



calculated and the results averaged for the most accurate



estimate.  Fecal pellets can be collected for chemical analysis



by using special exposure chambers such as the clambox with



Macoma or wormtubes with polychaetes (see Appendix IX-1).



A method to estimate ingested dose is given in Lee et al.  (in



press).





C. USE OF CONSERVATIVE TRACE ELEMENTS



     Another approach to correcting for gut sediment is to use



the concentration of a conservative, non-biologically active



element as a means to determine sediment mass in the gut



(Kennedy, 1986).  Knowing the sediment pollutant concentration,



it is then theoretically possible to calculate the amount of



pollutant associated with gut sediment.  Some of the conservative



elements common in minerals but not typically found in more than



trace amounts in tissues include silicon, aluminum,  and iron



(Kennedy, 1986).  The difficulty with this approach is that the



elemental content of gut sediment in selective deposit-feeders
                                 206

-------
may differ from that of the bulk sediment, especially if the



organism selectively ingests organic rather than mineral



particles.  Additionally, this method will underestimate the gut



pollutant mass unless the pollutant concentration of the ingested



sediment  (CPSi) is used rather than the bulk sediment



concentration.
                                 207

-------
                             GLOSSARY


Accumulation Factor  fAF) -  Ratio of lipid normalized tissue
     residue to carbon normalized sediment pollutant
     concentration.   (Appendix I-I).

ACE - Army Corps of Engineers.

Alpha - see Type I error.

Apparent Steady-State -  See Steady-State.

ASTM - American Society for Testing and Materials.

BAF - See Bioaccumulation Factor.

BaP - Benzo(a)pyrene

BCF - See Bioconcentration Factor.

Bedded Sediment - Consolidated sediment (i.e., not suspended).

Beta - see Type II Error.

Bioaccumulation - Uptake from all phases,  including water, food
     and sediment.

Bioaccumulation Factor - Ratio of tissue residue to sediment
     pollutant concentration.  (Appendix 1-1)

Bioaccumulation Potential - Qualitative assessment of whether a
     pollutant in a particular sediment is bioavailable.
     (Appendix IV-1)

Bioconcentration - Uptake from water.

Bioconcentration Factor (BCF) - Ratio of tissue residue to
     to water pollutant concentration.   (Appendix 1-1)

Block - Group of homogeneous experimental  units.  (Chapter III)

Coefficient of Variation -  A standardized  variance term; the
     standard deviation divided by the mean and expressed as a
     percent.  (Chapter III)

Comparison-wise Error - Type I error applied to a single
     comparison of two means.  Contrast with Experiment-wise
     error.  (Chapter XII)
                                208

-------
Compositing - The combining of separate tissue or sediment
     samples into a single sample.   (Chapter III)

Control Sediment - Sediment with very low pollutant concentrations
     which is compared with reference and/or test sediments.
     (Chapter II)

Control Treatment - Treatment  (i.e., sediment type) that is
     chosen to give a baseline value for comparison with results
     from test treatments.  May consist of either control or
     reference sediments.  (Chapter III)

Degradation - As used in the manual, it refers to the metabolic
     breakdown of parent pollutant by a test species.  Along
     with depuration, it is one of the processes by which
     pollutants are removed from an organism.

Depuration - Loss of the parent pollutant from an organism. See
     Degradation.

DDT - Common environmental pollutant.  Metabolites include DDD
     and DDE.

DOC - Dissolved organic carbon.   (Chapter VI)

DOM - Dissolved organic matter.   (Chapter VI)

Eh - Redox potential, which is a measure the oxidation state of
     a sediment.  (Chapter VI)

EPA - Environmental Protection Agency

Experiment-wise Error -  Type I error (alpha) chosen such that
     the probability of making any Type I error in a series of
     tests is alpha.  Contrast with Comparison-wise error.
     (Chapter XII)

Experimental error - Variation among experimental units given
     the same treatment.  (Chapter III and XII)

Experimental unit - Organism or organisms to which one trial of a
     single treatment is applied.   (Chapter III)

FDA - Food and Drug Administration.

Fines - Silt-clay fraction of a sediment.  (Chapter VI)

Gut Purging - Voiding of sediment contained in the gut.  (Chapter X
     and Appendix X-l)
                                 209

-------
Ha - The alternate hypothesis.   (Chapter III)

Hg. - The null hypothesis.   (Chapter III)

Hydrophobic pollutants - Low water solubility pollutants with a
     high Row, and usually a strong tendency to bioaccumulate.

Interstitial water - Water between the particles  (i.e., interstices)
     in sediment.  (Chapter VI)

kl - Uptake rate constant.  (Chapter IV)

k2 - Depuration rate constant.   (Chapter IV)

Kinetic Bioaccumulation Model  - Any model that uses uptake
     and/or depuration rates to predict tissue residues.  In
     this manual, it refers to the linear uptake, first-order
     depuration model.   (Chapter IV)

Koc - Organic-carbon partitioning coefficient.

Row - Octanol-water partitioning coefficient.

Long-Term Uptake Tests - Bioaccumulation tests with an exposure
     period greater than 28 days.   (Chapter IV)

LOI - Loss on ignition.  (Chapter VI)

Multiple Comparisons - Statistical comparison of several
     treatments simultaneously such as with ANOVA.  (Chapter XIII)

Metabolism - see Degradation.

Minimum Detectable Difference  - The smallest (absolute)
     difference between two means that is statistically
     distinguishable.  (Chapter III)

NOAA - National Oceanic and Atmospheric Administration.

No Further Degradation - Approach by which a tissue residue is
     deemed acceptable if it is not greater than those at a
     reference site.   (Chapter XIII)

PAH - Polyaromatic hydrocarbons.

Pairwise Comparisons -  Statistical comparison of two
     treatments.  Contrast with multiple comparisons
     (Chapter XII)

PCS - Polychlorinated biphenyls.  Consists of over 200 congeners.
                                 210

-------
Power -  Probability of detecting a difference between the
     treatment and control means when a true difference
     exists.  (Chapter III)

Pseudoreplication - Incorrect assignment of replicates, often due
     to biased assignment of replicates.   (Chapter III)

Reference sediment - Sediment used as an indicator of background
     pollutant levels and resulting tissue residues.  May have
     moderate levels of pollutants.  (Chapter II)

Replication - Assignment of a treatment to more than one
     experimental unit.   (Chapter III)

Sampling unit - The fraction of the experimental unit that is
     to be used to measure the treatment effect.  (Chapter III)

Spiking - Experimental addition of pollutants to a sediment.
     (Chapter V and Appendix V-l)

Standard Reference Sediment - Standardized sediment and pollutant
     used to determine the variability due to variation in the
     test organisms.   (Chapter II)

Steady State - A "constant" tissue residue as determined
     by no statistical difference in three sampling periods
     (Chapter IV)

TC - Total carbon, including organic and inorganic carbon
     (Chapter VI)

Test Sediment - The sediment or dredge material of concern.  This
     is the sediment on which the regulatory decision will be
     made.  Contrast with Test Treatment.   (Chapter II)

Test Treatment - Treatment that is compared to the control
     treatment.  It may consist of a test  sediment  (compared to a
     reference or control sediment) or a reference sediment
     (compared to the  control sediment).   (Chapter III)

Thermodynamic Partitioning Bioaccumulation Model - Bioaccumulation
     model based on pollutant equilibrium  partitioning among
     lipids and sediment carbon.   (Appendix 1-1)

Tissue residues - Pollutant concentration  in the tissues.

TOC - Total organic carbon.   (Chapter VI)
                                 211

-------
Toxicokinetic Bioaccumulation Model - Bioaccumulation model based
     on the feeding and ventilatory fluxes of the organism.
     (Appendix 1-1)

Treatment - The procedure (type of sediment) whose effect is to
     be measured.  (Chapter III)

TVS - Total volatile solids.   (Chapter VI)

Type I Error - Rate at which Ho is rejected falsely.   (Chapter III)

Type II Error - Rate at which Ho is accepted falsely.   (Chapter III)
                                 212

-------
                           BIBLIOGRAPHY
ACGIH. 1987.  Threshold limit values and biological exposure
    indices  for  1987-1988. American Conference  of  Governmental
    Industrial Hygienists. Cincinnati,  OH.

Adams, D.D., D.A. Darby, and R.J. Young. 1980.   Selected
    analytical techniques  for characterizing the metal  chemistry
    and geology  of  fine-grained sediments and interstitial  water.
    In R.A.  Baker  [ed.] Contaminants  and Sediments,  Vol.  2. Ann
    Arbor  Science,  Ann Arbor,  MI.  pp.3-28.

Adams, W.J., R.A. Kimerle, and R.G. Mosher. 1985.   Aquatic safety
    assessment of chemicals  sorbed to  sediments. In R.D.
    Cardwell, R.  Purdy,  and  R.C.  Bahner,  [eds.]. Aquatic
    Toxicology and  Hazard  Assessment:  Seventh Symposium,  ASTM STP
    854. ASTM, Philadelphia,  PA.  pp.429-453.

Adams, W.J. 1987.  Bioavailability of neutral lipophilic organic
    chemicals contained on sediments.  In K.L. Dickson,  A.W. Maki,
    and W.A.  Brungs [eds.].  Fate and Effects of Sediment-Bound
    Chemicals in Aquatic Systems.  Proceedings of the 6thtPellston
    Workshop, Florissant,  Colorado,  August  12-17,  1984.   Pergamon
    Press,  New York,  NY. pp.219-244.

Ahsanullah, M.,  M.  Mobley, and D. Negilski. 1984.  Accumulation of
    cadmium from contaminated water and sediment by the shrimp
    Callianassa  australiensis.  Mar.  Biol. 82:191-197.

American Public Health Association. 1985.   Standard Methods For
    the Examination of Water and Wastewater:  16th  Edition.  M.A.
    Franson [ed.] American Public Health Association, pp.  92-100.

ASTM. 1980. E 729-80.  Standard practice for conducting acute
    toxicity tests  with fishes,  macroinvertebrates,  and
    amphibians.  ASTM,  Philadelphia,  PA.

ASTM. 1984. E 1022-84. Standard practice for conducting biocon-
    centration tests with  fishes and saltwater  bivalve  molluscs.
    ASTM,  Philadelphia,  PA.

ASTM. 1988a. E1367.  Proposed guide for conducting solid-phase
    10-day static sediment toxicity tests with  marine and
    estuarine infaunal amphipods.  ASTM,  Philadelphia, PA.

ASTM. 1988b. E1391.  Proposed guide for sediment collection,
    storage,  characterization,  and manipulation.   ASTM,
    Philadelphia, PA.
                               213

-------
Augenfeld, J.M. 1980.  Effects of Prudhoe Bay crude oil
    contamination  on  sediment  working  rates  of Abarenicola
    pacifica. Mar.  Environ.  Res.  3:307-313.

Augenfeld, J.M., J.W. Anderson, R.G. Riley,  and B.L. Thomas.
    1982.  The fate of polyaromatic hydrocarbons  in  an  intertidal
    sediment exposure system:   Bioavailability to Macoma
    inquinata  (Mollusca: pelecypoda) and Abarenicola pacifica
    (Annelida:  polychaeta).  Mar.  Environ.  Res. 7:31-50.

Bascom, W. 1984.  The concept of assimilative capacity. In
    Willard Bascom [ed.] Southern California Coastal Water
    Research Project  Biennial  Report,  1983-1984. Long  Beach,  CA.
    pp.171-177.

Batley, G.E.  and M.S. Giles. 1980.   A solvent displacement
    technique for  the separation  of  sediment interstitial
    waters. In  R.A. Baker  [ed.] Contaminants and Sediments,   Vol.
    2.  Ann Arbor Science, Ann  Arbor, MI. pp.101-117.

Bayne,  B.L.,  D.A.  Brown.,  K. Burns., D.R.  Dixon, A. Ivanovici, D.
    R.  Livingston,  D.M. Lowe,  M.N. Moore., A.R.D Stebbing,  and J.
    Widdows. 1981.  The Effects of Stress  and Pollution on Marine
    Animals. Praeger  Scientific,  New York, NY. 384pp.

Berg, C.J. Jr.   [ed.]. 1983.  Culture of Marine Invertebrates.
    Selected Readings. Hutchinson Ross Publ.  Co., Stroudsburg,
    PA.  367pp.

Beers,  Y.  1957. Introduction to the Theory of Error. Addison-
    Wesley Publishing Co. Inc.  Reading, MA.  26pp.

Bligh,  E.G.  and W.J.  Dyer.  1959.   A rapid method of total lipid
    extraction  and purification.  Can.  J. Biochem. Physiol.
    37:911-917.

Boese,  B.L.,  H. Lee II,  and D.T.  Specht.  1988.   The efficiency of
    uptake of hexachlorobenzene from water by the tellinid clam
   Macoma nasuta.  Aquat.  Toxicol. 12:345-356.

Boese,  B.L.,  H. Lee II,  D.T. Specht, R.C.  Randall,  and M. Winsor.
    1990.  Comparison of aqueous  and solid-phase uptake for hexa-
    chlorobenzene  in  the tellinid clam.Macoma nasuta (Conrad):  A
    mass balance approach.  Environ.  Toxicol.  Chem.  9:221-232.

Branson, D.R.,  G.E. Blau,  B.C. Alexander,  and W.B.  Neely. 1975.
    Bioconcentration  of 2,2',4,4'-tetrachlororbiphenyl in  rainbow
    trout  as measured by an accelerated test. Trans. Am.  Fish.
    Soc. 4:785-792.
                               214

-------
Breteler, R.J. and F.I. Saksa. 1985.  The role of sediment
    organic matter  on  sorption-desorption reactions  and bioavail-
    ability of mercury and  cadmium in an intertidal  ecosystem.  In
    R.D.  Cardwell,  R.  Purdy,  and  R.C.  Bahner [eds.],  Aquatic
    Toxicology and  Hazard Assessment:  Seventh Symposium,  ASTM STP
    854.  ASTM, Philadelphia,  PA.  pp.454-468.


Brown, D.A.,  R.W. Gossett,  G.P. Hershelman, C.F. Ward, and J.N.
    Cross. 1984.  Metal and  organic contaminants  in sediments and
    animals.  In Willard Bascom [ed.]  Southern California Coastal
    Water Research  Project  Biennial Report,  1983-1984.  Long
    Beach, CA. pp.179-193.

Bryan, G.W.  1976.  Some aspects of heavy metal tolerance in
    aquatic organisms.  In A.P.M.  Lockwood [ed.]  Effects Of
    Pollutants On Aquatic Organisms.  Cambridge Univ.  Press.
    New York,  NY. pp.7-34.

Buchanan, J.B. 1984.  Sediment analysis. In N.A. Holme and A.D.
    Mclntyre  [eds.]  Methods for the Study of Marine  Benthos,  IBP
    Handbook  #16. Blackwell Scientific Publications,  Boston,  MA.
    pp.41-64.

Buchanan, J.B. and M.R. Longbottom. 1970. The determination of
    organic matter  in  marine muds:  The effect of the presence of
    coal  and  the  routine determination of protein. J.  Exp. Mar.
    Biol. Ecol. 5:158-169.

Byers, D.F.,  S.C. Mills, E.L. Mills, and P.L. Stewart. 1978.  A
    comparison of methods of determining organic carbon in marine
    sediments, with suggestions for a standard method.
    Hydrobiologia 58:43-47.

Bysshe, S.E.  1982.   Bioconcentration factor in aquatic organisms.
    In W.J. Lyman,  W.F.  Reehl,  and D.H.  Rosenblatt [eds.]
    Handbook  of Chemical Property Estimation Methods.  Environ-
    mental Behavior of Organic Compounds.  McGraw-Hill,  New York,
    NY. pp.5-1 -  5-30.

Clayton Jr.,  J.R.,  S.P. Pavlou, and N.F. Breitner. 1977.
    Polychlorinated biphenyls in  coastal marine  zooplankton:
    Bioaccumulation by equilibrium partitioning.  Environ. Sci.
    Technol.  11:676-681.

Cross, F.A.,  L.H. Hardy, N.Y. Jones, and R.T. Barber. 1973.
    Relation  between total  body weight and concentrations of
    manganese, iron, copper,  zinc,  and mercury in white muscle  of
    bluefish  (Pomatomus saltatrix)  and a bathyl-demersal fish
    Antimora  rostrata.  J. Fish. Res.  Board Can.  30:1287-1291.
                               215

-------
D'Agostino, R.B. and M.A. Stephens. 1986. Goodness-of-fit
    Techniques.  Marcel  Dekker,  Inc.  New York,  NY.  560  pp.

Dalela, R.C., S.K. Bansal, A.K. Gupta, and S.R. Verma. 1979.
    Effect  of solvents  on in vitro pesticides  inhibition of
    ATPase  in certain tissues of  Labeo rohita.  Water,  Air,  and
    Soil  Poll.  11:201-205.

Daniel, W.W. 1978.  Applied Nonparametric Statistics. Houghton
    Mifflin Co.  Boston,  MA.  503pp.

Davies, R.P. and A.J. Dobbs. 1984.  The prediction of
    bioconcentration in fish.  Wat.  Res.  18:1253-1262.

Davis, J.P. and J.G. Wilson. 1983.  Seasonal changes in tissue
    weight  and biochemical composition of the  bivalve  Nucula
    turgida in Dublin Bay with reference to gametogenesis.  Neth.
    J.  Sea  Res.  17:84-95.

Davis, M. and H. Lee II. 1983.  Recolonization of sediment-
    associated microalgae and effects  of estuarine infauna  on
    microalgal production.  Mar. Ecol.  Prog.  Ser.  11:227-232.

de Boer, J. 1988.  Chlorobiphenyls  in bound and non-bound  lipids
    of fishes:  Comparison of different extraction methods.
    Chemosphere 17:1803-1810.

Dean, D. and M. Mazurkiewicz. 1975. Methods of culturing poly-
    chaetes. In W.L.  Smith and M.H.  Chanley [eds.]  Culture  of
    Marine  Invertebrate Animals.  Plenum Press,  New York,  NY. pp.
    177-197.

Dean, W.E.  1974.  Determination of  carbonate and organic matter
    in calcareous sediments and sedimentary rocks by loss on
    ignition: Comparison with other methods. J.  Sed.  Pet.
    4:242-248.

DeWitt, T.H.  1988.  Adult vs larval colonization: a comparison of
    recruitment rates in soft-bottomed benthic communities.
    Presented at AAAS Annual Meeting,  Boston,  MA.  Feb.  1988.

DeWitt, T.H., G.R. Ditsworth, andR.C. Swartz. 1988. Effects  of
    natural sediment features on  survival of the phoxocephalid
    amphipod, Rhepoxynius abronius.  Mar.  Environ.  Res.  25:99-124.

DiToro, D.M., J.D. Mahony, D.J. Hansen,  K.J. Scott, M.B. Hicks,
    M.M.  Mayr, and M.S. Redmond.   1990.   Toxicity of  cadmium in
    sediments: The role of acid volatile sulfides.  Environ.
    Toxicol. Chem. 19:1487-1502
                               216

-------
Edmunds, W.M. and A.H. Bath. 1976.  Centrifuge extraction and
    chemical  analysis  of  interstitial  waters.  Environ.  Sci.
    Technol.  10:467-472.

Elliott, J.M. 1983.  Some methods for the statistical analysis of
    samples of benthic invertebrates.  Freshwater Biological
    Assoc. Scientific  Pub.  25.  2nd Edition.  156pp.

el Wakeel, S.K.  and J.P.  Riley. 1956.  The determination of
    organic carbon  in  marine muds.  J.  Con.  Internat.  Explor.  Mer.
    22:180-183.

Engler, R.M.  1980.   Prediction of pollution potential through
    geochemical  and biological  procedures:  Development  of
    regulation guidelines and criteria for  the discharge of
    dredged and  fill material.  In  R.A.  Baker [ed.]  Contaminants
    and Sediments,  Vol. 1.  Ann  Arbor Science,  Ann Arbor, MI.  pp.
    143-169.

Fenchel, T.  1969.  The ecology of marine microbenthos.  IV.
    Structure and function  of the  benthic ecosystem,  its chemical
    and physical factors  and the micro-fauna communities with
    special reference  to  ciliated  Protozoa.  Ophelia.  6:1-182.

Ferraro, S.P.,  H. Lee II,  R.J. Ozretich,  and D.T.  Specht. 1990.
    Predicting bioaccumulation  potential: A test  of a fugacity
    based model.  Arch.  Environ. Contamin. Toxicol.  19:  386-394.

Folch, J.,  M. Lees, and G.A. Sloane-Stanley. 1957.   A simple
    method for the  isolation and purification  of  total  lipids
    from animal  tissues.  J.  Biol.  Chem.  226:497-509.

Folk, Robert  L.  1980.   Petrology of Sedimentary Rocks.  Hemphill
    Publishing Co.  Austin,  TX.  184pp.

Foster, G.,  S.M. Baksi, and J.C.  Means. 1987.  Bioaccumulation of
    trace organic contaminants  from sediment Baltic clams  (Macoma
    balthica) and soft-shell (Mya  arenaria). Environ. Toxicol.
    Chem. 6:969-976.

Fox, K.R.  and K.R.  Rao. 1982.  Accumulation, tissue distribution,
    and depuration  of  Benzo-a-pyrene and Benzo-a-anthracene  in
    the grass shrimp,  Palaemonetes  pugio. In N.L. Richards and
    B.I.L. Jackson  [eds.],  Symposium:  Carcinogenic  polynuclear
    aromatic  hydrocarbons in the marine  environment.  Pensacola
    Beach, FL.,  August  14-18, 1978.  EPA  600/9-82/013.

Frankenburg,  D.  and A.S.  Leiper.  1977.  Seasonal cycles in
    benthic communities of  the  Georgia continental  shelf.  In  B.
    Coull  [ed.]  Ecology of  Marine  Benthos. University of South
    Caroline  Press,  Columbia, SC. pp.383-398.

                               217

-------
Gabbot, P.A. 1976.  Energy metabolism. In B.L. Bayne  [ed.],
    Marine Mussels:  Their Ecology and Physiology.  Cambridge Univ.
    Press, pp.294-355.

Gambrell, R.P., V. Collard, and W.H. Patrick, Jr. 1980.  Cadmium
    uptake by marsh  plants as  affected by sediment physico-
    chemical conditions.  In R.A.  Baker [ed.]  Contaminants and
    Sediments,  Vol.  2.  Ann Arbor  Science,  Ann Arbor,  MI.  pp.425-
    445.

Gardner, W.S., W.A.  Frez, E.A. Cichocki, and C.C. Parrish. 1985.
    Micromethods  for lipids in aquatic invertebrates.  Limnol.
    Oceanog.  30:1099-1105

Gobas, F.A.P.C.,  C.G. Derek, and D. MacKay. 1988.  Dynamics of
    dietary bioaccumulation and fecal elimination of  hydrophobic
    organic chemicals  in  fish.  Chemosphere 17:943-962.

Green, R.H. 1979.  Sampling design and statistical methods for
    environmental  biologists.  Wiley-Interscience.  New York, NY.
    257pp.

Herbes, S.E. and C.P. Allen. 1983.  Lipid quantification of
    freshwater invertebrates:  Method modification for microquan-
    tification. Can. J. Fish.  Aquat.  Sci.  40:1315-1317.

Holme, N.A. and A.D. Mclntyre  [eds.]. 1984.  In Methods for the
    Study of Marine  Benthos. IBP  Handbook #16,  2nd Edition.
    Blackwell Scientific  Publications.  Boston,  MA. 387pp.

Hughes, J.S. and A.G.  Vikas. 1983.  Toxicity of N,N-dimethyl-
    formamide used as  a solvent in toxicity tests  with the  green
    alga,  Selenastrum  capricornuturn.  Bull.  Environ. Contam. Tox.
    31:98-104.

Hurlbert, S.H. 1984.  Pseudoreplication and the design of
    ecological  field experiments.  Ecol. Mono.  54:187-211.

Jackim, E., G. Morrison,  and R. Steele. 1977.  Effects of
    environmental  factors on radiocadmium uptake  by four  species
    of marine bivalves. Mar. Biol.  40:303-308.

Jenne, E.A., V.C. Kennedy, J.M. Burchard, and J.W. Ball.  1980.
    Sediment collection and processing  for selective  extraction
    and for total  trace element analyses.  In R.A.  Baker [ed.]
    Contaminants and Sediments, Vol.  2. Ann Arbor Science,  Ann
    Arbor,  MI.  pp.169-190.

Jennett, J.C., S. W. Effler, and E.G. Wilson. 1980.   Mobilization
    and toxicological  aspects  of  sediment  contaminants.   In R.A.
    Baker [ed.] Contaminants and  Sediments,  Vol.  1. Ann Arbor
    Science,  Ann Arbor, MI.  pp.429-444.

                               218

-------
Jensen,  A.L., S.A. Spigarelli, and M.M. Thommes. 1982.  PCB
   uptake by species  of  fish  in  Lake  Michigan,  Green Bay of  Lake
   Michigan,  and  Cayuga,  New  York,  NY.  Can.  J.  Fish.  Aquat.  Sci.
   39:700-709.

Johns, D.M., T.C.  Ginn, and D.J.  Reish. 1989.  Interim protocol
   for  juvenile Neanthes bioassay.  (Draft).  Report  for Wash.
   Dept. Ecology,  Olympia,  Wash. Ecology  No.  C0089018.  PTI
   Contract C901-07.  18p +  appendices.

Johns, D.M., R. Guljahr-Gobell, and P. Schauer. 1985.  Use of
   bioenergetics  to investigate  the impact  of dredge material  on
   benthic  species: A study with polychaetes and  Black Rock
   Harbor material. Technical Report  D-85-07.  Prepared by U.S.
   EPA,  Environmental Research Laboratory,  Narraganett,  RI,
   Contribution # ERL-N  623,  EPA 600/x-86/106 for the U.S. Army
   Engineer Waterways Experiment Station, Vicksburg,  MS.

Karickhoff,  S.W.  1984.  Organic pollutant sorption in aquatic
   systems.  J. Hydraul.  Eng.  110:707-735.

Kates, M. 1986.  Techniques of lipidology (isolation, analysis
   and  identification of lipids),  2nd rev.  ed.  Elsevier Science
   Pub.  Co.,  New  York, NY.  464pp.

Kennedy, P.C. 1986.  The use of molluscs for monitoring trace
   elements in the marine environment in  New Zealand.  1.   The
   contribution of ingested sediment  to the trace element
   concentrations in  New Zealand molluscs.  New Zealand J. Mar.
   F. W. Res.  20:627-640.

King, J.M.  and S.  Spotte. 1974.  Marine Aquariums in the  Research
   Laboratory. Aquarium  Systems, Inc.,  Eastlake,  OH.  39pp.

Kinne, Otto W. 1976-77.  Cultivation of Marine Animals. In
   Marine Ecology. Vol.  III.  Part  2.  Wiley  and Sons.  New York,
   NY.  pp.579-1293.

Knezovitch,  J.P.  and F.L. Harrison. 1987.   A new method for
   determining the concentrations  of  volatile organic compounds
   in sediment interstitial water.  Bull. Environ.  Contain.
   Toxicol.  38:937-940.

Konasewich,  D.E.,  P.M. Chapman, E.  Gerencher, G. Vigers,  and N.
   Treloar.  1982.  Effects, Pathways,  Processes,  and Trans-
   formation of Puget Sound Contaminants  of Concern.  Prepared
   for  NOAA,  Boulder,  CO by E.V.S.  Consultants Ltd.,  Vancouver,
   B.C.  NOAA Technical Memorandum  OMPA-20.  357pp.

Kristensen,  E. 1981.   Direct measurement of ventilation and
   oxygen uptake  in three species  of  tubicolous polychaetes
    (Nereis  spp.).  J.  Comp.  Physiol. 145:45-50.

                               219

-------
Lake, J.L., G. Hoffman, and S. Schimmel. 1985.  Bioaccumulation
    of contaminants from Black Rock Harbor dredged material by
    mussels and polychaetes.  Technical Report D-85-02.  Prepared
    by U.S.EPA,  Environmental  Research Laboratory,  Narragansett,
    RI,  for the U.S.  Army Engineer Waterways Experiment Station,
    Vicksburg,  MS.

Lake, J.L., N.I. Rubinstein, and S. Pavignano. 1987.  Predicting
    bioaccumulation:  Development  of a  partitioning model for use
    as a screening  tool in regulating  oceandisposal of  wastes. In
    K.L.  Dickson, A.W.  Maki,  and  W.A.  Brungs [eds.]  Fate and
    Effects of  Sediment-bound  Chemicals in Aquatic Systems. Sixth
    Pellston Workshop,  Florissant,  CO. August 12-17,  1984.

Lake, J.L., N.I. Rubinstein, H. Lee II, C.A. Lake, J. Heltshe,
    and S.  Pavignano.  1990.   Equilibrium partitioning and
    bioaccumulation of  sediment-associated contaminants  by
    infaunal organisms.  Submitted to Environ.  Toxicol.  Chem.
    9:1095-1106.

Landrum, P.F.  1988.  Toxicokinetics of organic xenobiotics in
    amphipod Pontoporeia hoyjL:  role of physiological  and
    environmental variables. Aquat.  Toxicol.  12:245-271.

Landrum, P.F.  1989.  Bioavailability and toxicokinetics of
    polycyclic  aromatic hydrocarbons sorbed to sediments  for  the
    amphipod Pontoporeia hoyi.  Environ.  Sci.  & Technol.  23:588-
    595.

Landrum, P.P.,  S.R. Nihart, B.J.  Eadie, and W.S.  Gardner. 1984.
    Reverse-phase separation method for determining pollutant
    binding to  Aldrich  humic acid and  dissolved organic  carbon of
    natural waters.  Environ. Sci.  & Technol.  18:187-192.

Landrum, P.F.  and R. Poore. 1988.  Toxicokinetics of selected
    xenobiotics  in  Hexagenia limbata.  J.  Great  Lakes  Res. 14:427-
    437.

Langston, W.J.  1978.  Accumulation of polychlorinated biphenyls
    in the  cockle Cerastoderma  edule and the  tellinid Macoma
    balthica. Mar.  Biol.  45:265-272.

Lauenstein, G.G. and D.R. Young.  1986. Benthic Surveillance
    Project: Cycle  III  Field Manual. National  Status  and  Trends
    Program for Marine  Environmental Quality.  NOAA  Technical
    Memorandum  NOS  OMA  28.  Rockville,  MD. 26pp.

Lee, II, H., B.L.  Boese, J. Pelletier, R.C. Randall,  and D.T.
    Specht.  1990.   Method  to estimate  gut uptake  efficiencies
    for  hydrophobic organic pollutants.  Environ. Toxicol. Chem.
    9:215-220.
                               220

-------
Lee, II, H., B.L. Boese, M. Winsor, R.C. Randall, and D.T.
    Specht.  1988.  (Abstract)  Passive sorption of hexachloro-
    benzene from interstitial water by the  clam, Macoma nasuta.
    Presented at 9th Annual Meeting of the  Soc.  Environ. Toxicol.
    and Chem.  Nov.  13-17,  1988.  Arlington,  VA.

Lee, II, H. and R.C. Randall. 1988.  (Abstract)  Cancer  risk
    associated with sediment quality.  Proceedings of the First
    Annual  Meetingon Puget Sound Research.  Vol.  2.  March 18-19,
    1988. Puget Sound Water Quality Authority,  Seattle,
    Washington, p.676.

Lee, II, H. and R.C. Swartz. 1980.  Biological  processes
    affecting the  distribution of pollutants in marine sediments.
    Part II.  Biodeposition and Bioturbation. In R.A. Baker [ed.]
    Contaminants and Sediments,  Vol.  2.  Ann Arbor Science,  Ann
    Arbor,  MI.  pp.555-606.

Lee, J.J. and W.A. Muller. 1972.  Culture of marsh microorganisms
    and micrometazoa. In W.L.  Smith and M.H. Chanley [eds.]
    Culture of Marine Invertebrate Animals.   Plenum Press,  New
    York, NY.  pp.87-107.

Luoma, S.N. and G.W. Bryan. 1978. Factors controlling avail-
    ability of sediment-bound lead to the estuarine bivalve
    Scrobicularia  plana. J. Mar.  Biol.  Assoc.  U.K.  58:973-802.

Luoma, S.N. and G.W. Bryan. 1982.  A statistical study  of
    environmental  factors  controlling concentrations of heavy
    metals  in the  burrowing bivalve Scrobicularia plana and the
    polychaete Nereis diversicolor.  Estuarine Coastal Shelf Sci.
    15:95-108.

Luoma, S.N. and E.A. Jenne. 1976.  Factors  affecting the
    availability of sediment bound cadmium  to the estuarine
    deposit feeding clam,  Macoma balthica.  In C.E.  Gushing [ed.]
    Radioecology and Energy Resources.  Ecol. Soc.  America.  Spec.
    Publ. No.  1. pp.283-290.

Luoma, S.N. and D.J.H. Phillips. 1988.  Distribution,
    variability, and impacts of  trace elements  in San Francisco
    Bay. Mar.  Poll.  Bull.  19:413-425.

Mac, M.J.,  C. Edsall, R. Hesselberg, and R. Sayers. 1984.
    Flowthrough bioassay for measuring bioaccumulation of toxic
    substances from sediment.  Final report  to EPA.  905/3-84/007.
    17  pp.

Manheim, F.T. 1974.  Comparative studies on extraction  of
    sediment  interstitial  water  studies.  Clays  Clay Minerl.
    22:337-343.
                               221

-------
Mayer, L.M. 1976. Chemical water sampling in lakes and sediments
    using dialysis bags.  Limnol.  Oceanog.  21:909-911.

McElroy, A.E. and J.C. Means.  1988.  Factors affecting the
    bioavailability of hexachlorobiphenyls to benthic  organisms.
    In W.J.  Adams,  G.A.  Chapman,  and W.G.  Landis,  [eds.],
    Aquatic  Toxicology and Hazard Assessment:  10th Volume,  ASTM
    STP 971.  ASTM,  Philadelphia,  PA.  pp.149-158.

McFarland, V.A., J.U. Clark, and A.B. Gibson. 1985.  Bioaccum-
    ulation  of polychlorinated biphenyls  (PCBs)  from Sheboygan
    Harbor sediments in laboratory exposures.  Technical  Report D-
    85-1473.  Prepared by the Environmental Laboratory  of U.S.
    Dept.  of  the  Army for U.S.  Army Engineer Waterways Experiment
    Station,Vicksburg,  MS.  53pp.+ appendices.

McFarland, V.A. and J.U. Clarke. 1986.  Testing bioavailability
    of polychlorinated biphenyls  from sediments  using  a  two-level
    approach.  In  R.G.  Willey [ed.]  Successful  Bridging between
    Theory and Applications.  Proceedings  of  the   Seminar on Water
    Quality  R&D:Davis,  CA.  pp.220-229.

McFarland, V.A. and J.U. Clarke. 1989.  Environmental occurrence,
    abundance,  and potential toxicity of  polychlorinated biphenyl
    congeners:  considerations for a congener-specific  analysis.
    Environ.  Health Perspect.  81:225-239.

McLeese, D.W., C.D. Metcalfe, and S.S. Pezzack.  1980.  Uptake of
    PCB's  from sediment  by Nereis virens  and Crangon
    septemspinosa.  Arch.  Environm.  Contain. Toxicol.  9:507-518.

Mood, A.M.,  F.A. Graybill, and B.C. Boes. 1974.  Introduction to
    the Theory of Statistics,  3rd edition. McGraw-Hill Book
    Company.  New  York,  NY.  546pp.

Mook, D.H. and C.M. Hoskin. 1982.  Organic determination by
    ignition:  Caution advised.  Estuarine  Coastal  Shelf Sci.
    15:697-699.

Morales-Alamo, R. and D.S. Haven. 1983.  Uptake of kepone from
    sediment  suspensions  and subsequent loss by  the oyster
    Crassostrea virginica.  Mar. Biol.  74:187-201.

Morse, J.W., F.J. Millero, J.C. Cornwell, and D. Rickard. 1987.
    The chemistry of the  hydrogen sulfide  and  iron sulfide
    systems  in natural waters.  Earth  Sci.  Reviews.  24:1-42.

National Research Council Committee on Marine Invertebrates.
    1981.  Laboratory Animal Management; Marine Invertebrates.
    National Academy Press,  Washington, B.C.  382pp.
                               222

-------
Nelson, W.G., D.K. Phelps, W.B. Galloway, P.P. Rogerson, and R.J.
    Pruell.  1987   Effects  of  Black Rock Harbor dredge material  on
    the scope for  growth of the blue  mussel  Mytilus  edulis  after
    laboratory and field exposures. Technical  Report D-87-7,
    prepared by the U.S. Environmental  Protection Agency,
    Narragansett,  RI,  for  the U.S.  Army Engineer Waterways
    Experiment Station, Vicksburg,  MS (Available from NTIS,  5285
    Port Royal Road,  Springfield,  VA  22161).

Niimi, A.J. and C.Y.  Cho.  1981.  Elimination of hexachlorobenzene
    (HCB) by rainbow  trout (Salmo  gairdneri),  and an examination
    of its  kinetics in Lake Ontario salmonids.  Can.  J.  Fish.
    Aquat.  Sci. 38:1350-1356.

Niimi, A.J. 1983.   Biological and toxicological effects of
    environmental  contaminants in  fish  and their eggs.  Can.  J.
    Fish. Aquat. Sci.  40:306-312.

Nkedl-Kizza, P.,  P.S. C.  Rao, and A.G.  Hornsby. 1985.  Influence
    of organic cosolvents  on  sorption of hydrophobic organic
    chemicals by soils. Environ. Sci. Technol.  19:975-979.

NOAA. 1988.  A summary of selected data on chemical contaminants
    in sediments collected during  1984,  1985,  1986,  and 1987.
    Progress Report for Marine Environmental Quality.  NS&T
    Program.  NOAA  Technical Memorandum  NOS OMA 44. Rockville, MD.
    15pp.+  appendices.

Norstrom, R.J., A.E.  McKinnon, and A.S. deFreitas. 1976.  A
    bioenergetics  based model for  pollutant  accumulation by fish.
    Simulation of  PCB and  methylmercury residue levels in Ottawa
    River.  J.  Fish. Res. Board Can. 33:248-267.

Office of Federal Register,  National Archives and Records
    Administration, Code of Federal Regulations.  (21 CFR
    177.2600)

Oliver, B.J. and A.J. Niimi.  1983.  Bioconcentration of
    chlorobenzenes from water by Raibow trout:  Correlations with
    partition coefficients and environmental residues.  Environ.
    Sci.  Technol.  17:287-291.

Pearson, T.H. and R.  Rosenberg. 1978.  Macrobenthic succession in
    relation to organic enrichment and  pollution of  the marine
    environment. Oceanogr. Mar.  Biol. Ann. Rev.  16:229-311.

Pearson, T.H. and S.O. Stanley. 1979. Comparative measurement of
    the redox potential of marine  sediments  as a rapid means' of
    assessing the  effect of organic pollution.  Mar.  Biol. 53:371-
    379.
                               223

-------
Peddicord, R.K. 1980.  Direct effects of suspended sediments on
    aquatic  organisms.  In R.A.  Baker [ed.]  Contaminants and
    Sediments.  Vol.1. Ann Arbor Science,  Ann Arbor, MI. pp.501-
    536..

Pelletier, J.K., K.J. Williamson, and H. Lee II. 1988a.
    Bioenergetics  of  Abarenicola pacifica.  In Final Report to
    U.S.  EPA, Newport,  OR.  by K.  J.  Williamson,  S. Woods,  and
    P.O.  Nelson, Oregon State Univ.,  Corvallis,  OR. Project No.
    CR811118-01-3.  pp.1-18.

Pelletier, J.K., K.J. Williamson, and H. Lee II. 1988b.
    Bioaccumulation of  cadmium by Abarenicola pacifica. In Final
    Report toU.S.  EPA,  Newport,  OR.  by  K. J.  Williamson,  S.
    Woods, and  P.O. Nelson,  Oregon State  Univ.,  Corvallis,  OR.
    Project  NO.CR811118-01-3.  pp.18-50.

Pesch,  C. and D. Morgan. 1978. Influence of sediment in copper
    toxicity tests  with the polychaete  Neanthes  arenacedontata.
    Wat.  Res. 12:747-751.

Plumb,  R.H.  Jr. 1981.  Procedures for Handling and Chemical
    Analysis of Sediment and Water Samples.  Technical Committee
    on  Criteria for Dredged and Fill  Material. U.S. EPA/  U.S.
    ACE.  EPA-4805572010.  U.S.  EPA Great Lakes Laboratory.  Grosse
    lie,  MI.  pp.1-1 - 3-403.

Presley, B.J.,  R.R. Brooks, and H.M. Kapp.  1967. A simple
    squeezer for removal of interstitial  water from ocean
    sediments.  J. Mar.  Res.  25:355-357.

Pritchard, R.H., C.A. Monti, E.J. O'Neill,  J.P.  Connolly, and
    D.G.  Ahearn. 1986.   Movement  of  Kepone"  (chlordecone) across
    an  undisturbed  sediment-water interface  in laboratory
    systems.  Environ. Toxicol.  Chem.  5:647-657.

Pruell, R.J., J.L. Lake, W.R. Davis, and J.G. Quinn.  1986.
    Uptake and  depuration of organic  contaminants by blue mussels
    (Mytilus edulis)  exposed to contaminated sediment.  Mar.
    Biol.  91:497-507.

Randall, R.C.,  H.  Lee II, R.J. Ozretich,  J.L. Lake,  and R.J.
    Pruell.   1991.  Evaluation of selected  lipid methods  for
    normalizing pollutant bioaccumulation.   Environ. Toxicol.
    Chem.  10:10.

Reichert, W. L., B.L. Eberhart, and U.  Varanasi. 1985.  Exposure
    of  two species  of deposit-feeding amphipods  to
    sediment-associated BaP:  Uptake,  metabolism, and covalent
    binding  to  tissue macromolecules. Aquat.  Toxicol. 6:45-56.
                               224

-------
Reish, D.J. 1974.  The establishment of laboratory colonies of
    polychaetous  annelids.  Thalassia 10:181-195.

Reish, D.J. 1985. The use of the polychaetous annelid Neanthes
    arenaceodentata  as  a  laboratory  experimentalanimal.  Tethys.
    11:335-341.

Reish, D.J. and M.C. Alosi. 1968.  Aggressive behavior in the
    polychaetous  annelid  family  Nereidae.  Bull.  So.  Calif. Acad.
    Sci.  67:21-28.

Reish, D.J. and T.L. Richards. 1966.  A culture method for
    maintaining large populations of polychaetous  annelids in the
    laboratory. Turtox  News 44:16-17.

Renfro, W.C. and G.  Benayoun. 1975.  Sediment-worm interaction:
    transfer of Zn65 from marine silt by  the polychaete  Nereis
    diversicolor.  In Colbert E.  Gushing,  Jr.  ted.] Radioecology
    and Energy Sources. Proc.  Fourth Nat.  Symposium  on Radiology,
    Oregon  State  University.  4:250-255.

Rhoads, D.C.,  P.L. McCall,  and J.Y.  Yingst. 1978.   Disturbance
    and production on the estaurine  seafloor. Am.  Sci.   66:557-
    586.

River, J.B., J.E. Pearson,  and C.D.  Shultz. 1972.   Total and
    organic mercury  in  marine fish.  Bull.  Environ. Contam.
    Toxicol. 8:257-266.

Robbins,  J.A.  and J. Gustins. 1977.  A squeezer for efficient
    extraction of pore  water from small volumes  of anoxic
    sediment. Limnol. Oceanog. 16:905-909.

Roberts,  J.B.,  A.W.S.  deFrietas, and M.A.J. Gidney.  1977.
    Influence of  lipid  pool size on  bioaccumulation  of the
    insecticide Chlordane by Northern Redhorse suckers Moxostoma
    macrolepidotum.  J.  Fish.  Res.  Board Can. 34:89-97.

Roesijadi,  G.,  D.L.  Woodruff, and J.W.  Anderson. 1978.
    Bioavailability  of  naphthalenes  from  marine  sediments
    artificially  contaminated with Prudhoe Bay oil.  Environ.
    Pollut.  15:223-228.

Rubinstein, N.I., C.N.  D'Asaro,  C. Sommers, and F.G. Wilkes.
    1980.   The effects  of contaminated sediments on  representaive
    estuarine species and developing benthic communities. In R.A.
    Baker [ed.] Contaminants and Sediments. Volume 1. Ann Arbor
    Science, Ann  Arbor, MI.  pp.445-461.
                               225

-------
Rubinstein, N.I., E. Lores, and N. Gregory. 1983.  Accumulation
    of  PCBs, mercury,  and  cadmium by Nereis virens. Mercenaria
    mercenaria.  and  Palaemonetes  pugio  from contaminated harbor
    sediments. Aquat.  Toxicol.  3:249-260.

Rubinstein, N.I., W.T. Gilliam, and N.R. Gregory. 1984.  Dietary
    accumulation of  PCBs from a contaminated  sediment  source  by a
    demersal fish (Leiostomus xanthurus). Aquat.  Toxicol.  5:331-
    342.

Rubinstein, N.I., J.L. Lake, R.J. Pruell, H.  Lee II,  B. Taplin,
    J.  Heltshe,  R. Bowen,  and S.  Pavignano. 1987. Predicting
    bioaccumulation  of sediment-associated  organic  contaminants:
    Development  of a regulatory tool for dredged  material
    evaluation.  Internal report.  U.S. EPA 600/x-87/368,
    Narragansett, RI.  54pp.+ appendices.

Sax, N.I. 1984.   Dangerous Properties of Industrial Materials,
    6th edition.  Van Nostrand Reinhold  Company, New York,  NY.
    3124pp.

Schaeffer, D.J.  and K.G.  Janardan. 1978.  Theoretical comparison
    of  grab and  composite  sampling programs.  Biom.  J.   20:215-
    227.

Schulz-Baldes, M. 1974.  Lead uptake from seawater and food,  and
    lead  loss  in the common mussel Mytilus  edulis.  Mar.  Biol.
    25:177-193.

Sinnett,  J.C.  and W.R. Davis.  1983.   A programmable turbidistat
    for suspended particles in  laboratory aquaria.  J.  Exp. Mar.
    Biol.  Ecol.  73:167-174.

Smith,  K.L. Jr.  1978.  Benthic community respiration in the N.W.
    Atlantic Ocean:  In situ measurements from 40  to 5200 m.   Mar.
    Biol.  47:337-347.

Smith,  K.L. Jr., G.T. Rowe, and J.A.  Nichols. 1973.  Benthic
    community  respiration  near  the Woods Hole sewage outfall.
    Estuar. and  Coastal Mar. Sci.  1:65-70.

Sokal,  R.R. and F.J. Rohlf. 1981.  Biometry,  2nd edition. W.H.
    Freeman and  Company. New York, NY.  859pp.

Spacie, A. and J.L.  Hamelink.  1982.   Alternative models for
    describing the bioconcentration of  organics in  fish.
    Environ. Toxicol.  Chem.  1:309-320.

Specht, D.T.  and H.  Lee II. 1989.  Direct measurement technique
    for determining  ventilation rate in the deposit-feeding clam
    Macoma nasuta (Bivalvia, Tellinaceae).  Mar. Biol.  101:211-
    218.

                               226

-------
Stein, J.E., T. Horn, E. Casillas, A. Friedman, and U. Varanasi,
    1987.  Simultaneous exposure  of  English  sole  (Parophrys
    vetulus)  to sediment-associated  xenobiotics:  Part 2-Chronic
    exposure to an urban estuarine sediment  with  added 3H-
    benzo[a]pyrene and  14C-polychlorinated biphenyls.  Mar.
    Environ.  Res. 22:123-149.

Swartz, R.C., W.A.  DeBen, and F.A.  Cole. 1979.  A bioassay for
    Toxicity of sediment to marine macrobenthos.  J. Wat.  Poll.
    Control  Fed.  51:944-950.

Swartz, R.C., W.A.  DeBen, J.K.P.  Jones, J.O. Lamberson,  and F.A.
    Cole.  1985.   Phoxocephalid amphipod bioassay  for  marine
    sediment toxicity.  In R.D.  Cardwell,  R.  Purdy,  and R.C.
    Banner,  [eds.].  Aquatic Toxicology  and Hazard Assessment:
    Seventh  Symposium,  ASTM STP 854.  ASTM, Philadelphia,  PA. pp.
    284-307.

Swartz, R.C., G.R.  Ditsworth, D.W.  Schults,  and J.O.  Lamberson.
    1986.  Sediment  toxicity  to a marine infaunal amphipod:
    cadmium  and its  interaction with sewage  sludge. Mar.  Environ.
    Res.  18:133-153.

Swartz, R.C., P.P.  Kemp., D.W. Schults, and J.O.  Lamberson. 1988.
    Effects  of  mixtures of sediment  contaminants  on the  marine
    infaunal amphipod,  Rhepoxynius abronius.  Environ. Toxicol.
    7:1013-1020.

Swartz, R.C., P.P.  Kemp., D.W. Schults, G.R. Ditsworth, and R.G.
    Ozretich.  1989.   Acute toxicity  of  sediment from  Eagle
    Harbor,  Washington  to the infaunal  amphipod,  Rhepoxynius
    abronius.  Environ.  Toxicol. Chem. 8:215-222.

Swartz, R.C., D.W.  Schults,  G.R.  Ditsworth,  and W.A.  DeBen. 1984
    Toxicity of sewage  sludge to  Rhepoxynius abronius. a marine
    benthie  amphipod. Arch. Environ.  Contain.  Toxicol. 13:359-364.

Tetra Tech,  Inc. 1985a.  Bioaccumulation Monitoring  Guidance: 2.
    Selection of Target Species and  Review of Available
    Bioaccumulation  Data. Final Report  prepared for Marine
    Operations  Division, U.S.  EPA, Washington, D.C." EPA Contract
    NO. 68-01-6938,  TC  3953-03. Bellevue,  WA.

Tetra Tech,  Inc. 1985b.  Bioaccumulation Monitoring  Guidance:
    Recommended analytical detection limits.  Final  Report
    prepared for U.S. EPA, Washington,  D.C.  EPA Contract NO.
    68-01-6938,  TC 3953-03. Bellevue, WA.

Tetra Tech,  Inc. 1986a.  ODES Statistical Power Analysis. Draft
    Report,  prepared for Office of Marine and Estuarine
    Protection,  U.S.  EPA Contract NO. 68-01-6938, TC-3953-03.
    Bellevue,  WA.

                               227

-------
Tetra Tech, Inc. 1986b.  Quality Assurance/Quality Control
    (QA/QC)  for  301(h)  Monitoring  Programs:  Guidance  on Field and
    Laboratory Methods.  Final  Report  for Marine Operations
    Division, U.S. EPA,  Washington, DC.  EPA Contract  NO 68-01-
    6939, TC-3953-04.  Bellevue,  WA.

Tetra Tech, Inc. 1986c.  Analytical methods for U.S.  EPA piority
    pollutants and 301(h) pesticides  in  estuarine  and marine
    sediments. Final Report  prepared  for Marine Operations
    Division, U.S. EPA,  Washington, D.C.  EPA Contract NO.  68-01-
    6938, TC-3953-03.  Bellevue.  WA.

Tetra Tech, Inc. 1986d.  Guidance manual for health  risk
    assessment of chemically contaminated seafood.  Prepared for
    U.S.  EPA, Region X,  Seattle, WA.  TC  3991-97. Bellevue,  WA.

Tetra Tech, Inc. I986e.  Bioaccumulation Monitoring  Guidance:
    Analytical methods for U.S.  EPA priority pollutants and
    301(h)  pesticides  in tissues from estuarine and marine
    organisms. Prepared for  Marine Operation Division,  U.S. EPA,
    Washington,  D.C. EPA Contract  NO.  68-01-638, TC-3953-03.
    Bellevue, WA.

Tetra Tech, Inc. 1986f.  Recommended protocols for measuring
    metals  in Puget Sound water, sediment,  and tissue samples.
    Final Report prepared for  Resource Planning Associates for
    U.S.  Army Corps of Engineers,  Contract NO.  TC-3090-04.
    Seattle District,  Seattle, WA. Bellevue,  WA.

Tetra Tech, Inc. I986g.  Recommended protocols for measuring
    organic compounds  in Puget Sound  sediment and  tissue samples.
    Final Report prepared for  U.S. EPA.  Contract NO.  TC-3991-04,
    Region  X, Seattle  WA. Bellevue, WA.

Tetra Tech, Inc. 1987.  Bioaccumulation Monitoring Guidance:
    Strategies  for Sample Replication and Compositing.  Vol. 5.
    Prepared for U.S.  EPA, Office  of  Marine and Estuarine
    Protection.  Washington,  DC.  U.S.  EPA Contract  NO.430/09-87/-
    003.  Bellevue, WA.

The Bureau  of National Affairs,  Inc. 1986.  U.S. Environmental
    Protection Agency  General  Regulation for Hazardous Waste
    Management.  (40  CFR 260; 45  FR 12724, February 26,  1980.
    Effective August  26,  1980; Amended as shown in Code of
    Federal Regulations,  Volume  40,  Revised as of  July, 1,  1985;
    50 FR 28742, July  15,  1985;   51 FR 10174,  March 24, 1986,
    Effective September 22,  1986;   51 FR 16443, May 2,  1986,
    Effective October  29,  1986;  51 FR 25470,  July  14, 1986,
    Effective January  12,  1987;  51 FR 28682,  August 8,  1986,
    Effective November 8,  1986), SEC. 260.1 (b) (4).   Washington,
    D.C.
                               228

-------
Thomann, R.V. and J.P. Connolly. 1984. Model of PCB in the Lake
   Michigan  lake  trout  food chain.  Environ.  Sci.  Technol.  18:65-
   71.

Thompson, B.E., J. Cross, J. Laughlin, P. Hershelman, R. Gossett,
   and  D.  Tsukada.  1983.   Sediment  and biological conditions on
   coastal slopes.  In Willard  Bascom [ed.],  Southern California
   Coastal Water  Research Project.  Biennial  Report,  1983-1984.
   Long Beach,  California,  pp.37-67.

Tyler-Schoeder, D.B. 1978a.  Culture of the grass shrimp
   Palaemonetes pugio in the laboratory.  Ocean Disposal  Working
   Group.  Office  of  Research and Development.  Bioassay
   Procedures  for the Disposal  Permit Program.  U.S.  EPA-600/9-
   76/010. Gulf Breeze,  FL.  pp.69-72

Tyler-Schoeder, D.B. 1978b.  Entire life-cycle toxicity  tests
   using grass shrimp Palaemonetes  pugio Holthius.  Ocean
   Disposal  Working  Group.  Office of  Research  and Development.
   Bioassay  Procedures  for the  Disposal  Permit Program.  U.S.
   EPA-600/9-76/010.  Gulf Breeze, FL.  pp.83-88.

U.S.  Coast Guard.  1986.   CHRIS Hazardous Chemical Data,   Chemical
   Hazard  Response  Information  System. Washington,  D.C.  Phone
   #202/426-9568.

U.S.  ACE. 1988.  Puget Sound Dredged Disposal Analysis Report
   Technical Appendix Evaluation Procedures  Phase I (Central
   Puget sound).  U.S. Army Corps of Engineers,  Washington  State
   Department  of  Natural  Resources,  U.S.  EPA -  Region 10.

U.S.  EPA. 1978.  Ocean Disposal Working Group.  Office of Research
   and  Development.  Bioassay Procedures  for  the Disposal Permit
   Program.  U.S.  EPA-600/9-76/010.  Gulf  Breeze, FL.  121pp.

U.S.  EPA. 1986a.  NEIC Policies and Procedures Manual. National
   Enforcement Investigations  Center.  EPA-330/9-78/001-R.
   Denver, CO.  (Revised May, 1986).

U.S.  EPA. 1986b.  Techniques for study of effects and uptake of
   sediment  associated  chemicals. Office of  Toxic Substances.
   U.S.  EPA. Gulf Breeze,  FL.

U.S.  EPA. 1987.  Guidelines for the Selection of Chemical
   Protective  Clothing,  3rd. ed., Vol I  and  II. Washington,  D.C.
   Phone AC(202)382-3647.

U.S.  EPA. 1988.  USEPA Contract Laboratory Program Statement of
   Work for  Organics  Analysis  (Multi-Media), Multi-Concen-
   tration) . Washington,  D.C.  SOW NO.  2/88.
                               229

-------
U.S. EPA. 1989.  USEPA Contract Laboratory Program Statement of
    Work for Inorganic Analysis (Multi-Media,  Multi-Concentra-
    tion) .  Washington,  B.C.  SOW NO.  788 (Revised 2/89).

U.S. EPA/U.S. ACE. 1977.  Ecological Evaluation of Proposed
    Discharge of Dredged Material  into  Ocean Waters.  Technical
    Committee on Criteria for Dredged and Fill Material.
    Environmental Effects Laboratory. U.S.  Army Engineer
    Waterways Experiment Station.  Vicksburg,  MS.

U.S. EPA/U.S. ACE. 1988. Ecological evaluation of proposed
    discharge of dredgedmaterial into ocean waters.
    Implementation Manual. Draft #3.  Prepared by Battelle,  Ocean
    Sciences,  Duxbury,  MA. for U.S.  EPA Office of Marine and
    Estaurine Protection,  Washington, D.C.  Contract  No.  68-03-
    3319.

Varanasi, U., W.L. Reichert, J.E. Stein, D.B. Brown, and H.R.
    Sanborn.  1985.   Bioavailability  and transformation  of
    aromatic hydrocarbons in benthic organisms exposed  to
    sediment from an urban estuary.  Environ.  Sci.  Technol.
    19:836-841.

Veith, G.D. 1975.  Baseline concentrations of PCBs and DDT in
    Lake Michigan fish.  Pet.  Monit.  J.  9:21-29.

Veith, G.D. and V.M. Comstock. 1975.  Apparatus for continuously
    saturating water with hydrophobic organic chemicals.  J.  Fish.
    Res.  Board Can.  32:1849-1851.

Waldichuk,  M. 1985.  Biological availability of metals to marine
    organisms.  Mar.  Poll.  Bull.  16:7-11.

Walkley, A. and I.A. Black.  1934.  An examination of the
    Degtjareff method for determining soil  organic matter,  and  a
    proposed modification of the chromic  acid  titration method.
    Soil Sci.  37:29-38.

Williams, A.B. and H.J. Porter. 1971.  A ten-year study of
    meroplankton in North Carolina estuaries:  Occurrence of
    postmetamorphal bivalves.  Ches.  Sci.  12:26-32.

Williams, A.B. and K.H. Bynum. 1972.  A ten-year study of
    meroplankton in North Carolina estuaries:  Amphipods.  Ches.
    Sci.  13:175-192.

Winsor, M., B.L. Boese, H. Lee, II,  R.C. Randall, and D.T.
    Specht.  1990.   A method  of estimating the fluxes  of
    interstitial and overlying water by the clam Macoma  nasuta.
    Environ.  Toxicol.  Chem.  9:209-213.
                               230

-------
Word, J.Q. 1978.  The infaunal index. In W. Bascom [ed.]  Southern
    California Coastal  Water Research Project Annual Report,
    1978.  Long Beach,  CA.  pp.19-39.

Word, J.Q., J.A. Ward, L.M. Franklin, V.I.  Cullinan,  and  S.L.
    Kiesser.  1987.   Evaluation of the equilibrium partitioning
    theory for estimating  the toxicity of the nonpolar organic
    compound  DDT to  the sediment  dwelling amphipod Rhepoxynius
    abronius.  WA56,  Task 1.  Prepared for U.S.  EPA by Battelle,
    Washington,  D.C.

Young, D.R. 1988.  Report on the assessment and  application of
    pollutant biomagnification potential in near coastal waters.
    U.S.  EPA  Internal  Report #600/x-88/295.  U.S.  EPA
    Environmental Research Laboratory, Narragansett, RI, Pacific
    Division,  Newport,  OR.  185p.

Young, D.R.,  A.J. Mearns, and T.K.  Jan. 1987.  The cesium:
    potassium index  of  foodweb structure and biomagnification of
    trace elements in  a polluted  harbor of southern California.
    In S.E. Lindberg and T.C.  Hutchinson [eds.]  Proc.
    International Conference on Heavy Metals in the Environment.
    Vol.  2. CEP Consultants LTD.,  Edinborough, U.K. pp.74-76.

Yevich, P.P., C. Yevich, G. Pesch,  and W. Nelson.  1987.   Effects
    of Black  Rock Harbor dredged  material on the histopathology
    of the blue mussel  Mytilus edulis and polychaete worm Nephtys
    incisa after laboratory and field exposures.  Field
    Verification Program (Aquatic Disposal)  Technical Report D-
    87-8,  prepared by  the  U.S.  Environmental Protection Agency,
    Narragansett,  RI,  for  the U.S.  Army Engineer Waterways
    Experiment Station,  Vicksburg,  MS (Available from NTIS, 5285
    Port  Royal Road, Springfield,  VA 22161).

Zimmermann, C.F., M.T. Price, and J.R. Montgomery.  1978.
    Comparison of ceramic  and Teflon" In Situ  samplers for
    nutrient  pore water determinations.  Estuar.  Coastal Mar. Sci.
    7:93-97
                               231
                                        U.S. GOVERNMENT PRINTING OFFICE: 1993 - 750-002/80301

-------