-------
13. Given the increased hydrophobicity of alkylated PAHs, is the use of the
correction factor associated with the corresponding parent an appropriate
approach for estimating steady state residues of alkylated PAHs? If not, please
elaborate.
14. For the DDT derivatives anddieldrin, please comment on the appropriateness of
using M. nasuta data rather than N. virens-specific data in the estimation of steady
state multipliers.
15. Are the approaches taken to adjust organic contaminant bioaccumulation data to
steady state adequate? Do the proposed multipliers agree with previously
published studies (i.e., do they appear reasonable)? If not, please elaborate.
16. What are the major sources of uncertainty associated with the approaches? What
alternative approaches would reduce the uncertainties? How could these
uncertainties be described and accounted for in decision-making?
Adjustment to Steady State: Metals
17. In your opinion, is the methodology followed to derive the steady state multiplier
for non-essential metals (i.e., a factor of three) scientifically appropriate
(Appendix G)? Please elaborate. Do you have any recommendations of additional
or alternate methodologies or information that can be used to either supplement or
replace the proposed method?
Human Health Evaluations: Overall
18. Please comment on each factor listed above (and in Table 5) as to its
appropriateness for use in the equations listed above. Would you recommend
additional factors? Would you change or modify the equations as written above?
If so, how?
19. Are the methods used to derive the human health exposure parameters and
assigned values discussed in Section E appropriate (please review the referenced
appendices)? If not, please elaborate. How should these factors be factored into
the risk analyses and decision-making?
20. Is the approach taken to relate fish whole body and fillet concentrations
scientifically appropriate? If not, what method would you recommend?
21. Could the analysis be improved by focusing on key fish (seafood) species at the
HARS? What characteristics should be used to select these key species?
22. In your opinion, is the approach for assuming total metal to be in the most toxic
form appropriate and reasonable? Should metal speciation/complexation be
considered in the assessment of metals bioaccumulation, trophic transfer, and
human health risks? Is the proposed approach for evaluating methyl mercury
appropriate? Are there alternative analytical or risk assessment techniques
-------
available that would improve the risk assessment of metals? Is the multiplier
proposed for adjusting measured concentrations of arsenic appropriate and
reasonable?
23. Is the assumption that the potency of alkylated PAHs can be estimated by the
potency of the parent PAH appropriate? Is this assumption likely to result in an
under- or overestimate of the risk associated with the alkylated PAHs?
24. Please comment on the potential for human exposure to PAHs through
consumption of finfish and other seafood.
25. What are the major sources of uncertainty associated with the approaches
described in Section E? What alternative approaches would reduce the
uncertainties? How could these uncertainties be described and accounted for in
decision-making?
26. What is your recommendation for evaluating the potential toxicity of organotins?
Should they be evaluated as individual compounds? Summed as total? Should
there be some consideration of relative toxicity?
27. Please comment on the appropriateness of the proposed approach for converting
and using the analytical data for alkylated and parent PAHs to estimate risk from
all PAHs.
Human Health Evaluations: Comparison to HARS-Specific Values
28. Do you believe that the "disaggregate" modeling discussed above (and shown in
Figure 4) for estimating human health HARS-Specific Values for lead is
appropriate? Would you recommend an alternative risk assessment method be
used given the information and data available? Do you believe the method
described has appropriately taken uncertainty into account? Please elaborate.
Human Health Evaluations: Consideration of Combined effects
29. In your opinion, are the methodologies and equations described above appropriate
for estimating total carcinogenicity and combined non-cancer impacts of
contaminant mixtures accumulated from dredged materials proposed for use as
Remediation Material at the HARS?
30. Is the conceptual model for evaluating fish exposure to dredged material at the
HARS and human exposure through ingestion of seafood appropriate and
reasonable? How can the uncertainties associated with the assumptions in this
conceptual model be reduced? Please consider the spatial and temporal elements
of exposure in your discussion.
-------
PROPOSED
TESTING EVALUATION
FRAMEWORK (TEF)
-------
Proposed Bioaccumulation Testing
Evaluation Framework for Assessing the
Suitability of Dredged Material to be
Placed at the Historic Area Remediation
Site (HARS)
-------
Background for Reviewers 3
A. Brief Discussion on the HARS 3
B. Brief Description of Testing and Evaluation of Dredged Material 4
C. Brief History of Regional Bioaccumulation Test Evaluation Guidelines 7
Proposed Bioaccumulation Evaluation Process 9
A. Bioaccumulation Analytes 11
1. Proposed Additions to Analyte List 12
i. Alkylated PAHs 12
ii. Organotins 13
iii. Coplanar PCB Congeners 13
B. Comparison to Reference 14
C. Comparison to Regional Dioxin Values 15
D. Adjustment to Steady State 15
1. Steady State Multipliers for Organic Compounds 16
2. Steady State for Metals (Safety Factors) 19
E. Human Health Evaluations 19
1. Comparison of Individual Constituents to HARS-specific Values 25
2. Consideration of Combined Risk of Contaminants 28
i. Combined Effects Evaluation: Total Carcinogenicity 28
ii. Combined Effects Evaluation: Non-Cancer Hazard Index 28
References 30
Appendices 33
A. Definition of Remediation Material and Summary of Data and Rationale
Supporting the Need for Remediation of the HARS 33
B. Example Memorandum Documenting Current HARS Suitability Evaluation 39
C. Estimation of Total PCB Residue Based on Reported Concentration of 22
Congeners: Review of NY/NJ Harbor Regional Data 81
D. Proposed Analytical Protocol for Alkylated PAHs 83
E. EPA Region 2 Review of NY/NJ Harbor Data on Organotins 145
F. Proposed HARS-Specific Values for Assessing Potential Human Health
Risks Associated with Contaminants Accumulated from Dredged Material 147
G. Use of Ranges of Metals Body Residues Measured in Field-Collected
Polychaetes to Derive a Safety Factor to Account for Potential Underestimates
of Metals Uptake by 28-day Exposure Duration of Bioaccumulation Assay 149
H. Derivation of an HARS-Specific Tissue Guideline Value for Lead 153
I. Identification of Target Population and Estimation of Seafood 159
Consumption Rate
J. Consideration of HARS Site Use by Finfish 163
K. Whole Body to Fillet Correction Factors 167
L. Trophic Transfer 169
1
-------
This page intentionally left blank.
2
-------
I. BACKGROUND FOR REVIEWERS
A. BRIEF DISCUSSION ON THE HARS
EPA de-designated and terminated use of the New York Bight Dredged Material Disposal Site
[also known as Mud Dump Site (MDS)], and simultaneously designated the Historic Area
Remediation Site (HARS) (see 40 CFR 228.15(a)(d)(6)), in a final rule that became effective on
September 29,1997 (EPA, 1997c). Pursuant to the rule, the HARS is restricted to receive only
dredged material suitable for use as Material for Remediation (also referred to as Remediation
Material). Material for Remediation is defined in the HARS final rule preamble as
"uncontaminated dredged material (i.e., dredged material that meets current Category I1 standards
and will not cause significant undesirable effects including through bioaccumulation)."
The need for remediating the HARS was described in detail in the HARS SEIS (EPA, 1997a),
associated proposed (EPA, 1997b) and final (EPA, 1997c) rulemaking, and the Response to
Comments on the proposed rule (EPA, 1997d). In summary, the initial proposal to terminate and
de-designate the MDS, and to simultaneously re-designate the site and surrounding degraded
areas as the HARS, was supported by the presence of toxic effects in the HARS, dioxin
bioaccumulation exceeding Category I1 levels in worm tissue collected from the HARS, and
PCB/TCDD contamination in area lobster stocks. The SEIS (EPA, 1997a) and introductory
sections prefacing EPA's Response to Comments on the May 13, 1997 proposed rule (EPA
1997d) summarize the data and rationale supporting the decision to remediate the HARS by
capping the site with dredged materials shown to be Category I in laboratory testing. Relevant
sections of the Response to Comments document and a definition of Category I dredged material
is provided as Appendix A of this document.
Individual elements of the aforementioned data do not prove that sediments within the HARS are
imminent hazards to the New York Bight Apex ecosystem, living resources, or human health.
However, the collective evidence presents cause for concern, and justifies the finding that a need
for remediation exists, that the site is Impact Category I (see, 40 CFR 228.102), and that the site
1 Categories I, II, III were defined for the former Mud Dump Site. The Category I definition is used as part of the
definition of Remediation Material at the HARS. For a definition of Category I, see Appendix A.
2 40 CFR 228.10(b). The following types of effects, in addition to other necessary or appropriate considerations, will
be considered in determining to what extent the marine environment has been impacted by materials disposed of at
an ocean disposal site: (1) Movement of materials into estuaries or marine sanctuaries, or onto oceanfront beaches,
or shorelines; (2) Movement of materials toward productive fishery or shellfishery areas; (3) Absence from the
disposal site of pollution-sensitive biota characteristic of the general area; (4) Progressive, non-seasonal, changes in
water quality or sediment composition at the disposal site, when these changes are attributable to materials disposed
of at the site; (5) Progressive, non-seasonal, changes in composition or numbers of pelagic, demersal, or benthic
biota at or near the disposal site, when these changes can be attributed to the effects of materials disposed of at the
site; (6) Accumulation of material constituents (including without limitation, human pathogens) in marine biota at or
near the site.
40 CFR 228.10(c). The determination of the overall severity of disposal at the site on the marine environment,
including without limitation, the disposal site and adjacent areas, will be based on the evaluation of the entire body
3
-------
•J
should be managed to reduce impacts to acceptable levels fsee. 40 CFR 228.11(c) ]. (For more
information see the HARS SEIS)
B. BRIEF DESCRIPTION OF TESTING AND EVALUATION OF DREDGED MATERIAL
EPA's Ocean Dumping regulations (under which the HARS was designated) at 40 CFR 227.6(c)
indicate that evaluating the potential for significant undesirable effects must include the use of
biological assays. EPA and USACE believe that sediment contaminant concentrations alone
provide neither a measure of adverse biological effects nor an estimate of the potential for
effects. Therefore, the Agencies believe that sediment contaminant levels in dredged material do
not necessarily have to be lower than in exposed HARS sediments to be deemed suitable for use
as Remediation Material. The Ocean Dumping regulations also require a demonstration that
there are "no practicable alternative locations" for disposal of proposed dredged material before
they may be placed in the ocean.
of pertinent data using appropriate methods of data analysis for the quantity and type of data available. Impacts will
be categorized according to the overall condition of the environment of the disposal site and adjacent areas based
on the determination by the EPA management authority assessing the nature and extent of the effects identified in
paragraph (b) of this section in addition to other necessary or appropriate considerations.
The following categories shall be used: (1) Impact Category I: The effects of activities at the disposal site shall be
categorized in Impact Category I when one or more of the following conditions is present and can reasonably be
attributed to ocean dumping activities:
(i) There is identifiable progressive movement or accumulation, in detectable concentrations above normal
ambient values, of any waste or waste constituent from the disposal site within 12 nautical miles of any
shoreline, marine sanctuary designated under title III of the Act, or critical area designated under section
102(c) of the Act; or
(ii) The biota, sediments, or water column of the disposal site, or of any area outside the disposal site where any
constituent from the disposal site is present in detectable concentrations above normal ambient values, are
adversely affected by the toxicity of such waste or waste constituent to the extent that there are statistically
significant decreases in the populations of valuable commercial or recreational species, or of specific species
of biota essential to the propogation of such species, within the disposal site and such other area as compared
to populations of the same organisms in comparable locations outside such site and area; or
(iii) Solid waste material disposed of at the site has accumulated at the site or in areas adjacent to it, to such an
extent that major uses of the site or of adjacent areas are significantly impaired and the Federal or State
agency responsible for regulating such Uses certifies that such significant impairment has occurred and states
in its certificate the basis for its determination of such impairment; or
(iv) There are adverse effects on the taste or odor of valuable commercial or recreational species as a result of
disposal activities; or
(v) When any toxic waste, toxic waste constituent, or toxic byproduct of waste interaction, is consistently
identified in toxic concentrations above normal ambient values outside the disposal site more than 4 hours
after disposal.
3 40 CFR 228.11(c) When the EPA management authority determines that activities at a disposal site have placed the
site in Impact Category I, the Administrator or the Regional Administrator, as the case may be, shall place such
limitations on the use of the site as are necessary to reduce the impacts to acceptable levels.
4
-------
All dredged material proposed for use as Remediation Material at the HARS is evaluated for its
suitability through the application of bioassays consistent with guidance provided in the joint
EPA/US ACE 1991 manual entitled "Evaluation of Dredged Material Proposed for Ocean
Disposal" (or the 'Green Book') and joint EPA Region 2/USACE-New York District (CENAN)
guidance for regional implementation of the Green Book. Under this guidance, the potential for
significant undesirable effects due to contaminants in the solid phase of dredged material is
assessed through: (1) 10-day exposure of two sensitive benthic marine organisms to bedded
sediment to assess the potential for acute toxicity; and (2) 28-day exposure of two infaunal
organisms (the deposit feeding clam, Macoma nasuta and a burrowing, sediment-ingesting
worm, Nereis virens) to bedded sediment to assess the potential for bioaccumulation of
contaminants from the sediment.
In both assays, a reference comparison is made using sediment collected from a location in the
New York Bight Apex that is unaffected by activities at the HARS (Figure 1). HARS reference
sediment is predominately (>95%) sand and is therefore substantially free of contaminants.
Bioaccumulation of contaminants from HARS reference sediments is correspondingly low.
Typical bioaccumulation resulting from exposure to HARS reference sediment is reported in
Table 1 of Appendix B (Columns 2 and 3).
5
-------
HARS
• HARS Reference Site
Figure 1. Location of the Reference Site Relative to the Historic Area Remediation Site (HARS)
6
-------
The Green Book provides clear quantitative guidance for interpreting the results of acute toxicity
assays showing greater toxicity in dredged material exposures than in reference exposures.
Guidance provided in the Green Book for interpreting the results of bioaccumulation testing
however, only recommends qualitative (or semi-quantitative) factors to consider in the
assessment. Specifically, the Green Book recommends that US ACE and EPA "develop and
agree upon case-specific evaluative criteria" while considering the following factors:
• Number of species in which bioaccumulation from the dredged material is statistically
greater than bioaccumulation from the reference material;
• Number of contaminants for which bioaccumulation from the dredged material is
statistically greater than bioaccumulation from the reference material;
• Magnitude by which bioaccumulation from the dredged material exceeds
bioaccumulation from the reference material;
• Toxicological importance of the contaminants whose bioaccumulation from the dredged
material statistically exceeds that from the reference material;
• Phylogenetic diversity of the species in which bioaccumulation from the dredged material
statistically exceeds bioaccumulation from the reference material;
• Propensity for the contaminants with statistically significant bioaccumulation to
biomagnify within aquatic food chains;
• Magnitude of toxicity and number and phylogenetic diversity of species exhibiting greater
mortality in the dredged material than in the reference material;
• Magnitude by which contaminants whose bioaccumulation from the dredged material
exceeds that from the reference material also exceed the concentrations found in
comparable species living in the vicinity of the proposed disposal site.
C. BRIEF HISTORY OF BIOACCUMULATION TEST EVALUATION GUIDELINES
EPA Region 2/CENAN have derived and applied HARS-specific guidelines for evaluating
bioaccumulation test results. The current guidelines used by EPA Region 2/CENAN evaluate the
suitability of dredged material with respect to human health are briefly described below.
1. FDA Action Levels: There are FDA Action levels for seven compounds (aldrin, dieldrin,
a-chlordane, heptachlor, heptachlor epoxide, PCBs, and mercury). The source of FDA
Action levels is described in the Green Book Exceeding a FDA Action level results in a
conclusion that the placement of the dredged material would result in significant adverse
effects and the material would be unsuitable for use as Remediation Material.
2. Regional Matrix Values: In 1981, an assessment of ambient conditions in the New York
Bight Apex was made using field collected tissue residue levels or water contaminant
7
-------
concentrations. Tissue guidelines for assessing bioaccumulated levels of four
contaminants (i.e., mercury, cadmium, total PCBs and total DDT) in test organisms were
derived using this field data. When bioaccumulated levels of these compounds did not
exceed their respective guideline value, it was concluded that ocean disposal of the
dredged material would not be expected to further degrade conditions (i.e., cause
elevation of tissue residues in ambient organisms) in the New York Bight Apex and
therefore not result in significant undesirable effects.
3. Dioxin: Regional Dioxin Values were developed and detailed in a policy memorandum
dated March 15, 1997 (EPA, 1997e). That memorandum defined the current process for
Category I as dredged material in which wet weight test tissue concentrations of 2,3,7,8-
TCDD in test organisms exposed for 28 days do not exceed 1 part per trillion (pptr), and
in which the total toxicity equivalence of all non-2,3,7,8-TCDD dioxin and furan
congeners in test tissue do not exceed 4.5 pptr. The Regional Dioxin Value of 1 pptr
(2,3,7,8-TCDD) reflects the best available detection limit at the time the process was
developed. The Regional Dioxin Value for the total toxicity equivalence (TEQ) of non-
2,3,7,8-TCDD dioxins/furans is based on detection limits, rather than risk.
4. December 1996 Guidelines: In December 1996, human health and ecological benthic
tissue guideline values were derived to evaluate bioaccumulation test results for all
measured contaminants not having associated Regional Matrix or Dioxin Values. Human
health guideline values were derived assuming that the target population consumed
seafood at the national average rate and adopting several conservative exposure
assumptions.
An example of the use of current guidelines (described above) in making suitability
determinations regarding dredged material proposed for use as Remediation Material at
the HARS is attached as Appendix B.
In 1998, the guidelines and evaluative framework used to evaluate bioaccumulation test
results of material proposed for use as Remediation Material at the HARS were submitted
to a panel of experts for scientific peer review. In light of the comments of those peer
reviewers and comments received from New York-New Jersey Harbor stakeholders, the
EPA Region 2 has proposed a modification to the testing evaluation framework that,
when finalized EPA Region 2 and CENAN will use to evaluate future bioaccumulation
test results. The EPA and Corps executed a Memorandum of Agreement in September
2000 to implement several actions (e.g., technical and policy reviews by the Remediation
Material Workgroup, peer review, and public review) on an identified timeline to
culminate in a final updated TEF document.
A brief overview of the proposed evaluation process and a detailed description of the
technical elements supporting the proposed process for assessing human health risk
associated with contaminants bioaccumulated from dredged material proposed for use as
Remediation Material at the HARS are provided in the following sections and in the
8
-------
attached appendices. The process for evaluating ecological risks is not discussed in this
document and will be the subject of a separate peer review effort.
s
Charge:
1. Throughout the proposed process, there are various uncertainties introduced. Please
identify the key areas of uncertainty that need to be addressed. Are there additional data
sources or parameters that could be used to address these areas? What methods are
available for describing and accounting for these uncertainties in the calculation of
HARS-Specific Values? Of the methods available, which would you recommend for
consideration and why? Please consider the implications of implementing these methods
in the regulatory framework. Please include an evaluation of probabilistic and
deterministic methods in your discussion.
II. PROPOSED BIOACCUMULATION EVALUATION PROCESS
EPA Region 2 is proposing to use the framework described in Figure 2 to evaluate the potential
for ecological and human health effects associated with bioaccumulation of contaminants
associated with sediments proposed for placement as Remediation Material at the HARS.
As shown in Figure 2, evaluating bioaccumulation test results to assess the potential for human
health and ecological effects involves a four-step evaluation process (numbers in parentheses
correspond to the numbered steps shown in Figure 2). In the first three steps, accumulations of
contaminants measured in tissues of organisms exposed to project sediments are sequentially
evaluated by: (1) statistical comparison to accumulations in test organisms exposed to reference
sediment; (2) comparison to regional dioxin values; and (3) comparison to benthic tissue
guideline values (i.e., HARS-specific Values) associated with specified levels of risk/hazard. If
contaminant residues in test organism tissues statistically exceed residues in reference organisms,
but do not exceed Regional dioxin Category I values in Step 2 or any of the HARS-Specific
Values for specific contaminants in step 3, then a fourth evaluation (4) is performed. This step
considers the potential for human health risk/hazard (combined carcinogenic risk and combined
non-cancer effects hazard) based on the combined action of multiple contaminants accumulated
from the dredged material by test organisms.
The HARS-Specific Values and TEF are not binding regulatory criteria. EPA Region 2/CENAN
intend to use them as tools in considering the weight of evidence regarding the suitability of a
given dredged material for use as Remediation Material at the HARS. Factors that may be
considered in the weight of evidence include: variability around the mean accumulated residues
and the analytical results; uncertainties concerning the toxicological significance of compounds;
and the magnitude of accumulation of the compounds in light of uncertainty around the various
estimates and assumptions regarding exposure that support the HARS-Specific Values.
Therefore, it is possible that bioaccumulation test results for dredged material proposed for use as
Remediation Material at the HARS could exceed one or more of the HARS-Specific Values
9
-------
and/or TEF Steps and still be determined to be suitable for use as Remediation Material at the
HARS. Likewise, it is possible that bioaccumulation test results for dredged material proposed
for use as Remediation Material at the HARS could be below one or more of the HARS-Specific
Values, and be determined to be unsuitable after considering the material as a whole
(i.e., combined effects).
Individual Chemical
Effect Evaluation
Remediation
No
Material
Yes
No
No
Yes
No
No
Potentially
Suitable for
Use as
Remediation
Material
Yes
Yes
Yes
Not Remediation
Material
Not
Remediation Material
Chemical X
Measurement
(2a) Is Chemical
"X" Dioxin?
Adjust to
'Steady State', it
necessary
Adjust to
'Steady State', if
necessary
[1) Is Chemical
'X" greater than
reference?
(2b) Does Chemical
"X" Exceed Dioxin
Value?
(3) In comparing chemical X to its
HARS-Specific Value, does the
weight of evidence determine
there to be a potential for
significant undesirable effects
through its bioaccumulation in
accordance with 40 CFR 227.6?
(4) Does the weight of
evidence regarding the
Combined Effects
Evaluation (i.e.,
comparison of dredged
material as a whole to
CBR, total carcinogenicity
and non-cancer hazard
index) determine there to
be a potential for
significant undesirable
effects through
bioaccumulation in
accordance with 40 CFR
227.6?
Figure 2. Proposed EPA Region 2 Framework for Evaluating Human Health Risks
Associated with Bioaccumulation Test Results
aRisk assessment of certain contaminants considers their toxicological action as a group (e.g., total PCBs, total endosulfans, total
DDT, and total benzo(a)pyrene equivalency of PAHs). Comparison of test and reference accumulations for these compounds is
proposed to be conducted with the total accumulations (or toxic equivalency), rather than with the accumulations of the
individual compounds contributing to the totals.
10
-------
A. BIO ACCUMULATION TESTING ANALYTES
Tissues of organisms used for bioaccumulation testing of sediments proposed for placement at
the HARS are currently analyzed for residue levels of nine individual metals, 15 individual
chlorinated pesticides, the 16 parent PAHs, 22 specific PCB congeners, and the seventeen
2,3,7,8-substituted dioxins and furans. These analytes were identified as "of concern" in the
New York/New Jersey Harbor system by the Toxics Workgroup of the EPA NY/NJ Harbor
Estuary Program based on a review of available data by Squibb et al. (1991). Table 1 presents
analytes currently used for testing for HARS suitability. Total PCBs in tissue residues are
estimated as two times the sum of the 22 congeners measured (see Table 1). The basis for this
method of estimating total PCBs is described in Appendix C.
Table 1. Analytes for Current HARS Testing
Acenaphthene
PCB 8
2,3,7,8,-TCDD
Acenapthylene
PCB 18
1,2,3,7,8-PeCDD
Anthracene
PCB 28
1,2,3,4,7,8-HxCDD
Fluorene
PCB 44
1,2,3,6,7,8-HxCDD
Naphthalene
PCB 49
1,2,3,7,8,9-HxCDD
Phenanthrene
PCB 52
1,2,3,4,6,7,8-HpCDD
Benzo[a]anthracene
PCB 66
OCDD
Benzo[a]pyrene
PCB 87
2,3,7,8-TCDF
Benzo[g,h,i]perylene
PCB 101
1,2,3,7,8-PeCDF
Benzo[b]fluoranthene
PCB 105
2,3,4,7,8-PeCDF
Benzo[k]fluoranthene
PCB 118
1,2,3,4,7,8-HxCDF
Chrysene
PCB 128
1,2,3,6,7,8-HxCDF
Dibenzo[a,h]anthracene
PCB 138
1,2,3,7,8,9-HxCDF
Fluoranthene
PCB 153
2,3,4,6,7,8-HxCDF
Indeno[l ,2,3-c,d]pyrene
PCB 170
1,2,3,4,6,7,8-HpCDF
Pyrene
PCB 180
1,2,3,4,7,8,9-HpCDF
Aldrin
PCB 183
OCDF
Dieldrin
PCB 184
a-Chlordane
PCB 187
trans nonachlor
PCB 195
Heptachlor
PCB 206
Heptachlor epoxide
PCB 209
Endosulfan I
1,4-dichlorobenzene
Endosulfan II
Arsenic
Endosulfan sulfate
Cadmium
4,4-DDT
Chromium
2,4-DDT
Copper
4,4-DDD
Lead
2,4-DDD
Mercury
4,4-DDE
Nickel
2,4-DDE
Silver
Zinc
11
-------
In addition to these contaminants, EPA Region 2 is considering the need to require analysis of
organo(butyl)tins and alkylated PAHs in test tissue for all dredged material proposed for use as
Remediation Material at the HARS, as discussed below.
1. Proposed Additions to Analvte List
i. Alkylated PAHs
EPA Region 2 proposes that measurement of alkylated PAHs would more accurately represent
the bioaccumulation of total PAHs by test organisms than does measurement of the parent PAH
compounds, alone. EPA Region 2 proposes to require the method that is currently approved for
analysis of the parent PAH compounds {i.e., EPA Method 8270) to be performed with specific
analytical sample cleanup procedures and analytical instrument configurations that have been
optimized to allow detection and quantitation of parent PAHs and their alkylated homologues.
The method adaptation is intended to allow quantification of approximately 30 alkyl PAH
homologues representing thousands of alkylated PAH compounds (see Table 2). A copy of this
protocol is enclosed as Appendix D.
Table 2. PAH and Alkyl PAH Target Compound List
dg-Naphthalene11
Dibenzothiophene
Naphthalene
C rDibenzothiophenes
CrNaphthalenes
C2-Dibenzothiophenes
2-Methylnaphthalene
C3-Dibenzothiophenes
1 -Methylnaphthalene
Fluoranthene
C2-Naphthalenes
Pyrene
2,6-Dimethylnaphthalene
C i-Fluoranthene/pyrenes
C3-Naphthalenes
C2-Fluoranthene/pyrenes
2,3,5-Trimethylnaphthalene
C3-Fluoranthene/pyrenes
C4-Naphthalenes
di2-Chrysenea
d10-Acenaphthene
Benz|alanthracene
Acenaphthylene
Chrysene
Acenaphthene
Ci-benzfalanthracenes/chrysenes
Biphenyl
C2-benz[alanthracenes/chrysenes
d10-Fluorene
C3-benz[alanthracenes/chrysenes
Dibenzofuran
C4-benzralanthracenes/chrysenes
Fluorene
d12-benzofalpyreneb
Cpfluorenes
Benzo[b]fluoranthene
C2-fluorenes
Benzo|k]fluoranthene
C3-fluorenes
Benzo[e]pyrene
d10-Phenanthrenea
Benzofalpyrene
Phenanthrene
Perylene
Anthracene
Indeno[l,2,3-c,
-------
Charge:
2. Is measurement of the 16 priority pollutant PAHs (i.e., parent PAHs) sufficient for
characterizing the risks associated with the total PAH bioaccumulated by organisms exposed
to dredged material proposed for placement at the HARS? Does measurement of the
alkylated compounds significantly improve risk assessment of PAHs?
3. Is the proposed adaptation of EPA Method 8270 (Appendix D) acceptable and appropriate
for regulatory decision-making? If not, what is an acceptable and appropriate method?
4. Under what specific conditions would the testing for alkylated PAHs for a particular project
be appropriate and warranted?
5. What uncertainties would be introduced within the analysis of risk should alkylated PAHs
be included? What steps could be taken to account for these uncertainties in decision-
making? Given the likelihood the method for using non-detects (as described in
EPA/CENAN (1997) will result in an overestimate of risk, what are the implications?
ii. Organotins
EPA Region 2 reviewed available data for organotins in the NY/NJ Harbor system. The data
review conducted is summarized in Appendix E. As a result of this review, EPA Region 2 is
considering the need to require analysis of organo(butyl)tins for all dredged material proposed for
use at the HARS as Remediation Material. EPA Region 2 is considering requiring use of the
analytical method of Rice et al. (1987) for the analysis of organotins. This method is
recommended in the 1991 Green Book and the Inland Testing Manual (EPA/USACE 1991,
1998).
Charge:
6. It is recognized that additional methods have been used for the analysis of organotins
(e.g., Krone et al. (1989). Will the proposed analytical method (Rice et al., 1987) provide
adequate data of sufficient quality to assess relevant risks from organotins? Is this method
the most appropriate analytical method for organotins? If not, please provide
recommendations.
7. What special QA/QC procedures should be implemented to ensure the quality and usability
of the organotin data?
8. Under what specific conditions would the testing for organotins for a particular project be
appropriate and warranted?
Hi. Coplanar PCB Congeners
EPA Region 2 proposes to require analysis of coplanar PCBs in bioaccumulation organism
tissue, if and when the decision is made to replace the existing Regional Dioxin Values with a
risk-based evaluation approach based on the Agency's reassessment of dioxin. In the event that
the EPA Region 2 elects to evaluate dioxin-like activity using a risk-based approach based on the
Agency's reassessment, project proponents would be required to analyze and report tissue
residues of those congeners having 2,3,7,8-TCDD toxic equivalency factors assigned by the
13
-------
World Health Organization (1998). The 2,3,7,8-TCDD equivalent contribution of measured
residues of these congeners would be considered with the measured residues of the seventeen
2,3,7,8- substituted dioxin/furan congeners currently measured in any future risk-based
evaluation of dioxin-like activity.
Charge:
9. If the approach for evaluating dioxin is modified, should it include the contribution of
coplanar PCBs with dioxin-like activity as proposed? If so, how?
B. COMPARISON TO REFERENCE
Concentrations of contaminants in tissues of organisms exposed for 28 days to project sediments
are compared to concentrations in tissues of organisms exposed for 28 days to a reference
sediment. The reference sediment used for testing of proposed Remediation Material for the
HARS is collected in an area of clean, sandy sediments located in the New York Bight near the
HARS, where the sediments are unaffected by dredged material disposal activities (see Figure 1
in Section I. B. of this document). Statistical comparisons to reference treatments are made using
the methods recommended in the Green Book. In this analysis, tissue residues that are reported as
"< detection level" are assigned discrete values pursuant to joint EPA Region 2/CENAN policy
described in EPA/CENAN (1997). This policy dictates that undetected residues will be
estimated at half the detection limit if the reported limit of detection is below the method
detection limit. If the limit of detection achieved by the laboratory is higher than the method
detection limit, the residue will be treated as zero (in reference organisms) or as the reported
detection limit (in test organisms) in the statistical comparison of the treatments.
When bioaccumulation of a given contaminant by test organisms exposed to project sediments is
not statistically greater than bioaccumulation by test organisms exposed to HARS reference
sediments, this means that placement of the material would not result in bioaccumulation above
that found to occur in the reference sediment. [Human health risk assessment of certain
contaminants considers their toxicological action as a group (e.g., total PCBs, total DDT, total
endosulfans, and total benzo(a)pyrene equivalency of PAHs). Statistical comparison of test and
reference accumulations for these contaminants would be conducted with the total
accumulations, rather than with accumulations of the individual compounds contributing to the
totals (e.g., statistical analysis of PCB accumulation would be performed with the total residues
in each replicate rather than with individual congeners)] Accordingly, EPA Region 2/CENAN
conclude that placement of such material will not result in bioaccumulation of that contaminant
to levels that would cause unreasonable degradation of the environment or human health, or
significant adverse effects.
In cases where bioaccumulation levels are statistically greater (at the 95 percent confidence limit)
than in the reference, further evaluation of the potential for effects is conducted. The potential
for significant undesirable effects, due to statistically significant bioaccumulation of
contaminants, is evaluated by the comparisons outlined in Figure 3 (and discussed below).
14
-------
(Evaluation of contaminant residues reported as "< detection limit" is conducted using values
assigned pursuant to the policy outlined above).
Charge:
10. Please consider the policy for assigning values to tissue residues that are reported as "<
detection limit" (described in EPA/CENAN, 1997) as you review the proposed evaluation
methodologies. As you deem appropriate, please comment on the effects of this policy on
the outlined evaluations. If the current approach is considered inappropriate, what would be
a technically supportable alternative approach for evaluating tissue residues reported as
"< detection limit" in the risk assessment process?
11. Is the use of functional groupings in statistical comparisons to reference appropriate and/or
preferable to statistical comparisons using individual contaminants for the purposes of risk
analysis?
C. COMPARISON TO REGIONAL DIOXIN VALUES
Regional dioxin values are intended for comparison to 28-day bioaccumulation test results, and
the source and use of the values are described in the EPA (1997e). Twenty-eight day
bioaccumulation test results that equal or exceed the regional Category I dioxin value of 1 pptr
for 2,3,7,8-TCDD (or that exceed 4.5 pptr using the TEQ approach described in EPA (1997e) for
the non 2,3,7,8-TCDD congeners) indicate that the sediment is not Remediation Material under
EPA Region 2/CENAN guidance. Until EPA completes its dioxin reassessment, EPA Region 2
will not consider changing this evaluation policy.
D. ADJUSTMENT TO STEADY STATE
Long-term exposure studies have shown that 28-day exposures to bedded sediments may not be
sufficient for test organism tissue concentrations of all contaminants to attain 'steady state'
(i.e., dynamic equilibrium) with sediment concentrations to which the organisms are exposed.
Therefore, 28-day tests will underestimate the residues of certain compounds in organisms
exposed to dredged material in the field following placement at the HARS.
Proposed HARS-Specific Values (except dioxins and furans, which are evaluated by directly
comparing the 28-day tissue level to the Regional dioxin values) are intended for evaluating risk
associated with residues anticipated in organisms exposed to dredged material following
placement at the HARS (i.e., 'steady state'). Therefore, consideration must be given to whether
the 28-day test results represent 'steady state' levels. The proposed HARS-Specific Values are
provided in Appendix F and their derivation is discussed below in Section E.
Rather than require long-term exposures in dredged material bioassays to allow for attainment of
'steady state', measured 28-day body burdens of contaminants are proposed to be adjusted by a
multiplier that is based on the proportion of the steady state residue attained in 28 days.
15
-------
1. Steady State Multipliers for Organic Compounds
Multipliers for estimating 'steady state' concentrations of nonpolar organic compounds can be
derived theoretically using the log KoW of the specific compound and the relationship reported in
Figure 6-1 of the joint US ACE/EPA dredged material testing manual entitled "Evaluation of
Dredged Material Proposed for Discharge in Waters of the U.S. - Testing Manual. Inland
Testing Manual" (EPA/USACE, 1998). Alternatively, multipliers can be derived for certain
organic compounds from long-term exposure studies in which 28-day and long-term residues
were reported for the same sediment.
EPA Region 2 proposes to continue the current process of applying multipliers that are based on
the log KoW of the compound and the relationship reported in Figure 6-1 of EPA/USACE (1998)
to reported 28-day concentrations of organic compounds, unless data are available from long-
term exposure studies.
Organic compounds for which 'steady state' multipliers were derived using Figure 6-1 of
EPA/USACE (1998) include aldrin, chlordane derivatives (a-chlordane, frans-nonachlor,
heptachlor, heptachlor epoxide), and endosulfans (I, El, sulfate) and PAHs. Reported 28-day
residues of alkylated PAHs are proposed to be adjusted by the same multiplier as its parent
(unalkylated) homologue. Values recommended by Karickoff and Long (1995) were used to
estimate the log K<,w s for use in estimating the 'steady state' multiplier for each compound. The
log Kow, proportion of steady state (Pss) predicted by Figure 6-1, and the multiplier proposed for
use in evaluating these compounds are presented in Table 3.
Empirically derived multipliers from long-term exposure studies were used to estimate 'steady
state' of PCBs (Pruell et al., 1990), dieldrin (Lee et al., 1994), and DDT and its derivatives
(i.e., 2,4- and 4,4-DDD, DDE, and DDT) (Lee et al., 1994). The proportion of 'steady state' that
was attained after 28 days of exposure and the multiplier proposed for use in evaluating risks of
these compounds is presented in Table 4. The proportions of 'steady state' predicted by
Figure 6-1 for the DDT derivatives and dieldrin are also provided for reference in Table 4.
Where empirical 'steady state' multipliers were not reported for N. virens, the multiplier derived
using M. nasuta data was used for both organisms.
16
-------
Table 3. 'Steady state' Multipliers for Organic Compounds Derived from Figure 6-1 of
EPA/USACE (1998) Using log Kows Reported by Karickoff and Long (1995)
Contaminant
log K„w
Proportion of 'steady state' predicted by
Fig. 6-1 after 28 days exposure
Proposed
Multiplier
aldrin
6.50
0.37
3
a-chlordane
6.32
0.38
3
frans-nonachlor
6.87
0.33
3
heptachlor
6.26
0.42
3
heptachlor epoxide
5.00
0.93
total endosulfans
4.10
1.0
1
napthalene
3.36
1.0
1
acenapthene
3.92
1.0
1
acenaphthylene
4.08
1.0
1
fluorene
4.21
1.0
1
phenanthrene
4.55
1.0
1
anthracene
4.55
1.0
1
pyrene
5.11
0.9
1
fluoranthene
5.12
0.9
1
chrysene
5.70
0.62
2
benz[a]anthracene
5.70
0.62
2
benzo[a]pyrene
6.11
0.47
2
dibenz[a,h]anthracene
6.69
0.36
3
benzo[b+k]fluoranthene
6.20
0.44
3
benzo[g,h,i]perylene
6.70
0.36
3
indeno[l ,2,3-c,d]pyrene
6.65
0.36
3
17
-------
Table 4. 'Steady state' Multipliers for Organic Compounds Empirically Derived from Long-term
Exposure Studies Using Dredged Material Test Organisms (M. nasuta and N. virens)
Contaminant
P»
(Fig. 6-1)
Proportion of 'steady state'
attained after 28 days
Reference
Proposed
Multiplier
(Species used in assay)
o, p'-DDD (M. nasuta)
0.48
0.55
Lee et al. (1994)
2
p, p'-DDD (M nasuta)
0.48
0.33
Lee et al. (1994)
3
o, p'-DDE (M nasuta)
0.35
0.42
Lee et al. (1994)
3
p, p' -DDE (M nasuta)
0.35
0.49
Lee et al. (1994)
2
o, p'-DDT (M nasuta)
0.36
0.55
Lee et al. (1994)
2
p, p'-DDT (M nasuta)
0.36
0.09
Lee et al. (1994)
11
dieldrin (M. nasuta)
0.83
0.58
Lee et al. (1994)
2
total PCBs (N. virens)
N/A
0.58
Pruell et al. (1990)
2
total PCBs (M. nasuta)
N/A
0.81
Pruell et al. (1990)
1
Charge:
Please review the tables of organic compounds and for each individual compound comment on
the following:
12. Is it appropriate to apply a multiplier based on log KoW for these compounds (organics), or
are there other specific data that can be used to estimate steady state? If so, please identify.
13. Given the increased hydrophobicity of alkylated PAHs, is the use of the correction factor
associated with the corresponding parent an appropriate approach for estimating steady state
residues of alkylated PAHs? If not, please elaborate.
14. For the DDT derivatives and dieldrin, please comment on the appropriateness of using M.
nasuta data rather than N. virens-specific data in the estimation of steady state multipliers.
15. Are the approaches taken to adjust organic contaminant bioaccumulation data to steady state
adequate? Do the proposed multipliers agree with previously published studies (i.e., do they
appear reasonable)? If not, please elaborate.
16. What are the major sources of uncertainty associated with the approaches? What alternative
approaches would reduce the uncertainties? How could these uncertainties be described and
accounted for in decision-making?
18
-------
2. Steady State for Metals (Safety Factors')
Application of steady state correction factors as defined for the organic compounds is not
appropriate for metals, because a "true" steady state for metals does not appear to exist.
Therefore, EPA Region 2 has elected to derive a safety (uncertainty) factor to apply to 28-day
bioaccumulation test results for non-essential metals (i.e., silver, cadmium, mercury, and lead) in
the proposed HARS Framework. Safety factors are only proposed for application 28-day test
results for non-essential metals because there is less evidence for regulation of non-essential
metals by exposed marine organisms, and because of the higher level of human health and
ecological concern associated with non-essential (compared to essential) metals.
To develop an appropriate safety factor for application to 28-day test results for non-essential
metals, EPA Region 2 compiled and compared sediment concentrations and polychaete tissue
concentrations in co-located samples from the vicinity of the HARS, reported by Battelle (1996,
1997). EPA Region 2 assumed that metals concentrations measured in the organisms collected
in this effort represent a range of exposure durations and conditions that are typical of benthic
organisms at the HARS. The results of this analysis indicate that despite sediment
concentrations which varied by as much as two orders of magnitude, tissue concentrations of
non-essential metals in field collected benthic organisms varied within a factor of three
(i.e., maximum reported concentrations of all non-essential metals were approximately three
times higher than the lowest concentration reported). The data is available in Appendix G.
Therefore, EPA Region 2 is proposing to apply a safety factor of three to bioaccumulation test
results for non-essential metals.
Charge:
17. In your opinion, is the methodology (see Appendix G) followed to derive the steady state
multiplier for non-essential metals (i.e., a factor of three) scientifically appropriate? Please
elaborate. Do you have any recommendations of additional or alternate methodologies or
information that can be used to either supplement or replace the proposed method?
E. HUMAN HEALTH EVALUATIONS
Exposure of humans to contaminants in exposed sediments of the HARS (or in dredged material
that is proposed for placement as Remediation Material) is assumed primarily to occur via
consumption of contaminated seafood. Uptake of HARS contaminants by marine organisms was
assumed to occur through direct exposure of biota to the sediments and/or through trophic
transfer from contaminated prey. Because the HARS is located offshore and in deep water, and
because data shows that suspended and dissolved constituents of dredged material do not persist
in the water column following release from the barge, pathways of human exposure other than
consumption of seafood (e.g., inhalation, or direct exposure through bathing) were not
considered in assessing associated risks.
19
-------
For the proposed process for assessing risks to ecological receptors (and in support of assessing
risks to human consumers through the food web), a simplified description of the food web is
used to describe trophic relationships between species anticipated to be present at the HARS.
The New York Bight food web used in modeling trophic transfer of contaminants was described
by a simplified food chain consisting of three representative trophic levels. These trophic levels
were described in EPA (1995a) as: benthic organisms, benthic predators, and upper level
predators.
The proposed process for evaluating risks associated with contaminants in dredged material
proposed for use at the HARS assumes that recreational anglers represent a reasonably
maximally exposed (RME) population for assessing risks to humans. More explicitly stated,
EPA Region 2 assumes that there is a subpopulation of anglers that fishes exclusively at the
HARS and that these anglers consume similar quantities of recreationally-caught as the average
New Jersey angler (as represented by respondents to the NJMSC (1994) survey, see Appendix I).
The assessment further assumes that all recreationally caught fish reportedly consumed by this
hypothetical subpopulation of anglers are obtained by angling at the HARS. In addition, the
assessment assumes that fish are filleted prior to cooking (and consumption).
To determine whether a tested sediment would result in bioaccumulation that would cause
significant undesirable effects with regard to human health, standard human health risk
calculations were used to develop HARS-specific Values and combined effects evaluations for
cancer risk and noncancer effects. The basic risk assessment equations underlying all
calculations presented hereafter are as follows:
Cancer risk level = MV x CPF x RPF x FIR x CF x EF x ED x TTF
BWxATxBFR
Noncancer hazard quotient = MV x FIR x CF x EF x ED x TTF
RfD x AT x BFR x BW
where:
MV - Measured tissue value (mg/kg)
CPF - Cancer potency factor (Kg-day/mg)
RPF - Relative Benzo(a)pyrene Potency Factor [used only for carcinogenic PAHs] (unitless)
FIR - Fish Ingestion Rate (g/day)
CF - Conversion factor (kg/g)
EF - Exposure frequency (365 days/year)
ED - Exposure duration (70 years)
TTF - Trophic transfer factor (unitless)
BW - Body weight (70 Kg)
AT - Averaging time (25,550 days)
BFR - Whole body to fillet ratio (unitless)
RfD - Reference dose (mg/Kg-day).
For the purpose of this evaluation, it was assumed that fish consumption is the pathway of
concern for humans to be exposed to contaminants in dredged material proposed for use as
20
-------
Remediation Material at the HARS, and that the fish consumed would be exposed through
trophic transfer of contaminants from benthic invertebrate prey. The following specific guideline
measures and assumptions were applied to all human health risk/effects evaluations to estimate
human exposure to HARS contaminants (Summary also provided in Table 5).
• Cancer Potency Factor/Reference Dose- Available cancer potency factors and chronic
reference doses for oral exposure were obtained from the EPA Integrated Risk
Information System (IRIS) for each contaminant, except lead. The RfD for lead was
withdrawn from IRIS due to the lack of an established toxicity threshold for neurological
effects in children. In the current and proposed evaluation processes, the toxicity
assessment for lead has been refined through use of a biomarker (i.e., blood lead
concentration) that serves as both a marker of lead exposure and effect. The approach
used to derive a HARS-specific value for lead is described in Appendix H. In the absence
of available toxicological values (e.g., cancer slope factor, RfD) for the specific alkyl
PAH homologues, the toxicological values for the associated parent compound will be
used as a surrogate for the potency of the alkyl group.
• Seafood consumption - A factor of 7.2 grams per day (g/day) is proposed for use as a site-
specific estimate of daily fish consumption by high consumers (i.e., New Jersey
recreational anglers) in the vicinity of the HARS. Further information regarding the use
of data from the NJMSC (1994) study is provided in Appendix I.
• Exposure Duration: EPA Region 2 proposes to employ a default exposure duration of 70
years for its assessment of human health risks.
• Site Use Factor - A factor to express the proportion of time that fish predators may be
exposed to contaminated benthic prey residing at the HARS. A factor of 0.777
(i.e., 77.7 percent HARS-area foraging), was derived to estimate site use for a "generic"
fish in the diet of the target sub-population (i.e., New Jersey recreational fishers). Further
information regarding site use factor is provided in Appendix J.
• Whole-body to fillet factor - In assessing risks due to lipophilic, organic compounds and
selected metals, EPA Region 2 proposes to employ a whole-body to fillet correction
factor to estimate the concentration of contaminant in the whole body of the fish that is
associated with toxicological dose in the edible (fillet) portion of the fish. Further
information regarding the whole body to fillet factors is provided in Appendix K.
• Trophic transfer factor - Trophic transfer of contaminants from benthic prey to fish
predators was estimated by applying a discrete factor that expresses the ratio of the
residue concentration in predator as a function of the residue concentration in prey.
Further information regarding the trophic transfer factor is provided in Appendix L.
The nine metals that are currently analyzed in the EPA Region 2/CENAN dredged material
evaluation program are measured in bioaccumulation test organism tissue (and reported) as total
residue for each metal. With the exception of arsenic, the total metal residue is used in the
assessment of risks associated with each metal. Risk assessment of total metal residues in test
21
-------
organisms is proposed to be performed by applying the metals potency factors {i.e., cancer slope
factors and reference doses) and exposure factors (such as trophic transfer) associated with the
most toxicologically significant individual valence species or form of the metal that might be
present. The risk assessment of chromium and mercury is therefore conducted assuming that the
measured concentration {i.e., total chromium and total mercury) is present in its most toxic
and/or efficiently transferred form (i.e., hexavalent chromium and methylmercury). Proponents
of dredging projects with total chromium or mercury bioaccumulation test results that exceed
their respective HARS-Specific Values will be offered an opportunity to provide analytical
speciation data for the metal exceeding the value in order to reduce uncertainty in the risk
assessment.
The residue of total arsenic reported in bioaccumulation test organism tissue is proposed to be
adjusted by application of a 0.1 multiplier, pursuant to U.S. FDA guidance (see USFDA, 1993).
This multiplier is applied to reflect the typical proportion of total arsenic that exists in marine
benthic invertebrates as inorganic arsenic, the most toxicologically important forms of arsenic.
This estimated inorganic arsenic residue is then used with toxicological potency factors
associated specifically with inorganic arsenic in the risk assessment.
Risk assessment of certain organic compounds is conducted by considering the risks/hazards
associated with the total residue of all contaminants within the specific group of compounds to
which they belong. Specifically, human health risk/hazard for endosulfans (including I, II, and
sulfate), DDT derivatives (including p,p' and o,p' stereoisomers of DDD, DDE, and DDT), and
butyl tins (mono-, di-, tri- and tetra butyl tins) is assessed as total steady state residue of the
group using potency and/or exposure factors associated with the individual contaminant of
greatest potency in each group.
22
-------
Charge:
18. Please comment on each factor listed above (and in Table 5) as to its appropriateness for use
in the equations listed above. Would you recommend additional factors? Would you
change or modify the equations as written above? If so, how?
19. Are the methods used to derive the human health exposure parameters and assigned values
discussed in Section E appropriate (please review the referenced appendices)? If not, please
elaborate. How should these factors be factored into the risk analyses and decision-making?
20. Is the approach taken to relate fish whole body and fillet concentrations scientifically
appropriate? If not, what method would you recommend?
21. Could the analysis be improved by focusing on a smaller number of key fish (seafood)
species at the HARS? What characteristics should be used to select these key species?
22. In your opinion, is the approach for assuming total metal to be in the most toxic form
appropriate and reasonable? Should metals speciation/complexation be considered in the
assessment of metals bioaccumulation, trophic transfer, and human health risks? Are there
alternative analytical or risk assessment techniques available that would improve the risk
assessment of metals? Is the multiplier proposed for adjusting measured concentrations of
arsenic appropriate and reasonable?
23 . Is the assumption that the potency of alkylated PAHs can be estimated by the potency of the
parent PAH appropriate? Is this assumption likely to result in an under- or overestimate of
the risk associated with the alkylated PAHs?
24. Please comment on the potential for human exposure to PAHs through consumption of
finfish and other seafood.
25. What are the major sources of uncertainty associated with the approaches described in
Section E? What alternative approaches would reduce the uncertainties? How could these
uncertainties be described and accounted for in decision-making?
26. What is your recommendation for evaluating the potential toxicity of organotins? Should
they be evaluated as individual compounds? Summed as total? Should there be some
consideration of relative toxicity?
27. Please comment on the appropriateness of the proposed approach for converting and using
the analytical data for alkylated and parent PAHs to estimate risk from all PAHs.
23
-------
Table 5. Assumptions Used to Develop Human Health HARS-Specific Values
Compound
Cancer Potency
Factor
Reference
Dose
Trophic
Transfer
Whole
body: Filet
Seafood
Consumption
(g/day)
Site Use
Factor
PAHs
Acenaphthene
60
0.1
1.35
7.2
0.777
Anthracene
300
0.1
1.35
7.2
0.777
Fluorene
40
0.1
1.35
7.2
0.777
Naphthalene
20
0.1
1.35
7.2
0.777
Phenanthrene
300
0.1
1.35
7.2
0.777
Benzo(a)pyrene
7.3
0.1
1.35
7.2
0.777
Benzo(a)anthracene
0.1
1.35
7.2
0.777
Benzo(b)fluoranthene
0.1
1.35
7.2
0.777
Benzo(k)fluoranthene
0.1
1.35
7.2
0.777
Chrysene
0.1
1.35
7.2
0.777
Dibenzo(a,h)anthracene
0.1
1.35
7.2
0.777
Indeno(l ,2,3-cd)pyrene
0.1
1.35
7.2
0.777
Fluoranthene
40
0.1
1.35
7.2
0.777
Pyrene
30
0.1
1.35
7.2
0.777
PESTICIDES
Aldrin
17
0.03
3
1.35
7.2
0.777
Dieldrin
16
0.05
1.6
1.35
7.2
0.777
aChlordane
0.35
0.05
2.9
1.35
7.2
0.777
Heptachlor
4.5
0.5
2.9
1.35
7.2
0.777
Heptachlor epoxide
9.1
0.013
1.4
1.35
7.2
0.777
Total Endosulfans
6
1.1
1.35
7.2
0.777
Total DDT
0.34
0.5
3
1.35
7.2
0.777
TOTAL PCBs
2
0.02
3
1.35
7.2
0.777
1,4-Dichlorobenzene
0.024
30
1
1.35
7.2
0.777
METALS
Arsenic
1.5
0.3
0.25
1.4
7.2
0.777
Cadmium
1
0.25
1
7.2
0.777
Chromium (total)
3
1
1.2
7.2
0.777
Copper
37.1
0.21
1
7.2
0.777
Lead
1
0.23
1
7.2
0.777
Mercury
0.1
6.2
0.7
7.2
0.777
Nickel
20
1
1
7.2
0.777
Silver
5
1
1
7.2
0.777
Zinc
300
0.24
1
7.2
0.777
Tributyltin
NA
0.3
1.02
1.35
7.2
0.777
24
-------
1. Comparison of Individual Constituents to HARS-Specific Values
Except for dioxin-like action, risk-based or effects-based end point guideline values will be used
in evaluating bioaccumulation test results. These guideline values, or HARS-specific Values, are
developed as deterministic guidelines using standard risk assessment calculations. Figure 3
presents the standard calculations for deriving the HARS-Specific Values for cancer and non-
cancer endpoints. The derivations of cancer and non-cancer guidelines differ only in the way
toxicological dose is calculated (Equation 1).
Eauation 1 (for Cancer Risk)
Toxicological Dose (ng/day) =
fTareet Risk Levell x TBodv Weight (70 keVI x MO3 ue/mel
[Cancer Potency Factor (kg-day/mg)]
Eauation 1 (for Non-Cancer Hazard)
Toxicological Dose (ng/day) =
[Reference dose (mg/kg-day)] x [Body Weight (70 kg)] x [103 ng/mg]
Eauation 2
Cone, in Edible Fish Tissue (ng/kg) =
fToxicoloaical Dose (ue/davYI
[Seafood Cons. (7.2 g/day)] x [10"3kg/g]
Eauation 3
HARS-Specific Value (ng/kg) =
rConc. in Edible Fish Tissuel x 1 Whole Bodv/Fillet Factorl
[Trophic Transfer Factor]x [Site Use Factor (0.777)]
Figure 3. Calculation of HARS-Specific Values for Protection of Human Health
The HARS-Specific Values represent guidelines for evaluating the human health risk/hazard
associated with residues of contaminants in lower trophic level (benthic) organisms. The HARS-
Specific Values are back calculated from tissue concentrations in upper trophic level organisms
(i.e., fish) that correspond with specific human health risk/ hazard levels.
The HARS-specific value associated with the target cancer risk level for PAHs was derived as a
guidance level intended for comparison to the total benzo(a)pyrene carcinogenic equivalency of
the mixture. Specifically, the HARS-specific value was derived using the cancer slope factor for
benzo(a)pyrene. For purposes of comparison to this HARS-specific value the residue of each
PAH (including alkylated PAHs) measured in test organism tissues will be: (1) converted to
molar concentration using its molecular weight deriving this guidance level, and then
(2) expressed as benzo(a)pyrene equivalent using the specific "estimated order of potential
potency" listed for the PAH in EPA (1993). In the absence of available "estimated orders of
potential potency" for the specific alkyl PAH homologues, the toxicological values for the
associated parent compound will be used as a surrogate estimate for the potency of the alkyl
group. The benzo(a)pyrene equivalent molar concentrations will then be summed for all PAHs
and compared to the HARS-specific value.
25
-------
For lead, the potential for non-carcinogenic effects associated with the test tissue was estimated
using a "disaggregate" modeling approach, (which relates multi-media lead exposure to blood
lead concentration) employed in the 1986 EPA Air Quality Document (see Figure 4). The
calculations yield lead concentrations representing levels below which the concentrations do not
indicate a potential for significant undesirable effects {i.e., 95 percent of the blood lead
probability distribution below 10 |ig/dl). Further information on the derivation of the lead
guidance value is provided as Appendix H.
Charge:
28. Do you believe that the "disaggregate" modeling discussed above (and shown in Figure 4)
for estimating human health HARS-Specific Values for lead is appropriate? Would you
recommend an alternative risk assessment method be used given the information and data
available? Do you believe the method described has appropriately taken uncertainty into
account? Please elaborate.
26
-------
1 - Calculate current human exposure from all routes
A - Average current exposures
Drinking water
4 ppb
Soil/dust/paint
800 ppm
Air
0.1 ug/cuM
Dietary
5.5 ug/da
B - Media-specific blood lead coefficients
Drinking water
0.16 ug/dlperppb
Soil/dust/paint
0.002 ug/dlperppm
Air
2 ug/dl per ug/cuM
Dietary
0.16 ug/dl per ug/da
Calculate Blood lead contributions [A x B]
Drinking water
0.64 ug/dl
Soil/dust/paint
1.6 ug/dl
Air
0.2 ug/dl
Dietary
0.9 ug/dl
C - Sum of blood lead contributions to determine all-route current exposure [Sum(A x B)|
Current exposure
3.3 ug/dl
2 - Calculate acceptable lead contribution from fish spending time at MDS [Level of concern - current
exposure]
Level of concern
4.6 ug/dl
Current exposure
3.3 ug/dl
Acceptable HARS fish contribution
1.3 ug/dl
3 - Convert acceptable fish contribution to lead fish tissue concentration
A - Calculate acceptable daily intake [Acceptable fish contribution / food lead coefficient]
Acceptable fish contribution
1.3 ug/dl
Food lead coefficient
0.16 ug/dl per da
Acceptable daily intake
8.125 ug/da
B - Calculate acceptable tissue concentration in fish [Acceptable intake/Average fish consumption]
Average fish consumption
7.2 g/da
Acceptable tissue cone
1.1285 ug/g
C - Calculate acceptable benthic tissue concentration from acceptable concentration in fish
Whole-body: fillet
1
After adjustment for fillet
1.1285 ug/g
Site use percent
77.7 percent
After adjustment for site use
1.4523 ug/g
Trophic transfer factor
0.23
After adjustment for trophic transfer
6.3 ug/g
or
HARS-SPECIFIC VALUE
6.3 ppm
Figure 4. HARS-Speciflc Value Calculations for Non-Cancer Effects from Lead
27
-------
2. Consideration of Combined Risk of Contaminants
The evaluation of the potential for human health effects due to the combined action of multiple
contaminants (total cancer risk and combined hazard quotients) is proposed to be conducted by:
(1) using the same formulas (see Figure 5) and assumptions (see Table 5) in a conventional
forward calculation to estimate the cancer risk and noncancer hazard quotient associated with the
level of each contaminant accumulated by benthic organisms; and then (2) adding the products of
these calculations for the individual contaminants (i.e., cancer risk and hazard quotient) that
share modes of toxic action (e.g., carcinogens) or sites of toxic action (i.e., non cancer effects on
specific organ systems).
i. Combined Effects Evaluation: Total Carcinogenicity
Cancer risks associated with accumulated levels would be calculated using standard risk
assessment equations as described in Figure 5. EPA Region 2 then proposes to assess the total
carcinogenicity of the mixture of carcinogenic compounds that are accumulated by test
organisms by summing the individual cancer risks associated with accumulated concentrations of
each carcinogenic constituent and comparing that sum to the target risk level of 10"4. The
combined cancer risk of all contaminants would then be considered as part of the weight of
evidence regarding the suitability of the proposed dredged material for use as Remediation
Material at the HARS (see Figure 2).
Eauation 1
Estimated Cone, in Fish (ng/kg) =
fMeasured Tissue Level fue/kell x ITroDhic Transfer Factorl x rSite Use Factor (0.777)1
[Whole Body/Fillet Factor]
Equation 2
Toxicological Dose (ng/day) =
[Estimated Cone, in Fish (ue/keYI x [Seafood Consi7.2 e/davYI
[103 g/kg]
Eauation 3
Estimated Cancer Risk (unitless) =
[Toxicoloeical Dose (ua/dav^l x [Cancer Potency Factor (ke-dav/mell x RPF
[Body Weight (70 kg)fn3] x [103 ng/mg]
Figure 5. Calculation of Constituent-Specific Risk for Total Carcinogenicity
ii. Combined Effects Evaluation: Non-Cancer Hazard Index
A non-cancer hazard quotient associated with accumulated levels of each contaminant that shares
a similar toxicological mode of action or target organ system with other analyzed contaminants is
proposed to be calculated using standard risk assessment equations as described in Figure 6.
EPA Region 2 then proposes to assess the total non-cancer hazard indices of contaminants with
shared target organ systems that are accumulated by test organisms by summing the individual
non-cancer hazard associated with accumulated concentrations of each contributing compound
(see EPA 1996) and comparing that sum to the target hazard index of 1.0. In this summation, the
non-cancer hazard quotients of alkylated PAHs (i.e., fluorenes and fluoranthenes/pyrenes) will be
calculated using the RfD reported for the parent compounds. The combined non-cancer hazard
indices of all contaminants sharing target organ systems are then proposed to be considered as
28
-------
part of the weight of evidence regarding the suitability of the proposed dredged material for use
as Remediation Material at the HARS.
Contaminants having shared target organ systems (i.e., circulatory system, liver, and kidney) and
therefore evaluated using this combined hazard index approach are listed in Table 6.
Equation 1
Cone, in Fish (ng/kg) =
fMeasured Tissue Level fue/keYl x UroDhic Transfer Factorl x fSite Use Factor f0.777Y|
[Whole Body/Fillet Factor]
Equation 2
= [Cone, in Fish (ng/kg)] x [Seafood Cons. (7.2 g/day)] x [10"3kg/g]
Toxicological Dose (ng/day)
Equation 3
Hazard Quotient (unitless) =
FToxicoloeical Dose (^e/davll / fBodv Weieht (70 keYI
[Reference Dose (mg/kg-day)] x [103 ng/mg]
Figure 6. Calculation of Hazard Indices Associated with Individual Constituents
Table 6. Summary of Target Systems Evaluated in Hazard Index Approach
Target System
Chemical
Circulatory
Zinc, Fluorene, Fluoranthene
Liver
Acenaphthene, Fluorene
Kidney
Cadmium, Endosulfan, Pyrene
Charge:
29. In your opinion, are the methodologies and equations described above appropriate for
estimating total carcinogenicity and combined non-cancer impacts of contaminant mixtures
accumulated from dredged material proposed for use as Remediation Material at the HARS?
30. Is the conceptual model for evaluating fish exposure to dredged material at the HARS and
human exposure through ingestion of seafood appropriate and reasonable? How can the
uncertainties associated with the assumptions in this conceptual model be reduced? Please
consider the spatial and temporal elements of exposure in your discussion.
29
-------
III. REFERENCES:
Battelle Ocean Sciences. 1996. Sediment Survey of the Mud Dump Site and Environs. Submitted
to U.S. Environmental Protection Agency, Office of Wetlands, Oceans, and Watersheds and
Region II Under EPA Contract No. 68-C2-0134. Work Assignment 2-133. Final Data Report.
Battelle Ocean Sciences. Duxbury, MA.
Battelle Ocean Sciences. 1997. Contaminants in polychaetes from the Mud Dump Site and
environs. Prepared for U.S. EPA Region 2 (Work Assignment 3-133, Contract No. 68-C2-0134),
dated March 4,1997. Various pagings.
EPA (U.S. Environmental Protection Agency). 1986. Air Quality Criteria for Lead. Research
Triangle Park, NC: Office of Research and Development, EPA 600/8-83-028F.
EPA (U.S. Environmental Protection Agency). 1993. Provisional Guidance for Quantitative Risk
Assessment of Polycyclic Aromatic Hydrocarbons. EPA/600/R-93/089. Environmental Criteria
and Assessment Office, Office of Health and Environmental Assessment. Cincinnati, OH. 20 p.
EPA (U.S. Environmental Protection Agency). 1996. Soil Screening Guidance: Technical
Background Document. EPA/540/R-95/128. Environmental Criteria and Assessment Office,
Office of Health and Environmental Assessment. Cincinnati, OH. (May 1996)
EPA (U.S. Environmental Protection Agency, Region 2). 1997a. Supplement to the
Environmental Impact Statement on the New York Dredged Material Disposal Site Designation
for the Designation of the Historic Area Remediation Site (HARS) in the New York Bight Apex.
EPA (U.S. Environmental Protection Agency, Region 2). 1997b. Proposed Rule - Simultaneous
De-designation and Termination of the Mud Dump Site and Designation of the Historic Area
Remediation Site. May, 1997. 62 Federal Register 26267.
EPA (U.S. Environmental Protection Agency, Region 2). 1997c. Final Rule - Simultaneous De-
designation and Termination of the Mud Dump Site and Designation of the Historic Area
Remediation Site. August 1997. 62 Federal Register 46142.
EPA (U.S. Environmental Protection Agency, Region 2). 1997d. Response to Comments on the
May 13, 1997, Proposed Rule for the Simultaneous De-designation and Termination of the Mud
Dump Site (MDS) and Designation of the Historic Area Remediation Site (HARS). 369 p.
EPA (U.S. Environmental Protection Agency). 1997e. to File from A. Lechich. Re: Summary of
Dioxin Risk Evaluation Approach. March 15, 1997.
30
-------
EPA (U.S. Environmental Protection Agency). 1997d. Ecological Risk Assessment Guidance for
Superfund: Process for Designing and Conducting Ecological Risk Assessments. Interim Final
Report. EPA 540-R-97-006. United States Environmental Protection Agency, Office of Solid
Waste and Emergency Response. Washington, DC.
EPA (U.S. Environmental Protection Agency). 2000. "Integrated Risk Information System
(IRIS)". National Center for Environmental Assessment, Cincinnati, OH.
http://www.epa.gov/iris. Accessed Feb. 10,2000.
EPA/CENAN (U.S. Environmental Protection Agency, Region 2/ U.S. Army Corps of
Engineers, New York District) 1992. Guidance for Performing Tests on Dredged Material
Proposed for Ocean Disposal. New York District Corps of Engineers, U.S. Environmental
Protection Agency -Region 2.
EPA/CENAN (U.S. Environmental Protection Agency, Region 2/ U.S. Army Corps of
Engineers, New York District) 1997. (Joint Memorandum) Ocean Disposal of Dredged Material
Clarification of Two Procedural Elements of Interagency Coordination Between USEPA Region
2 and the New York District, USACE-Treatment of Non-Detects, Chemical Data, and Rules and
Responsibilities in Preparation of Ocean Disposal Regulatory Compliance Memorandum.
EPA/USACE (U.S. Environmental Protection Agency/ U.S. Army Corps of Engineers). 1991.
Evaluation of Dredged Material Proposed for Ocean Disposal - Testing Manual. (Green Book).
EPA- 503/8-91/001.
EPA/USACE (U.S. Environmental Protection Agency/ U.S. Army Corps of Engineers). 1998.
Evaluation of Dredged Material Proposed for Discharge in Waters of the U.S. - Testing Manual.
EPA-823-B-98-004, Washington, D.C.
FDA. (U.S. Food and Drug Administration). 1993. Guidance Document for Arsenic in
Shellfish. Center for Food Safety and Applied Nutrition, Food and Drug Administration.
Washington, DC. 31 p.
Karickhoff, S.W. and J.M. Long. 1995. Internal Report on Summary of Measured, Calculated,
and Recommended Log KoW Values. Prepared for E. Southerland, EPA Office of Water. Dated
April 10. 40 pgs.
Lee, H. n, A. Lincoff, B.L. Boese, F.A. Cole, S.P. Ferraro, J.O. Lamberson, R.J. Ozretich, R.C.
Randall, K.R. Rukavina, D.W. Schults, K.A. Sercu, D.T. Specht, R.C. Swartz, and D.R. Young.
1994. Ecological Risk Assessment of the Marine Sediments at the United Heckathorn Superfund
Site. ERL-N: N269. U.S.EPA , ERL - Narragansett. Newport OR. various pagings.
31
-------
NJMSC (New Jersey Marine Sciences Consortium). 1994. Fish Consumption Patterns by New
Jersey Consumers and Anglers. Prepared for New Jersey Dept. of Environmental Protection and
Energy, Division of Science and Research Under Contract No. P 30695. New Jersey Dept. of
Environmental Protection and Energy, Division of Science and Research. Trenton, NJ. Various
pagings.
Pruell, R.J., N.I. Rubinstein, B.K. Taplin, J.A. LiVolsi, and C.B. Norwood. 1990. 2,3,7,8-TCDD,
2,3,7,8-TCDF and PCBs in marine sediments and biota: Laboratory and field studies. Final
Report to: U.S. Army Corps of Engineers - New York District. Dated March 12,1990. 74 pp.
Rice, C., F. Espourteille and R. Huggett. 1987. A method for analysis of tributyltin in in
estuarial sediments and oyster tissue, Crassostrea virginica. Appl. Organomet. Chem. 1:541-
544.
Squibb, K.S., J.M. O'Connor, and Kneip, T.J. 1991. Toxics Characterization Report,
Module 3.1. Report prepared by Institute of Environmental Medicine, NY Univ. Medical Center
for the NY/NJ Harbor Estuary Program.
32
-------
I>
o
D
3
ex.
APPENDICES
-------
APPENDIX A
Definition of Remediation Material and Summary of Data
and Rationale Supporting the Need for Remediation of the
HARS
-------
Appendix A
DEFINITION OF REMEDIATION MATERIAL AND SUMMARY OF
DATA AND RATIONALE SUPPORTING THE NEED FOR
REMEDIATION OF THE HARS
The data and rationale supporting the need for remediating the HARS is provided in the SEIS
(EPA, 1997a) and the introductory section of EPA's Response to Comments received on the
proposed rule (EPA, 1997d). Relevant sections of that section are provided below.
I. INTRODUCTION
The U.S. Environmental Protection Agency (EPA) proposed on May 13, 1997, to de-designate
and terminate the New York Bight Dredged Material Disposal Site (also known as the Mud
Dump Site (MDS)). The MDS was designated in 1984 for the disposal of up to 100 million
cubic yards of dredged material from navigational dredging and other dredging projects
associated with the Port of New York and New Jersey and nearby harbors. Simultaneous with
closure of the MDS, the site and surrounding areas that have been used historically as disposal
sites for dredged materials would be redesignated under 40 CFR Part 228 as the Historic Area
Remediation Site (HARS). The HARS will be managed to reduce impacts of historical disposal
activities at the site to acceptable levels (in accordance with 40 CFR Section 228.11 (c)). The
proposed amendment would, when finalized, identify for remediation an area in and around the
MDS which has exhibited the potential for adverse ecological impacts. As discussed further
below, the HARS would be remediated with uncontaminated dredged material (i.e., dredged
material that meets current Category I standards and will not cause significant undesirable effects
including through bioaccumulation) (hereinafter referred to as "the Material for Remediation" or
"Remediation Material").
II. SUMMARY
Need for Remediation
As discussed and documented in the HARS SEIS accompanying the proposed rule, field studies
of the New York Bight Apex have found undesirable levels of bioaccumulative contaminants
and toxicity in the surface sediments of much of the MDS and in sediments immediately
surrounding the MDS. Further, it was found that some of these sediments cause toxicity in
amphipod bioassays. Amphipods are small-bodied crustaceans that live in the surface layers of
sediment, and are important prey items for many coastal marine organisms. These and other
organisms are used by EPA Region 2 and the USACE-NYD (and in many other places around
the country) to evaluate sediment samples from proposed dredging sites.
While it is impossible to quantify how much of New York Bight Apex contamination is the
direct result of past dredged material disposal, other ocean dumping activities (e.g., former
sewage sludge disposal at the 12-Mile Site), or other sources (e.g., via Hudson River plume or
atmospheric deposition), the presence of these degraded sediments in the Apex is cause for
33
-------
Appendix A
concern. Organisms living in or near these degraded surface sediments in nearshore waters will
be continually exposed to contaminants until the contaminants are buried by natural
sedimentation, placement of Remediation Material, or otherwise isolated or removed. Exposed
sediments can directly and indirectly impact benthic and pelagic organisms. Impacts to
terrestrial organisms (including human beings) are also possible if the contaminants were to
undergo trophic transfer.
EPA employed several types of evaluations to determine the extent and location of potential
environmental impacts in the vicinity of the MDS and historic dredged material disposal areas.
These included the type of amphipod bioassays normally conducted on sediment samples from
proposed dredging sites, contaminant-bioaccumulation evaluations of infaunal organisms and
sediment from the Study Area (a 30 square nautical mile area within the New York Bight Apex
encompassing benthic areas that showed evidence of dredged material disposal (presence of
craters and mounds)), and evaluation of the benthic community structure in the potentially
impacted areas. The results of these evaluations and the main factors that make remediation
necessary are summarized below.
Contaminant Toxicity
Sediment toxicity was evaluated using the same 10-day amphipod (Ampelisca abdita) bioassay
test used as part of the evaluation of the suitability of sediment for ocean disposal by EPA
Region 2 and the USACE-NYD. The data from amphipod bioassays of sediments from 1994
Study Area samples indicated widespread toxic conditions in sediment from areas around the
MDS. If these surface sediments from the Study Area were from a proposed EPA Region
2/USACE-NYD dredging project site, the sediments would have been categorized as Category
III and found to not meet the limiting permissible concentration (LPC) in EPA's Ocean Dumping
Regulations (40 CFR Section 227.27), and thus would not be permitted for disposal at the MDS.
Contaminant Bioaccumulation/Trophic Transfer
Contaminant bioaccumulation was evaluated by analyzing the tissues of infaunal worms
collected from the Study Area sediments. Infaunal organism bioaccumulation of sediment-
associated contaminants can, if accumulated to high enough levels, result in both acute and
chronic impacts and eventually transform benthic community structure. Such changes can affect
the food source of demersal predators. When demersal predators feed on infauna with
contaminated tissues, the contaminants can be transferred to and potentially accumulate in the
predator. These contaminants can then potentially be consumed by humans. EPA's evaluation
of contaminant bioaccumulation in the Study Area was similar to the national testing manual's
(Green Book) Tier IV "steady-state" evaluations, which can be used in determining compliance
with the ocean dumping criteria. The results showed that there were areas in the vicinity of the
MDS where these benthic worms were accumulating undesirable levels of contaminants from the
sediments.
34
-------
Appendix A
Contaminants in Sediments
Contaminant concentrations in sediments in the vicinity of the MDS were compared to National
Oceanic and Atmospheric Administration (NOAA) ER-L (Effects Range-Low) and ER-M
(Effects Range-Median) values which have been derived from a broad range of biological and
chemical data collected synoptically from field and laboratory experiments. Although ER-L/ER-
M values are not appropriate for regulatory decision making, they are useful in sediment
evaluations when considered concurrently with other data. In general, the comparisons of ER-
L/ER-M values to contaminant levels in sediments from parts of the Study Area indicated that,
based on contaminant levels in the sediment, negative biological effects could be possible at
many stations. This conclusion is corroborated by the results of the toxicity and contaminant
bioaccumulation tests described above.
Contaminant Levels in Area Lobsters
NOAA tissue data from lobsters that were harvested in the New York Bight Apex in 1994
revealed that PCB and 2,3,7,8-TCDD (dioxin) concentrations in the hepatic tissue (tomalley) of
the lobsters were above U.S. Food and Drug Administration consumption guidelines. Other
contaminants were also present in the hepatopancreas and other tissues, but the concentrations of
these contaminants were within consumption guidelines.
It must be kept in mind that the lobsters analyzed in the NOAA study were harvested from wild
stocks in the Apex, whose populations migrate seasonally through the region, including the SEIS
Study Area. Contamination of these animals cannot be definitively linked to specific areas of
dredged material disposal, to other past dumping activities, or to other ongoing pollution sources.
Nor does the study data indicate that human consumption of lobster muscle tissue (meat)
presents health risks. However, the lobster study data do show that contaminants are being
accumulated, and that concern about potential human-health risks is warranted. This
contaminant data set complements other evidence of benthic contamination in the New York
Bight Apex region.
Definition of Remediation Material
In order for dredged material to be suitable for use as Remediation Material to be placed at the
HARS, it must conform to the Ocean Dumping Regulations (the Regulations). The Marine
Protection, Research, and Sanctuaries Act (MPRSA) or "The Act" prohibits dumping of
materials into the ocean except as authorized by USEPA or, in the case of dredged materials, by
the U.S. Army Corps of Engineers (USACE). Section 102 of the Act directs the USEPA to
establish and apply criteria for reviewing and evaluating permit applications (33 U.S.C. Section
1412). The USEPA has adopted such criteria in the Regulations. 40 CFR Section 227.6(a) lists
constituents that are prohibited from being placed in the ocean unless only present as trace
contaminants in material otherwise suitable for dumping (hereinafter referred to as "listed
constituents"). Specifically listed constituents include: organohalogen compounds; mercury and
35
-------
Appendix A
mercury compounds; cadmium and cadmium compounds; oil of any kind or in any form; known
or suspected carcinogens, mutagens, or teratogens. Section 227.27 addresses compliance with
the Limiting Permissible Concentration (LPC). See also, Section 227.13(c).
Section 227.6(b) states that constituents are considered to be present as trace contaminants only
when they are present in such forms and amounts that the "dumping of the materials will not
cause significant undesirable effects, including the possibility of danger associated with their
bioaccumulation in marine organisms." The regulations set forth criteria for determining the
potential for significant undesirable effects in Section 227.6(c). In order to be found
environmentally acceptable for ocean placement, it must be found that the liquid phase does not
contain any of the listed constituents in concentrations that would exceed applicable marine
water quality criteria after allowance for initial mixing (Section 227.6(c)(1)). For the suspended
particulate phase (Section 227.6(c)(2)) and the solid phase (Section 227.6(c)(3)), bioassay results
must not indicate occurrence of significant mortality or significant adverse sublethal effects due
to the ocean placement of wastes containing the listed constituents.
Section 227.27 of the regulations addresses the LPC. For the liquid phase, Section 227.27(a)
provides that the LPC is that concentration which does not exceed applicable marine water
quality criteria after initial mixing, or when there are no applicable marine water criteria, that
concentration of material that, after initial mixing, would not exceed 0.01 of a concentration
shown to be acutely toxic to appropriate sensitive marine organisms in a bioassay carried out in
accordance with procedures approved by USEPA and USACE. For the suspended particulate
phase and the solid phase, Section 227.27(b) provides that the LPC is that concentration of
material which will not cause unreasonable acute or chronic toxicity or other sublethal adverse
effects based on results of bioassays using appropriate sensitive organisms and conducted
according to procedures that have been approved by USEPA and USACE, and which will not
cause accumulation of toxic materials in the human food chain.
The HARS encompasses an area which includes the Mud Dump Site (MDS), and which has
exhibited the potential for adverse ecological impacts and has been identified for remediation.
The HARS consists of a Priority Remediation Area (PRA), a Buffer Zone, and a No Discharge
Zone. The PRA is a 9.0 square nautical mile area to be remediated with at least a 1 meter cap of
the Material for Remediation. The site is to be remediated with uncontaminated dredged
material (i.e., dredged material that meets current Category I standards and will not cause
significant undesirable effects including through bioaccumulation) (referred to as "Remediation
Material" or "Material for Remediation"). Under 40 CFR 228.15 (d)(6)(v)(A) the site will be
managed to reduce impacts within the Primary Remediation Area (PRA) to acceptable levels in
accordance with 40 CFR Section 228.11(c). Use of the site is restricted to dredged material
suitable for use as Material for Remediation. This material shall be selected so as to ensure it
will not cause significant undesirable effects including through bioaccumulation or unacceptable
toxicity in accordance with 40 CFR 227.6.
The following discussion describes how dredged material proposed for placement at the HARS
as Remediation Material is to be evaluated for compliance with the requirements of 40 CFR
227.6,227.27, and 228.15(d)(6). Testing of dredged material is conducted following procedures
36
-------
Appendix A
approved by USEPA and USACE, and contained in the joint USEPA/USACE national guidance
"Evaluation of Dredged Material Proposed for Ocean Dumping - Testing Manual" (February,
1991) (the "Green Book") (USEPA/USACE, 1991), and the regional implementation manual
developed by the USEPA Region 2 and CENAN (USEPA/CENAN, 1992).
Dredged material test results are analyzed in accordance with the Regulations to ensure that the
proposed placement meets the criteria of Part 227 and the requirements of 228.15(d)(6). As
explained in the preamble to the HARS Rule, Remediation Material is uncontaminated dredged
material (i.e., dredged material that meets current Category I standards and will not cause
significant undesirable effects including through bioaccumulation). The determination of
whether materials meeting the Part 227 criteria are assigned to Category I is based upon
technical and scientific judgment as set out below.
Applying the current USEPA Region 2/CENAN guidance to a project, the material would be
Category I if it meets the Part 227 criteria (including the requirements regarding acute toxicity)
and:
• bioaccumulation test results are below the regional Matrix levels for cadmium, mercury,
total PCBs, and total DDT, and are below the regional Category I values for dioxin; and
• bioaccumulation test results for the other bioaccumulative chemicals of concern
identified in USEPA/CENAN (1992) do not indicate a potential for undesirable effects
using conservative assessment techniques.
Under the proposed process, Regional Matrix Values would be eliminated from the Testing
Evaluation Framework. Under the proposed framework, the material would be Category I if it
meets the Part 227 criteria (including the requirements regarding acute toxicity) and:
• bioaccumulation test results are below the regional Category I values for dioxin; and
• bioaccumulation test results for the other bioaccumulative chemicals of concern
identified in USEPA/CENAN (1992) do not indicate a potential for undesirable effects
using conservative assessment techniques.
Sediments that meet this Category I definition are suitable for placement at the HARS as
Remediation Material as they will improve sediment conditions at the HARS to reduce impacts
to acceptable levels in accordance with 40 CFR Section 228.11(c). Sediments that do not meet
this definition are not suitable for placement at the HARS.
37
-------
Appendix A
This page intentionally left blank.
38
-------
APPENDIX B
Example Memorandum Documenting
Current HARS Suitability Evaluation
-------
Appendix B
US ARMY CORPS
OF ENGINEERS
NEW YORK DISTRICT REGION 2
MEMO FOR THE RECORD
SUBJECT: Review of Compliance with the Testing Requirements of 40 CFR 227.6 and
227.27, and Site Designation Provisions of 40 CFR 228.15 for the XXXX Project
FROM:
Douglas Pabst, Team Leader
John Tavolaro
Dredged Material Management Team
Chief, Operations Support Branch
Division of Environmental Planning
New York District
and Protection
U.S. Army Corps of Engineers
EPA Region 2
I. SUMMARY
This memorandum provides comprehensive review and analysis of XXXX Project sediment test
results. This memorandum addresses compliance with the regulatory testing criteria of 40 CFR
Sections 227.6 and 227.27, and the requirements of the rule establishing the Historic Area
Remediation Site (HARS) set out in Section 228.15(d)(6). These requirements hereinafter are
referred to as the "Regulations."
As discussed in the HARS rulemaking preamble (See 62 Fed. Reg. 46142 (August 29, 1997) and
62 Fed. Reg. 26267 (May 13, 1997)) and its accompanying documentation, the need to
remediate the Historic Area Remediation Site is amply supported by the presence in the HARS
of toxic effects, dioxin bioaccumulation exceeding Category I levels in worm tissue, as well as
TCDD/PCB contamination in area lobster stocks. Individual elements of the aforementioned
data do not prove that sediments within the Study Area are imminent hazards to the New York
Bight Apex ecosystem, living resources, or human health. However, the collective evidence
presents cause for concern, justifies that a need for remediation exists, that the site is Impact
Category I (see, 40 CFR 228.10), and that the site should be managed to reduce impacts to
acceptable levels (see, 40 CFR 228.11(c)). Further information on the conditions in the Study
Area and the surveys performed may be found in the Supplemental Environmental Impact
Statement (EPA, 1997c).
39
-------
Appendix B
This evaluation confirms that: 1) all tests required under the Regulations were conducted; 2) this
project meets the criteria at 40 CFR Section 227.6 for trace contaminants and Section 227.27 for
Limiting Permissible Concentration (LPC); and 3) the dredged material is Category I under U.S.
Environmental Protection Agency (USEPA) Region 2/Corps of Engineers, N.Y. District
(CENAN) guidance and is suitable for placement at the HARS as Remediation Material.
II. PROJECT DESCRIPTION
The proposal is to dredge and place approximately 440,000 cubic yards (yd) of dredged material
at the HARS. The project encompassed one reach; sediment core samples were taken from 13
locations to characterize the sediment (see sampling plan (CENAN, 2000)).
HI. REGULATORY REQUIREMENTS
In order for dredged material to be suitable for use as Remediation Material to be placed at the
HARS, it must conform to the Regulations. The Marine Protection, Research, and Sanctuaries
Act (MPRSA) or "The Act" prohibits dumping of materials into the ocean except as authorized
by USEPA or, in the case of dredged materials, by the U.S. Army Corps of Engineers (USACE).
Section 102 of the Act directs the USEPA to establish and apply criteria for reviewing and
evaluating permit applications (33 U.S.C. Section 1412). The USEPA has adopted such criteria
in the Regulations. 40 CFR Section 227.6(a) lists constituents that are prohibited from being
placed in the ocean unless only present as trace contaminants in material otherwise suitable for
dumping (hereinafter referred to as "listed constituents"). Section 227.27 addresses compliance
with the LPC. See also, Section 227.13(c).
Section 227.6(b) states that constituents are considered to be present as trace contaminants only
when they are present in such forms and amounts that the "dumping of the materials will not
cause significant undesirable effects, including the possibility of danger associated with their
bioaccumulation in marine organisms." The regulations set forth criteria for determining the
potential for significant undesirable effects in Section 227.6(c). In order to be found
environmentally acceptable for ocean placement, it must be found that the liquid phase does not
contain any of the listed constituents in concentrations that would exceed applicable marine
water quality criteria after allowance for initial mixing (Section 227.6(c)(1)). For the suspended
particulate phase (Section 227.6(c)(2)) and the solid phase (Section 227.6(c)(3)), bioassay results
must not indicate occurrence of significant mortality or significant adverse sublethal effects due
to the ocean placement of wastes containing the listed constituents.
Section 227.27 of the regulations addresses the LPC. For the liquid phase, Section 227.27(a)
provides that the LPC is that concentration which does not exceed applicable marine water
quality criteria after initial mixing, or when there are no applicable marine water criteria, that
concentration of material that, after initial mixing, would not exceed 0.01 of a concentration
shown to be acutely toxic to appropriate sensitive marine organisms in a bioassay carried out in
accordance with procedures approved by USEPA and USACE. For the suspended particulate
40
-------
Appendix B
phase and the solid phase, Section 227.27(b) provides that the LPC is that concentration of
material which will not cause unreasonable acute or chronic toxicity or other sublethal adverse
effects based on results of bioassays using appropriate sensitive organisms and conducted
according to procedures that have been approved by USEPA and US ACE, and which will not
cause accumulation of toxic materials in the human food chain.
The HARS encompasses an area which includes the Mud Dump Site (MDS), and which has
exhibited the potential for adverse ecological impacts and has been identified for remediation.
The site will be remediated with uncontaminated dredged material (i.e., dredged material that
meets current Category I standards and will not cause significant undesirable effects including
through bioaccumulation) (hereinafter referred to as "Remediation Material" or "Material for
Remediation"). Under 40 CFR 228.15 (d)(6)(v)(A) the site will be managed to reduce impacts
within the Primary Remediation Area (PRA) to acceptable levels in accordance with 40 CFR
Section 228.11(c). Use of the site is restricted to dredged material suitable for use as Material
for Remediation. This material shall be selected so as to ensure it will not cause significant
undesirable effects including through bioaccumulation or unacceptable toxicity in accordance
with 40 CFR 227.6.
Section 228.15(d)(6) of the Regulations describes the locations where material from NY/NJ
Harbor and surrounding areas may be placed in the HARS, provided that it is suitable as
Remediation Material. The HARS consists of a PRA, a Buffer Zone, and a No Discharge Zone.
Under 228.15(d)(6) placement of Remediation Material is limited to the PRA.
IV. GUIDANCE FOR TESTING AND EVALUATION OF DREDGE D MATERIAL
The discussion below describes how the material proposed for placement at the HARS as
Remediation Material, resulting from the maintenance dredging of XXXX Project was evaluated
for compliance with the requirements of 40 CFR 227.6, 227.27, and 228.15(d)(6). Testing of the
material was conducted following procedures approved by USEPA and USACE, and contained
in the joint USEPA/USACE national guidance "Evaluation of Dredged Material Proposed for
Ocean Dumping - Testing Manual" (February, 1991) (the "Green Book") (USEPA/USACE,
1991), and the regional implementation manual developed by the USEPA Region 2 and CENAN
(USEPA/CENAN, 1992).
These test results were analyzed in accordance with the Regulations to ensure that the proposed
placement meets the criteria of Part 227 and the requirements of 228.15(d)(6). As explained in
the preamble to the HARS Rule, Remediation Material is uncontaminated dredged material
(i.e., dredged material that meets current Category I standards and will not cause significant
undesirable effects including through bioaccumulation). The determination of whether materials
meeting the Part 227 criteria are assigned to Category I is based upon technical and scientific
judgment as set out below.
41
-------
Appendix B
Applying the USEPA Region 2/CENAN guidance to this project, the material would be Category
I if it meets the Part 227 criteria (including the requirements regarding acute toxicity) and:
• bioaccumulation test results are below the regional Matrix levels for cadmium, mercury,
total PCBS, and total DDT, and are below the regional Category I values for dioxin; and
• bioaccumulation test results for the other bioaccumulative chemicals of concern
identified in USEPA/CENAN (1992) do not indicate a potential for undesirable effects
using conservative assessment techniques.
Sediments that meet this Category I definition are suitable for placement at the HARS as
Remediation Material as they will improve sediment conditions at the HARS to reduce impacts
to acceptable levels in accordance with 40 CFR Section 228.11(c). Sediments that do not meet
this definition are not suitable for placement at the HARS.
V. RESULTS OF EVALUATION OF THE MATERIAL
A. Evaluation of the liquid phase
The liquid phase of the material was evaluated for compliance with Sections 227.6(c)(1) and
227.27(a). There are applicable marine water quality criteria for constituents in the material,
including listed constituents, and the applicable marine water quality criteria would not be
exceeded after initial mixing. In addition, liquid phase bioassays run as part of the suspended
particulate phase on three appropriate sensitive marine organisms, show that after initial mixing
(as determined under 40 CFR 227.29(a)(2)), the liquid phase of the material would not exceed a
toxicity threshold of 0.01 of a concentration shown to be acutely toxic to appropriate sensitive
marine organisms. Accordingly, it is concluded that the liquid phase of the material would be in
compliance with 40 CFR 227.6(c)(1) and 227.27(a). The specific test results and technical
analysis of the data underlying this conclusion are described and evaluated in CENAN (2001).
B. Evaluation of the suspended particulate phase
The suspended particulate phase of the material was evaluated for compliance with Sections
227.6(c)(2) and 227.27(b). Bioassay testing of the suspended particulate phase of the material
has been conducted using three appropriate sensitive marine organisms: inland silversides
(Menidia beryllina), mysid shrimp (Mysidopsis bahia), and blue mussel (Mytilus edulis). That
information shows that when placed in the HARS and after initial mixing (as determined under
40 CFR 227.29(a)(2)), the suspended particulate phase of this material would not exceed a
toxicity threshold of 0.01 of a concentration shown to be acutely toxic in the laboratory
bioassays, and thus would not result in significant mortality. The material may be discharged on
either incoming or outgoing tides. However discharge may occur only at distances of at least
2,400 feet from the northern and southern boundaries and at least 2,200 feet from eastern and
western boundaries of the HARS, independent of the direction of tidal transport at the time of
disposal. The specific test results and technical analysis of the data underlying this conclusion
are described in CENAN (2001). The factor of 0.01 was applied to ensure that there would be no
42
-------
Appendix B
significant adverse sublethal effects. Moreover, the fact that after placement, the suspended
particulate phase would only exist in the environment for a short time, means the suspended
particulate phase would not cause significant undesirable effects, including the possibility of
danger associated with bioaccumulation, since these impacts require long exposure durations
(see USEPA, 1994). Accordingly, it is concluded that the suspended phase of the material would
be in compliance with 40 CFR 227.6(c)(2) and 227.27(b).
C. Evaluation of the solid phase
The solid phase of the material was evaluated for compliance with Sections 227.6(c)(3) and
227.27(b). This evaluation was made using the results of two specific types of evaluations on the
solid phase of the material, one focusing on the acute (10-day) toxicity of the material, and the
other focusing on the potential for the material to cause significant adverse effects due to
bioaccumulation. Both types of tests used appropriate sensitive benthic marine organisms
according to procedures approved by USEPA and the US ACE. The following sections address
the results of those tests and further analyze compliance with the regulatory criteria of Sections
227.6(c)(3), 227.27(b), and 228.15 and with EPA Region 2/CENAN guidance.
1. Solid phase toxicity evaluation
Ten-day toxicity tests were conducted on project materials using mysids (M. bahia) and
amphipods (Ampelisca abdita), which are appropriate sensitive benthic marine organisms. These
organisms are good predictors of adverse effects to benthic marine communities (see, USEPA,
1996a). The mortality in project sediments did not exceed mortality in the reference sediment by
10% for mysid shrimp or 20% for amphipods and was not statistically greater than reference for
either mysids or amphipods. These results show that the solid phase of the material would not
cause significant mortality and meets the solid phase toxicity criteria of Sections 227.6 and
227.27.
2. Solid phase bioaccumulation evaluation
USEPA/USACE (1991) describes an approved process of evaluating bioaccumulation potential
using comparative analysis of project sediment bioaccumulation to reference sediment
bioaccumulation, FDA Action levels and evaluation of eight additional factors for assessing the
significance of bioaccumulation. These factors are:
• number of species in which bioaccumulation from the dredged material is statistically
greater than bioaccumulation from the reference material
• number of contaminants for which bioaccumulation from the dredged material is
statistically greater than bioaccumulation from the reference material
• magnitude by which bioaccumulation from the dredged material exceeds
bioaccumulation from the reference material
43
-------
Appendix B
• toxicological importance of the contaminants whose bioaccumulation from the dredged
material exceeds that from the reference material
• phylogenetic diversity of the species in which bioaccumulation from the dredged material
statistically exceeds that from the reference material
• propensity for the contaminants with statistically significant bioaccumulation to
biomagnify within aquatic food webs
• magnitude of toxicity and number and phylogenetic diversity of species exhibiting
greater mortality in the dredged material than in the reference material
• magnitude by which contaminants whose bioaccumulation from the dredged material
exceeds that from the reference material also exceed the concentrations found in
comparable species living in the vicinity of the proposed site
In following this guidance, USEPA Region 2 and CENAN used a framework (described in
Figure 1) for evaluating project sediment bioaccumulation results. As shown in Figure 1, this
process involves four consecutive evaluations. In the first three evaluations, the project sediment
bioaccumulation test results for each compound of concern are sequentially compared to:
a) reference test results; bi) FDA Action levels; b2) Regional Matrix levels; bs) regional dioxin
values; and, c) general risk-based evaluations (including comparison to background tissue
concentrations). If these evaluations show that the project sediment does not exceed the
reference test results in step (a), the FDA levels in step (bi), and the Regional Matrix
levels/dioxin Category I values in steps (b2 to b3) for a particular compound, this indicates that
the placement of the material would not result in adverse effects due to that chemical, and there
is no need to further evaluate that individual chemical in the next step. Markings in columns 5 or
7 of Table 1 indicate where project test results were statistically greater than the reference levels
for the clam or the worm. If any species are marked for a particular compound, the evaluation
will proceed to the next step. General risk-based evaluations are conducted in step (c) for
compounds not resolved in steps (a) or (bi to b3). The fourth evaluation (d) uses all the
information and results of the individual chemical evaluations (particularly as these results relate
to the eight Green Book factors listed above), to evaluate the solid phase of the dredged material
as a whole. These evaluations for this project are discussed below in the order described in
Figure 1.
Bioaccumulation tests were conducted on the solid phase of the project material for contaminants
of concern identified in USEPA/CENAN (1992) and the project sampling plan (CENAN, 1997)
using two appropriate sensitive benthic marine organisms, sand worm (Nereis virens) and bent-
nosed clam (Macoma nasuta). These species are considered to be good representatives of the
phylogenetically diverse base of the marine food chain. Contaminants of concern were identified
for the regional testing manual from the NY/NJ Harbor Estuary Program Toxics Characterization
report (Squibb et al., 1991). That report was prepared as part of development of the Harbor
Estuary Program in order to identify and characterize contaminants in Harbor sediments. Those
compounds with the potential to bioaccumulate (KoW of approximately 4 or greater) are included
on the testing list and evaluated by use of bioaccumulation tests when expected to be present in
project sediments based upon the location of contaminant inputs and results of previous sediment
44
-------
Appendix B
sampling. The bioaccumulation test results were used in evaluating the potential impacts of the
material. The determination is that the combined results of the toxicity and bioaccumulation
tests indicated that the material meets the criteria of Sections 227.6(c)(3) and 227.27(b) and
228.15(d)(6)(v)(A) of the Regulations, and that the material is suitable for placement at the
HARS as Remediation Material.
a. Comparison of Bioaccumulation Test Results to Reference Sediment Test Results
Concentrations of contaminants in tissues of organisms exposed for 28 days to project sediments
were compared to concentrations in tissues of organisms exposed for 28 days to reference
sediment. Reference sediment serves as a point of comparison to identify potential effects of
contaminants in the dredged material (USEPA/USACE, 1991). In essence, exposing test
organisms to this sediment allows for the prediction of contaminant levels that would result in
the test organisms were they "in the wild" at the area from which the reference sediment was
taken.
The tissue concentrations in two species of appropriate sensitive benthic marine organisms
resulting from 28-day exposure to project sediments is compared to the tissue concentrations in
the same species of organisms resulting from 28-day exposure to reference sediment. In order to
make a statistically valid determination that the project sediment does/does not cause greater
bioaccumulation than the reference sediment, several sub-samples of the dredged material and
reference are run; these separate sub-samples are called replicates. A mean can then be
calculated with a standard deviation for each sediment {i.e., XXXX Project sediments, and
Reference sediment). The means and standard deviations are compared using a standard
statistical approach, and a determination is made, with 95 percent confidence, that there is or is
not a true difference between the test and reference sediments. A statistical analysis is merely a
quantification of the variability between the test and reference data, and a measure of the
probability of the difference being real. Throughout this memorandum, statements regarding
project sediment having "greater" or "less" bioaccumulation are referring to calculated
differences which are statistically significant at the 95% confidence level. To be
environmentally conservative, test values which were below detection levels were estimated at
very conservative levels for purposes of statistical comparisons (USEPA/CENAN, 1997)
The reference sediment used for this project was collected at the Reference Site, in an area of
clean, sandy sediments located in the New York Bight near the HARS, where the sediments are
unaffected by prior dredged material disposal (see reference values in Table 1, Columns 2 and
3).
When bioaccumulation in organisms exposed to project sediments is not greater than
bioaccumulation in organisms exposed to appropriate reference sediments, this means that
placement of the material would not result in bioaccumulation above that found to occur in the
"clean" reference sediment. Accordingly, such material would not result in bioaccumulation that
would cause unreasonable degradation of the environment or human health, or significant
adverse effects. In cases where bioaccumulation levels are statistically greater (at the 95%
confidence limit) than in the reference, further evaluation for potential effects is warranted.
45
-------
Appendix B
A statistically significant difference between test and reference bioaccumulation is not itself a
quantitative prediction that an impact would occur in the field, nor is it related to any cause and
effect. A key to understanding bioaccumulation and potential adverse impacts is that
bioaccumulation is a phenomenon and does not necessarily result in an effect. In addition,
depending upon the exposure (concentration and duration), bioaccumulation may cause no harm.
On the other hand, as exposure and subsequent bioaccumulation increases, the potential for
adverse effects increases.
The following text summarizes the test results comparing bioaccumulation from the project
sediments to that in the reference sediments. (Contaminants for which bioaccumulation from the
dredged material was statistically greater than the reference in the clam and/or the worm are
indicated by a mark in columns 5 and/or 7 for that compound in Table 1.)
Metals
• Of the nine metals tested, six in the clam and four in the worm were bioaccumulated
greater in the project sediment than the reference. Cadmium and mercury are the only
metals that are listed constituents in Section 227.6(a). Only mercury in the clam tissue
bioaccumulated greater in the project sediment than the reference.
Pesticides
• Of the 15 pesticides (including DDT congeners) tested, seven in the clam and six in the
worm were bioaccumulated greater in the project sediment than the reference.
Industrial Chemicals
• Total PCBs and 1,4-dichlorobenzene bioaccumulated greater than reference in both the
clam and the worm.
Dioxins
• Of dioxin congeners (including 2,3,7,8-TCDD and 16 other dioxin congeners), five in the
clam and 11 in the worm were bioaccumulated greater in the project sediment than the
reference.
PAHs
• Of the 16 PAHs tested, 15 in the clam and five in the worm were bioaccumulated greater
in the project sediment than the reference.
For all metals, PCBs, 1,4-dichlorobenzene, and dioxin/furan compounds, the magnitude of
exceedance is less than 10 times the reference. For the remaining contaminants that
bioaccumulated from project sediment to greater concentrations than the reference, 11 PAHs,
and two pesticides exceeded the reference greater than ten times in the clam and/or the worm.
Exceedance of the reference values is common when reference values are very low or Anon-
detect, @ as here. In such cases the potential for the actual tissue concentration to be related to
an effect on the organism or the food chain (including human health) is further evaluated.
46
-------
Appendix B
b. Comparison to FDA Action levels, Regional Matrix Levels and Dioxin Values
i.) Comparison to FDA Action levels (bi)
There are FDA Action levels for seven compounds (aldrin, dieldrin, a-chlordane, heptachlor,
heptachlor epoxide, PCBs, and mercury). None of the contaminants for which there are FDA
Action levels exceed such thresholds in the tissues of organisms exposed to project sediments for
28 days (see also Table 1). The source of FDA Action levels is described in USEP A/US ACE
(1991). Table 1, Column 18, identifies the relevant FDA Action levels.
Exceedance of an FDA Action level results in a conclusion that the placement of the dredged
material would result in significant adverse effects. No contaminants for which there are FDA
Action level exceeded any such level in either the clam or the worm tissues.
ii) Comparison to Regional Matrix Levels (b2)
There are regional Matrix levels for four compounds (cadmium, mercury, PCBs and total DDT).
The source of regional Matrix levels is described in USACE (1981). Table 1, Column 20,
identifies the relevant regional Matrix levels. Bioaccumulation results that exceed the regional
Matrix level indicate that the sediment is not Category I under USEPA Region 2/CENAN
guidance. Fulfilling a commitment in the New York - New Jersey Comprehensive Conservation
and Management Plan, in 1998 USEPA initiated a peer review of the testing evaluation
framework utilized to determine suitability of dredged material for use as Remediation Material
at the HARS. On September 27, 2000, the USEPA and USACE signed a memorandum of
agreement (MO A) outlining the steps to be taken to ensure that remediation of the HARS
continues in a manner appropriately protective of human health and the environment. The MO A
stated that "in order to ensure that EPA-R2 and US ACE-NAD continue to incorporate the most
up-to-date science in determining what material is appropriate for placement at the HARS, and to
ensure that the quality of material approved for placement at the HARS meets the remedial
purposes of the HARS designation, one element of the HARS-TEF [testing evaluation
framework] will immediately be changed. Specifically, one matrix value [PCBs in worm tissue]
in the HARS-specific effects levels will be revised." The MO A further states: "This interim
change shall be effective immediately upon signing of this MOA and be subject to future review
by the re-convened RMW [Remediation Material Workgroup]." The Matrix value for clam
tissue remains the same. Total DDT, total PCB, cadmium, and mercury did not bioaccumulate in
either clam or worm tissue at concentrations exceeding the Matrix or interim revised Matrix
levels.
iii) Comparison to Regional Dioxin Values (b3)
Regional dioxin values are intended for comparison to the results of 28-day bioaccumulation test
results, and the source and use of the values are described in the USEPA (1997a). Table 1,
Sheet B, Column 20, identifies the relevant regional dioxin values. Twenty-eight day
bioaccumulation test results that equal or exceed the regional Category I dioxin value of 1 pptr
for 2,3,7,8-TCDD (or 4.5 pptr using the TEQ approach described in USEPA (1997a) for the non
47
-------
Appendix B
2,3,7,8-TCDD congeners) indicate that the sediment is not Remediation Material under USEPA
Region 2/CENAN guidance. Bioaccumulation test results were below the regional Category I
dioxin values for both the worm or the clam.
iv) Steady State Considerations for Matrix Compounds
When the end point to which the test data is compared potentially represents a steady-state level,
rather than a 28-day level, consideration may need to be given to whether the 28-day test results
are representative of bioaccumulation levels that could be expected to occur in the field after
placement. The literature was reviewed to determine the degree to which the test results reached
steady state, as appropriate. The relevance of adjusting project data to steady state for
comparison to regional Matrix levels is discussed below.
PCBs
To assess the rate of bioaccumulation of PCBs and other compounds, Rubinstein et al. (1990)
and Pruell et al. (1993) exposed three species of organisms, the grass shrimp Palaemonetes
pugio; the sandworm Nereis virens; and the clam Macoma nasuta, to sediments collected from
the Passaic River, N.J. Sub-samples of the exposed organisms were removed on various days
into the study including days 0,10,28,42, 84, and 180. For the clam tissue, the variance in the
concentrations on day 28 and day 84 (by which point the maximum concentration had been
reached) overlap, thus indicating that the two are not statistically different and the
bioaccumulation on day 28 is at or very close to steady-state. Thus, the clam bioaccumulation
for the project sediments using 28-day exposures is acceptable for use as steady-state tissue
levels, and was below the Matrix level for total PCB. For the worm tissue, variances for days 28
and 180 do not overlap, thus indicating that steady-state was probably not reached in 28 days,
although the variance in the data makes it difficult to quantify a real difference. However, if the
means for days 28 and 180 from Rubinstein et al. (1990) are compared (approximately 1,750
ng/g (nanograms per gram or parts per billion, ppb) for 28 days, and 3,000 ng/g for 180 days)
this indicates approximately 58% of steady-state would have been reached in 28 days. If on this
basis the worm project data are conservatively adjusted upward by even a factor of two to
calculate a steady-state tissue concentration, the dredged material tissue concentration is still
below the Matrix level for PCBs in the worm.
Total DDT
With regard to DDT and its metabolites, the degree to which these compounds reached steady-
state was also evaluated. Table 1 contains the project test results for the total DDT, which is the
sum of the results for DDT and its metabolites (i.e., DDE and DDD). This level is compared to
the Matrix level for total DDT. To assess the rate of bioaccumulation of the DDTs and their
metabolites, Lee et al. (1994) exposed the clam Macoma nasuta, to sediments collected from the
vicinity of the United Heckathorn Superfimd site in Richmond California. The study measured
tissue residues and uptake kinetics from exposure to pesticide-contaminated sediments. Results
of the study indicate that one parent compound, 4,4-DDT, bioaccumulates much more slowly
than 2,4-DDT and the DDT metabolites. The results range from approximately 9 percent of
48
-------
Appendix B
steady state after 28 days for 4,4-DDT, to 55 percent of steady state after 28 days for 2,4-DDT.
(Lee et al., 1994) In the XXXXX project, the parent compounds (4,4-DDT and 2,4-DDT) were
not detected at levels statistically greater than in reference. The DDT metabolites were detected
and were statistically greater than the reference in the bioaccumulation test results for both the
clam and the worm tissue. In order to calculate a steady-state tissue concentration, based on the
above study a factor of 11 was applied to the project data for 4,4-DDT, a factor of three to the
project data for 4,4-DDD, and a factor of two for 2,4-DDT and the remaining DDT metabolites,
assuming the detection limit represents the amount present when "not detected." Using these
conservative assumptions, the dredged material tissue concentration is below the Matrix level for
total DDT in both the worm and the clam.
Cadmium and Mercury
Cadmium and mercury are not regulated in marine organisms as are essential metals, and, thus,
no adjustment for steady state is applicable. The Matrix levels for cadmium and mercury,
therefore, do not represent "steady state." Bioaccumulation of these metals is affected by many
complex factors, and is essentially linear (Dethlefsen, 1978; Giesy, et al., 1980; V-Balogh and
Salanka, 1984). Therefore, there are no adjustments that can be made to reproduce "steady
state," and so 28-day test results are used to compare to the Matrix levels.
c. Risk-based evaluations
The potential for impacts due to compounds that produced greater bioaccumulation from project
sediments than the reference sediments and for which Matrix levels or Regional dioxin values
did not exist, was determined using risk-based evaluations. As noted in Table 1 and the previous
discussions, for this project PAHs, chlordane, dieldrin, trans nonachlor, 1,4-dichlorobenzene,
arsenic, chromium, copper, lead, and nickel fall into that group for the worm and/or clam.
The toxicological significance of this bioaccumulation was evaluated by: i) consideration of
steady-state bioaccumulation and food-chain transfer; ii) comparison to background tissue
concentrations; iii) consideration of potential ecological effects; and, iv) consideration of
potential carcinogenic and non-carcinogenic effects on human health.
i) Consideration of Steady-State Bioaccumulation and Food-Chain Transfer
Bioaccumulation tests were conducted using 28-day exposure of appropriate sensitive benthic
marine organisms to sediment. As previously discussed, for bioaccumulation evaluations
involving comparisons with "steady-state" tissue concentrations (as opposed to evaluations using
other 28-day tissue concentrations such as the comparison to reference sediment), it may be
necessary to understand the extent to which the organism tissue concentration has reached
steady-state. Steady-state may be defined operationally as the lack of any significant difference
(ANOVA, alpha = 0.05) among tissue residues taken at three consecutive sampling intervals
(Lee, et al, 1989). The 28-day test exposure period was selected as appropriate because most
chemicals of concern will reach at least 80% of steady-state in benthic marine organisms within
that time frame (Boese and Lee, 1992). For the few chemicals that may not meet steady-state
tissue concentrations in 28 days, a factor may be used to adjust the data to steady-state when
49
-------
Appendix B
necessary. In order to better use the tissue concentration results of 28-day bioaccumulation
exposure tests to assess the risks posed to the environment from the chemicals requiring further
evaluation (see discussion above for the identification of such chemicals), consideration was
given to the steady-state concentration of these compounds that could occur in the HARS after
extended periods of time. In addition, the potential movement of these compounds through the
food chain was considered and appropriate trophic transfer factors applied to adjust the data
accordingly, as described below.
Metals
In general, metals bioaccumulate more rapidly than organics and 28-day tests are sufficient to
evaluate potential effects (see USEPA/USACE, 1991), for example, arsenic (Naqvi, et al., 1990;
Riedel, et al., 1987; Oladimeji, et al., 1984).
Trophic transfer of most metals is not sufficient to qualify as biomagnification (Brown and Neff,
1993). The lack of observed biomagnification for such metals as chromium, copper, lead, and
nickel is the result of incomplete absorption of metals across the gut, rapid excretion, and
dilution in muscle, which represents a large part of the total body weight of most marine animals
(Fowler, 1982; Suedel et al., 1994). For purposes of conducting the human health and ecological
evaluations below, a conservative trophic transfer coefficient equal to one will be used for these
non-biomagnifying metals (Suedel et al., 1994 and references cited therein). The available
evidence does indicate that biomagnification in aquatic food chains is possible for arsenic,
though biomagnification of this metal will not occur in all cases (Suedel et al., 1994). The
biomagnification potential of arsenic in aquatic systems results from the transformation of this
metal into organic forms, thereby increasing lipid solubility and transfer rates across biological
membranes (Lunde 1977; Bryan and Langston, 1992). Arsenic is found in marine organisms as
an organic complex which includes such compounds as arsenobetaine and arsenocholine (Abel
and Axiak, 1991). Organic arsenic in the tissues of aquatic organisms is not metabolized by
predators or humans and is readily eliminated from the body through excretion (Hrudey et al.,
1996). As a result, the toxicity of organic arsenic ingested from seafood is low and appears to
pose no significant hazard (Abernathy and Ohanian, 1992). Ecological effects of arsenic are
evaluated using a WQCTL, as discussed below.
Pesticides and Industrial Chemicals
Uptake of non-polar organic contaminants from food is highly dependent on its hydrophobicity,
a property measured by the octanol/water partition coefficient, K<>w. The higher the value of KoW,
the longer it takes to reach steady-state in benthic marine organisms. For the organochlorine
compounds aldrin, dieldrin, heptachlor, trans nonachlor, and alpha-chlordane that have log KoW >
6, it is possible that steady-state was not reached within 28 days. Information contained in Boese
and Lee (1992) indicates that 28-day bioaccumulation tests for these chemicals achieve at least
50% of steady-state for aldrin, dieldrin, and a-chlordane. The remaining compounds, heptachlor
and trans nonachlor, were not addressed in Boese and Lee (1992), however estimates of the
fraction of steady state achieved after 28-days can be calculated using the equations contained in
McFarland (1995). Results of such calculations indicate heptachlor and trans nonachlor reach
50
-------
Appendix B
approximately 50% of steady state after 28 days. Comparison of the project data for these
compounds with the effects data discussed below after using an appropriate factor (a factor of
two) to adjust the project data to an appropriate steady-state concentration, indicates that project
sediment results would be below conservative ecological and human effects levels. Calculating
the fraction of steady state achieved for the remaining pesticides with log KoW<6, and for 1,4-
dichlorobenzene (Kow=3.44) indicates that these compounds reach steady state within 28 days, so
no adjustment is necessary.
The potential for these chemicals to biomagnify was also evaluated. Although organic
contaminants with values of log KoW > 4 tend to biomagnify in the marine food chain, studies
(USACE, 1995) have shown that this is not true for higher molecular weight compounds such as
the most highly chlorinated PCBs or for easily metabolized compounds such as PAHs. Those
organic compounds which are not efficiently excreted, such as certain pesticides (including
dieldrin, a-chlordane, and trans nonachlor), can biomagnify in the food chain. For the organic
constituents with a potential to biomagnify in the marine food chain, trophic transfer factors were
calculated, using the approach described by Gobas (1993). The values are summarized in
Attachment B. These factors, which ranged from 1.4 to 2.3, were taken into account in assessing
potential human health and ecological risk effects of these compounds, as discussed below.
PAHs
The time required for a given PAH to attain a steady-state concentration following exposure to
bedded sediments (tss) is determined primarily by the log KoW of the compound in question
(McFarland, 1995; Meador et al, 1995). Meador et al. (1995) reviewed nine studies that
investigated the attainment of steady-state tissue concentrations of PAHs by various marine
invertebrates. In each case, tissue concentrations approached steady-state within several days to
two weeks after initiating exposure to both low molecular weight PAHs and high molecular
weight PAHs. McFarland (1995) estimated the time to steady-state (tss) for 15 PAHs based on
their hydrophobicity. The tss values ranged from 3.5 to 326 days. The estimated steady-state
concentration of the sum total of the 15 PAHs analyzed by McFarland for sediments collected
from typical harbor areas revealed that the mean concentration attained after 28-day
bioaccumulation tests was approximately 86% of steady-state. McFarland (1995) concluded that
28-day tests are likely to reflect steady-state. However, even using the conservative approach of
adjusting the data to calculate steady-state for the individual PAHs in the project based on
McFarland (1995) (using a factor of one, two, or three, as indicated) and summing the results, the
project data would still fall below the effects levels as discussed below.
With regard to the potential for biomagnification of PAHs, feeding studies show that assimilation
rates from ingested food are extremely low, e.g., more than 98% of the target contaminant
remained in an undigested form in fish gut 48 hours after feeding squid containing radio-labeled
benzo[a]pyrene to young cod (Corner et al., 1976) and juvenile Atlantic herring (Whittle et al.,
1977). PAH metabolites are also transferred through the marine food chain; however, they are
absorbed even less efficiently than their parent compounds (McElroy and Sisson, 1989; McElroy
et al., 1991). Up to 99% of the PAH compounds taken up by fish are metabolized and excreted
into bile, the usual elimination mode, within 24 hours of uptake (Varanasi et al, 1989). Similar
51
-------
Appendix B
results are described in Brown and Neff (1993) who evaluated various studies describing trophic
transfer. The studies cited in Brown and Neff (1993) indicate a trophic transfer rate for BaP
from invertebrates to fish of between 0.02 and 0.23 times the concentration in the ingested
invertebrates (Corner et al., 1976, O'Connor et al., 1988, McElroy et al., 1991). This was taken
into account when assessing the ecological and human health effects of the project material as
discussed below.
ii) Comparison of Test Results to Background Tissue Concentrations
Where data regarding tissue levels of organisms living in the general area of the HARS are
available ("background levels"), it is useful to compare those levels with the test levels as part of
the risk evaluation (Figure 1, Box c). However, this comparison is not, by itself, definitive.
When bioaccumulation in organisms exposed to project sediments is not greater than tissue
concentrations in organisms from the vicinity of the remediation site (the background levels),
this means that placement of the material would not result in bioaccumulation above existing
ambient levels in the general area and thus does not have a potential to cause undesirable effects.
When bioaccumulation in organisms exposed to project sediments is greater than these levels, it
may or may not be predictive of adverse effects {e.g., it may reflect extremely low "background"
levels). Depending on the exposure (concentration and duration), bioaccumulation may cause no
harm. However, as exposure increases, the potential for adverse effects increases.
Organisms collected from a broad area of the sea floor in the vicinity of (but not inside) the
HARS have been collected and analyzed for tissue concentration for bioaccumulative
contaminants of concern (Charles and Muramoto, 1990; USACE, 1994; USEPA, 1996f; USEPA,
1997b). These field-generated bioaccumulation results provide a measure of the tissue residues
for organisms living outside the HARS. Table 1, Columns 16 and 17 summarize the most recent
background data. For clam background, data were collected only for the following constituents:
all PAHs, aldrin, two DDT compounds, PCBs, and seven of the nine metals analyzed. Where
background values exist, the following constituents exceeding background are identified: only
two PAHs exceeded background levels in the worm, no PAHs exceed clam background values.
Total PAHs did not exceed background levels for either clam or worm tissue. Dieldrin, a-
chlordane, trans nonachlor, total residual chlordane/heptachlor, 4,4-DDD, and total PCBs
exceeded background levels for the worm.
iii) Consideration of Potential Ecological Effects
A review of scientific information was also done to further evaluate the test results with respect
to potential ecological impacts for the chemicals requiring further evaluation (above reference
and for which there is no Matrix level or dioxin value).
Metals, Pesticides, and Industrial Chemicals
The potential for ecological effects from the bioaccumulation over reference of dieldrin, total
chlordane, 1,4-dichlorobenzene, arsenic, chromium, copper, lead, and nickel was evaluated by
comparing to a Water Quality Criterion Tissue Level (WQCTL). The WQCTL is calculated by
multiplying the Clean Water Act Section 304(a)(1) Federal water quality criterion chronic value
52
-------
Appendix B
(CV) for the chemical by the empirically determined bioconcentration factor (BCF) for the
chemical for a representative marine organism (Lee et al., 1989). A BCF is the ratio of the
concentration of a contaminant in an organism to the concentration of the contaminant in water.
Thus, the WQCTL represents the tissue concentration that would be expected in an organism
exposed to water containing the chemical at the CV concentration. This level is set to protect
95% of all tested organisms included in the water quality criterion database, thus representing a
conservative level of protection (USEPA, 1985b). Table 1 lists the calculated WQCTLs.
Sources of CVs and BCFs are USEPA ambient water quality criteria documents (USEPA 1980b,
1980c, 1980d, 1980e, 1980f, 1984a, 1984b, 1985a, 1985c, 1986,1987b and 1992a) and
Calabrese (1984)(for silver). Calculations are shown in attachment A. None of the WQCTLs
were exceeded. Therefore, these bioaccumulation test results do not indicate a potential for
undesirable ecological effects.
PAHs
For PAHs, a more definitive method is available for evaluating the potential ecological effects.
This method makes use of a direct comparison of total PAH tissue residues and the Critical Body
Residue (CBR). This approach is supported by a review of the scientific literature. The CBR
approach described by McCarty (1991) was used to evaluate the potential impacts of total PAHs
accumulated in the dredged material bioaccumulation test organisms. CBRs are concentrations
of chemical residues in organisms which elicit a deleterious biological response associated with
narcosis, which is the primary non-cancer effect of PAHs. Narcotic responses measured can be
acute (e.g., immobilization or death) or chronic endpoints (e.g., reduced reproduction, fecundity
or growth). CBRs are represented as the ratio of the mass of toxicant to the mass of the
organism, such as millimoles or micrograms of toxicant per kilogram (mmole or ug/kg) of
organism. For the narcosis endpoint, each molecule of individual PAH congeners is generally
equipotent, thus the total PAH concentration is compared to the CBR. For example, a 400 ppb
dose of naphthalene would elicit a similar toxicity response as 400 ppb of fluorene; if both
chemicals are present together at these concentrations, then the dose would equal 800 ppb (see
Appendix for Table 1).
As shown in Table 1, total PAH levels in tissues from the dredged material bioaccumulation test
were below levels at which chronic adverse effects might be expected from a narcotic mode of
action in sensitive aquatic organisms (i.e., fish) as estimated by the CBR.
Effects of Mutagenic, Carcinogenic and Teratogenic PAHs. Applying the uncertainty factor (UF)
of 10 and a trophic transfer factor of 0.1 described in the Appendix for Table 1, to the no-effects
level for BaP calculated from Hannah et al. (1982), as discussed in the Appendix for Table 1
(8,021 ppb) results in a no-effect level for BaP of approximately 8,000 ppb in benthic tissue,
which is considerably greater than the highest tissue concentration of BaP found in the project
bioaccumulation test results (approx. 7 ppb). Even when applying the more conservative steady-
state factors for BaP and the other carcinogenic PAHs derived from McFarland (1995), as
identified above, the calculated concentrations (13 ppb for BaP only and 18 ppb for total BaP
equivalents) are still below the no-effects level; the project tissue concentrations would still be
below this no-effect level if the higher trophic transfer factor (0.23) reported by McElroy et al.
53
-------
Appendix B
(1991) was used. Therefore, the most relevant aquatic effects information reviewed indicates
that the highest tissue levels accumulated in the dredged material bioaccumulation tests are
below the no-effect level.
Another study that was reviewed considered the carcinogenicity of BaP in rainbow trout
resulting from embryo microinjection (Black et al., 1988). A statistically significant number of
liver neoplasms was found at a concentration of approximately 200,000 ppb, with non-significant
effects at up to one half that concentration. Therefore, using the above across-species UF of 10
and trophic transfer factor of 0.1 results in an aquatic no-effect level of 100,000 ppb. Since this
is several orders of magnitude above the highest tissue concentration of BaP for this project, as
described above (and even the highest BaP-equivalent levels for human health, as discussed
above), this provides additional support for a finding that the test results do not indicate a
potential for undesirable effects to the marine environment due to mutagenic, carcinogenic or
teratogenic contaminants.
Hall and Oris (1991) reported on experiments that exposed fathead minnows to anthracene
during long-term exposures and observed adverse effects on reproduction. The paper reported
that a concentration of anthracene in the tissue of the egg in the range of 3,750 to 8,000 ppb
resulted in no significant effects on egg hatching or survivorship. Using the same approach for
accounting for species-to-species uncertainty and food chain transfer described above and in the
Appendix for Table 1, yields a conservative benthic tissue level of 3,750 ppb. Anthracene tissue
concentrations from the project bioaccumulation tests are well below this level.
iv) Consideration of Potential Carcinogenic and Non-carcinogenic Effects on Human Health
Human health effects screening levels were developed for those chemicals requiring further
evaluation with risk-based methods using conservative estimates of exposure to assess whether
these contaminants would accumulate to levels in fish and shellfish that could lead to significant
adverse effects to humans. The approach assessed consumption of fish and shellfish to derive
conservative estimates of contaminant concentrations in benthic tissue protective of human
health using USEPA standard risk-assessment assumptions and the process described in the
Appendix for Table 1. Table 1, Column 14 lists conservative human cancer protection levels in
benthic organisms for the chemicals which are known or suspected carcinogens that would lead
to a human cancer risk level of 10"4. When the bioaccumulation test results for those chemicals
are adjusted for steady-state (as previously described), the results are below the human cancer
protection levels in Table 1.
Since the analysis used conservative methods, the result represents conservative estimates of
risk, or what are in effect plausible upper-bound estimates. Thus, the true risk is highly unlikely
to be greater than estimated and could be much lower. None of the human health cancer
protection levels were exceeded in the bioaccumulation test results.
The potential for non-cancer impacts can be expressed as a hazard quotient (HQ), which is the
ratio of the average daily intake divided by the toxicological reference dose for the chemical. If
the HQ is less than unity (i.e., 1), an adverse noncarcinogenic effect is highly unlikely to occur.
54
-------
Appendix B
If the HQ exceeds unity, an adverse health impact may occur. The higher the HQ, the more
likely that an adverse noncarcinogenic effect will occur as a result of exposure to the
contaminant in the dredged material after placement. Table 1, Column 15 includes the
noncancer protection levels in benthic organisms for the chemicals requiring further analysis that
are known to cause, or suspected of causing, non-carcinogenic effects, that would result in a
human HQ equal to unity. Those numbers were derived using the conservative assumptions and
source materials described in the Appendix for Table 1. The concentrations of the chemicals
requiring further evaluation were below the non-cancer protection level.
d. Evaluation of Solid Phase Bioaccumulation Results for Dredged Material as a Whole
The evaluation of the testing results performed above indicates that the material does not have a
potential to cause undesirable effects to aquatic marine biota due to chronic adverse effects
including such effects due to mutagenic, carcinogenic, or teratogenic contaminants, or to human
health due to cancer or non-cancer effects from the individual contaminants. That evaluation
includes the information relevant to the eight factors identified in the Green Book for assessing
bioaccumulation test results (USEPA/USACE, 1991). As a final and additional step in the
evaluative process, however, it is appropriate to go beyond assessing the individual test results in
order to look at the results as a whole so as to provide an opportunity for an integrated
assessment of the individual test results (Figure 1, Box d). For example, if a number of the
individual bioaccumulation test results were only marginally at or below the relevant levels of
concern, it is appropriate to consider this and the other relevant factors to evaluate whether, taken
as a whole, the material is unsuitable for placement at the HARS, even though no single
individual test result would indicate that outcome.
As indicated above, the following chemicals of concern were bioaccumulated above reference
for the clam and/or the worm: PAHs, dieldrin, chlordane, trans nonachlor, DDTs, total PCB,
1,4-dichlorobenzene, chromium, copper, lead, and nickel bioaccumulated in both the clam and
worm; arsenic and mercury bioaccumulated only in the clam; and no constituent bioaccumulated
only in the worm. In the case of those contaminants with test results exceeding reference, and
which have regional Matrix levels or dioxin values, none exceeded the relevant Category I value.
For the non-Matrix or non-dioxin contaminants with test results that exceeded reference levels,
except for PAHs, only dieldrin, chlordane, trans nonachlor, and total residual
chlordane/heptachlor bioaccumulated in the project sediments greater than background levels.
Total PAHs did not bioaccumulate to greater than background levels. Dieldrin and acenaphthene
bioaccumulated to about six and eight times background respectively in the worm, all others
were less than three times background. Although some of the contaminants that bioaccumulated
in the tests can be toxicologically important, in no case did they accumulate to toxicologically
important concentrations, even when conservative assumptions were used to evaluate the test
results exceeding reference, as described above. Dieldrin, chlordane, trans nonachlor, PAHs,
arsenic, chromium, copper, lead, and nickel exhibited bioaccumulation test results above
reference which were all below the acceptable human health risk range and acceptable aquatic
effects range using conservative approaches and analyses as described above to evaluate those
test results. Thus, an evaluation of the solid phase bioaccumulation test results for the dredged
55
-------
Appendix B
material as a whole considering the factors in the Green Book (Figure 1, Box d) would not
indicate a different outcome than that shown by the individual test results themselves; i.e., that
the material does not have the potential to cause undesirable effects due to bioaccumulation.
Taking into account all of the above information, it is determined that this material will not cause
undesirable effects due to bioaccumulation as a result of the presence of individual chemicals or
of the solid phase of the dredged material as a whole. Therefore, it is concluded that the solid
phase of the material proposed for HARS placement is classified as Category I under USEPA
Region 2/CENAN general guidance and meets the requirements of 40 CFR '227.6(c)(3),
227.27(b), and 228.15(d)(6).
VI. OVERALL CONCLUSION ON THE PROPOSED PROJECT
Based upon this review of the results of testing of the sediments proposed for dredging and ocean
placement from XXXX, the material meets the criteria for acceptability for ocean placement as
described in Sections 227.6, and 227.27 of the Regulations, is Category I under USEPA
Region 2/CENAN guidance, and is suitable for placement at the HARS under
Section 228.15(d)(6) as Remediation Material.
Placement of this material at the HARS will serve to reduce impacts to acceptable levels and
improve benthic conditions. Sediments in the HARS have been found to be acutely toxic to
appropriate sensitive benthic marine organisms, amphipods and mysids, in laboratory tests,
whereas project sediments used in laboratory acute toxicity tests with the same appropriate
sensitive benthic marine organisms were determined not to be toxic. Placement of project
material over existing toxic sediments would serve to remediate those areas for toxicity. In
addition, by covering the existing sediments in the site with this project material, surface
dwelling organisms will be exposed to sediments exhibiting Category I qualities
(e.g., 2,3,7,8-TCDD bioaccumulation less than 1 pptr) whereas the existing sediments exceed
these Category I levels.
Thus, this material meets the requirements for placement at the HARS as Remediation Material
as described in 40 CFR Section 228.15 (d)(6).
56
-------
PAGE NOT
AVAILABLE
-------
Appendix B
FOOTNOTES FOR TABLE 1:
*: Carcinogenic PAHs.
#: Levels represent the conservative level of protection for the sum of the related compounds and
their metabolites.
na: Not Available
1. A "X" in this column indicates that the analyte concentration in the test sediment is
statistically greater than that of the reference sediment. Means and statistical
comparisons were determined using conservative estimates of concentrations for analytes
that were below the detection limit (USEPA/CENAN, 1997).
2. Conversion factors from 28-day bioaccumulation results to steady state are obtained from
the following sources: for PAHs: from McFarland, 1995; for Aldrin, Dieldrin, Chlordane,
DDT, DDD, and DDE: from Lee and Lincroft et al, 1994; for PCBs: from Pruell et al.,
1993, and Rubinstein et air, for 1,4-Dichlorobenzene: from de Bruijn et al.; for
Endosulfan I, Endosulfan II, Endosulfan Sulfate, Heptachlor and trans nonachlor: from
Syracuse Research Corporation, 1996, and McFarland, 1995; for Heptachlor Epoxide:
from Veith et al., 1979.
3. PAH TEFs taken from: USEPA. 1993; Dioxin TEFs taken from: USEPA. 1989.
4. Toxic equivalence for the carcinogenic PAHs are from USEPA (1993).
5. This value represents the 10"4 cancer risk level for the carcinogenic PAHs. The total
concentration of carcinogenic PAHs is expressed in BaP equivalents (see discussion in
the text of the memo).
6. Cancer risk factor or reference dose are not assigned by USEPA in IRIS (USEPA, 1995).
7. FDA limits are from the USEPA/USACE, 1991.
8. This value represents the benthic level expected to result in a no-effect level for possible
mutagenic and teratogenic effects in fish from exposure to BaP, which is the most toxic
PAH.
9. This value represents the non-specific narcosis effects level (see discussion in Appendix).
This value is compared to the sum of all PAHs measured.
10. Calculations are included in the appendix to Table 1.
11. Means of five tissue replicates calculated using conservative estimates where analytes
were not detected (USEPA/CENAN, 1992); "U" indicates that all five replicates were not
detected.
12. Chemicals for which the bioaccumulation from the dredged material was greater than the
reference but less than the Matrix level are indicated by holding the Matrix level in
Column 20.
59
-------
Appendix B
13. Levels are based on the Regional Dioxin Values.
14. Level is the sum of all dioxin congeners other than 2,3,7,8-TCDD.
15. For this PAH, the no-effect level for possible mutagenic and teratogenic effects in fish is
estimated from exposure to BaP, which is the most toxic PAH.
Cadmium and mercury do not obey steady state kinetics, therefore, no adjustment is made (see
discussion in the text of the memo).
Cancer and non-cancer protection levels, based on inorganic arsenic as contained in EPA's IRIS
database, are not appropriate for evaluating the potential human health impacts of arsenic
bioaccumulation from dredged material, and therefore, are not included in Table 1 (see
discussion in Appendix to Table 1).
On September 27, 2000, the USEPA and USACE signed a memorandum of agreement (MOA)
outlining the steps to be taken to ensure that remediation of the HARS continues in a manner
appropriately protective of human health and the environment. The MOA stated that "in order to
ensure that EPA-R2 and US ACE-NAD continue to incorporate the most up-to-date science in
determining what material is appropriate for placement at the HARS, and to ensure that the
quality of material approved for placement at the HARS meets the remedial purposes of the
HARS designation, one element of the HARS-TEF will immediately be changed. Specifically,
one matrix value in the HARS-specific effects levels will be revised. The PCB matrix value will
be revised from 400 ppb to an interim valued of 113 ppb . .." The MOA further states: "This
interim change shall be effective immediately upon signing of this MOA and be subject to future
review by the re-convened RMW [Remediation Material Workgroup]." (see discussion in
Appendix for Table 1).
60
-------
Appendix B
VII. REFERENCES
Abel, P.D. and V. Axiak. 1991. Ecotoxicology and the marine environment. Ellis Horwood, New York,
pp. 269.
Abernathy, C.O. and E.V. Ohanian. 1992. Non-carcinogenic effects of inorganic arsenic. Environ.
Geochem. Health 14: 35.
Baumann, P.C., W.D. Smith, W.K. Parland. 1987. Tumor Frequencies and Contaminant Concentrations
in Brown Bullheads from an Industrialized River and a Recreational Lake. Transactions of the American
Fisheries Society 116:79-86.
Black, JB, A.E. Maccubbin, and C.J. Johnston. 1988. Carcinogenicity of benzo(a)pyrene in rainbow trout
resulting from embryo micro injection. AquaTox. 13,297-308.
Breteler, R. (ed.). 1984. Chemical Pollution of the Hudson-Raritan Estuary. NOAA Technical
Memorandum NOS OMA 7. National Oceanic and Atmospheric Administration, National Ocean
Service. Rockville, Md.
Brown B., J. Neff. 1993. Bioavailability of Sediment-Bound Contaminants to Marine Organisms.
Report #PNL-8761 UC-000 by Battelle/Marine Sciences Laboratory prepared for the National Ocean
Pollution Program Office, NOAA.
Bryan, G.W. and W.J. Langston. 1992. Bioavailability, accumulation and effects of heavy metals in
sediments with special reference to United Kingdom estuaries: A review. Environmental Pollution 76:
89-131.
Calabrese, A. 1984. "Effects of Long Term Exposure to Silver and Copper on Growth, Bioaccumulation
and Histopathology in the Blue Mussel (Mytilus edulis)." Mar. Envir. Res. 1, 253-274.
Call, D.J., L.T. Brooke, M.L. Knuth, S.H. Poirler, and M.D. Hoglund. 1985. Fish subchronic toxicity
prediction model from industrial organic chemicals that produce narcosis. Environ. Tox. Chem. 4, 335-
341.
CENAN.2000. Sampling/Testing Plan for XXXX Project, New York District Corps of Engineers.
CENAN.2001. Memorandum from M. Greges: for Chief, Technical Support Section, Subject: ADDAMS
Model Evaluation of Application No. 2000-XXXX-OD, XXXXX Project
Charles, JB and J. Muramoto. 1990. Assessment of Contaminants in Sediment and Biota at the Mud
Dump Site, New York Bight. Report No. SAIC-91/7608&256 by Science Applications International
Corp. (SAIC) for USEPA - Region 2.
Corner, E.D.S., R.P. Harris, K.J. Whittle, and P.R. Mackie. (1976). Hydrocarbons in marine zooplankton
and fish. In: Effects of Pollutants on Aquatic Organisms, Lockwood APM (ed), pp. 71- 106. Cambridge
University Press, Cambridge, England.
61
-------
Appendix B
de Bruijn, J., Busser, F., Seinen, W., and Hermens, J. 1989. Determination of octanol/water partition
coefficients for hydrophobic organic chemicals with the "slow stirring" method. Environ. Toxicol.
Chem., 8:499-512.
Dethlefsen, V. 1978. Uptake, retention, and loss of cadmium by brown shrimp. 1978.
Meeresforschung, 26:137 (reported in Giesy et al. 1980).
Feroz, M. And M.A.Q. Khan, 1979. Fate of 14C-cis-chlordane in goldfish, Casassius auratus (L.).
Bulletin of Environmental Contamination and Toxicology 23:64-69.
Finger, E.F., E.F. Little, M.G. Henry, J.F. Fairchild and T.P. Boyle. 1985. Comparison of laboratory and
field assessment of fluorene - Part 1: effects of fluorene on survival, growth, reproduction, and behavior
of aquatic organisms in laboratory tests. In: Validation and Predictability of Laboratory Methods for
Assessing the Fate and Effects of Contaminants in Aquatic Ecosystems. ASTM STP865, Philadelphia,
pp. 120-133.
Fowler, S.W. 1982. Biological transfer and transport processes. In: Pollutant Transfer and Transport in
the Sea. Vol. II, ed. G. Kullengerg, pp. 1-65. CRC Press, Boca Raton, Florida.
Giesy, J.P., Bowling, J.W., and Kania, H.J. 1980. Cadmium and zinc accumulation and elimination by
freshwater crayfish. Arch. Environm. Contam. Toxicol., 9:683-697.
Gobas, F. 1993. A Model for predicting the bioaccumulation of hydrophobic organic chemicals in
aquatic food webs: application to Lake Ontario. Ecological Modeling. 69,1-17.
Hall, A.T. and J.T. Oris. 1991. Anthracene reduces reproductive potential and is maternally transferred
during long-term exposure in fathead minnows. Aquatic Toxicol. 19,249-264.
Hannah, JB, J.E. Hose, M.L. Landolt, B.S. Miller, S.P. Felton, and W.T. Iwaoka. 1982. Benzo(a)pyrene-
induced morphologic and developmental abnormalities in rainbow trout. Arch. Environ. Contam.
Toxicol. 11,167-171.
Holcombe, G.W., G.L. Phipps, and J.T. Fiandt. 1983. Toxicity of selected priority pollutants to various
aquatic organisms. Ecotoxicol. Environ. Safety. 7,400-409.
Hose, J.E., J.B. Hannah, M.L. Landolt, B.S. Miller, S.P. Felton, and W.T. Iwaoka. 1981. Uptake of
benzo(a)pyrene by gonadal tissue of flatfish (family Pleuronectidae) and its effects on subsequent egg
development. J. Toxicol. Environ. Health. 7:991-1000.
Hose, J.E., J.B. Hannah, D. Dijulio, M.L. Landolt, B.S. Miller, W.T. Iwaoka, and S.P. Felton. 1982.
Effects of benzo(a)pyrene on early development of flatfish. Arch. Environ. Contam. Toxicol. 11:167-
171.
Hrudey, S.E., W. Chen, and C.G. Rousseaux. 1996. Bioavailability in environmental risk assessment.
CRC Press, Inc., Boca Raton, Florida, pp.294.
Jarman, W; K. Hobson, W. Sydeman, C. Bacon and E. McLaren. 1996. Influence of Trophic Position
and Feeding Location on Contaminant Levels in the Gulf of the Farallones Food Web Revealed by Stable
Isotope Analysis. Environmental Science & Tech. 30(2):654-660.
62
-------
Appendix B
Landrum P.E., B.J. Eadie, and W.R. Faust. 1988. Toxicity and toxicokinetics for a mixture of sediment
associated polycyclic aromatic hydrocarbons to the .amphipod Pontoporeia hovi. In: Poster Abstracts,
SETAC Ninth Annual Meeting, Society of Environmental Toxicology and Chemistry, Washington D.C.
p. 29.
Lee, R.F., J. Stolzenbach, S. Singer, and K.R. Tenore. 1981. Effects of crude oil on growth and mixed
function oxygenase activity in polychaetes, Nereis sp. In: Biological Monitoring of Marine Pollutants.
Ed. Vernburg, F.A. Calabrese, F. Thurberg, and W. Vernberg. Academic Press, pp. 323-334.
Lee, H., II, Boese, B.L., Pelletier, J., Winsor, M., Specht, D.T., and Randall, R.C., 1989. Guidance
Manual: Bedded Sediment Bioaccumulation Test. USEPA Pacific Ecosystem Branch Bioaccumulation
Team, Newport, OR.
Lee, H., II, Lincroft, A, et al, 1994. Ecological Risk Assessment of the Marine Sediments at the United
Heckathorn Superfund Site. USEPA Pacific Ecosystem Branch Bioaccumulation Team, Newport, OR.,
USEPA Region IX, San Francisco, CA.
Lunde, G. 1977. Occurrence and transformation of arsenic in the marine environment. Environmental
Health Perspectives 19: 47-52.
McCarty, L.S. 1986. The relationship between aquatic toxicity QSARs and bioconcentration for some
organic chemicals. Environ. Toxicol. Chem. 8:1071-1080.
McCarty, L.S. 1991. Toxicant body residues: implications for aquatic bioassays with some organic
chemicals. In: Aquatic Toxicology and Risk Assessment: Fourteenth Volume, ASTM STP 1124; M.A.
Mayes and M.G. Barron, Eds., American Society for Testing and Materials, Philadelphia; pp. 183-192.
McCarty, L.S., D. MacKay, A.D. Smith, G.W. Ozburn, and D.G. Dixon. 1992. Residue-based
interpretation of toxicology bioconcentration QSARs from aquatic bioassays: neutral narcotic organics.
Environ. Tox. Chem. 11:917-930.
McElroy A.E., J.M. Cahill, J.D. Sisson, and K.M. Kleinow. 1991. Relative bioavailability and DNA
adduct formation of Benzo[a]pyrene and metabolites in the diet of the winter flounder. J. Comp.
Biochem. Physiol. 100, 12-29.
McElroy. A.E. and J.D. Sisson. 1989. Trophic transfer of Benzo[a]pyrene metabolites between benthic
marine organisms. Mar. Environ. Res. 28, 265-269.
McElroy, A.E., J.M. Cahill, J.D. Sisson, and K.M.Kleinow. 1991. Relative bioavailability and DNA
adduct formation of Benzo[a]pyrene and metabolites in the diet of the winter flounder. J. Comp.
Biochem. Physiol. 100C:l-2,29-33.
McFarland, V.A. 1995. Evaluation of Field-Generated Accumulation Factors for Predicting the
Bioaccumulation Potential of Sediment-Associated PAH Compounds. USACE - WES Technical Report
D-95-2. July 1995.
Meador J.P., J.E. Stein, W.L. Reichert, and U.Varanasi. 1995. Bioaccumulation of polycyclic aromatic
hydrocarbons by marine organisms. Rev. Environ. Contam. Toxicol. 143, 79-165.
63
-------
Appendix B
Naqvi, S.M., Flagge, C.T., and Hawkins, R.L. 1990. Arsenic uptake and depuration by Red Crayfish,
Procambarus clarkii, exposed to various concentrations of monosodium methanearsonate (MSMA)
herbicide. Bull. Environ. Contam. Toxicol., 45:94-100.
O'Connor, J.M., A.R.Schnitz, and K.A. Squibb. 1988. In vivo kinetics of Benzo[a]pyrene and 7,12-
dimethylbanz[a]anthracene assimilation and metabolism in rainbow trout. Mar. Environ. Res. 24:63-67.
Oladimeji, A.A., Qadri, S.U., and deFreitas, S.W. 1984. Long-term effects of arsenic in rainbow trout,
Salmo gairdneri. Bull. Environ. Contam. Toxicol., 32:732-741.
Parrish, P.R., S.C.Schimmel, D.J. Hansen, J.M. Patrick, J. Forester. 1976. Chlordane: Effects on Several
Estuarine Organisms. Journal of Toxicology and Environmental Health, 1:485-494.
Pruell, R.J., N.I. Rubinstein, B.K. Taplin, J.A. LiVolsi, R.D. Bowen. 1993. Accumulation of
polychlorinated organic contaminants from sediment by three benthic marine species. Arch. Envir.
Contam. Toxicol. 24,290-297.
Rice, D.R., M.M. Babcock, C.C. Brodersen, J.A. Gharrett and S. Korn. 1987. Uptake and depuration of
aromatic hydrocarbons by reproductively ripe pacific herring and the subsequent effect of residues on egg
hatching and survival. In: Pollution Physiology of Estuarine Organisms. Ed. Vernberg, W., A.
Calabrese, F. Thruberg, and F. Vernberg. University of South Carolina Press, pp. 139-154.
Riedel, G.F., Sanders, J.G., and Osman, R.W. 1987. The effect of biological and physical disturbances
on the transport of arsenic from contaminated estuarine sediments. Estuarine, Coastal and Shelf Science,
25:693-706.
Rubinstein, N.I., Lores, E., and Gregory, N.R. 1983. Accumulation of PCBs, mercury and cadmium by
Nereis virens, Mercenaria mercenaria and Palemonetes pugio from contaminated harbor sediments.
Aquatic Toxicol., 3:249-260.
Rubinstein, N. I., R. J. Pruell, B. K. Taplin, J. A. LiVolsi, and C. B. Norwood. 1990. Bioavailability of
2,3,7,8-TCDD, 2,3,7,8-TCDF, and PCBs to marine benthos from Passaic River sediments. Chemosphere,
20, 1097-1102.
Squibb, K.S., J.M. O'Connor, and Kneip, T.J. 1991. Toxics Characterization Report, Module 3.1.
Report prepared by Institute of Environmental Medicine, NY Univ. Medical Center for the NY/NJ Harbor
Estuary Program.
Steimle, F.W., V.S. Zdanowicz, S.L. Cuneff and R. Terranova. 1994. Trace metal concentrations in
common benthic macrofaunal prey form the New York Bight. US National Marine Fisheries Service,
NOAA. Marine Pollution Bulletin. 28, 12, pp. 760-765.
Suedel, B.C., J.A. Boraczek, R.K. Peddicord, P.A. Clifford, and T.M. Dillon. 1994. Trophic transfer and
biomagnification potential of contaminants in aquatic ecosystems. Reviews of Environmental
Contamination and Toxicology 136: 21-89.
Sweeney, B., D. Funk and L. Standley. 1993. Use of the Stream Mayfly Cloeon Triangulifer as a
Bioassay Organism: Life History Response and Body Burden Following Exposure to Technical
Chlordane. Environ. Tox. andChem. 12:115-125.
64
-------
Appendix B
Syracuse Research Corporation, Environmental Science Center. 1996. Experimental Log P
(Octanol/water partition coefficient database). http://esc.syrres.com/~ESC/kowexpdb.htm.
Thomas, L.M. 1987. Letter from Lee M. Thomas, Administrator, U.S. Environmental Protection Agency
to Honorable Henry A. Waxman, Chairman, Subcommittee on Heath and the Environment, Committee on
Energy and Commerce, House of Representatives. May 29, 1987.
USACE. 1981. Final Interpretive Guidance for Bioaccumulation of Petroleum Hydrocarbon, DDT,
Cadmium, and Mercury in the New York Bight. Memorandum from North Atlantic Division Corps of
Engineers to G.R. Tobertson, Deputy Director of Civil Works, Dept. of Army.
USACE. 1994. Bioaccumulation Guidance Values for Selected Contaminants in Sediments and Biota of
the Sandy Hook Reference Site for the New York Bight Apex Mud Dump Site, (draft) Report by Corps of
Engineers Waterways Experiment Station (WES) for the New York District Corps.
USACE. 1995. Trophic transfer and biomagnification potential of contaminants in aquatic ecosystems.
In: Environmental Effects of Dredging Technical Notes. EEDP-01-33. USACE Waterways Experiment
Station (WES).
USEPA/CENAN. 1992. Guidance for Performing Tests on Dredged Material Proposed for Ocean
Disposal. New York District Corps of Engineers, U.S. Environmental Protection Agency -Region 2.
USEPA/CENAN. 1997. (Joint Memorandum) Ocean Disposal of Dredged Material Clarification of Two
Procedural Elements of Interagency Coordination Between USEPA Region 2 and the New York District,
USACE-Treatment of Non-Detects, Chemical Data, and Rules and Responsibilities in Preparation of
Ocean Disposal Regulatory Compliance Memorandum.
USEPA/USACE. 1991. Evaluation of Dredged Material Proposed for Ocean Disposal - Testing Manual.
(Green Book). EPA - 503/8-91/001.
USEPA. 1980a. Water quality criteria documents: availability. Federal Register, Vol. 45, No. 231.
November 28, 1980.
USEPA. 1980b. Ambient Water Quality Criteria for Aldrin/Dieldrin; EPA 440/5-80-019; December
1980.
USEPA. 1980c. Ambient Water Quality Criteria for Chlordane; EPA 440/5-80-027; October 1980.
USEPA. 1980d. Ambient Water Quality Criteria for Heptachlor; EPA 440/5-80-052; October 1980.
USEPA. 1980e. Ambient Water Quality Criteria for Endosulfan; EPA 440/5-80-046; October 1980.
USEPA. 1980f. Ambient Water Quality Criteria for Dichlorobenzenes; EPA 440/5-80-039; October 1980.
USEPA. 1984a. Ambient Water Quality Criteria for Lead - 1984; EPA 440/5-84-027; January 1985.
USEPA. 1984b. Ambient Water Quality Criteria for Copper - 1984; EPA 440/5-84-031; January 1985.
USEPA. 1985a. Ambient Water Quality Criteria for Chromium - 1984; EPA 440/5-84-029; January 1985.
65
-------
Appendix B
USEPA. 1985b. Guidelines for Deriving Numerical National Water Quality Criteria for the Protection of
Aquatic Organisms and Their Uses. NTIS # PB85-227049.
USEPA. 1985c. Ambient Water Quality Criteria for Arsenic - 1984; EPA 440/5-84-033; January 1985.
USEPA. 1986. Ambient Water Quality Criteria for Nickel - 1986; EPA 440/5-86-004; September 1986.
USEPA. 1987a. National primary drinking water regulations - synthetic organic chemicals; monitoring
for unregulated contaminants; final rule. Federal Register, Vol. 52, No. 130, 25690. July 8,1987.
USEPA. 1987b. Ambient Water Quality Criteria for Zinc - 1987; EPA 440/5-87-003.
USEPA. 1988. Guidance for state implementation of water quality standards for CWA section
303(c)(2)(B). Federal Register, Vol. 54, No. 346. November 12,1988.
USEPA. 1989. Interim Procedures for Estimating Risks Associated with Mixtures of Chlorinated
Dibenzo-p-Dioxins and -Dibenzofurans (CDDs and CDFs) and 1989 Update. U.S. Environmental
Protection Agency, Risk Assessment Forum, Washington, DC. EPA/625/3-89/016.
USEPA. 1991. National Primary Drinking Water Regulations; Final Rule. 40 CFR Part 141. January 30,
1991.
USEPA. 1992a. Draft Ambient Water Quality Criteria for Silver.
USEPA. 1992b. Water quality standards; establishment of numeric criteria for priority toxic pollutants;
states compliance. Federal Register, Vol. 57: 60848.
USEPA. 1993. Provisional Guidance for Qualitative Risk Assessment of Polycyclic Aromatic
Hydrocarbons. EPA/600/R-93/089.
USEPA. 1994. Final clarification of suspended particulate phase bioaccumulation testing requirements
for material dumped in ocean waters. Federal Register Vol. 59: 52650. October 18, 1994.
USEPA. 1995. On-Line. Integrated Risk Information System (IRIS). Cincinnati, OH: Office of
Research and Development, Environmental Criteria and Assessment Office.
USEPA. 1996a. Ocean dumping testing requirements; final rule. Federal Register, Vol. 61, No. 190,
51196. September 30, 1996.
USEPA. 1996c. Memo to File from A. Lechich. Subject: Issues Regarding Exposure and Uptake
Mechanisms for PAHs. (Discussion with V. McFarland). December 5, 1996.
USEPA. 1996d. Memo to File from A. Lechich. Subject: Discussion of PAHs With Regard to East River
Memo. (Discussion with D. Hansen). December 5, 1996.
USEPA. 1996e. Memo to File from C. Vogt. Subject: Acceptable Levels of Lead: East River
Bioaccumulation Tests. December 13, 1996.
USEPA. 1996f. Battelle Body Burden Study. Report prepared by Battelle Ocean Sciences, Duxbury,
MA, for USEPA - Region II.
66
-------
Appendix B
USEPA. 1997a. Memo to File from A. Lechich. Subject: Summary of Dioxin Risk Evaluation
Approach. March 15, 1997.
USEPA. 1997b. Contaminants in Polychaetes from the Mud Dump Site and Environs. March 4,1997.
Report prepared by Battelle Ocean Sciences, Duxbury, MA, for USEPA - Region II.
USEPA. 1997c. Supplemental to the Environmental Impact Statement on the New York Dredged
Material Disposal Site Designation for the Designation of the Historic Area Remediation Site (HARS) in
the New York Bight Apex. U.S. Environmental Protection, Region 2, New York, May 1997.
USEPA. 2000. Memorandum of Agreement: among the Department of the Army, the Environmental
Protection Agency, and the U.S. Army Corps of Engineers. To Strengthen Environmental Protection of
the Ocean Environment and to Promote Economic Progress in the Port of New York and New Jersey.
September 27, 2000.
USEPA. 2000a. Memorandum to the File from Douglas Pabst. Subject: Modification of the Matrix
Value for PCB in Worm Tissue. September 27, 2000.
Varanasi U., J.E. Stein, and M. Nishimoto. 1989. Biotransformation and disposition of polycyclic
aromatic hydrocarbons (PAH) in fish. In: Varanasi U. (ed) Metabolism of Polycyclic Aromatic
Hydrocarbons in the Aquatic Environment. CRC Press, Boca Raton, Fl, pp 94-149.
V-Balogh, K., and Salanka, J. 1984. The dynamics of mercury and cadmium uptake into different organs
of Anodonta cygnea L. Water res., 18(11):1381-1387.
Veith, G. D., DeFoe, D.L., and Bergstedt, B.V. 1979. Measuring and estimating the bioconcentration
factor of chemicals in fish. J. Fish. Res. Board Can., 36(9): 1040-1048.
Verschueren, K. 1983. Handbook of Environmental Data on Organic Chemicals, second edition. Van
Nostrand Reinhold Company.
Ward, G.S., P.R. Parrish, and R.A. Rigby. 1981. Early life stage toxicity tests with a saltwater fish:
Effects of eight chemicals on survival, growth, and development of sheepshead minnows (Cvprinidon
variegatus). J. Toxicol. Environ. Health. 8:225-240.
Whittle K.J., J. Murray, P.R. Mackie, R. Hardy, and J. Farmer. 1977. Fate of hydrocarbons in fish. In:
Petroleum Hydrocarbons in the Marine Environment. Cons. Intern. Explor. Mer. Vol. 171, Mclntyre
A.D. and Whittle K.J. (eds), pp 139-142. Charlottenlund Slot, Denmark.
WHO. 1993. Guidelines for Drinking Water Quality. World Health Organization. Geneva.
67
-------
Appendix B
This page intentionally left blank.
68
-------
Appendix B
Appendix for Table 1
I. CONSIDERATION OF ECOLOGICAL EFFECTS
A. Potential for ecological effects based on Water Quality Criteria (Column 19)
The potential for ecological impacts due to bioaccumulation of several compounds of concern
was evaluated by calculating a Water Quality Criterion Tissue Level (WQCTL). The WQCTL is
calculated by multiplying the Clean Water Act Section 304(a)(1) Federal water quality criterion
chronic value (CV) for the chemical by the empirically determined bioconcentration factor
(BCF) for the chemical for a representative marine organism (Lee et al., 1989). A BCF is the
ratio of the concentration of a contaminant in an organism to the concentration of the
contaminant in water. Thus, the WQCTL represents the tissue concentration that would be
expected in an organism exposed to water containing the chemical at the C V concentration. This
level is set to protect 95% of all tested organisms included in the water quality criterion database,
thus representing a conservative level of protection (USEPA, 1985b). Table 1 lists the calculated
WQCTLs. Sources of CVs and BCFs are from USEPA ambient water quality criteria documents
(USEPA 1980b, 1980c, 1980d, 1980e, 1980f, 1984a, 1984b, 1985a, 1985c, 1986, 1987b and
1992a) and Calabrese (1984)(for silver). Calculations are shown in attachment A.
Several pesticides were evaluated based on the sum of their primary constituents and associated
metabolites (e.g., total chlordane, total endosulfan, and total DDT). Alpha(trans)-chlordane,
trans nonachlor, heptachlor and heptachlor epoxide represent the primary components of
technical chlordane and its metabolites found in the tissue of aquatic organisms (Jarman, et. al.,
1996; Verschueren, 1983; Sweeney, et. al., 1993). These constituents are summed as total
chlordane as is consistent with current practice for chlordane (Jarman, et. al., 1996) and total
DDT. The WQCTL for total chlordane was calculated using the WQC for chlordane as a
conservative level of protection. While water quality criteria exist, and WQCTLs can be
calculated, for heptachlor (133 ppb) and chlordane (64 ppb), the sum total chlordane is compared
to the WQCTL for chlordane in order to be more environmentally conservative. The chlordane
WQCTL provides a conservative level of protection as indicated by published residue effects
levels (Sweeney, et. al., 1993; Bauman, et. al., 1987; Feroz, et. al., 1979; Parrish, et. al., 1976).
Consistent with the above approach, the tissue concentration for endosulfan I, endosulfan II and
endosulfan sulfate were also summed as total endosulfan and compared to the WQCTL for total
endosulfan.
The WQCTLs were also calculated for all metals of concern which don't have Matrix values.
For total chromium, the WQCTL was calculated based on chromium(VI), which is substantially
more toxic than chromium (III) and elemental chromium in order to provide a conservative level
of environmental protection.
69
-------
Appendix B
B. Potential for ecological effects based on PAH toxicity (Column 19).
The Critical Body Residue (CBR) approach described by McCarty (1991) was used to derive
values for use in evaluating the potential impacts of PAHs accumulated in the dredged material
bioaccumulation test organisms. CBRs are concentrations of chemical residues in organisms
which elicit a deleterious biological response associated with narcosis, which is the primary non-
cancer effect of PAHs. Narcotic responses measured can be acute (e.g., immobilization or death)
or chronic endpoints (e.g., reduced reproduction, fecundity or growth). CBRs are represented as
the ratio of the mass of toxicant to the mass of the organism, such as millimoles or micrograms
of toxicant per kilogram (mmole or ug/kg) of organism. For the narcosis endpoint, each
molecule of individual PAH congeners are generally equipotent, thus the total PAH
concentration is compared to the CBR. For example, a 400 ppb dose of naphthalene would elicit
a similar toxicity response as 400 ppb of fluorene; if both chemicals are present together at these
concentrations, then the dose would equal 800 ppb.
McCarty (1991) states that an average critical body residue of400,000 - 1,200,000 ppb can be
used as an estimate for acute effects for a narcosis-producing chemical (e.g., PAHs) on fish
populations. (Note: McCarty reports the CBR in units of millimoles per kilogram; this value has
been converted to ppb for PAHs using the average molecular weight of the PAHs analyzed in the
bioaccumulation test). Chronic effect critical body residues can be estimated by applying an
acute to chronic ratio of 10 to the acute CBR (McCarty, 1986; Call et al., 1985). Therefore, the
chronic critical body residue for PAHs can be estimated at 40,000 - 120,000 ppb of PAHs in
organism tissue, and Table 1 thus uses the 40,000 ppb level.
These CBRs were based on fish data. The use of CBRs based on fish toxicity represents a
conservative estimate of potential toxicity due to exposure to dredged material because: (1) it is
extremely unlikely that a fish would get its whole diet from the HARS; and (2) fish are generally
more sensitive than the benthic organisms in direct contact with the dredged material placed at
the HARS (e.g., Landrum et al. (1988) estimated an acute CBR for crustaceans of 800,000 ppb -
42,000,000 ppb).
C. Potential ecological impacts of mutagenic, carcinogenic and teratogenic PAHs
(Column 19)
USEPA and the USACE reviewed eleven scientific journal articles to obtain information about
the potential for adverse effects to the marine environment due to the observed bioaccumulation
of PAHs in the marine worm, Nereis virens, and the clam, Macoma nasuta. These articles
reported the results of laboratory experiments that sought to relate the concentration of a
contaminant(s) in water, as injected doses, or tissue concentrations, to mutagenic, carcinogenic,
teratogenic and/or reproductive effects to fish. These studies all used fish species which are
considered to be among the most sensitive organisms in the marine environment to exhibit the
above effects (USEPA, 1996c). In addition, most of these studies focussed on the PAH most
believed to cause such effects for which there is data, benzo(a)pyrene (BaP). One study
(Breteler, 1984), discussed the possible sources and distribution of PAHs in the Hudson/Raritan
70
-------
Appendix B
estuary, and ranked the threat of PAHs to aquatic biota and humans. The main threat was
believed to be carcinogenicity, with a greater threat ranking assigned to humans than biota.
However, Breteler (1984) did not provide specific effects-based levels that could be used in the
following analysis. Two articles evaluated the effects of crude oil, and thus were not useful for
evaluating the effects of specific PAHs measured in the bioaccumulation test (Rice et al., 1987;
Lee et al., 1981). Three studies considered the effects of specific PAHs, but did not synoptically
measure tissue concentrations in the organisms (Ward et al., 1981; Holcombe et al., 1983; Finger
et al., 1985) and were not used, because the lack of tissue data for these studies makes their
utility in evaluating the tissue concentration resulting from the dredged material bioaccumulation
tests highly uncertain.
The remaining five papers reported measured tissue concentrations and observed reproductive
effects in organisms exposed to PAH-spiked water. One article reported the tissue
concentrations of adult fish and the observed effect on survival and health of the fish's offspring.
Hose et al. (1981) reported that adult English sole injected with benzo(a)pyrene (BaP)
accumulated the chemical in the gonad and mature gametes. The amount of BaP taken up by the
ovary ranged from 16,800 to 49,700 ppb. Two samples of ripe eggs contained 51,200 and
263,000 ppb of BaP and its metabolites. No adverse effects were reported for these
concentrations. Hose et al. (1981) also reported the results of injecting female flathead sole with
BaP. Adverse effects to egg hatching success were reported for each female. The paper does not
report tissue concentrations in either the parent fish or the egg of the flathead sole. Effects on
reproductive success were reported but could only be correlated with the external dose injected
into the parent. Therefore, since concentrations and effects were not synoptic in this report, it
was not useful in the evaluation of the dredged material bioaccumulation results.
Three papers reported the results of experiments which measured fish egg or alevin
concentrations of BaP and associated reproductive or carcinogenic effects (Hose et al, 1982;
Hannah et al., 1982; Black et al., 1988). Hose et al. (1982) exposed three species of sole, sand
sole, English sole, and flathead sole, to BaP-spiked water. Tissue concentrations of 2,100 ppb
were measured in sand sole on day 6 (24 hours after hatching) and were associated with reduced
hatching success. However, we did not consider the results to be appropriate for use in setting
effects levels because they may have been compromised by the methods of replication used in
the experimental design.
Hannah et al. (1982) estimated a concentration of BaP in tissue that caused abnormalities in
development of rainbow trout eggs, using aqueous exposures and actual measured tissue
concentrations in alevin tissues. An exposure to a 2.4 ppb mean aqueous BaP concentration
accumulated an average of 12,340 ppb in alevins. This concentration was associated with an
increase in percentage of abnormalities from approximately 6% at lower water concentrations
(0.08, 0.21,0.37 and 1.48 ppb) to approximately 13% at higher concentrations. From 0.08 to
1.48 ppb in the water, there were no increasing effects exhibited, therefore, the effects were
apparently Areal@ (i.e., significantly greater than the threshold effect level of 6%) only at the
aqueous 2.4 ppb concentration. The Hannah et al. (1982) study is considered the most reliable
study for this evaluation since it used exposure series and measured tissue concentrations
71
-------
Appendix B
associated with observed effects, and therefore allows for the calculation of a no-effects level
directly from the measured results.
In applying these studies to evaluations of dredged material, consideration must be given to
uncertainties in converting these kinds of results to concentrations protective of other biota.
Three uncertainties needing to be considered are: (1) those associated with converting effect to
no-effect concentrations, (2) across-species uncertainties, and (3) uncertainties in estimating the
dose of contaminants to which the organism is exposed. These uncertainties are discussed
below.
With respect to uncertainty when converting effect to no-effect concentrations, an uncertainty
factor of one order of magnitude is often used when only an effect measure is reported.
However, in Hannah et al. (1982), the no-effect level can be estimated to be the next lowest
concentration below the lowest-observed effect level, since the range of concentrations below
this level did not exhibit significantly different responses. In this case, the no-effect level
occurred at the water exposure concentration of 1.48 ppb. Although a tissue concentration was
not measured at the 1.48 ppb water concentration, it can be calculated from the concentration
measured at the effect level (i.e., the no-effect water concentration (1.48 ppb) is close to 65
percent of the observed effect concentration (2.4 ppb) so the no-effect tissue concentration
should be about 65 percent of the lowest-observable effect tissue concentration (0.65 x 12,340
ppb = 8,021 ppb)). Thus, a factor to adjust these data from lowest observed effect tissue
concentration to the calculated no-observed effect tissue concentration is obtained directly from
the data.
There can also be uncertainty as to the proximity to the site of toxic action in the organism that a
dose or concentration is measured, and with respect to species-to-species variability. Hannah
(1982) reported dose concentrations in the tissue and, therefore, there is no need to account for
variability associated with the large uncertainties encountered in typical water-only exposure
studies where the actual concentration at the site of toxic action is unknown. When measured in
the tissues, as was done for this project, concentrations of narcotic chemicals causing effects
(i.e., critical body residues, CBRs) in aquatic organisms are reported to range only from 1.4 to 21
umoles/g wet weight (a factor of about one order of magnitude) for organisms as diverse as
insects, crustaceans, and fish (McCarty et al., 1992). Therefore, from a tissue concentration
perspective, the species-to-species uncertainty factor appropriate for both total PAHs operating
as narcotics and individual PAHs having teratogenic effects would be one order of magnitude, or
a value of 10 (USEPA, 1996d).
In summary, a factor of 10 (representing species to species uncertainty) is an appropriate UF to
use in these evaluations. Also, as described in memo SectionV, subsection C2(c)(i) above,
Brown and Neff (1993) show that trophic transfer of PAHs up the food chain to fish decrease
tissue levels by over an order of magnitude. Given this data and the fact that these studies
included fish that spent 100% of their time feeding in the test sediment, whereas this would be
highly unlikely to occur at an ocean site, a trophic transfer factor of 0.1 is used in this analysis.
Applying this UF of 10 and a trophic transfer factor of 0.1 to the no-effects level for BaP
72
-------
Appendix B
calculated from Hannah et al. (1982), as discussed above (8,021 ppb) results in a no-effect level
for BaP of approximately 8,000 ppb in benthic tissue.
II. CONSIDERATION OF POTENTIAL EFFECTS ON HUMAN HFAITH ^Columns
14 and 15)
Human effects screening levels were developed with risk-based methods using conservative
estimates of exposure to assess whether these contaminants would accumulate to levels in fish
and shellfish that could lead to significant adverse effects to humans. The approach assessed
consumption of fish and shellfish to derive conservative estimates of contaminant concentrations
in benthic tissue protective of human health using the following USEPA standard risk-
assessment assumptions: a 70-kilogram adult eats 6.5 grams of fish and shellfish per day over a
70-year lifetime. This assessment considered potential for both cancer and non-cancer effects in
humans. USEPA IRIS (USEPA, 1995) and effects information from USEPA's National Toxics
Rule (USEPA, 1992b) were used in the human health assessment to calculate acceptable levels
in fish and shellfish to protect human health. Trophic transfer factors, as discussed earlier, were
then used to convert these fish and shellfish levels into benthic tissue concentrations.
For regulatory purposes, USEPA utilizes 10"4 to 10"6 (one in ten thousand to one in one million)
as an acceptable incremental risk range for activities with potential for causing cancer in human
beings (USEPA, 1980a; USEPA, 1988; USEPA, 1987a; Thomas, 1987; USEPA, 1991). USEPA
considers a cancer risk within this range to be safe and protective of public health. This is
supported by the World Health Organization Guidelines for Drinking Water Quality (WHO,
1993), where it selected a 10"5 guideline value, and then explained that the application could vary
by a factor of ten (e.g., 10"4 to 10"6). Since this analysis uses conservative methods, the results
represent conservative estimates of risk, or what are in effect plausible upper-bound estimates.
Thus, the true risk is highly unlikely to be greater than estimated and could be much lower.
Table 1, Column 14 lists human cancer protection levels in benthic organisms for chemicals
which are known or suspected carcinogens that would lead to a human cancer risk level of 10"4.
For PAHs, this analysis used BaP-equivalents derived from the toxic equivalence factor for each
carcinogenic PAH (from USEPA (1993); note: these factors are listed in Column 11 of Table 1
for each of the compounds).
The potential for non-cancer impacts can be expressed as a hazard quotient (HQ), which is the
ratio of the average daily intake divided by the toxicological reference dose for the chemical. If
the HQ is less than unity (e.g., 1), an adverse noncarcinogenic effect is highly unlikely to occur.
If the HQ exceeds unity, an adverse health impact may occur. The higher the HQ, the more
likely that an adverse noncarcinogenic effect will occur as a result of exposure to the
contaminant in the dredged material after placement. Table 1, Column 15 lists noncancer
protection levels in benthic organisms for the chemicals that are known to cause, or suspected of
causing, non-carcinogenic effects, that would result in a human HQ equal to unity. Those
numbers were derived using the conservative assumptions and source materials described in the
introductory paragraph to this section.
73
-------
Appendix B
For the following compounds, the following special considerations were used in evaluating the
results in Table 1.
Metals
No reference dose has been established for lead. EPA has adopted a blood lead level of lOug/dl
as the level of concern and EPA policies are that regulatory programs should seek to minimize
the number of children with blood lead levels above a target of 10 ug/dl (Final Rule for Lead and
Copper NPDWR, 56FR26468, June 7, 1991), and this value was used to calculate the effects
level in Table 1 (see USEPA 1996e).
When interpreting the importance of arsenic tissue concentrations for human health,
consideration was given to the arsenic form present (i.e., inorganic vs. organic). Arsenic is found
in marine organisms as an organic complex which includes such compounds as arsenobetaine
and arsenocholine (Abel and Axiak, 1991). Organic arsenic in the tissues of aquatic organisms is
not metabolized by predators or humans and is readily eliminated from the body through
excretion (Hrudey et al., 1995). As a result, the toxicity of organic arsenic ingested from seafood
is low and appears to pose no significant hazard (Abernathy and Ohanian, 1992). For this
reason, cancer and non-cancer protection levels, based on inorganic arsenic as contained in
EPA's IRIS database, are not appropriate for evaluating the potential human health impacts of
arsenic bioaccumulation from dredged material, and therefore, are not included in Table 1.
Pesticides
Alpha(trans)-chlordane, trans nonachlor, heptachlor and heptachlor epoxide represent the
primary components of technical chlordane and its metabolites found in the tissue of aquatic
organisms (Jarman, et. al., 1996; Verschueren, 1983; Sweeney, et. al., 1993). These constituents
are summed as total chlordane as is consistent with current practice for chlordane (Jarman, et. al.,
1996) and other pesticides (e.g., total DDT). Total chlordane is evaluated using the 10"4 cancer
risk level and non-cancer level for heptachlor epoxide, which has the greatest potency of the
chlordane constituents or metabolites. Similarly, endosulfan I, endosulfan II, and endosulfan
sulfate are summed and the total is compared to the conservative non-cancer protection level for
endosulfan.
PAHs
For PAHs, this analysis used BaP-equivalents derived from the toxic equivalence factor for each
carcinogenic PAH (from USEPA (1993); note: these factors are listed in column 11 of Table 1
for each of the compounds).
74
-------
Appendix B
III. REVISION OF THE WORM MATRIX VALUE FOR PCBS (Column 20)
This PCB interim revision of the worm Matrix value was made effective by the MOA signed by
the USEPA and USACE on September 27, 2000, and reflects EPA Region 2's interpretation and
ongoing review of the science associated with responding to the peer review comments. This
risk-based value was calculated using exposure assumptions chosen to represent specific
conditions associated with consuming fish from the HARS. Technical derivation of this HARS-
Specific Value is described below.
Derivation of HARS-Specific Total PCB Values
The following tissue residue levels have been developed by EPA Region 2 for use in evaluating
the potential for adverse effects due to bioaccumulation of polychlorinated biphenyls (PCBs) by
benthic organisms exposed to sediments proposed for use as Remediation Material at the HARS:
Human Health Non-Cancer*
113 ppb
*guidelines are reported in parts per billion (ppb) of total PCBS, wet weight.
1. HUMAN HEALTH RISK
For human health, the Value (for protection from non-cancer effects)was back-calculated using
standard risk assessment equations (EPA 1987, 1989) to identify tissue concentrations associated
with a non-cancer hazard quotient of one (see Figure 1).
Table 1 presents a summary of assumptions used by EPA in deriving the HARS-Specific Values
for PCBs that were derived for the protection of human health. For the purpose of evaluating
human health risks of PCBs accumulated by benthic organisms at the HARS, it was assumed
that: 1) fish consumption is the pathway of concern by which humans are exposed to
contaminants in dredged material proposed for use as Remediation Material at the HARS; and 2)
that the fish consumed would be exposed to contaminants through trophic transfer from
contaminated benthic invertebrate prey. (USEPA. 2000a)
Table 1. Assumptions Used to Develop Human Health HARS-Specific Values
Compound
Reference
Trophic
Whole
Seafood
Site
Dose
Transfer
body:
Consumption
Use
Filet
(g/day)
Factor
TOTAL PCBS
0.02
3
1.35
7.2
77.7%
75
-------
Appendix B
Attachment A: Tissue Concentration is Calculated Using BCF * Water Quality Criteria
(WQC) Ambient Aqueous Concentration, and Assuming 1 kg = 1L.
Compound
Ambient
Cone.
(ug/L)'
BCF
Tissue Cone.
(ug/Kg)
Remarks
Aldrin
0.13
2,300
299
WQC was reduced by a factor of 10 to account for chronic effects; BCF
estimate is based on Dieldrin since Aldrin rapidly transformed to Dieldrin
in the environment; BCF is based on 1.1% lipid level for marine fish, Spot
(Leiostomus xanthurus).
Dieldrin
0.0019
2,300
4.37
BCF is based on 1.1% lipid level for marine fish, Spot (Leiostomus
xanthurus).
Total Chlordane
0.004
16,000
64
Total Chlordane includes alpha-chlordane, trans nonachlor, heptachlor,
heptachlor epoxide; WQC for Chlordane is used for Total Chlordane; BCF
is based on 3.6% lipid level for sheepshead minnow (Cyprinodon
variegatus).
Total Endosulfan
0.0087
328
2.85
Total Endosulfan includes endosulfan I, endosulfan II and endosulfan
sulfate; WQC for Endosulfan is used for Total Endosulfan; BCF is based
on 3.6% lipid level for sheepshead minnow (Cyprinodon variegatus).
1,4-Dichlorobenzene
197
60
11820
Ambient conc. is based on lowest observed effect level (LOEL) for
saltwater species from WQC, and reduced by a factor of 10 to account for
chronic effects; BCF is based on the whole body for bluegill (Lepomis
macrochirus).
Arsenic
36
350
12600
Ambient conc. is based on the saltwater criteria continuous conc. for
arsenic (III); BCF is based on the Eastern Oyster (Crassostrea virginica).
Chromium
50
236
11800
Ambient conc. is based on Chromium (VI) since it is substantially more 1
toxic than Chromium (111); BCF is based on polychaete worm.
Copper
2.9
3,300
9570
WQC based on a hardness value of 100; BCF is based on soft shell clam.
Lead
8.5
1,400
11900
Ambient conc. based on saltwater criteria continuous conc.; BCF based on
the Eastern Oyster (Crassostrea virginica).
Nickel
8.3
458
3802
Ambient conc. is based on saltwater criteria continuous conc.; BCF is
based on the Eastern Oyster (Crassostrea virginica).
Silver
0.23
6,500
1495
Water Quality Criterion (WQC) was reduced by a factor of 10 to account
for chronic effects; BCF is based on the Blue Mussel (Mytilus edulis).
Zinc
86
17,640
1517040
Ambient conc. based on saltwater criteria continuous conc.; BCF based on
Eastern Oyster (Crassostrea virginica).
'The following documents were used to obtain the water quality criteria values.
Calabrese, A. 1984. "Effects of Long Term Exposure to Silver and Copper on Growth, Bioaccumulation and Histopathology
in the Blue Mussel (Mytilus edulis)" Mar. Envir. Res. 1,253-274.
USEPA. 1980b. Ambient Water Quality Criteria for Aldrin/Dieldrin; EPA 440/5-80-019; December 1980.
USEPA. 1980c. Ambient Water Quality Criteria for Chlordane; EPA 440/5-80-027; October 1980.
USEPA. 1980d. Ambient Water Quality Criteria for Heptachlor; EPA 440/5-80-052; October 1980.
USEPA. 1980e. Ambient Water Quality Criteria for Endosulfan; EPA 440/5-80-046; October 1980.
USEPA. 1980f. Ambient Water Quality Criteria for Dichlorobenzenes; EPA 440/5-80-039; October 1980.
USEPA. 1984a. Ambient Water Quality Criteria for Lead - 1984; EPA 440/5-84-027; January 1985.
USEPA. 1984b. Ambient Water Quality Criteria for Copper - 1984; EPA 440/5-84-031; January 1985.
USEPA. 1985a. Ambient Water Quality Criteria for Chromium - 1984; EPA 440/5-84-029; January 1985.
USEPA. 1985c. Ambient Water Quality Criteria for Arsenic - 1984; EPA 440/5-84-033; January 1985.
USEPA. 1986. Ambient Water Quality Criteria for Nickel - 1986; EPA 440/5-86-004; September 1986.
USEPA. 1987b. Ambient Water Quality Criteria for Zinc - 1987; EPA 440/5-87-003.
USEPA. 1992a. Draft Ambient Water Quality Criteria for Silver.
76
-------
Appendix B
Attachment B: Benthic Cancer Protection Level Calculations for the Protection of Human
Health from the Consumption of Fish Exposed to Dredged Material at the
Historic Area Remediation Site
Basis61:
10"4 Benthic Tissue Level (ug/kg) = T10"4 Cone, in Fish 1 x TWhole Bodv/fillet Factor CI.35Ylfo5
—
Trophic Transfer Factor*™2
10"4 Cone, in Fish (ug/kg) = Toxicological Dose (ug/dav)
[Seafood Consumption (6.5 g/day)fo3] x [10~3kg/g]
Toxicological Dose (ug/dav) =fRisk Level CIO"4)! x IBodv Weight (70 kg)*"3! x 1103 ug/mgl
Potency Factor, q/ (kg-day/mg)614
Cancer Potency
Factor
qi*
(kg-day/mg)
Acceptable
Concentration in Fish
(ug/kg)
Trophic
Transfer
Factor
Benthic
Protection
Level
(«g/kg)
Pesticides
Aldrin
17
63
2.6
33
Chlordane
1.3
828
2.3
486
Dieldrin
16
67
1.4
65
Heptachlor
4.5
239
2.7
120
Heptachlor
epoxide
9.1
118
1.4
114
Industrial
Organics
1.4-
Dichlorobenzene
0.024
44,872
1
60,577
PAHs
Benzo(a)pyrene
7.3
147
0.1
2,000
METALS
Arsenic
1.5
718
3
323
77
-------
Appendix B
Attachment C: Benthic Non-Cancer Protection Level Calculations for the Protection of
Human Health from Consumption of Fish Exposed to Dredged Material at
HARS
Basis6":
Benthic Tissue Level fus/lce) = TConc. in Fish 1 x rWhole Bodv/fillet Factor d.35Vlfil5
Trophic Transfer Factor1"2
Cone, in Fish fue/ke") = Toxicoloeical Dose fue/dav)
[Seafood Consumption (6.5 g/day)fr3] x [10"Jkg/g]
Toxicological Dose (ug/day) = [Reference dosefn4] x [Body Weight (70 kg)*"3]
Reference Dose
(fig/kg-day)
Acceptable Concentration
in Seafood
(Hg/kg)
Trophic
Transfer
Factor
Benthic
Protection Level
(fig/kg)
Compound
Arsenic
0.3
3,231
3
1,454
Chromium
5
54,000
1
73,000
Copper
37.1
400,000
1
540,000
Nickel
20
215,000
1
290,000
Silver
5
54,000
1
73,000
Zinc
300
3,230,769
1
4,361,538
Aldrin
0.03
323
2.6
167
Chlordane
0.06
592
2.3
350
Dieldrin
0.05
538
1.4
518
Endosulfan
6
64,615
1
87,231
Heptachlor
0.5
5,385
2.7
2,692
Hept. epoxide
0.013
140
1.4
135
Acenaphthene
60
650,000
0.1
8,775,000
Fluorene
40
430,000
0.1
5,805,000
Phenanthrene
300
3,230,000
0.1
43,605,000
Anthracene
300
3,230,000
0.1
43,605,000
Fluoranthene
40
430,000
0.1
5,805,000
Pyrene
30
325,000
0.1
4,387,000
78
-------
Appendix B
NOTES:
fnl Human health cancer and non-cancer assessments adapted from Guidance for Assessing Chemical Contaminant Data for
use in Fish Advisories: Volume II: Risk Assessment and Fish Consumption Limits. U.S. Environmental Protection
Agency, EPA823-B-94-004, Office of Science and Technology, Washington, DC, June 1994.
fn2 Trophic transfer factors were calculated by Mr. Lawrence Burkhard, EPA Duluth, using the food chain transfer model
developed by Gobas (1993).
fn3 Default values were taken from EPA's national toxics rule for setting water quality criteria, USEPA (1992b).
fn4 Cancer potency factors and non-cancer reference doses are taken from USEPA (1995).
fnS The acceptable concentration in seafood is defined on the basis of the fillet or edible portion for humans. Trophic
transfer, however, was defined on the basis of whole body characteristics, including lipid concentrations. Experience in
New York State indicates a whole body to fillet ration ranging from 1.2 to 1.5 is applicable to lipophilic substances. The
mid range value of 1.35 is used in this analysis.
79
-------
Appendix B
This page intentionally left blank.
80
-------
APPENDIX C
Estimation of Total PCB Residue Based on
Reported Concentration of 22 Congeners: Review of
NY/NJ Harbor Regional Data
-------
Appendix C
Estimation of total PCB residue based on reported concentration of 22 congeners: Review
of NY/NJ Harbor Regional Data
Three data sets that measured an extensive list of PCB congeners (i.e., 79 or more congeners) in
environmental media in the New York/New Jersey Harbor area were reviewed by EPA Region 2
to determine whether total PCB concentrations could be reliably estimated from the 22
congeners that are routinely quantified in the dredging program. These studies (Durell and
Lizotte 1998; Battelle 1998; and EPA 1992) were conducted on sewage influents/effluents,
dredged material, and bottom sediments, respectively. In all cases, the subset of 22 congeners
that is routinely quantified in the dredging program was highly correlated with, and predictive of
(R2 > 0.97), the total PCB concentrations actually measured in those studies.
In addition, three data sets that measured an extended list of PCB congener (i.e., 106 or more
congeners) residues in tissues of organisms sampled from the NY/NJ Harbor region were also
examined. As shown in Table C-l, total PCB tissue residues (as estimated by 106 or more
congeners) could be reliably estimated by doubling the subtotal of the 22 PCB congeners that are
routinely quantified in the dredging program. The ratio of the subtotal of the 22 congeners to the
total PCB concentration was consistent across four species of infaunal worms, five species of
molluscs, and five species of finfish. This ratio is shown by the slopes reported in the fifth
column of Table C-l.
Based on the consistency of this relationship across species, taxonomic groups and geographic
subregions in the NY/NJ Harbor area, quantification of an extended list of congeners appears
unnecessary to provide a reasonable estimate of total PCBs. Therefore, EPA Region 2 proposes
to continue to estimate total PCB residues in test organism tissues by doubling the sum of the
measured residues of the 22 PCB congeners specified in the Regional Testing Manual.
Table C-l. Relationship between subtotal of 22 PCB congeners (x) and total measured PCB concentrations
(y) in aquatic organisms collected from the NY/NJ Harbor region (y = mx + b, b = Oj
Study
No. obs.
Species
Area
slope (m)
R2
No. of Congeners
Measured
NOAA, 1987
21
M. edulis
NY/NJ
Harbor
2.0
>0.97
106
McFarland et
al. 1994
26
4 worms; 4
molluscs
NY Bight
Apex
1.89
0.96
106
EPA, 1998.
Hudson River
RI/FS
31
5 finfish
lower
Hudson
1.98
0.99
138
It is important to note that although EPA Region 2 proposes to require quantitation and reporting
of coplanar PCB residues if a risk-based approach for assessing ecological and human health
risks associated with dioxin-like activity in dredged sediments proposed for use as Remediation
81
-------
Appendix C
Material at the HARS, their residues do not need to be included in the sum of congeners used to
predict total PCB residue concentration.
EPA Region 2 believes this to be appropriate because the method for estimating total PCBs from
the subtotal of 22 measured congeners (i.e., doubling measured residues of the congeners listed
in the RTM) is only an estimate of total PCB mass in the sample. Furthermore, the coplanar
PCB congeners generally occur at much lower levels than other congeners (i.e., at pptr levels)
and therefore would not substantially alter the sum obtained using the 22 PCB congeners
currently measured. Therefore, EPA Region 2 does not believe that adding the residues of the
three coplanar PCBs to the residue of the 22 PCBs before doubling would not significantly
improve the estimate of total PCB mass. Therefore, Region 2 is proposing to continue to
estimate total PCBs using the current procedure of doubling the summed residue of the 22 PCB
congeners currently considered and not adding the coplanar PCBs to this sum.
References Cited:
Battelle - Duxbury. 1998. Aroclor Analysis and Extended Congener Analysis for Mamaroneck Harbor,
Raritan River, and Kill van Kull. Sediment and Water Data Organics. Data Report Prepared for U.S.
Army Corps of Engineers New York District. Battelle - Duxbury. Duxbury, MA. 35 p.
Durell, G.S. and R.D. Lizotte, Jr. 1998. PCB levels at 26 New York City and New Jersey WPCPs that
discharge to the New York/New Jersey Harbor Estuary. Environ. Sci. Techno!. 32(8): 1022-1031.
EPA (U.S. Environmental Protection Agency). 1998. Database for the Hudson River PCBs Reassessment
RI/FS. Release 4.1. [CD-ROM]. TAMS Consultants, Inc. Bloomfield, NJ.
McFarland, V.A., C.H. Lutz, and F.J. Reilly. 1994. Bioaccumulation data and analysis for selected
contaminants in sediments and biota of the New York Bight Apex Mud Dump Reference Site. U.S.
Army Corps of Engineers, Waterways Experiment Station. Final Report. February 1994. 65 pgs
(appendices).
NOAA. 1987. Status and Trends Mytilus edulis tissue data for the NY/NJ Harbor region. Compiled by
Battelle Ocean Sciences.
82
-------
APPENDIX D
Proposed Analytical Protocol for Alkylated PAHs
-------
DRAFT REPORT
PROCEDURES FOR
ANALYSIS OF PAH AND ALKYL PAH IN SEDIMENT AND TISSUE AT
RISK-BASED DETECTION LIMITS
submitted to
US EPA Region 2
290 Broadway
24th Floor
New York, NY 10007
November 24,1999
prepared by
Battelle
397 Washington Street
Duxbury, MA 02332
(781) 934-0571
83
-------
This page intentionally left blank.
84
-------
Appendix D
Measurement of Poly nuclear Aromatic Hydrocarbons by GC/MS
TABLE OF CONTENTS
1. Introduction 87
1.1 PAH Compounds of Potential Concern 87
1.2 Detection Limit Considerations 88
2. Analytical Methods 90
3. Quality Contol and Data Quality Objectives 92
Sediment/Soil
4. Initial and On-Going Demonstration of Method Proficiency 94
5. References 95
LIST OF FIGURES
Figure 1. Methods Flowchart 90
LIST OF TABLES
Table 1. PAH and alkyl PAH target compound list 89
Table 2. Considerations for adaptation of the methods for analysis of PAH in sediment/soil and
biological tissue 91
Table 3. Data Quality Objectives for the measurement of PAH and alkyl PAH in sediment/soil and
tissues 93
LIST OF ATTACHMENTS
Attachment 1. Procedures for Sediment Extraction for Trace-Level Semi-Volatile
Organic Contaminant Analysis
Attachment 2. Procedures for Tissue Extraction for Trace-Level Semi-Volatile
Organic Contaminant Analysis
Attachment 3. Procedures for Gel Permeation HPLC Cleanup of Sediment and Tissue Extract for
Semi-Volatile Organic Pollutants
Attachment 4. Procedures for Identification and Quantification of Polynuclear Aromatic Hydrocarbons
by Gas Chromatography/Mass Spectrometry
85
-------
Appendix D
Measurement of Polynuclear Aromatic Hydrocarbons by GC/MS
This page intentionally left blank.
86
-------
Appendix D
Measurement of Polynuclear Aromatic Hydrocarbons by GC/MS
1. INTRODUCTION
This document describes analytical methods suitable for the measurement of polycyclic aromatic
hydrocarbons (PAH) and many of their alkylated homologues in marine and estuarine sediment and
biological tissues. The analytical methods described in this document are adaptations of National
Oceanic and Atmospheric Administration (NOAA) National Status and Trends (NS&T) Program
methods. The NS&T Program has, since 1985, continually refined a series of analytical protocols for the
measurement of PAH (and other contaminants) in sediment and biological tissues at environmentally
relevant concentrations. These methods include robust sample cleanup procedures and analytical
instrument configurations optimized for the detection and quantitative analysis of PAH and other
compounds. These methods are described in NOAA (1993) and a later update (NOAA, 1998).
In addition to being the methods used for generating 15 years of coastal monitoring data for PAH in
sediment and biota in US coastal waters for the NOAA NS&T Program, these methods have been used to
support dredged material testing prescribed by the US EPA/Army Crops of Engineers Testing Manual,
Evaluation of Dredged Material Proposedfor Ocean Disposal (EPA, 1991:often referred to as the
"Green Book"). Importantly, the NOAA NS&T methods are those designated by the US Army Corps of
Engineers New York District/EPA Region 2 Guidance for Performing Tests on Dredged Material
Proposedfor Ocean Disposal [December 18, 1992 and subsequent revisions] for the measurement of
PAH in sediment and biological tissues as part of its testing regiment. This means that the adaptations of
the NOAA NS&T Program methods proposed in this document for measurement of PAH and alkylated
PAH will generate data consistent with historical monitoring of many sediments in the New York Harbor
area and in biological tissues exposed to these sediments as part of bioaccumulation testing of these
sediments.
1.1 PAH Compounds of Potential Concern
Polycyclic aromatic hydrocarbons (PAH) are ubiquitous contaminants in urban coastal marine
environments. PAH can be transported to aquatic environments from accidental spills, waste dumping,
industrial and domestic wastewater discharges, runoff from land, river outflows, and atmospheric fallout
(Neff, 1979). Because of their nature of formation and similar physical/chemical properties, groups of
petrogenic or pyrogenic PAH tend to co-occur in sediments. That is to say that genetically-related PAH
will occur together in sediments. PAH originate from a large number of sources which can be broadly
classified as either (1) diagenetic, (2) petrogenic, or (3) pyrogenic.
• Diagenetic sources are natural sources of PAH that are not ordinarily recognized as
significantly impacting sediment quality (Wakeham et al., 1980).
• Petrogenic sources are anthropogenic (def, derived from man's activities) sources of
PAH that are derived directly from crude oil or refined petroleum products.
• Pyrogenic sources are anthropogenic sources of PAH which include those derived from
combustion of petroleum products, or combustion and conversion of coal.
PAH, as the name implies, contain multiple 'ring' structures which are aromatic in nature. While PAH
are comprised principally of hydrogen and carbon, other non-carbon atoms, e.g., nitrogen, sulfur can be
present in the aromatic ring structures). The arrangement and number of rings is used to distinguish
different PAH. In addition to the ring structures, many PAH contain carbon side-chains of varying
numbers, lengths, and locations. Those PAH without any side-chains are considered as "parent" or C0-
87
-------
Appendix D
Measurement of Polynuclear Aromatic Hydrocarbons by GC/MS
PAH. The substituted PAH are collectively referred to as alkylated PAH. Within each series of alkylated
PAH are a number of theoretical structural isomers, e.g., compounds with the same basic structure but
differing in the arrangement of the alkyl side-chains. While it is possible to measure many of the
individual isomers, it is common to measure the collection of isomers within an alkyl substitution group
as the summation of those isomers, For example, the alkylated PAH with one, single carbon side chain
are said to be Ci-PAH, two additional carbons attached are C2-PAH, and so on.
Among the PAH, the most common that can be characterized by conventional means, e.g., gas
chromatography, contain between 2 and 6 aromatic rings. Sixteen 2 through 6 ring- unsubstituted PAH
have been designated by the US EPA as Priority Pollutant PAH (the bold compounds in Table 1).
Standard methods for the analysis of PAH in environmental media (e.g., EPA 8270, EPA 8310) focus
exclusively on this list of PAH compounds. However, PAH found in environmental media, as mentioned
above, are often complex mixtures whose total PAH concentrations are dominated by the alkyl substituted
PAH (e.g., Youngblood and Blumer, 1975; Laflamme and Hites, 1978)—it is not uncommon for 50 or
more percent of total PAH found in environmental samples to be comprised of the alkylated PAH. One
of the objectives of this document is to provide a convenient means to measure PAH and their major
alkylated homologues in sediment and tissue samples to assist risk assessors in developing a more
quantitative understanding of the occurrence and potential hazard of PAH in the marine environment.
The list of target PAH we propose to measure are listed in Table 1. This list includes the major
homoatomic 2- through 6-ring parent PAH, and the important sulfur-containing thiophene and
dibenzothiophene which are found in almost all PAH mixtures in environmental samples. The important
Ci-, C2-, C3-, and C4- alkylated homologues of these PAH that are known to be dominant constituents of
these such assemblages are included in the target compound list.
1.2 Detection Limit Considerations
The adaptations of the NOAA NS&T methods described in this document will provide method detection
limits (MDLs) for individual PAH compounds, on a dry weight basis, of approximately 10 |xg/Kg for
sediment and 5 p.g/Kg for biological tissues. The need for low detection limits are driven by several facts:
(1) it is not uncommon to encounter clean sediment with very low concentrations of PAH (<10-
20 (J.g/Kg)
(2) it is common to encounter biological tissues (especially 'uncontaminated' biota used in
bioaccumulation investigations) to contain <5-20 ng/kg of individual PAH
(3) biological tissues exposed to contaminated sediments generally have upper bioaccumulation
potentials of approximately 100-500 |a.g/Kg
Clearly, it is advantageous to have MDLs below or as close as practical to these thresholds in order to
measure PAH at environmentally relevant concentrations. Note that standard EPA Methods such as EPA
Method 8270 have detection limits of approximately 330 Hg/Kg for the Priority Pollutant PAH in
sediment. Detection limits for PAH in tissues are likely higher that 330 |ag/Kg because the standard do
not provide for adequate cleanup of lipid from tissue matrices to achieve such detection limits. Hence,
application of the NOAA NS&T Program methods provide the best opportunity for accurately measuring
PAH in sediment and biological tissue at concentrations likely to be encountered in many estuarine and
marine sediments, or in the tissues of organisms exposed to such sediment.
88
-------
Appendix D
Measurement of Polynuclear Aromatic Hydrocarbons by GC/MS
Table 1. PAH and Alkyl PAH target compound list.
Decalin
Anthracene
Ci-Decalins
C i -Phenanthrenes/ anthracenes
C2-Decalins
1 -Methylphenanthrene
C3_Decalins
C2-Phenanthrenes/anthracenes
C4-Decalins
C3-Phenanthrenes/anthracenes
Benzothiophene
C4-Phenanthrenes/anthracenes
C i -Bnzoth iophenes
Dibenzothiophene
C2-Benzothiophenes
Ci-Dibenzothiophenes
Cj-Benzothiophenes
C2-Dibenzothiophenes
C4-Benzothiophenes
C3-Dibenzothiophenes
dg-Naphthalene3
Fluoranthene
Naphthalene
Pyrene
Ci-Naphthalenes
C i -Fluoranthene/pyrenes
2-Methylnaphthalene
C2-Fluoranthene/pyrenes
1 -Methylnaphthalene
C3-Fluoranthene/pyrenes
C2-Naphthalenes
di2-Chrysenea
2,6-Dimethylnaphthalene
Benz[a] anthracene
C3-Naphthalenes
Chrysene
2,3,5 -T rimethy Inaphthalene
C i -benz[a]anthracenes/chrysenes
C4-Naphthalenes
C2-benz[a]anthracenes/chrysenes
dio-Acenaphtheneb
C3-benz[a]anthracenes/chrysenes
Acenaphthylene
C4-benz[a]anthracenes/chrysenes
Acenaphthene
d 12-benzo[a]pyreneb
Biphenyl
Benzo [b] fluoranthene
dio-Fluoreneb
Benzo[k]fluoranthene
Dibenzofuran
Benzo[e]pyrene
Fluorene
Benzo[a] pyrene
Cpfluorenes
Perylene
C2-fluorenes
Indeno[1^2^-c,
-------
Appendix D
Measurement of Polynuclear Aromatic Hydrocarbons by GC/MS
2. ANALYTICAL METHODS
The methods for the analysis of sediment/soil and biological tissues for PAH and alkylated PAH are
presented as four separate attachments to this document. The methods described in Attachments one
through four are adaptations of the NOAA National Status and Trends Program (NOAA, 1993,1998),
The method documents in the four
attachments are:
• Procedures For Sediment
Extraction for Trace-Level Semi-
Volatile
Organic Contaminant Analysis
(Including PAH)
• Procedures For Tissue Extraction
for Trace-Level Semi-Volatile
Organic Contaminant Analysis
(Including PAH)
• Procedures For Gel Permeation
HPLC Cleanup Of Sediment
And Tissue Extracts For Semi-
Volatile Organic Pollutants
• Procedures For Identification And
Quantification Of Polynuclear
Aromatic Hydrocarbons By Gas
Chromatography/Mass
Spectrometry
Figure 1 captures the logical application of
these methods. The appropriate sample
extraction procedure is selected first,
depending on the matrix under
investigation. Crude extracts, prepared
either from sediment or tissue samples, are
further purified by Gel Permeation HPLC,
then analyzed for PAH compounds by gas
chromatography with mass spectrometry.
The method descriptions in Attachments
one through four contain specific information regarding their application. Quality Control and Data
Quality Objectives for the application of these methods are presented below in Section 3. Overlying these
methods is the assumption that the laboratory using these methods can apply them within the boundaries
of expected method detection limits (MDLs) and adherence to appropriate quality control criteria.
Table 2 presents a synopsis of information germane to laboratories interested in adapting these methods.
While not data quality objectives (which are described below in Section 3), this information is presented
to allow a potential laboratory a perspective on certain technical requirements that will assist them in the
practical expectations of applying the methods.
Matrix?
Sediment
Tissue
I
Extraction
"Sediment Extraction for
Trace-Level Semi-Volatile
Organic Contaminant
Analysis"
L
Extraction
Tissue Extraction for
Trace-Level Semi-Volatile
Organic Contaminant
Analysis"
Cleanup
" Procedures For Gel Permeation HPLC
Cleanup Of Sediment
And Tissue Extracts For Semi-Volatile
Organic Pollutants"
Instrumental Analysis
"Procedures For Identification And
Quantification Of Polynuclear Aromatic
Hydrocarbons By Gas
Chromatography/Mass Spectrometry"
Figure 1. Method Flowchart
90
-------
Appendix D
Measurement of Polynuclear Aromatic Hydrocarbons by GC/MS
Table 2. Considerations for adaptation of the methods
for analysis of PAH in sediment/soil and biological tissues.
Parameter
Tissue,
Sediment/Soil
Target MDL
5 ng/Kg
Dry weight
basis
10 ng/Kg
Dry weight
basis
Compounds used to
demonstrate MDLs
16 Priority Pollutant
PAH
Clean tissue
(clam, oyster)
spiked 3-5
times expected
instrument
detection limit
16 Priority
Pollutant PAH
Clean
sediment
spiked 3-5
times expected
instrument
detection limit
Sample mass extracted
(wet weight)
20-25 g
—
20-30 g
—
Surrogate Internal
Standards (SIS) added to
each sample
Naphthalene-dg
Phenanthrene-d | o
Chrysene-dn
1 jxg each
Naphthalene-dg
Phenanthrene-dio
Chrysene-dn
1 pgeach
Recovery Internal
Standards (RIS) added to
each sample
Fluorene-dio
Acenaphthene-dio
Benzo[a]pyrene-d]2
1 |ag each
Fluorene-dio
Acenaphthene-dio
Benzofalpyrene-dn
1 |ig each
Matrix Spike &
Matrix Spike Duplicate
16 Priority Pollutant
PAH
0.5 |ig each
compound
16 Priority
Pollutant PAH
100-500 ng
each
compound
Laboratory Control
Sample (LCS)
16 Priority Pollutant
PAH
0.5 ng each
compound
16 Priority
Pollutant PAH
50-100 ng
each
compound
Standard Reference
Material (SRM)
NIST1974A
Use 5 grams
freeze-dried
material
NIST 1941A
Use 5 grams
freeze dried
material
GC/MS calibration
curve range
Low Standard
High Standard
0.05 ng/fiL
10 ng/nL
Low Standard
High Standard
0.05 ng/(iL
10 ng/nL
Sample Pre-Injection
Volume (PIV)
0.5 mL
—
0.5 mL
—
91
-------
Appendix D
Measurement of Polynuclear Aromatic Hydrocarbons by GC/MS
3. QUALITY CONTROL AND DATA QUALITY OBJECTIVES
A thorough quality control (QC) regiment must be followed in the application of the analytical methods
described in this document. Every batch of sediment/soil or tissue samples (defined as 20 or fewer field
samples) must be accompanied by appropriate quality control samples, each with specific data quality
objectives. Specifically, the following QC samples must accompany each analytical batch:
Sediment/Soil
• Procedural Blank: 30 grams of sodium sulfate, carried through the entire analytical process.
Used to assess impact of any laboratoiy-based contamination, if any, on method performance.
• Matrix Spike/Spike Duplicate: 20-30 grams of clean sediment, fortified with 100 |iL of Matrix
Spike Solution (0.5 |ag of each Priority Pollutant PAH), and carried through the entire analytical
process. Used to assess method accuracy and precision with matrices similar to the field samples.
• Laboratory Control Sample: 20 grams of sodium sulfate, fortified with 100 p.L of Matrix Spike
Solution (0.5 fig of each Priority Pollutant PAH), and carried through the entire analytical
process. Used to assess method accuracy in the absence of matrix.
• Standard Reference Material: 5 grams of NIST 1941 A, carried through the entire analytical
process, used to assess method accuracy.
• Field Duplicate: A second aliquot of a sediment sample, prepared and analyzed to assess sample
homogeneity and method precision.
Tissue
• Procedural Blank: 30 grams of sodium sulfate, carried through the entire analytical process.
Used to assess impact of any laboratory-based contamination, if any, on method performance.
• Matrix Spike/Spike Duplicate: 20-25 grams of clean tissue (clam, oyster, or project-specific
matrix), fortified with 100 |a.L of Matrix Spike Solution (0.5 fig of each Priority Pollutant PAH),
and carried through the entire analytical process. Used to assess method accuracy and precision
with matrices similar to the field samples.
• Laboratory Control Sample: 20 grams of sodium sulfate, fortified with 100 p.L of Matrix Spike
Solution (0.5 |ig of each Priority Pollutant PAH), and carried through the entire analytical
process. Used to assess method accuracy in the absence of matrix.
• Standard Reference Material: 5 grams of NIST 1974A, carried through the entire analytical
process, used to assess method accuracy.
• Field Duplicate: A second aliquot of a tissue sample, prepared and analyzed to assess sample
homogeneity and method precision.
In addition to QC samples, requirements for the initial calibration and on-going calibration of the GC/MS
system must be met. Note that the specific concentration range for instrument calibration standards are
described in Attachment 4. The low level calibration standard for the GC/MS system must be used in
order to properly bracket environmentally relevant concentrations of PAH in sediment and biological
tissues.
The data quality objectives for the methods and QC samples described in this Section are summarized in
Table 3.
92
-------
Appendix D
Measurement of Polynuclear Aromatic Hydrocarbons by GC/MS
Table 3. Data Quality Objectives for measurement of PAH and Alkyl PAH
in sediment/soil and tissues.
QC Type/Parameter
DQO Targets
Comments
Procedural Blank
5 x
background concentration to be used for
data quality assessment.
Matrix Spike/Spike Duplicate
Precision
<30% Relative Percent Difference
Analyte concentration in MS must be >5 x
background concentration to be used for
data quality assessment.
Field Duplicate
<30% Relative Percent Difference
for target analytes detected in sample
Analyte concentration in native sample
must be >5 x MDL to be used for data
quality assessment.
Standard Reference Material
Accuracy
Within 30% of certified concentration range
Analyte concentration must be >5 x MDL
to be used for data quality assessment.
Surrogate Compound Recovery
40- 120% recovery
Instrument Control Check
% Difference <15%
Independent check standard to confirm
instrument accuracy.
Instrument Calibrations
Initial (minimum of 5-point)
Continuing Calibration Check
<25% Relative Standard Deviation
in RRFs (average <15%)
<25% Percent Difference
vs reference concentration for all analytes.
Initial calibration required after all
instrument maintenance and alter a failing
continuing calibration.
Continuing calibrations must run every 10
samples and at the end of each sample
batch.
93
-------
Appendix D
Measurement of Polyttuclear Aromatic Hydrocarbons by GC/MS
4. INITIAL AND ON-GOING DEMONSTRATION
OF METHOD PROFICIENCY
A laboratory proposing to implement the methods described in this document should demonstrate initial
proficiency with both the sample preparation and determinative methods by generating data of acceptable
accuracy and precision for target analytes spiked into a clean matrix. The laboratory should also repeat
the following proficiency demonstration whenever new staff are trained or significant changes in the
instrumentation are made.
Gas Chromatography/Mass Spectrometry. The laboratory must demonstrate the ability to analyze and
resolve the 16 Priority Pollutant PAH target compounds listed in Table 1 at the lowest concentration in
the instrument calibration curve: 0.05 ng/^L. Furthermore, the laboratory must be able to demonstrate the
ability to extract the ion current profiles for the alkylated PAH of concern in Table 1, and favorably
compare them to the EIPs in Attachment 4 of this document. This latter demonstration should be
accomplished through the analysis of a crude oil standard (e.g., North Slope Crude Oil) dissolved in
dichloromethane at a concentration of approximately 5 mg/mL.
Initial Precision and Accuracy should be demonstrated through the analysis of replicate (3 to 5) matrix
spike samples of both sediment/soil and biological tissues, fortified with the 16 priority pollutant PAH at
a concentration of approximately 100 (J.g/kg (sediment/soil) and 5 to 10 p.g/Kg (tissue) for each of the
individual target compounds. This concentration corresponds to 10 times the MDL for each matrix. The
samples should be processed according to the methods described within this document. The measured
concentrations of the individual spiked compounds should be 40 to 120% (accuracy) of the true value,
and the relative standard deviation (RSD - precision) in the replicate measurements should be less than
25%.
Furthermore, the laboratory should analyze National Institute of Standards and Technology (NIST)
Standard Reference Material (SRM) sediment and tissue, respectively, and meet the certification limits
for these materials as documented by NIST. Appropriate SRMs include NIST SRM 1974A (tissue) and
NIST SRM 1941A (sediment).
Method Detection Limits (MDL) for each of the 16 Priority Pollutant PAH should be determined in
both sediment/soil and tissue. MDLs should be determined according to procedures outlined in Test
Methods for Evaluating Solid Waste, Physical/Chemical Methods (SW-846). Final Update III, 1996.
Eight replicate samples should be fortified at 3 to5 times the expected MDL (r.e., fortified at ca. 50 (o.g/kg
for sediment/soil, ca. 25 jxg/Kg for tissue) using the 16 Priority Pollutant PAH compounds listed in Table
1. The samples must be extracted and analyzed according to the methods described in this document, and
the MDL calculated from the equation
MDL = tx a
Where
t is the one-sided t statistic at the 99% confidence level for
n-1 degrees of freedom
n is the number of replicate analyses
a is the standard deviation in the concentration of the replicate measurements
94
-------
Appendix D
Measurement of Polynuclear Aromatic Hydrocarbons by GC/MS
MDLs of 5 |ag/Kg (tissues) and 10 |J.g/Kg (sediment/soil), on a diy weight basis, should be achieved to
demonstrate acceptable method performance.
Procedural Blanks processed with this initial demonstration of method performance must be free from
any target analytes of interest at concentrations less than the method detection limit.
5. REFERENCES
EPA and US Army Corps of Engineers. 1991. Evaluation of dredged material for ocean disposal—
testing manual. Environmental Protection Agency/U.S. Army Corps of Engineers. U.S. Army Engineer
Waterways Experiment Station, Vicksburg, MS.
Laflemme, R.E. and R.A. Hites. 1978. The global distribution of polycyclic aromatic hydrocarbons in
recent sedimets. Geochim. Cosmochim. Acta 42:289-303.
Neff, J.M. 1979. Polycyclic Aromatic Hydrocarbons in the Aquatic Environment. Sources, Fates, and
Biological Effects. Appl. Sci. Publ., Ltd., London.
NOAA, 1993. Sampling and Analytical Methods of the National Status and Trends Program National
Benthic Surveillance and Mussel Watch Project. NOAA Technical Memorandum NOS ORCA 71.
National Oceanic and Atmospheric Administration, Silver Spring, MD.
NOAA, 1998. Sampling and Analytical Methods of the National Status and Trends Program Mussel
Watch Project: 1993-1996 Update. NOAA Technical Memorandum NOS/ORCA/CMBAD 130.
National Oceanic and Atmospheric Administration, Silver Spring, MD.
Wakeham, S.G., Schaffner, C., and Giger, W. 1980. Polycyclic aromatic hydrocarbons in Recent lake
sediments - II. Compounds derived from biogenic precursors during early diagenesis. Geochim.
Cosmochim. Acta 44:415-429.
Youngblood, W.W. and M. Blumer. 1975. Polycyclic aromatic hydrocarbons in the environment.
Homologous series in soils and recent marine sediments. Geochim. Cosmochim. Acta 39: 1303-1314.
95
-------
Appendix D
Measurement of Polynuclear Aromatic Hydrocarbons by GC/MS
This page intentionally left blank.
96
-------
ATTACHMENT 1
Procedures for
Sediment Extraction for Trace-Level Semi-Volatile
Organic Contaminant Analysis
97
-------
This page intentionally left blank.
98
-------
Appendix D
Measurement of Polynuclear Aromatic Hydrocarbons by GC/MS
1.0 OBJECTIVE
This procedure documents standardardized methods for extracting trace levels of semi-volatile organic
analytes from sediment/soilmatrices for analysis by gas chromatography. The extraction procedures are
suitable for analysis of low levels of semi-volatile organic pollutants including polycyclic aromatic
hydrocarbons (PAH), chlorinated pesticides, polychlorinated biphenyls (PCB) and others such as those
listed those listed in EPA Methods 608, 610, 625, 8081, 8082, and 8270. The method may be suitable
for other analytes once acceptable extraction efficiency has been demonstrated. This procedure is used to
prepare extracts that are further cleaned up by liquid chromatography procedures prior to their
instrumental analysis.
Note: Soils originate on land (terrestrial environments) and are typically dry mixtures of geological and
organic materials, while sediments originate from the bottom of open aqueous environments
(e.g., coast, rivers, lakes) and are typically wet mixtures; both will be treated the same in this
SOP.
2.0 PREPARATION
2.1 APPARATUS AND MATERIALS
Apparatus for soil/sediment sample extraction
• Orbital shaker table (Lab-Line Instruments Inc. Model 3520 or equivalent) or roller extractor
• Teflon jar or centrifuge bottle, 250-mL capacity
• 500-mL Erlenmeyer flask
Apparatus for determining wet weight and dry weight
• Top-loading balance capable of weighing to 0.01 g (SOP No. 3-160)
• Aluminum weighing pans, stored in an aluminum foil "package" at 105°C
• Stainless steel spatula
• Drying oven maintained at 105-120 °C, Blue M Model SW-17TA or equivalent
Centrifuge
Balance accurate to 0.01 g
Glass wool heated to 400°C for at least 4 h, then stored in a covered glass container at 105°C.
Glass fiber filters heated to 400°C for at least 4 h, then stored in aluminum foil "package" at 105°C.
19/21-mm chromatography column with -200 mL reservoir and Teflon stopcock
Kuderna-Danish (K-D) apparatus (listed below) or Turbovap concentration units and tubes
• Reservoir, 500 mL
• Snyder column, three-ball macro
• Concentrator tube, 10- or 20-mL
99
-------
Appendix D
Measurement of Polynuclear Aromatic Hydrocarbons by GC/MS
Hot water bath capable of reaching 100 °C, located in fume hood.
Boiling chips, solvent rinsed
Nitrogen evaporation apparatus, N-Evap or equivalent, with water bath maintained at about 25 °C
Glass graduated cylinders, 100- and 500-mL
Erlenmeyer flasks, 250- and 500-mL
Microliter syringes
2.2 REAGENTS
Dichloromethane (DCM), pesticide grade or equivalent
Acetone, pesticide grade or equivalent
Milli-Q and de-ionized (DI) water
HC1, 12N
Sodium sulfate—anhydrous, reagent grade, heated to 400°C for at least 4 h, then cooled and stored in a
tightly-sealed glass container at room temperature.
Alumina, F-20 (Aldrich Chemical. CAS #B44-28-l. 80-200 mesh).
Activate the alumina by heating in a shallow dish to 400 °C for at least 4 hr. Allow the alumina to cool in
an oven (do not keep in open lab atmosphere for extended periods). Store the alumina in a covered glass
container (with an activation date label) at 105 °C, and use within 1 week of activation.
Deactivate the alumina prior to use with 2% water. Prepare the 2% deactivated alumina in batches by
adding 5 mL of Milli-Q water (accurately measured with a syringe or volumetric pipette) to 250 g of
alumina in a 1,000 mL round bottom glass flask. Seal with a glass or Teflon stopper, shake vigorously by
hand for 1 min to begin the process. Vent the flask, put the stopper back on, seal with Teflon tape, and
allow to vigorously mix on a shaker table for 2 hr (± 15 min). Pack columns within 2 hr of completing
the deactivation and use the same day. Unused deactivated alumina should be discarded.
Granular Copper, 99% pure (copper J.T. Baker CAS #7440-50-8 or equivalent), activated.
To activate copper: add 20-25 g copper to a beaker (should be no more than 1 inch "height" of copper).
Add Milli-Q or DI water to beaker, then add an equal volume of 12N HC1. Stir the mixture for 2-3 min
(until uniformly bright copper colored) with a spatula, and decant to acid waste. Rinse with Milli-Q or DI
water until there is no evidence of acid (typically at least 8-10 times), rinse with acetone until there is no
evidence of water (typically at least 3-5 times), and rinse with DCM until there is no evidence of acetone
(typically at least 3-5 times). If the DCM appears cloudy after the first addition, there is still water
present in the copper; repeat process from the addition of acetone. Activate copper no more than one
hour before use. Unused activated copper should be discarded.
Surrogate Internal Standards (SIS) spiking solution. Prepare a Working Standard solution containing 10
Hg/mL each of Naphthalene-dg, Phenanthrene-dio, and Chrysene-d]2 in dichloromethane.
100
-------
Appendix D
Measurement of Polynuclear Aromatic Hydrocarbons by GC/MS
Recovery Internal Standards (RIS) spiking solution. Prepare a Working Standard solution containing
5 ng/mL each of Fluorene-dio, Acenaphthene-di0, and Benzo[a]pyrene-d]2 in dichloromethane.
2.3 LABWARE PREPARATION
All glassware must be thoroughly washed with detergent and water, rinsed with deionized water, and
baked in a 400°C muffel oven for a minimum of 4 hr. prior to use. All other labware should be cleaned
with detergent and water and thoroughly rinsed with deionized water.
3.0 PROCEDURES
3.1 GENERAL
Samples must be extracted in batches of 20 or fewer field samples. Quality control samples
accompanying each batch include a procedural blank, laboratory control sample (LCS), matrix spike,
matrix spike duplicate, standard reference material (NIST 1941 A), and a field sample duplicate. (See
Section 5.0 for more detail).
Each sample, including QC samples, should be spiked with SIS. Add 100 p.L of the SIS Working
Standard so that approximately 1 jxg of each SIS compound is added to each sample.
The soil/sediment samples should be thoroughly homogenized prior to any aliquotting for chemical or
physical characterization. Any overlying water on samples should be decanted prior to stirring and
subsampling unless otherwise specified in the project work plan.
3.2 PERCENT MOISTURE DETERMINATION
Weigh approximately 10 g of well mixed soil/sediment, into a pre-weighed, pre-baked, aluminum
weighing pan and record to the nearest 0.01 g. Place the sample in a dicing oven and dry overnight at ca.
105 °C. After approximately 24 h, allow the sample to cool at room temperature for at least 30 min.
Record dry weight to the nearest 0.01 g. Calculate the percent moisture as described in Section 4.
3.3 EXTRACTION, CONCENTRATION, AND CLEANUP
Soil/Sediment Sample Extraction
• Weigh 30 to 40 g of wet, well mixed, soil/sediment into a 250 mL Teflon jar to the nearest 0.01 g.
Record sample weight.
• Add approximately 60 g sodium sulfate, mix well with spatula. Add 100 |j.L of the SIS Working
Standard (ca. 1 fig each SIS compound into the sample) and 100 mL DCM. Add more sodium
sulfate if sample is lumpy — some sodium sulfate should move freely in the solvent indicating
that no additional water is available. Hardened sodium sulfate should be broken up periodically
to ensure optimum extraction and efficient solvent rinsing of the sample.
101
-------
Appendix D
Measurement of Polynuclear Aromatic Hydrocarbons by GC/MS
• Cap bottle and shake by hand until contents are loose. Shake sample on a shaker table for at least
12 hr (overnight), checking the bottles after 30-60 min and "breaking up" the sample, if needed.
• Centrifuge the sample for 5 min. at 1500-2000 rpm. Longer centrifugation may be needed for
samples with fine particulates. Carefully decant the extract into an Erlenmeyer flask and cap with
aluminum foil.
• Repeat the extraction with an additional 100 mL DCM for approximately 4 hr, after first "breaking
up" the sample, if needed (i.e., ensure that sample is free-flowing when swirled). Centrifuge and
combine the extracts in the Erlenmeyer flask.
• Repeat the extraction a third time with 100 mL DCM for ca. 0.5 hr, after first "breaking up" the
sample, if needed. Centrifuge sample and decant and combine the extracts in the Erlenmeyer
flask. Break up, if needed, and rinse the extracted sample mixture with 25 mL of DCM,
combining it with the rest of the extract in the Erlenmeyer flask.
• Add approximately 50 g of sodium sulfate to the combined extract, swirl, and allow to sit for at
least 15 min. Add more sodium sulfate if it all becomes lumpy — some sodium sulfate should
move freely in the solvent indicating that no additional water is available.
Note 1: One of the first two extractions must last for approximately 12 hr, or more. The other extraction
must last 4 hr, or longer. The third extraction is essentially a rinse; a shaking period of approximately
0.5 hr should be performed.
Note 2: The final combined extract should not be stored in the Erlenmeyer flask for more than 1 day. If
longer storage is needed before completing the sample preparation, the extract should be concentrated to
several milliliters (see below), transferred to a glass vial, securely capped, and stored in darkness in a
refrigerator or freezer.
Extract Concentration — Kuderna-Danish (K-D) Technique
• Transfer sample extract to a K-D flask and receiver (filter through glass fiber filter or glass wool if
there is evidence of suspended particulates). Rinse the Erlenmeyer twice with about 10 mL of
DCM and add to K-D.
• Add 3-5 boiling chips to the K-D receiver and insert a Snyder column. Pre-wet the condenser
column with approximately 5-mL DCM. Place the K-D apparatus in a hot water bath maintained
at 60-65 ° C (monitored by a mercury thermometer), such that the concentrator tube is partially
immersed in hot water and the entire lower rounded surface of the flask is bathed in hot water
vapor. At the proper rate of evaporation, the balls of the Snyder column will actively chatter, but
will not flood with condensed solvent. Continue concentration until the sample volume is
reduced to approximately 10 mL.
• Remove the K-D apparatus and allow it to drain and cool for at least 10 min. Transfer the
concentrator tube to the N-Evap unit and concentrate the sample until the volume is 1 -2 mL,
maintaining a water bath temperature of approximately 25 °C.
Note 1: The water bath for the K-D is maintained at 60-65 ° C for concentrating DCM, 70-75 ° C for
concentrating acetone, and a boiling bath is used for concentrating hexane or toluene.
Note 2: Adjust the flow of nitrogen on the N-Evap to a gentle stream — do not allow the sample to
bubble or splatter, or have a large "dimple " on the surface, as this will result in the loss of
sample. A small dimple is expected on the surface of the solvent.
102
-------
Appendix D
Measurement of Polynuclear Aromatic Hydrocarbons by GC/MS
Extract Concentration — TurboVap® Technique (optional)
A TurboVap® may be used for concentrating the extracts rather than using K-D.
• Transfer sample extract to a 200 mL TurboVap® tube with a 1 mL collection reservoir.
TurboVap® waterbath temperature should be set at approximately 25 °C; nitrogen pressure should
be approximately 4 to 6 psi.
• Follow manufacturers instructions for operation and maintenance. After initial TurboVap®
concentration the extract should be quantitatively transferred to a 4 mL vial (with DCM rinses),
and concentrated until the volume is 1-2 mL.
Extract Cleanup — Alumina
Alumina cleanup is performed on soil/sediment extracts to remove gross biogenic material that is co-
extracted with target compounds of interest. This step is performed prior to HPLC/GPC cleanup to
prevent overloading of the size exclusion column.
Column Preparation
• Prepare the 2% deactivated F-20 alumina as described in Section 2.2.
• Pack the tip of the 19/21-mm ("fat") chromatography column with a small piece of glass wool.
Add approximately 10 mL DCM and tap glasswool with clean glass rod to remove bubbles.
Drain solvent into waste cup.
• Add approximately 25 mL DCM to the column. Weigh 20-g, 2% deactivated alumina into a
beaker. Add -30 mL DCM and swirl to remove bubbles. Slowly pour alumina slurry into the
column while rinsing beaker with DCM
• Place approximately 1 g of sodium sulfate on top of the alumina. Tap the column to remove
bubbles. Drain column to top of packing and discard solvent. Column is now ready for use;
columns must be used the day they are prepared.
Column Elution
• Put a clean glass collection flask under the column {e.g., Erlenmeyer flask, K-D apparatus, or
TurboVap tube).
• Load the 1 to 2 mL sample extract (must be in DCM) onto the column.
• Slowly drain and stop at the top of the alumina packing. Rinse the sample vial with about 1 mL
DCM and load onto column. Slowly drain and stop at the top of the packing. Repeat vial rinse
one time.
• Load with 100 mL of DCM.
• Drain the column slowly (about 2 mL/min), collecting the column eluent containing the target
analytes, and stop at the top of the column — do not drain the column dry into the collection vial.
• Open up and drain the column dry into a solvent waste jar.
103
-------
Appendix D
Measurement of Pofynuclear Aromatic Hydrocarbons by GC/MS
Extract Cleanup — Copper Treatment
Activated granular copper (prepared as described in Section 2.2) is added to the extract to remove sulfur,
which interferes with GC analyses.
• Concentrate the alumina purified extract by K-D or TuboVap (described earlier) to a volume of
approximately 10 mL.
• Add a "scoop" (about 1-2 g) of activated granular copper to the extract using a clean stainless steel
spatula. Allow to react for at least 10 min and observe. Add additional activated copper if the
copper in the extract has turned black color (sulfur causes the copper to become black), and
observe.
• Repeat the activated copper addition step until the lastly added copper remains unchanged in color.
The extracts should remain in the vial with copper for at least 1 hour, but not over night (extended
contact with copper is not good for some less stable compounds).
Extract Cleanup — GPC
The sample is now ready for further cleanup and processing by Size exclusion HPLC/GPC (described in
Attachment 3 of this document). GPC must be carried out for effective low-level measurement of
contaminants.
The DCM extract volume should be adjusted to approximately 900 nL and centrifuged, if necessary, to
settle any particulates. The volume is measured with a gas tight syringe (to within 5 |iL), the volume
recorded, and the sample transferred to a 2-mL GC vial.
Process the samples by GPC as described in Attachment 3 of this document.
Following GPC cleanup, the samples are concentrated by K-D or Turbovap® procedures, spiked with
appropriate recovery internal standards (RIS), and submitted for GC/MS analysis of PAH. Each purified
sample extract is concentrated to about 400 fiL, then fortified with 100 (iL of the RIS Working Standard
(1 |j.g each RIS compound).
4.0 CALCULATIONS
Calculate percent dry weight with the following equation:
% dry weight -
aliquot wet wt
aliquot dry wt?
* 100
Calculate percent moisture with the following equation:
% moisture =
aliquot wet wt f - aliquot dry wt
* 100
aliquot wet wt
104
-------
Appendix D
Measurement of Polynuclear Aromatic Hydrocarbons by GC/MS
a corrected for pan weight
Calculate sample dry weight with the following equation:
Sample dry weight (g) = % dry wt. * sample wet wt. (g)
5.0 QUALITY CONTROL
Quality control methods for this procedure consists of preparation and analysis of a select QC samples,
prepared as described below. These QC samples are prepared and analyzed with every set of 20 or fewer
field samples of similar matrix:
• Procedural Blank: 30 grams of sodium sulfate, carried through the entire analytical process.
Used to assess impact of any laboratoiy-based contamination, if any, on method performance.
• Matrix Spike/Spike Duplicate: 20-30 grams of clean sediment, fortified with 100 (iL of Matrix
Spike Solution (0.5 jag of each Priority Pollutant PAH), and carried through the entire analytical
process. Used to assess method accuracy and precision with matrices similar to the field samples.
• Laboratory Control Sample: 20 grams of sodium sulfate, fortified with 100 p.L of Matrix Spike
Solution (0.5 (ig of each Priority Pollutant PAH), and carried through the entire analytical
process. Used to assess method accuracy in the absence of matrix.
• Standard Reference Material: 5 grams of NIST 1941 A, carried through the entire analytical
process, used to assess method accuracy.
• Field Duplicate: A second aliquot of a sediment sample, prepared and analyzed to assess sample
homogeneity and method precision.
The Data Quality Objectives for these QC procedures are described in the body of the document that
accompanies these method descriptions.
105
-------
Appendix D
Measurement of Polynuclear Aromatic Hydrocarbons by GC/MS
This page intentionally left blank.
106
-------
ATTACHMENT 2
Procedures for
Tissue Extraction for Trace-Level Semi-Volatile
Organic Contaminant Analysis
107
-------
This page intentionally left blank.
108
-------
Appendix D
Measurement of Polynuclear Aromatic Hydrocarbons by GC/MS
1.0 OBJECTIVE
This procedure documents standardized methods for extracting trace levels of semi-volatile organic
analytes from biological tissue. The extraction procedures are suitable for analysis of low levels of semi-
volatile organic pollutants including polycyclic aromatic hydrocarbons (PAH), chlorinated pesticides,
polychlorinated biphenyls (PCB) and others such as those listed those listed in EPA Methods 608, 610,
625, 8081, 8082, and 8270. The method may be suitable for other analytes once acceptable extraction
efficiency has been demonstrated. This procedure is used to prepare extracts that are further cleaned up
by gel permeation liquid chromatography (GPC) procedures prior to their instrumental analysis.
2.0 PREPARATION
2.1 APPARATUS AND MATERIALS
Apparatus for homogenizing tissue
• Tekmar Tissuemizer with probes, or equivalent
• Teflon jar or centrifuge bottle, 250-mL capacity
• 500-mL Erlenmeyer flask
Apparatus for determining wet weight and dry weight (tissue and sediment)
• Top-loading balance capable of weighing to 0.01 g (SOP No. 3-160)
• Aluminum weighing pan, stored in an aluminum foil "package" at 105°C
• Stainless steel spatula
• Drying oven maintained at 105-120 °C, Blue M Model SW-17TA or equivalent
Apparatus for determining lipid weight:
• Class A volumetric pipette or 10 mL syringe
• Aluminum weighing pan
Centrifuge
Balance accurate to 0.01 g
Glass wool heated to 400°C for at least 4 h, then stored in a covered glass container at 105°C.
Glass fiber filters heated to 400°C for at least 4 h, then stored in aluminum foil "package" at 105°C.
19/21-mm chromatography column with -200 mL reservoir and Teflon stopcock
Kuderna-Danish (K-D) apparatus (listed below) or Turbovap concentration units and tubes
• Reservoir, 500 mL
• Snyder column, three-ball macro
• Concentrator tube, 10- or 20-mL
109
-------
Appendix D
Measurement of Polynuclear Aromatic Hydrocarbons by GC/MS
Hot water bath capable of reaching 100 °C, located in fume hood.
Boiling chips, solvent rinsed
Nitrogen evaporation apparatus, N-Evap or equivalent, with heated water bath maintained at about 25 °C
Glass graduated cylinders, 100- and 500-mL
Erlenmeyer flasks, 250- and 500-mL
Microliter syringes
2.2 REAGENTS
Dichloromethane (DCM), pesticide grade or equivalent
Milli-Q water
Sodium sulfate—anhydrous, reagent grade, heated to 400°C for at least 4 h, then cooled and stored in a
tightly-sealed glass container at room temperature.
Alumina, F-20 (Aldrich Chemical. CAS #B44-28-l. 80-200 mesh).
Activate the alumina by heating in a shallow dish to 400 °C for at least 4 hr. Allow the alumina to cool in
an oven (do not keep in open lab atmosphere for extended periods). Store the alumina in a covered glass
container (with an activation date label) at 105 °C, and use within 1 week of activation.
Deactivate the alumina prior to use with 2% water. Prepare the 2% deactivated alumina in batches by
adding 5 mL of Milli-Q water (accurately measured with a syringe or volumetric pipette) to 250 g of
alumina in a 1,000 mL round bottom glass flask. Seal with a glass or Teflon stopper, shake vigorously by
hand for 1 min to begin the process. Vent the flask, put the stopper back on, seal with Teflon tape, and
allow to vigorously mix on a shaker table for 2 hr(+ 15 min). Pack columns within 2 hr of completing
the deactivation and use the same day. Unused deactivated alumina should be discarded.
Surrogate Internal Standards (SIS) spiking solution. Prepare a Working Standard solution containing 10
(ig/mL each of Naphthalene-dg, Phenanthrene-dio, and Chrysene-di2 in dichloromethane.
Recovery Internal Standards (RIS) spiking solution. Prepare a Working Standard solution containing
5 (J.g/mL each of Fluorene-dio, Acenaphthene-dio, and Benzo[a]pyrene-d12 in dichloromethane.
2.3 LABWARE PREPARATION
All glassware must be thoroughly washed with detergent and water, rinsed with deionized water, and
baked in a 400°C muffel oven for a minimum of 4 hr. prior to use. All other labware should be cleaned
with detergent and water and thoroughly rinsed with deionized water.
110
-------
Appendix D
Measurement of Poly nuclear Aromatic Hydrocarbons by GC/MS
3.0 PROCEDURES
3.1 GENERAL
Samples should be extracted in batches of 20 or fewer field samples unless otherwise stated in the project
specific work plan. Quality control samples accompanying each batch may include a procedural blank,
laboratory control sample, matrix spike, reference material and/or field sample duplicate. (See Section
5.0 for more detail).
Each sample, including QC samples, should be spiked with SIS. Add 100 |iL of the SIS Working
Standard so that approximately 1 p.g of each SIS compound is added to each sample.
The tissue samples should be thoroughly homogenized prior to any aliquotting for chemical or physical
characterization, to ensure that a representative sub-sample is taken.
3.2 PERCENT MOISTURE DETERMINATION
Weigh approximately 5 g of well mixed tissue homogenate into a pre-weighed, pre-baked, aluminum
weighing pan and record to the nearest 0.01 g. Place the sample in a drying oven and dry overnight at ca.
105 °C. After approximately 24 h, allow the sample to cool at room temperature for at least 30 min.
Record dry weight to the nearest 0.01 g in sample batch book. Calculate the percent dry weight and/or
percent moisture as in Section 4.0.
3.3 EXTRACTION, CONCENTRATION, AND CLEANUP
Tissue Sample Extraction
• Weigh 20-30 g of well mixed tissue into a 250 mL centrifuge jar (Teflon), weighing the sample to
the nearest 0.01 g. Record sample weight.
• Add approximately 50 g sodium sulfate, mix well with spatula. Add more sodium sulfate if
sample is lumpy — some sodium sulfate should move freely in the solvent indicating that no
additional water is available.
• Add 100 p.L of the SIS Working Standard (ca. 1 jxg each SIS compound into the sample) and 75
mL DCM. Macerate/extract the sample with Tissumizer (or equivalent) at high speed for 2 min.
Avoid spattering the sample.
• Centrifuge the sample for 5 min. at 1500-2000 rpm. Longer centrifugation may be needed for
samples containing fine particulates or suspended matter. Carefully decant the extract into an
Erlenmeyer flask and cap the flask with aluminum foil.
• Repeat the maceration/extraction once more with an additional 75 mL DCM. Centrifuge as above
and decant extract into Erlenmeyer, combining the sample extract.
• Add 50 mL DCM to the sample, seal centrifuge bottle and shake on a shaker table for about 1 hour.
Centrifuge as above, then decant solvent into the Erlenmeyer with the rest of the sample extract.
• Determine the lipid weight (DCM extractable) as follows:
111
-------
Appendix D
Measurement of Polynuclear Aromatic Hydrocarbons by GC/MS
Gently swirl the Erlenmeyer to mix the extract. Mark the meniscus on the outside of the Erlenmeyer.
With a clean 10 mL class "A" volumetric pipette or 10 mL syringe, remove 10 mL of the extract and
place in a pre-weighed aluminum weighing pan. Cover weighing pan with foil, and allow extract to air
dry overnight or until DCM is no longer visible. After drying, weigh pan and record weight. Record the
total extract volume (see under Extract Concentration below) and calculate the lipid content as described
in Section 4.
• Add approximately 20-50 g sodium sulfate to the Erlenmeyer and swirl. Add more sodium
sulfate if sample is lumpy — some sodium sulfate should move freely in the solvent indicating
that no additional water is available. Wait approximately 1 h.
Note: The final combined extract should not be stored in the Erlenmeyer flaskfor more than 1 day. If
longer storage is needed before completing the sample preparation, the extract should be concentrated to
several milliliters (see below), transferred to a glass vial, securely capped, and stored in darkness in a
refrigerator or freezer.
Extract Concentration —Kuderna-Danish (K-D) Technique
• Transfer the sample extract to a K-D flask and receiver (filter through glass fiber filter or glass
wool if there is evidence of suspended particulates or suspended solids). Rinse the Erlenmeyer
twice with about 10 mL of DCM and add to K-D.
• Add 3-5 boiling chips to the K-D receiver and insert a Snyder column. Pre-wet the condenser
column with approximately 5-mL DCM. Place the K-D apparatus in a hot water bath maintained
at 60-65 0 C (monitored by a mercury thermometer), such that the concentrator tube is partially
immersed in hot water and the entire lower rounded surface of the flask is bathed in hot water
vapor. At the proper rate of evaporation, the balls of the Snyder column will actively chatter, but
will not flood with condensed solvent. Continue concentration until the sample volume is
reduced to approximately 10 mL.
• Remove the K-D apparatus and allow it to drain and cool for at least 10 min. Transfer the
concentrator tube to the N-Evap unit and concentrate the sample until the volume is 1-2 mL,
maintaining a water bath temperature of approximately 25 °C.
• Lipid weight extract volume determination: After extract is poured from the original Erlenmeyer,
rinse flask and remove sodium sulfate. Add tap water to the meniscus line drawn prior to
removal of aliquot for lipid weight determination, and determine, and record, the original volume
by measuring it with a 250 mL graduated cylinder. Calculate the lipid weight according to
formulae presented in Section 4.0. Also, determine the lipid weight correction factor so the final
data can be corrected for the proportion of extract removed for the lipid determination.
Note 1: The water bath for the K-D is maintained at 60-65 °C for concentrating DCM, 70-75 ° C for
concentrating acetone, and a boiling bath is used for concentrating hexane or toluene.
Note 2: Adjust the flow of nitrogen on the N-Evap to a gentle stream — do not allow the sample to
bubble or splatter, or have a large "dimple " on the surface, as this will result in the loss of
sample. A small dimple is expected on the surface of the solvent.
112
-------
Appendix D
Measurement of Polynuclear Aromatic Hydrocarbons by GC/MS
Extract Concentration — TurboVap® Technique (optional)
If specified in the work plan, TurboVap® may be used for concentrating the extracts rather than using
K-D.
• Transfer sample extract to a 200 mL TurboVap® tube with a 1 mL collection reservoir.
TurboVap® waterbath temperature should be set at approximately 25 °C; nitrogen pressure should
be approximately 4 to 6 psi.
• Follow manufacturers instructions for operation and maintenance. After initial TurboVap®
concentration the extract should be quantitatively transferred to a 4 mL vial (with DCM rinses),
and concentrated until the volume is 1-2 mL.
Extract Cleanup — Alumina
Alumina cleanup must be performed on tissue extracts to remove gross biogenic material that is co-
extracted with target compounds of interest. This step is performed prior to HPLC/GPC cleanup to
prevent overloading of the size exclusion column.
Column Preparation
• Prepare the 2% deactivated F-20 alumina as described in Section 2.2.
• Pack the tip of the 19/21-mm ("fat") chromatography column with a small piece of glass wool.
Add approximately 10 mL DCM and tap glasswool with clean glass rod to remove bubbles.
Drain solvent into waste cup.
• Add approximately 25 mL DCM to the column. Weigh 40-g, 2% deactivated alumina into a
beaker. Add -50 mL DCM and swirl to remove bubbles. Slowly pour alumina slurry into the
column while rinsing beaker with DCM
• Place approximately 1 g of sodium sulfate on top of the alumina. Tap the column to remove
bubbles. Drain column to top of packing and discard solvent. Column is now ready for use;
columns must be used the day they are prepared.
Column Elution
• Put a clean glass collection flask under the column {e.g., Erlenmeyer flask, K-D apparatus, or
TurboVap® tube).
• Load the 1 to 2 mL sample extract (must be in DCM) onto the column.
• Slowly drain and stop at the top of the alumina packing. Rinse the sample vial with about 1 mL
DCM and load onto column. Slowly drain and stop at the top of the packing. Repeat vial rinse
one time.
• Load with 150 mL of DCM.
• Drain the column slowly (about 2 mL/min), collecting the column eluent containing the target
analytes, and stop at the top of the column — do not drain the column dry into the collection vial.
• Open up and drain the column dry into a solvent waste jar.
113
-------
Appendix D
Measurement of Polynuclear Aromatic Hydrocarbons by GC/MS
Extract Cleanup — GPC
The sample is now ready for further cleanup and processing by Size exclusion HPLC/GPC (described in
Attachment 3 of this document). GPC must be carried out for effective low-level measurement of
contaminants.
• The DCM extract volume should be adjusted to approximately 900 |iL and centrifuged, if
necessary, to settle out any particulates. The volume is measured with a gas tight syringe (to
within 5 fiL), the volume recorded, and the sample transferred to a 2-mL GC vial.
• Process the samples by GPC as described in Attachment 3 of this document.
Following GPC cleanup, the samples are concentrated by K-D or Turbovap© procedures, spiked with
appropriate recovery internal standards (RIS), and submitted for GC/MS analysis of PAH. Each purified
sample extract is concentrated to about 400 jiL, then fortified with 100 |iL of the RIS Working Standard
(1 ng each RIS compound).
4.0 CALCULATIONS
Calculate percent dry weight with the following equation:
% dry weight =
aliquot dry wt
—- - * 100
aliquot wet wt
Calculate percent moisture with the following equation:
% moisture =
aliquot wet wt- aliquot dry wt
* 100
aliquot wet wt
Calculate sample lipid weight with the following equations:
Total lipid weight (mg) =
volume of sample extract (mL) ^
aliquot dry wt. (mgf
aliquot vol(mL)
a corrected for pan weight
Calculate sample dry weight with the following equation:
114
-------
Appendix D
Measurement of Polynuclear Aromatic Hydrocarbons by GC/MS
Sample dry weight (g) = % dry wt. * sample wet wt. (g)
5.0 QUALITY CONTROL
Quality control methods for this procedure consists of preparation and analysis of a select QC samples,
prepared as described below. These QC samples are prepared and analyzed with every set of 20 or fewer
field samples of similar matrix:
• Procedural Blank: 30 grams of sodium sulfate, carried through the entire analytical process.
Used to assess impact of any laboratory-based contamination, if any, on method performance.
• Matrix Spike/Spike Duplicate: 20-25 grams of clean tissue, fortified with 100 |aL of Matrix
Spike Solution (0.5 p.g of each Priority Pollutant PAH), and carried through the entire analytical
process. Used to assess method accuracy and precision with matrices similar to the field samples.
• Laboratory Control Sample: 20 grams of sodium sulfate, fortified with 100 |_iL of Matrix Spike
Solution (0.5 |j.g of each Priority Pollutant PAH), and carried through the entire analytical
process. Used to assess method accuracy in the absence of matrix.
• Standard Reference Material: 5 grams of NIST 1974A, carried through the entire analytical
process, used to assess method accuracy.
• Field Duplicate: A second aliquot of a tissue sample, prepared and analyzed to assess sample
homogeneity and method precision.
The Data Quality Objectives for these QC procedures are described in the body of the document that
accompanies these method descriptions.
115
-------
Appendix D
Measurement of Polynuclear Aromatic Hydrocarbons by GC/MS
This page intentionally left blank.
116
-------
ATTACHMENT 3
Procedures for Gel Permeation HPLC Cleanup of
Sediment and Tissue Extracts for
Semi-Volatile Organic Pollutants
117
-------
This page intentionally left blank.
118
-------
Appendix D
Measurement of Polynuclear Aromatic Hydrocarbons by GC/MS
1.0 OBJECTIVE
The objective of this document is to describe the cleanup of sediment/soil and tissue sample extracts by
Gel Permeation High Performance Liquid Chromatography (HPLC/GPC) prior to analysis of PAH and
potentially other semi-volatile organic compounds. This document describes the operation of the HPLC
as it relates to sample cleanup as well as sample preparation procedures associated with HPLC cleanup.
The HPLC system uses a pump which provides a constant isocratic or gradient flow (depending on the
application), a UV detector, an autosampler, a programmable fraction collector, a recorder or integrator
for monitoring the UV response, and appropriate HPLC column(s) for the particular cleanup/fractionation
to be performed.
2.0 PREPARATION
2.1 APPARATUS AND MATERIALS
1. A liquid chromatography pump capable of providing a constant flow of 5 or 10 mL/min,
depending on the application at hand (Thermo-Separations PI000, Spectra-Physics 8800, or
equivalent).
2. HPLC columns:
(1) GPC. A 22.5 x 300 mm, 100 A pore size, Phenogel GPC/size exclusion column, and
one 7.8 x 50 mm Phenogel precolumn (Phenomenex, Rancho Palos Verde, CA), or
equivalent. Columns are arranged so that the precolumn can be backflushed periodically.
Other column dimensions, and columns connected in series, may be used if acceptable
performance is demonstrated. The GPC is commonly used for separating PCB, chlorinated
pesticide, PAH, and other semivolatile compounds from other tissue and sediment matrix
components, thus purifying the extract before instrumental analysis.
3. Auto sampler capable of injecting 500 to 600 (J.L of sample with minimum sample loss in transfer
lines, loop flushing, or elsewhere (Spectra-Physics AS3000, Gilson 231/401, or equivalent). A
1000 or 2000 (xL sample loop is used to assure that none of the 500 to 600 |iL sample is flushed
to "waste"during sample loading.
4. Programmable fraction collector which is synchronized with the injector (Foxy 200, Gilson 201,
or equivalent).
5. Recorder or integrator (H-P 3396, or equivalent) which is synchronized with the injector and
records the signal from a UV detector (Spectra-Physics 8450, or equivalent) fixed at a suitable
wavelength for the target compounds (typically 254 nm).
6. Microliter syringes
7. Glass vials with Teflon lined caps
1.8 mL HPLC autosampler vials, with Teflon/Silicone/Teflon septa
60 mL fraction collector vials/tubes
119
-------
Appendix D
Measurement of Polynuclear Aromatic Hydrocarbons by GC/MS
2.2 REAGENTS
1. Dichloromethane, hexane, and/or methanol, as applicable (standard laboratory grade). These are
the three solvents routinely used with the system. The actual solvent delivery programs vary with
application, and the most commonly used programs are discussed in Attachment 5. Non-routine
applications and solvent programs will be described in the QAPP.
2. HPLC calibration solution compounds. In general, the first and last eluting compounds should be
in the calibration solution [e.g., 4,4'-dibromooctafluorobiphenyl (DBOFB) and perylene or
benzo(g,h,i)perylene; d8-naphthalene or naphthalene and perylene or benzo(g,h,i)perylene; Cli(l)
and Clio(209)]. Potential interferants/contaminants from which the samples are to be separated
and cleaned (e.g., corn oil and sulfur) may also be added.
2.3 LABWARE PREPARATION
AH glassware must be thoroughly washed with detergent and water, rinsed with deionized water, and
baked in a 400°C muffel oven for a minimum of 4 hr. prior to use. All other labware should be
cleaned with detergent and water and thoroughly rinsed with deionized water.
3.0 PROCEDURE
3.1 GENERAL
Sediment/soil and tissue samples are extracted according to procedures specified in Attachment 1 and
2, respectively, of this document. The extracts are run through a rapid clean-up, filtering, and/or
centrifugation procedure, as needed, concentrated, and submitted for HPLC/GPC cleanup. The
retention time window for the analytes of interest is determined and the instrument calibrated using a
calibration standard. A portion of the sample is injected on the HPLC, the UV response monitored,
and the analyte fraction(s) collected, concentrated and submitted for further instrumental analysis.
3.2 STANDARDS PREPARATION
Prepare an HPLC fractionation time standard appropriate for the cleanup/fractionation to be
performed. The following are examples.
GPC Column - NS&T type PAH, PCB/pesticide analytes. Prepare a standard of approximately 20
fig/mL DBOFB, 20 |ig/mL perylene or 10 (o.g/mL benzo[g,h,i]perylene, 20 mg/mL corn oil, and 40
(i.g/mL sulfur in dichloromethane. This standard will be used as a retention time (RT) marker since
DBOFB elutes towards the beginning and perylene/benzo[g,h,i]perylene towards the end of the
analytes of interest in these analyses, and corn oil and sulfur represent potential interferences.
Perylene elutes just before benzo[g,h,i]perylene.
120
-------
Appendix D
Measurement of Polynuclear Aromatic Hydrocarbons by GC/MS
3.3 PRE-HPLC SAMPLE PREPARATION
1. Pre-HPLC sample cleanup and/or filtration may or may not be necessary, depending on the lipid
content and amount of particulates in the extracts. Specific cleanup/filtration techniques to be
used will be determine on a project by project basis by the project team, and documented in the
sample preparation records.
2. Extracts are concentrated to approximately 900 to 1000 (iL using at first a Kuderna-Danish
apparatus and then nitrogen evaporation. Transfer the sample to a 1.8 mL HPLC autosampler
vial. If particulate matter is present in the sample it may be necessary to filter the.sample.
Alternatively, the sample may be separated from the particulates by centrifugation and carefully
removing the extract with a pipette or syringe, leaving the particulates, and a minimum amount of
sample, behind and then dispensing the sample into a 1.8 mL HPLC autosampler vial. It is
imperative that all samples be totally free of particulate matter and suspended solids. The
presence of particulates and any project specific sample filtration/separation procedures will be
documented in the sample preparation records.
Measure the sample volume to the nearest 10 |j.L and record the volume on a HPLC Split
Documentation Form. Place the samples in a refrigerator until shortly before HPLC analysis, but
do not store for more than 1 day because evaporation may alter the measured volume.
3.4 HPLC FRACTIONATION TIME DETERMINATION
1. Check the frit and distribution disk on the in-line filter, and replace if needed.
2. Turn on the power for HPLC pump, autosampler, fraction collector, detector, and integrator.
Allow at least 15 min for the detector to warm up before use.
3. Rinse out and put fresh solvent in all solvent-holding bottles at least monthly. As needed, de-
aerate the mobile phase and autosampler rinse solvent bottle solvent by bubbling a gentle stream
of clean nitrogen or helium through them for at least 1 hour.
4. Purge air out of mobile phase and auto sampler lines by opening the bypass valve and allowing
the solvent to be purged into a beaker. If necessary, prime the pump prior to purging.
5. Purge air out of syringe/injector rinse solvent lines and flush several milliliters through the dilutor
and sample syringes. No air bubbles should be visible at top of syringe or elsewhere in sample
delivery lines. Minor bubbles in the solvent delivery line are generally acceptable.
6. Make sure the waste bottle can hold the volume of mobile phase that will be used.
7. GPC Column. Before each day of HPLC analysis with the GPC column, backflush the pre-
column with DCM at a flow rate of 2-5 mL/min for 10 to 20 min with the backflushing valve in
the RINSE position. Return the valve to the RUN position before analyzing any standards or
samples. Alternatively, use an electronic switching valve and program it to backflush the
precolumn for 2-4 min at the end of each sample run.
Note: Never backflush the main GPC column. The flow through this column must always be in
the forward direction.
8. Program the autosampler, fraction collector, HPLC pump, and integrator for analysis of the
HPLC fractionation time standard and the samples to be fractionated.
121
-------
Appendix D
Measurement of Polynuclear Aromatic Hydrocarbons by GC/MS
Detailed information on operation and maintenance of the HPLC pump, autosampler, fraction
collector, and detector can be found in the respective instrument manuals.
9. Start the pump. Record the pressure once it has equilibrated. Should the pressure be
unacceptably high (over 700 PSI for the GPC column) it may be necessary to replace the in-line
filter, and possibly the precolumn. Other possible causes of elevated pressure are clogged tubing
at a fitting (reverse flow through tubing) and accumulations in the detector cell (clean cell with 5-
10% nitric acid, as per description in detector manual). Should the pressure be low (under 100
PSI) it may be necessary to replace one of the HPLC pump piston seals.
10. Start the auto sampler program for analysis of HPLC fractionation time standards. Analyze the
HPLC fractionation time standard at least twice. If the RT differences between runs is greater
than 0.1 min (GPC) reanalyze until acceptable RTs are obtained from two consecutive runs.
Average the RTs from the reproducible HPLC runs for the retention time marker compounds
separately. Pressure readings are recorded on the HPLC Sample Fractionation Log Form at the
start of each set of fractionations. Documentation of corrective action and maintenance needed to
correct RT drift is kept with the HPLC.
11. Determine the analyte fractionation window, and program the fraction collector accordingly. The
programmed fractionation time will be recorded, and the fractionation time programming will be
witnessed by another qualified operator, if available. If no-one is available to witness the
programming at the beginning of the run then someone should witness what was programmed in
at the end of the run before samples are removed from the system.
GPC. The following table provides general guidelines for establishing a fractionation window,
for a project. These assume that the commonly used GPC conditions (e.g., single 300 mm long
column, pre-column, 5 mL/min flow rate of 100% DCM) are used. The values presented below
are instrument-specific, and must be determined independently for every HPLC/GPC unit.
Analytes
Start Time
Relative to DBOBF RT
End Time
Relative to Perylene
RT
PAHs
+0.5 min
+1.2 min
3.5 SAMPLE FRACTIONATION/CLEANUP
1. Place a rack with the appropriate number of 50 mL fraction collection tubes on the fraction
collector, and place the HPLC samples on the auto sampler. Complete a HPLC Sample
Fractionation Log Form. Confirm the order and labeling of the HPLC samples (on the
autosampler) and fraction collection tubes (on the fraction collector).
2. Include an HPLC fractionation time standard for analysis approximately every 10-12 samples.
3. Check that the HPLC pump, autosampler, fraction collector, and integrator programs are correct.
4. Start the HPLC sample fractionation/cleanup run by starting the autosampler program. It is
important that the HPLC pump runs continuously after the running of the retention time standards
used for fractionation window determination, because if the pump is shut off and restarted it may
equilibrate at a different pressure and RTs may shift.
122
-------
Appendix D
Measurement of Polynuclear Aromatic Hydrocarbons by GC/MS
3.6 POST HPLC SAMPLE PREPARATION FOR INSTRUMENTAL ANALYSIS
The collected fractions are removed from the fraction collector and concentrated by nitrogen
evaporation or TurboVap®. Concentrate the samples to the appropriate project-specific final volume,
transfer to an autosampler vial, add the internal standard, cap, vortex, and submit for instrumental
analysis.
Visually (with a reference vial) check the volumes remaining in the autosampler vials as soon as the
analysis sequence is completed. If the volume appears to be significantly different from the expected,
measure and record all volumes and adjust the HPLC Injection Volume form, as appropriate. Identify
the injection problem and resolve.
3.7 REFRACTIONATION OF SAMPLES
Should the fractionated sample not be usable for final instrumental analysis (e.g., pump shut off in the
middle of a run), the unfractionated portion of the original extract can be used for HPLC processing.
The unfractionated portion is then brought to a volume of approximately 900 to 1000 (J.L with the
appropriate solvent, the exact volume indicated on an HPLC Split Documentation Form, and the
sample analyzed/fractionated as described in previous sections. Unusable sample fractionations will
be so indicated on the HPLC Sample Fractionation Form, explained, dated, and initialed.
Refractionations will be indicated as such on the new HPLC Sample Fractionation Form that is
prepared, and the number of the refractionation will also be documented {e.g., -R: first
refractionation; -R2: the second time the samples are refractionated).
4.0 CALCULATIONS
Determination of fractionation window is the only calculation required as part of this procedure, and
was discussed in Section 3.4 - items #9 and #10. The portion of the sample processed through the
HPLC is calculated and recorded on the HPLC Split Documentation Form using the following
equation.
First Fractionation: [(IVi)/(EVi)] x 100%
Re-fractionation: [lO((IV,+AV)/EV,)] x [(IVR)/(EVR)] x 100%
IVi = Sample injection volume (fiL) in original/first fractionation
EV] = Pre-HPLC sample extract volume (|iL) in original/first fractionation
AV = Autosampler volume difference between aspiration and dispension (generally 0 |j.L with the
S-P AS3000 and 20 |iL with the Gilson 231/401).
IVR = Sample injection volume (|u.L) in re-fractionation
EVr = Pre-HPLC sample extract volume (|iL) in re-fractionation
123
-------
Appendix D
Measurement of Polynuclear Aromatic Hydrocarbons by GC/MS
5.0 QUALITY CONTROL
1. Quality control methods for this procedure will consist of establishing the RTs of the
fractionation window calibration analytes before each batch of samples are run by multiple
injections of an HPLC fractionation time standard until two analyses result in RT differences of
less than 0.1 min (0.2 min for silica column). The average RTs from these analyses will be used
to program the fraction collector. An HPLC fractionation time check standard will be analyzed
every 10-12 samples to check the RTs of the calibration compounds, which should be ±0.5 min
of the RTs used to program the fraction collector.
2. After significant HPLC maintenance (e.g., new columns installed or major system
DreplumbingQ), or when performance dictates, the performance will be checked against recently
achieve good performance. This will be done qualitatively/semiquantitatively by comparing the
chromatography obtained with the calibration standard (e.g., peak shape and absolute and relative
retention times of the first and last eluting target analytes are reviewed). From these data it can
be determined if the system is performing well (i.e., good peak shape, approximately the same
retention times of the peaks, and approximately the same separation between the peaks).
If the system appears to be performing poorly or differently from before, this should be
investigated and resolved before proceeding with sample analysis. If new
chromatography/performance appears to be correct, the procedure described under item #3 below
should be performed to check/establish retention time windows. If "atypical" system
performance is accepted without additional quantitative check (e.g., item #3 below) this should be
documented and approved by the Laboratory Manager or Project Manager before proceeding with
sample analysis.
3. After major HPLC system changes (e.g., column(s) from different manufacturer, new column
dimensions installed, new dimension or length of lines/tubing), or when performance dictates,
calibrations will be done to determine the size of the fractionation window needed to collect
>95% of, at a minimum, the first and last eluting target analytes. This calibration will be done by
fractionating an injection of the HPLC fractionation time standard and collecting small fractions
(e.g., 0.5 min) from several minutes before to several minutes after the HPLC RTs of these
compounds (a single larger window can be collected between the two RT marker analytes) and
analyzing the fractions by the appropriate instrumental method.
Alternatively, the recoveries of all analytes will be determined by running an analyte standard
mixture through the HPLC fractionation procedure and analyzing for the analytes by the
applicable instrumental method. This recovery check will always be carried out if the HPLC
cleanup procedure is to be used in the analysis of analytes for which the procedure has previously
not been used (i.e., first and last eluting compound is not known). Attachment 7 list recent data
from such a calibration.
From these data it can be determined how long before to how long after the RTs of the RT marker
compounds fraction collection must be performed.
The quality control procedure described under item #1 above will be carried out for every batch of
samples. The procedure under item #2 when column replacement, or other significant maintenance
has been performed. The procedure described under items #3 will be performed when major HPLC
system changes have been performed, when performance dictates, when procedures or analytes have
changed, and at the discretion of the Laboratory Manager. Additionally, relevant sample and analysis
specific information must be recorded on a HPLC Sample Fractionation Log Form.
124
-------
Appendix D
Measurement of Polynuclear Aromatic Hydrocarbons by GC/MS
6.0 MAINTENANCE
This procedural document is not intended to include detail on system maintenance. Maintenance
considerations and procedures are presented in the various instrument manuals. However, a few
specific items that need be considered are listed below. Non-routine maintenance (e.g., pump seal,
column or other hardware replacement, or dislodging of a system clog) should be recorded in the
HPLC System's Log Book. Routine daily maintenance, such as pre-column backflushing and filter
and frit changes, need not be recorded.
Column/Precolumn Cleanup/Conditioning. Before each day of HPLC analysis with the GPC
column, backflush the precolumn with DCM at a flow rate of 2-5 mL/min for 10 to 20 min with the
backflushing valve in the RINSE position. Return the valve to the RUN position before analyzing
any standards or samples. Alternatively, a 2-4 min automatic backflush may be programmed in to the
end of each sample run, if an electronic switching valve is used. In a 30 min GPC run the
backflushing would, typically, be from 26 to 29 min, which allows for a 1 min re-equilibration in the
normal flow direction.
With time the backpressure due to accumulations in the precolumn (see below) will become
unacceptably high, and the precolumn will need to be replaced. It is important to replace it before the
main column begins deteriorating. The main column could last several years if the system is
maintained well.
High/Low Pressure. The system pressure is checked each day before samples are analyzed. Should
the pressure be unacceptably high (over 700 PSI for the GPC column and over 400 PSI for the silica
column) it may be necessary to clean or replace the in-line filter and/or pre-column. Accumulations
in the main column and detector (clean cell with 5-10% nitric acid, as per description in detector
manual) may also cause elevated pressure. Other possible causes of elevated pressure are clogged
tubing at a fitting (reverse flow through tubing) and accumulations in the detector cell. Should the
pressure be low (under 100 PSI) there may be a leak or it may be necessary to replace one of the
HPLC pump piston seals.
125
-------
Appendix D
Measurement of Polynuclear Aromatic Hydrocarbons by GC/MS
This page intentionally left blank.
126
-------
ATTACHMENT 4
Procedures for Identification and Quantification of
Polynuclear Aromatic Hydrocarbons by
Gas Chromatography/Mass Spectrometry
127
-------
This page intentionally left blank.
128
-------
Appendix D
Measurement of Polynuclear Aromatic Hydrocarbons by GC/MS
1.0 OBJECTIVE
This procedure describes the identification and quantification of unsubstituted polynuclear aromatic
hydrocarbons (PAH) and selected alkylated homologues of those PAH by gas chromatography/mass
spectrometry (GC/MS). Table 1 lists potential target PAH. Other PAH and semi-volatile organic
compounds can be added to this list as needed with appropriate demonstration of method applicability.
This method is a modification of EPA methods 8270C and 625 and therefore certain criteria (i.e., initial
calibrations and daily verifications) are different than those referenced in EPA methods
1.1 SUMMARY OF METHOD
Extracts of sediment/soil or tissue samples, prepared as described in Attachments 1-3 of this document,
are analyzzed for polynuclear aromatic hydrocarbons via high resolution capillary gas chromatography
and electron impact mass spectrometry. A data system interfaced to the GC/MS is used to control
acquisition and to store, retrieve, and manipulate mass spectral data. In order to meet risk-based detection
limits, this method requires the operation of the mass spectrometer in selected ion monitoring mode
(SIM).
2.0 PREPARATION
The following procedures document initial procedures to establish the GC/MS system for sample
analysis.
2.1 GAS CHROMATOGRAPH/MASS SPECTROMETER TUNING
Prior to the analysis of analytical standards and/or samples the mass spectrometer must be tuned prior to
starting a new sequence to ensure that the electronic mass to charge (m/z) values are precise to within
+/- 0.5 atomic mass units (AMU) and that the relative abundances of ions across the mass range 50 - 650
AMU meet accepted criteria. The calibration solution used to tune the mass spectrometer is
perfluorotributylamine (PFTBA). PFTBA tuning can be performed automatically through autotuning or
manually with operator control of tune parameter settings.
Hewlett-Packard's EnviroQuant software can automatically tune the instrument using PFTBA to achieve
certain predefined performance criteria. Use of the autotune program minimizes operator-to-operator
variation in tuning procedures and can maximize instrument sensitivity. The autotune program sets target
values for six PFTBA ions(m/z 51, 69, 131, 219, 414, 502, and 614) and adjusts the various tuning
parameters to achieve the relative values listed in Table 2. These target values are the levels that the
masses would have if the instrument were tuned so that a spectrum of DFTPP conformed to EPA criteria.
2.1.1 The AUTOTUNE program is always used as the default tuning method after source cleanings.
2.1.2 AUTOTUNE programs should be run with the "standard values/settings" as a starting point.
2.1.3 The AUTOTUNE program adjusts the voltages of the repeller, ion focus lens, entrance lens,
electron multiplier, and x-ray lens. It also adjusts the AMU gain and offset and the mass gain and
offset.
129
-------
Appendix D
Measurement of Polynuclear Aromatic Hydrocarbons by GC/MS
Table 1. Selected PAH Quantification And Confirmation Ions For GC/MS SIM Analysis
Quantification Recommended
Analyte Ions (AMU) Confirmation Ions (AMU)
[% of Base Peak]"
d8-naphthaleneb
136
134 [15]
naphthalene
128
127 [15]
Crnaphthalenes
142
141 [80]
2-methylnaphthalene
142
141 [80]
1 -methylnaphthalene
142
141 [80]
C2-naphthalenes
156
141
2,6-dimethylnaphthalene
156
141
C3-naphthalenes
170
155
2,3,5-trimethylnaphthalene
170
155
C4-naphthalenes
184
169,141
dio-acenaphtheneb
164
162 [95]
acenaphthylene
152
153 [15]
acenaphthene
154
153 [98]
biphenyl
154
152 [30]
d,0-biphenylb
164
162
dio-fluorene1'
176
174 [85]
dibenzofuran
168
169 [20]
fluorene
166
165 [95]
Crfluorenes
180
165 [100]
C2-fluorenes
194
179 [25]
C3-fluorenes
208
193
dio-phenanthrene"
188
184
phenanthrene
178
176 [20]
anthracene
178
176 [20]
C i -phenanthrenes/anthracenes
192
191 [60]
1 -methylphenanthrene
192
191 [60]
C2-phenanthrenes/anthracenes
206
191
C,-phenanthrenes/anthracenes
220
205
C4-phenanthrenes/anthracenes
234
219, 191
dibenzothiophene
184
152 [15] 139
Crdibenzothiophenes
198
184 [25] 197
C2-dibenzothiophenes
212
197
Cj-dibenzothiophenes
226
211
fluoranthene
202
101 [15]
pyrene
202
101 [15]
C rfluoranthene/pyrenes
216
215 [60]
C2-fluoranthene/pyrenes
230
215
Cj-fluoranthene/pyrenes
244
229,215
d,2-chryseneb
240
236
benz[a] anthracene
228
226 [20]
chrysene
228
226 [30]
C,-benz[a]anthracenes/chrysenes
242
241
C2-benz[a]anthracenes/chrysenes
256
241
C3-benz[a]anthracenes/chrysenes
270
255
C4-benz[a]anthracenes/chrysenes
284
269,241
du-benzo[e]pyreneb
264
260 [20]
d,2-benzo[a]pyreneb
264
260 [20]
130
-------
Appendix D
Measurement of Polynuclear Aromatic Hydrocarbons by GC/MS
TablelTCont^^SelecteclPAHQuantificationAndConfirmatioi^on^oi^rC/M^IIIN^Analyste
Quantification Recommended
Analyte Ions (AMU) Confirmation Ions (AMU)
[~% of Base Peak]*
benzo[b]fluoranthene
252
253 [30], 125 [10]
benzo[k]fluoranthene
252
253 [30], 125 [10]
benzo[e]pyrene
252
253 [20]
benzo[a]pyrene
252
253 [30], 125 [10]
perylene
252
253 [20]
indeno[ 1,2,3-c,d]pyrene
276
277 [25], 138 [30]
dibenz[a,h]anthracene
278
279 [25], 139 [20]
benzo[g,h,i]perylene
276
277 [25], 138 [2.0]
Other:
decalin
138
Cpdecalins
152
C2-decalins
166
C3.decalins
180
C4-decalins
194
benzothiophene
134
C pbenzothiophenes
148
C2-benzothiophenes
162
C3-benzothiophenes
176
C4-benzothiophenes
190
"Relative abundance of ions within any given isomer group will vary considerably, depending on isomer of interest.
Relative abundances should be determined from analysis of crude oil solution. ""Representative spiking compound.
131
-------
Appendix D
Measurement of Polynuclear Aromatic Hydrocarbons by GC/MS
Table 2. PFTBA Tune Criteria
m/z
Ion Abundance2
51
1 - 6 %
69
100%
131
30 - 50%
219
30 - 60%
414
1 - 2 %
502
1 - 6%
614
detectable
"Relative to base peak m/z 69
2.1.4 The relative ion abundances of the three masses m/z 69, 219, and 502 from the tune report should
be compared to the target abundances listed in Table 1. If the abundances are not within range, tuning
should be restarted using standard autotune. If the abundances are still not within range, the instrument
can be tuned manually, as described in Section 2.3, below, to achieve the desired ion abundances. In all
cases, a hard-copy of the final tune report will be placed in the GC/MS Tune Logbook
2.2 The gas chromatograph is fitted with a 60-m x 0.25-mm DB-5MS (J & W Scientific, Inc.) or
equivalent (0.25-p.m film thickness) fused silica capillaiy column. The GC oven temperature
program will be as follows, unless otherwise specified in the project-specific protocol:
Injection Port: 300°C
Transfer Liner: 280°C
Initial Temp: 40°C
Initial Hold: 1 minute
Ramp Rate: 6°C/min
Final Temp: 290°C
Final Hold: 20 minutes or until benzo(g,h,i)perylene elutes
2.2.1 All gas chromatographs are fitted with Electronic Pressure Control (EPC). The EPC ramping
program will be as follows, unless otherwise specified in the project-specific protocol:
Initial Pressure: 30 psi Initial Time: 1 min
Level 1 Rate: 99 psi/min Final Pressure: psi equivalent to flow 1 mL/min
Vacuum Compensation: On
2.3 PREPARATION FOR SELECTED ION MONITORING (SIM) ANALYSIS
The oven temperatures and electronic pressure programs outlined in sections 2.2.1 and 2.2.2 are
used for SIM analysis. A 2-/iL injection volume is analyzed. The mass spectrometer should be
operated in electron impact mode at a manifold pressure of less than 6 x 10"5 torr. The electron
multiplier voltage should be set a minimum of 100 V above the tune value.
132
-------
Appendix D
Measurement of Poly nuclear Aromatic Hydrocarbons by GC/MS
2.3.1 The quantification and confirmation ions used in the analysis of selected PAH are listed in Table 1.
If possible ion groups should be selected so that no more than 20 ions are monitored in a single
group. It should be noted that as the number of ions scanned per group increases and the individual
dwell times decrease, sensitivity will also decrease. Each ion in a group should have identical
dwell times to ensure that correct ion ratios are preserved. Total group dwell time should not
exceed 400 ms, individual dwell times should be a minimum of 20 ms.
2.3.2 Prior to the first analysis of analytical standards and/or samples in SIM, a 10 ng//uL analytical
standard and reference sample (e.g., North Slope Crude Oil) should be analyzed in full-scan mode,
under the conditions presented in Section 2.2. The full-scan total ion chromatograms- (TIC's) from
these analyses should be used to determine the proper group start and stop times in the SIM
acquisition method.
2.3.3 In the event that a section of the analytical column is removed, or the analytical column is
replaced, a 10 ng/fxL analytical standard and reference sample should be analyzed in SIM mode
under the conditions presented in Section 2.3. The extracted ion profiles (EIP's) from these
analyses should be used to reassign the proper group start and stop times of the SIM acquisition
method.
3.0 PROCEDURES
3.1 CALIBRATION
Demonstration of a linear initial calibration is required prior to the analysis of samples. A
calibration verification is required at the beginning and end of each 12 hour period during which
analyses are performed (approximately every eight injections). Sample data may be used if
bracketed by a passing calibration standard. Initial calibrations from previous sequences may be
used if the calibration verifications still show acceptable linearity and no major instrument
maintenance has been performed (i.e., source cleaning, column change, etc.).
3.1.1 Initial Calibration
Analyze a minimum of five analytical standards that will represent sample concentration
range.
3.1.1.1 The concentration of the low standard should be approximately 0.05 ng/|iL (this it about 3-5
times the instrument detection limit when the mass spectrometer is operated in SIM). The
medium concentration standard should be near the concentration range expected in the
samples. The high concentration standard should be approximately 10 ng/|xL. The
concentration of the deuterated internal standards should be approximately at the mid-
concentration range of the calibration curve.
3.1.1.2 For each level of calibration, calculate the response factors for each analyte of interest using
the software supplied with the GC/MS data system. For each analyte, calculate the average
response factor (RF) and the percent relative standard deviation (RSD). For a valid initial
calibration, the percent RSD for each analyte should be less than 25 percent, unless specifically
stated otherwise in project protocols (see Section 4.1 for calculation of RF and RSD). The
average RSD for all targets should be <15%.
133
-------
Appendix D
Measurement of Poly nuclear Aromatic Hydrocarbons by GC/MS
3.1.2 Calibration Verification
Calibration verification will be performed using the following procedure at the beginning and
end of each 12 hour period in which samples are analyzed.
3.1.2.1 Analyze a medium concentration analytical standard under the same analytical conditions used
in the initial calibration. Calculate response factors for each analyte of interest and compare
them to the average response factor calculated in Section 3.1.1.2, above.
3.1.2.2 The percent difference between the average response factor and the calibration verification
response factor is less than 25 percent, the initial calibration is still valid and sample analyses
may continue (see Section 4.2 for calculation of percent difference). If the percent difference
of a target RF exceeds 25 percent, remedial action should be taken and the medium
concentration standard reanalyzed. If the calibration verification fails again, then sample
analyses should be terminated, maintenance performed, and a new initial calibration analyzed,
unless specifically stated otherwise in project protocols. In addition the average percent
difference of all the targets should be <15%. Analysts should be visually monitoring the
retention time (RT) deviation (in minutes) and percent area deviation recorded on the
calibration verification reports. Ideally, these values should not exceed 0.5 minutes for RT, and
<50% or >200% area deviation.
3.2 SAMPLE ANALYSIS
Samples are analyzed under the same analytical conditions used in the analysis of the
analytical standards. A set of Recovery Internal Standards, added to the samples immediately
preceding instrumental analysis, are used to quantify the concentration of target compounds
and Surrogate Internal Standards in samples. The addition and concentration of the RIS are
described in Attachments 1 and 2 of this document.
The criteria presented in Sections 3.2.1 through 3.2.7 must be satisfied to verify identification
of an analyte in a sample. Analyte peaks that do not satisfy these criteria in the preliminary
quant are deleted from the final quantification reports/files.
3.2.1 Relative Retention Time
The sample component relative retention time (RRT) should be within ±0.1 min (6 sec) of the
standard component. The standard must have been run within 12 hours of the sample.
3.2.2 Sample Spectrum
For each analyte suspected to be present in the sample, the corresponding ions listed in Table 1
must be present and the relative intensities must agree to within ±20 percent of the relative
intensities found in spectra of the standard component. This criteria may be modified for trace
level analysis and is left to the project manager's/analyst's discretion. Modifications to sample
spectrum criteria will be documented. If the intensity of the quantification ion in a sample
spectrum is greater than the intensity of the same ion in the high concentration analytical
standard the sample should be diluted and reanalyzed.
134
-------
Appendix D
Measurement of Polynuclear Aromatic Hydrocarbons by GC/MS
3.2.3 Signal to Noise Ratio
A quantifiable analyte peak should (but is riot limited to) exhibit a signal:noise ratio of 3:1 or
greater unless specified otherwise in the project protocol.
3.2.4 Peak Symmetry
Suspected peaks should be examined for proper shape and symmetry.
3.2.5 Patterns
Suspected sample components should display established patterns. See Attachment 1A.
3.2.6 Minimum Area
It is recommended that each of the sample components have a minimum area of 1/2 the
abundance count of the component in the low standard. Quantitation of a component is not
limited to this criteria however. For instance, a component may have an area abundance
slightly less than 1/2 the low standard area abundance but exhibit a signal to noise ratio of 5:1.
This component would be considered for quantitation.
3.2.7 Alkyl Homologues
Alkyl homologues are identified by the procedures in Sections 3.2.8. Quantification of alkyl
homologues is covered in Section 3.2.9, below. Response factors for each alkyl homologue is
assigned that of the parent (unsubstituted) PAH.
3.2.8 QUANTIFICATION OF ANALYTES
Quantification of analytes identified in samples will be performed by the internal standard
method, using the average response factor from the initial calibration, unless otherwise
specified in the project protocols. See Section 4.3 for additional information regarding
calculations used for the determination of target analyte concentrations in samples.
3.2.9 Quantification of Alkyl Homologues
Alkyl homologue groups are quantified by the internal standard method. The molecular ion of
the alkyl homologue should be extracted, and the areas of the individual isomers summed by a
straight line, baseline integration.
3.2.9.1 The response factor used to quantify a specific homologous series (e.g., Crnaphthalenes)
should be the response factor of parent compound (e.g., naphthalene) unless specified
otherwise in the project protocols.
3.2.10 Manual Integrations
Manual integration is an important component of sample analysis at Battelle to ensure that the
most accurate data are generated. Given the complex nature of environmental samples,
Battelle considers it critical that a trained analyst reviews each individual analyte in each
135
-------
Appendix D
Measurement of Polynuclear Aromatic Hydrocarbons by GC/MS
sample using a chromatography data system, and manually alters and optimizes the integration
as needed. Each analyst is trained in the proper integration of peaks.
The data system baseline integration can be used to quantify samples. If, during the review of
the chromatogram, the analyst notices improper integration by the data system, he/she will
manually reintegrate the peak using proper integration techniques.
Proper integration techniques will account for near co-eluted peaks, negative peaks, and other
peak shape or baseline anomalies which often occur in environmental samples.
4.0 CALCULATIONS
4.1 INITIAL CALIBRATION
The average response factors from the initial calibration are calculated using the following
equation:
RF = (Aa/Ab) x (CJCt)
where:
Aa = Area of quantification ion for target analyte in the standard
Ab = Area of quantification ion for internal standard in the standard
Ca = Concentration of target analyte in the standard
Cb = Concentration of internal standard in the standard
The percent relative standard deviation (RSD) is evaluated with following equation:
RSD = (SD/RF)x 100
where:
SD = standard deviation of the mean RF, calculated as n-1.
See Section 3.1.1.2 for details regarding initial calibration acceptability criteria.
4.2 CALIBRATION VERIFICATION
The response factors determined from the calibration verifications are checked against those
determined from the initial calibration. The percent difference is calculated using the
following
Percent Difference = [(RFi - RFc)/RFi] x 100
where:
RFi = Average response factor from initial calibration
RFC = Response factor from continuing calibration
See Section 3.1.2.2 for calibration verification acceptability criteria.
136
-------
Appendix D
Measurement of Polynuclear Aromatic Hydrocarbons by GC/MS
4.3 QUANTIFICATION OF SAMPLES
Samples are quantified as detailed in Section 3.2. The concentration of target analytes is
determined using the following equation:
Ca = [(Aa/Ai) x (Amti/RFi) x D]/Va
where:
Ca = Concentration target analyte
Aa = Area quantification ion for target analyte
Aj = Area quantification ion for internal standard
Amtj = Amount internal standard added to sample
RFi = Average RF for analyte determined from initial calibration
D = Dilution factor if applicable
Va = Sample size
Sample size may refer to sample volume or sample dry/wet weight. The project protocol will
specify reporting criteria.
5.0 QUALITY CONTROL
The GC/MS Facility should be operated and documented by facility specific SOPs that document
system operation, maintenance, and calibration.
Hardcopies of all initial and continuing calibrations are maintained with the project data. The number
of procedural blanks, spiked blanks, matrix spikes, and other types of quality control samples will be
specified in individual project protocols.
5.1 Each day that analysis is performed, the daily calibration standard should be evaluated to
determine if the chromatographic system is operating properly.
5.1.1 Peak shape should be evaluated for proper peak shape and symmetry.
5.1.2 The instrumental response should be comparable to previous calibrations.
5.1.3 The system must be recalibrated if the analytical column is replaced.
5.2 There must be an initial calibration of the instrument as specified in Section 3.1.1.
137
-------
Appendix D
Measurement of Polynuclear Aromatic Hydrocarbons by GC/MS
ATTACHMENT 1A— Alkyl Homologue Extracted Ion Profile Patterns
Abundance
Ion 128.00 (127.70 to 128.70): A5457.D
120000
100000
80000
60000
40000
20000
0
Time—>
Abundance
NA 3HTHALENES
rmj i i i i | n i rp
19.0CEO.OC21.0CE2.0<23.0(E4.0(E5.0CE6.C(E7.0(E8.0(E9.0CBO.OGB1.0(B2.00
Ion 142.00 (141.70 to 142.70): A5457.D
150000
100000
50000
Time—> ®
Abundance
80000
60000
40000
20000
Time—>
Abundance
Ci-MAPHTHALENES
I
19.0C20.0CE1.0CE2.0(E3.0(E4.0CE5.0(E6.0C27.0G28.0(E9.0CBO.OCB1.0CB2.00
Ion 156.00 (155.70 to 156.70): A5457.D
ICS-NAPHTHALENES
I !
! il
n-'' 'vi ¦
A k
'TmrT1
19.0CE0.0CB1 .OC22.0C23.0C24.0C25.0CE6.0CB7.0(E8.0(29.0(BO.OQ81.0CB2.00
Ion 170.00 (169.70 to 170.70): A5457.D
40000
30000
20000
10000
0
Time-->
Abundance
C3-NAPHTHALENES
fr ¦ ¦ ' A h a f
19.0CE0.0C21 .OCE2.OCE3.OCB4.OC25.OCE6.OCE7.OCE8.OCE9.OCBO.OCB1.0CB2.00
Ion 184.00 (183.70 to 184.70): A5457.D
Time
40000
30000
20000
10000
0
C4-NAPHTHALENES
JUUA.A.U'. i.
19.0C20.0Cei ,OCE2.0C23.0C24.0CE5.0C26.0CE7.0CE8.0(E9.0CBOOCB1 0CB2.00
138
-------
Appendix D
Measurement of Polynuclear Aromatic Hydrocarbons by GC/MS
Abundance
ATTACHMENT 1A (continued)
Ion 166.00 (165.70 to 166.70): A5457.D
12000
10000
8000
6000
4000
200°L_/Lyu
ILUORENE
Time—>
Abundance
29.00 30.00 3100 32^00 33!oo 34I0O 35!00 36!00 37.00
Ion 180.00 (179.70 to 180.70): A5457.D
15000
10000
5000
0
Time—>
Abundance
-A-
1-FLUORENES
29.00 30.00 31.00 32.00 33.00 34.00 35.00 36.00 37.00
Ion 194.00 (193.70 to 194.70): A5457.D
8000
6000
4000
2000
Time->
Abundance
j^-FLUORENES
vi; j[J ' v v\. ¦../*- > '
"fTf, 1"T"l"f'T I I | I I I I | I I I 1 | 1 1 r 1 | I 1 I I | I ' I I | I I I 1 |
29.00 30.00 31.00 32.00 33.00 34.00 35.00 36.00 37.00
Ion 208.00 (207.70 to 208.70): A5457.D
5000
4000
3000
2000
1000
0
Time->
C3-FLUORENES |
llll iWI fi
hi rJ
~l—1—'—1—1—I-
29.00 30.00 31.00 32.00 33.00 34.00 35.00 36.00 37.00
139
-------
Appendix D
Measurement of Polynuclear Aromatic Hydrocarbons by GC/MS
Abundance
ATTACHMENT 1A (coutinued)
Ion 178.00 (177.70 to 178.70): A5457.D
40000
20000
0
Time—>
Abundance
PHENANTHRENE
33.00 34.00 35.00 36.00 37.00 38.00 39^00 40!00 41 loo 42!00 43^00
Ion 192.00 (191.70 to 192.70): A5457.D
30000
20000
10000
0
Time—>
Abundance
C1-^H|NANTHRENES/ANTHRACENES
mi ii
/'t-A.
33.00 34.00 35.00 36.00 37.00 38.00 39.00^40.00 41 !oO 42.00 43.00
Ion 206.00 (205.70 to 206.70): A5457.D
30000
20000
10000
0
Time—>
Abundance
C2-PHENANTHRENES\ANTHRACENES
33.66 34!00 35^00 36^00 37^00 38.00 39.00 4o!oO 41.00 4Z00 43!oO
Ion 220.00 (219.70 to 220.70): A5457.D
15000
lOOOQ^HENANTHRENESNANTHRACENES I
5000
0
Time—>
Abundance
33.00 34.00 35.00 36.00 37.00 38.00 39.00 40.00 41.00 42.00 43.00
Ion 234.00 (233.70 to 234.70): A5457.D
5000
0
C4-PHENANTHRENES/ANTHRACENES
-t T-T-p
Time—>
33.00 34.00 35.00 36.00 37.00 38.00 39.00 40.00 41.00 42.00 43.00
140
-------
Appendix D
Measurement of Polynuclear Aromatic Hydrocarbons by GC/MS
ATTACHMENT 1A (continued)
Abundance
Ion 184.00 (183.70 to 184.70): A5457.D
40000 -
30000^)IB4NZOTHIOPHENES
20000
10000
0
4-U
'i ''|vv t i i | i i i r| i i i i | i v i*r | r-TTi"|' r i r rq~r~T i i | i i . i | i i i i | i i
32.00 33.00 34.00 35.00 36.00 37.00 38.00 39.00 40.00 41.00
Time—>
Abundance
40000
30000
20000
10000
0
Time—>
Abundance
Ion 198.00 (197.70 to 198.70): A5457.D
Ci-DIBENZOTHIOPHENES
ii || ii
i - i |" 1 "i i Vs i i^l i | i v rr | ¦ n i*"i "r [¦ r f i i | ~i ,-*r i i | i i i i | i i i i [ i i i i
32.00 33.00 34.00 35.00 36.00 37.00 38.00 39.00 40.00 41.00
Ion 212.00 (211.70 to 212.70): A5457.D
| G2-DIBENZOTHIOPHENES
UXMju,
r ' C r-
,—p.
Time—>
Abundance
32.00 33.00 34.00 35.00 36.00 37.00 38.00 39.00 40.00 41.00
Ion 226.00 (225.70 to 226.70): A5457.D
10000
5000
0
Time—>
Abundance
C3-DIBENZOTHIOPHENES
,LA»
32.00 33.00 34.00 35.00 36.00 37.00 38.00 39.00 40.00 41.00
Ion 240.00 (239.70 to 240.70): A5457.D
3000
2000
1000
0
C4-DIBENZOTHIOPHENES
ii A
/ vV
4 ii
¦ WV !«! \ .v^
32.00 33.00 34.00 35.00 36.00 37^00 38!00 39^00 4o!oO 4l!oO
Time—>
141
-------
Appendix D
Measurement of Polynuclear Aromatic Hydrocarbons by GC/MS
Abundance
ATTACHMENT 1A (continued)
Ion 216.00 (215.70 to 216.70): A5457.D
ORANTHENES/PYRENES
1-FLl
~~1 I I ~j I I I I | I—I—I I | I I I I "J 1 I I I~~j I T I I J I I I I-1 I I I I | I I I I J I I I r~
38.00 39.00 40.00 41.00 42.00 43.00 44.00 45.00 46.00
Time—>
Abundance
Ion 230.00 (229.70 to 230.70): A5457.D
3000'
2500
2000
1500
1000
500"
0"
di
-fll|oranthenes/pyrenes
!
Abundance
Ion 244.00 (243.70 to 244.70): A5457.D
3000
2500
2000
1500
1000
500
0
C3-FL(!JORANTHENES/PYRENES
38!00 39^00 40!00 41 !00 42^00 43^00 44^00 45^00 46^00
Time—>
142
-------
Appendix D
Measurement of Polynuclear Aromatic Hydrocarbons by GC/MS
Abundance
6000
4000
2000
ATTACHMENT 1A (continued)
Ion 228.00 (227.70 to 228.70): A5457.D
CHRYSENES
Time—>
Abundance
—i—i—r—j—>—i—i—i—|—i—i-*t—i—|—i—i—i—«—|—?—i—i—i—|—i—r—i—i—|—i—i—i—i—|—i—i—i—i—j—i—i—i—r—|
42.00 44.00 46.00 48.00 50.00 52.00 54.00 56.00 58.00
Ion 242.00 (241.70 to 242.70): A5457.D
5000
4000
3000
2000
1000
0
1-CHRYSENES
1)!
,
42.00 44.00 46.00 48.00 50.00 52.00 54.00 56.00 58.00
Time—>
Abundance
Ion 256.00 (255.70 to 256.70): A5457.D
1500
1000
500
0
ill
C2-CHRYSENES
Time—>
Abundance
800
600
400
200
1 1 > ' i 1 1 1 1 i 1 1 1 1 i 1 ' ' 1 l 1 ' 1 1 l 1 1 1 1 I ' 1 1 1 l i ' 1 1 l i 1 1 1 l
42.00 44.00 46.00 48.00 50.00 52.00 54.00 56.00 58.00
Ion 270.00 (269.70 to 270.70): A5457.D
M
J
If
Vfc\j
C3-CHRYSENES
L
Time—>
Abundance
400
300]
200
100
1 1 1 ! 1 1 1 ' I 1 ' ' ' I ' 1 1 1 I ' 1 ' 1 I 1 1 1 ' I 1 ' 1 1 I 1 1 1 ' I 1 1
42.00 44.00 46.00 48.00 50.00 52.00 54.00 56.00 58.00
Ion 284.00 (283.70 to 284.70): A5457.D
„ 4it jLiM C4-|chrysenes
Vw V Vw*
***
Time—>
42.00 44loO 46 loo 48!oO 5o!oO 52!oO 54loO 56!oO 58!oO
143
-------
Appendix D
Measurement of Polynuclear Aromatic Hydrocarbons by GC/MS
ATTACHMENT 1A (continued)
Ion 138.00 (137.70 to 138.70): A5457.D
Abundance
10000
5000
"T T i i | 'l i i i -J t~i—I I p l l l | I I I—i -j i—n—r~|—i I I I—|
Time—> 16 00 17.00 18.00 19.00 20.00 21.00 22.00 23.00 24.00 25.00 26.00 27.00
Abundance
DECALIN
10000
5000
U,
Ion 152.00 (151.70 to 152.70): A5457.D
-DECALINS
¦ i. (vVl|~*rli'VT'>p'i'"iT l11 i'| ll"r"T'ryT~r'i~r'prTT'i~[T'W r p i T"'"j W i i | i"'i i^i |
Mi
Time—> 16-00 17.00 18.00 19.00 20.00 21.00 22.00 23.00 24.00 25.00 26.00 27.00
Abundance
Ion 166.00 (165.70 to 166.70): A5457.D
4000
2000
;2-DECALINS
I 1 ¦ ¦ ¦ I
"T'm . i
Time—> 16 00 17 00 18 00 19 00 20 00 21 00 22 00 23 00 24 00 25 00 26 00 27 00
Abundance
Ion 180.00 (179.70 to 180.70): A5457.D
2000
1500
1000
500
C3-DECALINS
i ¦ ' -r
if' \vV\/' Aju
l 1 1 1 1 l
Time—>
Abundance
16.00 17.00 18.00 19.00 20.00 21.00 22.00 23.00 24.00 25.00 26.00 27.00
Ion 194.00 (193.70 to 194.70): A5457.D
2000
1000
C4-DECALINS
,AvV*\
.kJl
I ' 1 ' 1 I
Time—> 16 00 17 00 18 00 19 00 20 00 21 00 22 00 23 00 24 00 25 00 26 00 27 00
144
-------
APPENDIX E
EPA Region 2 Review of NY/NJ Harbor
Data on Organotins
-------
Appendix E
EPA Region 2 Review of NY/NJ Harbor Data on Organotins
Squibb etal. (1991) compiled available data on chemical concentrations in sediment, water, and fish
collected from the New York/New Jersey Harbor Estuary. Chemicals were categorized as "of concern" or
"not of concern", based on comparison of measured concentrations in water or biota with available federal
and state marine water quality or fish tissue standards for the protection of marine life and human health,
and with NOAA effects-range values for sediment concentrations. Due to the varying confidence in the
data supporting these classifications, the identification of certain chemicals (including lindane and
hexachlorobenzene) as "of concern" by the authors was intended to indicate a need for further evaluation
and monitoring before a definitive conclusion could be reached. Squibb et al. (1991) documented their
conclusions in a December, 1991 report to the NY/NJ Harbor Estuary Program's Toxics Workgroup
entitled "NY-NJHarbor Estuary Program Module 3.1: Toxics Characterization Report".
EPA Region 2 reviewed relevant data sets that have been generated in the NY/NJ Harbor estuary and the
New York Bight since Squibb et al. (1991) along with other scientific information to evaluate the need to
add constituents (i.e., lindane, hexachlorobenzene (HCB), organotins (TBT), alkylated PAHs and
endocrine disruptors) that were specifically mentioned by 1998 peer reviewers to the list of required
analytes for the evaluation of dredged material that is proposed for use as Remediation Material at the
HARS. Based on this review, EPA Region 2 concluded that there is insufficient evidence to require
analysis of lindane and HCB, and that the data supports the need to require analysis of organotins (TBT).
The results of the review conducted for organotins (TBT) review are summarized below.
Organotins
Tributyl tin (TBT) as the sum of TBT-related compounds (monobutyl-, dibutyl-, tributyl-, and
tetra-n-butyl tin) [CAS Registry Number: 56-35-9] is one of several man-made organotin compounds
with various industrial uses. Tri-substituted organotins find uses as biocides in agriculture and industry.
Tributyl tin may be used as an anti-fouling agent in marine paints within the limits set by the Organotin
Antifouling Paints Control Act of 1988, which restricts the release rate of organotin paints used on ships
in the U.S. TBT has been linked to a disturbance in sex hormone production and damaged immune
systems. Organotins are now being recorded in fish at the top of food chains which spend their lives in
deeper water well away from crowded shipping lanes and coastal maritime traffic. In fact, marine wildlife
the world over is now contaminated with organotins. Organotins are known to cause not only
reproductive disturbances in marine molluscs, but also damage to the central nervous system in mammals.
They have also been linked with disturbances in immune systems in humans.
TBT concentrations in polychaete tissues from 10 stations located in and around the former Mud Dump
Site were measured and reported in Battelle (1997). Although sampling of this contaminant was limited,
accumulated concentrations of TBT were measured in polychaetes from all samples and were generally
higher in polychaetes sampled within the HARS than in polychaetes sampled outside the HARS
(i.e., average concentrations of 34.5 ppb [63.2 ppb total organotins] and 17.2 ppb [34.1 ppb total
organotins], respectively). Organotin concentrations measured in tissues of these organisms were as
much as an order of magnitude higher than the sediments from which they were collected. This data
suggests that bioaccumulation of TBT is occurring in NY Bight polychaetes and that historic disposal of
New York/New Jersey Harbor dredged material may be a contributing factor.
TBT concentrations in environmental media of the NY/NJ Harbor estuary were not evaluated by Squibb
et al. (1991). EPA Region 2 reviewed the regional NPDES database for dischargers in the lower Hudson
River estuary (Bronx, NY to Sandy Hook, NJ) and determined that there are no known permitted
discharges of TBTs to the lower estuary. However, TBT concentrations in surficial sediments collected
145
-------
Appendix E
throughout New York/New Jersey Harbor in 1993-94 are reported in EPA (1998). Harborwide-average
sediment concentrations of TBT and total butyltins \tfere 30 ppb and 50 ppb, respectively. Sediment
concentrations of total butyltins in New York/New Jersey Harbor ranged (from not detected) to 520 ppb.
Sediments exceeding 100 ppb were concentrated in the Newark Bay/Kills complex, and Raritan and
Jamaica Bays. Limited data for TBT residues in benthic organisms is available in the NY/NJ Harbor
system. Eighteen mussel samples from NY/NJ Harbor were analyzed for TBT (total butyltin) body
burdens by NOAA (Status and Trends) from 1989-95. These data showed a gradient in wet tissue
residues that decreased from an average of 541 ppb in the Upper/Newark bays, to an average of 210 ppb
in the Lower/Raritan bays, and an average of 100 ppb in Jamaica Bay. These concentrations have been
associated with adverse effects to sensitive ecological receptors (see Jarvinen and Ankley, 1999; Meador,
2000)
Based on the above, EPA Region 2 is proposing to add organotins (TBT and butyltins) to the list of
required analytes for evaluation of dredged material proposed for use as Remediation Material at the
HARS.
References Cited:
Adams, D.A., J.S. O'Connor, and S.B. Weisberg. 1998. Sediment Quality of the NY/NJ Harbor System.
An Investigation under the Regional Environmental Monitoring and Assessment Program (R-EMAP).
Final Report. EPA/902-R-98-001. March 1998.
Battelle (Battelle Ocean Sciences). 1997. Contaminants in polychaetes from the Mud Dump Site and
environs. Prepared for U.S. EPA Region 2 (Work Assignment 3-133, Contract No. 68-C2-0134), dated
March 4, 1997. Various pagings.
Jarvinen, A.W. and G.T. Ankley. 1999. Linkage of Effects to Tissue Residues: Development of a
Comprehensive Database for Aquatic Organisms Exposed to Inorganic and Organic Chemicals. Society
of Environmental Toxicology and Chemistry (SETAC) Press. Pensacola, FL. 364 pp.
Meador, J.P. 2000. Predicting the fate and effects of tributyltin in marine systems. Rev. Environ.
Contam. Toxicol. 166:1-48.
NOAA. 1995. Magnitude and Extent of Sediment Toxicity in the Hudson-Raritan Estuary. NOAA
Technical Memorandum NOS ORCA 88. Silver Spring, MD. 230 pgs.
Squibb, K.S., J.M. O'Connor, and Kneip, T.J. 1991. Toxics Characterization Report, Module 3.1.
Report prepared by Institute of Environmental Medicine, NY Univ. Medical Center for the NY/NJ Harbor
Estuary Program.
146
-------
APPENDIX F
Proposed HARS-Specific Values for Assessing
Human Health Risks Associated with Contaminants
Accumulated from Dredged Material
(Ecological risks are the subject of a separate peer review
effort)
-------
Appendix F
Proposed HARS-Specific Values for assessing human health risks associated with contaminants
accumulated from dredged material. (Ecological risks are the subject of a separate peer review
effort).
Proposed HARS-Specific Values
Compound
Human Health Cancer Level
Human Health Non-Cancer
Level (HQ=1)
PAHS
(Hg/Kg)
(Ug/Kg)
Acenaphthene
10,135,135
Anthracene
50,675,676
Fluorene
6,756,757
Naphthalene
3,378,378
Phenanthrene
50,675,676
Benzo(a)pyrene
2,314"
Fluoranthene
6,756,757
Pyrene
5,067,568
Chlorinated Organic Contaminants
(Ug/Kg)
(Ug/Kg)
Aldrin
33
169
Dieldrin
66
528
aChlordane
1,664
291
Heptachlor
129
2912
Heptachlor epoxide
133
157
Total Endosulfansb
92,138
Total DDT"
1,656
2,815
Total PCBs"
282
113
1,4-Dichlorobenzene
70,383
506,757
METALS
(mg/Kg)
(mg/Kg)
Arsenicd
4.67
21
Cadmium
50
Chromium'
45
Copper
2,211
Lead
6.3
Mercury'
0.141
Nickel
250
Silver
63
Zinc
15,641
(Mg/Kg)
(Mg/Kg)
Tributyltin (new constituent)
4,954
' Cancer risks of entire PAH mixture is assessed using benzo(a)pyrene potential potency equivalence and molar concentrations of
individual PAHs
b Measured concentrations of individual contaminants contributing to this total will be adjusted by multipliers that reflect differences
in trophic transfer of the contributing compounds to calculate adjusted total residues for comparison to HARS-Specific Values (see
Section E)
c In cases where a project tissue data for a total metal exceeds the HARS-Specific Value, the applicant would be given the
opportunity to re-analyze for the speciated forms of the metals to allow a direct comparison to the Value.
d A factor of 0.1 will be applied to the measured total arsenic for comparison to the HARS-Specific Value.
147
-------
Appendix F
This page intentionally left blank.
148
-------
APPENDIX G
Use of Ranges of Metals Body Residues Measured in
Field-Collected Polychaetes to Derive a Safety Factor to
Account for Potential Underestimates of Metals
Uptake by 28-day Exposure Duration of
Bioaccumulation Assay
-------
Appendix G
To develop an appropriate safety factor for application to 28-day test results for non-
essential metals to account for potential differences from 'steady state', data on sediment
metals content and resident infaunal benthos (i.e., polychaetes) in the vicinity of the
HARS were compiled and compared. Sediment concentrations and polychaete tissue
concentrations in 14 co-located samples from the vicinity of the HARS, reported by
Battelle (1996,1997), were used in this effort.
It is assumed that metals concentrations measured in the organisms collected in this effort
represent a range of exposure durations and conditions that are typical of benthic
organisms at the HARS. The results of this analysis indicate that despite sediment
concentrations which varied by as much as two orders of magnitude, tissue
concentrations of non-essential metals in field collected benthic organisms only varied
within a factor of three (i.e., maximum reported concentrations of all non-essential metals
were approximately three times higher than the lowest concentration reported).
Therefore, a safety factor of three will be applied to the results of the 28-day
bioaccumulation test for non-essential metals (i.e., silver, cadmium, mercury, and lead) to
reflect this pattern in field tissue accumulation. Actual metals concentrations and the
calculated ranges are reported in the table below (Table G-l).
149
-------
Table G-l. Metals concentrations measured in co-located polycheate tissue and sediment samples collected from the
vicinity of the HARS.
Silver
Arsenic
Cadmium
Chro
mium
Cop
per
Site
Tissue
Sediment
Tissue
Sediment
Tissue
Sediment
Tissue
Sediment
Tissue
Sediment
(ug/g, dry)
(ug/g, dry)
(ug/g, dry)
(ug/g, dry)
(ug/g, dry)
(ug/g, dry)
(ug/g, dry)
(ug/g, dry)
(ug/g, dry)
(ug/g, dry)
4
0.63
0.17
16.59
4.14
0.33
0.11
15.89
44.70
22.34
17.90
7
0.75
1.59
32.67
15.00
0.25
0.82
11.34
81.30
27.00
56.60
9
0.51
0.33
14.75
6.00
0.31
0.17
8.11
46.10
15.48
23.10
10
0.40
1.70
16.44
14.00
0.28
1.38
6.44
104.60
11.33
68.90
11
0.83
1.40
24.32
10.90
0.28
0.89
10.12
87.40
26.46
55.80
14
0.60
1.87
18.40
18.80
0.70
1.35
9.02
90.70
12.89
66.70
15
0.41
7.33
24.24
21.50
0.31
3.22
8.18
187.20
11.42
178-20
19
0.80
0.88
17.34
22.60
0.38
0.70
17.55
121.40
12.12
109.40
27
0.36
0.80
16.05
11.40
0.25
0.46
8.03
142.60
8.24
99.90
28
0.79
0.84
31.01
20.30
0.27
0.50
15.35
107.20
23.21
71.00
29
0.49
2.65
30.28
12.60
0.27
1.49
12.36
82.90
14.02
57.00
32
0.44
0.11
17.33
3.30
0.66
0.08
9.14
40.40
14.06
16.10
33
0.56
0.71
15.76
11.70
0.59
0.33
12.38
91.20
13.92
62.20
49
0.68
0.44
14.41
4.79
0.42
0.42
9.74
45.40
16.18
27.60
Min
0.36
0.11
14.41
3.30
0.25
0.08
6.44
40.40
8.24
16.10
Max
0.83
7.33
32.67
22.60
0.70
3.22
17.55
187.20
27.00
178.20
Max/Min
2.3
66.6
2.3
6.8
2.8
40.3
2.7
4.6
3.3
11.1
-------
Table G-l. Metals concentrations measured in co-located polycheate tissue and sediment samples collected from the
vicinity of the HARS. (cont)
Mercury
Nickel
Lead
Zi
DC
Site
Tissue
Sediment
Tissue
Sediment
Tissue
Sediment
Tissue
Sediment
(ug/g, dry)
(ug/g, dry)
(ug/g, dry)
(ug/g, dry)
(ug/g, dry)
(ug/g, dry)
(ug/g, dry)
(ug/g, dry)
4
0.21
0.14
6.49
14.40
28.69
25.30
126.25
59.20
7
0.28
1.28
4.63
26.00
20.77
84.60
110.60
131.00
9
0.12
0.18
3.88
13.30
13.15
37.90
137.70
65.90
10
nd
1.09
3.62
24.10
12.22
98.10
135.45
151.60
11
0.24
0.82
4.24
37.30
18.10
80.50
87.50
129.70
14
0.13
1.10
4.36
23.10
12.86
86.80
114.42
151.40
15
nd
2.47
5.55
40.70
12.39
194.00
165.14
329.00
19
0.33
0.88
6.69
35.20
25.40
402.00
102.28
255.00
27
nd
0.30
11.48
88.90
11.68
111.10
137.63
213.00
28
0.34
0.52
5.38
32.60
24.12
110.80
95.39
166.00
29
0.21
1.37
6.98
51.20
18.47
71.70
152.64
121.70
32
0.13
0.06
5.46
27.80
14.82
39.10
153.50
55.90
33
nd
0.41
5.53
29.40
22.46
91.60
114.69
134.70
49
0.21
0.32
4.92
12.10
22.54
30.60
134.73
57.10
Min
0.12
0.06
3.62
12.10
11.68
25.30
87.50
55.90
Max
0.34
2.47
11.48
88.90
28.69
402.00
165.14
329.00
Max/Min
2.8
41.2
3.2
7.3
2.5
15.9
1.9
5.9
-------
Appendix G
This page intentionally left blank.
152
-------
APPENDIX H
Derivation of an HARS-Specific Tissue
Guideline Value for Lead
-------
Appendix H
Derivation of an HARS-Specific tissue guideline value for lead
Provided below is the methodology used to calculate acceptable levels of lead in benthic test
tissues for dredged material proposed for use as Remediation Material at the HARS. The
methodology uses the best available information for lead from EPA's Air Quality Criteria for
Lead Document (EPA (1986)) and an approach similar to that used by the Agency in the 1991
National Primary Drinking Water Regulations for lead and copper (EPA (1991a)). The
calculations estimate blood lead contributions from the different routes of exposure through use
of regression coefficients derived from various epidemiological studies. Baseline blood lead
values are derived from NHANES III data (Brody, et. al., (1994)). While other approaches have
been used by the Agency for various purposes, we used the disaggregate approach based on its
reliance on published, peer reviewed studies and its relative simplicity.
The calculations are based on the "disaggregate" modeling approach in Chapter 13 (Table 13-6)
of the 1986 Air Quality Criteria Document (EPA (1986)) using Table 5-1 of the 1989 OAQPS
(Office of Air Quality Planning and Standards) Staff Report: "Review of the National Ambient
Air Quality Standards for Lead: Exposure Analysis Methodology and Validation" (EPA (1989)).
Both documents relied on published literature and were extensively peer reviewed by the
Agency's Clean Air Science Advisory Committee.
Consistent with EPA policy on regulation of lead, the analysis is protective of children. This
analysis was conducted for lead because no reference dose or cancer risk potency factors have
been established for lead. For assessment of potential human health effects of exposure to lead,
EPA adopted levels of concern for lead (see first bullet below) in human blood, and these were
used as the basis of this analysis. Key factors in the analysis are the following:
1. The EPA Level of Concern for Lead in Blood is a level of 10 |ig/dl (EPA (1991a); EPA
(1991b)).
2. To keep as many children as practical (i.e., 95.0%) under 10 ng/dl, EPA has determined that
the population geometric mean blood lead level in children should be approximately 4.6
ug/dl (EPA (1989)).
3. Based upon the above, 4.6 (ig/dl was used as the basis for deriving acceptable benthic test
tissue levels; 4.6 |j.g/dl was used instead of the EPA level of concern of 10 |ig/dl so as not to
result in an overall increase in blood lead levels in the general population.
4. The basic formula used is:
Acceptable lead = Acceptable blood lead - Current human exposure
exposure from fish levels (i.e., 4.6 ug/dl) from all routes
5. The specific calculations are set forth below:
153
-------
Appendix H
I. Calculate the current human exposure from all routes (i.e., background) regarding their
contributions to average blood lead levels. •
(A.) First estimate average current exposure concentrations in different media:
Drinking water: 4 ppb (EPA (1994a); EPA (1994b)); EPA's OGWDW estimate based upon
lead compliance monitoring data of average lead levels if the lead action level of 15 ppb
is met.
Soil/dust/paint (approximate background for NYC): 800 ppm (EPA (1986); EPA (1996)).
Air: 0.1 ng/m3 (EPA (1989)).
Dietary intake: 5-6 |xg lead/day (EPA (1994a); EPA (1994b); EPA (1986)). This estimate of
dietary intake already includes a measure of fish consumption.
(B.) Calculate blood lead contributions by multiplying media-specific concentrations
or intakes by media-specific blood lead coefficients for children, derived from
EPA Air Quality Criteria Document (EPA (1986), EPA OAQPS Staff Paper (EPA
1989), and preamble to lead drinking water regulation (EPA (1991a)).
drinking water: 4 ppb X 0.16 ng/dl per ppb = 0.64 (xg/dl
soil/dust/paint: 800 ppm X 0.002 |xg/dl per ppm -1.6 fxg/dl
air: 0.1 |ag/m3 X 2 jxg/dl per p.g/m3 = 0.2 jxg/dl
diet: 5.5 |ig/day X 0.16 jag/dl per |xg/day = 0.9 |ig/dl
(C.) Sum the above blood lead concentration contributions to determine average
current human exposure from all routes:
0.64 ^g/dl + 1.6 ng/dl + 0.2 jj.g/dl + 0.9 ng/dl = 3.3 |ag/dl
II. Calculate the acceptable contribution of lead from fish consumption to blood lead levels.
Level of Concern - Current exposures = Acceptable fish contributions
4.6ug/dl - 3.3 ng/dl = 1.3 M-g/dl
III. Convert acceptable lead contributions from fish to lead fish tissue concentrations.
(A.) Calculate acceptable daily intake from fish:
Divide 1.3 jxg/dl by the blood lead to food lead coefficient (0.16 ng/dl per |ig/day)
154
-------
Appendix H
Equals: 8.125 |ig lead/day from fish
(B.) Calculate acceptable tissue concentration in fish:
Using the EPA standard average consumption rate of fish of 7.2 grams/day:
1. 8.125 (ig lead/day divided by 7.2 grams fish/day = 1.1285 jig/gram of fish
or 1.1285 ppm
IV. Determine acceptable level in benthic test tissue:
Using an assumption that fish consumed spend 77.7% of their time foraging at the HARS and
an estimate of trophic transfer efficiency of lead to fish from benthic prey of 23%, the
acceptable concentration in fish (1.13 ppm) corresponds to an acceptable tissue concentration
in benthic prey organisms of 6.32 ppm (i.e., 1.13 ppm/(0.777)(0.23)).
In this analysis, it was conservatively assumed that:
• 100% of daily fish intake of 7.2 grams per day was from the Mud Dump Site;
• fish spent 77.7% of their time foraging at the Mud Dump Site;
• 23% of lead from lower trophic levels would be absorbed into fish tissue;
• lead in fish was uniformly distributed between tissues; and
• lead was in addition to the normal assumed dietary intake of lead.
155
-------
Appendix H
HARS-Specific Value Calculations for Non-Cancer Effects from Lead
1 - Calculate current human exposure from all routed
A - Average current exposures
Drinking water 4 ppb
Soil/dust/paint 800 ppm
Air 0.1 ng/cuM
Dietary 5.5 ng/da
B - Media-specific blood lead coefficients
Drinking water 0.16 ng/dl per ppb
Soil/dust/paint 0.002 ng/dl per ppm
Air 2 |ig/dl per ug/cuM
Dietary 0.16 |ig/dl per ug/da
Calculate Blood lead contributions [A x B]
Drinking water 0.64 jxg/dl
Soil/dust/paint 1.6 |ig/dl
Air 0.2 ng/dl
Dietary 0.9 (ig/dl
C - Sum of blood lead contributions to determine all-route current exposure [Sum(A x B)]
Current exposure 3.3 fig/dl
2 - Calculate acceptable lead contribution from fish spending time at MDS [Level of concern - current
exposure]
Level of concern 4.6 ng/dl
Current exposure 3.3 ng/dl
Acceptable HARS fish contribution 1.3 ng/dl
• Convert acceptable fish contribution to lead fish tissue concentration
A - Calculate acceptable daily intake [Acceptable fish contribution / food lead coefficient]
Acceptable fish contribution 1.3 ng/dl
Food lead coefficient 0.16 (-ig/dl per da
Acceptable daily intake 8.125 |ig/da
B - Calculate acceptable tissue concentration in fish [Acceptable intake/Average fish consumption]
Average fish consumption 7.2 g/da
Acceptable tissue cone 1.1285 (ig/g
C - Calculate acceptable benthic tissue concentration from acceptable concentration in fish
Whole-body: fillet 1
After adjustment for fillet 1.1285 ng/g
Site use percent 77.7 percent
After adjustment for site use 1.4523 |ig/g
Trophic transfer factor 0.23
After adjustment for trophic 6.3 |ig/g
transfer
or
HARS-SPECIFIC VALUE 6.3 ppm
156
-------
Appendix H
References Cited:
Brody, D.J.; Pirkle, J.L.; Kramer, R.A.; Flegal, K.M.; Matte, T.D.; Gunter, E. W.; Paschal, D.C.; Blood
lead levels in the US population: Phase 1 of the Third National Health and Nutrition Examination Survey
(NHANES III, 1988 to 1991). Journal of the American Medical Association, 1994, pages 272:277-283.
EPA (1986). Air Quality Criteria for Lead. Research Triangle Park, NC: Office of Research and
Development, EPA 600/8-83-028F.
EPA (1989). OAQPS Staff Report: Review of the National Ambient Air Quality Standards for Lead:
Exposure Analysis Methodology and Validation. 1989.
EPA (1991a). Lead and Copper NPDWRs/56 FR26468, June 7, 1991.
EPA (1991 b). Lead Strategy Document; CDC, 1991.
EPA (1994a). Guidance Manual for the Integrated Exposure Uptake Biokinetic Model for Lead in
Children. Washington, DC. Office of Solid Waste and Emergency Response. EPA/540/R-94/039.
EPA (1994b). Technical Support Document: Parameters and Equations used in Integrated Exposure
Uptake Biokinetic Model for Lead in Children. Office of Solid Waste and Emergency Response.
Washington, DC. EPA/540/R-94/040.
EPA (1994). OSWER Soil Lead Directive.
EPA (1996). Urban Soil Lead Abatement Demonstration Project, EPA Integrated Report, Volume I, US
EPA 1996, 600/P93/001AF.
157
-------
Appendix H
This page intentionally left blank.
158
-------
APPENDIX I
Identification of Target Population and Estimation of
Seafood Consumption Rate
-------
Appendix I
Identification of Target Population and, Estimation of Seafood Consumption Rate
A study of fish consumption published in 1994 by the New Jersey Department of Agriculture, New Jersey
Marine Sciences Consortium and the Eagleton Institute for Public Polling of Rutgers University (NJMSC,
1994) was reviewed by EPA Region 2 to identify an appropriate target subpopulation of seafood
consumers for assessing human health risk/hazard associated with contaminants accumulated by benthic
organisms at the HARS. Data presented in this study suggested that, overall, recreational fishermen
consumed approximately 62 grams of finfish/day (based on an averaging time of one year). EPA Region
2 selected the average consumption of recreationally caught fish self-reported by recreational fisherman
as a reasonable estimate of fish consumption by an RME (reasonably-maximally exposed) population.
Use of this value, 7.2 g/day, for assessing risk/hazard assumes that all of the recreational fish in the RME
individual's diet is comprised of fish obtained from recreational fishing at the HARS.
Due to its offshore location and relative inaccessibility to subsistence anglers, EPA Region 2 assumed
that a specific subsistence fishery does not exist at the HARS. EPA Region 2 instead proposes to use
recreational fishermen as an appropriate subpopulation for estimating seafood consumption rates that
reflect reasonably maximum exposure (RME) of humans to contaminants at the HARS.
EPA Region 2 contacted the New Jersey Department of Health, New Jersey Department of
Environmental Protection and the New Jersey Sea Grant College Program to obtain copies of relevant
studies that assess fish consumption in New Jersey in order to identify an appropriate ingestion rate for
use in the evaluation process. Two pertinent studies were received and reviewed. These studies were:
• Belton et al. 1985. A Study of Toxic Hazards to Urban Recreational Fishermen and Crabbers.
New Jersey Department of Environmental Protection. Office of Science and Research.
• New Jersey Marine Sciences Consortium. 1994. Fish Consumption Patterns by New Jersey
Consumers and Anglers. Prepared for the NJDEPE- Division of Science and Research. Contract
No. P 30695 00962.
The Belton et al. (1985) study focused on urban anglers active on the Hudson River, Upper Bay, and
Newark Bay shorelines and did not attempt to quantify actual consumption rates of the anglers. The
study, therefore, was judged to be inappropriate and of little utility in estimating human consumption
rates of fish from the HARS area.
Despite certain limitations that were identified in the report, the second study (NJMSC, 1994) provided
useful data for estimating an appropriate ingestion rate for the target group of consumers (i.e., anglers).
One of the primary limitations of the study was that calculated fish ingestion rates and patterns were
based on consumption in the week preceding the survey (October-November). This could potentially
result in the introduction of a biased estimate that may not truly represent annual consumption behavior.
Calculation of Overall Seafood Consumption of NJ Recreational Anglers
New Jersey consumers were classified into two groups based on whether or not those respondents had
consumed fish in the week preceding the survey. Overall seafood consumption by anglers that reported
having consumed fish in the preceding week were somewhat higher than those that did not report having
consumed fish in the preceding week. Average consumption of those anglers consuming fish in the
preceding week were used to obtain a more conservative estimate of consumption by New Jersey anglers.
This group of anglers reported consuming an average of 2.42 meals and 15.23 ounces of fish per week.
159
-------
Appendix I
This level of consumption (i.e., 15.23 oz/wk) is equivalent to an average daily seafood consumption rate
of approximately 61 grams, confirming that recreational anglers indeed consume significantly more fish
than the national average (i.e., 6.5 g/day) and, therefore, appear to be an appropriate (i.e., RME) target
population for use in evaluating potential for human health risks due to consumption of fish from the area
of the HARS.
Estimate of Recreationallv Caught Fish Consumed bvNJ Recreational Anglers (Based on Self-Reporting')
Anglers responding to the NJMSC (1994) survey indicated that they annually consume an average of 5.76
lbs. of recreationally caught fish. Information on angling habits in New Jersey marine waters is not
available. Therefore, EPA Region 2 conservatively assumed that there may be a subpopulation of
recreational anglers that preferentially fishes at the HARS and obtains all of the recreationally caught fish
in their diet from fishing at the HARS. This equates to an average daily consumption of 7.2 grams of
recreationally caught fish. Given that such a population exists, their consumption of finfish that are
potentially exposed to the HARS could be estimated at 7.2 g/day.
Estimate of Recreationallv Caught Fish Consumed bvNJ Recreational Anglers (Based on Overall
Consumption)
As part of the NJMSC survey, respondents identified the quantities and types/species of fish that they
consumed (see page 4-69, Table 4-14 of NJMSC, 1994). Of the 15.23 ounces consumed weekly by
recreational anglers, 7.57 ounces were reported to be saltwater finfish. Approximately 37% (2.8 ounces)
of the saltwater finfish consumed by recreational anglers were reported to be fresh (i.e., not canned or
processed) fish of varieties that are listed in the HARS SEIS (EPA 1997) as species that may occur in the
vicinity of the HARS. In decreasing contribution to angler diet, the reported fish species include:
flounder/fluke, cod, sea bass, haddock, whiting, blackfish, porgy, bluefish, striped bass1, and weakfish.
The weekly consumption rate of 2.8 ounces, or 79.38 grams, equates to a daily consumption rate of 11.34
grams of finfish that could potentially occur at the HARS. This consumption rate, therefore, does not
include consumption of processed fish or of species that are not expected to occur at the HARS, such as
red snapper, orange roughly, and off-shore species (e.g., tuna, swordfish). NJMSC (1994) angler
respondents indicated that 60% of the fish they consume is prepared in the home. EPA Region 2 assumed
that recreationally caught fish is consumed by recreational anglers in their home. If the percentage of fish
that occurs at the HARS that is consumed by recreational anglers in their home is similar to the overall
percentage of home-prepared fish (i.e., 60%) in their diet, then the daily consumption rate of
recreationally-caught (i.e., home-prepared) fish potentially occurring at the HARS by New Jersey anglers
would be estimated to be 6.8 grams (i.e., 11.34 g/day x 60 percent).
Selected Estimate of Consumption of Recreationallv-Caught Fish for Risk Assessment
The two estimates of consumption of finfish that are potentially exposed to HARS and caught
recreationally by New Jersey anglers (outlined above) agree well. EPA Region 2 proposes to use the
higher of the two estimates (i.e., 12 grams/day) as an appropriate estimate of fish consumption for
assessing the risks to a reasonably maximally exposed (RME) human subpopulation associated with
contaminants in sediments proposed for use as Remediation Material at the HARS.
Approximately 75 percent of anglers in the survey reported consuming less than three meals of seafood
per week. More significantly, all of the recreationally caught fish consumed by anglers is assumed to be
'The contribution of striped bass to anglers' diets was not reported in NJMSC (1994). Its contribution to
anglers' diets was assumed to be equal to that of bluefish (i.e., 4 percent).
160
-------
Appendix I
composed of species occurring at, and obtained from, the HARS. Certain species that contributed to the
consumption rate and that were assumed to be recreationally caught at the HARS are not generally
targeted at the HARS. Examples include: structure-associated species, such as porgy, blackfish, and sea
bass; deeper water bottom species, such as cod and haddock; and species generally targeted within bays
and estuaries, such as winter flounder. However, EPA Region 2 also recognizes that the contribution of
lobster to the human diet (estimated at 3.2 g/day in NJMSC (1994)) is not reflected in the estimated
consumption rate. There is, however, no directed recreational lobster fishery at the HARS and therefore
the assumption that all (or 60%) of consumed lobster is obtained at the HARS cannot be supported.
Eighty-five to 90 percent of the survey participants with consumption rates in this range {i.e., two to three
fish meals/week) also reported that this rate was typical or slightly more than their usual consumption.
This suggests that the potential bias associated with using a single week's consumption (identified above)
to extrapolate annual fish consumption may not be significant, using the NJMSC (1994) study.
References Cited:
Belton et al. 1985. A Study of Toxic Hazards to Urban Recreational Fishermen and Crabbers. New
Jersey Department of Environmental Protection. Office of Science and Research.
EPA (U.S. Environmental Protection Agency, Region 2). 1997. Supplement to the Environmental Impact
Statement on the New York Dredged Material Disposal Site Designation for the Designation of the
Historic Area Remediation Site (HARS) in the New York Bight Apex.
NJMSC (New Jersey Marine Sciences Consortium). 1994. Fish Consumption Patterns by New Jersey
Consumers and Anglers. Prepared for the NJDEPE- Division of Science and Research. Contract
No. P 30695 00962.
161
-------
Appendix I
This page intentionally left blank.
162
-------
APPENDIX J
Consideration of HARS Site Use by Finfish
-------
Appendix J
Consideration of HARS Site Use by Finfish
This factor is intended to express the proportion of time that fish predators may be exposed to
contaminated benthic prey residing at the HARS. EPA Region 2 derived an estimate of site use for a
"generic" fish in the diet of the target sub-population (i.e., New Jersey recreational fishers). This estimate
of site use was derived primarily by examining quarterly commercial landings from the area of the New
York Bight encompassing the HARS for each species of interest and expressing the potential site use for
the species as the number of quarters required to encompass 95% of annual landings of the species (In this
way, site use for each species was expressed as a discrete quartile variable). Site use for a "generic"
HARS-exposed fish was then estimated as an overall average of the quartiles of the species weighted by
the relative contribution of each species to the diet of the target subpopulation. This approach resulted in
a weighted average site use for "fish" of 77.7%.
EPA Region 2 used data on the seasonal presence of fish species in New York Bight waters to derive a
site use exposure factor to reflect the potential foraging at the HARS of fish species that are consumed by
recreational fishermen (Consumed species are listed in Appendix H). Derivation of the proposed site use
factor is described below.
EPA Region 2 reviewed 1993 commercial catch data reported by NOAA/NMFS, along with species
summaries from Bigelow and Schroeder (1953), and Smith (1982) to establish the presence or absence of
species in the New York Bight on a quarterly (seasonal) basis throughout the year. The NOAA/NMFS
data are summarized in Appendix A of the HARS SEIS, (EPA 1997). With the exception of cod, EPA
Region 2 found that 95 percent (by weight) or greater of all fish caught were restricted to three or less of
the 4 quarters for each year. Catches of cod in the New York Bight were distributed throughout the year.
No seasonal data were available for haddock; therefore, its seasonal presence was assumed to be similar
to that of cod.
To derive the site use exposure factor, the seasonal presence of species (i.e., minimum percent of year (as
quartiles required to account for 95 percent of species landings) was weighted by the relative contribution
of that species to the total estimated consumption of fish by recreational fishermen. A single weighted
average was obtained that reflects the seasonal presence (and potential exposure) of consumed fish at the
HARS, considered collectively. The calculation of this seasonal fish foraging exposure is summarized in
Tables J-l, J-2, and J-3, below. The weighted seasonal residence factor (i.e., Site Use Factor) for fish in
New York Bight waters (i.e., in the vicinity of the HARS) is estimated to be 77.7 percent of the year.
Where the duration of exposure of fish is important to the calculation of potential for risk (e.g., human
health-based values), EPA Region 2 proposes to incorporate a site use exposure factor of 0.777 (or its
reciprocal (i.e., 1.29), as appropriate) to account for this seasonality.
163
-------
Appendix J
Table J-l. 1993 Commercial Catch per Quarter (Metric Tons)
Fish Name
Q1 (Jan-Mar)
Q2 (Apr-Jun)
Q3 (Jul-Sep)
Q4 (Oct-Dec)
Total Annual
Winter Flounder
8.9
61.1
14.9
42.6
127.5
Summer Flounder
3.8
94.3
240.4
43.4
381.9
Yellowtail Flounder
M
8,5
OJ.
0J.
18.2
Flounders (Totals)
22.2
163.9
255.4
86.1
527.6
Cod
8.7
2.7
2.9
4.5
18.8
Whiting
24.3
50.4
7.1
275.2
357.0
Bluefish
0
202
101.8
56.1
359.9
Porgy
0
4.6
1.8
36
42.4
Blackfish
0.5
13.3
23.3
15.5
52.6
Weakfish
0
7.2
12.5
11.6
31.3
Striped Bass - Present primarily in Spring Fall (Smith 1982)
Sea Bass - Present primarily in Spring Fall (Bigelow and Schroeder, 1953)
Haddock - Assumed to be similar to Cod
Commercial catch data reported by NOAA/NMFS 1993
Table J-2. 1993 Commercial Catch per Quarter (Percent of Total Annual)
Fish Name
Q1 (Jan-Mar)
Q2 (Apr-Jun)
Q3 (Jul-
Sep)
Q4 (Oct-
Dec)
Minimum
quarters to
explain
95% of
presence
Percent of
catch
Flounders (Totals)
4.2
31.1
48.4
16.3
3
95.8
Cod
46.3
14.4
15.4
23.9
4
100.0
Whiting
6.8
14.1
2.0
77.1
3
98.0
Bluefish
0.0
56.1
28.3
15.6
3
100.0
Porgy
0.0
10.8
4.2
84.9
2
95.8
Blackfish
1.0
25.3
44.3
29.5
3
99.0
Weakfish
0.0
23.0
39.9
37.1
3
100.0
Commercial catch data reported by NOAA/NMFS 1993
164
-------
Appendix J
Table J- 3. Seasonal Residence Weighted by Contribution to Fish Consumption
Fish Name
Contribution to HARS diet
(%)
Seasonal residence at
HARS
(%)
Seasonal Residence
Weighted by
Contribution to HARS
diet (%)
Flounders (all spp.)
48.6
75
36.49
Cod
10.8
100
10.81
Whiting
2.7
75
2.03
Bluefish
10.8
75
8.11
Striped Bass
10.8
75
8.11
Haddock
2.7
100
2.70
Porgy
2.7
50
1.35
Blackfish
2.7
75
2.03
Weakfish
2.7
75
2.03
Sea Bass
5.4
75
4.05
99.9
Site Use Factor (%> Year Present at the HARS)
77.70
165
-------
Appendix J
References Cited:
Bigelow, H.B. and W.G. Schroeder. 1953. Fishes of the Gulf of Maine. First Revision. Fishery Bulletin of
the Fish and Wildlife Service, Volume 53. U.S. Government Printing Office, Washington, DC. 577 p.
EPA (U.S. Environmental Protection Agency, Region 2). 1997. Supplement to the Environmental Impact
Statement on the New York Dredged Material Disposal Site Designation for the Designation of the
Historic Area Remediation Site (HARS) in the New York Bight Apex.
Smith, W.G. 1982. Striped Bass. Pages 79-82jn: M.D. Grosslein and T.R. Azarovitz, Fish Distribution.
MESA New York Bight Atlas Monograph 15. New York Sea Grant Institute, Albany, NY. 181 p.
166
-------
APPENDIX K
Whole Body to Fillet Correction Factors
-------
Appendix K
Whole Body to Filfet Correction Factors
EPA Region 2 proposes to employ a 1.35 whole-body to fillet ratio for organic, lipophilic compounds
(including PAHs, pesticides, PCBs and butyl tins). This whole-body to fillet value is based on fish tissue
data collected from New York State and the Great Lakes for lipophilic chlorinated organic substances,
such as PCBs and DDT.
Only limited information was identified in the literature regarding the whole-body to fillet ratios for
inorganic compounds. Bevelhimer et al. (1997) investigated the relationship between fillet and whole-
body concentrations of inorganic contaminants in several finfish species and developed ratios for specific
inorganic chemicals . Clearly defined, statistically significant relationships between fillet and whole-body
concentrations were reported for arsenic, chromium, and mercury. For other inorganic contaminants
(i.e., Cd, Cu, Pb, Ni, and Zn) the data suggested that whole-body and fillet concentrations differed
significantly, however, whole body concentrations could not consistently be predicted from
concentrations in the fillet.
The slopes of the lines relating the whole body and fillet residues for arsenic, chromium, and mercury
reported by Bevelheimer et al. (1997) are proposed to be used in the assessment of risks to human
consumers associated with metals accumulated by benthic organisms from sediments proposed for
placement at the HARS. These slopes are listed in Table K-l below. All other metals (Ag, Cd, Cu, Ni,
Pb, Zn) will be considered assuming equivalent concentrations in the fillet and whole body of the fish in
the risk assessment process.
Table K-l. Metals Whole Body to Fillet Ratios
Chemical
RatlO (Cwh0|e/Cfiiiet)
Arsenic
1.4
Chromium
1.2
Mercury
0.7
References Cited:
Bevelhimer, M.S., J.J. Beauchamp, B.E. Sample, and G.R. Southworth. 1997. Estimation of Whole-Fish
Contaminant Concentrations From Fish Fillet Data. Prepared by the Risk Assessment Program, Oak
Ridge National Laboratory, Oak Ridge, TN for the U.S. Department of Energy. ES/ER/TM-202.
167
-------
Appendix K
This page intentionally left blank.
168
-------
APPENDIX L
Trophic Transfer Factors
-------
Appendix L
Trophic Transfer
Trophic transfer of contaminants from benthic prey to fish predators was estimated by applying a discrete
factor that expresses the ratio of the residue concentration in predator as a function of the residue
concentration in prey. Trophic transfer coefficients for chlorinated organic contaminants were assigned
using the food web model of Gobas (1993), as run by EPA (1995a). Trophic transfer coefficients for
most other compounds (PAHs, and six metals (arsenic, cadmium, copper, lead, mercury, and zinc) were
selected from literature values. A trophic transfer coefficient of 1 was applied in the calculation of risks
associated with chromium, nickel, and silver.
PAHs
EPA Region 2 proposes to apply a trophic transfer coefficient of 0.1 to residues of PAHs (alkylated and
parent) accumulated by bioaccumulation test organisms for the purpose of assessing risk/hazard of PAHs
to humans consuming finfish exposed to contaminated benthic prey. Selection of this trophic transfer
coefficient for PAHs was based on the results of two studies that documented inefficient trophic transfer
of PAHs. Burns and Teal (1979) estimated a trophic transfer of 0.1 for total PAHs between mummichog
(Fundulus heteroclitus) and American eel (Anguilla rostrata). Broman et al. (1990) reported a trophic
transfer factor of 0.1 for benzo(a)pyrene between zooplankton and mussels (Mytilus edulis).
Chlorinated Organics
Trophic transfer coefficients proposed for use in assessing risks/hazards of chlorinated organic
contaminants at the HARS were derived using the food web model of Gobas et al. (1993), run at
equilibrium (EPA 1995b). The food web used in the model was a simplified New York Bight food chain
consisting of three representative trophic levels: benthic organisms, benthic predators, and upper level
predators (for additional information on the trophic levels see EPA (1995b)). Average lipid contents of
each trophic level were calculated from lipid contents of representative organisms in each level, as
reported by NYSDEC (1996).
Karickhoff and Long (1995) reviewed log Kows and derivation methods published for various compounds
and the methods used to derive those values. Log Kows obtained using the "slow-stirring" or "shake-flask"
methods were usually recommended for use by the authors, depending on the compound. These
recommended log Kow values were used to run the Gobas et al. (1993) model.
Table L-l. Log K^s assigned to chlorinated organic pesticides and resultant trophic transfer
coefficients predicted by the Gobas et al. (1993) model are reported in the following
table
Compound
log KoW
Trophic Transfer
Factor
Compound
log Kow
Trophic Transfer
Factor
aldrin
6.5
3.0
Heptachlor
epoxide
5.0
1.4
dieldrin
5.3
1.6
DDD
6.1
2.7
a-chlordane
6.32
2.9
DDE
6.76
3
trans nonachlor
6.87
2.6
DDT
6.5
3
heptachlor
6.26
2.9
Endosulfans
4.1
1.1
169
-------
Appendix L
In deriving guidance values for evaluating risk/hazard associated with organochlorine mixtures (e.g., total
DDT, total PCB), a single trophic transfer coefficient was used to estimate the potential for trophic
transfer of the mixture to upper level predators. The trophic transfer coefficient used in generating
reflects the most efficiently transferred of the compounds contributing to the mixture, based on its log
KoW. In order to better predict and consider the trophic transfer of these organochlorine mixtures in
assessing risks associated with bioaccumulation from dredged material, EPA Region 2 proposes to
calculate an adjusted total residue of the mixture by applying a multiplier to residues of contributing
compounds that will correct for the difference in the actual trophic transfer predicted by the Gobas model
for the individual compound (based on its log KoW) and that used for deriving the HARS-Specific value
for assessing the risks associated with the entire mixture. Figure L-l illustrates how this correction would
be made for a mixture of DDD, DDE, and DDT compounds.
Figure L-l. Calculation of an adjusted total DDT residue level (based on differences in log KoWs of
contributing compounds) for comparison to an HARS-Specific Value that was derived to assess
risks of accumulated DDT compounds. (Benthic Organisms to Upper Level Predators)
Trophic Transfer Factor
Compound
Log Kow
Predicted By Gobas
Assigned in HARS-
Correction
Model
Specific Value Dev't.
Factor
DDD
6.1
2.7
3
0.9
DDE
6.8
3
3
1
DDT
6.5
3
3
1
Therefore, the adjusted total DDT residue used for comparison to HARS-Specific Value for total DDT
would be calculated as ([DDD]r*0.9) + ([DDE]r* 1.0) + ([DDT]r* 1.0). Where the correction factor is
calculated as predicted/assigned trophic transfer factor, and [x]r is the reported test organism tissue
concentration after adjustment for steady state.
EPA Region 2 proposes to use the log Kows reported for individual PCB congeners (Hawker and Connell,
1988) to similarly predict and consider the trophic transfer of the specific PCB mixtures that are
accumulated by test organisms from dredged material EPA Region 2 proposes to adjust total PCB
residues in test organism tissues in a similar manner using log Kows reported by Hawker and Connell
(1988) for individual PCB congeners prior to comparison to HARS-Specific Values (The HARS-Specific
Values for total PCBs assume a trophic transfer coefficient of 3. This trophic transfer coefficient reflects
the most efficiently transferred of the PCB congeners). EPA Region 2 believes that this approach will
minimize the uncertainty and better estimate trophic transfer of specific mixtures of organochlorine
contaminants accumulated by test organisms.
Mercury:
Trophic transfer coefficients for estimating mercury transfer from prey to fish predators was reported in
the EPA Mercury Study Report to Congress (EPA 1997). In this report, a trophic transfer coefficient of
6.2 was recommended for planktivorous fish; and a trophic transfer coefficient of 5.0 was recommended
for piscivorous fish. EPA Region 2 is proposing to employ the higher of these trophic transfer
coefficients (i.e., 6.2) to express the potential trophic transfer of mercury from benthic prey to finfish in
assessing risk/hazard of total mercury accumulated from sediments proposed for placement at the HARS.
170
-------
Appendix L
Metals:
Based on its review of available scientific literature on dietary transfer of metals to finfish from
contaminated benthic invertebrate prey, EPA Region 2 has developed and is proposing trophic transfer
coefficients for four cationic metals (i.e., copper, cadmium, lead and zinc) and arsenic.
EPA Region 2 focused on recent laboratory studies conducted with field-collected benthic organisms
from areas known to be contaminated by metals (Woodward et al., 1994; 1995; Farag et al., 1994; 1999;
2000) in order to derive trophic transfer coefficients for metals. The trophic transfer factors (i.e., benthic
invertebrates to fish) presented in Table L-2 were derived by EPA Region 2 and are proposed for use in
characterizing the potential for trophic transfer of these metals to fish from contaminated benthic prey at
the HARS.
Table L-2. Proposed Trophic Transfer Factors
Chemical
Trophic Transfer: Prey to
Fish
Trophic Transfer: Fish to
Prey
Arsenic
0.25 ([As]fiSh / [As]prey)
4.00 ([As]prey / [As]fish)
Cadmium
0.25 ([Cd]fish/[Cd]prey)
4.00 ([Cd]prey / [Cd]flsh)
Copper
0.21 ([Cu]fjsh / [Cu]prey)
4.76 ([Cu]prey / [Cu]fish)
Lead
0.23 ([Pb]fish/[Pb]prey)
4.35([Pb]prey/[Pb]fish)
Zinc
0.24 ([Zn]fIsh / [Zn]prey)
4.17 ([Zn]prey / [Zn]flsh)
EPA Region 2 is proposing to use a trophic transfer coefficient of 1 for assessing human health risk of the
other cationic metals (i.e., silver, chromium, nickel) that are measured in bioaccumulation test organism
tissue.
Background and State of the Science
Much of the early work that examined the importance of the dietary pathway for transfer of metals to fish
in metals-contaminated systems focused on the relative concentrations of metals in fish and prey collected
from within the same system or area (e.g., Metayer et al. 1980, Dallinger and Kautzky 1985, and
references in Table L-3). These field studies suggested that dietary transfer of metals from prey could
significantly contribute to fish body burdens of metals and in certain exposure situations could be of
greater importance than absorption from the water. Results of certain field studies, however, were
inconclusive (see Metayer et al. 1980, and Dallinger and Kautzky 1985). Field-derived ratios suggested
that the transfer of metals to fish from prey, was generally inefficient (i.e., ratios of consumer/prey
concentrations tended to be less than one). Suedel et al. (1994) reviewed available data in efforts to
assess the potential for trophic transfer of metals in aquatic foodwebs. The data that was considered by
Suedel et al. is summarized in Table L-3. Based on this data they concluded that trophic transfer of
metals to fish should not generally be described as biomagnification and noted that "concentrations of
most metals were often higher in tissues of producers and primary consumers...than carnivorous fish".
Trophic transfer ratios of less than 1 are consistent with the findings of Reinfelder et al. (1998). Using
the kinetic model approach, they concluded that trophic transfer of cationic metals (however, only
cadmium was specifically modeled in that paper) to fish is expected to be less than one.
171
-------
Appendix L
Since publication of Suedel et al. (1994), the potential for trophic transfer of metals has been increasingly
investigated in the laboratory under controlled conditions. Laboratory studies have been conducted with
commercial feeds or live prey (e.g., Artemia) that were contaminated with known amounts of metals in
the laboratory and fed to fish (e.g., Handy et al. 1992; Cockell and Hilton 1988; Hatakeyama and Yasuno
1982; Kumada et al. 1973; Mount et al. 1994;) or with benthic organisms that were collected from the
field from areas known to be contaminated by metals and fed to fish (Woodward et al. 1994,1995; Farag
et al. 1994, 1999). Results of these studies are summarized in Table L-4.
In laboratory studies, body burdens of copper in fish ranged from 0 to 20.6% (n = 28, ave. 4.6%, median
2.1%) of the dietary copper concentration to which the fish were exposed. Body burdens of cadmium in
fish ranged from 0 to 25%2 (n = 31, ave. 6.6%, median 5.0%) of the dietary cadmium concentration to
which the fish were exposed. Body burdens of lead in fish ranged from 0 to 22.7%3 (n = 19, ave. 5.5%,
median 3.9%) of the dietary lead concentration to which the fish were exposed. Body burdens of zinc in
fish ranged from 0 to 89.9% (n = 16, ave. 18.4%, median 11.4%) of the dietary zinc concentration to
which the fish were exposed. Body burdens of arsenic in fish ranged from 0 to 29.7% (n = 37, ave. 8.6%,
median 6.1%) of the dietary arsenic concentration to which the fish were exposed.
Farag et al. (2000) showed that the degree of association of the metal with organic compounds (proteins)
within prey significantly effects the efficiency of trophic transfer of metals to fish (i.e., increased covalent
bonding and complexation of metals enhances the bioavailability of metals to fish consumers). Harrison
and Curtis (1992) demonstrated that uptake of cadmium is higher from natural diets raised in
contaminated environments than from Cd-contaminated commercial feeds. Farag et al. (2000) also
demonstrated that metals in laboratory-dosed and field-collected invertebrates are processed differently
by fish consumers during digestion and metals in laboratory-dosed prey are less available to fish.
Therefore, results of studies that are conducted using contaminated feeds or laboratory-contaminated prey
differ significantly from those of studies using natural prey and may underestimate the potential for
trophic transfer of metals to fish.
In light of the above considerations, studies such as those conducted by Farag et al. (1994, 1999) and
Woodward et al. (1994,1995), that used field-collected and contaminated prey were deemed to be the
most relevant and appropriate studies for use in deriving a conservative estimate of trophic transfer
potential of metals to fish from benthic invertebrate prey. Table L-5 lists results of those studies in which
field-collected contaminated prey was used to estimate trophic transfer.
The maximum trophic transfer values reported for these compounds can be used as appropriately
conservative estimates of the potential for trophic transfer of metals to fish from benthic invertebrates
exposed to dredged material for use in interpreting the results of 28 day laboratory bioaccumulation tests.
2Excludes outlier of 156%, initial/control concentrations were not reported by authors
3Excludes outlier of 375%, initial/control concentrations not reported; diet concentrations questionable
172
-------
Appendix L
These factors are:
Copper: 0.21 ([Cu]fjSh/[Cu]prey)
Cadmium: 0.25 ([Cd]fiSh/[Cd]prey)
Lead: 0.23 ([Pb]fish/[Pb]prey)
Zinc: 0.24 ([Zn]f1Sh/[Zn]prey)
Arsenic: 0.25 ([As]fish/[As]prey)
4.76 ([Cu]prey/[Cu]fjSh)
4.00 ([Cd]prey/[Cd]fish)
4.35 ([Pb]prey/[Pb]fish)
4.17 ([Zn]prey/[Zn]flsh)
4.00 ([As]prey/[As]fish>
While these factors are less conservative than the factor of one that is used in the current evaluation
process, they still reflect the results of these studies in a conservative manner. Median trophic transfer
factors reported in these studies for copper, cadmium, lead and arsenic ranged from 0.06 to 0.12 (median
factor for zinc was 0.21). Because it is impossible to assess how the availability of metals in the prey
species used in these studies (e.g., caddisfly and stonefly larvae) relates to availability in dominant prey
species at the HARS (i.e., polychaetes and amphipods) and how uptake by trout may differ from fish
species at the HARS, EPA Region 2 believes that the conservative interpretation of this dataset is
reasonable.
173
-------
Appendix L
References Cited:
Broman, D., C. Naf, I. Lundbergh, and Y. Zebuhr. 1990. An in situ study on the distribution,
biotransformation and flux of polycyclic aromatic hydrocarbons (PAHs) in an aquatic food chain (Seston-
Mytilus edulis L.- Somateria Mollissima L.) from the Baltic: an ecotoxicological perspective. Environ.
Toxicol. Chem. 9:429-442.
Burns, K.A. and J.M. Teal. 1979. The West Falmouth oil spill: hydrocarbons in the salt marsh ecosystem.
Estuar. Coast. Mar. Sci 8:349-360.
Cockell, K.A. and J.W. Hilton. 1988. Preliminary investigations on the comparative chronic toxicity of
four dietary arsenical to juvenile rainbow trout (Salmo gairdneri R.). Aquat. Toxicol. 12:73-82.
Dallinger, R. and H. Kautzky. 1985. The importance of contaminated food for the uptake of heavy metals
by rainbow trout (Salmo gairdneri): a field study. Oecologia (Berlin) 67:82-89
EPA (U.S. Environmental Protection Agency). 1995a. Memo from L. Burkhard. Biomagnification
Factors for the NY Bight Apex. 9 pgs.
EPA (U.S. Environmental Protection Agency). 1995b. Memo from A. Lechich. Preliminary Food Chain
Summary Information for New York Bight Food Chain Modeling. 2 pgs.
EPA (U.S. Environmental Protection Agency). 1997. Mercury Study Report to Congress. Volume III:
Fate and Transport of Mercury in the Environment. December 1997. EPA 452/R-97-005. Various
Pagings.
Farag, A.M., C.J. Boese, D.F. Woodward, and H.L. Bergman. 1994. Physiological changes and tissue
metal accumulation in rainbow trout exposed to foodborne and waterborne metals. Env. Tox. Cont.
13(12):2021-2029.
Farag, A.M., M.J. Suedkamp, J.S. Meyer, R. Barrows, and D.F. Woodward. 2000. Distribution of metals
during digestion by cutthroat trout fed benthic invertebrates contaminated in the Clark Fork River,
Montana and the Coeur d'Alene River, Idaho, U.S.A., and fed artificially contaminated Artemia. J. Fish
Biol. 56:173-190.
Farag, A.M., D.F. Woodward, W. Brumbaugh, J.N. Goldstein, E. MacConnell, C. Hogstrand, and F.T.
Barrows. 1999. Dietary effects metals-contaminated invertebrates from the Coeur d'Alene River, Idaho,
on cutthroat trout. Trans. Am. Fish. Soc. 128:578-592.
Gobas, F.A.P.C. 1993. A model for predicting the bioaccumulation of hydrophobic organic chemicals in
aquatic food-webs; application of Lake Ontario. Ecol. Modell. 69:1-17.
Handy, R.D. 1992. The assessment of episodic metal pollution. II. The effects of cadmium and copper
enriched diets on tissue contaminant analysis in rainbow trout (Oncorhynchus mykiss). Arch. Env.
Contam. Toxicol. 22:82-87.
Harrison, S.E. and P.J. Curtis. 1992. Comparative accumulation efficiency of 109cadmium from natural
food (Hyallela azteca) and artificial diet by rainbow trout (Oncorhynchus mykiss). Bull. Env. Cont. Tox.
49:757-764
174
-------
Appendix L
Hatakeyama, S. and M. Yasuno. 1982. Accumulation and effects of cadmium on guppy (Poecilia
reticulata) fed cadmium-dosed cladocera (Moina macrocopa). Bull. Env. Contam. Toxic. 29:159-166.
Hawker, D.W. and D.W. Connell. 1988. Octanol-water partition coefficients of polychlorinated biphenyl
congeners. Environ. Sci. Technol. 22:382-387.
Karickhoff, S.W. and J.M. Long. 1995. Internal Report on Summary of Measured, Calculated, and
Recommended Log Kow Values. Prepared for E. Southerland, EPA Office of Water. Dated April 10. 40
Pgs-
Metayer, C., J.C. Amiard, C. Amiard-Triquet, and J. Marchand. 1980. Etude de transfert de quelques
oilgo-elements dans les chaines trophiques neritique et estuariennes: accumulation biologique chez le
poissons omnivores et super-carnivores. Helg. Meeresunters. 34:179-191.
Mount, D.R., A.K. Barth, T.D. Garrison, K.A. Barten and J.R. Hockett. 1994. Dietary and waterbome
exposure of rainbow trout (Oncorhynchus mykiss) to copper, cadmium, lead and zinc using a live diet.
Env. Toxicol. Chem. 13(12):2031 -2041.
NYSDEC. 1996. Chemicals in Fish, Shellfish and Crustaceans from the New York-New Jersey Harbor
Estuary. PCB, Organochlorine Pesticides, and Mercury. Division of Fish, Wildlife and Marine
Resources. November 1996. 150 pgs.
Reinfelder, J.R., N.S. Fisher, S.N. Luoma, J.W. Nichols, W.-X. Wang. 1998. Trace element trophic
transfer in aquatic organisms: A critique of the kinetic model approach. Sci. Total Env. 219:117-135.
Suedel B.C., J.A. Boraczek, R.K. Peddicord, P.A. Clifford, and T.M. Dillon. 1994. Trophic transfer and
biomagnification potential of contaminants in aquatic ecosystems. Rev. Env. Cont. Toxic. 136:21-89.
Woodward, D.F., A.M. Farag, H.L. Bergman, A.J. DeLonay, E.E. Little, C.E. Smith and F.T. Barrows.
1995. Metals-contaminated benthic invertebrates in the Clark Fork River, Montana: Effects on age-0
brown trout and rainbow trout. Can. J. Fish Aq. Sci. 52:1994-2004.
Woodward, D.F., W.G. Brumbaugh, A.J. DeLonay, E.E. Little, and C.E. Smith. 1994. Effects on rainbow
trout fry of a metals-contaminated diet of benthic invertebrates from the Clark Fork River, Montana.
Trans. Am. Fish. Soc. 123:51-62.
175
-------
Appendix L
Table L-3. Trophic Transfer of Metals to Fish (Suedel et al., 1994)
Metai
Species
Common Name
SW/FW
Field/Lab
TTC
Reference
Arsenic
Hexanchus griseus
Shark
SW
field
20.9
LeBlanc and Jackson (1973)
H. griseus
Shark
sw
field
10
LeBlanc and Jackson (1973)
Hexagrammos spp
Greenling
sw
field
0.3
LeBlanc and Jackson (1973)
Hydrolagus colliei
Ratfish
sw
field
15.2
LeBlanc and Jackson (1973)
Diaphus dumerili
Headlightfish
sw
field
0.1
Leatherland et al. (1973)
Carassius auratus
Goldfish
FW
laboratory
0.2
Maeda etal. 1990
Cadmium
Omnivorous fish
FW
field
1.1
Ward etal. (1986)
D. dumerili
Headlightfish
SW
field
0.1
Leatherland etal. (1973)
Chromium
Carpiodes cyprinus
Quillback
FW
field
0.03
Mathis and Cummings (1973)
M. dolomieu
Smallmouth bass
FW
field
0.5
Mathis and Cummings (1973)
Postlarval fish
SW
laboratory
0.1
Baptist and Lewis (1969)
Fundulus heteroclitus
Mummichog
SW
laboratory
1.6
Baptist and Lewis (1969)
Copper
C. cyprinus
Quillback
FW
field
0.02
Mathis and Cummings (1973)
M. dolomieu
Smallmouth bass
FW
field
0.7
Mathis and Cummings (1973)
Pleuronectes platessa
Plaice
SW
laboratory
0.5
Saward etal. (1975)
Lead
Etheostoma flabellare
fantail darter
FW
field
0.3
Enk and Mathis (1977)
M. dolomieu
Smallmouth bass
FW
field
0.9
Enk and Mathis (1977)
C. cyprinus
Quillback
FW
field
0.1
Mathis and Cummings (1973)
M. dolomieu
Smallmouth bass
FW
field
0.9
Mathis and Cummings (1973)
Omnivorous fish
SW
field
2.6
Ward et al. (1986)
Helotes sexlineatus
Trumpeter
SW
field
0.4
Ward et al. (1986)
Platichthyes flesus
Flounder
SW
field
0.7
Hardisty etal. (1974)
Nickel
C. cyprinus
Quillback
FW
field
0.03
Mathis and Cummings (1973) I
M. dolomieu
Smallmouth bass
FW
field
0.7
Mathis and Cummings (1973) I
O. mykiss
rainbow trout
FW
field
0.01
Dallinger and Kautzky (1985)
M. dolomieu
Smallmouth bass
FW
field
1.6
Wren ef al. (1983)
S.namaycush
lake trout
FW
field
1
Wren ef al. (1983)
Zinc
C. cyprinus
Quillback
FW
field
0.06
Mathis and Cummings (1973)
M. dolomieu
Smallmouth bass
FW
field
1
Mathis and Cummings (1973)
Gobius spp
Omnivorous fish
SW
field
0.1
Ward et al. (1986)
H. sexlineatus
Trumpeter
SW
field
0.4
Ward ef al. (1986)
P. flesus
Flounder
SW
field
1.4
Hardisty ef al. (1974)
Diaphus dumerili
Headlightfish
SW
field
0.1
Leatherland etal. (1973)
Postlarval fish
SW
laboratory
0.68
Baptist and Lewis (1969)
F. heteroclitus
Mummichog
SW
laboratory
0.11
Baptist and Lewis (1969)
L. xanthurus
Spot
SW
laboratory
0.17
Willis and Sunda (1984)
176
-------
Appendix L
Table L-4. Trophic Transfer of Metals to Fish from Contaminated Prey/Food
Concentrations
'
Species
Metal
Diet
Initial
Control
Final
TTC (%)
Reference
Notes
Copper
R. trout
Copper
110
2.9
4.6
1.55
Mount etal 1994
60 d, live, combined aqueous/diet
R.trout
Copper
140
2.9
4.7
1.29
Mount et al 1994
60 d, live, combined aqueous/diet
R.trout
Copper
200
2.9
4.4
0.75
Mount etal 1994
60 d, live, combined aqueous/diet
R.trout
Copper
250
2.9
5.9
1.20
Mount etal 1994
60 d, live, combined aqueous/diet
R.trout
Copper
440
5.7
19.6
3.16
Mount etal 1994
60 d, live, combined aqueous/diet
R. trout
Copper
830
5.7
22.4
2.01
Mount etal 1994
60 d, live, combined aqueous/diet
R.trout
Copper
1000
5.7
27.7
2.20
Mount ef a/1994
60 d, live, combined aqueous/diet
R.trout
Copper
55
2.9
3.4
0.91
Mount et al 1994
60 d, live, combined aqueous/diet
R.trout
Copper
110
2.9
5.1
2.00
Mount et al 1994
60 d, live, combined aqueous/diet
R.trout
Copper
200
2.9
6.4
1.75
Mount et al 1994
60 d, live, combined aqueous/diet
R.trout
Copper
340
2.9
7.1
1.24
Mount et al 1994
60 d, live, combined aqueous/diet
R.trout
Copper
320
2.9
8.8
1.84
Mount et al 1994
60 d, live, combined aqueous/diet
R. trout
Copper
200
5.75
17
5.63
Handy 1992
32 d, feed, no depuration
R.trout
Copper
200
5.75
5.5
0.00
Handy 1992
32 d, feed, w/ 12 d depuration
B.trout
Copper
87
8.5
11.5
3.45
Woodward et al. 1995
88 d, collected, dead, 24h no feed before analysis
B.trout
Copper
178
6
26
11.24
Woodward et al. 1995
88 d, collected, dead, 24h no feed before analysis
B.trout
Copper
174
7.5
34
15.23
Woodward et al. 1995
88 d, collected, dead, 24h no feed before analysis
R. trout
Copper
381
6.25
33.5
7.15
Woodward et al. 1994
91d, collected, dead, starved 24h before analysis
R.trout
Copper
14
6.25
3.5
0.00
Woodward et al. 1994
9Id, collected, dead, starved 24h before analysis
R.trout
Copper
12
6.25
4.35
0.00
Woodward et al. 1994
80d, feed, starved 24h before analysis
R.trout
Copper
109
6.25
16
8.94
Woodward etal. 1994
80d, collected, dead, starved 24h, vitamins
R.trout
Copper
415
6.25
39
7.89
Woodward etal. 1994
80d, collected, dead, starved 24h, vitamins
R.trout
Copper
38.8
8
20.62
Farag e/cr/. 1994
2Id, collected, dead, starved 24 h
R.trout
Copper
185.7
6.5
3.50
Farag etal. 1994
2Id, collected, dead, starved 24 h
C.trout
Copper
9.9
5.2
3.5
0.00
Farag et al. 1999
90d, feed, starved 24h
C.trout
Copper
32.9
5.2
6.1
2.74
Farag et al. 1999
90d, collected, starved 24h, vitamins
C.trout
Copper
61.5
5.2
9
6.18
Farag et al. 1999
90d, collected, starved 24h, vitamins
C.trout
Copper
43.8
5.2
12.3
16.21
Farag etal. 1999
90d, collected, starved 24h, vitamins
Cadmium
R.trout
Cadmium
7.6
0.36
0.69
4.34
Mount et al 1994
60 d, live, combined aqueous/diet
R.trout
Cadmium
16
0.36
0.95
3.69
Mount et al 1994
60 d, live, combined aqueous/diet
R.trout
Cadmium
23
0.36
1.08
3.13
Mount etal 1994
60 d, live, combined aqueous/diet
R.trout
Cadmium
21
0.36
1.29
4.43
Mount et al 1994
60 d, live, combined aqueous/diet
R.trout
Cadmium
9.5
0.76
1.31
5.79
Mount et al 1994
60 d, live, combined aqueous/diet
R.trout
Cadmium
36
0.76
2.76
5.56
Mount et al 1994
60 d, live, combined aqueous/diet
R.trout
Cadmium
69
0.76
6.83
8.80
Mounter al 1994
60 d, live, combined aqueous/diet
R.trout
Cadmium
150
0.15
5.45
3.53
Handy 1992
32 d, feed, no depuration
R.trout
Cadmium
150
0.15
0.985
0.56
Handy 1992
32 d, feed, w/ 12 d depuration
Guppy
Cadmium
69.5
3
4.32
Hatekeyama and Yasuno 1982
30 d, live, 1 d water depuration
Guppy
Cadmium
125.9
5
3.97
Hatekeyama and Yasuno 1982
30 d, live, 1 d water depuration
Guppy
Cadmium
170.6
6
3.52
Hatekeyama and Yasuno 1982
30 d, live, 1 d water depuration
R.trout
Cadmium
3
0.05
0.3
8.33
Kumada etal. 1973
12 wks, feed, no depuration
R.trout
Cadmium
3
0.04
0.1
2.00
Kumada era/. 1973
12 wks, feed, 6 wks depuration
R.trout
Cadmium
10
0.05
0.65
6.00
Kumada et al. 1973
12 wks, feed, no depuration
R.trout
Cadmium
10
0.04
0.09
0.50
Kumada et al. 1973
12 wks, feed, 6 wks depuration
R.trout
Cadmium
30
0.05
1.9
6.17
Kumada etal. 1973
12 wks, feed, no depuration
R.trout
Cadmium
30
0.04
0.12
0.27
Kumada etal. 1973
12 wks, feed, 6 wks depuration
R.trout
Cadmium
100
0.05
5.6
5.55
Kumada et al. 1973
12 wks, feed, no depuration
R.trout
Cadmium
100
0.04
0.27
0.23
Kumada et al. 1973
12 wks, feed, 6 wks depuration
B.trout
Cadmium
nd (<0 27)
0.075
0.15
na
Woodward et al. 1995
88 d, collected, dead, 24h no feed before analysis
B.trout
Cadmium
nd (<0.27)
0.225
na
Woodward et al. 1995
88 d, collected, dead, 24h no feed before analysis
B.trout
Cadmium
0.26
0.044
0.45
156.25
Woodward et al. 1995
88 d, collected, dead, 24h no feed before analysis
R.trout
Cadmium
3.12
0.05
0.8
24.04
Woodward etal. 1994
9Id, collected, dead, starved 24h before analysis
R.trout
Cadmium
0.36
0.05
0.05
0.00
Woodward et al. 1994
9Id, collected, dead, starved 24h before analysis
R.trout
Cadmium
0.5
0.05
0.095
9.00
Woodward el al. 1994
80d, feed, starved 24h before analysis
R.trout
Cadmium
1.2
0.05
0.11
5.00
Woodward et al. 1994
80d, collected, dead, starved 24h, vitamins
R.trout
Cadmium
2.39
0.05
0.6
23.01
Woodward et al. 1994
80d, collected, dead, starved 24h, vitamins
R.trout
Cadmium
0.9
0.225
25.00
Farag et al. 1994
2Id, collected, dead, starved 24h
R.trout
Cadmium
1
0.085
8.50
Farag etal. 1994
2Id, collected, dead, starved 24h
177
-------
Appendix L
Table L-4. Trophic Transfer of Metals to Fish from Contaminated Prey/Food (cont.)
Concentrations
'
Species
Metal
Diet
Initial
Control
Final
TTC (%)
Reference
Notes
C.trout
Cadmium
0.21
0.04
0.04
0.00
Farag etal. 1999
90d, feed, starved 24h
C.trout
Cadmium
0.97
0.04
0.1
6.19
Farag etal. 1999
90d, collected, starved 24h, vitamins
C.trout
Cadmium
29.9
0.04
2.88
9.50
Farag et al 1999
90d, collected, starved 24h, vitamins
C.trout
Cadmium
29.1
0.04
4.33
14.74
Farag etal. 1999
90d, collected, starved 24h, vitamins
Lead
R.trout
Lead
33
0.98
1.93
2.88
Mount et al 1994
50 d, live, combined aqueous/diet
R. trout
Lead
58
0.98
2.37
2.40
Mount et al 1994
50 d, live, combined aqueous/diet
R.trout
Lead
90
0.98
2.31
1.48
VIount et al 1994
50 d, live, combined aqueous/diet
R.trout
Lead
82
0.98
3.09
2.57
vlount et al 1994
50 d, live, combined aqueous/diet
R.trout
Lead
88
1.74
6.29
5.17
Mount et al 1994
50 d, live, combined aqueous/diet
R.trout
Lead
130
1.74
8.96
5.55
Mount et al 1994
50 d, live, combined aqueous/diet
R.trout
Lead
210
1.74
10
3.93
Mount et al 1994
50 d, live, combined aqueous/diet
B.trout
Lead
7
1
1.2
2.90
Woodward et al. 1995
!8 d, collected, dead, 24h no feed before analysis
B.trout
Lead
15
1.1
2.5
9.33
Woodward et al. 1995
88 d, collected, dead, 24h no feed before analysis
B.trout
Lead
15
0.95
4.35
22.67
Woodward et al. 1995
!8 d, collected, dead, 24h no feed before analysis
R.trout
Lead
nd (<2.0)
0.5
0.6
na
Woodward et al. 1994
9Id, collected, dead, starved 24h before analysis
R. trout
Lead
32.7
0.5
2.5
6.12
Woodward et al. 1994
9Id, collected, dead, starved 24h before analysis
R.trout
Lead
0.36
0.5
nd (<0.2)
0.00
Woodward et al. 1994
SOd, feed, starved 24h before analysis
R.trout
Lead
9.69
0.5
1
0.00
Woodward et al. 1994
SOd, collected, dead, starved 24h, vitamins
R.trout
Lead
28.4
0.5
2.4
6.69
Woodward et al. 1994
SOd, collected, dead, starved 24h, vitamins
R.trout
Lead
0.2
0.75
375.00
Farag etal. 1994
2 Id, collected, dead, starved 24 h
R.trout
Lead
8.6
0.25
2.91
Farag etal. 1994
2Id, collected, dead, starved 24 h
C.trout
Lead
0.2
0.2
0.2
0.00
Farag etal. 1999
90d, feed, starved 24h
C.trout
Lead
7.4
0.2
1.2
13.51
Farag etal.. 1999
90d, collected, starved 24h, vitamins
C.trout
Lead
792
0.2
36.8
4.62
Farag etal. 1999
90d, collected, starved 24h, vitamins
C.trout
Lead
452
0.2
52.3
11.53
Farag etal.. 1999
90d, collected, starved 24h, vitamins
Zinc
R.trout
Zinc
300
88
101
4.33
Mount et al. 1994
60 d, live, combined aqueous/diet
R.trout
Zinc
460
88
104
3.48
Mount etal. 1994
60 d, live, combined aqueous/diet
R.trout
Zinc
720
88
92
0.56
Mount etal. 1994
60 d, live, combined aqueous/diet
R.trout
Zinc
740
88
107
2.57
Mount et al. 1994
60 d, live, combined aqueous/diet
R.trout
Zinc
920
116
163
5.11
Mount ef al. 1994
60 d, live, combined aqueous/diet
R.trout
Zinc
930
116
189
7.85
Mount etal. 1994
60 d, live, combined aqueous/diet
R.trout
Zinc
1900
116
303
9.84
Mount et al 1994
SO d, live, combined aqueous/diet
R.trout
Zinc
185
165
89.19
Woodward etal.. 1994
80d, feed, starved 24h before analysis
R.trout
Zinc
655
155
23.66
Woodward et al. 1994
SOd, collected, dead, starved 24h, vitamins
R.trout
Zinc
1070
180
16.82
Woodward et al.. 1994
80d, collected, dead, starved 24h, vitamins
R.trout
Zinc
148.2
nd
nd
Farag etal. 1994
I Id, collected, dead, starved 24h
R.trout
Zinc
320.9
nd
nd
Farag et al. 1994
2Id, collected, dead, starved 24h
C.trout
Zinc
135
78
130
38.52
Farag etal.. 1999
90d, feed, starved 24h
C.trout
Zinc
384
78
160
21.35
Farag etal. 1999
90d, collected, starved 24h, vitamins
C.trout
Zinc
2336
78
380
12.93
Farag etal. 1999
90d, collected, starved 24h, vitamins
C.trout
Zinc
2119
78
520
20.86
Farag etal. 1999
90d, collected, starved 24h, vitamins
Arsenic
R.trout
Arsenic
35
3.1
4.6
4.29
Mount etal. 1994
60 d, live, combined aqueous/diet
R.trout
Arsenic
40
3.1
5.3
5.50
Mount et al 1994
60 d, live, combined aqueous/diet
R.trout
Arsenic
51
3.1
5.4
4.51
Mount ef a/. 1994
60 d, live, combined aqueous/diet
R.trout
Arsenic
63
3.1
6.7
5.71
Mounter al. 1994
60 d, live, combined aqueous/diet
B.trout
Arsenic
6.5
0.8
0.95
2.31
Woodward et al. 1995
88 d, collected, dead, 24h no feed before analysis
B.trout
Arsenic
19
1.85
3.55
8.95
Woodward et al. 1995
88 d, collected, dead, 24h no feed before analysis
B.trout
Arsenic
19
1.45
3.9
12.89
Woodward et al. 1995
88 d, collected, dead, 24h no feed before analysis
R.trout
Arsenic
6.5
0.15
1
13.08
Woodward et al. 1995
88 d, collected, dead, 24h no feed before analysis
R.trout
Arsenic
19
0.45
2.9
12.89
Woodward et al. 1995
88 d, collected, dead, 24h no feed before analysis
R.trout
Arsenic
19
0.25
3.6
17.63
Woodward et al. 1995
88 d, collected, dead, 24h no feed before analysis
R.trout
Arsenic
3.5
1.25
1
0.00
Woodward et al. 1994
9Id, collected, dead, starved 24h before analysis
R.trout
Arsenic
43.1
1.25
12
24.94
Woodward et al. 1994
9Id, collected, dead, starved 24h before analysis
R.trout
Arsenic
2.8
1.25
1.15
0.00
Woodward et al. 1994
80d, feed, starved 24h before analysis
R.trout
Arsenic
5
1.25
1.05
0.00
Woodward etal. 1994
80d, collected, dead, starved 24h, vitamins
R.trout
Arsenic
42
1.25
7.5
14.88
Woodward et al. 1994
80d, collected, dead, starved 24h, vitamins
R.trout
Arsenic
1.5
nd
nd
Farag etal. 1994
2Id, collected, dead, starved 24h
178
-------
Appendix L
Table L-4. Trophic Transfer of Metals to Fish from Contaminated Prey/Food (cont.)
Concentrations
Species
Metal
Diet
Initial
Final
TTC (%)
Reference
Notes
Control
R.trout
Arsenic
15.4
nd
nd
Farag et al. 1994
2Id, collected, dead, starved 24h
C. trout
Arsenic
3.5
0.76
1.8
29.71
Farag et al. 1999
90d, feed, starved 24h
C.trout
Arsenic
2.6
0.76
0.9
5.38
Farag et al. 1999
90d, collected, starved 24h, vitamins
C.trout
Arsenic
50.8
0.76
3.3
5.00
Farag et al. 1999
90d, collected, starved 24h, vitamins
C.trout
Arsenic
13.5
0.76
2.4
12.15
Farag etal. 1999
90d, collected, starved 24h, vitamins
R.trout
Arsenic
ISO
4.5
15.5
6.11
Cockell and Hilton, 1988
56 d, feed, no depuration
R.trout
Arsenic
360
4.5
44
10.97
Cockell and Hilton, 1988
56 d, feed, no depuration
R.trout
Arsenic
732
4.5
89.5
11.61
Cockell and Hilton, 1988
56 d, feed, no depuration
R.trout
Arsenic
1477
4.5
108
7.01
Cockell and Hilton, 1988
56 d, feed, no depuration
R.trout
Arsenic
137
4.5
34.5
21.90
Cockell and Hilton, 1988
56 d, feed, no depuration
R.trout
Arsenic
262
4.5
45.5
15.65
Cockell and Hilton, 1988
56 d, feed, no depuration
R.trout
Arsenic
500
4.5
56
10.30
Cockell and Hilton, 1988
S6 d, feed, no depuration
R.trout
Arsenic
1053
4.5
72.5
6.46
Cockell and Hilton, 1988
56 d, feed, no depuration
R.trout
Arsenic
163
2.5
15
7.67
Cockell and Hilton, 1988
56 d, feed, no depuration
R.trout
Arsenic
362
2.5
22
5.39
Cockell and Hilton, 1988
56 d, feed, no depuration
R. trout
Arsenic
793
2.5
34.5
4.04
Cockell and Hilton, 1988
56 d, feed, no depuration
R.trout
Arsenic
1497
2.5
57
3.64
Cockell and Hilton, 1988
56 d, feed, no depuration
R.trout
Arsenic
193
2.5
10
3.89
Cockell and Hilton, 1988
56 d, feed, no depuration
R. trout
Arsenic
405
2.5
16
3.33
Cockell and Hilton, 1988
56 d, feed, no depuration
R.trout
Arsenic
735
2.5
19
2.24
Cockell and Hilton, 1988
56 d, feed, no depuration
R. trout
Arsenic
1503
2.5
30.5
1.86
Cockell and Hilton, 1988
56 d, feed, no depuration
179
-------
Appendix L
Table L-5. Trophic Transfer of Metals from Field Collected Benthic Invertebrates
Concentrations
Diet | Initial/Control
Final
TTC (%)
Reference
Copper
B. trout
87
8.5
11.5
3.45
Woodward et al. 1995
B.trout
178
6
26
11.24
Woodward et al. 1995
B. trout
174
7.5
34
15.23
Woodward et al. 1995
R.trout
381
6.25
33.5
7.15
Woodward et al. 1994
R.trout
14
6.25
3.5
0.00
Woodward et al. 1994
R.trout
109
6.25
16
8.94
Woodward et al. 1994
R.trout
415
6.25
39
7.89
Woodward et al. 1994
R.trout
38.8
8
20.62
Farag et al. 1994
R.trout
185.7
6.5
3.50
Farag et al. 1994
C. trout
32.9
5.2
6.1
2.74
Farag etal. 1999
C.trout
61.5
5.2
9
6.18
Farag etal. 1999
C.trout
43.8
5.2
12.3
16.21
Farag etal. 1999
Cadmium
B.trout
nd (<0.27)
0.075
0.15
na
Woodward et al. 1995
B.trout
nd (<0.27)
0.225
na
Woodward et al. 1995
R.trout
3.12
0.05
0.8
24.04
Woodward et al. 1994
R.trout
0.36
0.05
0.05
0.00
Woodward et al. 1994
R.trout
1.2
0.05
0.11
5.00
Woodward et al. 1994
R.trout
2.39
0.05
0.6
23.01
Woodward et al. 1994
R.trout
0.9
0.225
25.00
Farag et al. 1994
R.trout
1
0.085
8.50
Farag et al. 1994
C.trout
0.97
0.04
0.1
6.19
Farag etal. 1999
C.trout
29.9
0.04
2.88
9.50
Farag etal. 1999
C.trout
29.1
0.04
4.33
14.74
Farag etal. 1999
Lead
B.trout
7
1
1.2
2.90
Woodward et al. 1995
B.trout
15
1.1
2.5
9.33
Woodward et al. 1995
B.trout
15
0.95
4.35
22.67
Woodward et al. 1995
R.trout
nd (<2.0)
0.5
0.6
na
Woodward et al. 1994
R.trout
32.7
0.5
2.5
6.12
Woodward et al. 1994
R.trout
9.69
0.5
1
0.00
Woodward et al. 1994
R.trout
28.4
0.5
2.4
6.69
Woodward et al. 1994
R.trout
8.6
0.25
2.91
Farag etal. 1994
C.trout
7.4
0.2
1.2
13.51
Farag etal.. 1999
C.trout
792
0.2
36.8
4.62
Farag etal. 1999
C.trout
452
0.2
52.3
11.53
Farag etal. 1999
Zinc
R.trout
655
155
23.66
Woodward etal. 1994
R.trout
1070
180
16.82
Woodward et al. 1994
R.trout
148.2
nd
nd
Farag et al. 1994
R.trout
320.9
nd
nd
Farag etal. 1994
C.trout
384
78
160
21.35
Farag etal. 1999
C.trout
2336
78
380
12.93
Farag etal. 1999
C.trout
2119
78
520
20.86
Farag etal. 1999
Arsenic
B.trout
6.5
0.8
0.95
2.31
Woodward et al. 1995
B.trout
19
1.85
3.55
8.95
Woodward et al. 1995
B.trout
19
1.45
3.9
12.89
Woodward et al. 1995
R.trout
6.5
0.15
1
13.08
Woodward et al. 1995
R.trout
19
0.45
2.9
12.89
Woodward et al. 1995
R.trout
19
0.25
3.6
17.63
Woodward et al. 1995
R.trout
3.5
1.25
1
0.00
Woodward et al. 1994
R.trout
43.1
1.25
12
24.94
Woodward et al. 1994
R.trout
5
1.25
1.05
0.00
Woodward et al. 1994
R.trout
42
1.25
7.5
14.88
Woodward et al. 1994
R.trout
1.5
nd
nd
Farag etal. 1994
R.trout
15.4
nd
nd
Farag et al. 1994
C.trout
2.6
0.76
0.9
5.38
Farag etal. 1999
C.trout
50.8
0.76
3.3
5.00
Farag et al. 1999
C.trout
13.5
0.76
2.4
12.15
Farag etal. 1999
180
-------
Appendix L
References Cited:
Cockell, K.A. and J.W. Hilton. 1988. Preliminary investigations on the comparative chronic toxicity of
four dietary arsenical to juvenile rainbow trout (Salmo gairdneri R.). Aquat. Toxicol. 12:73-82.
Dallinger, R. and H. Kautzky. 1985. The importance of contaminated food for the uptake of heavy metals
by rainbow trout {Salmo gairdneri)-. a field study. Oecologia (Berlin) 67:82-89
Farag, A.M., C.J. Boese, D.F. Woodward, and H.L. Bergman. 1994. Physiological changes and tissue
metal accumulation in rainbow trout exposed to foodborne and waterborne metals. Env. Tox. Cont.
13(12):2021-2029.
Farag, A.M., M.J. Suedkamp, J.S. Meyer, R. Barrows, and D.F. Woodward. 2000. Distribution of metals
during digestion by cutthroat trout fed benthic invertebrates contaminated in the Clark Fork River,
Montana and the Coeur d'Alene River, Idaho, U.S.A., and fed artificially contaminated Artemia. J. Fish
Biol. 56:173-190.
Farag, A.M., D.F. Woodward, W. Brumbaugh, J.N. Goldstein, E. MacConnell, C. Hogstrand, and F.T.
Barrows. 1999. Dietary effects metals-contaminated invertebrates from the Coeur d'Alene River, Idaho,
on cutthroat trout. Trans. Am. Fish. Soc. 128:578-592.
Handy, R.D. 1992. The assessment of episodic metal pollution. II. The effects of cadmium and copper
enriched diets on tissue contaminant analysis in rainbow trout (Oncorhynchus mykiss). Arch. Env.
Contam. Toxicol. 22:82-87.
Harrison, S.E. and P.J. Curtis. 1992. Comparative accumulation efficiency of 109cadmium from natural
food (Hyallela azteca) and artificial diet by rainbow trout (Oncorhynchus mykiss). Bull. Env. Cont. Tox.
49:757-764
Hatakeyama, S. and M. Yasuno. 1982. Accumulation and effects of cadmium on guppy (Poecilia
reticulata) fed cadmium-dosed cladocera (Moina macrocopa). Bull. Env. Contam. Toxic. 29:159-166.
Metayer, C., J.C. Amiard, C. Amiard-Triquet, and J. Marchand. 1980. Etude de transfer! de quelques
oilgo-elements dans les chaines trophiques neritique et estuariennes: accumulation biologique chez le
poissons omnivores et super-carnivores. Helg. Meeresunters. 34:179-191.
Mount, D.R., A.K. Barth, T.D. Garrison, K.A. Barten and J.R. Hockett. 1994. Dietary and waterborne
exposure of rainbow trout (Oncorhynchus mykiss) to copper, cadmium, lead and zinc using a live diet.
Env. Toxicol. Chem. 13( 12):2031 -2041.
Reinfelder, J.R., N.S. Fisher, S.N. Luoma, J.W.Nichols, W.-X. Wang. 1998. Trace element trophic
transfer in aquatic organisms: A critique of the kinetic model approach. Sci. Total Env. 219:117-135.
Suedel B.C., J.A. Boraczek, R.K. Peddicord, P.A. Clifford, and T.M. Dillon. 1994. Trophic transfer and
biomagnification potential of contaminants in aquatic ecosystems. Rev. Env. Cont. Toxic. 136:21-89.
Woodward, D.F., A.M. Farag, H.L. Bergman, A.J. DeLonay, E.E. Little, C.E. Smith and F.T. Barrows.
1995. Metals-contaminated benthic invertebrates in the Clark Fork River, Montana: Effects on age-0
brown trout and rainbow trout. Can. J. FishAq. Sci. 52:1994-2004.
181
-------
Appendix L
Woodward, D.F., W.G. Brumbaugh, A.J. DeLonay, E.E. Little, and C.E. Smith. 1994. Effects on rainbow
trout fry of a metals-contaminated diet of benthic invertebrates from the Clark Fork River, Montana.
Trans. Am. Fish. Soc. 123:51-62.
182
-------
SUPPLEMENTAL
INFORMATION
-------
Methods Presented for
Consideration by the Army Corps of
Engineers (USACE)
-------
White Paper
The Use of Probabilistic Exposure Modeling to Derive HARS-Speciflc Values
Submitted by: U.S. Army Corps of Engineers
USEPA Region II completed, during the summer of 1998, a peer-review of the risk-based
evaluation process used by the NY District of the Corps and USEPA Region II since
1996 to make permit decisions for dredging projects. Several comments and
recommendations made by those peer reviewers concerned the importance of
characterizing, in probabilistic terms, the uncertainty associated with the analyses
involved. In response to these comments and to advance the state-of-the-art in
assessments of dredged material and contaminated sediments, the U.S. Army Engineer
Research and Development Center undertook an effort to demonstrate the application and
potential utility of probabilistic modeling approaches to assess risks posed by
bioaccumulation of contaminants from dredged material.
The results of this effort, summarized in the two attached draft manuscripts submitted for
peer-review publication, has led the Corps to the conclusion that the HARS-Specific
Values and the evaluative process could be more confidently applied to make permit
decisions if probabilistic methods and more site-specific approaches were employed. The
commitment of additional effort to develop such an evaluative process is justified given
the serious environmental and economic consequences of permit decisions.
There are significant benefits to be had by approaching the exposure modeling for the
HARS using probabilistic methods. As it currently stands, the Corps and EPA would be
hard pressed to justify the single values used for several key input parameters. The
uncertainties associated with some of these inputs are simply too large. When faced with
parameter uncertainty USEPA Region II has endeavored to select "conservative" values
as input parameters, i.e. values that should error on the side of protecting the
environment. However, it is unlikely that the subjective approach employed by the
Region will result in an optimal or "ideal" amount of protection. It is much more likely
that the approach taken will result in an uncontrolled and highly uncertain level of
protection. The most accessible solution to this problem is to approach uncertainty in as
quantitative a manner as possible using data and methods described in USEPA guidance
and reports and the scientific literature.
The studies represented by the two attached manuscripts used a NY dredging and
disposal scenario and regional data. However, these studies were not intended to result in
HARS-Specific Values, but to provide examples of how probabilistic exposure modeling
and site-specific and species-specific information could be used to assess risks posed by
contaminated sediment.
-------
MANUSCRIPT FOR SUBMITTAL TO RISK ANALYSIS
Importance of Uncertainty and Variability to Predicted Risks from Trophic Transfer of
PCBs in Dredged Sediments
von Stackelberg, Katherine, E. Menzie-Cura & Associates, Inc. 1 Courthouse Lane, Suite 2,
Chelmsford, MA 01824
Burmistrov, Dmitriy Menzie-Cura & Associates, Inc.
Vorhees, Donna, J. Menzie-Cura & Associates, Inc.
Bridges, Todd, S. US Army Corps of Engineers, Vicksburg, MS.
Linkov, Igor Arthur D. Little, Inc., Cambridge, MA
Running Title: Uncertainty in Trophic Transfer
1
-------
MANUSCRIPT FOR SUBMITTAL TO RISK ANALYSIS
Abstract
Biomagnification of organochlorine and other persistent organic contaminants by higher
trophic level organisms represents one of the most significant sources of uncertainty and
variability in evaluating potential risks associated with disposal of dredged materials. While it is
important to distinguish between population variability (e.g., true population heterogeneity in
fish weight, and lipid content) and uncertainty (e.g., measurement error), they can be
operationally difficult to define separately in probabilistic estimates of human health and
ecological risk. We propose a disaggregation of uncertain and variable parameters based on: a)
availability of supporting data; b) the specific management and regulatory context (in this case,
of the US Army Corps of Engineers/US Environmental Protection Agency tiered approach to
dredged material management); and, c) professional judgment and experience in conducting
probabilistic risk assessments. We describe and quantitatively evaluate several sources of
uncertainty and variability in estimating risk to human health from trophic transfer of
polychlorinated biphenyls (PCBs) using a case study of sediments obtained from the New York-
New Jersey Harbor and being evaluated for disposal at an open water off-shore disposal site
within the Northeast region.
The estimates of PCB concentrations in fish and dietary doses of PCBs to humans ingesting
fish are expressed as distributions of values, of which the arithmetic mean or mode represents a
particular fractile. The distribution of risk values is obtained using a food chain biomagnification
model developed by Gobas (1,2) by specifying distributions for input parameters disaggregated
to represent either uncertainty or variability. Only those sources of uncertainty that could be
quantified were included in the analysis.
2
-------
MANUSCRIPT FOR SUBMITTAL TO RISK ANALYSIS
Results for several different two-dimensional Latin Hypercube analyses are provided to
evaluate the influence of the uncertain versus variable disaggregation of model parameters. The
analysis suggests that variability in human exposure parameters is greater than the uncertainty
bounds on any particular fractile, given the described assumptions.
Keywords: biomagnification, probabilistic risk assessment (PRA), polychlorinated biphenyls,
dredged material, trophic transfer, uncertainty and variability
1. Introduction
Sediment-associated contaminants represent a significant source of environmental risks
associated with the dredging and disposal of sediments due, in part, to the potential for
bioaccumulation and biomagnification of some contaminants in aquatic food chains.
Accumulation of these contaminants in tissues of aquatic organisms can lead to adverse effects to
the aquatic organisms themselves (3,4,5,6,7,8,9) as well as to the ecological and human receptors
ingesting them (10,11,12,13,14). The USACE has developed an Environmental Effects Residue
Database (ERED), available at http://www.wes.armv.mil/el/ered, that compiles and makes
accessible information on tissue concentrations of contaminants associated with adverse
ecological effects or, in some cases, with no adverse effects.
USEPA and USACE have established a tiered approach for assessing the potential for
environmental impacts of open-water disposal of dredged materials. These approaches are
outlined in two documents, Evaluation of Dredged Material Proposed for Ocean Disposal -
Testing Manual (15), and Evaluation of Dredged Material Proposed for Discharge in Waters of
the U.S. - Testing Manual (16).
3
-------
MANUSCRIPT FOR SUBMITTAL TO RISK ANALYSIS
Using the approach described in USEPA/USACE (15,16), a Tier I analysis examines
whether a determination of potential environmental impact can be made based on existing
information. Tier II provides rapid chemical screening for potential impacts. In Tier ID, toxicity
and bioaccumulation tests are performed on dredged material to determine whether it would be
expected to cause unacceptable impacts. When Tier III analysis results in a highly uncertain
conclusion, Tier IV may include a risk assessment.
A qualitative analysis of the many sources of uncertainty involved in evaluating impacts of
open-water disposal is presented by Vorhees et al. (17). The goal of that analysis was to improve
dredged material management decision-making by identifying and ranking sources of uncertainty
that could contribute to costly or inappropriate decisions by managers. Vorhees et al. (17)
concluded that trophic-chain transfer is one of the most important sources of uncertainty in the
current assessment framework for dredged material management.
There is uncertainty in the estimated concentrations of sediment and water to which aquatic
organisms are exposed and also variability in parameters contributing to contaminant
bioaccumulation. Uncertainty and variability should be viewed separately in risk assessment
because they have different implications to regulators and decision makers (18,19). Variability is
a population measure, and provides a context for a deterministic average or reasonable maximum
exposure (RME) point estimate. Variability cannot be reduced, only better understood. In
contrast, uncertainty represents unknown but often measurable quantities. Typically, uncertainty
can be reduced by obtaining additional measurements of the uncertain quantity. Quantitatively
separating uncertainty and variability allows an analyst to determine the fractile of the population
for which a specified risk occurs and the uncertainty bounds or confidence interval around that
4
-------
MANUSCRIPT FOR SUBMITTAL TO RISK ANALYSIS
predicted risk. If uncertainty is large relative to variability (i.e., contributes most to the range of
risk estimates) and if the differences in cost among management alternatives are high, additional
collection and evaluation of information can be recommended before making management
decisions on contaminated sediments. On the other hand, including variability in risk estimates
allows decision makers to quantitatively evaluate the likelihood of risks both above and below
selected reference values or conditions (for example, average risks as compared to 95th percentile
risks).
We developed a spreadsheet model based on the modeling approach of Gobas (1,2) for a
typical open water aquatic food web in the Northeast region. This model relies on steady-state
solutions of differential equations describing time-varying uptake of PCBs. The model
incorporates both sediment and water sources in predicting PCB uptake based on prey
consumption, direct water uptake across the gill, and food web dynamics. This analysis uses
sediment chemistry and 28-day bioaccumulation test results from New York-New Jersey Harbor
(NY-NJ) sediments and literature values (20,21) to develop distributions for input parameters.
These input parameters include: sediment and water column concentrations of PCBs, weight and
lipid content of aquatic organisms, total organic carbon in sediment, dissolved organic carbon
and particulate organic carbon in the water column, and KoW- While the model has been used in
numerous regulatory contexts (22,23,24), there is uncertainty as to whether this is the appropriate
conceptual model to use as well as both uncertainty and variability in the specific rate constants.
Conceptual model uncertainty, while assumed to be low, is not directly quantifiable. Uncertainty
and variability in the rate constants is addressed by providing distributions for the variables that
drive the rate constants (for example, KoW, fish weight, etc.).
5
-------
MANUSCRIPT FOR SUBMITTAL TO RISK ANALYSIS
Two versions of the model were developed: the first, called "calculated sandworm model,"
uses measured sediment PCB concentrations from NY-NJ Harbor sediments proposed for
dredging and estimates concentrations at the base of the foodweb under equilibrium assumptions.
The second, called "measured sandworm model," uses the 28-day bioaccumulation test results in
which sandworms were exposed to the two sediments and tissue concentrations were measured
following the exposure. Using sandworm concentration data eliminates the need to estimate
concentrations at the base of the benthic food web from measured sediment concentrations. The
USACE guidance calls for bioaccumulation testing at Tier 3 and these data are routinely
collected (15,16).
The data set selected for this example is from the Kill Van Kull tributary in New Jersey.
Although the current analysis is restricted to PCBs, the approach and conclusions are generally
applicable to hydrophobic, lipophilic organic contaminants (1,2).
2. Methods
2.1 Conceptual Model
To evaluate the potential for an adverse impact from trophic transfer of contaminants in
dredged materials, a simplified sediment-based food web was designed to represent conditions at
a hypothetical offshore open water dredged material disposal site. The offshore disposal site is
assumed to be located in the NY-NJ region. We assume predominant food chain exposures
originate via sediments rather than water because it is the sediment-associated contaminants that
are of most concern in a bioaccumulative context when making decisions about whether to
dispose of dredged materials (15,16,25,26,27).
6
-------
MANUSCRIPT FOR SUBMITTAL TO RISK ANALYSIS
The selection of fish species was based on several criteria including: 1) importance to
commercial fishing, 2) abundance in the New York/New Jersey coastal area, 3) importance in the
diet of piscivorous fish and other consumers, and 4) representative of particular habitats or
trophic levels. Resident species with small foraging areas were specifically selected because
these fish are likely to be more highly exposed to sediments in open-water disposal sites than
species with larger ranges.
In our conceptual model, the polychaete worm Nereis virens (sandworm) represents the base
of the food web. The sandworm is a deposit feeder that burrows and lives in sediment and moves
partially out of its burrow to feed (28). Mummichogs (Fundulus heteroclitus) represent the next
trophic level after sandworms and feed primarily on invertebrates in the sediment. Mummichogs
have limited home and foraging ranges and occupy tidal creeks, coves and inlets (21,29,30). The
top-level predator, summer flounder (Paralichthys dentatus) is a piscivorous fish, with adults
occupying the top of the aquatic food chain (31,32).
Feeding preferences for all of the organisms will vary with the age and size of the individual.
This analysis assumes the sandworm is representative of a number of invertebrates that
mummichog might consume, and Nereis virens is specifically referenced as a favored dietary
item for mummichog (30). Predicted (or measured) concentrations in sandworms are considered
as representative of the expected concentrations via bioaccumulation from sediment. This is
consistent with the information contained in the US ACE database of biota-sediment
bioaccumulation factors obtained from the results of 28-day sediment bioaccumulation tests and
other sources. The BSAF database is available through the Dredging Operations Technical
7
-------
MANUSCRIPT FOR SUBMITTAL TO RISK ANALYSIS
Support (DOTS) Web page of the USACE Waterways Experiment Station
(http://www.wes.armv.mil/el/dots/).
Likewise, summer flounder will consume a variety of invertebrates and smaller fish (32), but
are known to preferentially consume mummichog (32). Predicted concentration distributions in
mummichog represent a surrogate population of fish occupying similar trophic levels and
utilizing similar foraging strategies as a number of fish consumed by summer flounder. Humans
then consume summer flounder (under the assumption that predicted concentrations in summer
flounder are representative of a number of fish species that might be consumed). This simplified
food web is itself uncertain since the exact feeding preferences of the fish are not known.
However, this uncertainty is not quantitatively evaluated and these simplifications are consistent
with an evaluation of a sediment-driven food web.
The models presented in this paper are designed to estimate body burdens in adult fish
because adult fish are more likely to be consumed by humans than juveniles. Human exposures
are estimated using a distribution for fish ingestion rate developed from the USEPA Exposure
Factors Handbook (33,34) for adult recreational consumption of marine fish in the mid-Atlantic
region. Body weight and exposure duration are estimated for men and women combined. The
cancer slope factor is obtained from the Integrated Risk Information System (13) and is specified
as a point estimate following USEPA guidance (35); thus, this important source of uncertainty is
not quantitatively described. Averaging time for carcinogenic effects is a lifetime. Table I
provides a summary of these parameters.
2.2 Modeling Framework
8
-------
MANUSCRIPT FOR SUBMITTAL TO RISK ANALYSIS
The model was developed using the framework first conceptualized by Gobas (1,2). The
steady state solution to the general form of the differential equation describing the change in
concentration of PCBs in biota with respect to time is given by:
_ kl* Cwd + kd* Cdiet ,,
C/ = (1
kl + ke + km + kg
ki
= gill uptake rate (L/Kg/d)
Cwd
= freely dissolved concentration in water (ng/L)
kd
= dietary uptake rate (d"1)
Cjiet
= concentration in the diet (jAg/kg)
k2
= gill elimination rate (d"1)
ke
= fecal egestion rate (d"1)
km
= metabolic rate (d"1)
kg
= growth rate (d1)
cf
= concentration in fish (jug/kg)
Several sources provided equations for the rate constants (1,2,36). The rate constants
themselves are uncertain, however, this uncertainty is incorporated in the distributions for the
input parameters which drive the rate constants. Input parameters (Tables I through HI) are
represented either as single values or distributions. The fish body burden model is used to predict
point estimates or distributions of concentrations to which humans are exposed via fish ingestion.
The model requires sediment and freely dissolved water column concentrations as inputs.
Since the sediments are being evaluated for potential disposal, obtaining water column
9
-------
MANUSCRIPT FOR SUBMITTAL TO RISK ANALYSIS
concentrations in overlying water from the area from which the sediments were obtained is not
useful to evaluating concentrations at the disposal site, particularly since the sediments are
obtained from an estuary with numerous influences while the disposal site is in an offshore, open
water environment. To obtain an upperbound water column concentration, this analysis assumes
that once the dredged sediments are placed on the sea floor, sediment and overlying water will
come into equilibrium with one another. Predicted upperbound water concentrations of each
chemical were estimated from the organic carbon-water partition coefficient, KoC and the
distribution of organic carbon-normalized sediment concentrations for the New York/New Jersey
locations as follows:
(Co<
Ctv =
yKocj
(2)
where
Cw = concentration of freely dissolved chemical in the water (jig/L)
Coc = the organic carbon-normalized sediment concentration (jj,g/kg dry wt sediment) and
Koc = organic carbon-water partition coefficient (L/kg organic carbon)
The Koc for each chemical was estimated from its octanol-water partition coefficient, KoV
according to the following regression relationship (37):
log Koc = 0.00028 + 0.983 logi0Kow (3)
10
-------
MANUSCRIPT FOR SUBMITTAL TO RISK ANALYSIS
The model can also accept as a measured invertebrate concentration resulting from the
standard Tier 3 28-day bioaccumulation test results (15,16). To account for the fact that these
measured concentrations may not have achieved steady-state, a Kow-dependent adjustment is
made (15,16,38,39):
log tss = 6.9 x 10"3(log Kow)4 - 1.85 x 10"'(log K^)3 + 1.65(log K^)2 - 5.34(log K^) + 5.93 (4)
where:
tss = time required to reach steady-state
Finally, the predicted risk to adults (men and women combined) is given as:
n. , CSF * IRf * C/ * ED (5)
Risk =
BW* 1000000 *AT
where:
Risk -
incremental lifetime cancer risk
CSF =
cancer slope factor (mg/kg-day)"1
IRf —
annualized fish ingestion rate (g/day)
cf =
concentration in fish (|xg/kg)
ED -
exposure duration (days)
BW =
body weight (kg)
AT =
averaging time (days)
11
-------
MANUSCRIPT FOR SUBMITTAL TO RISK ANALYSIS
Many of the model parameters are described by distributions. The construction of these
distributions considered whether a parameter should be characterized as uncertain (i.e., there is
insufficient information concerning a true, but unknown, value), or variable (i.e., inherent
population heterogeneity). Development of these distributions is described in Section 2.3.
2.3 Parameterization of the Models
Separating uncertain and variable parameters in a simulation allows decision-makers to
distinguish the uncertainty in predicted risk due to lack of knowledge and the variation caused by
natural variability in a measurement or population (18,19,40,41,42,43). In this exercise, model
parameters are categorized as either predominantly variable or uncertain in different
combinations to evalaute the impact on predicted risks. Although this distinction is artificial
since all parameters, to some extent, will exhibit both variability and uncertainty (and/or
uncertainty in the specific parameterization of variability), it is made for a number of reasons: i)
uncertainty and variability are separated in the context of achieving particular management
goals; ii) in many cases, one or the other category will dominate; and iii) the availability of data
to be able to characterize both uncertainty and variability simultaneously (40). In this example,
the management goal is to reach a decision about the suitability of dredged materials for open-
water disposal within the USEPA/USACE tiered approach.
In this analysis, we attempt to quantify the uncertainty associated with the variable
parameters listed in Table I to the extent possible. For example, lipid content, considered
variable, also reflects some uncertainty attributable to measurement error. However, this source
of uncertainty is expected to be low relative to observed population heterogeneity in lipid values
12
-------
MANUSCRIPT FOR SUBMITTAL TO RISK ANALYSIS
(22,25). The same is true for body weight (33,34,41). This may not necessarily be the case for
the human fish ingestion rate, but the available data did not support quantifying the uncertainty
in this rate without a high degree of subjectivity. Such subjectivity may not be informative
unless supported by a broader consensus.
This analysis did not quantitatively assess uncertainty associated with the models
themselves. This analysis assumes that the models (both bioaccumulation in aquatic organisms
and the USEPA model for estimating carcinogenic and noncarcinogenic risks) are valid and does
not explicitly consider any conceptual model uncertainties.
2.3.1 Aquatic Food Web Parameters
Sediment Concentrations
Sediment concentrations were obtained from an actual USACE dredging project and are
expressed as total PCBs (|ig/kg dry weight). Over the long term, aquatic organisms and humans
consuming aquatic organisms will integrate exposure over larger temporal and spatial scales.
Therefore, it is exposure to the distribution of average sediment concentrations that is of
regulatory interest, and, this source of uncertainty can be reduced by collecting additional
samples or by refining the spatial extent of dredged materials. Uncertainty in observed sediment
concentrations is characterized by the standard error on the mean.
Freely Dissolved Water Concentrations
The Gobas model requires a freely dissolved water concentration. As described previously,
the water column concentration of interest is that which is expected at the disposal site. These
13
-------
MANUSCRIPT FOR SUBMITTAL TO RISK ANALYSIS
were obtained by running a submodel based on the conservative assumption that the overlying
water at the disposal site comes into equilibrium with the sediment according to equation 3.
Equation 3 contains a term for Koc, which is obtained from equation 4 based on Kow- Since K<>w
is described by a distribution, Koc is also described by a distribution. There is statistical
uncertainty in the regression relationship described in equation 3, but this source of uncertainty
could not be quantitatively evaluated (it is expected to be low based on the published R ).
Predicted freely dissolved water column concentrations are characterized as uncertain. It is
inherent in the equilibrium assumption that sediment and water concentrations are correlated, but
it is unlikely that sediment and water are in equilibrium in the environment. Therefore, the
analyses were run first assuming a correlation of one and then of zero.
Measured Sandworm Concentrations
Measured sandworm concentrations were obtained from the results of the 28-day
bioaccumulation tests. These results were adjusted by the relationship between time to achieve
steady-state and Log KoW as described by equation (4). Measured sandworm concentrations are
considered uncertain rather than variable given that aquatic receptors exposed to invertebrates
through the diet will integrate exposure and therefore be exposed to an average and some
uncertainty around that average.
Octanol-Water Partition Coefficient (Kow)
Kow is evaluated first as an uncertain parameter and then as a variable parameter. It is not
possible to analytically separate the uncertainty in the variability for this parameter. We argue
14
-------
MANUSCRIPT FOR SUBMITTAL TO RISK ANALYSIS
that KoW should be treated as variable rather than uncertain for several reasons. First,
theoretically, KoW for any given contaminant can be determined in the laboratory and these data
are available. Second, PCBs are being evaluated as "total" PCBs. Total PCBs represent a
mixture of individual congeners, each of which has its own KoW. The congener distribution
between exposure media and exposed organisms is typically quite different, with aquatic
organisms showing a shift in the congener distribution toward higher chlorinated congeners than
that of the underlying exposure media (49). Although the true distribution of congeners and
associated K<,ws in the mixture is technically unknown, KoW should still be considered variable
because differences in KoW among individual congeners contribute to the differential uptake of
the congeners in the mixture.
The distribution for KoW was specified as a triangular distribution where the range is given by
the minimum and maximum KoW for the individual PCB congeners analyzed, and the mode is
estimated as the average of all the congeners in the mixture. The KoW data were obtained from
Mackay et al. (57). The model was run again assuming the same distribution for KoW as
uncertain to evaluate the effect on predicted risks.
Lipid Content in Aquatic Organisms
The lipid content of fish for a particular age class (for example, adult fish) vary greatly due to
temperature, seasonality, prey availability, and other biological factors. The variability is
typically much greater than the uncertainty in the estimates (22,25,49,50). Percent lipid
distributions were specified as uniform for the mummichog (21,30), and as triangular for the
15
-------
MANUSCRIPT FOR SUBMITTAL TO RISK ANALYSIS
sandworm (28, 51,52,53,54,55,56) and the summer flounder (31,32,55,56) based on information
from the literature. Body weight of mummichog (21,30) and summer flounder (55,56) were
specified as normal.
Weight of Fish
Fish weight, as with lipid content, shows significant seasonal and lifetime variability,
depending on temperature, prey availability, and other biological factors. Distributions for body
weight of fish were obtained from literature sources (21, 28-32)
Residence Time of Fish
Residence time of the fish was specified as a uniform distribution and reflects the amount of
time per year that a fish spends exposed to the disposal site (29).
Total Organic Carbon in Sediment
TOC was obtained from the sediment data results and determined to follow a normal
distribution.
2.3.2 Human Exposure Parameters
Human exposure parameters reflecting variability in the population include body weight, fish
consumption rate, and exposure duration. There were no correlations specified for these
parameters as there was no evidence in the literature for such correlations.
Body Weight
16
-------
MANUSCRIPT FOR SUBMITTAL TO RISK ANALYSIS
The body weight distribution was characterized as lognormal based on USEPA (33) for adult
men and women combined.
Fish Ingestion Rate
Similarly, the fish ingestion rate, combined for men and women, was derived from data
presented in USEPA (34) for the recreational consumption of marine fish in the mid-Atlantic
region.
Exposure Duration
The selected exposure duration assumes 30 years of fish consumption because USEPA
typically uses this time period to reflect the reasonable maximum time that someone lives in a
single location.
PCB Carcinogenicity
Although there is considerable uncertainty in the estimate of PCB carcinogenicity
(13,14,35,48), this source of uncertainty was not explicitly modeled as EPA regulatory programs
do not permit probabilistic analyses for toxicity values. Further, within the USACE tiered
approach, decision makers do not have the opportunity to reduce this source of uncertainty
(although other research programs could clearly be designed for this purpose). The toxicity
values provided by the USEPA (13,14) were used. For potential carcinogenic effects, this value
is 2.0 (mg/kg-day)"1. For potential non-cancer effects, the Reference Dose is 2 x 10~5 mg/kg-day,
assuming that the PCB congener mixture in fish is similar to Aroclor 1254.
17
-------
MANUSCRIPT FOR SUBMITTAL TO RISK ANALYSIS
2.4 Sensitivity Analysis
Following the estimation of incremental lifetime cancer risks and hazard indices, sensitivity
analyses were run using rank correlation and percent contribution to variance to determine the
model sensitivity to individual parameters. For each combination of input parameters, the output
of the model was recorded and combinations of all possible inputs and outputs were plotted
against each other to show the influence of the parameters on the output values. If the Spearman
or partial rank regression coefficient is close to 1 or -1 for a specific input model parameter, this
parameter significantly influences model output.
2.5 Software Implementation
The food chain models were constructed using Crystal Ball™ software. Crystal Ball™
permits key input variables to be represented by a distribution of values rather than a single point
estimate. The simulations were conducted using Latin Hypercube sampling and the sensitivity
analyses used the algorithms incorporated in the Crystal Ball™ software.
The model uses a nested analysis to evaluate uncertainty and variability separately by
selecting values from the uncertain distributions and then freezing uncertainty values and
running the entire model using the variability distributions for a given number of iterations
(48,58). This "inner loop - outer loop" process repeats for a number of uncertainty simulations,
describing how the predicted risk distribution varies due to the uncertain parameters. Figure 1
provides a schematic of this process. First, uncertain variables are fixed. Then, 1,500 variability
iterations are run, holding the uncertain variables constant. Finally, the process is repeated for
250 simulations of uncertain variables.
18
-------
MANUSCRIPT FOR SUBMITTAL TO RISK ANALYSIS
To test the hypothesis that this number of iterations is sufficient, we ran 500 simulations,
each with 10,000 iterations and compared the results (i.e. cumulative distributions of noncancer
and cancer risks) to those obtained using 250 simulations, each with 1,500 iterations. Although a
small difference in risks between simulations with lower and higher sample numbers was
detected, the Kolmogorov-Smirnov test (59) and the t-test indicated that the difference between
these distributions and the means, respectively, were not statistically significant. All statistical
analyses were conducted with Statistica software (StatSoft).
3. Modeling Results
This section presents the results of the modeling and sensitivity analyses.
Under most regulatory programs, hazard indices exceeding 1 are considered to indicate
potential risk, while carcinogenic effects are evaluated within a range of 10"4 and 10"6. Predicted
risks below 10"6 are considered not to pose a risk, and predicted risks above 10"4 generally
require some action be taken. There is room for judgment within this range.
3.1 Incremental Lifetime Cancer Risks
Figure 2 presents the results of the incremental lifetime cancer risk estimates. The x-axis
shows the individual fractiles (i.e., reflects the variability in the population estimates), and the y-
axis displays risk expressed as number per million exposed individuals. Box and whisker plots
are used to represent the uncertainty in risk estimates for each fractile of the population. The
median point estimate for incremental lifetime cancer risk for fish ingestion at the 50th percentile
(median of the population) is just above 2 in one million for the measured sandworm PCB
concentrations (cases 1 and 2), and almost exactly 2 in one million for the calculated PCB
concentrations (cases 3 and 4). An incremental lifetime cancer risk of 3 additional cancer cases
19
-------
MANUSCRIPT FOR SUBMITTAL TO RISK ANALYSIS
per million exposed individuals can also be interpreted as the median exposed individual facing a
3 in a million incremental lifetime cancer risk.
The context for these point estimates is provided by the probabilistic output. The box and
whisker plots shown in Figure 2 are interpreted as follows. For example, using case 1 (top plot
in Figure 2) we expect 65% of the population to have a 95% probability of experiencing an
incremental lifetime cancer risk of close to 6 in one million or less, the top bar of the box and
whisker plot for the 65th percentile. At the top of the box, we expect 65% of the population to
have a 75% probability or less of experiencing an incremental lifetime cancer risk of 4 in one
million or less.
These plots show that when Log KoW is considered uncertain (case 1) rather than variable
(case 2), as expected the uncertainty bounds increase. For example, at the 95th percentile, the
uncertainty in the predicted incremental lifetime risk increased from 6 to 12 in one million
(roughly a factor of two in case 2), to less than 1 to 14 in one million (or an order of magnitude
in case 1). However, the median risk estimates in terms of variability appear more stable. That
is, the median predicted risk at each percentile is very close between the two cases.
The results are somewhat different for the calculated sandworm case. In this case, the
uncertainty bounds increase when Log KoW is specified as uncertain rather than variable
(compare cases 3a with 4a, 3b with 4b), and there is somewhat more of a difference in the
median estimate. For example, the median predicted incremental lifetime cancer risk for the 95th
percentile is approximately 9 in one million for case 3 a and over 11 in one million for case 4a.
3.2 Sensitivity Analysis
20
-------
MANUSCRIPT FOR SUBMITTAL TO RISK ANALYSIS
By plotting each individual input value selected for a particular iteration against the
corresponding output (incremental lifetime cancer risk), the sensitivity of the output to the input
can be estimated using rank correlation. Table IV presents the results of the ranking exercise.
The fish ingestion rate is the dominant parameter across all scenarios. For the measured
sandworm cases (1 and 2), Log K
-------
MANUSCRIPT FOR SUBMITTAL TO RISK ANALYSIS
is, the difference between the predicted median and 95th fractile risk estimates is greater than the
uncertainty in the average estimate. Thus, decision makers explicitly see the impact of
population variability on risk. However, this model assumes that the variability in fish tissue
concentration within the aquatic environment is the same as the variability in fish tissue
concentration in the human diet.
Under the measured sandworm model, the fish ingestion rate contributes the most to the
variance in the predicted risk distribution, followed by Log KoW, exposure duration, body weight,
and temperature. In the calculated sandworm model, fish ingestion rate and sandworm lipid
contribute most to the variance. When the correlation between sediment and water is set to 1
(cases 3b and 4b), then sediment and water concentrations contribute as much to the variance as
the fish ingestion rate.
Although this model did not evaluate the uncertainty associated with variability (i.e.,
uncertainty in the variability distributions), neither the management goals nor the available data
warranted such an analysis. The separation of uncertain versus variable parameters within the
context of achieving management goals together with an evaluation of the individual parameter
attributes allows decision makers to determine the impact of uncertainty that can be reduced
within the US ACE tiered approach. Although there are additional sources of uncertainty that
were not quantified (for example, in the toxicity estimates), these cannot be addressed within the
management context of the decision to be made. Certainly a research program to reduce that
source of uncertainty could be designed, but it is beyond the scope of the specific management
objective in this case.
22
-------
MANUSCRIPT FOR SUBMITTAL TO RISK ANALYSIS
To evaluate the reasonableness of the assumptions underlying the analysis, we evaluated
combinations of parameters leading to specific risk estimates. For example, what fish ingestion
rate, exposure duration, and body weight are associated with particular fractiles of incremental
lifetime cancer risk or hazard indices? Evaluating the reality of the assumptions to avoid the
pitfall of "compounding conservatism" is an important component of any probabilistic analysis
(60).
th
In this case, the 95 fractile hazard index and incremental lifetime cancer risk estimates are
associated with the following combinations of parameters: 680 fish meals (assumed to be 280
grams) over the course of 12 years for an individual with a body weight of 63 kg; 1,372 fish
meals over the course of 25 years for an individual with a body weight of 44 kg or 589 fish meals
over the course of 15 years for an individual weighing 52 kg. The maximum calculated hazard
index and incremental lifetime risk are associated with 81 fish meals per year for 30 years for an
individual weighing 62 kg.
In general, the measured sandworm case showed less of a range in uncertainty than the
calculated sandworm model. Several factors contribute to this, notably the assumptions at the
base of the food web. Since predicted sandworm concentrations are obtained from benthic lipid,
TOC, and sediment concentrations, there can be greater than an order of magnitude difference in
these predictions. Table IV presents the predicted sandworm concentrations using discrete
values from the distributions for percent lipid in benthos and TOC.
The results of the sensitivity analysis, together with the specific estimated incremental
lifetime cancer risks in the context of trophic transfer, suggest that given the assumptions
presented here, decision makers are better off obtaining better fish ingestion rate information
23
-------
MANUSCRIPT FOR SUBMITTAL TO RISK ANALYSIS
than collecting additional sediment samples prior to making a disposal decision if we assume that
sediment and water concentrations are not correlated. The framework presented here illustrates
the usefulness of separating uncertainty and variability, even if it is an artificial separation. The
artificial separation is performed in the context of specific management goals, professional
judgment regarding the predominance of uncertainty versus variability for each parameter, and
the availability of data for characterizing uncertainty and variability.
5. Acknowledgement
The authors wish to thank Drs. L. Rosman, J. Link, and K. Able for providing
information on lipid concentration and weight for summer flounder and sandworms. This study
was supported by the US Army Corps of Engineers, Dredging Operations Environmental
Research Program (DOER). Permission was granted by the Chief of Engineers to publish this
material.
6. References
1. F.A.P.C. Gobas, "A Model for Predicting the Bioaccumulation of Hydrophobic Organic
Chemicals in Aquatic Food-Webs: Application to Lake Ontario," Ecological Modelling 69,
1-17(1993).
2. F.A.P.C. Gobas, M.N. Z'Graggen and X. Zhang, "Time Response of the Lake Ontario
Ecosystem to Virtual Elimination of PCBs," Environmental Science and Technology 29,2038-
2046(1995).
3. D.E. Black, D.K. Phelps, and R.L. Lapan. "The effect of inherited contamination on egg and
larval winter flounder, Pseudopleuronectes americanus." Marine Environmental Research.
25,45-62(1988).
24
-------
MANUSCRIPT FOR SUBMITTAL TO RISK ANALYSIS
4. D.E. Black, Gutzahr-Gobell, R., Pruell, R.J., Bergen, B., and McElroy, A.E. "Effects of a
mixture of non-ortho- and mono-ortho-polychlorinated biphenyls on reproduction in fundulus
heteroclitus (linnaeus)." Environmental Toxicology and Chemistry, 17(7), 1396-1404
(1998).
5. D.E. Black, R. Gutjahr-Gobell, R.J. Pruell, B. Bergen, L. Mills, and A.E. McElroy.
"Reproduction and polychlorinated biphenyls in Fundulus heteroclitus (Linnaeus) from New
Bedford Harbor, Massachusetts, USA." Environmental Toxicology and Chemistry, 17(7),
1405-1414(1998).
6. U. Borgmann, W.P. Norwood, and K.M. Ralph. "Chronic toxicity and bioaccumulation of
2,5,2',5 - and 3,4,3',4'-tetrachlorobiphenyl and Aroclor 1242 in the amphipod Hyalella
azteca." Arch. Environ. Contam. Toxicol. 19, 558-564(1990).
7. E. Casillas, D. Misitano, L.L. Johnson, L.D. Rhodes, T.K. Collier, J.E. Stein, B.B. McCain
and U. Varanasi. "Inducibility of spawning and reproductive success of female English sole
('Parophrys vetulus) from urban and nonurban areas of Puget Sound, Washington." Marine
Environmental Research. 31, 99-122 (1991).
8. S.W. Fisher, S.W. Chordas III, P.F. Landrum. "Lethal and sublethal body residues for PCB
intoxication in the oligochaete, Lumbriculus variegatus." Aquatic Toxicology. 45, 115-126
(1999).
9. A.V. Nebeker and F.A. Puglisi. 1974. "Effect of polychlorinated biphenyls (PCBs) on
survival and reproduction of daphnia, gammarus, and tanytarsus, Trans. Am. Fish. Soc."
103(4), 722-728 (1974).
25
-------
MANUSCRIPT FOR SUBMITTAL TO RISK ANALYSIS
10. S.N. Wiemeyer, C.M. Bunck, and C.J. Stafford. Environmental Contaminants in bald eagle
eggs - 1980-84- and further interpretations of relationships to productivity and shell
thickness. Archives of Environmental Contamination and Toxicology. 24, 213-227 (1993).
11 .D.E. Tillitt, G.T. Ankley, J.P. Giesy, J.P. Ludwig, H.Kurita-Matsuba, L. Sileo, K.L.
Stromborg, J. Larson, T.J. Kubiak. Polychlorinated biphenyl residues and egg mortality in
double-crested cormorants from the Great Lakes. Environmental Toxicology and Chemistry.
11, 1282-1288 (1992).
12. J.C. Restum, S.J. Bursian, J.P. Giesy, J.A. Render, W.G. Helferich, E.B. Shipp, and D.A.
Verbrugge. Multigenerational study of the effects of consumption of PCB-contaminated carp
from Saginaw Bay, Lake Huron, on mink. 1. Effects on mink reproduction, kit growth and
survival, and selected biological parameters. Journal of Toxicology and Environmental
Health, Part A, 54, 343-375 (1998).
13. United States Environmental Protection Agency, Integrated Risk Information System
Database (IRIS), http://www.epa.gov/iris. (2000).
14. Agency for Toxic Substances Disease Registry, Toxicological Profile for Polychlorinated
Biphenyls (PCBs), U.S. Department of Health and Human Services (December 1998).
15. United States Environmental Protection Agency and United States Army Corps of Engineers,
Evaluation of Dredged Material Proposed for Ocean Disposal: Testing Manual. (EPA-
503/8-91/001, 1991).
16. United States Environmental Protection Agency and United States Army Corps of Engineers,
Evaluation of Dredged Material Proposed Discharge in Waters of the U.S. Testing Manual:
Inland Testing Manual. (EPA 823-B-98-004, 1998).
26
-------
MANUSCRIPT FOR SUBMITTAL TO RISK ANALYSIS
17. D.J.Vorhees, S.B.Kane Driscoll, K.von Stackelberg, and T.S. Bridges, "Improving Dredged
Material Management Decisions with Uncertainty Analysis," Dredging Operations and
Environmental Research Program, Technical Report DOER-3, (December 1998).
18. K.M. Thompson, and J.D. Graham, "Going Beyond the Single Number: Using Probabilistic
Risk Assessment to Improve Risk Management," Human and Ecological Risk Assessment 2,
1008-1034 (1996).
19. M.G. Morgan, and M. Henrion, Uncertainty: A Guide to Dealing with Quantitative Risk and
Policy Analysis (Cambridge University Press, New York, NY 1990).
20. T.S. Bridges, Data from several sites in the NY-NJ Harbor from US Army Corps of
Engineers (1999).
21. T.J. Iannuzzi, N.W. Harrington, N.M. Shear, C.L. Curry, H. Carlson-Lynch, M.H. Henning,
S.H. Su, and D.E. Rabbe, "Distributions of Key Exposure Factors Controlling the Uptake of
Xenobiotic Chemicals in an Estuarine Food Web," Environmental Toxicology and Chemistry
15,1979-1992 (1996).
22. United States Environmental Protection Agency (USEPA). Great Lakes Water Quality
Initiative Technical Support Document for the Procedure to Determine Bioaccumulation
Factors, Office of Water, Washington, DC. EPA/820/B-95/005. (March 1995).
23. Thermo Retec Consulting Corporation. "Baseline Human Health and Ecological Risk
Assessment for the Lower Fox River," Prepared for Wisconsin Department of Natural
Resources, February 24, 1999. Available online at:
www.dnr.state.wi.us/org/water/wm/lowerfox/rifs/
27
-------
MANUSCRIPT FOR SUBMITTAL TO RISK ANALYSIS
24. United States Environmental Protection Agency, Revised Baseline Modeling Report for the
Hudson River Remedial Investigation/Feasibility Study Volume 2D. Prepared by Limno-
Tech, Inc., Menzie-Cura & Associates, Inc. and Tetra-Tech, Inc. for US EPA. (January
2000).
25. United States Environmental Protection Agency. Ambient Water Quality Criteria Derivation
Methodology Technical Support Document, Final Draft. Office of Science and Technology,
EPA/822/B-98/005, (July 1998).
26. United States Environmental Protection Agency. Methodology for Deriving Ambient Water
Quality Criteria for the Protection of Human Health (2000). Office of Water and Office of
Science and Technology, EPA/822/B-00/004, (October 2000).
27. United States Environmental Protection Agency (USEPA). Bioaccumulation testing and
interpretation for the purpose of sediment quality assessment - status and needs. Office of
Water, Office of Solid Waste, EPA-823-R-00-001. (February 2000).
28. W.H. Wilson, Jr. and R.E. Ruff, "Species Profiles: Life Histories and Environmental
Requirements of Coastal Fishes and Invertebrates (North-Atlantic) - Sandworm and
Bloodworm," (FWS/OB2-82/11.80, US Fish and Wildlife Service, Washington, D.C., 1988).
29. V.A. Lotrich, "Summer Home Range and Movements of Fundulus heteroclitus (Pisces:
Cyprinodontidae) in a Tidal Creek," Ecology 56, 191-198 (1975).
30. B.J. Abraham, "Species Profiles: Life Histories and Environmental Requirements of Coastal
Fishes and Invertebrates (Mid-Atlantic) - Mummichog and Killifish," (FWS/OB2-82/11.40,
US Fish and Wildlife Service, Washington, D.C., 1985).
28
-------
MANUSCRIPT FOR SUBMITTAL TO RISK ANALYSIS
31. B.H. Grimes, M.T. Huish, J.H. Kerby, J.S. Fish, and D. Moran, "Species Profiles: Life
Histories and Environmental Requirements of Coastal Fishes and Invertebrates (Mid-
Atlantic) - Summer and Winter Flounder," (FWS/OB2-82/11.112, US Fish and Wildlife
Service, Washington, D.C., 1989).
32. L.M. Dery, "Post workshop age and growth study of young summer flounder," In: Smith,
R.W., L.M. Dery, P.G. Scarlett, and A. Jearld, Jr. (eds.), Proceedings of the summer flounder
(Paralichthys dentatus) age and growth workshop, 20-21 May 1980, Woods Hole, MA, p. 7-
11. Tech. Memo. NMFS-F/NEC-11, Northeast Fish. Cent., Natl. Mar. Fish. Serv., NOAA,
Woods Hole, MA 02543, 30 p. (1981).
33. U.S. Environmental Protection Agency (USEPA), Exposure Factors Handbook, Volume I:
General Factors (EPA, Office of Research and Development, Washington D.C., EPA/600/P-
95/002Fa, 1997a).
34. U.S. Environmental Protection Agency, Exposure Factors Handbook, Volume II: Food
Ingestion Factors (EPA, Office of Research and Development, Washington D.C.,
EPA/600/P-95/002Fb, 1997b).
35. U.S. Environmental Protection Agency, Risk Assessment Guidance for Superfund, Volume 1
-Human Health Evaluation Manual, Part A, Interim Final (EPA, Office of Emergency and
Remedial Response, Washington, D.C., EPA/540/1-89/0002, 1989).
36. L.P. Burkhard," Comparison of Two Models for Predicting Bioaccumulation of
Hydrophobic Organic Chemicals in a Great Lakes Food Web," Environmental Toxicology
and Chemistry 17, 383-393 (1998).
29
-------
MANUSCRIPT FOR SUBMITTAL TO RISK ANALYSIS
37. D.M. DiToro, C.S. Zarba, D.J. Hansen, W.l Berry, R.C. Swartz, C.E. Cowan, S,P. Pavlou,
H.E.Allen, N.A. Thomas, and P.R. Paquin, P. R "Technical basis for establishing sediment
quality criteria for nonionic organic chemicals using equilibrium partitioning." Environmental
Toxicology and Chemistry 10, 1541-1583 (1991).
38. Connell DW, Hawker DW. 1988. "Use of polynomial expressions to describe the
bioconcentration of hydrophobic chemicals by fish." Ecotoxicol Environ Saf, 16(3):242-57
39. McFarland, V. A. (1994). "Evaluation of field-generated accumulation factors for predicting
the bioaccumulation potential of sediment-associated PAH compounds," Ph.D. diss.,
Northeast Louisiana University, Monroe.
40. E.J. Kelly and K. Campbell. "Separating Variability and Uncertainty in in Environmental
Risk Assessment - Making Choices." Human Ecol. Risk Assess. 6, 1-13(2000).
41. U.S. Environmental Protection Agency, Risk Assessment Guidance for Superfund, Volume 3
- Part A, Process for Conducting Probabilistic Risk Assessment. Draft, Revision No. 5,
(1999).
42. T.E. McKone, "Uncertainty and Variability in Human Exposures to Soil Contaminants
Through Home-Grown Food: A Monte Carlo Assessment" Risk Analysis 14,449-463 (1994).
43. P.S. Price, S.H. Su, J.R. Harrington, and R.E. Keenan, " Uncertainty and variation in indirect
exposure assessments: An Analysis of Exposure to Tetrachlorodibenzo-p-dioxin from a Beef
Consumption Pathway," Risk Analysis 16, 263-277 (1996).
44. P.M. Gschwend and S. Wu. On the Constancy of Sediment-Water Partition Coefficients of
Hydrophobic Organic Pollutants. Environ. Sci. Technol. 19, 90-96 (1985).
30
-------
MANUSCRIPT FOR SUBMITTAL TO RISK ANALYSIS
45. P.M. Cook, R.J. Erickson. R.L. Spehar, S.P. Bradbury and G.T. Ankley. Interim Report on
Data and Methods for Assessment of 2,3,7,8-tetrachlorodibenzo-p-dioxin Risks to Aquatic
Life and Associated Wildlife. Duluth, MN: USEPA, Environmental Research Laboratory.
EPA/600/R-93/055 (1993).
46. T.R. Parsons, M. Takahashiand and B. Hargrave, Biological Oceanographic Processes
(Pergamon Press, Oxford, 1984).
47. H. Morrison, F.A.P.C. Gobas, R. Lazar, D.M. Whittle, and G.D. Haffner, "Development and
Verification of a Benthic/Pelagic Food Web Bioaccumulation Model for PCB Congeners in
Western Lake 'Erie" Environmental Science and Technology, 31,3267-3273 (1997).
48. J.T. Cohen, M.A. Lampson, and T.S. Bowers," The Use of Two-stage Monte Carlo
Simulation Techniques to Characterize Variability and Uncertainty in Risk Analysis,"
Human and Ecological Risk Assessment 2, 939-971 (1996).
49. United States Environmental Protection Agency, Baseline Ecological Risk Assessment for
the Hudson River Remedial Investigation/Feasibility Study. Prepared by Menzie-Cura &
Associates, Inc. and TAMS Consultants, Inc. for US EPA. Appendix K: Examination of
Congener Patterns to Determine Exposure. (August 2000).
50. United States Environmental Protection Agency, Peer Review Bioaccumulation Workgroup,
Peer Review for the Ambient Water Quality Criteria Derivation Methodology Technical
Support Document. (1999).
51. M.E. Schrock, E.S. Barrows, and L.B. Rosman, "Biota to Sediment Accumulation Factors
for TCDD and TCDF in Worms from 28-Day Bioaccumulation Tests," Chemosphere 34,
1333-1339(1997).
31
-------
MANUSCRIPT FOR SUBMITTAL TO RISK ANALYSIS
52. H. Lemieux, P.U. Blier, F.Dufresne, and G. Desrosiers, "Metabolism and Habitat
Competition in the Polychaete Nereis virens," Marine Ecology Progress Series 156,151-156
(1997).
53. D.E.G Briggs and A.J. Kear," Decay and Preservation of Polychaetes: Taxonomic
Thresholds in Softbodied Organisms," Paleobiology 19, 107-135(1993).
54. L.B. Rosman, Personal communication with Igor Linkov of Menzie-Cura & Associates, Inc.
regarding sandworms. (1999).
55. Northeast Fisheries Science Center (NFSC), "25th Northeast Regional Stock Assessment
Workshop (25th SAW): Stock Assessment Review Committee (SARC) Consensus Summary
of Assessments," Northeast Fish. Science Center Reference Document 97-14,143 p. (1997).
56. W.O Watanabe, E.P. Ellis, S.C. Ellis, and M.W. Feeley, "Progress in Controlled Maturation
and Spawning of Summer Flunder Paralichthys dentatus brookstock," Journal of the World
Aquaculture Society 29, 393-403 (1998).
57. D. Mackay, W.Y Shiu, and K.C. Ma, Illustrated Handbook of Physical-Chemical Properties
and Environmental Fate for Organic Chemicals. Volume I. Monoaromatic Hydrocarbons,
Chlorobenzenes, and PCBs (Lewis Publishers, Inc., Chelsea, Michigan, 1992).
58. D.E. Burmaster, and A. M. Wilson," An Introduction to Second-Order Random Variables in
Human Health Risk Assessments," Human and Ecological Risk Assessment 2, 892-919
(1996).
59. B. Rosner, Fundamentals of Biostatistics (Duxbury Press, Wadsworth Publishing Company,
Belmont, Massachusetts, 1995).
32
-------
MANUSCRIPT FOR SUBMITTAL TO RISK ANALYSIS
60. A.C. Cullen, " Measures of Conservatism in Probabilistic Risk Assessment," Risk Analysis
14, 389-393 (1994).
33
-------
MANUSCRIPT FOR SUBMITTAL TO RISK ANALYSIS
Table I: Cases Evaluated
Case Number
Uncertain
Variable
Correlation
1: Kill Van Kull measured
sandworm
Water and
sandworm
concentrations,
KoW, and TOC.
Lipid content,
body weight (fish
and human),
ingestion rate,
temperature
la: Sediment
and water = 0;
lb. Sediment
and water = 1
2. Kill Van Kull measured
sandworm
Water and
sandworm
concentrations, and
TOC
Lipid content,
body weight (fish
and human),
ingestion rate,
temperature, Kow
2a. Sediment
and water = 0;
2b: Sediment
and water = 1
3. Kill Van Kull calculated
sandworm
Sediment and water
concentrations,
KoW, and TOC
Lipid content,
body weight (fish
and human),
ingestion rate,
temperature
3 a. Sediment
and water = 0;
3b. Sediment
and water = 1
4. Kill Van Kull calculated
sandworm
Sediment and water
concentrations, and
TOC
Lipid content,
body weight (fish
and human),
ingestion rate,
temperature, KoW
4a. Sediment
and water = 0;
4b: Sediment
and water = 1
Notes:
NA - not applicable
-------
MANUSCRIPT FOR SUBMITTAL TO RISK ANALYSIS
Table II: Distributions for Kill Van Kull Dataset
Variable
Category
Units
Distribution
Shape1
Parameterization
Reference
Sediment
concentration
Uncertain
ng/g bulk dry
weight
Normal (n, o)
(19.1,8.53)
20
Water
concentration
Uncertain
ng/L freely
dissolved
Lognormal (n, c)
(0.43,0.65)
20
Sandworm
concentration
Uncertain
ng/g wet
weight
Normal (ja, o)
(38.9,10.7)
20
Total organic
carbon
(TOC)
Uncertain
Percent
Normal (|A, c)
(3.84, 0.89)
20
Kow
Variable,
Uncertai
n
Triangular (min,
mode, max)
(5.24, 6.74, 7.36)
Assumed,
see text
Residence
time
Variable
Fraction
Uniform (min,
max)
(0.4,0.6)
Assumed,
see text
Surface water
temperature
Variable
DegC
Triangular (min,
mode, max)
(8,15,22)
Assumed,
see text
Sandworm
Percent lipid
Variable
Percent
Triangular (min,
mode, max)
(1.0,1.2, 2.0)
21,28,51-
54
Mummichog
Body weight
Variable
Grams
Normal (n, o)
(3.0,2.2)
21,36,29,3
0
Percent lipid
Variable
Percent
Uniform
(0.01,0.04)
21,36,29,3
0
Summer
Flounder
Body weight
Variable
Grams
Triangular (min,
mode, max)
(200, 574, 5000)
31,32,56
Percent lipid
Variable
Percent
Triangular (min,
mode, max)
(0.23, 0.72, 2.0)
31,32,56
Human
Body weight
Variable
kg
Lognormal (ja, a)
(4.3,2.7)
33
Fish
ingestion rate
Variable
g/day
Lognormal (n, a)
(2.1,1.1)
34
Exposure
duration
Days
Uniform
(3650,10950)
33
Cancer slope
factor
Point
estimate
(mg/kg-day)"1
Point estimate
2.0
13
35
-------
MANUSCRIPT FOR SUBMITTAL TO RISK ANALYSIS
Notes:
1 - distributions parameterized as follows:
Normal (fx, a): arithmetic mean and standard deviation (reflects standard error of data)
Lognormal (|i, a): geometric mean and geometric standard deviation
Triangular (min, mode, max): minimum, likeliest value, maximum
Uniform (min, max): minimum, maximum
-------
DRAFT MANUSCRIPT FOR SUBMITTAL TO RISK ANALYSIS
06/28/01
Table III. Ranking of Contribution to Variance in Outcome
Cases la & 2a
Cases lb & 2b
Case 3a & 4a
Cases 3b & 4b
Correlation
Correlation
Correlation
Correlation
Coefficients
Coefficients
Coefficients
Coefficients
Parameter
Cancer Risk
Cancer Risk
Cancer Risk
Cancer Risk
Measured
Measured
Calculated
Calculated
Fish Ingestion Rate (g/day)
0.59
0.58
0.51
0.48
Sandworm Lipid (%)
NA
NA
0.44
0.43
log Kow (kg/L)
0.54
0.52
0.35
0.31
Exposure Duration (days)
0.29
0.30
0.24
0.25
Body Weight (kg)
-0.19
-0.19
\o
o
I
-0.16
Temperature
0.16
0.16
0.13
0.13
TOC (%)
NA
NA
-0.17
-0.15
Sediment (ng/g dry wt.)
NA
0.11
0.33
0.44
Dissolved PCBs in Water (ng/L)
0.01
0.11
0.01
0.43
Summer Flounder Lipid (%)
0.05
0.08
0.06
0.06
Mummichog Lipid (%)
0.06
0.06
0.07
0.06
Summer Flounder Weight (gr)
0.05
0.05
0.05
0.05
Mummichog Weight (gr)
0.02
0.02
0.03
0.01
Residence Time for Summer
-0.01
0.01
-0.03
0.02
Flounder (fraction)
Sandworm Weight (gr)
0.01
0.01
0.01
0.02
Measured Sandworm PCB (ug/kg) 0.14
0.15
NA
NA
37
-------
DRAFT MANUSCRIPT FOR SUBMITTAL TO RISK ANAL YSIS 06/28/01
Table IV: Combinations of Percent Lipid in Benthos and Total Organic Carbon Leading to Predicted Sandworm
Concentrations in |Xg/kg
Measured Sediment Concentration =19.1 Hg/kg
TOC-> and Percent
Lipid in Benthos
TOC = 1.2%
TOC = 2.0%
TOC = 3.8%
TOC = 5%
Benthic Lipid = 0.5%
8.0
4.8
2.5
1.9
Benthic Lipid = 1.0%
15.9
9.6
5.0
3.8
Benthic Lipid = 1.2%
19.1
11.5
6.0
4.6
Benthic Lipid = 2.0%
31.8
19.1
10.1
7.6
Measured Sediment Concentration = 28.1 jUg/kg
TOC-> and Percent
Lipid in Benthos
TOC = 1.2%
TOC = 2.0%
TOC = 3.8%
TOC = 5%
Benthic Lipid = 0.5%
11.7
7.0
3.7
2.8
Benthic Lipid = 1.0%
23.4
14.1
7.4
5.6
Benthic Lipid = 1.2%
28.1
16.9
8.9
6.7
Benthic Lipid = 2.0%
46.8
28.1
14.8
11.2
38
-------
DRAFT MANUSCRIPT FOR SUBMITTAL TO RISK ANALYSIS
Figure 1. Diagram of Nested Latin Hypercube Approach
06/28/01
Select Uncertainty Parameters
(PCB concentrations, TOC, POC, DOC)
Select trophic chain variability parameters
(Kow, weights, and lipids)
Select human exposure variability paramerters
(weight, fish ingested, exposure duration)
>
1
Calculate risk
O
# Iterations: >1,000
Specify Risk / Hazard Distribution
# Simulations: >250
§•
t
a>
I
39
-------
DRAFT MANUSCRIPT FOR SUBMITTAL TO RISK ANALYSIS
06/28/01
Figure 2: Predicted Incremental Lifetime Cancer Risks - See Table 1 for Description of Cases
Incremental Lifetime Cancer Risks:
Results of Two-Dimensional Latin-Hypercube
J*
(/)
in
L_
0)
o
T3
c
-------
DRAFT MANUSCRIPT FOR SUBMITTAL TO RISK ANALYSIS
06/28/01
26
22
18
14
^ s 10
Incremental Lifetime Cancer Risks:
Results of Two-Dimensional Latin-Hypercube
Percentile of Uncertainty
-i- 95%-tile
75%-tile
median
25%-tiic
5%-tile
-2
¦ CASE4A
~ CASE4B
5% 10% 15% 20% 25% 30% 35% 40% 45% 50% 55% 60% 65% 70% 75% 80% 85% 90% 95%
Fractile of Population Distribution (Variability)
41
-------
Attachment 1
Risk-Based Management of Contaminated Sediments: Consideration of Spatial and
Temporal Patterns in Exposure Modeling
Linkov1*, Igor, Burmistrov1, Dmitriy, Cura1, Jerome, Bridges2, Todd, S.
'Menzie-Cura & Associates, Inc. One Courthouse Lane, Suite Two, Chelmsford, MA 01824
United States Army Corps of Engineers, Vicksburg, MS.
$
Corresponding Author:
tel: (978)-322-2855
fax: (978)-970-2791
e-mail: ilinkov@menziecura.com
Key Words:
-------
Abstract
This paper addresses interactions among foraging behavior, habitat preferences, site
characteristics, and the spatial distribution of contaminants in developing PCB exposure
estimates for winter flounder at a hypothetical open water dredged material disposal site in
the coastal waters of New York and New Jersey (NY-NJ). The implications of these
interactions on human health risk estimates for local recreational anglers who fish for and
eat flounder are described. The models implemented in this study include a spatial sub-
model to account for spatial and temporal characteristics of fish exposures and a
probabilistic adaptation of the Gobas bioaccumulation model that accounts for temporal
variation in concentrations of hydrophobic contaminants in sediment and water. We
estimated the geographic distribution of a winter flounder sub-population offshore of NY-
NJ based on species biology and its vulnerability to local recreational fishing, the foraging
area of individual fish, and their migration patterns. We incorporated these parameters
and an estimate of differential attraction to a management site into a spatially explicit
model to assess the range of exposures within the population. The output of this modeling
effort, flounder PCB tissue concentrations, provided exposure point concentrations for an
estimate of human health risk through ingestion of locally caught flounder. The risks
obtained for the spatially non-explicit case are as much as one order of magnitude higher
than those obtained with explicit consideration of spatial and temporal characteristics of
winter flounder foraging and seasonal migration. This practice of "defaulting" to extremely
conservative estimates for exposure parameters in the face of uncertainty ill serves the
decision-making process for management of contaminated sediments in general and
specifically for disposal of dredged materials. Consideration of realistic spatial and
temporal scales in food chain models can help support sediment management decisions by
providing a quantitative expression of the confidence in risk estimates.
2
-------
INTRODUCTION
Exposure estimates for wildlife in areas containing spatially localized contaminants are
a function of spatial factors such as foraging area, size of the habitat and the distribution of
contamination. Species exhibiting different foraging behavior may experience significantly
different chemical exposures from the same site, even if their foraging areas overlap. Currently,
exposure estimates and subsequent human health and ecological risk projections usually assume
a static and continuous exposure of an ecological receptor to a contaminant concentration
described by some descriptive statistic such as the mean or maximum contaminant concentration
in sediment. These assumptions are thus overly conservative and ignore some of the major
advantages offered by risk assessment, the ability to account for site-specific conditions and to
conduct iterative analyses.
The importance of consideration of the spatial extent of site contamination in the
terrestrial environment has recently attracted the attention of individual researchers (1,2,3,4,5,6)
as well as regulatory agencies (7,8). These studies have called for explicit incorporation of
habitat sizes and foraging behavior for terrestrial receptors. Nevertheless, spatial scales in the
risk assessment for aquatic ecosystems has not been widely considered.
This paper proposes a framework for spatially explicit risk assessment associated with
contaminated sediments. Many sediments are contaminated as a result of industrial
development and management decisions have to be made on their safe use as well as on their
disposal. For example, the United States Army Corps of Engineers (USACE) or their permit
recipients dredge about 400 million cubic yards of sediment annually to maintain 25,000 miles of
navigational channels. About 60 million cubic yards of dredged material, including sediments
that receive urban or agricultural runoff are placed in more than 150 open water sites designated
by the US Environmental Protection Agency (US EPA). The fact that open water management
facilities are geographically restricted invites an ecological exposure analysis that accounts for
the spatial and temporal aspects of a receptor's biology. The proposed framework can be used to
support risk-based decision making in the regulation and management of contaminated sediment.
3
-------
The USACE and USEPA apply a tiered approach (10) to judge the suitability of dredged
materials for open-water disposal. Tiers I and II use existing or easily obtained information and
apply relatively inexpensive, rapid tests to predict environmental effects using site specific
information and sediment chemistry. Tiers ID and IV involve biological and chemical
evaluations that require field sampling, laboratory analyses, and risk assessment.
For some sediments and sites, bioaccumulation and biomagnification of hydrophobic
organic contaminants, such as polychlorinated biphenyls (PCBs), may represent the primary
source of environmental risks to the aquatic organisms and their higher-order predators,
including humans. This paper addresses the interactions of various aspects of foraging behavior,
habitat characteristics, and the spatial distribution of contaminants in developing PCB exposure
estimates for winter flounder at a hypothetical open water dredged material disposal site in the
coastal waters of New York and New Jersey (NY-NJ). It then considers the implications of these
interactions for human health risk estimates for local recreational anglers who fish for and eat
those flounder. We also address the advantages of such spatially explicit modeling in
environmental decision making where sediment contamination poses risk to wildlife.
MODELING APPROACH AND PARAMETERS
Conceptual Model
We developed a conceptual model to represent a predominantly sediment-driven food
web that is common for sites with contaminated sediments. The conceptual model is a simple
food chain in which the contaminant of concern is total PCBs. We selected PCBs, which are
highly lipophilic and hydrophobic, for this analysis because they: 1) are often found in
contaminated sediments; 2) are known to biomagnify through food chains; and, 3) pose risk to
both humans and ecological receptors. Although the current analysis addresses only PCBs, the
general methodology and conclusions are applicable to a wide range of organic contaminants.
The exposure media are surface water and sediments. A common polychaete, Nereis virens
(sandworm), represents the prey base for Pseudopleuronectes americanus (winter founder). The
human receptors are recreational anglers eating the flounder.
4
-------
The analysis employs a spatially explicit foraging sub-model (Figure 1) that provides a
time series of sediment and water concentrations that a fish may encounter within its habitat.
The model inputs are information on: seasonal abundance of fish; habitat size for a species, size
and location of the management area within the species habitat; size of the species foraging area;
and, sediment and water concentrations over the management site and in the surrounding areas.
The model also uses a site specific attraction factor that accounts for differential attraction to the
management area. The outputs of the spatial sub-model are combinations of sediment and
surface water concentrations that the fish population may encounter while foraging in this habitat
over time.
The analysis then applies a bioaccumulation sub-model for transfer of PCBs from
sediments and surface water through a fish food chain. This analysis uses sandworms as the base
of the food chain. The sandworm's exposure to sediment contaminants represents a conservative
estimate of invertebrate exposure because it is a deposit feeder that burrows and lives in
sediment and moves only partially out of its burrow to feed (11).
Winter flounder, which represent the next trophic level above sandworms, feed primarily
on invertebrates in the sediment. Their feeding preferences vary with the age and size of the
individual. The adult fish mostly consume annelids, molluscs, and cnidaria. Several investigators
(12,13,14) noted that they are omnivorous, opportunistic feeders and prey upon various sediment
dwelling organisms such as polychaete worms, amphipods and isopods (crustaceans),
pelecypods, and plant material. Steimle and Terranova (15) found that winter flounder from
contaminated and cleaner control areas fed primarily on the tentacular crowns of tube-dwelling
anemones and large polychaete worms. Within this conceptual model, we assumed that the
flounder feed solely on the polychaete Nereis virens. This is a conservative assumption in terms
of potential exposures because the other major component of their diet, anemones, are likely to
be less exposed to sediment than the sandworms.
Winter flounder is a reasonable representative fish species because it: 1) is an important
recreational and commercial species; 2) occurs abundantly in the New York/New Jersey coastal
area; 3) represents a higher order, bottom-feeding predator; and, 4) is a resident species with a
relatively small foraging area. Therefore, it will more frequently encounter localized
5
-------
contaminated sediments than other recreationally obtained species such as bluefish or striped
bass that forage over larger ranges.
The output of the bioaccumulation sub-model is a time series of fish tissue residues of
PCBs. The human risk sub-model (Figure 1) averages these time series and then calculates
human risk due to ingestion of these fish based on accepted human exposure parameters.
Spatial Submodel
The approach for the spatial sub-model is an extension and modification of a prior
method (3). The habitat area is divided into a grid of one meter by one meter cells across the
management area and surrounding habitat. In the current study, all cells within the limits of the
disposal site are assigned a probability distribution for PCB concentration (see Table 1 and
discussion below). While all cells outside the management area are assumed to be free of PCBs,
other model parametrization (for example, inclusion of background concentration outside the
facility) is possible.
The spatial sub-model calculates exposure point concentrations for fish utilizing the
habitat. At specified time periods each individual fish in the simulation is modeled to forage in
randomly selected areas within the habitat. The exposure point concentration for each time step
is the average concentration of the cells that a fish encounters within its foraging area for a
specified time period. In the current simulation, a monthly time step is used because the data for
fish abundance is available only in monthly increments. However, a time interval shorter than 1
month can be implemented.
The exposure point concentration depends on the sizes of foraging area, facility and
habitat: Individuals with a small foraging area may feed exclusively within the facility
boundaries and thus have a high accumulation of contaminants, while individuals feeding outside
the facility may not be affected. On the other hand, species foraging widely over the habitat can
be continuously exposed to contaminated areas, but to a much smaller extent.
The outputs of the spatial sub-model are the exposure point concentrations that individual
fish encounter over time. These time series data are used in the bioaccumulation sub-model to
calculate PCB body burdens in fish over time. The input parameters for the spatial sub-model
6
-------
include sediment and water PCB concentrations, size and location of the management site within
the habitat, site attraction factor, seasonal abundance of fish, fish foraging area, and habitat size.
Sediment and Water Concentrations.The key input parameter for sediment is the total
PCB concentration in Hg/kg dry weight. The sediment data in the analysis are from several sites
in NY-NJ Harbor (collectively referred to as Dredged Material Management Plan, or DMMP,
data). The relatively high concentrations of PCBs in the DMMP sediments represent a
reasonably high conservative estimate for sediments proposed for dredging in NY-NJ Harbor.
Greges (personal communication) and Wisemiller (USACE, unpublished data) provided these
site-specific sediment concentrations. We assumed that sediment PCB concentrations outside
the management area were zero, though alternative approaches are feasible.
The key input parameter for the water column is the freely dissolved concentration of
PCBs in water. This analysis assumes that the dredged sediments and overlying water achieve
equilibrium. For model inputs, we estimated water concentrations of PCBs from sediment
concentrations based on equilibrium partitioning.
Table 1 describes the probability distributions for sediment and water concentrations
derived from individual measured concentrations. In deriving distributions, we considered that
over the long term, humans consuming aquatic organisms would be exposed to average
concentrations in the environment. The mean sediment and water concentrations were thus
characterized by normal distributions for the mean, with uncertainties characterized by the
standard error of the mean. The correlation coefficient between the water and sediment
concentrations was set to 1.0 because the water concentrations were calculated under an
equilibrium assumption.
Size of the Management Site and Attraction Factor. We assumed that the sediments from
these areas are dredged and subsequently placed at a management site offshore of NY-NJ
Harbor. The assumed size of this site is 3.75 square kilometers which is close to the size of a
typical management area (Lutz, USACE, personal communication).
The model uses a differential attraction factor that quantifies the effects of increased
flounder population density at the management site due to disturbance, presence of topographic
7
-------
features, or organic enrichment of sediments at the site. Exposure assessments frequently
recognize such differential attraction as a source of uncertainty, but rarely address it explicitly
other than to assume that a receptor spends all its time within the boundaries of a site. This is a
common approach that can produce large overestimates of risk with large uncertainty.
We define the attraction factor as the ratio of fish abundance within the boundaries of the
management site (number of fish per unit area or catch per unit effort) to the fish abundance
outside the facility boundaries. In the absence of habitat-specific information that describes the
potential differential attraction of flounder to a management site or other disturbed sites, we
assumed that an examination of spatial variation in fish abundance among historical sampling
stations would reflect the range for such an attraction factor. We reviewed the difference in
winter flounder abundance among several sampling stations within their habitat as reported in
the literature. The data on distribution and abundance of adult winter flounder collected during
the NEFSC bottom trawl survey in 1963-1997 (16) shows that fish abundance can vary across
sampling station within the habitat, but the ratio of winter flounder abundance among many
stations did not exceed 10. This ratio is also confirmed by a study of winter flounder abundance
in Narragansett Bay (17). Therefore we assumed that the difference among stations represents
the magnitude of the potential attraction and we used 10 as the estimate for the attraction factor,
but we varied it within the spatial sub-model from 1 to 100 to assess sensitivity of the model to
this uncertain parameter.
Winter Flounder Seasonal Movements and Abundance. Studies of seasonal migration of
winter flounder show that adults live in cooler offshore waters during the summer and then move
to shallower inshore waters in winter and early spring. The extent of offshore-inshore
movements varies geographically. Many tagging studies (18,19,20,21,22) showed that flounders
remain in bays and harbors year-round, moving into deeper waters during the warmest weather.
The model incorporates fish abundance (number of individual fish per unit area) as well
as seasonal changes in population abundance due to seasonal migration between in-shore and
off-shore areas that are characteristic of many fish species. Reported abundance for winter
flounder varies widely. Pearcy (12) reported that the average abundance for juvenile winter
flounder in the Upper Mystic River estuary ranged between 0.1 to 0.01 juvenile fish/m2. The
8
-------
abundance of adult fish is likely to be about 10 times less than this density based on the survival
curve presented in their study. Haedrich and Haedrich (23) report 0.004,0.017,0.027 and 0.013
fish/m2 in June, August, November and May for the Mystic River Estuary. Black and Miller
(24) observed about 0.005 fish/m2 near Lower Argyle, Nova Scotia. We used data from the most
detailed study of winter flounder abundance in Narragansett Bay, Rhode Island (17). This study
reports average monthly abundance for winter flounder at 10 stations within the bay. The
abundance varies from 0.005 fish/m2 in August-October to about 0.02 fish/m2 in January. We
assumed that this range and pattern of seasonal change describes the flounder population in the
NY-NJ area.
Foraging Area and Habitat Size. Foraging area is the spatial area foraged by individuals
over specified time period. The habitat is the area that provides a population of individual species
with food and shelter. The size of the population's habitat (i.e. the area occupied by a fish
population) and foraging area have a strong influence on fish exposure to the site contamination.
If the area over which one defines the local population is large (i.e. fish routinely migrate large
distances over a short time period) spatially localized contamination would not result in
significant fish exposure and, thus, risks. On the other hand, a relatively large, with respect to
habitat size, contaminated site could result in significant exposure to all of a population.
We assumed that the size of the foraging area is an undirected component of fish
movement and is characterized by a dispersion coefficient (25). We use dispersion coefficients
based on tagging experiments in Rhode Island Sound (21) to represent a foraging area. These
estimated dispersion coefficients ranged from 1.74 to 2.85 km per day. This study employs the
average dispersion coefficient of 2.33 km2 or about 250 hectares as typical for a winter flounder
foraging area. We varied the size of the foraging area from 25 to 2500 hectares to study the
sensitivity of the model to this uncertain parameter.
The size of the habitat can be defined based on a biological basis (i.e. the migration area
over which individual fish move over a specified time period) or operationally to reflect a
combination of ecological considerations and risk management judgments.
The size of a winter flounder habitat defined biologically can be very large. Flounder can
travel large distances with an average speed of about 1.44 km/hr, if constantly moving (26). A
9
-------
study of winter flounder movement in New York Bight using tagged fish found that fish can
travel over 40 km in one season; one tagged fish traveled 328km from the tagging site (27). In
another study (28), fish tagged in Barnegat Bay, New Jersey, were recaptured over an area of
5,000 square miles over two years. This large habitat size for winter flounder has been observed
in other geographic areas as well. The average distance traveled by winter flounder around Cape
Cod ranged from 5.7 to 42.2 km across 15 locations where fish were tagged (18).
The above data indicate that the habitat size for a winter flounder population can range
from several hundred to several thousand square kilometers. If the spatial sub-model used such a
large habitat, large areas of relatively clean habitat surrounding a small management site would
dilute the effect of localized contamination from a management facility and risks would always
be minimal. This may be the appropriate approach to use when addressing population level
ecological effects on the flounder. However, when considering the winter flounder as an element
of exposure to humans through ingestion, the definition of local habitat should incorporate that
portion of the flounder habitat over which the human population is likely to obtain fish.
Therefore, to provide a conservative (i.e. human health protective) estimate of risk, we
adopted an operational definition of the sub-population to which local consumers might be
exposed. A conservative estimate for habitat size for this sub-population can be derived based
on the total consumption of winter flounder by recreational anglers likely to consume locally
caught flounder. The average annual catch reported for New Jersey is about 500,000 adult fish
(29). The smallest spatial area required to support the production of this number of fish can be
estimated by dividing the total catch by average fish abundance. Using an average abundance of
0.01individuals/m2 (12,17,23,24) results in a total habitat size of about 50 km2. We used 25,50
and 100 km2 to test the model's sensitivity to the size of winter flounder habitat.
Bioaccumulation Model
We developed a mechanistic, time-varying model based on the approach of Gobas
(30,31). The model predicts PCB accumulation in fish through direct gill uptake of PCBs from
water and dietary consumption of contaminated prey. It relies on solutions for the following
10
-------
differential equations that describe the time-varying uptake of PCBs using time series data for
sediment and surface water PCB concentrations.
— = kl*CWd + fo*Cdie,-(k2 + ke + km + kg)*Cf (1)
dt
where:
ki
= gill uptake rate (L/Kg/d)
Cwd
= freely dissolved concentration in water (ng/L)
kd
= dietary uptake rate (d1)
Cjiet
= concentration in the diet (jj-g/kg)
k2
= gill elimination rate (d"1)
ke
= fecal egestion rate (d"1)
km
= metabolic rate (d"1)
kg
= growth rate (d"1)
Cf
= concentration in fish (fxg/kg)
The model can be run deterministically to predict point estimates of bioaccumulation in
the food web or probabilistically by incorporating distributions for input parameters. These input
parameters include: time series for sediment and water concentrations for PCBs, weight and lipid
content of aquatic organisms, food ingestion rate and body weight of fish, total organic carbon in
sediment, and KoW. The bioaccumulation model uses time series for sediment and water
concentrations based on the output of the spatial exposure model as explained above. Water
concentrations were calculated from sediment concentrations using equilibrium partitioning.
This is conservative assumption since equilibrium is not likely in marine ecosystem. Data from
the literature were used to develop distributions for species-specific input parameters.
11
-------
Model Constants. Several sources provided equations for the rate constants used in the
model(30,31,32).
TOC. We used site-specific measurements for TOC (Wisemiller, USACE, unpublished data)
to derive its probability distribution (Table 1).
Body Weight and Lipid concentrations. Lipid content and weight of fish for a particular age
class (for example, adult fish) varies greatly and thus were treated using probability distributions
derived from lipid concentrations and body weights available in the literature (33). Percent lipid
distributions were specified as triangular for the winter flounder using measurements in the New
York/New Jersey area (34).
Octanol-water partition coefficient (Kow). PCBs are being evaluated as "total" PCBs. "Total
PCBs" represents a mixture of individual congeners, each of which has its own Kow- The KoW
was treated as a model variable with an assigned triangular distribution, where the range is given
by the minimum and maximum KoW for the individual PCB congeners analyzed, and the mode is
estimated as the average of all the congeners in the mixture. The KoW data were obtained from
Mackay et al. (35).
Human Health Exposure and Risk Model
The cancer risk to adults defined as:
n. , CSF *IRf *Cf*ED (2)
Risk =
BW *1000000* AT
where:
Risk =
incremental individual lifetime cancer risk
CSF =
cancer slope factor (mg/kg-day)-l
IRf =
annualized fish ingestion rate (g/day)
cf =
concentration in fish (Mg/kg)
ED =
exposure duration (days)
12
-------
BW = body weight (kg)
AT = averaging time (days)
The non-cancer risk was estimated using the hazard index approach defined as:
IRf *Cf* ED (3)
Hi =
BW*RdD*AT* 10"
where:
HI = toxicity hazard index
RdD = reference dose (mg/kg-day)
BW = body weight (kg), and
106 = unit conversion factor.
Exposure duration. The output of the spatial and bioaccumulation sub-models are
predictions of PCB concentrations in the tissue of an individual fish over time. The averaging
time and number of fish consumed are required to provide input to an estimate of human health
risk from exposure through fish ingestion. An averaging time of 7300 days (i.e. 365 day/yr for
30 yr) was used to characterize lifetime exposure for cancer risk calculations. Annual averages
(365 days) were used in characterizing non-cancer risks.
Fish Ingestion. The USEPA Exposure Factors Handbook (36) provides a distribution for
fish ingestion rates for adult recreational consumption of marine fish in the mid-Atlantic region.
The fish ingestion rate for the average consumer was set at 6.5 grams/day; the rate for a
reasonably maximally exposed individual (RME) was set at 18.9 grams/day (36).
Body weight. Body weight is set to 70 kg. This weight is commonly used in USEPA risk
assessments and is assumed in the derivation of CSFs (USEPA Exposure Factors Handbook, (36)
Toxicity Factors. The cancer slope factor and reference dose are from the Integrated Risk
Information System (37). These values are specified as point estimates following USEPA
guidance (33).
13
-------
RESULTS AND DISCUSSIONS
Figure 2 presents the modeled spatial distribution of fish foraging around the management
site with different degrees of attraction for the facility (AF varies from 1 to 100). Each dot
corresponds to the spatial location on a specific month of one of the 1,000 individual fish
considered in this simulation. For a facility with no differential attraction (top graph in the
figure), the fish are evenly distributed across the habitat. For an AF=10 (middle graph in the
figure), most of the predicted foraging occurs within the management site. For the site with an
AF=100, the model predicts that only about two percent of the fish forage at a significant
distance from the site.
The bars on Figure 3 represent the time-varying exposure point concentrations for three
individual fish over the three year time interval simulated by the spatial sub-model. The high
exposure point concentration corresponds to more frequent exposure to the contaminated site
within the habitat. For example, fish 1 foraged within the site boundaries in August of Year 1
and in August of Year 2, while fish 2 happened to forage within the site boundaries in April of
Year 3. In each plot, the lower monthly averages for the exposure point concentrations
correspond to partial exposure to the site. These partial exposures occurred when fish foraged
only in clean areas for some period , excluding contaminated sediments from the foraging
habitat.
The PCB tissue concentration in fish (simulated by the bioaccumulation sub-model) is
presented as solid lines on Figure 3. The Figure clearly illustrates rapid PCB accumulation when
fish forage in contaminated areas with much slower depuration after leaving the site. For
example, when Fish 3 foraged close to the facility in January-February of Year 1, the PCB
concentration in its tissue slowly decreased to the background level over the remainder of the
year when the fish foraged only over clean sediments.
Figure 4 compares cancer risk and hazard indices for the unrealistic (but often employed)
scenario in which there are no spatial considerations for exposure to three scenarios that
accommodate various spatially explicit assumptions about fish foraging and/or site
14
-------
characteristics. The x-axis presents exposure scenarios with various modeling assumptions for
habitat size, attraction factors, and foraging areas. The y-axis shows cancer risk and hazard
indices resulting from fish consumption by recreational anglers, under the varying assumptions.
Under most regulatory programs, a hazard index exceeding 1 and a cancer risk between 10"4 and
10"6 indicate potential risk. Box and whisker plots represent the distribution of risks
corresponding to each exposure scenario.
For example, consider the leftmost plot in Figure 4. It represents the distribution of risks
under the biologically unrealistic scenario that does not incorporate spatial or temporal aspects of
fish exposure. The plot shows an expected 75% probability of exceeding a hazard index of
about 27, and a 95% probability or less of experiencing a hazard index of about 31 or less.
Risk estimates decrease under the various scenarios that incorporate data on the spatial and
temporal dimensions of flounder biology and the physical characteristics of the management site.
The box and whisker plots demonstrate that increasing habitat size from 24 to 96 km2 decreases
the median hazard index and cancer risk by a factor of three. A site with a high attraction factor
(100) could result in almost 7 times the median cancer risk and hazard index compared to a
facility with no differential attraction (AF=1). The 95th percentile shows a higher sensitivity to
the size of the foraging area for winter flounder. A change from a 25 to 2500 hectare foraging
area decreases the 95th percentile by almost a factor of 7 for the cancer risk and hazard index.
The median risk value changes by a factor of 3.
Risk assessments provide risk managers with estimates to evaluate potential human health
risks associated with exposure to contaminants in dredged sediments. Most often, the risk
assessment defaults to conservative exposure assumptions. For example, food chain models
often use the average concentration in contaminated media without considering the spatial and
temporal behavior of the receptors. USEPA guidance explicitly requires that risk assessments
address uncertainty in the underlying assumptions (33). The pragmatic question facing dredged
material managers is "How confidant can risk managers be that these estimates realistically
represent exposure and risk?"
The present analysis shows that spatial factors of fish behavior (size of foraging area and
habitat) and characteristics of the management site (size and differential attraction) are important
15
-------
components in evaluating realistic exposure and risk to human receptors. We have presented a
model that: 1) is useful in examining these factors, 2) demonstrates the variation in risk estimates
that they engender, and 3) provides a model framework for incorporating realistic assumptions
into risk estimates.
Our analysis assumed that all fish consumption for recreational fishermen comes from a
flounder population whose habitat is conservatively restricted on the basis of local fish catch
statistics. This assumption is conservative because a biologically defined habitat for flounder
would be much larger, resulting in much lower risk estimates. Even under this conservative
assumption, the incorporation of rational (i.e. data driven) parameters in the exposure models
results in significantly lower median health risks as compared to a spatially non-explicit model.
To obtain median risks close to the prediction of the spatially non-explicit case, all spatial
parameters would have to be taken at conservative extremes simultaneously.
It is important to note that the spatially explicit approach does not ignore the possibility that
some individuals may ingest fish that have foraged mostly in the contaminated area. The
advantage of the approach is that it assigns a probability to the occurrence of this scenario. For
example, Figure 4 shows that when the model incorporates a small foraging area (25 hectares),
the 95th percentile for cancer risk and the hazard index are close to the 95th percentile observed in
the no-spatial considerations scenario. The reason is that if the foraging area is small, a fraction
of the fish population will forage exclusively within the site boundaries and thus receive higher
exposure. There is some probability, however small, that some individuals may eat only these
fish.
We also tested the confidence of our model prediction by varying spatially explicit
parameters over a wide range. Our results suggest that 95% estimates for cancer and non-cancer
risk are nearly always lower than the median risk estimate for the non-spatially explicit case.
Scenarios with varying habitat sizes and attraction factors also result in 95th percentile values
lower than in the spatially non-explicit case.
Even though the presented model incorporates simplified assumptions about the nature of
spatial behavior of ecological receptors, it is useful for capturing some of the major components
of an exposure and risk analysis for contaminated sites. Since the current paper illustrates the
16
-------
general framework for a hypothetical case, rigorous model validation cannot be presented here
and is the subject of a subsequent publication. Nevertheless, a qualitative evaluation of the
available concentration measurements for New York - New Jersey harbor estuary supports the
validity of the model.
Figure 5 presents PCB measurements in eel, winter flounder and blue fish collected in six
sampling areas in the New York - New Jersey harbor estuary in fall 1993 or early winter 1994
(38). The six areas have varying sediment contamination by PCBs resulting from industrial
discharges and disposal of waste materials. The three fish species were selected to represent
different foraging strategies: eel is a well-known localized fish spending most of the time
foraging in the same area while the bluefish is known to cover large distances within short time
periods. As discussed in this paper, winter flounder is a residential fish, but its foraging area is
quite large compared to that of eel, however much smaller than that of bluefish. The figure
shows that the average PCB concentration in eel varies over three orders of magnitude among
sampling areas, while the range for bluefish contamination is less than one order of magnitude.
The range of contamination within the same sampling area shows a similar trend: winter flounder
caught within the same general area exhibit quite a wide range of PCB concentrations, while
individual bluefish show similar concentration values1.
These trends in concentration variation can be explained by the fact that fish with small
foraging areas are likely to reflect local sediment contamination. Eel that happen to forage in a
contamination hotspot are likely to be heavily contaminated, while other eel collected in a non-
contaminated area are likely to be clean. Since bluefish forage over large spatial areas as well as
consume fish that forage over extended spatial areas, they are affected by both clean and
contaminated areas. No matter where bluefish are captured, they reflect the average
contamination of a large habitat. Figure 5 shows that winter flounder falls between these two
extremes.
1 Individual eels exhibit a smaller range of PCB concentrations as compared to winter flounders, which seem to
contradict model prediction. The reason is that individual eels were collected in exactly the same location, so
variation among individual fish is small and cannot be compared to variation among individual bluefish and winter
flounder collected at different locations within the same general area.
17
-------
This paper illustrates the use of spatial modeling in risk analysis. If used in a
conservative but realistic fashion, it can more fully inform the decision-making process for
the management of contaminated sediments. The model could be also modified to
incorporate additional complexities and numbers of sites within a habitat, different site
shapes and contamination profiles, and preferential migration of ecological receptors.
However, we note that the ability to use and interpret such models is often limited by the
state of knowledge concerning the spatial behavior of ecological receptors. Nevertheless,
probabilistic treatment of the model parameters coupled with sensitivity analyses should
provide a rigorous basis for making sound environmental decisions.
ACKNOWLEDGEMENTS
The authors wish to thank Mr. Monte Greges and Mr. Bryce Wisemiller of the NY
District office of US ACE for access to the DMMP data sets. Fruitful discussions and paper
review by Ms. K. von Stackelberg, Drs. D. Vorhees, S. Kane Driscoll and Ms. Williams is also
gratefully acknowledged. This study was supported by the US Army Corps of Engineers,
Dredging Operations Environmental Research Program (DOER). Permission was granted by the
Chief of Engineers to publish this material.
LITERATURE CITED
1. Suter, G.W. Ecological Risk Assessment. Lewis Publishers: Ann Arbor, MI, 1993.
2. Clifford, P.A., Barchers, D.E., Ludwig, D.F., Sielken, R.L., Klingensmith, J.S., Graham,
R.V., Banton, M.I.. 1995. An approach to quantifying spatial components of exposure
for ecological risk assessment. Environ. Toxicol. Chem. 14:895-906.
3. Freshman, J.S., Menzie, C.A. 1996. Two wildlife exposure models to assess impacts at
the individual and population levels and the efficacy of remedial actions. Human and
Ecological Risk Assessment. 2(3): 481-496.
18
-------
4. Ak
-------
Fisheries Science Center, Woods Hole, Massachusetts. NOAA Technical Memorandum
NMFS-F/NEC-84. July 1991.
16. Pereira, J. J., Goldberg, R., Ziskowski, J. J., Berrien, P.L., Morse, W.W., Johnson, D.L.
1999. Essential Fish Habitat Source Document: Winter Flounder, Pseudopleuronectes
americanus, Life History and Habitat Characteristics. U.S. Department of Commerce,
National Oceanic and Atmospheric Administration, National Marine Fisheries Service,
Northeast Region, Northeast Fisheries Science Center. NOAA Technical Memorandum
NMFS-NE-138.
17. Oviatt, C. A., Nixon, S.W. 1973. The demersal fish of Narragansett Bay: an analysis of
community structure, distribution and abundance. Estuarine Coastal Mar. Sci. 1: 361-
378.
18. Howe, A. B., Coates, P.G. 1975. Winter flounder movements, growth and mortality off
Massachusetts. Transactions of the American Fisheries Society. 104: 13-29.
19. Azarovitz, T. R. 1982. Winter flounder, Pseudopleuronectes americanus. In: Fish
Distribution MESA New York Bight Monograph No. 15. M. D. Grosslein and T.
Azarovitz, eds. New York Sea Grant Institute, Albany, New York.
20. Pierce, D. E., Howe, A.B. 1977. A further study on winter flounder group identification
off Massachusetts. Transactions of the American Fisheries Society. 106: 131-139.
21. Saila, S. B. 1961. A study of winter flounder movements. Limnology and Oceanography.
6: 292-298.
22. Van Guelpen, L., Davis, C.C. 1979. Seasonal movements of the winter flounder
(Pseudopleuronectes americanus) in two contrasting inshore locations in Newfoundland.
Trans. Am. Fish. Soc. 108(1): 26-37.
23. Haedrich, R. L., Haedrich, S.O. 1974. A seasonal survey of the fishes in the Mystic
River, a polluted estuary in downtown Boston, Massachusetts. Estuarine Coastal Mar.
Sci. 2: 59-73.
24. Black, R, Miller, R.J. 1991. Use of the intertidal zone by fish in Nova Scotia.
Environmental Biology of Fishes. 31:109-121.
25. Jones, R. 1959. A method of analysis of some tagged haddock returns. J. Cons. Int.
Explor. Mer. 25: 58-72.
26. MacDonald, J. S. 1983. Laboratory observations of feeding behavior of the ocean pout
(Macrozoarces americanus) and winter flounder (Pseudopleuronectes americanus) with
20
-------
reference to niche overlap of natural populations. Canadian Journal of Zoology. 61(3):
539-585.
27. Phelan, B. A. 1992. Winter Flounder Movements in the Inner New York Bight.
Transactions of the American Fisheries Society. 121: 777-784.
28. Scarlett, P. G. 1988. Life history investigations of marine fish: occurrence, movements,
food habits and age structure of winter flounder from selected New Jersey estuaries. New
Jersey Department of Environmental Protection, Division of Fish, Game, and Wildlife,
Marine Fisheries Administration, Bureau of Marine Fisheries. Technical Series 88-20.
29. New Jersey Department of Environmental Protection. (NJDEP). 1994. Fish
consumption patterns by New Jersey consumers and anglers. Prepared by New Jersey
Marine Sciences Consortium, Sandy Hook, NJ and New Jersey Department of
Agriculture, Trenton, NJ.
30. Gobas, F.A.P.C. 1993. A Model for Predicting the Bioaccumulation of Hydrophobic
Organic Chemicals in Aquatic Food-Webs: Application to Lake Ontario. Ecological
Modelling. 69, 1-17.
31. Gobas, F.A.P.C., Z'Graggen, M.N., Zhang, X. 1995. Time Response of the Lake Ontario
Ecosystem to Virtual Elimination of PCBs. Environmental Science and Technology. 29,
2038-2046.
32. Burkhard, L.P. 1998. Comparison of Two Models for Predicting Bioaccumulation of
Hydrophobic Organic Chemicals in a Great Lakes Food Web. Environmental Toxicology
and Chemistry. 17, 383-393.
33. United States Environmental Protection Agency, 1989. Risk Assessment Guidance for
Superfund, Volume 1 - Human Health Evaluation Manual, Part A, Interim Final (EPA,
Office of Emergency and Remedial Response, Washington, D.C., EPA/540/1-89/0002.)
34. Northeast Fisheries Science Center (NFSC). 1997. "25th Northeast Regional Stock
Assessment Workshop (25th SAW): Stock Assessment Review Committee (SARC)
Consensus Summary of Assessments," Northeast Fish. Science Center Reference
Document 91-14, 143 p.
35. Mackay, D., Shiu, W.Y., Ma, K.C. Illustrated Handbook of Physical-Chemical
Properties and Environmental Fate for Organic Chemicals. Volume I. Monoaromatic
Hydrocarbons, Chlorobenzenes, and PCBs. Lewis Publishers, Inc.: Chelsea, Michigan,
1992.
21
-------
36. United States Environmental Protection Agency (USEPA), 1997a. Exposure Factors
Handbook, Volume I: General Factors (EPA, Office of Research and Development,
Washington D.C., EPA/600/P-95/002Fa).
37. United States Environmental Protection Agency, 2000. Integrated Risk Information
System Database (IRIS), http://www.epa.gov/iris.
38. Skinner, L.C., Jackling, S.J., Kimber, G., Waldman, J., Shastay Jr., J., Newell, A.J. 1996.
Chemicals in Fish, Shellfish and Crustaceans from the New York - New Jersey Harbor
Estuary: PCB, Organochlorine Pesticides and Mercury. New York State Department of
Environmental Conservation, Division of Fish, Wildlife and Marine Resources,
November 1996.
39. Schrock, M.E., Barrows E.S., Rosman, L.B. 1997. Biota to Sediment Accumulation
Factors for TCDD and TCDF in Worms from 28-Day Bioaccumulation Tests.
Chemosphere. 34, 1333-1339.
40. Lemieux, H., Blier, P.U., Dufresne, F., Desrosiers, G. 1997. Metabolism and Habitat
Competition in the Polychaete Nereis virens. Marine Ecology Progress Series. 156, 151-
156.
41. Briggs, D.E.G., Kear, A.J. 1993. Decay and Preservation of Polychaetes: Taxonomic
Thresholds in Softbodied Organisms. Paleobiology. 19, 107-135.
42. Rosman, L.B. 1999. Personal communication with Igor Linkov of Menzie-Cura&
Associates, Inc. regarding sandworms.
22
-------
CAPTIONS
Figure 1. Modeling Approach
Figure 2. Foraging of Winter Flounder in the vicinity of the management site with different
degrees of attraction. Each dot represents one of 1,000 simulated fish in the population. The top
plot represents a site that is equally attractive as compared to the neighboring areas. Foraging
around more attractive sites (AF=10 and AF=100) are also shown.
Figure 3. PCB bioaccumulation in Winter Flounder. Bars in each graph represent exposure point
concentrations (i.e. average sediment concentration the individual fish exposed to in the
corresponding month) calculated in the spatial sub-model for three random individual fish. Solid
lines present resulting temporal pattern for PCB bioaccumulation.
Figure 4. Hazard index and cancer risk for human consuming winter flounder. Scenarios with
different assumptions for winter flounder spatial behavior are presented. Box and whisker plots
are used to represent the uncertainty in risk estimates for each scenario.
Figure 5. PCB Concentration in fish collected within the New-York-New Jersey Harbor Estuary
by Skinner et al (38). Sampling Stations: 1-Upper Bay, 2-East River, 3-Kills, 4-Jamica Bay, 5-
Lower Bay, 6-NY Bight).
23
-------
Table 1. Input Parameters
Parameter
Distribution
Shape
Mode
Mean
Min
Max
Standard
Deviation
Reference
Facility
Facility Size (km2)
point
3.75
Lutz, personal
communication
Attraction Factor
3 cases
10
1
100
As derived in text
Sandworm
Lipid Content (%)
Triangular
1.2
1.4
1.0
2.0
(39)
(40)
(41)
(42)
Winter Flounder
Body Weight (g)
triangular
263
115
631
(34)
Lipid Content (%)
traingular
1.04
0.33
2.09
(34)
Seasonal
Abundance (#/ha)
point
J 185
F 59
M 74
A 101
M 77
J 143
J 84
A 54
S 47
O 49
N 131
D 116
(17)
Foraging Range
(ha)
250
none
(21)
Habitat Size (km2)
3 cases
48
24
96
estimated
Sediment and
Water
Log-KoW
Triangular
6.74
6.44
5.24
7.36
Measured congeners
Sediment
Concentration
(total PCBs, ng/g
dry wt)
264.1,346.6,
391.8, 475.5,
546.4,364.1,
1256.2,
1657.9
687
normal
1657
175
Wisemiller (USACE,
unpublished data)
Water
Concentration
(total PCBs, ng/1)
Equilibrium
estimate
30.9
normal
6.16
estimates
24
-------
TOC, %
3.84
normal
0.89
Wisemiller (USACE,
unpublished data)
Human Ingestion
Body Weight (kg)
lognormal
71.5
15.0
(36)
Fish Ingestion
(g/day)
lognormal
6.5
8.03
18.9
50th and 95th percentiles
from (36)
Exposure Duration
(days)
uniform
9125
7300
3650
10950
assumed
25
-------
Spatial Characteristics of:
• habitat
• foraging
• management facility
time series for
Spatial
exposure point concentration
Sub-model
Biochemical characteristics of:
• contaminant
• fish
• sediment and water
Characteristics of
population at risk
Bioac c umulation
Sub-model
time series for PCB
concentration in fish
HI
cancer risk
Sub-model
26
-------
AF=
~ t» V / 1 » »4y» V» ~% *
«*!? L>« ~V.F £.V*. <;
i> ;\i 2& £«.»?•*$
i 3000
6000
5000
4000
¦E 3000
>-
2000
1000
0
AF= 10
WW \
~~~~; ~~ ~
^ ~ ~ «
~ ~~ ~~
/ ~ *<5®
IfHSflQfEfV ~ ~ * ~
mhVT
* ~ ^ ^ 4.
w^ -*w +
~~ ~ ~' 4 / ~*
*~ * .~
* ~~~ • *
~ *¥# ~
6000
5000
4000
-i- 3000
>-
2000
1000
0
AF= 100
i i
1000 2000 3000 4000 5000 6000 7000 8000
X (m)
27
-------
I 400 -
* 300
ffl 100
Fish 1
I" I I I I11!*1! I I lMl
I I lnlMl I I
j 3.5
- 3
E
2.5
a
a
2
.c
U)
u.
1.5
c
- 1
<0
00
O
-- 0.5
£L
¦- 0
700
Fish 2
| 400
59 100
3.5
3
E
- 2.5
a.
a.
-- 2
¦C
(0
-- 1.5
LL
C
-- 1
<0
CO
-- 0.5
o
a.
700
O)
i*
600 --
o>
500 -
c
0
F
400 -:
0)
300 -
V)
c
200 -
10
CO
O
100 -
Q.
Fish 3
^ ^ cf5" & ^ ^ d3" ^ ^ ^ o&
3.5
28
-------
95^0-tile
No Spatial Considerations
75%-tiIe
median
5%-tile
Habitat Size (sq. km)
Attraction Factor
Foraging area (ha)
24
48
96
100
10
25
250 2500
1e-4
V)
a>
o
c
CO
O
1e-5
1e-6
No Spatial Considerations
s75%-tile
piedian
25%-tile
5%-tile
Habitat Size (sq. km)
Attraction Factor
Foraging area (ha)
24 48 96 100 10 1 25 250 2500
-------
Wet Weight
8.000
I
£ 3.005
§ 0.705
IE
£ 0.205
.4=
0.065
Eel " • t t T
! Bluefish
i • i i i I
0.015
1 2 3 4 5 6
1 2 3 4 5 6
Sampling Stations
1 2 3 4 5 6
Max
Min
I I 75th %
25th %
~ Median
400.0
E
Q.
Q.
J=
W
il
c
c
o
e
a>
o
c
o
o
m
o
a.
Lipid Normalized
. T ! i-T"!
: 1: iilo
I
Winter Flounder
; i i i L
I ; i .j...
Blue Fish i
1 2 3 4 5 6
1 2 3 4 5 6
Sampling Stations
1 2 3 4 5 6
I Max
Min
I I 75th %
25th %
~ Median
-------
White Papers Submitted by
Remediation Materials Workgroup
(RMW) Members
-------
Paper Submitted by Menzie-Cura
Behalf of the Port Authority
-------
Science of the Total Environment
In Press
Uncertainty and Variability in Risk from Trophic Transfer of Contaminants in Dredged
Sediments
Linkov1*, Igor, von Stackelberg1, Katherine, E., J., Burmistrov1, Dmitriy, Bridges2, Todd, S.
'Menzie-Cura & Associates, Inc. One Courthouse Lane, Suite Two, Chelmsford, MA 01824
United States Army Engineer Research and Development Center, Vicksburg, MS.
* Corresponding Author:
tel: (978)-322-2855
fax: (978)-970-2791
e-mail: ilinkov@menziecura.com
Key Words: Uncertainty, Variability, Probabilistic Risk Assessment, Ecological Risk
Assessment, 2-D Monte-Carlo Simulations, Bioaccumulation
1
-------
Science of the Total Environment
In Press
Abstract
Risks associated with bioaccumulative contaminants must be considered when evaluating
dredged material disposal alternatives. Bioaccumulation of organochlorines and other
contaminants by higher trophic level organisms represents one of the most significant sources of
uncertainty in risk assessment. Both population variability (e.g., true population heterogeneity in
body weight, lipid content, etc.) and uncertainty (e.g., measurement error) in trophic transfer can
lead to large errors in predicted risk values for ecological receptors. This paper describes and
quantitatively evaluates sources of uncertainty and variability in estimating risk to an ecological
receptor (osprey) from trophic transfer of polychlorinated biphenyls (PCBs) in sediments from
the New York - New Jersey (NY-NJ) Harbor. The distribution of toxicity quotients is obtained
using a food chain model for the osprey and specifying distributions for input parameters, which
are disaggregated to represent either uncertainty or variability. PCB concentrations in sediment
and water are treated as predominantly uncertain, whereas lipid content in fish, feeding
preferences, and fish weight are assumed to contribute primarily to population variability in PCB
accumulation. The analysis shows that point estimates of reasonable maximum exposure (RME)
exceed the uncertainty bounds on the 95th percentile of variability. The analysis also shows that
uncertainties in the sediment and water contaminant concentrations contribute more to the range
of risk estimates than does the variability in the population exposure parameters. Separation of
uncertainty and variability in food chain models can help support management decisions
regarding dredged material disposal by providing a quantitative expression of the confidence in
ecological risk estimates. A rationale is provided for the distinction between uncertain and
variable parameters based on management goals and data availability.
2
-------
Science of the Total Environment
In Press
1. Introduction
Separate consideration of uncertainty and variability is important in environmental policy
assessment because each has very different implications for decision making (Thompson and
Graham, 1996; Morgan and Henrion, 1990). Variability (i.e. heterogeneity in nature) is a
population measure, and provides a context for a deterministic point estimate or probabilistic
distribution for central tendency (CTE) or reasonable maximum exposure (RME). Variability
cannot be reduced, only better understood. In contrast, uncertainty pertains to unknown but
measurable quantities. Parameter uncertainty can be reduced by making additional
measurements of the uncertain quantity. Separating uncertainty and variability allows the analyst
to determine the fractile of the population for which a specified dose occurs, as well as the
uncertainty bounds around that dose.
Explicit treatment of uncertainty and variability in risk assessment has recently attracted
significant attention (Kelly and Campbell, 2000). EPA has recommended separate consideration
of uncertainty and variability in ecological and human health risk assessments. The current
version of Volume 3 of the Risk Assessment Guidance for Superfund (RAGS) suggests several
specific modeling approaches, such as two dimensional Monte-Carlo simulation and
microexposure modeling, that can be used to separate consideration of uncertainty and variability
in risk assessment for Superfund sites (USEPA, 1999). However, these approaches are computer-
intensive and their application should be justified.
To study the utility of explicit treatment of uncertainty and variability in environmental
decision making, we conducted a quantitative analysis of uncertainty and variability in risks
associated with disposal of contaminated sediments. Sediment contamination represents the
primary source of environmental risks associated with dredging and disposal, due in part to the
potential for bioaccumulation and biomagnification of sediment contaminants in aquatic food
chains. Accumulation of these contaminants in tissues of aquatic organisms can lead to adverse
effects to the aquatic organisms themselves, as well as to higher trophic level ecological
receptors. PCBs, which are highly lipophilic and hydrophobic, were selected for this analysis
because they are often found in sediments proposed for dredging, are known to biomagnify
3
-------
Science of the Total Environment
In Press
through food chains, and may contribute to risk to humans and ecological receptors consuming
fish and other aquatic organisms. Although our current approach is restricted to PCBs, the
general methodology and conclusions are applicable to a wide range of organic contaminants.
We have developed a model to estimate risk to osprey from feeding on fish exposed to
contaminated dredged materials. This spreadsheet model is based on the approach of Gobas
(1993; Gobas et al., 1995) and relies on steady-state solutions to differential equations that
describe the time-varying uptake of PCBs. Using both sediment and water PCB concentrations,
the model predicts PCB accumulation by ecological receptors through direct gill uptake of PCBs
from water and dietary consumption of contaminated prey. The model can be run in a
deterministic mode, designed to predict point estimates of bioaccumulation in the food web, as
well as in a probabilistic mode, designed to incorporate and predict distributions instead of point
estimates for input parameters and predicted concentrations. These input parameters include:
sediment and water concentrations for PCBs, weight and lipid content of aquatic organisms,
ingestion rate and body weight of osprey, total organic carbon in sediment, dissolved organic
carbon and particulate organic carbon in the water column, and KoW. Our analysis uses site-
specific measurements as well as data from the literature (Greges, personal communication;
Palermo, unpublished data; Iannuzzi et al., 1996; Sample et al., 1996) to develop distributions for
input parameters.
One data set used in this analysis originates from the Kill Van Kull tributary in New Jersey,
and the other derives from a number of sites in the NY-NJ Harbor (collectively referred to as
Dredged Material Management Program, or DMMP, data). Kill Van Kull sediments contain
relatively low concentrations of PCBs, while DMMP sediments contain relatively high
concentrations of PCBs. These two data sets represent a reasonable range for PCB
concentrations commonly observed in sediments proposed for dredging in NY-NJ Harbor.
Our analyses address the question of whether the contribution of uncertainty to the
overall risk is large relative to the contribution of variability in this very specific case and using
these assumptions. If this is the case, and the differences in cost among management alternatives
are high, additional collection and evaluation of information may be warranted before making
management decisions regarding contaminated sediments.
4
-------
Science of the Total Environment
In Press
2. Methods
2.1 Conceptual Model
The conceptual model is designed to represent a predominantly sediment-driven food web.
The predominant food chain exposures originate via sediments rather than water, as is common
at disposal sites contaminated with organochlorines and pesticides. Water concentrations of
PCBs and other lipophilic contaminants are expected to be low over the long-term because of
dilution by mixing processes.
The osprey was chosen to represent the highest trophic level of the aquatic food chain.
Osprey's diet consists of more than 99% fish (Poole, 1989), thus it can be exposed to especially
high levels of bioaccumulative contaminants. Like other raptors, osprey populations are also
extremely sensitive to certain types of bioaccumulative contaminants, and some populations have
experienced rapid and significant declines due to eggshell thinning and other adverse effects
associated with exposure to DDT and organo-chlorine chemicals (Terres, 1995). For these
reasons, ospreys are designated as species of concern in some states.
Selection of fish species was based on several criteria including: 1) abundance in the New
York/New Jersey coastal area, 2) importance in osprey's diet, and 3) representativeness of
particular habitats or trophic levels. Resident species with small foraging areas were specifically
selected because these fish are likely to be more highly exposed to sediments in open-water
disposal sites than species with larger ranges.
In our conceptual model, the polychaete worm Nereis virens (sandworm) represents the base
of the food web. The sandworm is a deposit feeder that burrows and lives in sediment and moves
partially out of its burrow to feed (Wilson and Ruff, 1988). Mummichogs {Fundulus
heteroclitus) which represent the next trophic level above sandworms, feed primarily on
invertebrates in the sediment. Mummichogs have limited home ranges and occupy tidal creeks,
coves and inlets (Iannuzzi et al., 1996; Lotrich, 1975). Summer flounder is a piscivorous fish
which occupies the top of the aquatic food chain. Flounder are typically found on muddy or
sandy bottoms, particularly around structures like rock piles, reefs, and bridge pilings, where
5
-------
Science of the Total Environment
In Press
they prey upon small finfish. The top-level predator in this analysis is the osprey (.Pandion
haliaetus).
Feeding preferences for all of the organisms vary with the age and size of the individual.
Mummichogs typically feed on a number of benthic invertebrates, but for the purposes of this
paper are assumed to feed solely on sandworms. Likewise, summer flounder consumes a variety
of invertebrates and smaller fish, but to simplify the modeling effort, they are assumed to feed
exclusively on mummichogs. In our model, Osprey consume only summer flounder, under the
assumption that summer flounder is representative of a number of fish species in its diet.
The model presented in this paper is designed to estimate body burdens in adult fish because
fish of adult size are more likely to be consumed by osprey than are juveniles (Poole, 1989).
Osprey exposures are estimated using distributions for fish ingestion rate and body weight.
These estimates were taken from the Wildlife Exposure Factors Handbook (USEPA, 1993) and
the literature (Sample et al., 1996). Model parameters are discussed in Section 2.3 below.
2.2 Modeling Framework
The model was developed using the framework first conceptualized by Gobas (1993; Gobas
et al., 1995). The general form of the steady-state solution for the differential equation describing
the change in concentration of PCBs in fish with time is given by:
_ k\*Cwd + kd*Cdiet
Lf — (1)
k 2 + ke + km + kg
where:
ki = gill uptake rate (L/Kg/d)
Cwd = freely dissolved concentration in water (|_ig/L)
kd = dietary uptake rate (d"1)
Cdiet = concentration in the diet (fag/kg)
6
-------
Science of the Total Environment
In Press
k2 = gill elimination rate (d"1)
ke = fecal egestion rate (d"1)
km = metabolic rate (d"1)
kg = growth rate (d"1)
Cf = concentration in fish (ng/kg)
The fish body burden model is then used to predict point estimates or distributions of
concentrations to which osprey are exposed. The risk to osprey is expressed as a toxicity
quotient:
where:
IRf*Cf (2)
BW*TRV*\tf
TQ
= predicted toxicity quotient
TRV
= toxicity reference value (mg/kg-day)
IRf
= fish ingestion rate (g/day)
cf
= concentration in fish (fig/kg)
BW
= body weight (kg), and
106
= unit conversion factor.
Two versions of the model were developed. The first, called the "calculated sandworm
model," uses measured sediment PCBs concentrations from Kill van Kull and DMMP sediments
and estimates concentrations at the base of the foodweb (i.e. in sandworms) under equilibrium
assumptions (Bierman, 1990). The second version, called "measured sandworm model," uses
results from site-specific 28-day bioaccumulation tests in which tissue concentrations in
sandworms exposed to the sediments were measured following exposure (USEPA, 1991). Using
7
-------
Science of the Total Environment
In Press
sandworm concentration data eliminates the need to numerically estimate concentrations at the
base of the food web from measured sediment concentrations.
2.3 Parameterization of the Models
Several sources provided equations for the rate constants (Gobas, 1993; Gobas et al., 1995;
Burkhard, 1998). Input model parameters are represented either as single values or distributions.
The construction of these distributions considers whether a parameter should be characterized as
uncertain (i.e., there is insufficient information concerning a true, but unknown, value), or
variable (i.e., inherent population heterogeneity). This distinction is artificial (Kelly and
Campbell, 2000; USEPA, 1999) as all parameters, to some extent, will exhibit both variability
and uncertainty. For example, lipid content, considered variable, also reflects some uncertainty
attributable to measurement error. However, this source of uncertainty is expected to be low
relative to observed population heterogeneity in lipid values. The same is true for body weight.
This may not necessarily be the case for the fish ingestion rate, but the available data did not
support quantifying the uncertainty in the variability without a high degree of subjectivity.
Any separation between uncertainty and variability is thus artificial and can be made only: (i)
in the context of achieving particular management goals; (ii) when one or the other category will
dominate; and (iii) if data are available to characterize both uncertainty and variability. In this
example, the management goal is to reach a decision within the USEPA/USACE tiered approach
to analyzing the suitability of dredged materials for open-water disposal (USEPA, 1991). To
address this management goal, our analysis considers parameters related to sediment
contamination prior to dredging (such as water and sediment PCB concentration, TOC, etc.) as
uncertain since (i) exposure to sediment concentrations is of regulatory interest—management
decisions will depend on extent of contamination and risk, and, (ii) uncertainty regarding PCB
concentrations in sediment can be reduced by collecting additional samples, or by refining the
spatial extent of dredging (i.e. selecting locations with specified contamination). Parameters
relating to trophic transfer and contributing to population heterogeneity in the uptake of PCBs
are treated as variable (Figure 1). The evaluation as to whether variability or uncertainty or both
8
-------
Science of the Total Environment
In Press
will be important for any particular parameter depends on results from other analyses,
discussions in the literature, and subjective professional judgment. The development of our
distributions is presented below.
2.3.1 Input Parameters Treated as Predominantly Uncertain
Sediment and water concentrations of PCBs, as well as particulate organic carbon (POC),
dissolved organic carbon (DOC) and total organic carbon (TOC), are described as uncertain
parameters. Sediment and water concentrations were obtained from both the Kill van Kull and
the DMMP data sets. Total PCB concentrations are represented as the sum of 26 individual
congeners (8,18,28,44,49, 52, 66, 77, 87,101,105,118,126,128,138,153, 156,169,170,
180,183, 184, 187, 195, 206, and 209) for which measurements were done. This approach may
underestimate true total PCB concentration and is a source of uncertainty not explicitly
quantified in this study.
The key input parameter for sediment is the total PCB concentration in |ig/kg dry weight.
The key input parameter for the water column is the freely dissolved concentration of PCBs in
water. This analysis assumes that once the dredged sediments are placed at disposal location,
sediment and overlying water will come into equilibrium with each other. Thus, water
concentrations of PCBs for model inputs were estimated from sediment concentrations by
assuming equilibrium partitioning with the sediment. After obtaining a total water concentration,
the following submodel was used to estimate the dissolved water concentration of PCBs:
Cwd — * Cww /"j\
(1 + 0.1* DOC * DEoc * Koc + POC * DEoc * Koc (3)
where:
Cwd = truly dissolved concentration in water (ng/L)
DE0C = density of organic carbon (mg OC/mg)
9
-------
Science of the Total Environment
In Press
DOC = dissolved organic carbon (mg/L)
POC = particulate organic carbon (mg/L)
Koc = organic carbon/water partition coefficient (L/mg OC)
Cww = whole water concentration (ng/L)
Individual sediment and water concentrations as well as derived probability distributions are
provided in Table I for the Kill Van Kull data set and in Table II for the DMMP data set. In
deriving distributions, we considered that over the long term, aquatic organisms and ospreys
consuming aquatic organisms will be exposed to average concentrations in the environment. The
mean sediment and water concentrations were thus characterized by normal distributions for the
mean, with uncertainties characterized by the standard error of the mean. Because the water
concentrations were calculated under an equilibrium assumption, the correlation coefficient
between the water and sediment concentrations was set to 1.0.
Table I and Table II list other uncertain parameters used in the model. Distributions for DOC
and POC were obtained from Parsons et al. (1984), based on typical observed concentrations in
coastal waters. DOC and POC have a significant impact on the estimated freely dissolved
concentrations of PCBs in water, so obtaining site-specific measurements for these parameters
would significantly reduce uncertainty in the estimates of dissolved PCB concentrations. TOC
data were not available for the Kill Van Kull data set, so the distribution for the DMMP data set,
which was determined to be a normal distribution, was used for both sites.
2.3.2 Input Parameters Treated as Predominantly Variable
Parameters contributing to population variability include lipid content and weight of aquatic
organisms, KoW, and osprey exposure parameters (Table III). Lipid concentrations and weights
in each of the aquatic species were derived from the available literature. Lipid content and
weight of fish for a particular age class (for example, adult fish) varies greatly and variability is
typically much greater than the uncertainty in the estimates (USEPA, 1989). Percent lipid
distributions were specified as uniform for the mummichog (Iannuzzi et al., 1996), and as
10
-------
Science of the Total Environment
In Press
triangular for the sandworm and the summer flounder (NFSC, 1997). Residence time of the fish
was specified as a uniform distribution and reflects the amount of time per year that a fish spends
in off-shore areas (Lotrich, 1975) and thus can be potentially exposed to the disposal site.
Kow is treated as variable rather than uncertain for several reasons. PCBs are being evaluated
as "total" PCBs. "Total PCBs" represents a mixture of individual congeners, each of which has
its own Kow. Although, theoretically, KoW can be determined in the laboratory, the congener
distribution between exposure media and exposed organisms is typically quite different, with
aquatic organisms containing congeners that are higher in chlorine content than congeners in the
exposure media (US EPA 1999b). The differences in KoW among individual congeners
contribute to the differential uptake of the congeners in the mixture. Therefore, in modeling the
total PCB mixture in this manner, it is impossible to determine the appropriate KoW distribution
because of the variability in partitioning of individual congeners. The variability distribution for
Kow was specified as a triangular distribution, where the range is given by the minimum and
maximum KoW for the individual PCB congeners analyzed, and the mode is estimated as the
average of all the congeners in the mixture. The KoW data were obtained from Mackay et al.
(1992). The KoC was estimated from Morrison et al. (1997) as 0.41 * KoW-
Osprey exposure parameters reflecting variability in the population include body weight, fish
consumption rate, and toxicity reference values (TRVs). In our analysis, TRV for osprey were
modeled as uniform distribution for PCB body burdens at or above which adverse effects may
occur. The body weight distribution was characterized as lognormal based on USEPA (1993)
and Poole (1989). Similarly, the fish ingestion rate was derived from data presented in USEPA
(1993). There were no correlations specified for these parameters as there was no evidence in
the literature for such correlations. A uniform distribution was derived for the TRV values
(Sample et al., 1996).
2.4 Sensitivity Analysis
Following the estimation of toxicity quotients for the two data sets, sensitivity analyses were
run using rank correlation to determine the model's sensitivity to individual parameters. For each
11
-------
Science of the Total Environment
In Press
combination of input parameters, the output of the model was recorded and combinations of all
possible inputs and outputs were plotted against each other to show the influence of the
parameters on the output values. If the Spearman or partial rank regression coefficient is close to
1 or -1 for a specific input model parameter, this parameter significantly influences model
output.
2.5 Software Implementation
The food chain models were constructed using Crystal Ball™ software (Decisioneering).
Crystal Ball™ permits key input variables to be represented by a distribution of values rather
than a single point estimate. The simulations were conducted using Latin Hypercube sampling
and the sensitivity analyses used the algorithms incorporated in the Crystal Ball™ software.
The model uses a nested analysis to evaluate uncertainty and variability separately by selecting
values from the uncertain distributions and then freezing uncertainty values and running the
entire model using the variability distributions for a given number of iterations (Cohen et al.,
1996; Burmaster and Wilson, 1996). This "inner loop-outer loop" process repeats for a number
of uncertainty simulations, describing how the predicted risk distribution varies due to the
uncertain parameters. Figure 2 provides a schematic of this process. First, uncertain variables
are fixed. Then, 1,500 variability iterations are run, holding the uncertain variables constant.
Finally, the process is repeated for 250 simulations of uncertain variables.
To test the hypothesis that this number of iterations is sufficient, we ran a simulation with
10,000 iterations for variability and 500 uncertainty simulations. The cumulative distributions of
toxicity quotients for these two simulations were then compared. Although a small difference in
risks between simulations with lower and higher sample numbers was detected, the Kolmogorov-
Smirnov test (Rosner, 1995) indicated that the difference between these distributions was not
statistically significant. Furthermore, more powerful t-tests were used to compare the mean risks
in these two simulations and no statistically significant changes were detected. All statistical
analyses were conducted with Statistica™ software (StatSoft).
12
-------
Science of the Total Environment
In Press
3. Modeling Results
Figure 3 presents toxicity quotients for the heavily contaminated DMMP sediment. The x-
axis shows the individual fractiles (i.e., reflects the variability in the population estimates), and
the y-axis displays toxicity quotients. Under most regulatory programs, toxicity quotients
exceeding 1 are considered to indicate potential risk. Box and whisker plots are used to
represent the uncertainty in each variability fractile of toxicity quotients. For example, using the
calculated sandworm model (top plot in Figure 3), we expect 90% of the population to have a
95% probability of exceeding a toxicity quotient of 0.45 (the top bar of the box and whisker plot
for the 90th percentile). At the top of the same box, we expect 90% of the population to have a
75% probability or less of experiencing toxicity quotient of 0.35 or less.
Figures 3 and 4 compare probabilistic CTE and RME estimates (i.e. 50 and 95 population
fractiles) with point estimates that are derived using conservative exposure parameters. The CTE
point estimate for the osprey's toxicity quotient is approximately 0.25 for calculated sandworm
PCB concentrations, and approximately 0.18 for measured PCB concentrations. The uncertainty
range for the probabilistic CTE estimates (i.e. 50th population fractile) falls entirely below 1 for
every osprey population fractile. In contrast, the point estimates for RME (3.6 for calculated
sandworm and 2.6 for measured sandworm models) are well above 1. The risk for the 95th
population variability fractile (that is often used as RME estimated using probabilistic modeling)
ranges from 0.25 to 0.7 for the calculated sandworm case and is always less than 1. Figure 3
shows that although the point estimate toxicity quotient for the RME exceeds 1, the uncertainty
bounds for 95th population variability fractile (i.e. probabilistic estimate for RME) fall well
below 1.
The toxicity quotients for the less contaminated Kill van Kull sediment (Figure 4) show a
similar pattern. However, the point RME estimates are significantly lower than the probabilistic
estimates. The point RME estimates for the toxicity quotients are well above the uncertainty
ranges for all fractiles of the exposed osprey populations. Nevertheless, the PCB contamination
at Kill van Kull is low, meaning that even these conservative RME estimates fall entirely below
1 for both measured and calculated sandworm concentration models.
13
-------
Science of the Total Environment
In Press
By plotting each individual input value selected for a particular iteration against the
corresponding hazard quotient, the sensitivity of the output to the input can be estimated using
rank correlation. Table IV presents the results of the ranking exercise. Each input parameter is
ranked according to the correlation of the input value to the output value. In all cases, the TRV
contributes most to the variance in the output distribution. Log KoW and sediment and water
concentrations are also important relative to other model parameters.
4. Discussion and Conclusions
Risk managers are typically provided with point estimates to evaluate potential ecological
risks associated with exposure to contaminants in dredged sediments. A question that arises is,
how confidant can risk managers be that these point estimates realistically represent risk? This
paper clearly illustrate that the use of point estimates may significantly overestimate risk.
Results for the DMMP dataset show that the point estimates for CTE and the 95th percentile
of estimated toxicity quotients fall within the acceptable risk range. The point estimates for
RME, by contrast, are well above 1 showing potential risk to the receptor. In situations when
point estimates shows risk of concern, the USEPA (1999) guidance allows for applying
probabilistic analysis. Here, the probabilistic analysis demonstrates that the uncertainty range for
the 95th population variability percentile in both the calculated and measured sandworm cases is
below 1. Our analysis thus demonstrates that although the RME point estimate may be
considered protective, it may not be representative of the true uncertainty range for the 95th
population fractile of the osprey population at risk. In contrast, for Kill van Kull, neither point
nor probabilistic estimates of the RME show significant risk from exposure to PCB-
contaminated sediments (toxicity quotient < 1).
The results of our simulations provide information about the uncertainty and variability
associated with estimates of contaminant transfer from sediments and water to higher trophic
level organisms (e.g., fish and piscivorous birds). Using sediment data from the NY-NJ Harbor,
the simulation showed that uncertainty in risk estimates increases with increasing fractile of the
exposed osprey population. That is, the error bounds for the 95th fractile were larger than for the
14
-------
Science of the Total Environment
In Press
median. Also, we found that variability was somewhat greater than uncertainty under the
assumptions incorporated in the modeling. The range between the 5th percentile and the 95th
percentile (a measure of variability) was much greater than the uncertainty range within a
particular percentile. That is, the difference between the toxicity quotients for median (50th
population fractile) and 95th population fractiles was greater than the uncertainty in the estimates
for the median toxicity quotient. Such modeling could inform the decision to collect more data to
refine estimates for variable parameters. Collection of such information could be more efficient
in overall uncertainty reduction and thus is more cost effective.
This analysis did not quantitatively assess model uncertainty. We assumed that the models
(both our model of bioaccumulation in aquatic organisms and the USEPA model for estimating
risks) were valid and did not explicitly consider any conceptual model uncertainties. This study
did not evaluate the uncertainty associated with variability (i.e., uncertainty in the variability
distributions). Although uncertainty associated with models and uncertainty about variability
may be important, neither the management goals nor the available data warranted such an
analysis.
The separation of uncertain versus variable parameters within the context of achieving
management goals, together with an evaluation of the individual parameter attributes, allows
decision-makers to determine the impact of uncertainty that can be reduced within the US ACE
tiered approach. There are parameters for which decision-makers can readily reduce uncertainty
within a typical application of the tiered approach to dredged material management (for example,
by collecting more samples, by dividing a given site into reaches and analyzing them separately,
and so on). Although there are additional sources of uncertainty that were not quantified in our
study, these cannot be addressed within the management context of the decision to be made.
Certainly a research program to reduce that source of uncertainty could be designed, but it is
beyond the scope of the specific management objective in this case.
Our sensitivity analysis showed that for this food chain model estimated toxicity quotients
are most sensitive to the osprey TRVs, followed by Log KoW, and sediment and water
concentrations (both received equal rank due to the correlation coefficient of 1). This results are
consistent with both Burkhard (1998) and Ianuzzi et al. (1996) who concluded that the lipid
15
-------
Science of the Total Environment
In Press
content of the exposed organisms and the Kow of the contaminant influence estimates of tissue
concentrations more than other parameters. These results, together with the specific estimated
toxicity quotients in the context of trophic transfer, suggest that given the assumptions presented
here, decision makers are better off obtaining better TRV information than collecting additional
sediment samples (i.e. obtaining more information on concentrations) prior to making a disposal
decision. The framework presented here illustrates the usefulness of separating uncertainty and
variability to provide managers with information needed to refine estimates and reduce overall
uncertainty.
5. Acknowledgements
The authors wish to thank Drs. L. Rosman, J. Link, and K. Able for providing
information on lipid concentration and weight for summer flounder and sandworms. Our special
thanks to Mr. Monte Greges and Mr. Bryce Wisemiller of the NY district office of USACE for
access to KvK and DMMP data sets. Fruitful discussions and paper review by Drs. D. Vorhees
and S. Kane Driscoll and Laurel Williams is also gratefully acknowledged. This study was
supported by the US Army Corps of Engineers, Dredging Operations Environmental Research
Program (DOER). Permission was granted by the Chief of Engineers to publish this material.
6. References
Bierman, V.J. Jr. Equilibrium partitioning and biomagnification of organic chemicals in benthic
animals. Environmental Science and Technology 1990,24:1407-1412.
Briggs, D.E.G. and Kear, A.J. Decay and preservation of polychaetes: taxonomic thresholds in
softbodied organisms. Paleobiology 1993, 19:107-135.
16
-------
Science of the Total Environment
In Press
Burkhard, L.P. Comparison of two models for predicting bioaccumulation of hydrophobic
organic chemicals in a Great Lakes food web. Environmental Toxicology and Chemistry 1998,
17(3):383-393.
Burmaster, D.E. and Wilson, A. M. An introduction to second-order random variables in human
health risk assessments. Human and Ecological Risk Assessment 1996, 12(4):892-919.
Cohen, J.T., Lampson, M.A., and Bowers T.S. The use of two-stage Monte Carlo simulation
techniques to characterize variability and uncertainty in risk analysis. Human and Ecological
Risk Assessment 1996, 2(4):939-971.
Gobas, F.A.P.C. A model for predicting the bioaccumulation of hydrophobic organic chemicals
in aquatic food-webs: application to Lake Ontario. Ecological Modelling 1993, 69:1-17.
Gobas, F.A.P.C., Z'Graggen M.N. and Zhang X. Time response of the Lake Ontario ecosystem to
virtual elimination of PCBs. Environmental Science and Technology. 1995,29(8):2038-2046.
Iannuzzi, T.J., Harrington, N.W., Shear, N.M., Curry, C.L., Carlson-Lynch, H., Henning, M.H.,
Su, S.H. and Rabbe D.E.. Distributions of key exposure factors controlling the uptake of
xenobiotic chemicals in an estuarine food web. Environmental Toxicology and Chemistry 1996,
15(11): 1979-1992.
Kelly, E.J. and Campbell K. Separating variability and uncertainty in environmental risk
assessment- making choices. Human and Ecological Risk Assessment 2000, v.6 pp 1-13.
Lemieux, H., Blier, P.U., Dufresne, F., and Desrosiers, G. Metabolism and habitat competition in
the polychaete Nereis virens. Marine Ecology Progress Series 1997, 156:151-156.
17
-------
Science of the Total Environment
In Press
Lotrich, V.A.. Summer home range and movements of Fundulus heteroclitus (Pisces:
Cyprinodontidae) in a tidal creek. Ecology 1975, 56:191-198.
Mackay, D., Shiu, W.Y and Ma, K.C. Illustrated handbook of physical-chemical properties and
environmental fate for organic chemicals. Volume I. Monoaromatic hydrocarbons,
chlorobenzenes, and PCBs. Chelsea, Michigan: Lewis Publishers, Inc. 1992.
Morgan, M.G., and. Henrion M. Uncertainty: A guide to dealing with quantitative risk and policy
analysis. Cambridge University Press: New York, NY. 1990.
Morrison, H., Gobas F.A.P.C., Lazar R., Whittle D.M., and Haffner G.D. Development and
verification of a benthic/pelagic food web bioaccumulation model for PCB congeners in Western
Lake Erie. Environ. Sci. Technol. 1997,31:3267-3273.
Northeast Fisheries Science Center (NFSC). (1997). 25th Northeast Regional Stock Assessment
Workshop (25th SAW): Stock Assessment Review Committee (SARC) consensus summary of
assessments. Northeast Fish. Sci. Cent. Ref. Doc. 97-14; 143 p.
Parsons, T.R., Takahashi, M., and Hargrave B. Biological oceanographic processes. Pergamon
Press, Oxford, 1984.
Poole, Alan F. Ospreys: A Natural and Unatural History. Cambridge University Press,
Cambridge, England, 1989.
Rosman, L.B. (1999). Personal communication with Igor Linkov of Menzie-Cura & Associates,
Inc. regarding sandworms.
Rosner, B. Fundamentals of Biostatistics. Belmont, MA: Duxbury Press, Wadsworth Publishing
Company, 1995.
18
-------
Science of the Total Environment
In Press
Sample, B.E., Opresko, D.M., Suter II, G.W. (1996). Toxicological Benchmarks for Wildlife:
1996 Revision. ES/ER/TM-86/R3
Schrock, M.E., Barrows, E.S., and Rosman, L.B. Biota to sediment accumulation factors for
TCDD and TCDF in worms from 28-day bioaccumulation tests. Chemosphere 1997, 34:1333-
1339.
Terres, John K. (1995). The Audubon Society Encyclopedia of North American Birds.
Thompson, K.M. and Graham J.D.. Going beyond the single number: Using probabilistic risk
assessment to improve risk management. Human and Ecological Risk Assessment 1996,
2(4): 1008-1034.
United States Environmental Protection Agency and United States Army Corps of Engineers
(1991). Evaluation of Dredged Material Proposed for Ocean Disposal: Testing Manual. EPA-
503/8-91/001, 1991.
United States Environmental Protection Agency (USEPA). (1999) Risk assessment Guidance for
Superfund: Volume 3 - (Part A, Process for Conducting Probabilistic Risk Assessment). Drfat.
Revision No.5. 1999.
United States Environmental Protection Agency (USEPA). (1999b). Phase 2 Review Copy
Baseline Ecological Risk Assessment for the Hudson River Remedial Investigation/Feasibility
Study. Prepared by Menzie-Cura & Associates and TAMS Consultants, Inc. for US EPA.
Appendix K: Examination of Congener Patterns to Determine Exposure. July, 1999.
United States Environmental Protection Agency (USEPA). (1993). Wildlife Exposure Facotrs
Handbook. Vol. I. Office of Research and Development. EPA/600/R-93/187a. December 1993.
19
-------
Science of the Total Environment
In Press
U.S. Environmental Protection Agency (USEPA). (1989). Risk Assessment Guidance for
Superfund, Volume 1 - Human Health Evaluation Manual, Part A, Interim Final. EPA/540/1-
89/0002. Publication 9285.7-01A. Office of Emergency and Remedial Response, Washington,
D.C.
Watanabe, W.O, Ellis, E.P., Ellis, S.C., and Feeley M.W. Progress in controlled maturation and
spawning of summer flounder Paralichthys dentatus broodstock. Journal of the World
Aquaculture Society 1998,29(4):393-403, December.
Wilson, W.H., Jr. and Ruff R.E. Species profiles: Life histories and environmental requirements
of coastal fishes and invertebrates (mid-Atlantic) - Sandworm and bloodworm. FWS/OB2-
82/11.80. US Fish and Wildlife Service, Washington, D.C. 1988.
20
-------
Science of the Total Environment
In Press
Table I. Input Parameters Treated as Predominantly Uncertain for the Kill Van Kull Case
Parameter
Data
Point
Estimate
(mean)
Distribution
Min
Max
Standard
Error
Reference
Sediment
Concentration
(total PCBs, ng/g
dry wt)
8.47
13.55
10.20
46.1
19.9
normal
46.1
8.9
measurements
Water
Concentration
(total PCBs, ng/1)
Equilibrium
estimate
1.54
normal
3.55
0.69
measurements
Sandworm
Concentration
(total PCBs, ng/g)
11.57
24.03
12.44
13.49
15.4
normal
24.03
2.9
measurements
TOC, %
3.84
normal
0.89
measurements
POC, mg/1
Uniform
0.059
Uniform
0.02
0.1
Parsons et al.,
1984
DOC,mg/l
Uniform
1.2
Uniform
0.4
2
Parsons et al.,
1984
21
-------
Science of the Total Environment
In Press
Table II. Input Parameters Treated as Predominantly Uncertain for the DMMP Case
Parameter
Data
Point
Estimate
(mean)
Distribution
Min
Max
Standard
Error
Reference
Sediment
Concentration
(total PCBs, ng/g
dry wt)
264.1,
346.6,
391.8,
475.5,
546.4,
364.1,
1256.2,
1657.9
687
normal
1256.21
175
measurements
Water
Concentration
(total PCBs, ng/1)
Equilibrium
estimate
30.9
normal
6.16
measurements
Sandworm
Concentration
(total PCBs, ng/g)
51.3,72.9,
89.76, 57.1,
70.4,50.6,
106.8
71.3
normal
106.8
7.9
measurements
TOC, %
3.84
normal
0.89
measurements
POC, mg/1
Uniform
0.06
Uniform
0.02
0.1
Parsons et al.,
1984
DOC,mg/l
Uniform
1.2
Uniform
0.4
2
Parsons et al.,
1984
'Sediment concentration of 1256.2 was used in RME calculations. The value of 1657.9 was considered as an
outliers.
22
-------
Science of the Total Environment
In Press
Table III. Input Parameters Treated as Predominantly Variable
Parameter
Distribution
Shape
Mode
Mean
Min
Max
Standard
Deviation
Reference
Kow
Triangular
6.74
6.44
5.24
7.36
Measured congeners
Residence Time
Uniform
0.5
0.4
0.6
Lotrich, 1975
Sandworm
Lipid Content (%)
Triangular
1.2
1.4
1.0
2.0
Schrock et al., 1997
Lemieux et al., 1997
Briggs and Kear, 1993
Rosman, 1999
Mummichog
BW (g)
Truncated
Normal
3.0
0.2
12.0
2.2
Iannuzzi et al., 1996
Lipid Content (%)
Uniform
2.25
1.0
3.5
Iannuzzi et al., 1996
Summer Flounder
Body Weight (g)
Triangular
574
1924
200
5000
Watanabe et al., 1998
NFSC, 1997
Lipid Content (%)
Uniform
0.25
0.72
1.79
NFSC, 1997
Osprey
Body Weight (g)
Triangular
1568
1200
1950
USEPA, 1993; Poole,
1989
Fish Ingestion
(g/day)
Triangular
300
250
350
USEPA, 1993
TRV (mg/kg/day)
Uniform
0.7 |
0.14
3.5
Sample et al., 1996
23
-------
Table IV. Ranking of Contribution to Variance in the Toxicity Quotients
DMMP Ranks
Kill van Kull Ranks
Parameter
Calculated
sandworm
Measured
sandworm
Calculated
sandworm
Measured
sandworm
TRV
-0.78
-0.81
-0.7
-0.77
log Kow
0.32
0.3
0.26
0.33
Total PCBs in Water (ng/L)
0.29
0.24
0.5
0.33
Sediment (ng/g dry wt.)
0.28
0.24
0.5
0.33
Osprey Body Weight
-0.12
-0.12
-0.1
-0.11
TOC
-0.1
0
-0.07
0
Mummichog Lipid
0.1
0.1
0.08
0.09
Summer Flounder Weight
0.09
0.08
0.08
0.09
Fish Ingestion Rate by Osprey
0.07
0.07
0.07
0.07
DOC (mg/L)
-0.07
-0.08
-0.07
-0.06
Summer Flounder Lipid
0.06
0.06
0.06
0.06
Mummichog Weight
-0.03
-0.07
-0.04
-0.02
Sandworm Lipid
0.03
-0.03
0.01
-0.03
POC (mg/L)
-0.03
-0.04
-0.03
-0.03
Measured Sandworm PCB (ug/kg)
NA
0.04
NA
0.11
Residence Time for Summer Flounder
-0.02
-0.01
-0.02
-0.02
-------
Figure Captions
Figure 1. Uncertainty and variability in model parameters.
Figure 2. Diagram of Nested Latin Hypercube Approach
Figure 3. Toxicity Quotient: Results of Two-Dimensional Latin-Hypercube Simulation Using
DMMP Data
Figure 4. Toxicity Quotient: Results of Two-Dimensional Latin-Hypercube Simulations Using
Kill van Kull Data
25
-------
• Body weights
< • Percent lipid
&>
=¦ • TRV
~ • Kow (mixture)
^ • Fish ingestion rate
•Management goals
C -Data availability
§ -DOC
% • POC
§• • TOC
^ • PCBs concentration (sediment, water, sandworms)
26
-------
Select Uncertainty Parameters
(PCB concentrations, TOC, POC, DOC)
Select trophic chain variability parameters
(Kow, weights, and lipids)
# Iterations: >1,000
Specify Risk Distribution
Select osprey exposure variability parameters
(weight, fish ingested)
>
f
Calculate risk
No
o
a
• fH
¦s
•n
£
# Simulations: >250
No
End
-------
3.5
3.0
2.5
c
©
*5
§ 2.0
a
&
% 1.5
o
i-
1.0
0.5
0.0
NP
>9
>8
vO
>9
sg
>8
sP
yO
^5
nP
o"
>8
sO
o^
o"-
0s
0s
0s'
o**
0s*
©**
0s*
o>»
O""
0s-
0s*
0s-
o"»
0s
LO
O
m
O
m
o
If)
O
in
o
u>
O
in
o
m
O
m
O
in
CM
CM
CO
CO
in
m
CD
CO
h-
GO
00
G>
o>
Fractile of Population Distribution (Variability)
3.0
2.5
2.0
c
o
"¦*3
§ 1.5
O
o
| 1.0
0.5
0.0
sO
v$
>9
vP
V?
"^9
>9
yO
sO
yO
vp
sO
Np
vO
sO
-*9
vO
0s
0s-
0s-
0s
0s
©"•
0s
0s"
0s
0s
O""
0s
0s-
0s-
0s
0s*
in
O
m
O
in
O
in
O
in
O
LO
O
in
O
in
O
m
O
in
¦*—
¦*-
CM
CN
CO
CO
^r
m
Uf>
CO
co
h-
h-
CO
00
O)
o>
Fractile of Population Distribution (Variability)
| RME |
j Calculated Sane
iwor
m
cent!
—|
TJ
CD
—!¦
e
Uncertain
95%-tjle
y i i
-
75%-tjle
median
25%-tjle
5%-tiJe
j
I j
|t
j CTE
iL
SlT
— j , . x ; g . ~ ^
T
i i
! RME
; Measured S;
jndv
irornr
rcenl
ile o
fUnc
95%-til
ertainty
75%-til
' i '¦ 1 \
median:
25%-tife ;
5%-tas r j |
;CTE
£
£
[ _ ; _ | :
j.
• i • i • : * i ¦& j •
28
-------
0.3
RME
0.2
c
a
*3
o
3
a
£
o
¦R
o
0.1
0.0
CTE
Calculated Sandworm
Percentile of Uncertainty
-r-i 95%-lile
75%-tBe
median
25%1-tile
-H 5%-tile
w
sO
0s
0s*
0s"
0s*
0s-
0s*
0s"
0s*
0s*
0s-
O
lO
O
lO
o
lO
o
lO
O
in
T-
C\l
OJ
CO
CO
LO
m
¦S ; ¦a'
o
CD
in
CD
o
N.
If)
h-
o
CO
in
00
o m
o> O)
Fractile of Population Distribution (Variability)
0.3
0.2
c
o
a
£
o
X
o
h-
0.1
RME
CTE
0.0 ' i —
Percentile of Uricertajinty
95%-tile
75%
tile
median
¦nr1 25%j-tUe
- -j 5%-iile
Measured Sandworm
s
lO
o
CM
m
CM
sP so
0s 0s
o to
CO CO
o
in
o
m
to
in
o LO
CD CD
o
1^-
lO
h-
o
00
in
CO
o
o>
m
o>
Fractile of Population Distribution (Variability)
29
-------
DRAFT
12/07/01
Risk-Based Management of Contaminated Sediments: Consideration of Spatial and
Temporal Patterns in Exposure Modeling
Linkov1*, Igor, Burmistrov1, Dmitriy, Cura1, Jerome, Bridges2, Todd, S.
'Menzie-Cura & Associates, Inc. One Courthouse Lane, Suite Two, Chelmsford, MA 01824
2United States Army Corps of Engineers, Vicksburg, MS.
* Corresponding Author:
tel: (978)-322-2855
fax: (978)-970-2791
e-mail: ilinkov@menziecura.com
Key Words:
-------
DRAFT
12/07/01
Abstract
This paper addresses interactions among foraging behavior, habitat preferences, site
characteristics, and the spatial distribution of contaminants in developing PCB exposure
estimates for winter flounder at a hypothetical open water dredged material disposal site in the
coastal waters of New York and New Jersey (NY -NJ). The implications of these interactions on
human health risk estimates for local recreational anglers who fish for and eat flounder are
described. The models implemented in this study include a spatial sub-model to account for
spatial and temporal characteristics of fish exposures and a probabilistic adaptation of the Gobas
bioaccumulation model that accounts for temporal variation in concentrations of hydrophobic
contaminants in sediment and water. We estimated the geographic distribution of a winter
flounder sub-population offshore of NY-NJ based on species biology and its vulnerability to
local recreational fishing, the foraging area of individual fish, and their migration patterns. We
incorporated these parameters and an estimate of differential attraction to a management site into
a spatially explicit model to assess the range of exposures within the population. The output of
this modeling effort, flounder PCB tissue concentrations, provided exposure point concentrations
for an estimate of human health risk through ingestion of locally caught flounder. The risks
obtained for the spatially non-explicit case are as much as one order of magnitude higher than
those obtained with explicit consideration of spatial and temporal characteristics of winter
flounder foraging and seasonal migration. This practice of "defaulting" to extremely
conservative estimates for exposure parameters in the face of uncertainty ill serves the decision-
making process for management of contaminated sediments in general and specifically for
disposal of dredged materials. Consideration of realistic spatial and temporal scales in food
chain models can help support sediment management decisions by providing a quantitative
expression of the confidence in risk estimates.
2
-------
DRAFT
12/07/01
INTRODUCTION
Exposure estimates for wildlife in areas containing spatially localized contaminants are
a function of spatial factors such as foraging area, size of the habitat and the distribution of
contamination. Species exhibiting different foraging behavior will may experience.significantly
different chemical exposures from the same site, even if their foraging areas overlap. Currently,
exposure estimates and subsequent human health and ecological risk projections usually assume
a static and continuous exposure of an ecological receptor to a contaminant concentration
described by some descriptive statistic such as the mean or maximum contaminant concentration
in sediment. These assumptions are thus overly conservative and ignore some of the major
advantages offered by risk assessment, the ability to account for site-specific conditions and to
conduct iterative analyses.
The importance of consideration of the spatial extent of site contamination in terrestrial
environment has recently attracted attention of individual researchers as well as regulatory
agencies. Several studies have called for explicit incorporation of habitat sizes and foraging
behavior for terrestrial receptors. Nevertheless spatial consideration in the risk assessment for
aquatic ecosystems has not been widely considered.
This paper proposes framework for spatially explicit risk assessment associated with
contaminated sediments. Many sediments are contaminated as the result of industrial
development and management decision have to be done on their safe use as well as on their
disposal. For example, the United States Army Corps of Engineers (USACE) or their permit
recipients dredge about 400 million cubic yards of sediment annually to maintain 25,000 miles of
navigation channel. About 60 million cubic yards of dredged material, including sediments that
receive urban or agricultural runoff are placed in more than 150 open water sites designated by
the US Environmental Protection Agency (US EPA). The fact that open water management
facilities are geographically restricted invite an ecological exposure analysis that accounts for the
spatial and temporal aspects of a receptor's biology. The proposed framework can be used to
support risk-based decision making in the regulation and management of contaminated sediment.
3
-------
DRAFT
12/07/01
The USACE and USEPA apply a tiered approach (1) to judge the suitability of dredged
materials for open-water disposal. Tiers I and II use existing or easily obtained information and
apply relatively inexpensive, rapid tests to predict environmental effects using site specific
information and sediment chemistry. Tiers III and IV involve biological and chemical
evaluations that require field sampling, laboratory analyses, and risk assessment.
For some sediments and sites, bioaccumulation and biomagnification of hydrophobic
organic contaminants, such as polychlorinated biphenyls (PCBs), may represent the primary
source of environmental risks to the aquatic organisms and their higher-order predators,
including humans. This paper addresses the interactions of various aspects of foraging behavior,
habitat characteristics, and the spatial distribution of contaminants in developing PCB exposure
estimates for winter flounder at a hypothetical open water dredged material disposal site in the
coastal waters of New York and New Jersey (NY-NJ). It then considers the implications of these
interactions on human health risk estimates for local recreational anglers who fish for and eat
those flounder. We also address the advantages of such spatially explicit modeling in
environmental decision making where sediment contamination poses risk to wildlife.
MODELING APPROACH AND PARAMETERS
Conceptual Model
We developed a conceptual model to represent a predominantly sediment-driven food
web that is common at for sites with contaminated sediments. The conceptual model is a simple
food chain in which the contaminant of concern is total PCBs. We selected PCBs, which are
highly lipophilic and hydrophobic, for this analysis because they: 1) are often found in
contaminated sediments; 2) are known to biomagnify through food chains; and, 3) pose risk to
humans and ecological receptors. Although the current analysis addresses only PCBs, the general
methodology and conclusions are applicable to a wide range of organic contaminants. The
exposure media are surface water and sediments. A common polychaete, Nereis virens
(sandworm), represents the prey base for Pseudopleuronectes americanus (winter founder). The
human receptors are recreational anglers eating the flounder.
4
-------
DRAFT
12/07/01
The analysis employs a spatially explicit foraging sub-model (Figure 1) that provides a
time series of sediment and water concentrations that a fish may encounter within its habitat.
The model inputs are information on: seasonal abundance of fish; habitat size for a species, size
and location of the management area within the species habitat; size of the species foraging area;
and, sediment and water concentrations over the management site and in the surrounding areas.
The model also uses a site specific attraction factor that accounts for differential attraction to the
management area. The outputs of the spatial sub-model are combinations of sediment and
surface water concentrations that the fish population may encounter while foraging in this habitat
over time.
The analysis then applies a bioaccumulation sub-model for transfer of PCBs from
sediments and surface water through a fish food chain. This analysis uses sandworms as the base
of the food chain. The sandworm's exposure to sediment contaminants represents a conservative
estimate of invertebrate exposure because it is a deposit feeder that burrows and lives in
sediment and moves only partially out of its burrow to feed (2).
Winter flounder, which represent the next trophic level above sandworms, feed primarily
on invertebrates in the sediment. Their feeding preferences vary with the age and size of the
individual. The adult fish mostly consume annelids, molluscs, and cnidaria. Several investigators
(3,4,5) noted that they are omnivorous, opportunistic feeders and prey upon various sediment
dwelling organisms such as polychaete worms, amphipods and isopods (crustaceans),
pelecypods, and plant material. Steimle and Terranova (6) found that winter flounder from
contaminated and cleaner control areas fed primarily on the tentacular crowns of tube-dwelling
anemones and large polychaete worms. Within this conceptual model, we assumed that the
flounder feed solely on the polychaete Nereis virens. This is a conservative assumption in terms
of potential exposures because the other major component of their diet, anemones, are likely to
be less exposed to sediment than the sandworms.
Winter flounder is a reasonable representative fish species because it: 1) is an important
recreational and commercial species; 2) occurs abundantly in the New York/New Jersey coastal
area; 3) represents a higher order, bottom-feeding predator; and, 4) is a resident species with a
relatively small foraging area. Therefore, it will more frequently encounter localized
5
-------
DRAFT
12/07/01
contaminated sediments than other recreationally obtained species such as bluefish or striped
bass that forage over larger ranges.
The output of the bioaccumulation sub-model is a time series of fish tissue residues of
PCBs. The human risk sub-model (Figure 1) averages these time series and then calculates
human risk due to ingestion of these fish based on accepted human exposure parameters.
Spatial Submodel
The approach for the spatial sub-model is an extension and modification of a prior
method (7). The habitat area is divided into a grid of one meter by one meter cells across the
management area and surrounding habitat. In the current study, all cells within the limits of the
disposal site are assigned a probability distribution for PCB concentration (see Table 1 and
discussion below), while all cells outside the management area are assumed to be free of PCBs,
but other model parametrization (for example, inclusion of background concentration outside the
facility) is possible.
The spatial sub-model calculates exposure point concentrations for fish utilizing the
habitat. At specified time periods (we used a time step of one month because the data for fish
abundance is available only in monthly increments) each individual fish in simulation is modeled
to forage in a different area within the habitat selected at random. The exposure point
concentration for each time step is the average concentration of the cells that a fish encounters
within its foraging area for a specified time period. In the current simulation, monthly time step
are used, but time interval shorter than 1 month can be used.
The outputs of the spatial sub-model are the exposure point concentrations that individual
fish encounter over time. These time series data are used in the bioaccumulation sub-model to
calculate PCB body burden in fish over time. The input parameters for the spatial sub-model
include sediment and water PCB concentrations, size and location of the management site within
the habitat, site attraction factor, seasonal abundance of fish, fish foraging area, and habitat size.
Sediment and Water Concentrations. The key input parameter for sediment is the total
PCB concentration in jag/kg dry weight. The sediment data in the analysis are from several sites
in NY-NJ Harbor (collectively referred to as Dredged Material Management Plan, or DMMP,
6
-------
DRAFT
12/07/01
data). The relatively high concentrations of PCBs in the DMMP sediments represent a
reasonably high conservative estimate for sediments proposed for dredging in NY-NJ Harbor.
Greges (personal communication) and Wisemiller (unpublished data) provided these site-specific
sediment concentrations. We assumed that sediment PCB concentrations outside the
management area to be zero, though alternative approaches are feasible.
The key input parameter for the water column is the freely dissolved concentration of
PCBs in water. This analysis assumes that the dredged sediments and overlying water achieve
equilibrium. For model inputs, we estimated water concentrations of PCBs from sediment
concentrations based on equilibrium partitioning.
Table 1 describes the probability distributions for sediment and water concentrations
derived from individual measured concentrations. In deriving distributions, we considered that
over the long term, humans consuming aquatic organisms would be exposed to average
concentrations in the environment. The mean sediment and water concentrations were thus
characterized by normal distributions for the mean, with uncertainties characterized by the
standard error of the mean. The correlation coefficient between the water and sediment
concentrations was set to 1.0 because the water concentrations were calculated under an
equilibrium assumption.
Size of the Management Site and Attraction Factor. We assumed that the sediments from
these areas are dredged and subsequently placed at a management site offshore of NY-NJ
Harbor. The assumed size of this site is 3.75 square kilometers which is close to the size of a
typical management area (Lutz, personal communication).
The model uses a differential attraction factor that quantifies the effects of increased
flounder population density at the management site due to disturbance, presence of topographic
features, or organic enrichment of sediments at the site. Exposure assessments frequently
recognize such differential attraction as a source of uncertainty, but rarely address it explicitly
other than to assume that a receptor spends all its time within the boundaries of a site. This is a
common approach that can produce large overestimates of risk with large uncertainty.
We define the attraction factor as the ratio of fish abundance within the boundaries of the
management site (number of fish per unit area or catch per unit effort) to the fish abundance
7
-------
DRAFT
12/07/01
outside the facility boundaries. In the absence of habitat-specific information that describes the
potential differential attraction of flounder to a management site or other disturbed sites, we
assumed that an examination of spatial variation in fish abundance among historical sampling
stations would reflect the range for such an attraction factor. We reviewed the difference in
winter flounder abundance among several sampling stations within their habitat as.reported in
the literature. We found that fish abundance can vary across sampling station within the habitat,
but the ratio of winter flounder abundance among many stations did not exceed 10 (8). Therefore
we assumed that the difference among stations represents the magnitude of the potential
attraction and we used 10 as the estimate for the attraction factor, but we varied it within the
spatial sub-model from 1 to 100 to assess sensitivity of the model to this uncertain parameter.
Winter Flounder Seasonal Movements and Abundance. Studies of seasonal migration of
winter flounder show that adults live in cooler offshore waters during the summer and then move
to shallower inshore waters in winter and early spring. The extent of offshore-inshore
movements varies geographically. Many tagging studies (9,10,11,12,13) showed that
flounders remain in bays and harbors year-round, moving into deeper waters during the warmest
weather.
The model incorporates fish abundance (number of individual fish per unit area) as well
as seasonal changes in population abundance due to seasonal migration between in-shore and
off-shore areas that are characteristic of many fish species. Reported abundance for winter
flounder varies widely. Pearcy (3) reported that the average abundance for juvenile winter
-y
flounder in the Upper Mystic River estuary ranged between 0.1 to 0.01 juvenile fish/m . The
abundance of adult fish is likely to be about 10 times less than this density based on the survival
curve presented in their study. Haedrich and Haedrich (14) report 0.004,0.017,0.027 and 0.013
fish/m in June, August, November and May for the Mystic River Estuary. Black and Miller
(15) observed about 0.005 fish/m2 near Lower Argyle, Nova Scotia. We used data from the most
detailed study of winter flounder abundance in Narragansett Bay, Rhode Island (8). This study
reports average monthly abundance for winter flounder at 10 stations within the bay. The
abundance varies from 0.005 fish/m2 in August-October to about 0.02 fish/m2 in January. We
assumed that this range and pattern of seasonal change describes the flounder population in the
8
-------
DRAFT
12/07/01
NY-NJ area.
Foraging Area. We assumed that the size of the foraging area is an undirected component
of fish movement and is characterized by a dispersion coefficient (16). We use dispersion
coefficients based on tagging experiments in Rhode Island Sound (12) to represent a foraging
area. These estimated dispersion coefficients ranged from 0.67 to 1.1 square miles per day. This
study employs the average dispersion coefficient of 0.9 square miles or about 250 hectares as
typical for a winter flounder foraging area. We varied the size of the foraging area from 25 to
2500 hectares to study the sensitivity of the model to this uncertain parameter.
Habitat Size. The size of the population's habitat (i.e. the area occupied by a fish
population) has a strong influence on the fish exposure to the site contamination. If the area over
which one defines the local population is large (i.e. fish routinely migrates large distances over a
short time period) spatially localized contamination would not result in significant fish exposure
and thus risks. On the other hand, a relatively large, with respect to habitat size, contaminated
site could result in significant exposure to all population.
The size of the habitat can be defined based on a biological basis (i.e. the migration area
over which individual fish move over a specified time period) or operationally to reflect a
combination of ecological considerations and risk management judgments.
The size of a winter flounder habitat defined biologically can be very large. Flounder can
travel large distance swith an average speed of about 1.44 km/hr, if constantly moving (18). A
study of winter flounder movement in New York Bight using tagged fish found that fish can
travel over 40 km in one season; one tag fish traveled 328km from the tagging site (19). In
another study (20), fish tagged in Barnegat Bay, New Jersey, were recaptured over an area of
5,000 square miles. Large habitat size for winter flounder has been observed in other geographic
areas as well. The average distance traveled by winter flounder around Cape Cod ranged from
5.7 to 42.2 km across 15 locations where fish were tagged (9).
The above data indicate that the habitat size for a winter flounder population can range
from several hundred to several thousand square kilometers. If the spatial sub-model used such a
large habitat, large areas of relatively clean habitat surrounding a small management site would
dilute the effect of localized contamination from a management facility and risks would always
9
-------
DRAFT
12/07/01
be minimal. This may be the appropriate approach to use when addressing population level
ecological effects on the flounder. However, when considering the winter flounder as an element
of exposure to humans through ingestion, the definition of local habitat should incorporate that
portion of the flounder habitat over which the human population is likely to obtain fish.
Therefore, to provide a conservative (i.e. human health protective) estimate of risk, we
adopted an operational definition of the sub-population to which local consumers might be
exposed. A conservative estimate for habitat size for this sub-population can be derived based
on the total consumption of winter flounder by recreational anglers likely to consume locally
caught flounder. The average annual catch reported for New Jersey is about 500,000 adult fish
(21). The smallest spatial area required to support the production of this number of fish can be
estimated by dividing the total catch by average fish abundance. Using an average abundance of
O.Olindividuals/m results in a total habitat size of about 50 km . We used 25, 50 and 100 km to
test the model's sensitivity to the size of winter flounder habitat.
Bioaccumulation Model
We developed a mechanistic, time-varying model based on the approach of Gobas (22,
23). The model predicts PCB accumulation in fish through direct gill uptake of PCBs from water
and dietary consumption of contaminated prey. It relies on solutions for the following
differential equations that describe the time-varying uptake of PCBs using time series data for
sediment and surface water PCB concentrations.
— = kl* Cwd + kd* Cdie, ~(kl + ke + kn, + kg)* Cf (1)
dt
where:
ki = gill uptake rate (L/Kg/d)
Cwd = freely dissolved concentration in water (ng/L)
ka = dietary uptake rate (d"1)
10
-------
DRAFT
12/07/01
Cdiet
= concentration in the diet (fag/kg)
k2
= gill elimination rate (d"1)
ke
= fecal egestion rate (d"1)
km
= metabolic rate (d"1)
kg
= growth rate (d"1)
Cf
= concentration in fish (|ig/kg)
The model can be run deterministically to predict point estimates of bioaccumulation in
the food web or probabilistically by incorporating distributions for input parameters. These input
parameters include: time series for sediment and water concentrations for PCBs, weight and lipid
content of aquatic organisms, food ingestion rate and body weight of ecological receptors, total
organic carbon in sediment, and KoW. The bioaccumulation model uses time series for sediment
and water concentrations based on the output of the spatial exposure model as explained above.
Water concentrations were calculated from sediment concentrations using equilibrium
partitioning. Data from the literature were used to develop distributions for species-specific
input parameters.
Model Constants. Several sources provided equations for the rate constants used in the
model(l,2,17).
TOC. We used site-specific measurements for TOC to derive its probability distribution
(Table 1).
Body Weight and Lipid concentrations. Lipid content and weight of fish for a particular age
class (for example, adult fish) varies greatly and thus were treated using probability distributions
derived from lipid concentrations and body weights available in the literature (24). Percent lipid
distributions were specified as triangular for the winter flounder using measurements in the New
York/New Jersey area (25).
-------
DRAFT
12/07/01
Octanol-water partition coefficient (Kow). PCBs are being evaluated as "total" PCBs. "Total
PCBs" represents a mixture of individual congeners, each of which has its own Kow- Although,
theoretically, Kow can be determined in the laboratory, the congener distribution between
exposure media and exposed organisms is typically quite different, with aquatic organisms
containing congeners that are higher in chlorine content than congeners in the exposure media
(26). Therefore, the KoW was treated as variable with an assigned triangular distribution, where
the range is given by the minimum and maximum KoW for the individual PCB congeners
analyzed, and the mode is estimated as the average of all the congeners in the mixture. The Kow
data were obtained from Mackay et al. (27).
Human Health Exposure and Risk Model
The cancer risk to adults defined as:
where:
Risk = incremental individual lifetime cancer risk
CSF = cancer slope factor (mg/kg-day)-l
IRf = annualized fish ingestion rate (g/day)
Cf = concentration in fish (p.g/kg)
ED = exposure duration (days)
BW = body weight (kg)
AT = averaging time (days)
The non-cancer risk was estimated using the hazard index approach defined as:
. CSF * IRf *Cf* ED
Risk =
BW *1000000* AT
(2)
IRf * C/ * ED
(3)
12
-------
DRAFT 12/07/01
where:
toxicity hazard index
reference dose (mg/kg-day)
body weight (kg), and
unit conversion factor.
Exposure duration. The output of the spatial and bioaccumulation sub-models are
predictions of PCB concentrations in the tissue of an individual fish over time. The averaging
time and number of fish consumed are required to provide input to an estimate of human health
risk from exposure through fish ingestion. An averaging time of 7300 days (i.e. 365 day/yr for
30 yr) was used to characterize lifetime exposure for cancer risk calculations. Annual averages
(365 days) were used in characterizing non-cancer risks.
Fish Ingestion. The USEPA Exposure Factors Handbook (29) provides a distribution for
fish ingestion rates for adult recreational consumption of marine fish in the mid-Atlantic region.
The fish ingestion rate for the average consumer was set at 6.5 grams/day; the rate for RME
exposed individual was set at 18.9 grams/day.
Body weight. Body weight is set to 70 kg. This weight is commonly used in USEPA risk
assessments and is assumed in the derivation of CSFs (USEPA Exposure Factors Handbook, (29)
Toxicity Factors. The cancer slope factor and reference dose are from the Integrated Risk
Information System (30). These values are specified as point estimates following USEPA
guidance (24).
RESULTS AND DISCUSSIONS
Figure 2 presents the modeled spatial distribution of fish foraging around the management
site with different degrees of attraction for the facility (AF varies from 1 to 100). Each dot
corresponds to the spatial location on a specific month of one of the 1,000 individual fish
13
HI
RdD =
BW =
106 =
-------
DRAFT
12/07/01
considered in this simulation. For a facility with no differential attraction (top graph in the
figure), the fish are evenly distributed across the habitat. For an AF=10 (middle graph in the
figure), most of the predicted foraging occurs within the management site. For the site with an
AF=100, the model predicts that only about two percent of the fish forage at a significant
distance from the site.
The bars on Figure 3 represent the time-varying exposure point concentrations for three
individual fish over the three year time interval simulated by the spatial sub-model. The
increased concentration in sediment corresponds to more frequent exposure to the contaminated
site within the habitat. For example, fish 1 foraged within the site boundaries in August of Year
1 and in August of Year 2, while fish 2 happens to forage within the site boundaries in April of
the Year 3. In each plot, the lower monthly averages for the exposure point concentrations
correspond to partial exposure to the site, when fish foraged only in clean areas for some period.
The PCB tissue concentration in fish (simulated by the bioaccumulation sub-model) is
presented as solid lines on Figure 3. The Figure clearly illustrates rapid PCB accumulation when
fish forage in contaminated areas with much slower depuration after leaving the site. For
example, for Fish 3 that foraged close to the facility in January-February during year one the
PCB concentration in tissue slowly decreased to the background level over the remainder of the
year when the fish foraged only over clean sediments.
Figure 4 compares cancer risk and hazard indices for the unrealistic (but often employed) fish
ingestion scenario with no spatial considerations to three spatially explicit scenarios that
accommodate various assumptions about fish foraging and/or site characteristics. The x-axis
presents exposure scenarios with various modeling assumptions for habitat size, attraction
factors, and foraging areas. The y-axis shows cancer risk and hazard indices resulting from fish
consumption by recreational anglers, under the varying assumptions. Under most regulatory
programs, a hazard index exceeding 1 and a cancer risk between 10"4 and 10"6 indicate potential
risk. Box and whisker plots represent the distribution of risks corresponding to each exposure
scenario.
For example, consider the leftmost plot in Figure 4. It represents the distribution of risks
under the biologically unrealistic scenario that does not incorporate spatial or temporal aspects of
14
-------
DRAFT
12/07/01
fish exposure. The plot shows an expected 75% probability of exceeding a hazard index of
about 27, and a 95% probability or less of experiencing a hazard index of about 31 or less.
Risk estimates decrease under the various scenarios that incorporate data on the spatial and
temporal dimensions of flounder biology and the physical characteristics of the management site.
The box and whisker plots demonstrate that increasing habitat size from 24 to 96 km2 decreases
the median hazard index and cancer risk by a factor of three. A site with a high attraction factor
(100) could result in almost 7 times the median cancer risk and hazard index compared to a
facility with no differential attraction (AF=1). The 95th percentile shows a higher sensitivity to
the size of the foraging area for winter flounder. A change from a 25 to 2500 hectare foraging
area decreases the 95th percentile by almost a factor of 7 for the cancer risk and hazard index.
The median risk value changes by a factor of 3.
Risk assessments provide risk managers with estimates to evaluate potential human health
risks associated with exposure to contaminants in dredged sediments. Most often, the risk
assessment defaults to conservative exposure assumptions. For example, food chain models
often use the average concentration in contaminated media without considering the spatial and
temporal behavior of the receptors. USEPA guidance explicitly requires that risk assessments
address uncertainty in the underlying assumptions. The pragmatic question facing dredged
material managers is "How confidant can risk managers be that these estimates realistically
represent exposure and risk?"
The present analysis shows spatial factors of fish behavior (size of foraging area and habitat)
and characteristics of the management site (size and differential attraction) are important
components in evaluating realistic exposure and risk to human receptors. We have presented a
model that: 1) is useful in examining these factors, 2) demonstrates the variation in risk estimates
that they engender, and 3) provides a model framework for incorporating realistic assumptions
into risk estimates.
Our analysis assumed that all fish consumption for recreational fishermen comes from a
flounder population whose habitat is conservatively restricted on the basis of local fish catch
statistics. This assumption is conservative because a biologically defined habitat for flounder
would be much larger; resulting in much lower risk estimates. Even under this conservative
15
-------
DRAFT
12/07/01
assumption, the incorporation of rational (i.e. data driven) parameters in the exposure models
results in significantly lower median health risks as compared to a spatially non-explicit model.
To obtain median risks close to the prediction of the spatially non-explicit case, all spatial
parameters would have to be taken at conservative extremes simultaneously.
It is important to note that the spatially explicit approach does not ignore the possibility that
some individual may ingest fish that have foraged mostly in the contaminated area. The
advantage of the approach is that it assigns a probability to the occurrence of this scenario. For
example, Figure 4 shows that when the model incorporates a small foraging area (25 hectares),
the 95th percentile for cancer risk and the hazard index are close to the 95th percentile observed in
the no spatial considerations scenario. The reason is that if the foraging area is small, a fraction
of the fish population will forage exclusively within the site boundaries and thus receive higher
exposure. There is some probability, however small, that some individuals may eat only these
fish.
We also tested the confidence of our model prediction by varying spatially explicit
parameters over a wide range. Our results suggest that 95% estimates for cancer and non-cancer
risk are nearly always lower than the median risk estimate for the non-spatially explicit case.
Scenarios with varying habitat sizes and attraction factors result in 95th percentile values lower
than in the spatially non-explicit case because the maximally exposed flounder with relatively
large foraging areas still have some exposure to non-contaminated sediments.
Even though the presented model incorporates simplified assumptions about the nature of
spatial behavior of ecological receptors, it is useful for capturing some of the major components
of an exposure and risk analysis for contaminated sites. If used in a conservative but realistic
fashion, it can more fully inform the decision-making process for the management of
contaminated sediments.
This paper illustrates the use of the spatial modeling in risk analysis. The model could be
also modified to incorporate additional complexities and numbers of sites within a habitat,
different site shapes and contamination profiles, and preferential migration of ecological
receptors. However, we note that the ability to use and interpret such models is often limited by
the state of knowledge concerning the spatial behavior of ecological receptors. Nevertheless,
16
-------
DRAFT
12/07/01
probabilistic treatment of the model parameters coupled with sensitivity analyses should provide
a rigorous basis for making sound environmental decisions.
ACKNOWLEDGEMENTS
The authors wish to thank Mr. Monte Greges and Mr. Bryce Wisemiller of the NY
District office of USACE for access to the DMMP data sets. Fruitful discussions and paper
review by Drs. K. von Stackelberg, D. Vorhees, S. Kane Driscoll and Laurel Williams is also
gratefully acknowledged. This study was supported by the US Army Corps of Engineers,
Dredging Operations Environmental Research Program (DOER). Permission was granted by the
Chief of Engineers to publish this material.
LITERATURE CITED
1. United States Environmental Protection Agency and United States Army Corps of
Engineers, (1991). Evaluation of Dredged Material Proposed for Ocean Disposal:
Testing Manual. (EPA-503/8-91/001.
2. Wilson, W.H. Jr., and R.E. Ruff. (1988). Species Profiles: Life Histories and
Environmental Requirements of Coastal Fishes and Invertebrates (Mid-Atlantic) -
Sandworm and Bloodworm," (FWS/OB2-82/11.80, US Fish and Wildlife Service,
Washington, D.C.
3. Pearcy, W. G. (1962). Ecology of an estuarine population of winter flounder,
Pseudopleuronectes americanus (Walbaum). Bulletin of the Bingham Oceanographic
Collection, Yale University, 18: 1-78.
4. MacPhee, G. K. (1969). Feeding habits of the winter flounder, Pseudopleuronectes
americanus (Walbaum), as shown by stomach content analysis. M. A. Thesis, Boston
Univ., Boston. 66 p.
17
-------
DRAFT
12/07/01
5. Frame, D. W. (1972). Biology of young winter flounder Pseudopleuronectes americanus
(Walbaum): Feeding habits, metabolism and food utilization. Ph. D. Thesis. University of
Massachusetts, Amherst. 109 pp.
6. Steimle, Frank W., and Russell Terranova. (1991). Trophodynamics of select demersal
fishes in the New York Bight. U.S. Department of Commerce, National Oce.anic and
Atmospheric Administration, National Marine Fisheries Service, Northeast Region,
Northeast Fisheries Science Center, Woods Hole, Massachusetts. NOAA Technical
Memorandum NMFS-F/NEC-84. July 1991.
7. Freshman, J.S., C.A. Menzie. (1996). Two wildlife exposure models to assess impacts at
the individual and population levels and the efficacy of remedial actions. Human and
Ecological Risk Assessment. 2(3): 481-496.
8. Oviatt, C. A., and S. W. Nixon. (1973). The demersal fish of Narragansett Bay: an
analysis of community structure, distribution and abundance. Estuarine Coastal Mar. Sci.,
1:361-378.
9. Howe, A. B., and P. G. Coates. (1975). Winter flounder movements, growth and
mortality off Massachusetts. Transactions of the American Fisheries Society, 104: 13-29.
10. Azarovitz, T. R. (1982). Winter flounder, Pseudopleuronectes americanus. In: Fish
Distribution MESA New York Bight Monograph No. 15, M. D. Grosslein and T.
Azarovitz, eds. New York Sea Grant Institute, Albany, New York.
11. Pierce, D. E., and A. B. Howe. (1977). A further study on winter flounder group
identification off Massachusetts. Transactions of the American Fisheries Society, 106:
131-139.
12. Saila, S. B. (1961). A study of winter flounder movements. Limnology and
Oceanography, 6: 292-298.
13. Van Guelpen, L., and C. C. Davis. (1979). Seasonal movements of the winter flounder
(Pseudopleuronectes americanus) in two contrasting inshore locations in Newfoundland.
Trans. Am. Fish. Soc., 108(1): 26-37.
14. Haedrich, R. L., and S. O. Haedrich. (1974). A seasonal survey of the fishes in the Mystic
River, a polluted estuary in downtown Boston, Massachusetts. Estuarine Coastal Mar.
Sci., 2: 59-73.
15. Black, R., and R.J. Miller. (1991). Use of the intertidal zone by fish in Nova Scotia.
Environmental Biology of Fishes, 31: 109-121.
18
-------
DRAFT
12/07/01
16. Jones, R. (1959). A method of analysis of some tagged haddock returns. J. Cons. Int.
Explor. Mer., 25: 58-72.
17. Bigelow, H. B., and W. C. Schroeder. (1953). Fishes of the Gulf of Maine. U.S. Fish and
Wildlife Service Fisheries Bulletin, 74(53): 1-577.
18. MacDonald, J. S. (1983). Laboratory observations of feeding behavior of the ocean pout
(Macrozoarces americanus) and winter flounder (Pseudopleuronectes americanus) with
reference to niche overlap of natural populations. Canadian Journal of Zoology, 61(3):
539-585.
19. Phelan, B. A. (1992). Winter Flounder Movements in the Inner New York Bight.
Transactions of the American Fisheries Society, 121: 777-784.
20. Scarlett, P. G. (1988). Life history investigations of marine fish: occurrence, movements,
food habits and age structure of winter flounder from selected New Jersey estuaries. New
Jersey Department of Environmental Protection, Division of Fish, Game, and Wildlife,
Marine Fisheries Administration, Bureau of Marine Fisheries. Technical Series 88-20.
21. New Jersey Department of Environmental Protection. (NJDEP). (1994). Fish
consumption patterns by New Jersey consumers and anglers. Prepared by New Jersey
Marine Sciences Consortium, Sandy Hook, NJ and New Jersey Department of
Agriculture, Trenton, NJ.
22. Gobas, F.A.P.C. (1993) "A Model for Predicting the Bioaccumulation of Hydrophobic
Organic Chemicals in Aquatic Food-Webs: Application to Lake Ontario," Ecological
Modelling 69, 1-17.
23. Gobas, F.A.P.C., M.N. Z'Graggen and X. Zhang, (1995). Time Response of the Lake
Ontario Ecosystem to Virtual Elimination of PCBs, Environmental Science and Technology
29, 2038-2046.
24. United States Environmental Protection Agency, (1989). Risk Assessment Guidance for
Superfund, Volume 1 — Human Health Evaluation Manual, Part A, Interim Final (EPA,
Office of Emergency and Remedial Response, Washington, D.C., EPA/540/1-89/0002.
25. Northeast Fisheries Science Center (NFSC). (1997). "25th Northeast Regional Stock
Assessment Workshop (25th SAW): Stock Assessment Review Committee (SARC)
Consensus Summary of Assessments," Northeast Fish. Science Center Reference
Document 91-\4, 143 p.
19
-------
DRAFT
12/07/01
26. United States Environmental Protection Agency, (1999). Phase 2 Review Copy Baseline
Ecological Risk Assessment for the Hudson River Remedial Investigation/Feasibility
Study. Prepared by Menzie-Cura & Associates, Inc. and TAMS Consultants, Inc. for US
EPA. Appendix K: Examination of Congener Patterns to Determine Exposure. July,
1999.
27. Mackay, D., W.Y Shiu, and K.C. Ma, (1992).Illustrated Handbook of Physical-Chemical
Properties and Environmental Fate for Organic Chemicals. Volume I. Monoaromatic
Hydrocarbons, Chlorobenzenes, and PCBs (Lewis Publishers, Inc., Chelsea, Michigan.
28. Morrison, H., F.A.P.C. Gobas, R. Lazar, D.M. Whittle, and G.D. Haffner, (1997).
Development and Verification of a Benthic/Pelagic Food Web Bioaccumulation Model for
PCB Congeners in Western Lake Erie, Environmental Science and Technology, 31,3267-
3273.
29. United States Environmental Protection Agency (USEPA), (1997a). Exposure Factors
Handbook, Volume I: General Factors (EPA, Office of Research and Development,
Washington D.C., EPA/600/P-95/002Fa.
30. United States Environmental Protection Agency, (2000). Integrated Risk Information
System Database (IRIS), http://www.epa.gov/iris.
31. Schrock, M.E., E.S. Barrows, and L.B. Rosman, (1997). Biota to Sediment
Accumulation Factors for TCDD and TCDF in Worms from 28-Day Bioaccumulation
Tests," Chemosphere 34, 1333-1339.
32. Lemieux, H., P.U. Blier, F.Dufresne, and G. Desrosiers. (1997). Metabolism and Habitat
Competition in the Polychaete Nereis virens, Marine Ecology Progress Series 156, 151-
156.
33. Briggs, D.E.G., and A.J. Kear. (1993). Decay and Preservation of Polychaetes:
Taxonomic Thresholds in Softbodied Organisms, Paleobiology 19,107-135.
34. Rosman, L.B., (1999). Personal communication with Igor Linkov of Menzie-Cura &
Associates, Inc. regarding sandworms.
20
-------
DRAFT
12/07/01
CAPTIONS
Figure 1. Modeling Approach
Figure 2. Foraging of Winter Flounder in the vicinity of the management site with different
degrees of attraction. Each dot represent one of 1,000 fish in the population. The top plot
represents a site that is equally attractive as compared to the neighboring areas. Foraging around
more attractive sites (AF=10 and AF=100) are also shown
Figure 3. PCB bioaccumulation in Winter Flounder. Bars present the exposure point
concentrations calculated in the spatial sub-model for three random individual fish. Solid lines
present resulting temporal pattern for PCB bioaccumulation.
Figure 4. Hazard index and cancer risk for human consuming winter flounder. Scenarios with
different assumptions for winter flounder spatial behavior are presented.
21
-------
DRAFT
12/07/01
Table 1. Input Parameters
Parameter
Distributio
n Shape
Mode
Mean
Min
Max
Standard
Deviation
Reference
Facility
Facility Size (km2)
point
3.75
Bridges, personal
communication
Attraction Factor
3 cases
10
1
100
As derived in text
Sandworm
Lipid Content (%)
Triangular
1.2
1.4
1.0
2.0
(31)
(32)
(33)
(34)
Winter Flounder
Body Weight (g)
triangular
263
115
631
(25)
Lipid Content (%)
traingular
1.04
0.33
2.09
(25)
Seasonal Abundance
(#/ha)
point
J 185
F 59
M 74
A 101
M77
J 143
J 84
A 54
S 47
O 49
N 131
D 116
Foraging Range (ha)
250
none
Habitat Size (km2)
3 cases
48
24
96
Bridges, personal
communication
Sediment and Water
Log-Kow
Triangular
6.74
6.44
5.24
7.36
Measured congeners
Sediment
Concentration
(total PCBs, ng/g dry
wt)
264.1,
346.6,
391.8,
475.5,
546.4,
364.1,
1256.2,
1657.9
687
normal
1657
175
measurements
Water Concentration
Equilibriu
30.9
normal
6.16
measurements
22
-------
DRAFT
12/07/01
(total PCBs, ng/1)
m estimate
TOC, %
3.84
normal
0.89
measurements
Human Ingestion
Body Weight (kg)
point
70
13
Fish Ingestion (g/day)
point
6.5
18.9
50th and 95th percentiles
from 14
Exposure Duration
(days)
point
7300
10950
assumed
23
-------
DRAFT
12/07/01
Spatial Characteristics of: Biochemical characteristics of:
• habitat • contaminant Characteristics of
• foraging • fish population at risk
sediment and water
management facility
time series for PCB
concentration in fish
time series for
exposure point concentration
Risk
Sub-model
Bioaccumulalion
Sub-model
Spatial
Sub-model
24
-------
DRAFT
12/07/01
£ 3000
AF= 1
~
~ l* <%y ~% ?
AF= 10
-§- 3000
.~ ~
AF=400
¦§¦ 3000
"*i 1 1 i r
1000 2000 3000 4000 5000 6000 7000 8000
X (m)
-------
DRAFT
12/07/01
700
Fish 1
B 100
I I l"l I I I "l"l I I I I'M
I I I I I" I" I I I
H—Ft
- 3
E
-- 2.5
a
a
2
sz
0)
li.
-- 1.5
c
1
w
m
o
-- 0.5
CL
-I 0
Fish 2
o>
5- 500
99 100
0.5 Q.
I I'M I I
on
600 --
a>
3_
500 --
C
d)
E
400
¦a
a>
300 -
(0
c
200 -
(0
CD
O
100 -
Q.
Fish 3
- 3
E
- 2.5
a
a.
2
JZ
w
u.
-- 1.5
c
-- 1
(0
m
o
-- 0.5
a.
^ ^ ^ o* ^ ^ ^ d5" ^ ^ ^ od
_rS
^ _ rS>
-------
35
30
25
20
15
10
95"/o-tile
No Spatial Considerations
Habitat Size (sq. km)
Attarction Factor
Foraging area (ha)
24 48 96
100 10 1
25 250 2500
No Spatial Considerations
1e-4
1e-5
75%-tile
5%-tile
1e-6
Habitat Size (sq. km)
Attarction Factor
Foraging area (ha)
24 48
96
100 10
25 250 2500
-------
Papers Submitted by
URS Corporation on Behalf of
Nation's Port
-------
NPFS01REV4 12/6/00
REVIEW OF PROPOSED CHANGES TO THE EPA/USACE REVIEW
BIO ACCUMULATION TESTING EVALUATION FRAMEWORK (TEF) FOR
DETERMINING THE SUITABILITY OF DREDGED MATERIAL DISPOSAL
AT THE HARS
FACT SHEET No. 1: HARS-Specific PCB Decision Values
Author: Spyros P. Pavlou/URS Corporation - Technical Representative, Nation'sPort
Date: December 6,2000
TEF Revision: Section I.C.5.B, (p. 11-13), Tables 2-5 (p. 14-16), Development of PCB
HARS-Specific Bioaccumulation Decision Values. Calculation of allowable PCB benthic
tissue concentrations for protection of human health: fish ingestion scenario for cancer
and non-cancer risks. Appendix D, Section VI, HARS Benthic Tissue Remediation
Values, Polychlorinated Biphenyls (tPCB).
Evaluation: The approach and associated computations described in the sections of the
Revisions document listed above were reviewed and evaluated to determine reliability of
proposed revisions, defensibility of the computational framework, assumptions used,
equation parameter values, and reproducibility of results. In performing this evaluation
EPA responses to comments provided by earlier peer reviewers (Section II, Responses to
Scientific Peer Review Comments) were also considered.
Identification of Deficiencies: The computational framework used to estimate HARS-
specific PCB values for human health (cancer and non-cancer) and ecological protection
is inconsistent with the state-of-risk-assessment-practice, EPA risk assessment guidance,
and site specific exposure considerations. As currently derived in the 10/19/00 TEF
Revisions Document, the PCB decision values are indefensible. Specific deficiencies
contributing to the indefensibility of the proposed values are as follows:
1. In the TEF Revisions Document, bioaccumulation of PCBs in fish is estimated only
via ingestion of prey (worms). Direct uptake from ambient water (pore or interfacial
water) through equilibrium partitioning is ignored.
2. An area use factor was not incorporated in the calculation of the decision value to
account for the home range of the fish species encountered at the HARS. The
computational framework must be revised accordingly to include a combined
seasonal and area use factor in estimating the decision value.
3. For the species considered to be representative of the fish occupying the HARS, no
evidence was provided to determine the extent to which worms constitute the
dominant (or a significant) component of the individual species' diet. For example,
whiting, bluefish, striped bass, weakfish and sea bass are all upper-water-column
pelagic fish and would not be expected to directly feed on worms at the HARS.
Therefore, any bioaccumulation of PCBs by these pelagic fish species would only
result from consumption of resident demersal species at HARS. Based on a
preliminary review of available literature, the Nation'sPort has classified the species
-------
NPFS01REV4 12/6/00
of concern identified in the TEF Revisions document in two categories: non-worm
feeders (black sea bass, bluefish, striped bass, and weakfish), and worm feeders
(flounders, cod, haddock, and porgy). A decision should be made whether a
significant portion of the PCB burden in these species comes from HARS, hence their
inclusion or exclusion from the combined area and seasonal use factor calculations
should be determined.
4. The mathematical equations and associated parameters for the risk models presented
in the TEF Revisions Document do not follow accepted EPA risk assessment
guidance nomenclature. In fact, standard parameters of the EPA risk model equations
for fish ingestion (including fish PCB uptake from water and ingestion of prey are
missing from the equations used to compute the decision value. These parameters
include:
- averaging time for cancer and non-cancer estimates,
- target hazard quotient appropriate to the HARS specific exposure conditions,
- fraction of fish ingested,
- exposure frequency and duration for cancer and non-cancer estimates,
- PCB-bioconcentration factors (from water) for fish and prey, and
- fraction of prey in fish diet (not all fish diet consists of worms)
5. The selection of many risk model parameter values used in the TEF Revisions
Document is unsupportable by the existing literature and site specific exposure
conditions (e.g., area and seasonal use factors; see comment 3 above). In addition, in
selecting parameter values, no evidence of adherence to the EPA Data Quality
Objectives (DQO) requirements is provided. These requirements are specified under
Order 5360.1 CHG 1, Policy and Program Requirements for the Mandatory Agency-
wide Quality System (EPA, 1998), particularly regarding specification of acceptable
limits on decision errors given the uncertainty inherent in the equation parameters.
Proposed Action for the MRW Technical Committee:
1. The Nation'sPort requests that the Technical Committee of the HARS RMW reviews
the computational framework of the risk model as presented in the 10/19/00 TEF
Revisions Document and reaches a consensus on what would comprise a technically
defensible framework.
2. To initiate the technical discussions, the Nation'sPort requests that the computational
framework presented in Exhibit-1 of this Fact Sheet be discussed at the meeting and
considered as an alternative methodology for estimating PCB decision values.
3. The Nation'sPort requests that the technical committee discusses how EPA's DQO
protocols can be incorporated in the estimation of the chemical-specific decision
values, to establish tolerance limits for the proposed PCB (and potentially other driver
contaminants) HARS-specific decision values.
4. The Nation'sPort requests that the technical committee reaches consensus on a
preferred approach for quantifying an uncertainty range for the PCB decision values
or for establishing upper and lower bounds for these values (see also Item 3 above).
-------
NPFS01REV4 12/6/00
EXHIBIT-1
FACT SHEET NO. 1: PROPOSED RESOLUTION TO DEFICIENCIES
IDENTIFIED IN THE DEVELOPMENT OF HARS-SPECIFIC PCB DECISION
VALUES
Alternative Computational Framework
A revised equation for estimating decision values for PCBs appropriate to the protection
of human health through the fish consumption pathway is proposed as follows:
The basic premise in estimating the allowable residual concentration of PCB in fish for
protection of human health through fish ingestion is that PCBs are accumulated in fish
tissue through two uptake routes: directly from ambient water through equilibrium
partitioning, and bioacumulation through consumption of contaminated prey. The
concentration of PCBs in fish can be expressed mathematically as:
CF(x) = {BCFf*CW[x] + TTF*CP[x]*FR}*ASUF (1)
Assuming equilibrium partitioning between prey and water, CW[x] can be related to
CP[x] through a proportionality constant as: CW[x] = CP[x]/BCFp. Substituting this
expression into equation (1) yields
CF(x) = {BCFf*CP[x]/BCFp + TTF*CP[x]*FR}*ASUF (2)
CF(x) = concentration of PCB in fish tissue at point of exposure x (mg/kg)
CW[x] = average concentration of PCB in water at point x (mg/L)
CP[x] = average concentration of PCB in prey at point x (mg/kg)
BCFp = fish bioconcentration factor of PCB from water (L/kg)
BCFp = prey bioconcentration factor of PCB from water (L/kg)
TTF = trophic transfer factor of PCB from prey to fish (unitless)
FR = Fraction of fish diet consisting of specific prey species (unitless)
ASUF = area/seasonal use factor (unitless); see Exhibit-2
Solving equation (2) for the PCB prey concentration at point of exposure equation (2) can
be rewritten as
CP(x) = (CF(x)/ASUF)*(BCFp/BCFf)* {l/[ 1 +TTF*FR*(BCFp/BCFf)] } (3)
Setting CP(x) and CF(x) as their respective maximum allowable values, MACP and
MACF respectively, equation (3) takes the form of
MACP= (MACF/ASUF)*(BCFp/BCFf)* { 1/[1+TTF*FR*(BCFp/BCFf)]} (4)
For human health protection (cancer effects),
-------
NPFS01REV4 12/6/00
MACFc={(TR*BW*ATc)/(FIR*FI*ED*EF*SF*Q)}* BFR (5)
Substituting into equation (4),
MACPc = (1/ASUF)* {(TR*BW*ATc)/(FIR*FI*ED*EF*SF*0}*
BFR*(BCFP/BCFF)* {l/[ 1 +TTF*FR*(BCFp/BCFf)] } (6)
For human health protection (noncancer effects),
MACFnc= {(RfD*THQ*BW*ATNC)/(FIR*FI*<2*EF*ED)}*BFR (7)
MACPnc = (1/ASUF)* {(RfD*THQ*BW* ATnc)/(FIR*FI*0*EF*ED)} *BFR*
(BCFp/BCFp)* { 1/[1+TTF*FR*(BCFp/BCFf)] } (8)
where,
MACP = Maximum allowable PCB prey concentration (mg/kg)
MACF = Maximum allowable PCB fish concentration (mg/kg)
TR= Target Cancer Risk level (unitless)
RfD = Reference Dose (mg/kg/day)
THQ = Target Hazard Quotient (unitless)
BW = Body weight (kg)
ATc = Averaging Time for Carcinogenic Effects (days)
ATnc = Averaging Time for Noncarcinogenic Effects (days)
FIR = Fish Ingestion Rate (g/day)
FI = Fraction of fish ingested
EF = Exposure Frequency (days/year)
ED = Exposure Duration (years)
SF = Cancer Slope Factor (mg/kg/day)
BFR = body to fillet ratio for fish (unitless)
Q = Conversion Factor (kg/g)
Equations (7) and (8) are proposed as alternatives to the equations presented in the TEF
Revisions Document. Using similar nomenclature as above, the TEF equations are shown
below
MACPc = (1/SUF)* {(TR*BW*)/(FIR*SF*Q)} *BFR*(1/TTF) (9)
MACPNc = (1/SUF)* {(RfD*BW)/(FIR)*Q} *BFR*(1/TTF) (10)
where, SUF is the seasonal use factor and all other parameters are as previously defined.
The Nations'Port believes that equations (7) and (8) comprise a preferred computational
framework for estimating HARS-specific PCB decision values, and requests that the
RMW technical committee considers implementing this framework in the Final TEF
Document.
-------
NPFS02REV3 12/6/00
REVIEW OF PROPOSED CHANGES TO THE EPA/USACE REVIEW
BIO ACCUMULATION TESTING EVALUATION FRAMEWORK (TEF) FOR
DETERMINING THE SUITABILITY OF DREDGED MATERIAL DISPOSAL AT THE
HARS
FACT SHEET No. 2: Incorporation of Fish Life History in Estimating the Allowable PCB
Concentration in Fish Tissue (Consideration of Area and Seasonal Use Factors)
Author(s): Spyros P. Pavlou, Burt K. Shephard/URS Corporation - Technical Representative,
Nation'sPort
Date: December 6, 2000
TEF Revision: HARS-specific Value Calculations for Protection of Human Health from Cancer
and Non-cancer Effects (Calculation of HARS-Specific Values, Equation 2-Concentration in
Fish, Figures 3 and 4, page 15. Agency's response to reviewers' comments regarding exposure
assumptions, Response 15-1, p. 122-124; Table 15-3, p.124
Evaluation: The Nation'sPort has identified and reviewed available publications documenting
long-range seasonal migration and nearshore/offshore spawning patterns for resident fish in the
New York Bight. These publications include some species of concern listed in the 10/19/00 TEF
Revisions Document (e.g., striped bass, bluefish, weakfish, winter flounder, summer flounder,
Atlantic cod). They also include a number of Essential Fish Habitat source documents published
by the NMFS, and a series of Species Profiles (Life Histories and Environmental Requirements
of Coastal Fishes and Invertebrates) published by the USFWS. The substantial body of seasonal
use data contained in these reports could be used to refine the seasonal residence values
presented in TEF Revisions Document.
Identification of Deficiencies: The TEF Revisions Document does not account for the
occupancy of the HARS by individual fish species relative to their respective home range. The
site use factor value of 77.7 used in Equation 2 (Figures 3 and 4, p.15) is inappropriate. The site
use factor is defined as the sum of the product of a species specific seasonal residence at HARS
by its contribution to the HARS diet of a hypothetical fisher consuming only fish caught within
the HARS boundaries (see p. 124, Table 15-3). By limiting the site use factor to represent only
seasonal movement, fish mobility within their individual home range is not accounted for, thus
resulting in an overestimate of the HARS occupancy and, in-turn, in an overestimate of the
amount of PCBs bioaccumulated in fish within the HARS. This deficiency can be corrected by
incorporating the home range (i.e., the size of the entire habitat occupied by a species) in the
calculation of the site use factor. The area of the HARS (15.7 mi2) divided by the home range of
the fish would provide an appropriate area use factor (AUF) for the species. Multiplying the
seasonal residence factor (SUF) by the AUF yields a better representation of the proportion of the
time during which a resident fish would be exposed to PCBs at the HARS. In summary, failure
to incorporate area use factors in the estimation of allowable residuals in prey (worm)
erroneously overestimates the proportion of the total body burden of PCBs accumulated in fish.
-------
NPFS02REV3 12/6/00
Proposed Action for the MRW Technical Committee:
1. The Nation'sPort requests that the Technical Committee reexamines the defensibility of the
SUF for the fish species presented in Table 15-3 of the TEF Revisions document (Response
to Comment 15-1, p. 122-124) in light of the deficiencies outlined above. The Nation'sPort
also requests that the Technical Committee members coordinate their work to achieve a
consensus in selecting scientifically defensible values for the SUF and AUF parameters that
should be incorporated in the computation of the HARS-specific PCB Decision Values.
2. To facilitate the initiation of these technical evaluations, the Nation'sPort is providing a
preliminary list of references (Exhibit 1) that include species-specific life history and
movement directly relevant to their projected occupancy of the HARS. The Nation'sPort also
requests that in addition to the references listed in Exhibit-1, the Technical Committee
identifies other sources of information available through the open scientific literature. Further
more, reports published by the Atlantic States Marine Fisheries Commission (ASFMC), the
New York State Department of Environmental Conservation (NYSDEC), the New Jersey
Department of Environmental Protection (NJDEP), the Hudson River Foundation, and
regional academic institutions are also recommended as sources of relevant information to be
reviewed.
3. The Nation'sPort performed preliminary computations to demonstrate how the available fish
life history data can be used to enhance the computation of the decision values. These test
computations are presented in Exhibit 2 using home range and dispersion data limited to
flounder, bluefish and striped bass. The Nation'sPort requests that the Technical Committee
reviews the computations as part of their effort to strengthen the defensibility of the TEF
computational framework.
-------
NPFS02REV3 12/6/00
Exhibit 1 - Fish Life History References
Information on the life histories of the ten fish species listed in the TEF Revisions Document
includes diet composition, food habits, trophic status, and daily and seasonal movements. This
information was obtained from several sources of which the Nation'sPort reviewed the
following.
- The Marine and Coastal Species Information System (MACSIS), a computerized database
located at Virginia Tech which holds species accounts for 450 vertebrates and invertebrates
- A series of Essential Fish Habitat Source Documents prepared by the National Marine
Fisheries Service (NMFS)
- A series of Species Profiles prepared by the U.S. Fish and Wildlife Service (USFWS)
- The Fishery Management Plans of the Atlantic States Marine Fisheries Commission
(ASMFC).
The MACSIS database is publicly available on the Internet at:
http://fwie.fw.vt.edu/www/macsis/index.htm
The information in MACSIS has all been compiled from reviews of the primary fisheries
literature. The most accurate information on the diet composition, feeding habits, seasonal
migration patterns and home range of fish species is only attainable through a review of the
primary literature cited in MACSIS, the U.S. Fish and Wildlife Service Species Profiles, and the
NOAA/NMFS Essential Fish Habitat Documents.
Example citations for the USFWS Species Profiles, the NOAA/NMFS Essential Fish Habitat
documents, and the ASMFC Fishery Management Plans are provided at the end of this section.
The individual fish species report citations from USFWS, NOAA/NMFS and ASMFC vary only
in the species discussed in the reports, the document number, the author(s) and the date of issue.
All of the USFWS, NOAA/NMFS and ASMFC reports are available on the Internet at:
USFWS: http://www.nwrc.gov/publications/specindex.html
NOAA/NMFS: http://www.wh.whoi.edu/nefsc/habitat/efh/
ASMFC: http://www.asmfc.org/serv02.htm
The ASMFC management plans make only limited references to the primary literature.
These various sources provide a starting point from which the feeding habits, home ranges and
seasonal migrations of the various fish species can be obtained. In some cases, these documents
summarize information on feeding habits, seasonal migration and home range. In most cases,
however, the primary fisheries literature will have to be consulted before seasonal and area use
factors can be derived. Before deriving seasonal use factors and area use factors, the
Nation'sPort recommends that the original primary fisheries literature be obtained and reviewed
to allow for an accurate compilation of seasonal and area use factorsSeveral examples of the type
of readily available information in these reports are provided below.
Auster (1989, p. 7) describes the daily feeding range of tautog (blackfish) to be on the order of
-------
NPFS02REV3 12/6/00
several hundred meters. This value could be used as a home range for tautog during the portion
of the year that this species is not migrating. Steimle et al. (1999, p. 3) gives a brief description
of the food habits of scup (porgy), including their consumption of polychaetes. The summer
flounder fact sheet prepared by ASMFC in conjunction with their summer flounder Fishery
Management Plan (Beal 1998) summarizes their seasonal inshore and offshore migration
patterns. Evaluation of the literature citations within these summary reports would provide
refined estimates of seasonal and area use factors at the HARS. The table below provides a
summary of literature sources that provide references to the original literature from which
seasonal and area use factors could be derived.
Fish species
MACSIS
USFWS
NOAA/NMFS
ASMFC
Flounder
X
X
X
X
Cod
X
X
Whiting (silver hake)
X
X
Bluefish
X
X
X
X
Striped bass
X
X
X
Haddock
X
X
Porgy (Scup)
X
X
X
Blackfish (tautog)
X
X
X
Weakfish
X
X
X
Sea bass (black)
X
X
X
X
Literature Cited
Auster, P.J. 1989. Species Profiles: Life Histories and Environmental Requirements of Coastal
Fishes and Invertebrates (North Atlantic and Mid-Atlantic). Biological Report 82(11.105), Fish
and Wildlife Service, National Wetlands Research Center, Washington, D.C. August 1989.
Beal, R. 1998. 1998 Review of the Atlantic States Marine Fisheries Commission Fishery
Management Plan for Summer Flounder (Paralichthys dentatus). Atlantic States Marine
Fisheries Commission, Washington, D.C.
Steimle, F.W., C.A. Zetlin, P.L. Berrien, D.L. Johnson and S. Chang. 1999. Essential Fish
Habitat Source Document: Scup, Stenotomus chrysops, Life History and Habitat Characteristics.
NOAA Technical Memorandum NMFS-NE-149, National Oceanic and Atmospheric
Administration, National Marine Fisheries Service, Woods Hole, MA.
-------
NPFS02REV3 12/6/00
Exhibit 2 - Estimation of Site Specific Area and Seasonal Use Factors for Selected Fish
Species at HARS
Using the home range for winter flounder and striped bass reported by the USFWS (1989) and
Waldman et al.(1990), a composite area/seasonal use factor (ASUF) was computed. The AUF
was calculated by assuming that the movement of a fish (3 miles in the case of winter flounder)
is the radius of a circle of area Ttr2. The area of the HARS (15.7 mi2) was then divided by the
calculated home range (28.3 mi2), to estimate an AUF value of 0.56. For the purpose of these test
calculations, species (other than striped bass) were assumed to have 100% occupancy at the
HARS (i.e., AUF=1). Since the computations in this exhibit are intended primarily for illustrative
purposes, a more extensive review of the fisheries literature than what is provided in Exhibit 1
would be required in order to provide refined use factors for each species. However, it is the
reviewers' opinion that further examination of the literature will support values <100% for some
species. The preliminary ASUF values computed from this limited literature review are
summarized in the Table below. The Table is the revised Table 15-3 of the 10/19/00 TEF
Revisions Document. The ASUF for each species is estimated as the product of the values shown
in columns (A) through (B). The cumulative ASUF is the sum of all entries in column (D).
S£L
(B)
(C)
Fish
Contribution
to HARS diet
(%)
Seasonal
residence at
HARS (%)
Area Use
Factor
(%)
Seasonal residence
weighted by
contribution to HARS
diet (%)
TEF
New
TEF
New
New
TEF
New
Flounders (all spp.)
48.6
48.6
75
50a
56e
36.49
13.6
Cod
10.8
10.8
100
50b
100
10.81
5.4
Whiting
2.7
2.7
75
75
100
2.03
2.03
Bluefish
10.8
10.8
75
50°
100
8.11
5.4
Striped bass
10.8
10.8
75
25d
20d
8.11
0.54
Haddock
2.7
2.7
100
100
100
2.7
2.7
Porgy
2.7
2.7
50
50
100
1.35
1.35
Blackfish
2.7
2.7
75
75
100
2.03
2.03
Weakfish
2.7
2.7
75
75
100
2.03
2.03
Sea bass
5.4
5.4
75
75
100
4.05
4.05
77.7
39.13
-------
NPFS02REV3 12/6/00
a/ National Oceanic and Atmospheric Administration. 1999a. Essential Fish Habitat Source
Document: Winter Flounder, Pseudopleuronectes americanus. Life History and Habitat
Characteristics. NOAA Technical Memorandum NMFS-NE-138, Northeast Fisheries Science
Center, Woods Hole, MA.
b/ National Oceanic and Atmospheric Administration. 1999b. Essential Fish Habitat Source
Document: Atlantic Cod, Gadus morhua. Life History and Habitat Characteristics. NOAA
Technical Memorandum NMFS-NE-124, Northeast Fisheries Science Center, Woods Hole, MA.
c/ National Oceanic and Atmospheric Administration. 1999c. Essential Fish Habitat Source
Document: Bluefish, Pomatomus saltatrix. Life History and Habitat Characteristics. NOAA
Technical Memorandum NMFS-NE-144, Northeast Fisheries Science Center, Woods Hole, MA.
d/ Waldman, J.R., D.J. Dunning, Q.E. Ross and M.T. Mattson. 1990. Range dynamics of
Hudson River striped bass along the Atlantic coast. Trans. Amer. Fish. Soc. 119:910-919.
e/ Buckley, J. 1989. Species Profiles: Life Histories and Environmental Requirements of
Coastal Fishes and Invertebrates (North Atlantic): Winter Flounder. U.S. Fish and Wildlife
Service Biological Report 82(11.87).
-------
NPFS03REV2 12/7/00
REVIEW OF PROPOSED CHANGES TO THE EPA/USACE REVIEW
BIO ACCUMULATION TESTING EVALUATION FRAMEWORK (TEF) FOR
DETERMINING THE SUITABILITY OF DREDGED MATERIAL DISPOSAL
AT THE HARS
FACT SHEET NO. 3: Maximum Allowable Dioxin (2,3,7,8-TCDD) Residues in
Worm Tissue
Author(s): Burt K. Shephard, Spyros P. Pavlou/URS Corporation - Technical
Representative, Nation'sPort
Date: December 7,2000
TEF Revision: Section I.C.5.A, Consideration of Dioxins, (p. 10-11); EPA response to
review comment 5-1. EPA proposes to maintain the current dioxin regional value of 1
part per trillion (ppt) in worm tissue for protection of ecological receptors
Evaluation: The justification provided by EPA for maintaining thelppt dioxin value was
reviewed. The Nation'sPort examined the dioxin residue-effects data from laboratory
studies where aquatic species were exposed only to 2,3,7,8-TCDD. These studies
quantified the adverse effects on survival, growth and reproduction associated with
various dioxin residues. The Nation'sPort preformed the review by using the Tissue
Residue Effects Association Database (TREAD). This database comprises the source for
nearly all literature citations in the USACE ERED database.
Identification of Deficiencies: The justification for maintaining the lppt dioxin value
provided in the TEF Revisions Document has not demonstrated that an objective and
thorough review of the literature was conducted. Dioxin residues of about 20 ppt, have
been associated with induction of various enzymes (Fisk et al. 1997). However, available
data on dioxin residues in aquatic biota tissues suggests that adverse effects on survival,
reproduction and growth of aquatic species are typically associated with residue values of
approximately 50 ppt (USEPA 1993). This level is substantially higher than the
maximum permissible dioxin concentration of 1 ppt in worm tissue proposed in the TEF
Revisions document. Therefore even if one assumes a trophic transfer factor >1, the 1 ppt
threshold appears to be overly conservative as a decision value for the suitability of open
water disposal.
Proposed Action for the MRW Technical Committee:
The Nation'sPort requests that the Technical Committee compiles and examines the
available tissue residue-effects literature to establish a dioxin threshold value in worm
tissue for protection of ecological receptors at HARS. This effort may require the
development of a dioxin trophic transfer factor as well as identification of the lowest
residue associated with a 50% change in survival, growth or reproduction.
-------
NPFS07: Clarification on Data Sources Used by the Nation'sPort to Estimate the Areal/Seasonal Use
Factor for HARS
REVIEW OF PROPOSED CHANGES TO THE EPA/USACE REVIEW
BIOACCUMULATION TESTING EVALUATION FRAMEWORK (TEF) FOR
DETERMINING THE SUITABILITY OF DREDGED MATERIAL DISPOSAL
AT THE HARS
FACT SHEET No. 7: Clarification on Data Sources Used by the Nation'sPort to
Estimate the Areal/Seasonal Use Factor for HARS
Author(s): Burt K. Shephard, Spyros P. Pavlou/URS Corporation - Technical
Representative, Nation'sPort
Date: January 16,2001
Revisions to NPFS02, ASUF Calculations: This Fact Sheet summarizes the
Nation'sPort response to the comments received by Andrew Draxler, NOAA NMFS
during the RMW Technical Meeting, January 10, 2001. The responses were prepared by
B. Shephard and communicated verbally by S. Pavlou to Mr. Draxler and meeting
participants. The response also reflects a phone discussion between Pavlou, Shephard,
and Mark Reiss (USEPA) on the same topic a few days prior to the meeting.
It is the Nation'sPort's opinion that a key issue should be addressed and resolved among
the RMW Technical Committee members, and potentially the USEPA's peer reviewers.
Namely, whether population movements represented by a "typical" fish of each species or
individuals of each fish species are considered to be the most appropriate basis for
deriving the ASUF. The Nation'sPort acknowledges that it is possible to find at least a
few individuals of each species at HARS most of the year, if not 100% of the time.
In estimating the ASUF the Nation'sPort has adopted a population based approach. It
should be noted here that an individual fish from just about any NY Bight resident
species could show up at any time at any location, so by taking a single fish approach,
one might as well assign a residence factor of 100% at the HARS. The Data Sources for
the estimation of the ASUF were obtained from the references cited in the Nation'sPort
Fact Sheet NO. 2 (Document Reference, NPFS02REV3) as follows. For:
Bluefish: Essential Fish Habitat Source Document. NOAA-NMFS-NE-144
(1999). p. 2, 8, and Figs. 26,28, 30, 31
Cod: Essential Fish Habitat Source Document. NOAA-NMFS-NE-124 (1999).
p. 29, Fig. 8
Flounder (summer): Essential Fish Habitat Source Document. NOAA-NMFS-
NE-151 (1999). Table 1, Fig. 3.
Flounder (winter): Essential Fish Habitat Source Document. NOAA-NMFS-NE-
138 (1999). p. 4,9, Table 1, Fig. 3 (this figure tends more to support a 9 month
residency factor, not a 6 month as do the other citations.)
-------
NPFS07: Clarification on Data Sources Used by the Nation'sPort to Estimate the Areal/Seasonal Use
Factor for HARS
Striped bass: Waldman et al. 1990, entire paper, note especially his various tag
recovery maps and the complete absence of recovered tags from the HARS. More
generally, the striper estimate comes from my having worked with stripers in the
late 80s - early 90s while at Harza, doing Hudson River work for GE. This
includes field collection of the fish at numerous locations in the New York area,
although not specifically at HARS.
It should be noted, that the Nation'sPort has only reviewed secondary literature (except
for striped bass) available at URS. The Natoion'sPort suggests that a review of the
primary literature/data should be performed by NOAA NMFS regional experts •
to refine the ASUF estimates.
-------
NPFS09: Bioaccumulation Steady State Adjustments
REVIEW OF PROPOSED CHANGES TO THE EPA/USACE REVIEW
BIOACCUMULATION TESTING EVALUATION FRAMEWORK (TEF) FOR
DETERMINING THE SUITABILITY OF DREDGED MATERIAL DISPOSAL
AT THE HARS
FACT SHEET No. 9: Bioaccumulation Steady State Adjustments
Author(s): Burt K. Shephard, Spyros P. Pavlou/URS Corporation - Technical
Representative, Nation'sPort.
Date: January 17, 2001
TEF Revision: USEPA's response to Reviewers' Comments (Charge 15), and RMW
Technical Meeting 01/10/01, Discussion Item 3a. USEPA is proposing to use a multiplier
to increase the measured tissue residue at the completion of the standard 28-day
bioaccumulation test for certain chemicals. This is to take into account the fact that some
chemicals (non-essential metals, high Kow organic compounds) do not reach steady state
concentrations within 28 days. For organic compounds, the recommended adjustment
factor is derived from Figure 6-1 of the Inland Testing Manual (EPA/ACOE 1998). The
Nation'sPort is proposing an alternative approach for adjusting the results of 28-day
bioaccumulation tests to account for organic chemicals that don't reach steady state in 28
days.
Evaluation: The Nation'sPort recognizes that not all chemicals reach steady state within
a 28-day period and believes that a more quantitative approach to deriving the adjustment
factors is available and should be used in lieu of the USEPA method. This alternate
approach is reported in the literature by Feijtel et al. (1997). The methodology is based on
information on the relationship between Kow and the kinetic elimination rate constant
(ke) for hydrophobic organic chemicals.
The basic equation in Feijtel et al. relates ke to Kow as follows
log ke = (-0.414 x log Kow) + 1.47 (1)
The time (days) required to reach 80% and 95% of steady state concentration is given by
the following equations
Tg0=1.6/ke (2)
T95 = 3.0/ke (3)
Finally, the proportion of the steady state body burden bioaccumulated during any length
of exposure time t can be expressed as
Proportion of steady state residue (at time t) = 1 - e("ke x l) (4)
-------
NPFS09: Bioaccumulation Steady State Adjustments
The Nation'sPort has entered equations (1) through (4) into an Excel spreadsheet, which
can be used to calculate the percent of steady state obtained during a t = 28 day exposure
for a chemical of any KoW. These calculations indicate that steady state should be reached
within 28 days for any chemical with a log K<,w less than 5.3 and that 95% of steady state
will be reached within 2 days for chemicals with a log Kow<5.9. For organic chemicals
that do not reach steady state within 28 days, the percent of steady state reached in 28
days can be estimated. This quantity can be in-turn used to derive a bioaccumulation
adjustment factor for chemicals that are not at steady state in 28 days.
Literature Cited
Feijtel, T, P. Kloepper-Sams, K. Den Haan, R. Van Egmond, M. Comber, R. Heusel, P.
Wierich, W. TenBerge, A. Gard, W. De Wolf andH. Niessen. 1997. Integration of
bioaccumulation in an environmental risk assessment. Chemosphere 34:2337-2350.
U.S. Environmental Protection Agency and U.S. Army Corps of Engineers. 1998.
Evaluation of Dredged Material Proposed for Discharge in Waters of the U.S. - Testing
Manual. Inland Testing Manual. EPA 823-B-98-004, Office of Water, Washington,
D.C.
-------
Maximum Allowed Concentration in Prey (MAPC) Calculations
Parameter
Abbreviation
Units
Proposed Factor
Target Cancer Risk Goal
TR
unitless
1.00E-04
0.0001
Target Hazard Quotient (non-cancer health
goal)
THQ
unitless
1
1
Average Adult Body Weight
BW
kg
70
70
Cancer Averaging Time
ATc
days
25,550
25550
Non-cancer Averaging Time
ATti
days
10,950
10950
Fish Ingestion Rate
F-IR
g/day
7.2
6.5
Fraction of HARs worm ingested by fish
FR
unitless
0.5
0.5
Fraction of HARS fish ingested by people
FI
unitless
0.5
0.5
People's Exposure Frequency
EF
days/year
365
365
People's Exposure Duration
ED
years
30
30
Cancer Slope Factor
SF
(mg/kg-day)1
2
2
Non-cancer Reference Dose
RfD
mg/kg-day
0.00002
0.00002
Ratio of Bioconcentration Factor (BCF) from
Water to Prey over BCF from Water to Fish
BCFp/BCFf
unitless
1
1
Combined area and seasonal use factor
ASUF
unitless
0.391
0.498
Conversion Factor
CF,
kg/g
0.001
0.001
Whole Body to Filet Ratio
BFR
unitless
1.35
1.35
Trophic Transfer Factor
TTF
unitless
3
3
Worm Tissue Concentrations in mg/kg
MAPC w/
Proposed Factors
Original TEF Value
Based on Cancer Effects
3.133
0.282
Based on Noncancer Effects
0.537
0.113
Formulas
Cancer MAPC = (1/ASUF) x [(TRx BW x ATc) / (F-IR xFIxEFxEDxSFx CFJ] x (BCFp/BCFf) x 1/ [ 1 + TTF x FR x (BCFp/BCFf)] x BFR
Noncancer MAPC = (1/ASUF) x [(THQ x BW x Atn x RfD) / (F-IR xFIxEFxEDx CFJ] x (BCFp/BCFf) x 1/ [1 + TTF x FR x (BCFp/BCF[)] x BFR
-------
Estimated proportion of steady state bioaccumulation reached in a 28-day bioaccumulation test
To estimate, enter the log KqW in cell B4
log KqW 5.9 L/kg
Elimination rate constant (ke) 0.107 day1
Time to 80% of steady state (T80) 15.0 days
Time to 95% of steady state (T95) 28.2 days
Percentage of final steady state
concentration attained in 28 days 94.9 %
Multiplier for use with measured 28-day
tissue residues to estimate residue at
steady state 1.05
-------
COMMENTS ON SCIENTIFIC PEER REVIEW PACKAGE, CHARGE, AND
PROPOSED SOWs
Authors: Spyros P. Pavlou, Burt K. Shephard, and Sharon J. Quiring/URS Corporation
Technical Representative Nation'sPort
Date: August 24,2001
Comments on the Scientific Peer Review Package and Charge
Comment 1 - Section 2(A¥P(ii>) Organotins. page 10: Why is the analytical method of
Rice, et al (1987) being recommended over Krone, et al (1989)? We recommend
considering Krone, et al as the preferred method because it provides for analysis of all
butyl-tin species, not just tributyl-tin. There is documented evidence that the other
organo-tin species are also of toxicological interest (Environment Canada and Health and
Welfare Canada, 1993). Furthermore, Krone's method provides for better quality
assurance, thus producing more reliable data. References:
Krone C. A, D. W. Brown, D. G. Burrows, R. G. Bogar, Sin-Lam Chan, And U.
Varanasi. 1989. A Method for Analysis and Measurement of Butyltins in Sediment
and English Sole Livers from Puget Sound. Marine Environmental Research, 27,
(1989), p. 1-18.
Environment Canada and Health and Welfare Canada. 1993. Priority Substances List
Assessment Report. Non-Pesticidal Organotin Compounds. Prepared under
Canadian Environmental Protection Act, Ottawa, Ontario, Canada. 32 pp.
Comment 2 - Section 2(E) Human Health Evaluations, page 17: EPA default values for
Exposure Frequency (EF) and Exposure Duration (ED) are 350 days/year and 30 years,
respectively. The document lists 365 days and 70 years for these parameters. The EPA
default values are based on the fact that most people spend two weeks per year away
from home and that 30 years is the upper bound length of time that a person spends at one
location. Since the EPA values are sufficiently health-protective, what is the justification
for the higher values used for HARS? While it is possible to set the EF, ED, and AT
terms so that they cancel out of the risk equations for non-carcinogens, this assumption is
overly protective for carcinogens (except for special circumstances). For the non-
carcinogenic equations, the AT term is 365 days/year x ED (USEPA, 1989, 1991). If an
ED of 30 years is used (the EPA default), then the non-carcinogenic AT would be 10,950
days. If the EF is 365 days/year rather than the default of 350 days/year then the EF and
the ED terms are cancelled out by the AT term. References:
USEPA. 1989. Risk Assessment Guidance for Superfund: Volume 1 - Human Health
Evaluation Manual. Part A. Interim Final. U.S. EPA Office of Emergency and
Remedial Response. Washington, DC. EPA 540/1-89/002.
Page 1
NPort General.doc
-------
USEPA.1991. Risk Assessment Guidance for Superfund: Volume 1 - Human Health
Evaluation Manual. Supplemental Guidance: Standard Default Exposure Factors.
Interim Final. OSWER Directive: 9285.6-03. March 25.
Comment 3 - Section 2(E) Human Health Evaluations, page 18: We recognize that the
default assumption of 10% of total arsenic in fish is inorganic has been the standard.
However, recently new EPA research has found that 1% inorganic arsenic in fish is more
appropriate for some species (Quiring, personal communication with R. Lorenzana, EPA
Region 10, 8/20/01). We suggest that EPA Region 2 confirms these findings with EPA
Region 10 and determines the applicability of this information into the HARS human
health evaluations.
Comment 4 - Section 2(E) Human Health Evaluations, page 23: The fish whole body to
filet ratio used in the lead calculation is listed as one on this page, rather than four as
presented in the earlier document. We believe the original value of four is sufficiently
health protective and is well supported by empirical studies. Available information
indicates that lead concentrates in the bone, organs, and blood of most animals and
comparatively little is found in tissues. A recent study of three fish species in the
Spokane River in Washington State found that lead concentrations in filets averaged 12
percent of the whole body concentrations (Johnson, 2000). References:
Johnson, A. 2000. Results from analyzing metals in 1999 Spokane River fish
and crayfish samples. Washington State Department of Ecology.
Comment 5 - Section 2(E) Human Health Evaluations. Trophic Transfer Factors, p. 18 -
20 and Appendix K. Trophic Transfer Factors: The Gobas model should be rerun after
SOW #3 is executed and the mixture of fish species identified. The actual lipid content
and weights of the species measured should be used to obtain updated TTFs. For
completeness, the data for recomputing the TTFs should include:
- Feeding preferences, weights, and lipid contents of each species in the food web
- Water temperature
- Organic carbon content of the sediment and overlying water
- Concentration of chemical in the sediment and freely dissolved concentration of
the chemical in the water column
- Densities of lipid and organic carbon
- Metabolic transformation rate constant
- Octanol-water partition coefficients (Kow) for chemicals of concern
Comments on Proposed SOWs
Comment 6 - Statement of Work (SOW) #1. Harbor Sediment Survev/Bioaccumulation:
SOW #1 proposes to conduct bioaccumulation tests on sediment samples with elevated
concentrations of target analytes. It is recommended that the lipid content of all
bioaccumulation test organisms be determined during each test. The availability of lipid
content data will permit more accurate estimates of chemical concentrations in higher
Page 2
NPort General.doc
-------
trophic-level organisms potentially consumed by anglers at the HARS. These improved
bioaccumulation and trophic transfer estimates should permit more accurate estimates of
human health risks from consumption of fish captured at HARS (see also comment 5
above).
Comment 7 - Additional Contaminants of Concern: The organotin detection limits (DL)
currently achieved by Battelle as listed in SOW #1 are appropriate for use at the HARS in
the evaluation of potential ecological risks to aquatic species from bioaccumulation of
these chemicals. These limits are lower than the lowest observed adverse effect residues
of any organotin compound measured in field surveys of chemical effects on freshwater
or marine biota. For example, the DL for TBT (1.442 |ig/kg wet weight) is lower than
the lowest TBT residue in the Tissue Residue Effects Association Database (TREAD)
associated with an adverse toxicological response in either freshwater or marine species
tested in laboratory studies with individual organotin compounds. An amount of 10 (ig/kg
TBT is associated with immune system response in channel catfish (Rice et al. 1995), and
73 fig/kg TBT is the reported induction threshold in the periwinkle Littorina littorea
(Bauer et al. 1997). References:
- Bauer, B., P. Fioroni, U. Schulte-Oehlmann, J. Oehlmann and W. Kalbfus. 1997.
The use of Littorina littorea for tributyltin (TBT) effect monitoring - results from
the German TBT survey 1994/1995 and laboratory experiments. Environ. Pollut.
96:299-309.
- Rice, C.D., M.M. Barnes and T.C. Ardelt. 1995. Immunotoxicity in channel
catfish, Ictalurus punctatus, following exposure to tributyltin. Arch. Environ.
Contain. Toxicol. 28:464-470.
Comment 8 - Statement of Work (SOW) #2. Determination of Spatial Ranges for Fish
found at the HARS (Site Use Factors'): The goals and objectives of the study and the
methods for selecting the fish species (for which area use factors will be estimated) must
be better defined. It is not clear which of the following are the real goals for calculating
area use factors:
(1) Fish species consumed by anglers fishing in the vicinity of the HARS
(2) Fish species which comprise a food web from the benthic invertebrates used in
laboratory bioaccumulation studies (trophic level 2 of the Gobas model) up to fish in
(trophic levels 3 and 4 of the model), whether or not they are consumed by humans
(3) Both of the above as is implied in the first sentence of SOW #2.
It is not clearly stated why home ranges are needed in the ecological food-web model.
Once the goals of the tagging study are defined, the criteria for selecting the species to be
tagged could be refined after discussions between human health and ecological risk
assessors, and the fishery biologists who will be performing the tagging study. In this
manner, areas of conflict between selection criteria for ecological and human health
analysis can be resolved, and the limitations of data obtained from tagging studies
understood. These discussions will also serve to give direction to the fishery biologists
performing the tagging studies, particularly in the areas of the species to tag and the
Page 3
NPort General.doc
-------
duration of the tagging studies for species that spend only part of the year or their
lifetimes at the HARS. Support to these suggestions follows.
Selection of Target Fish (human health concerns): Although all of the factors listed under
Subtask 2a have bearing on the exposure of humans to contaminants within fish captured
at the HARS, not all of the factors will influence the selection of fish species to be
monitored during a tagging study. Humans will consume fish with a wide variety of lipid
contents, so lipid content of the different fish species per se is not a basis for selecting a
fish for tagging.
If the high lipid content fish are the only species consumed, they will likely contain
higher concentrations of hydrophobic organic contaminants, and thus represent a worst
case exposure scenario. However, if the species most commonly consumed by HARS
anglers are low lipid content species such as winter flounder, the low lipid species should
be included in the tagging study, based on a large contribution to the fish ingestion rate of
the human exposure scenario.
Similarly, cooking and preparation methods, and handling of the species prior to
consumption affect human exposure to contaminants, but have little or no bearing on the
selection of species to be tagged. If the objective of the tagging study is to determine
area use factors for species comprising the largest proportion of species caught at the
HARS and consumed by humans, the fish species caught at the HARS with the largest
human consumption rates should be the species selected for tagging.
The fish species comprising the largest proportion of the fish ingested will likely have a
range of lipid contents. Therefore, lipid content and handling and cooking methods,
should not be primary determinants in the selection of species to be tagged.
Selection of target fish (ecological concerns): Under Subtask 2a, the third criterion (that
the fish is in a predominantly sediment-driven food web) should be the only basis for
selection of the species to be tagged from a solely ecological standpoint. Even this
criterion may be overridden by human health concerns (for example, if striped bass
happen to be the fish captured at HARS which accounts for the largest portion of the fish
ingested by HARS anglers).
The criterion of tagging fish species with limited home ranges is tuntamount to the
criterion that a fish be resident at HARS throughout most of its life cycle. While both of
these criteria serve to to maximize its exposure to HARS contaminants, fish caught and
consumed by anglers at HARS may not be resident at the site for a significant portion of
their lifespan. The fact that a fish species spends a significant portion of its life outside of
the vicinity of the HARS should not by itself disqualify that species from consideration
for a tagging study. Such a species is either ecologically important (such as an apex
predator), or comprises a significant portion of the fish caught and consumed by HARS
anglers, such species should receive substantial consideration for inclusion in tagging
studies.
Page 4
NPort General.doc
-------
The proposed ecological selection criterion of tagging only top level predators is
questionable from both ecological and human health standpoints. If the goal is to refine
area use factors for the food web, these quantities should be defined for species at all
levels of the food web, not just for the top predators. If the goal is to define area use of
the species consumed by humans, not all fish species caught at the HARS by anglers will
be top predators.
The criterion specifying that fish are resident throughout most of their life cycle and
especially during critical life stages is not realistic, and in fact would exclude species
such as summer and winter flounder from tagging studies. Spawning of these two species
(a critical life stage) takes place in nearshore waters miles away from the HARS.
Comment 9 - Section 6. Scope of Work (SOW) #3: This SOW is too general to evaluate.
We recommend adding as many site-specific procedures as possible in order for peer
reviewers to provide meaningful comments. Our suggestions are as follows:
- Specify how the population of anglers who eat fish from HARS will be identified
- which local/state agencies will be contacted, which local fishing groups will be
involved?
- The SOW states that "because of the expense involved", HARS is not used for
subsistence fishers. Specify the reasons why HARS fishing is expensive and
where the information was obtained as this will have a direct bearing on the
angler population using the area.
- The SOW provides a laundry list of information a fish consumption survey "may
include." Specify how it will be decided which questions are pertinent and the
agencies/groups who will have input on the question list.
- The SOW states that once a question list has been identified it should be tested on
a focus group. Specify how such a focus group will be selected and how the
focus groups responses will be evaluated.
- The SOW lists two differing types of creel surveys and says either may be used.
Sufficient information is available to propose one over the other for the HARS
site.
Page 5
NPort General.doc
-------
1
?!
Paper Submitted by
New York State Department of
Environmental Conservation (NYSDEC)
-------
NYSDEC provides the following comments, addressing the expressed need for expedited submittals for
the next round of Federal review. Our comments are in three areas of EPA's proposed approach; 1 fish
consumption factor, 2. site use factor, and 3. integrated risk assessment (cumulative exposure).
1. Fish Consumption
Various aspects of this topic were discussed in the meetings, including the current method used by EPA
in developing the new PCB value, other possible methods such as conducting an area-specific
consumption survey, and general factors to consider such as appropriate at-risk groups. We are generally
in favor of conducting an area-specific survey to generate better information. If the currently proposed
EPA method were to stand however, our initial comments provided in the meetings remain, as outlined
below:
EPA referenced the study "Fish Consumption Patterns by New Jersey Consumers and Anglers" prepared
by NJMSC. The EPA Memo of September 27,2000 outlined the method, which uses various
information in the report to calculate a fish consumption term (7.2 grams/day) for the risk based non-
cancer equations.
The last adjustment used in the EPA approach does not appear to be warranted, which would leave a non-
conservative value of 11.3 grams/day for the consumption term. This last adjustment applies a statistic
that anglers reported 60% of the fish consumed is prepared in the home (11.3x0.6=7.2). The survey
methods described in the report, however (Study Design, pps. vi & viii), do not indicate that this result
should be applied to the saltwater, fresh caught fish reported to be consumed, but rather to all of the fish
reported to be consumed by the anglers. The recreational anglers reported consuming 15.23 ounces per
week (ostensibly of recreationally caught fish). The other factors EPA used to adjust the overall reported
consumption, saltwater finfish only and fresh vs. canned or processsed, could be appropriately adjusted
to the recreational consumption value (the study design does not make this clear) since these are all
reported on a weight basis. The home preparation adjustment belongs in a different set of statistics than
were grouped in the proposed approach, since it can be reasonably assumed that most (>60%) of
recreationally caught fish are prepared at home (i.e., as opposed to a restaurant, at work or school, or
another home, as described in the report).
2. Site Use Factor
There was extensive discussion of this issue which was facilitated through alternative methods proposed
by the consultants for NationsPort and the NY&NJ Port Authority. We are generally in support of using
more well-defined factors such as some that were proposed in the alternative approaches, including
adjustments for typical home ranges of target fish species. The development and application of any such
factors should be subject to expert review, though, since there were concerns raised from NMFS
regarding the relevance of the existing information in this regard.
We agree with EPA that the FR term (fraction of HARS worm ingested by fish) in one of the proposed
approaches is not appropriate since it is assumed that the test worms and clams are surrogates for any
benthic prey ingested by HARS fish.
3. Integrated risk assessment - cumulative exposure.
This issue was raised late in the meetings, since many issues had to be dealt with regarding specific
HARS disposal exposures and risk. It is NYSDEC's position that if a risk-based human health approach
is taken, then as is standard practice (EPA and DEC) cumulative risk needs to be considered. An
assessment can and should be done that evaluates the information available on a specific contaminant
basis that may identify a potential for cumulative concerns. Should the assessment indicate a potential for
-------
other dietary or other exposure (e.g., high regional mercury levels through exposures including air intake),
then policy decisions need to be formulated for allowable incremental exposure to selected at-risk groups.
Alex Lechich
Marine Resources
NYSDEC
-------
NYSDEC RMW Representative's Comments on Draft TEF Submittal to Peer Reviewers
September 14. 2001
Two main comments are provided below that are in regard to the inclusion of dioxin-like
coplanar PCBs in the regional assessment of dioxins/furans, and the specific consideration of exposure of
higher trophic level organisms (fish, crustaceans) to background sediment conditions outside of the
HARS.
1. EPA Region 2 proposes to include consideration of the dioxin-like coplanar PCBs in the regional
dioxin protocol only after the national dioxin reassessment is completed, the Region has reviewed the
findings, and the entire regional dioxin protocol potentially revised. The problem is that the whole
process can take an inordinate amount of time, considering that the issue can be dealt with by a relatively
minor and reasonable revision of the protocol. The dioxin-like properties of certain coplanar PCBs have
been accepted for years now, and dioxin toxic equivalency factors (TEFs) have been developed for them
by the World Health Organization (WHO), most recently revised in 1997. EPA Region 2 has utilized
the WHO (1997) dioxin/furan TEFs in their regional dioxin protocol, and it would be an appropriate
update to now include the dioxin-like PCBs. This can be done relatively simply, as discussed below, and
should not greatly upset the level of conservativeness of the overall evaluation approach.
The current regional dioxin approach cited in the testing evaluation memos (USEPA. 1997a.
Memo to File from A. Lechich. Subject: Summary of Dioxin Risk Evaluation Approach. March 15,
1997.) includes evaluation of dioxin/furan congener test residues by comparison to a TEQ (toxic
equivalent) calculated by applying the WHO TEFS and a detection limit protocol defined in a joint
EPA/Corps memo also cited in the decision memos (USEPA/CENAN. 1997. (Joint Memorandum)
Ocean Disposal of Dredged Material Clarification of Two Procedural Elements of Interagency
Coordination Between USEPA Region 2 and the New York District USACE - Treatment of Non-Detects,
Chemical Data, and Rules and Responsibilities in Preparation of the Ocean Disposal Regulatory
Compliance Memorandum.) This application resulted in the establishment of a threshold of 4.5 pptr
TEQ for all the non-2,3,7,8-TCDD dioxins and furans. The test organism tissue residues for these
congeners are then compared to the threshold by summing the product of their residues and respective
TEFs. This approach updated the original (1990) protocol that had established a threshold for only
2,3,7,8-TCDD (1 pptr for Category 1,10 pptr for Category 2). The dioxin-like coplanar PCBs can be
included in the same way as the other congeners were, with only a minor change in the regional protocol.
The resulting TEQ value would be some number greater than 4.5 pptr (will leave the calculation for the
competent EPA staff, don't have the MDLs), against which the sum of non-2,3,7,8-TCDD dioxins,
dioxin-like furans and dioxin-like PCB residues times their TEFs would be measured.
It should be acknowledged that this inclusion will place an additional burden on the original
conservative assumption that detectable amounts of 2,3,7,8-TCDD (and hence its toxic equivalents) are a
matter of concern. That assumption was based on an EPA model applied in 1987 that resulted in a 10 4
human cancer risk for 6.5 gm/d consumption of fish contaminated with 7 pptr dioxin, and considered that
a different acceptable risk level or consumption rate could result in acceptable benthic tissue levels
below detection (also considering other exposure and uptake factors). The Region has since established
that the 10 4 risk level is appropriate for their proposed evaluation framework. In addition, EPA proposes
to use a weight-of-evidence approach with the HARS-specific values, which would include consideration
of things like the variability around the mean accumulated residues and the uncertainties and
assumptions in development of the values. If these considerations are applied for the other HARS-
specific values, there is no reason why they should not apply for the dioxin values. Although it may
appear to be reasonable to await the completion of the reassessment, implementing these changes along
-------
with the rest of the TEF would update the dioxin evaluation according to currently accepted scientific
practice and maintain the Region's status at the forefront in dredged material testing and evaluation
methods.
2. It is NYSDEC's position that fish that may be exposed to HARS conditions are also exposed to the
conditions in the surrounding environment, and that this should be accounted for in a risk-based
approach. Since EPA has developed and proposes to adopt a site use factor in the exposure calculations,
it would be appropriate to also account for the exposure to the site surroundings when fish are not
assumed to be using the HARS. It is therefore suggested that if a site use factor of 0.77 is applied to
develop HARS-specific values, a factor of 0.23 should be applied to the benthic invertebrate background
data, by contaminant, and added to the test residue results. This is somewhat analogous to the current
practice of applying steady state factors to the test residue results. The steady state factors are also
applied for the purpose of obtaining a more correct estimate of actual exposure conditions in the field. In
the case of site use, the test residue values represent the potential accumulation by fish when they use the
HARS, and the adjusted background values would represent the potential accumulation when they are in
the site surroundings. The testing evaluation memos provide the arithmetic means of the benthic tissue
background data. Either these can be used or some other statistic that may be proposed by other RMW
members or peer reviewers can be used.
A review of the potential effects of the above suggested change using the background means
with information available thus far from EPA (proposed human health values) indicates that it might
effect mainly the dioxin evaluation. This would give added weight to the need for the considerations
described near the end of 1. above. It should also be noted that the current regional dioxin protocol uses
a site use factor of 0.5, based on discussions that were held with NOAA/NMFS staff at the Sandy Hook
laboratory.
Alex Lechich
Marine Resources
NYSDEC
-------
Paper Submitted by
Clean Ocean Action
-------
Review of Proposed Changes to the EPA/USACE Review of the Bioaccumulation
Evaluation Framework (TEF) for Determining Suitability of Dredged Sediments
as Material for Remediation at the Historic Area Remediation Site (HARS)
Prepared by: Clean Ocean Action, P.O. Box 505, Sandy Hook, NJ 07732; phone: 732-872-
0111; fax: 732-872-8041 on October 4,2001. Please address correspondence to Kristen
Milligan, Staff Scientist (e-mail: Science@CleanOceanAction.org).
Abstract
The Historic Area Remediation Site (HARS) is a 15-square nautical mile area determined by the
Environmental Protection Agency to be contaminated and in need of remediation. Remediation
plans include capping the site with at least 1 meter of Material for Remediation (i.e., cap
material) in a priority remediation area of approximately 9-square nautical miles. The site is
used for recreational and commercial fishing for species such as bluefish and lobster.
Current regulations require liquid, suspended, and solid phase testing of all sediments proposed
as HARS cap material. The solid phase tests include acute toxicity and bioaccumulation
bioassays. Bioaccumulation bioassay results are evaluated to determine if there is the potential
for unacceptable negative effects to marine life and human health. No bulk sediment chemistry
evaluations are performed. There are limitations to this current regional approach, including that
the guidance can not ensure that impacts will be reduced at the site since it does not consider
pre-remediation contamination levels (in benthic biota and sediments) at the site and does not
require that contaminant levels are reduced. Moreover, currently-used and proposed methods for
selecting HARS cap material do not consider ambient background contaminant levels. This peer
review specifically shall address the proposed methods for determining potential for risk from
sediments that may be placed at the HARS as cap material.
This paper summarizes issues of concern found within the proposed methods. Issues of concern
about the proposed twenty-eight (28)-day bioaccumulation framework are highlighted within
three parts: specific shared factors in ecological and human health guidelines, components of the
human health risk-based guidelines, and the total proposed framework (ecological and human
health risk-based guidelines).
Recommendations include:
• lobsters must be a target species, and use high exposure species if finfish target species are
chosen;
• incorporate background levels of contamination in fish and adopt more protective cancer risk
levels;
• reject spatial models of fish movement that separate HARS from the NY Bight Apex; and
• adopt validated methods that accurately reflect risks and support remedial goals. In the cases
where validated methods do not exist, apply additional procedures and err on the
precautionary side.
Other recommendations relate to the proposed analyte list, steady-state models, trophic transfer
models, and factors within the human health risk-based equations (e.g., site use factor, risk level,
body weight).
1
-------
1. General Background and Concerns with the Regional Assessment Methods
1.1 Introduction
On September 1,1997 the Historic Area Remediation Site (HARS), located in the New York
Bight Apex was designated under the Marine Protection, Research, and Sanctuaries Act (Figure
1). This federal law requires that only Material for Remediation {i.e., cap material) be placed at
the site, so as to reduce impacts to acceptable levels (in accordance with 40 C.F.R. 228.11(c)).
US Environmental Protection Agency (EPA) and US Army Corps of Engineers (ACE) manage
this site. Since its designation, over four
million cubic yards of material (Table I) have
been approved as cap material for the site.
However, regional guidance for the evaluation
system to determine Material for Remediation
has not been finalized. This peer review is
part of the finalization of guidance for
interpreting bioaccumulation bioassay test
results for selecting Material for Remediation.
1.2 Contamination Conditions at the
Historic Area Remediation Site
The Historic Area Remediation Site (HARS) is
a 15-square nautical mile area that was
concluded by the EPA to be contaminated
(USEPA, 1997). It surrounds the original Mud
Dump Site, where contaminated dredged
material was dumped for decades. As
described by EPA in the introduction to the
HARS in the peer review packet, the HARS
was designated "Impact Category I" under
the Marine Protection Research Sanctuaries
Act (40 C.F.R. Section 228.10 (c)(1)). Some
effects that can determine "Impact Category
I" status are an identifiable progressive
movement or accumulation of contaminants in detectable concentrations above normal ambient
values from the disposal site (Mud Dump site) and adverse impacts to marine populations within
the site (refer to endnote 1). In the Supplemental Environmental Impact Statement finalized in
May 1997 for the closure of the Mud Dump site and the designation of the HARS, the executive
summary concluded that remediation is necessary for four reasons:
(1) Contaminant toxicity—data taken in 1994 show acute toxicity in sediment from areas around
the Mud Dump Site (USEPA, 1997),
2
Figure 1. Location of the Mud Dump Site and the
Historic Area Remediation Site (HARS) in the
New York Bight. The Mud Dump Site and its
surrounding area were used for decades for dredged
material disposal. In 1997, the Mud Dump Site was
closed due to contamination-related impacts, and the
Historic Area Remediation Site (HARS) was
designated. The HARS is to be capped with clean
sediments. Source: USEPA, 1997.
-------
Table I. Projects approved by USEPA and USACE as HARS Material for Remediation, listed in chronological order. "Date of Approval" is
the date that the agencies sign a memo on the suitability of the sediments for HARS and is not the date of permit issuance "Type" of project refers
to whether the dredging is to maintain current channel depth (M, maintenance) or to deepen the channel (D, deepening). Data on particle sizes
are averages from composited oores from each reach of the project; in instances where more than one reach was evaluated, reach data were
averaged. (All data provided by USEPA, Region 2 and USACE, NYD.)
Project Name
Location
Date approved
Volume
(yd3)
Type
%
silt/clay
%
sand/gravel
Passenger Ship Terminal
Hudson River (piers 88-94, Borough of
Feb-98
440,000
M
96
4
Manhattan)
Refined Sugars
Hudson River (City of Yonkers, NY)
Aug-98
60,000
M
8
92
Kill Van Kull Contracts 1
am/1
lower Kill Van Kull
Jan-99
1,268,000
D
19
81
ana z
Raritan River, Reach A
Raritan River and upper Raritan Bay (off of
Feb-99
310,000
M
97
3
South Amboy)
Kill Van Kull/Newark Bay Kill Van Kull and Newark Bay
Apr-99
unlimited
D*
100
0
Complex *
Castle Astoria Terminals
Steinway Creek, upper East River (across from
Apr-99
110,000
M
87
13
Rikers Isl.)
Ellis Island
New York Harbor
May-99
70,000
M
100
0
Raritan-Arthur Kill Cutoff
upper Raritan Bay (off of Perth Amboy)
Jun-99
112,000
M
94
6
Brooklyn Marine Terminal
East River, Upper New York Bay
Sep-99
280,000
M
99
1
Buttermilk Channel
Buttermilk Channel (b/w East River and Red
Feb-00
112,000
M
68
32
Hook Channel)
Kill Van Kull C3 R1
mid-Kill Van Kull (near Constable Hook,
Mar-00
16,000
D
14
86
Bayonne NJ)
Kill Van Kull C3 R2
mid-Kill Van Kull (near Constable Hook,
Mar-00
251,000
D
27
73
Bayonne NJ)
Kill Van Kull C3 R3
mid-Kill Van Kull (near Constable Hook,
Mar-00
1,160,000
D
18
82
Bayonne NJ)
Kill Van Kull C4 R1
upper Kill Van Kull (near Bergen Point)
Mar-00
7,000
D
19
81
Kill Van Kull C5 R2
upper Kill Van Kull (near Bergen Point)
Mar-00
341,000
D
13
87
Perth Amboy/Ward Point
Lower NY Bay, Raritan Bay (Perth Amboy)
Jan-01
850,000
M
97
3
IMTT-Bayonne, Inc.
Kill Van Kull (city of Bayonne)
Apr-01
94,800
M
43
57
Red Hook Flats
Upper NY Harbor (south of Governor's Island)
Jun-01
445, 060
M
85
15
Densely consolidated Pleistocene Red Clay, which is found beneath recent sediments
3
-------
(2) Contaminant Bioaccumulation/Trophic Transfer— results showed that there were areas in
the study area [Mud Dump site and surrounding areas] where infaunal worms were
accumulating undesirable levels of contaminants from sediments (USEPA, 1997),
(3) Contaminants in Sediments— based on contaminant levels in the sediment, negative
biological effects could be possible at many sampling stations within the study area (Mud
Dump site and surrounding areas; USEPA, 1997), and
(4) Contaminant Levels in Area Lobsters—PCB and 2,3,7,8-TCDD (dioxin) concentrations in
the hepatic tissue of the lobsters were above the US FDA consumption guidelines and lobster
study data revealed that food sources of Bight Apex lobsters are contaminated, that
contaminants are being accumulated, and that concern about potential human-health risks is
warranted (USEPA, 1997).
Contaminants specifically found to be elevated in
infaunal worms living in sediments at HARS in
comparison to those living in sediments outside
of HARS were total polycyclic aromatic
hydrocarbons (PAHs), polychlorinated biphenyls
(PCBs), and the dioxin 2,3,7,8-TCDD (USEPA,
1997). Only total PAHs and PCBs were elevated
in HARS in comparison to ocean background
levels (those sites that are outside of HARS and
presumably not impacted by ocean dumping of
contaminated dredged material). The dioxin
2,3,7,8-TCDD was found to be elevated
throughout the New York Bight Apex and not
specifically elevated in HARS in comparison to
outside HARS. Both inside and outside HARS,
2,3,7,8-TCDD dioxin levels exceeded the current
acceptable limit of 2,3,7,8-TCDD dioxin (i.e., 1
ppt. wet weight) in benthic tissue (USEPA,
1997).
Even though there are high levels of
contamination within HARS, there is a diversity
of species that migrate through, inhabit, feed
within, and lay eggs in the HARS. The largest
percentage of fish species found in the area
studied by the Supplemental Environmental Impact
Statement (USEPA, 1997) spend most of their
lifecycle on or near the bottom and amount to 21
species that are considered by the EPA to be
commercially, recreationally, or ecologically
important. There are also species of invertebrates
including crustaceans (e.g., rock crabs, lobster)
dependent on the sea-floor at HARS, some of which
are currently commercially harvested (Figure 2).
73 5?W
w
—-* ¦ - ^ .. - -
i-swig types
Hook & Une
1996Banyme8y
A/ <20 meters
A/20 metes
/W> 20 meters
2 * 3 Ktonahn
73 WW
2 Mbs
Figure 2. Major fishery areas of the Historic
Area Remediation Site (HARS). Recreational
(hook and line) and commercial (lobstering,
trawling) fishing is frequent at the HARS and
vicinity. During baseline HARS surveys, trap
lines and trawl scour marks were observed in
the HARS area. There are no current or
planned fishing restrictions during or after t
remediation work. Source: USEPA, 1997.
4
-------
1.3 Goals of Remediation
The aim of remediation was to reduce the potential human-health and ecological impacts
presented by sediments in HARS. Specifically, the Supplemental Environmental Impact
Statement (USEPA, 1997; p. 4-31) concluded that:
"When remediation operations are completed in the PRA [priority remediation
area], the potential for contaminant bioaccumulation will be reduced as well as the
potential for sublethal effects in benthic marine organisms and their predators
(including human consumers of fish and shellfish from the area)."
The preferred alternative of remediation chosen in 1997 was intended to not only cap acutely
toxic sediments but to lower contaminant levels in organisms that inhabit the HARS. According
to the EPA (USEPA, 1997; p. 4-35), this alternative would result in the following scenario:
"Organisms such as crabs, lobsters, and demersal fish that currently feed on HARS
infauna with high body burdens of contaminants will receive decreasing
contaminant exposure as the PRA [priority remediation area] is remediated. This
exposure-reduction will be a beneficial effect on Bight Apex organisms, and human
beings will have less risk of adverse effects from consumption of Bight Apex
seafood."
However, specific targets for contaminant levels in benthic tissue and in sediments have not been
articulated by the federal agencies, and a debate continues whether levels of those contaminants
that were found elevated within HARS must be eventually lowered to levels approximating those
in background areas outside of HARS.
Even though no quantitative goals have been articulated by the federal agencies for HARS, there
is the implication that the ecological and human health risk-based guidelines under development
for benthic tissue will be acceptable target concentrations, or "goals" for post-remediation
conditions at the HARS. These goals can cause persisting elevated levels of contamination at the
HARS in relation to areas outside of the HARS because current policy in this region states that
contaminant guidelines (for benthic tissue) developed for HARS cap material may exceed
baseline levels at HARS and ambient background, including for those contaminants found
specifically elevated within the site. Indeed, previous sediments that have been approved as
HARS cap material caused levels of bioaccumulation in benthic worms and clams exceeding
ambient background levels for many contaminants and, some cases exceeding those levels in
HARS. Currently-used guidelines for many contaminants also exceed ambient background
levels (data available upon request).
The purpose of deriving ecological and human health risk-based guidelines for benthic tissue is
to ensure that whatever sediments are used, these sediments will not persist or create unsafe
levels of food chain contamination leading to unsafe levels in seafood. However, there is still
the outstanding concern that this effects-level approach needs to be combined with approaches
specifically tailored to reducing those elevated contaminant levels at HARS.
5
-------
One question for the peer review at this time is whether the proposed guidelines are protective of
ecological and human health for the purposes of HARS remediation and how these guidelines
can be improved via methodological or other types of changes.
1.4 Solid Phase Tests: Concerns about the tests and guidance that are the bases for
selecting HARS Material for Remediation
The current regional guidance for selecting HARS Material for Remediation relies on acute
toxicity and bioaccumulation test results for the solid phase testing with no direct comparisons to
baseline levels at the HARS. There are no limits or guidance for levels of contaminants in the
bulk sediment. There are several disadvantages to this current approach, including: acute
toxicity tests may not correspond to benthic health (Long et al. 2001; see endnote 2), twenty
eight-day (28) bioaccumulation test results do not correlate well with contaminant loads in the
sediments, and static laboratory bioaccumulation testing may not provide an accurate measure of
bioavailability and trophic transfer once sediments are placed at an open ocean site. Finally, the
guidance as discussed in sections 1.2 and 1.3 does not ensure that those contaminants found
elevated in biota and sediments at HARS will be reduced by the selection of Material for
Remediation.
Given the above reservations, the remainder of this paper discusses the proposed framework for
assessing ecological and human health risk-based guidelines for the 28-day bioaccumulation
bioassays. This paper does not discuss the other methods which could be applied in conjunction
with acute toxicity and bioaccumulation bioassays to improve the prediction of potential for
adverse effects and better ensure reduction of contaminant impacts within HARS.
2. Discussion on the Proposed Framework for
Evaluating 28-day Bioaccumulation Test Results
2.1 Introduction
The proposed framework can be viewed as a series of "sub-models" to approximate effects of
sediments and fit into a larger model that is the final decision process for selecting HARS
Material for Remediation. For bioaccumulation evaluations, these "sub-models" are the various
components of the ecological and human health risk assessments used for determining risk-based
guidelines.
Figure 3 illustrates life stages of a fish that may be affected by contamination. Other marine
organisms have similar or more complex life cycles. These effects (i.e., reproductive failure,
fertilization failure, recruitment failure) may have significant population-level impacts. The
ecological risk-based guidelines must be sufficiently protective to minimize effects in marine
organisms throughout the life cycle stages. The human health risk-based guidelines must protect
seafood consumers of adult fish (or molluscs, crustaceans). The remainder of this paper
6
-------
Adults with abnormalities and high body burdens of contaminants
juveniles- developmental
deformities may result in
mortalities
unfertilized eggs spawned with
contaminant load already present
second meiotic division of egg is especially
sensitive to contaminants
fertilized egg
larval development-
developmental deformities may
result in mortalities
sperm sensitive to
contaminants in
water column
blastodisc stage with cytologic and
chromosomal abnormalities
hatching- another
critical mortality period
gastrula stages may be especially
sensitive to contaminants
Embryo with developmental anomalies
Figure 3. Points in the life cycles when fish are especially sensitive to
pollutants. Adapted from NOAA, 1994. The peer review for human health risk
assessment considers the body burden in the adult phase of the fish and
consequent trophic transfer to seafood consumers. The review for ecological risk
assessment should consider potential for adverse effects at all life stages within the
cycle.
-------
discusses: factors in the proposed framework that have implications for the ecological and
human health risk-based guidelines (i.e., "shared factors"), the human health risk-based
guidelines, and validation of the overall framework.
2.2 Shared factors: Proposed analytes, adjustment to steady state, trophic transfer
2.2.1 Proposed analytes
Dioxin-like coplanar PCBs: The EPA (Region 2) recognizes the need to include dioxin-like
coplanar PCBs in the assessments of dredged sediments. The current proposal is to delay
inclusion of these contaminants until the dioxin reassessment has been completed by USEPA.
This process is open-ended with no discrete timeline. The dioxin-like properties of some
coplanar PCBs have been long-recognized by health officials and toxicologists. Given the
significant toxicological properties, it is important to include these co-planar PCBs at this time in
the assessments of dredged sediments that are being placed as remediation cap material at the
HARS to protect against future harm at that site and in NY Bight Apex fisheries resources.
Polycyclic Aromatic Hydrocarbons (PAHs) and organotins: The proposal to include alkylated
PAHs and organotins in bioaccumulation assessments marks a significant improvement in the
target analyte list, especially since dredged sediments will be originating from a heavily
industrialized and urbanized harbor system where PAH and organotin impacts are most likely to
be evident. These proposed analytes are important not only from a human health perspective but
also from an ecological perspective. For example, evidence has emerged over the past decade on
effects of PAHs on organisms that PAH contamination can cause significant effects at low
levels. Carls et al. (1999) demonstrated that tissue levels from 22 to 1,400 ppb in Pacific Herring
can result in mortality and/or structural abnormalities (lack of jaw development, small jaws,
spinal defects, failure to develop fins), yolk sac swelling, pericardial (heart) sac swelling,
alterations in swimming behavior, and chromosomal damage. Concentrations of PAHs in HARS
biota fall within this adverse effects range. Cases of significant effects to marine organisms can
also be made for organotins.
Mercury: The proposal to assess effects from mercury is to allow the measurement of
methylmercury in benthic tissues in the cases where bioaccumulation test results for total
mercury exceed ecological or human health risk-based guidelines. This approach is
underprotective because it assumes that the methylmercury tissue concentration after a twenty-
eight (28)-day bioassay accurately reflects methylmercury bioavailability and trophic transfer in
the field.
Areas of uncertainty for the proposal's assumption include: (1) Tissue storage requirements are
not articulated and may affect mercury speciation; (2) Bioaccumulation test conditions are not
the same as conditions during ocean placement of the sediments; and (3) Assuming that 28-day
bioassay test results could accurately reflect benthic tissue concentrations in the ocean
environment, trophic transfer models of methylmercury may not be accurate (e.g., due to
transformations in the gastrointestinal tract).
8
-------
For these reasons, total mercury should continue to be the analyte for assessing mercury effects
to the ecology and human health and methylmercury should not be considered a reliable proxy.
This recommendation is consistent with EPA recommendations for assessing chemical
contaminant data for use in fish advisories, where the determination of methylmercury is not
recommended (see endnote 3).
2.2.2 Adjustment to Steady State
An issue of concern is that organisms in the 28-day bioaccumulation bioassays are depurated to
remove contaminants from the gastrointestinal tract. Regional guidance requires that test
organisms must be depurated for twenty-four hours in "clean sand" prior to freezing and
subsequent tissue analysis (USEPA/USACE, 1992). A model for extrapolating benthic tissue
concentrations to higher trophic levels must include considerations of gut content, since
predators of benthic prey are not selective feeders and will consume the whole prey item. One
potential area for inclusion of this consideration is the steady-state correction factor.
Another issue of concern in the approach using adjustment to steady state is that the steady-state
concept assumes that contaminant levels in an organism will stabilize. This assumption,
however, is typically conditional on constant environmental conditions as defined in each
published experiment or theoretical model. For example, research results from which steady-
state approximations were taken would be dependent on environmental and biological conditions
such as water temperature (e.g., seasonality of sampling), ingestion and digestion rates,
metabolic rates, synergistic effects, and lipid concentrations. Results from these studies or
theoretical models can not be realistically applied to a fluctuating, natural system to predict
tissue levels. For these reasons, any steady state adjustment must err on the most conservative
side (for example, by including a conservative "safety factor").
2.2.3 Trophic transfer
Trophic transfer factors for transfer of contaminants from benthic prey to fish predators are
proposed for ecological and human health risk-based guidelines. In the proposed approach,
these factors are derived using the same methods for both purposes, which may be an
inappropriate approach. For protecting ecological health, these factors must protect the benthic
prey, various life stages of benthic and pelagic species in HARS, and top level predators. This
may require the derivation of several ecological risk-based guidelines (e.g., for protecting
reproductive health, for protecting benthic assemblages, for protecting organisms at higher
trophic levels).
Other concerns about the proposed trophic transfer model that must be addressed include: (1)
whether or not PAH transfer factors are appropriate in light of the proposed PAH analyte
additions, (2) whether the Gobas model (Gobas, 1993) at equilibrium is protective in an open
ocean scenario, and (3) if the lipid values used for deriving the transfer factors are appropriate,
protective, and represent high-exposure scenarios (e.g., those species with high lipid content or
peak season lipid concentrations at a trophic level).
9
-------
In addition, as discussed in section 2.2.2, organisms in the twenty-eight (28)-day
bioaccumulation bioassays are depurated to remove contaminants from the gastrointestinal tract.
Regional guidance requires that test organisms must be depurated for twenty-four hours in
"clean sand" prior to freezing and subsequent tissue analysis (USEPA/USACE, 1992). A model
for extrapolating benthic tissue concentrations to higher trophic levels must include
considerations of gut content, since predators of benthic prey are not selective feeders and will
consume the whole prey item. The trophic transfer correction factor is another area for inclusion
of this consideration.
2.3 Human health risk-based guidelines
Recreational and commercial fishing are important economic and cultural activities in the NY
Bight Apex. Active fishing (recreational and commercial) still continues in the HARS vicinity
(Figure 2). The model proposed for human health risk assessment explicitly assumes that finfish
consumption by recreational anglers is the pathway of concern for human exposure from the
HARS. Issues of concern with this proposed model are:
- cumulative exposure to contamination by finfish and crustacea,
- protections for the general public and/or high risk individuals who may be exposed to fish
exposed to contaminant levels at or to be placed at HARS, and
- target species of highest exposure.
The following section will discuss these points and their importance for consideration when
developing guidelines for HARS cap material to protect human consumers of seafood from the
NY Bight Apex.
2.3.1 Cumulative exposure to contamination: finfish and crustacea
Motile species and contaminant exposure: Pre-remediation HARS had elevated levels of
particular contaminants of concern. Specifically, one reason for the site's designation as a
remediation site was on-going contamination (of PCBs and the dioxin 2,3,7,8-TCDD) in local
lobster stocks (USEPA, 1997). Additionally, there were elevated levels of particular
contaminants in benthic animals. The goal of deriving human health risk-based guidelines for
benthic animal tissue is to ensure that sediments used for remediation will not persist or create
unsafe levels of food chain contamination leading to unsafe levels in seafood.
One challenge is that finfish and many large crustacea (i.e., lobsters) are motile and therefore act
as "integrators" of exposure, integrating contaminant exposure over time. For example, lobsters
migrate inshore to shallower water in the spring and summer and migrate to the offshore in early
winter. However there is evidence that a subpopulation may reside and spawn on continental
shelf areas instead of migrating to inshore areas and another subpopulation may reside inshore
and conduct only localized inshore migrations (Uzmann et al., 1977 as cited in EPA 1997).
According to NOAA (1996 as cited in USEPA, 1997), total PCB (ZigPCB) concentrations in
muscle and hepatic tissue were elevated in lobsters from three sampling areas closest to the
10
-------
historical mud dump site (now within the HARS area, Figure 1; endnote 4) in comparison to the
Hudson Shelf Valley, and this spatial trend was consistent over time. Concentrations ranged
from 1.8 to 7.4 ppm (wet weight) in the hepatic tissue (NOAA, 1996 as cited in USEPA, 1997).
Similar patterns attributable to HARS contamination were not clearly evident for finfish species,
however there are ambient elevated levels of contamination in particular species, and fish
consumption advisories have been issued by New York and New Jersey for some species (e.g.,
striped bass based on size, bluefish based on size) caught in the NY Bight Apex area.
The current proposals for establishing risk-based guidelines do not account for ambient
background levels of contamination. This approach is inappropriate and inadequate for selecting
HARS cap material. It should be realistically assumed that fish have concentrations of
contaminants prior to entering and foraging at HARS and that fish foraging within HARS are
likely to transit out of the site. Thus, contaminant levels accumulated at HARS would compound
any levels already bioaccumulated by species prior to foraging within HARS. Ambient
background levels of contamination in the lobster and/or finfish populations should be
incorporated into the risk models to ensure that post-remediation conditions contribute to a
reduction of PCBs in these food-chains. Any model that does not account for this cumulative
exposure (bioaccumulation due to HARS cap material plus background exposure) will not
adequately protect human consumers who catch and consume these seafood species.
Accounting for fish movement: As discussed above, any model applied for the purposes of
selecting HARS cap material must protect consumers who consume fish that have foraged within
HARS. These consumers may be both those eating fish caught within HARS boundaries as well
as those fish caught outside of HARS boundaries. This concern applies to both recreational and
commercial fisheries. Recreational fisheries include species such as summer flounder, winter
flounder, and bluefish that may commonly forage in HARS. Commercial fisheries include these
species as well as lobster.
Current proposals (by EPA and by other members of the Remediation Material Workgroup)
focus on inclusion of types of "spatial correction factors" (e.g., site use factor in EPA's proposal)
to account for the fact that motile seafood species move in and out of HARS. However, at this
time, spatial models of fish movement are not adequate to account for various exposures
resulting from the various spatial ranges of individuals and populations. Fish movement can
widely vary throughout the fish's range. For example, tagging data from the American Littoral
Society for bluefish (see endnote 5) demonstrate that some individual fish travel great distances
within a short period (~1 month) of time whereas other individuals may remain in the same
vicinity during the same or longer periods of time (+ 1 month). These movements are difficult to
quantitatively model and predict for any given population. Furthermore, spatially explicit and
individual-based models attempting to estimate population exposures have not been
operationally validated for their predictive abilities, supporting a more conservative approach
(see section 2.4 for discussion of operational validation).
To our knowledge, no spatial model (such as the types proposed within the Remediation material
Workgroup) of fish movement has been used for this type of regulatory purpose to date in the
United States. Challenges in validation make general spatial models unreliable for use in
regulatory decision-making. For example, one proposed spatial model for finfish suffers from
ll
-------
lack of conceptual and operational validation (i.e., it has not been adequately tested for accuracy
in representing real spatial relationships and individual and population level effects, it includes
assumptions that are difficult to defend, and it has not been adequately tested for accuracy in
prediction). The accuracy and utility of spatially explicit models for determining contaminant
exposures in fish is unknown, making them ineffective and unreliable management tools. See
section 2.4 for discussion of conceptual and operational validation terms.
More straightforward approaches for protecting humans against cumulative exposure risks can
include the adoption of a human health risk-based benthic tissue level for the entire NY Bight
Apex. One way to achieve this regional perspective is to use a "site use factor" of one (1) and
not adopt any spatial models of fish movements that attempt to separate HARS from its NY
Bight Apex ambient environment. This approach could also work towards a remediation goal of
reducing contamination in Bight-wide resources (e.g., PCB contamination in local lobster
stocks). Any models that predict fish movement for purposes of this exposure assessment should
be reviewed by experts in the field of spatial modeling in aquatic systems.
2.3.2 Protections for the general public and/or high risk individuals who may be exposed to
fish exposed to contaminant levels at or to be placed at HARS
Some significant factors make the currently proposed sub-model deficient for HARS human
health risk-based guidelines. The proposal assumes a general population consumer type and
does not appear to differentiate between consumers of higher risk and lower risk. This is
concerning because the approach is not consistent with many fish consumption advisory
approaches, which typically differentiate between high and low risk consumer groups. High-risk
consumer groups include pregnant and nursing women and children. As summarized in the EPA
peer review document (appendix chapter on "Identification of Target Population and Estimation
of Seafood Consumption Rate), recreational fishermen consume their catch at home, meaning
that children are also likely exposed to those fish. Human health risk-based guidelines
developed for an adult (as proposed, 70 kg) will not sufficiently protect children (average 14.5
kg as provided in USEPA, 1995). For certain site-specific situations, EPA Office of Water
recommends that states may use body weights less than 70 kg for women and children to
calculate fish consumption advisory screening values for carcinogens and noncarcinogens (in
edible tissue) in order to protect the health of high risk subpopulations (USEPA, 1995). Human
health risk-based guidelines for benthic tissue at HARS should thus include protections for
pregnant and nursing women and children.
Furthermore, the proposal assumes that only recreational fishermen who fish in HARS will be
affected. The proposal does not consider that commercial catches are brought to New York city
and broader markets for general distribution, and as discussed above, recreational fishermen
share their catch. The proposed 10"4 cancer risk level is unacceptable and not consistent with
approaches used for other risk-based guidelines, which utilize a 10"5 or 10"6 cancer risk level.
For example, EPA Office of Water recommends that a risk level of 10"5 be used to calculate fish
consumption advisory screening values (for carcinogens in edible fish tissue) for the general
adult population (USEPA, 1995). For certain site-specific situations, EPA Office of Water
recommends that states may use a 10"6 or 10"7 cancer risk level (USEPA, 1995).
12
-------
Since both recreational and commercial fishing occurs in the HARS vicinity and no fishing bans
are planned for during-or post-remediation conditions at HARS, it should be assumed that fish
foraging within HARS will be distributed to a wide consumer group including sensitive
subpopulations. The appropriate risk level should be 10"6. The adoption of conservative and
precautionary assumptions for other model parameters in lieu of changing the current risk level
(10"4) must not be considered equivalent to the adoption of a more protective risk level.
2.3.3 Target species of highest exposure
Finfish: The current proposal by EPA uses a variety of finfish species to estimate site use and
other factors. Alternative proposals include the use of target fish species, such as winter
flounder. While winter flounder is an important recreational and commercial fish, it may not be
a protective estimate of human exposure risks from contaminants passed through the food chain.
Thus, while this species may be one acceptable target species for ecological protections it does
not provide a protective estimate of risks to which human consumers of NY Bight Apex seafood
are exposed.
For example, the flounder has low lipid
content and doesn't accumulate
contaminants to as high levels as a species
such as bluefish, striped bass, or lobster (see
Table II). In 1993, NOAA conducted a
survey to collect and analyze four common
recreationally caught fish species (bluefish,
summer flounder, black sea bass, tautog)
during the fall (September to December)
when fish physiological condition and lipid
levels are highest. A range of contaminants
was evaluated, including PCBs. This survey
demonstrated that species with strikingly
different lipid contents (for example,
summer flounder= 0.55% wet weight ± 0.28
(SD), n=14; bluefish= 3.68% wet weight
± 3.69 (SD), n=14), have very different PCB
loads. Ninety-one percent of summer
flounder sampled had congener
concentrations that were below the
Minimum Detection Level (MDL) whereas
twenty percent of bluefish sampled had
congener concentrations below the MDL.
Additionally, trends in PCB concentrations
among species followed the trends in lipid
Table II. Total Polychlorinated Biphenyl
(PCB) concentrations and lipid content
in bluefish and summer flounder. Data
are reported as they are presented in
NOAA, 2000. Bluefish and summer
flounder were surveyed in fall/early winter
1993 in the NY Bight. Trends in PCB
concentrations followed trends in lipid
content for bluefish (r=0.724, p=0.003,
n=14).
bluefish
summer
flounder
sample size
14
lipid content (% 3.68
wet weight)
EPCBs, mean
(ppb, wet
weight)
2x I™ PCBs,
mean (ppb, wet
weight)
±3.69
(SD)
3681158
(SD)
624 ± 260
(SD)
14
0.55 ± 0.28
(SD)
< MDL t
< MDL f
t MDL= minimum detection limit; EPCBs= 47.1
ppb; 2x Lis PCBs= 73.2 ppb
13
-------
contents and PCBs were correlated with lipids in bluefish (r=0.724; p=0.003; n=14; see Table
II).
The EPA Fish Contaminant Workgroup (as cited in USEPA, 1995) outlined specific factors for
selecting target species, including that species should have the potential to bioaccumulate high
concentrations of chemical contaminants. For HARS, these priority species would include
bluefish, striped bass, tautog, sea bass, and lobster.
Lobster
Lobster must be considered as a target species, since (1) it is the most important crustacean
fishery in the NY Bight Apex, (2) commercial harvesting occurs in the HARS vicinity, and (3) it
was specifically mentioned in the HARS remediation plan as a seafood species of concern.
There are no current or planned fishing restrictions for during- or post-remediation HARS.
Seafood consumption advisories of "no hepatopancreas consumption" have been placed on
lobsters caught in the NY Bight Apex area by New York and New Jersey states, although these
advisories are largely ineffective. Lobsters are therefore a most appropriate species to be
included as a target species- the goal of reducing PCBs in lobsters in and around HARS requires
use of lobsters as a target species. The human-health risk guideline model would need to be
significantly altered to account for consumption of the hepatopancreas to ensure that PCB
contamination in lobsters and risks from HARS are indeed reduced and no new risks are created.
Molluscs
Other appropriate target group may include bivalve molluscs. Mollusc fisheries in the HARS
vicinity include those for hard clams and scallops. There are no planned shellfishing
prohibitions for during- and post-remediation HARS, so any sediments placed must also protect
for these fisheries. Similar to the case of lobster consumption, the human-health risk guideline
model would be significantly altered to account for uptake of water, sediments, and plankton
and, most importantly, the consumption of the whole organism (in the case of clams/
2.4 Framework for assessing twenty-eight (28)-day bioaccumulation bioassays and
protecting against adverse effects: validation of the overall framework
In order for the final decision for selecting HARS cap material is protective and accurate, the
proposed framework should be tailored to the overall goals of the remediation program, be
consistent in its application for a wide variety of sediment types, and have protective and
accurate outputs. The term "validation" implies that a model, within a designated range of
applicability, possesses a satisfactory range of accuracy (e.g., Sargent, 1984 and Curry et al,
1989 in Rykiel, 1996). The process of validating a model can be separated into three main
components: operational, data, and conceptual validation (as reviewed by Rykiel, 1996). These
components can be used when determining the applicability and suitability of the proposed
methods for assessing risk from 28-day bioaccumulation bioassays for the purposes of evaluating
proposed HARS cap material.
Each type of validation (i.e., operational, data, and conceptual validation) describes a different
aspect of models that may be of use to prediction. Operational validation is primarily concerned
14
-------
with how well the model can predict the system output and does not attempt to validate
mechanisms built into the model. In this process, if the output corresponds with observed data,
then the model is operationally validated; however, operational validation does not evaluate the
data against which the model output is compared {i.e., it does not ensure that the data used to
operationally validate the model are appropriate measures). Data validation is the process by
which the data from the real system (and interpretations) are justified as being appropriate testers
of the model predictions. Conceptual validation asks if the parameters and functions within the
model accurately represent relationships and processes occurring in the natural system.
Additionally, because modeling is an abstraction of the real world, conceptual validation
includes the justification of why known ecological processes have been simplified in the model.
It is important to recognize that a model may be operationally correct yet conceptually incorrect,
and vice versa. Also, a model may be deemed operationally incorrect if the data validation is
unsatisfactory.
For the purposes of application such as HARS remediation, it is imperative that conceptual and
operational validations of the proposed models are performed for a wide range of sediment types
{e.g., predominantly silt/clay to predominantly sand/gravel). For example, the proposed
framework does not take into account that predators eat benthic prey whole, including gut
contents. The proposed trophic transfer model implicitly assumes a depuration period since all
bioaccumulation assays require a 24-hour depuration period prior to tissue analysis
(USEP A/US ACE, 1992). Thus, the trophic transfer proposals may be conceptually and
operationally invalid since they do not account for gut contents. Another example of a possible
invalid proposal is the exclusion of lobster as a target species. Given concerns about on-going
contamination in lobsters from HARS and in the NY Bight Apex, the proposed models are
conceptually and likely operationally invalid for the purpose of HARS remediation.
Furthermore, it is ineffective to have a conceptually strong model that can not be or is not
operationally validated for the regulatory purposes of selecting HARS cap material {e.g., using a
depurated animal model may not achieve the regulatory goal of accurately characterizing risks
from sediments placed in the NY Bight Apex as remediation material). In cases when data are
not available to operationally validate the proposed sub-models or uncertainty is too great, data
needs should be clearly stated, precautionary assumptions made, and other sediment assessment
methods and interpretive tools added when needed to ensure protection of benthic assemblages,
the food chain, and human health.
3. Conclusion
To ensure that post-remediation HARS conditions will not pose a threat to the food chain,
recommendations for the proposed models are as follows (listed in order as the recommendations
appear in this document):
- Dixon-like coplanar PCBs should be included without delay in assessments for HARS to
protect against future harm at the site and in the NY Bight Apex fisheries resources.
- Alkylated PAHs and organotins should be added to the analyte list for protecting ecological
and human health.
15
-------
Total mercury should be the analyte for assessing mercury effects to the ecology and human
health, and methylmercury should not be considered a dependable substitute.
- Limitations to the steady state models must be recognized. Any steady state adjustment
should err on the most conservative side (for example, provide a conservative "safety
factor").
- Trophic transfer factors must protect benthic prey, various life stages of benthic and pelagic
species in HARS, and top level predators, and represent high exposure risk species.
- A model for extrapolating benthic tissue concentrations to higher trophic levels must include
considerations of gut content, since predators of benthic prey are non-selective tissue feeders
and will consume the whole prey item.
- Models for human health risk-based guidelines should account for cumulative exposure
(bioaccumulation due to HARS cap material plus background exposure) to adequately
protect human consumers who consume finfish and lobster, caught within or outside HARS.
- A benthic tissue level to ensure protection of seafood consumers for the entire NY Bight
Apex should be adopted. One way to achieve this regional perspective is to use a "site use
factor" of one (1) and not adopt any spatial models of fish movement. If any spatial model is
considered, it should be reviewed for this specific regulatory application by experts of spatial
modeling in aquatic systems.
- Human health risk-based guidelines should include protection for pregnant and nursing
women and children (by including a risk-based level that includes a lower consumer body
weight for assessment of carcinogens and noncarcinogens in fish tissue).
- A cancer risk level of 10"6 must be used.
- If target seafood species are considered for a human health risk-based guideline, species that
have the potential to bioaccumulate high concentrations of chemical contaminants should be
considered. For HARS, these species may include bluefish, striped bass, tautog, sea bass,
and lobster. Another appropriate target group is bivalve molluscs.
- Given the concern of on-going contamination throughout the NY Bight Apex lobster
populations and contaminant contributions from HARS, lobsters must be considered in the
human health risk-based models and guidelines.
- If target species are used, then concepts, assumptions, and validation issues within the
models for trophic transfer and steady-state should be re-considered to ensure that they are
suitable for those specific species.
- For the purposes of application such as HARS, conceptual and operational validations for the
proposed ecological and human health models should be performed for a wide range of
16
-------
sediment types. Where data are not available to operationally validate the proposed models,
these data needs should be clearly stated, precautionary assumptions made, and other
sediment assessment or interpretive methods added.
In conclusion, the HARS is a remediation site that must receive cap material to reduce impacts at
the site and protect the NY Bight Apex environment. Assessment methods for selecting cap
material must be tailored to the goals of reducing impacts and not perpetuating or creating new
impacts. This paper and the above listed recommendations reflect concerns about the proposed
methods for developing risk-based guidelines and suggest ways to address these concerns.
Endnotes
1 The Marine Protection Research and Sanctuaries Act (MPRSA, 40 C.F.R. 228.10(b)) states that:
"The following types of effects, in addition to other necessary or appropriate considerations, will be
considered in determining to what extent the marine environment has been impacted by materials
disposed of at an ocean disposal site:
(1) Movement of materials into estuaries or marine sanctuaries, or onto oceanfront beaches, or
shorelines;
(2) Movement of materials toward productive fishery or shellfishery areas;
(3) Absence from the disposal site of pollution-sensitive biota characteristic of the general area;
(4) Progressive, non-seasonal changes in water quality or sediment composition at the disposal site,
when these changes are attributable to materials disposed of at the site;
(5) Progressive, non-seasonal, changes in composition or numbers of pelagic, demersal, or benthic
biota at or near the disposal site, when these changes can be attributed to the effects of materials
disposed of at the site;
(6) Accumulation of material constituents (including without limitation, human pathogens) in marine
biota at or near the site."
The MPRSA further states (40 CFR, 228.10(c)(1)) that effects of activities at a site be categorized as
Impact Category I when one or more of the following conditions is present and can reasonably be
attributed to ocean dumping activities:
"(i) There is an identifiable progressive movement or accumulation, in detectable concentrations
above normal ambient values, of any waste or waste constituent from the disposal site within 12
nautical miles of any shoreline, marine sanctuary designated under Title III of the Act, or critical area
designated under section 102(c) of the Act; or
(ii) The biota, sediments, or water column of the disposal site, or of any area outside the disposal site
where any waste or waste constituent from the disposal site is present in detectable concentrations
above normal ambient values, are adversely affected by the toxicity of such waste or waste
constituent to the extent that there are statistically significant decreases in the populations of valuable
commercial or recreational species, or of specific species of biota essential to the propagation of such
species, within the disposal site and such other area as compared to populations of the same
organisms in comparable locations outside such site and area; or
(iii) Solid waste material disposed of at the site has accumulated at the site or in areas adjacent to it, to
such an extent that major uses of the site or of adjacent areas are significantly impaired and the
Federal or State agency responsible for regulating such uses certifies that such significant impairment
has occurred and states in its certificate the basis for its determination of such impairment; or
(iv) There are adverse effects on the taste or odor of valuable commercial or recreational species as a
result of disposal activities; or
17
-------
(v) When any toxic waste, toxic waste constituent, or toxic byproduct of waste interaction, is
consistently identified in toxic concentrations above normal ambient values outside the disposal site
more than 4 hours after disposal."
2 Long et al. 2001 report that 10-day amphipod acute toxicity test results for New York Harbor sediments
poorly correlated with benthic health, as measured by benthic indices. The study found that relationships
between acute toxicity and changes to the benthos were poorest in large, multiestuary systems such as
New York/New Jersey Harbor. Specifically, none of the indices of benthic community abundance or
diversity agreed with amphipod survival rates in New York/New Jersey Harbor.
3 "Because most mercury in fish and shellfish tissue is present as methylmercury (NAS, 1991; Tollefson,
1989) and because of the relatively high analytical cost for methylmercury, it is recommended that total
mercury be determined....This approach is deemed to be most protective of human health and most cost-
effective."; USEPA, 1995.
4 USEPA (1997) concluded that even though the HARS is in close proximity to the Hudson River Plume
(which may also contribute contaminant loads to the lobster population), the HARS is likely an important
contributor to the elevated PCB levels in area lobsters. The site is a foraging area for lobsters and
contains distinctly elevated PCB levels in sediments, benthic tissue, and lobsters taken from that area.
(USEPA, 1997).
5 American Littoral Society (ALS) manages and directs a fish-tagging program for recreational anglers.
Data are stored by ALS and National Marine Fisheries Service in Woods Hole, MA.
18
-------
References
Carls, M.G., S.D. Rice, J.E. Hose. 1999. Sensitivity of fish embryos to weathered crude
oil: Part 1. Low-level exposure during incubation causes malformations, genetic damage, and
mortality in larval pacific herring (Clupea pallasi). Environmental Toxicology and Chemistry.
18:481-493.
Curry, G.L., B.L. Deuermeyer, and R.M. Feldman. 1989. Discrete Simulation. Holden-
Day: Oakland, CA (as cited in Rykiel, 1996).
Gobas, F.A. 1993. A model for predicting the bioaccumulation of hydrophobic organic
chemicals in aquatic food webs: Application to Lake Ontario. Ecological Modelling. 69: 1-17.
Long, E.R., C.B. Hong, C.G. Severn. 2001. Relationships between acute sediment
toxicity in laboratory tests and abundance and diversity of benthic infauna in marine sediments: a
review. Environmental Toxicology and Chemistry. 20: 46-60.
NAS (National Academy of Sciences). 1991. Seafood Safety. Committee on Evaluation
of the Safety of Fishing Products. National Academy Press: Washington, DC. (as cited in
USEPA, 1995)
NOAA (National Oceanic and Atmospheric Administration). 1994. Quantitative Effects
of Pollution on Marine and Anadromous Fish Populations. U.S. Department of Commerce,
National Oceanic and Atmospheric Administration (NOAA), National Marine Fisheries Service
Northeast Region, Woods Hole, MA. NOAA Technical Memorandum NMFS-F/NEC-104.
NOAA (National Oceanic and Atmospheric Administration). 1996. Contaminant Levels
in Muscle and Hepatic Tissues of Lobsters from the New York Bight Apex. National Oceanic and
Atmospheric Administration, Northeast Fisheries Science Center, National Marine Fisheries
Service, Highlands, NJ. May 1996. (as cited in USEPA, 1997)
NOAA (National Oceanic and Atmospheric Administration). 2000. Contaminant Levels
in Muscle of Four Species of Recreational Fish from the New York Bight Apex. U.S. Department
of Commerce, National Oceanic and Atmospheric Administration (NOAA), National Marine
Fisheries Service Northeast Region, Woods Hole, MA. NOAA Technical Memorandum NMFS-
NE-157.
Rykiel, E.J. 1996. Testing ecological models: the meaning of validation. Ecological
Modelling. 90:229-244.
Sargent, R.G. 1984. A tutorial on verification and validation of simulation models. In: S.
Sheppard, U. Pooch, and D. Pegden (Eds.). Proceedings of the 1984 Winter Simulation
Conference. IEEE 84CH2098-2. pp. 115-122 (as cited in Rykiel, 1996).
Tollefson, Linda. 1984. Methylmercury in fish: Assessment of risk for U.S. consumers.
19
-------
In: The Risk Assessment of Environmental and Human Health Hazards: A Textbook of Case
Studies. Dermis J. Pasustenbach (ed.). John Wiley and Sons: New York, NY. (as cited in
USEPA, 1995)
USEPA (U.S. Environmental Protection Agency). 1997. Guidance for Assessing
Chemical Contaminant Data for Use in Fish Advisories. Volumes. 1 to 3. U.S. Environmental
Protection Agency, Office of Water. EPA 823-R-95-007. September 1995.
USEPA (U.S. Environmental Protection Agency). 1997. Supplemental to the
Environmental Impact Statement on the New York Dredged Material Disposal Site Designation
for the Designation of the Historic Area Remediation Site (HARS) in the New York Bight Apex.
U.S. Environmental Protection, Region 2, New York. May 1997.
USEPA/ USACE. (U.S. Environmental Protection Agency/ U.S. Army Corps of
Engineers). 1992. [draft] Guidance for performing tests on dredged material proposedfor ocean
disposal. U.S. Environmental Protection, Region 2, New York and U.S. Army Corps of
Engineers, New York District. December 1992.
Uzmann, J.R., R.A. Cooper, K.J. Pecci. 1977. Migration and Dispersion of Tagged
American Lobsters, Homarus americanus, on the Southern New England Continental Shelf. U.S.
Department of Commerce, National Oceanic and Atmospheric Administration, National Marine
Fisheries Service. NOAA Technical Report NMFS SSRF-705.
20
-------
I
Paper Submitted by
New Jersey Department of
Environmental Protection (NJDEP)
-------
New Jersey Department of Environmental Protection Staff Comments Offered for
Consideration of the Peer Reviewers of the "Proposed Changes to the
Bioaccumulation Testing Evaluation Framework and Response to Scientific Peer
Reviewer's Comments on the Existing Framework for Determining the suitability of
Dredged Material for Placement at the Historic Area Remediation Site (HARS)"
October 19,2000.
Comments prepared by Gary Buchanan, Ph.D., Lawrence Baier and Suzanne Dietrick
Please be advised that the following comments on the proposed changes to the
Bioaccumulation Testing Evaluation Framework (TEF) are submitted for consideration
by the peer review group. These comments are merely intended to discuss the scientific
merit of the proposed changes to the TEF. These comments are not intended to represent,
nor shall they be construed as a position by the New Jersey Department of Environmental
Protection concerning the proposed TEF. A final agency position requires policy
development with full consideration of both the environmental and economic impacts of
the proposal. Such an analysis has not been completed at this time, and would be
premature pending the completion of the peer review process.
Acceptable Risk Value
All three reviewers agree that the proposed TEF modifications are an improvement over
the previous version. Additional and more appropriate methods have been included which
should provide a more detailed and robust risk assessment of the bioaccumulation of
contaminants from sediments. However, the peer reviewers did indicate several areas of
uncertainty with the proposed Framework's methods. EPA lists and effectively discusses
these uncertainties in Appendix F. This appendix provides a thorough summary of the
weaknesses and concerns with the methods and assumptions used in the Framework.
In general, the TEF utilizes evaluation criteria which are within acceptable limits, but
which are at the lower end of the acceptable range. For example, acceptable risks for
cancer range from PIO"4 to 1*10"6. A policy decision was made to use the PIO"4 risk
level for HARS remediation material. This decision is not subject to peer review.
Similarly using an EC50 (a concentration at which 50 percent of a particular taxonomic
group are adversely impacted) as an ecological benchmark cannot be considered
praetorian. It must be recognized that the NJDEP routinely uses a risk level of 1 * 10"6 in
establishing its soil cleanup criteria (N.J.A.C. 7:26D). Given the policy decision to use a
risk level of 1 * 10"4, it is recommended that the EPA consider the use of more
conservative assumptions discussed below when calculating HARS specific values. We
believe that if the risk level remains at 1 * 10"4, then the use of less conservative
assumptions recommended by other representatives of the RMW Workgroup is not
appropriate. If EPA makes a decision to change the risk level to 1*10"6, then it may be
appropriate to take into consideration the less conservative assumptions recommended by
other RMW members.
-------
Fish Consumption Rate
The NJMSC report appears to be the best information presently available on fish
consumption. However, there is concern that the fish consumption rate may be
underestimated by the survey. We suggest a site-specific assessment in the future as a
means of removing uncertainty concerning fish consumption. (Should EPA decide to
undertake such a study the Department's Bureau of Risk Analysis would be interested in
participating, e.g., assist in the design of the survey.) It is unclear whether the amount of
saltwater finfish known to reside at the HARS should be reduced by the 40% of fish not
prepared in the home. Of all the species reportedly consumed, it seems logical to assume
that inshore species are more likely to have been caught by recreational anglers and
prepared than offshore species (tuna for example). Since the offshore species have
already been discounted because they do not reside at the HARS, the reduction may be
accounting for these species twice. To preserve the P10"4 risk value, we suggest that the
11.34 g/day attributable to HARS resident fish be used in the calculation of risk.
Site Use Factor
In considering whether a site use factor is appropriate at all, we are compelled to remind
the peer reviewers that the goal of placing material at the HARS is to reduce impacts of
historical disposal to acceptable levels, not to designate a new dredged material disposal
site. Material to be used for this purpose must not cause significant undesirable effects
including through bioaccumulation. Much of the discussion surrounding the site use
factor assumes a null value for exposure when fish are outside of the HARS. This
effectively dilutes the bioaccumulation effect of the material placed at the HARS. Given
that the material and not the setting bears the burden of demonstrating no significant
undesirable effects it seems that a constant exposure would be the most appropriate
yardstick. Beyond that, the assumption that there is a zero exposure to contaminants
outside of the HARS cannot be validated (see EPA reported background values in table
on page 144 of response to peer reviewers and MDS/HARS SEIS). In fact those species
which migrate into the estuary for part of their life cycle may be subject to greater
exposure than that represented at the HARS. At most, it may be appropriate to
incorporate a site use factor that accounts for species documented to seasonally migrate to
the outer continental shelf or beyond, where exposure is assumed to be minimal.
However, if home ranges and seasonal use are to be used to reduce the duration of
exposure within the HARS, then the exposure to contaminants outside of the HARS must
be factored into the risk assessment to ensure that the integrity of the minimum risk level
is maintained.
-------
Whole Body to Fillet Factor
As noted in the Technical Topics Summary for the RMW meetings held on January 10,11
and 12, one reviewer supports a reevaluation of the proposed changes to the whole body
to fillet factor for metals. The Bevelhimer study cited in Response to Comment #7 as the
basis for the change in the factor indicates that the ability to predict whole body
concentrations from concentrations in the fillet was limited for a number of inorganic
compounds.
Proposed Trophic Transfer Factor for Hg
The trophic transfer factor of 1.95 is an improvement over the prior value (i.e., 1).
However, this may still underestimate the biomagnification of mercury from one trophic
level to the next. In EPA's Mercury Study Report to Congress (1997, Volume HI,
Appendix D) predator-prey factors ranged from 5.0 to 8.1 for trophic level 2 fish, and
from 2.6 to 15.5 for trophic level 3 fish. This information should be reviewed for
assumptions regarding mercury bioaccumulation/biomagnification by fish.
Inclusion of Co-planar PCBs
It is well documented that co-planar PCBs exhibit dioxin like effects. The fact that EPA
is conducting a dioxin reassessment does not change this fact. The potency of these
compounds is sufficient cause for heightened concern. Therefore, to ensure that
remediation material does not produce a significant undesirable effect, the effects of co-
planar PCBs cannot be ignored.
PAH and PCB Toxicity Through Narcosis
The position proffered by Nation's Port is that PAHs and PCBs are being double counted
by considering both their contribution to specific modes of action and narcosis. The fact
that PAHs exhibit toxicity by specific modes of action at lower concentrations than they
exhibit narcosis alone does not mean that PAHs at lower levels will not contribute to
narcosis in combination with other contaminants. This same argument holds true for
PCBs, which may contribute to a narcotic effect at lower concentrations when combined
with compounds with similar modes of action. We believe that the approach suggested
by in the TEF appropriately considers the contribution of PAHs and PCBs to these
various effects.
Similarly, Nation's Port asserts that due to the wide range of PAH residues that exhibit
toxicity in aquatic species a tissue residue approach for defining acceptable levels of
PAHs is inappropriate. Toxicity induced by PAHs is an unacceptable effect and therefore
the risk associated with PAHs should be evaluated using the best available scientific
information.
-------
HARS Specific Values Based on Ecological Effects
EPA is encouraged to periodically review the availability of new data sets involving more
appropriate species to eliminate reliance on species that are less representative of the
HARS (e.g. freshwater species) due to variable exposure conditions and physiology that
affect the body burden at which adverse effects occur in establishing HARS specific
values. Perhaps freshwater species should not be considered when data on saltwater
species are available which are assumed to be more representative of taxonomic groups
that actually reside at the HARS.
Page 12, Section C, Consideration of Potential Ecological Effects - A paragraph should
be added to this section to provide more details as to how the ecological effects values
will be used by the EPA in the evaluation of dredged material proposed for use at the
HARS. Please refer to the discussion found in the last paragraph in Appendix F, Page 37,
item f.
Combined Effects Evaluations
Page 17, sections A, B, and C: The phrase ".. .the material may not be suitable for use as
Remediation Material" is used in each section to describe comparisons exceeding
thresholds (e.g., hazard index >1). If the intent is to use these thresholds as triggers, than
these sections should be reworded to say "...the material would not be suitable for use as
Remediation Material." Otherwise, additional explanation should be added to the
Framework document if supplementary examination of the data will occur subsequent to
the Combined Effects Evaluations.
-------
Paper Submitted by
Surfers Environmental Alliance (SEA)
-------
FINAL DOCUMENT
To: HARS Remediation Material Workgroup, EPA Region II
From: Kevin ODriscoll, Ph.D, Science Consultant to the Surfers Environmental Alliance
Re.: August 2001, Scientific Review of the Charges to Peer Reviewers of the Human
Health-Related Aspects
The intent of this statement is to provide critical guidance for the extended scientific
review of the Bioaccumulation Testing Evaluation Framework and Response to Scientific
Peer Reviewers Comments.
The Historic Area Remediation Site (HARS) presents a unique opportunity to develop
new strategies to cap a historical dump site with remediation material that will reduce
impacts, including those of elevated contaminant levels as outlined in the Supplemental
Environmental Impact Statement prepared for the HARS (EPA 1997). The current
charge to the peer reviewers to define the framework used for evaluating bioaccumulation
test results towards the selection of dredged material to be used in the remediation
process.
The designation of HARS is a precedent, since it is the only ocean dump site categorized
as a remediation site. Since its designation, capping has occurred with non-peer reviewed
guidelines for selection of cap sediments. One immediate and obvious problem with this
approach is that it is difficult to define what measures would reduce negative impacts on
human health from the HARS if those impacts were not first well understood. The
proximity of the HARS to a major metropolitan population, and its extensive use as a
commercial and recreational fishery both argue that conservative assumptions be used
throughout the risk assessment of the HARS. Thus there are numerous sources of
uncertainty in this framework, many of which may be addressed successfully in the
ongoing development of risk-based guidelines, site remediation, and site monitoring
plans. In addition, there are further complicating factors, including the need for all
parties to agree upon an appropriate mechanism by which the scientific review process
can be used effectively to inform policy decisions.
Implicit in this charge is the determination of the appropriate cancer and non-cancer risk
levels for both the general population of the NY/NJ metropolitan area as well as for that
sector of the population most at risk. The proposed use of a cancer risk level of 10E-4
for these risk-based guidelines is contrary to the EPA guidelines for the use of
conservative assumptions. Given the unique aspects of the HARS, the latter
determination is arguably not merely a "policy issue" but a scientific issue that is
determined by relative impact on a large population, and as such should be fully
considered at the level of scientific expertise in public health. Although the proposed use
of the 10E-4 risk level is nominally consistent with EPA's recognition of cancer risk
levels between 10E-4 and 10E-6 it is not conservative given the large population at risk.
In addition, the policy implementation of a cancer risk level of 10E-4 in the HARS risk-
based guidelines is unacceptable from a scientific standpoint. Specifically, its use would
-------
not fully ensure that that the actual cancer risk is within EPA guidelines, given the
considerable uncertainty in the parameters of the risk assessment.
It is significant in this regard that the EPA has published fact sheets such as
"Polychlorinated Biphenyls (PCBs) Update: Impact on Fish Advisories" (EPA 1999) that
specify the use of 10E -5 as a maximum acceptable cancer risk level in calculating
monthly fish consumption limits for PCBs, as well as for other contaminants at the
HARS including dioxin. According to EPA's guidelines any consumption of fish with
greater than 97 ppb PCBs is unacceptable for this cancer health endpoint. EPA Region II
currently recognizes a level of 113 ppb PCBs as an upper bound limit for
bioaccumulation in test organisms assayed using sediments to be used as remediation
material. Therefore, the upper limit of PCB levels for bioaccumulation for the HARS is
virtually identical to the upper limit for acceptable finned fish consumption assuming a
10E -5 cancer risk level. Since the charge to the peer reviewers suggests a trophic
transfer factor of 3 for total PCBs, it follows that actual cancer risk from human
consumption of ANY fish (including lobster) exposed to PCB contamination at the
remediated site is going to be above the 10E -5 level. The EPA-sponsored formal risk
assessments of the Hudson River Superfund sites found risks from PCB exposures due to
the consumption of fin fish caught in the Hudson River to be within this range. Taken
together, these considerations suggest that if similar assumptions are used in developing
the risk-based guidelines for selection of HARS remediation material, that upper limit
values for PCBs should be set at significantly lower levels than what EPA is currently
specifying, i.e. less than 113 ppb total PCBs.
In developing the risk-based guidelines for the HARS, EPA Region II proposes to use a
recreational angler as the target population. Using this target population ignores several
important facts. Firstly, the HARS lies within an area that is among the most intensive
commercial fisheries in the world. Therefore, members of the population at large, which
is more than 20 million individuals in the NY Metro region, are likely toconsume any one
of a variety of species that are exposed to toxic chemicals by foraging at the HARS.
Secondly, it ignores the more highly-impacted sub-population of subsistence and partial
subsistence fishermen, which is likely to include a significant number of individuals.
Finally, the recreational angler scenario ignores the fact that family members of the
fishermen are similarly impacted when meals are shared, and that these family members
include children and pregnant women. Given these considerations, a 10E -6 cancer risk
level is justified by scientific considerations of negative public health impacts on a much
larger population than merely the recreational anglers proposed by EPA. While these
considerations may not be included in the present charge to the peer reviewers, this issue
is clearly important in policy considerations.
The initial peer review of the HARS designation determined that there were fundamental
problems in several aspects relevant to human heath impacts. These are addressed in the
Proposed Bioaccumulation Evaluation Process document. The Dredged Material
Management Group at EPA Region II has attempted to define key parameters for a
deterministic approach to the human health risk assessment of the HARS. The Proposed
Bioaccumulation Evaluation Process document presents independent peer reviewers with
-------
Specific Concerns:
1.) Hazard Identification: Coplanar PCBs should be considered a dioxin-like
compound(s) for the purposes of the risk assessment since this is well documented in the
literature. Mixtures of chemicals in various dredged materials should be considered in
terms of potential synergy or other interactions between various genotoxic carcinogens,
and/or non-genotoxic classes of carcinogens in terms of the risk assessment. Further
chemical characterization of the existing sediments at the HARS and their exact
distributions are required in order to effectively remediate the site.
2.) Bioaccumulation Testing: A meaningful statistical assessment of each
bioaccumulation assay should be required including baselines for detection limits
determined empirically. The data set supporting the extrapolation to steady state from a
30 day assay need to be expanded and shown to be reproducible over a 120 day time
course using variable sediments and organisms. Chronic endpoint toxicity testing should
be performed for all test species.
3.) Fish Consumption Rates: Consideration of impacts of commercial fisheries should be
made in determining risk-based guidelines for the HARS. A conclusive survey of general
population seafood consumption, recreational fishermen, and subsistence fishermen
should be performed. A better understanding of body loading and clearance rates for
various suspect and known human carcinogens is required for accurate toxicokinetic
estimations.
4.) Risk Estimates: Quantitative limits of credibility should be expressed for human
health and ecological risk impacts. Conservative, protective assumptions should be
utilized throughout the food chain modeling procedures. Target human subpopulations
and acceptable risk levels need to be identified using conservative estimations, e.g.
children and nursing infants. Exact values of, and methods for risk-based
determinations of, acceptable contaminant levels in potential remediation materials need
to be subjected to ongoing scientific review.
-------
defined questions that attempt to reduce uncertainties in certain key aspects of the HARS
framework for selection of remediation material. Further definition of these key
parameters, which include the mechanisms of action of certain contaminants, the methods
by which they are assayed, and assumptions involved in modeling the human food chain,
will make possible more reliable risk-based assumptions for the HARS.
One over-arching theme in the peer review process has been the suitability of a
probabilistic risk assessment approach in contrast to the EPA's deterministic approach to
setting risk-based guidelines for the HARS. In response to comments by the initial
reviewers, several presenters at the meeting proposed methodologies to introduce
probabilistic values into the risk assumptions. Specifically, probabilistic models for
trophic transfer and for fish foraging are included in the charge to peer reviewers. These
proposals suggest a significant departure from the deterministic approach presented by
the EPA. While these methods have limited theoretical advantages over a strictly
deterministic method for modeling human health risk assessments, they are not at this
time suitable for implementation. This conclusion is warranted because of the need for
consistency throughout a given risk assessment.
A probabilistic risk assessment approach would require that all parameters be determined
within defined ranges of accuracy using validated modeling approaches and testing
methodologies. For instance the statistical methods used to address population
variability would have to be designed to avoid a given parameter becoming erroneously
limiting because it is put in the wrong context. An example of this would be to assume
that site use by foraging fish is a limiting factor while at the same time ignoring the fact
that many fish are foraging in adjacent or distant areas where there is also contamination.
In this example, food chain models such as that proposed by Linkov et al. (Linkov et al.
Ms. enclosed) assume that foraging outside the site is free from exposure to
contaminants. This may in fact be an erroneous assumption given the extensive
industrial contamination in the river estuaries and in the general benthic environment of
the NY bight apex. Examples such as this suggest that uncertainty in various parameters
both within and outside the study area must be fully considered. Furthermore, and more
troubling perhaps, a fuller consideration of the numerous sources of uncertainty in
developing risk-based guidelines for the HARS suggests the likelihood that application of
probabilistic approaches, in the wrong context, could underestimate risk.
The Surfers'Environmental Alliance recommends that deterministic methods that
incorporate background levels of contamination in target seafood species that forage in
the HARS be applied using conservative assumptions to ensure that risk-based guidelines
are protective for human consumers. Finally, it must be emphasized that the HARS
designation and remediation are ongoing processes involving many economic and social
factors above and beyond the science. To best serve the various interests it is absolutely
critical that all valid scientific considerations are taken into account during policy
administration and implementation. The HARS remediation process in particular will
hopefully be subjected to continual peer review especially as regards a future site-
monitoring plan.
-------
Paper Submitted by the
Army Corps of Engineers (USACE)
§2
a
-------
MENZIE • CURA & ASSOCIATES, INC.
Environmental Consultants
One Courthouse Lane, Suite Two • Chelmsford, Massachusetts 01824
(978) 453-4300 • Fax (978) 453-7260
www.menziecura.com
MEMORANDUM
Date: September 13, 2001
File: 753B
To: Todd Bridges
From: Jerry Cura, Katie Cloonan
Subject: Fish Ingestion Risks
This memorandum provides some estimates of the possible magnitude of risk for people
consuming fish from the New York Bight Apex, based on exposure factors, toxicity factors, and
exposure equations in USEPA Region 2 (2001), and on chemical concentrations in fish muscle
tissue provide by NOAA in Deshpande et al. (2000). The actual risks may differ given
uncertainties in patterns of exposure and chemical toxicity.
Our assumptions included:
That people ingest only one fish species throughout a lifetime (we used the four species
provided by NOAA to calculate four separate estimates of cancer risk and noncancer
hazard, and recognize the artificiality in this assumption);
That average chemical concentrations provided for bluefish (muscle and skin with scales
removed), black sea bass (muscle and skin with scales removed), summer flounder
(muscle only) and tautog (muscle only) represent the exposure of individuals ingesting
fish (we recognize that the average derives from a relatively small sample size of 14
composites per species, and the risk calculations could benefit from a closer examination
of the variability associated with these data);
That the toxicity factors for each chemical in Table 5 of USEPA Region 2 (2001) are
appropriate (we did not conduct a separate toxicity review); and
That the exposure factors in USEPA Region 2 (2001) are appropriate (we did not
independently evaluate these factors).
-------
Note that:
The estimates are deterministic and probabilistic analyses could provide improved
estimates;
The estimates assume 100% bioavailability of chemicals;
The assessment estimates risk and hazard from exposure to those compounds listed in
Table 5 of US ACE (2001) unless they were ne\er detected or their average concentrations
were reported as less than the MDL;
NOAA collected the four fish species from various locations among fifteen total stations
(not all species were caught at all stations) in the New York Bight Apex. The stations
were off the New Jersey coast from Sandy Hook to Long Branch and were approximately
0.3 to 7.4 nautical miles offshore. NOAA selected the locations based on fishermen
popularity. The locations are approximately 1.7 to 9.2 nautical miles from the HARS.
One station, BL-1, was within the HARS, however, bluefish was the only species
collected at that location; and
We did not conduct an extended search for chemical data other than that provided by
NOAA.
The equations used to calculate risk and the toxiciy factors (cancer potency factors and reference
doses) are also consistent with USEPA Region 2 (2001). The equations for calculating risk and
hazard are:
, EPCxCPFxFIRxCFxEFxED
Cancer Risk and
BWxAT
.. tt j ^ ¦ EPC xFIRxCF xEF xED
Noncancer Hazard Quotient=
RfDxAT xBW
where:
EPC
= Exposure Point Concentration in Fish (mg/kg);
CPF
= Cancer potency factor (kg-day/mg);
FIR
= Fish ingestion rate (g/day);
CF
= Conversion factor (kg/g);
EF
= Exposure frequency (days/year);
ED
= Exposure duration (years);
BW
= Body weight (kg);
AT
= Averaging time (days); and
RfD
= Reference dose (mg/kg-day).
-------
Attachment 1 presents the average concentrations in fish tissue from Deshpande et al. (2000) that
were used as the EPCs in the cancer risk and noncancer hazard estimates. Attachments 2 and 3
present the exposure factors and toxicity factors from USEPA Region 2 (2001).
Deshpande et al. (2000) present total arsenic concentrations in fish tissue. However, the majority
of arsenic in marine fish is usually in the nontoxic organic form of arsenobetaine. USEPA
Region 2 (2001) assume in their calculations of trophic level transfer that 10% of the total arsenic
in tissue is inorganic. Scientific literature suggests that the percentage of inorganic arsenic in
marine fish tissue may be even lower. Donahue and Abernathy (1999) state that concentrations
of inorganic arsenic seldom exceed 4% of the total arsenic. Neff (1997) reports that inorganic
arsenic represents only 0.5 to 1% of the total arsenic in the edible portions of most marine fish.
Therefore, separate cancer risks and noncancer hazard indices were estimated assuming that
100%, 10%, and 1% of the total arsenic in fish tissue is inorganic.
Attachment 4 summarizes the noncancer hazard quotients for each compound and Attachment 5
summarizes the cancer risks for each compound. The total cancer risks and the total noncancer
hazard indices are the sums of the cancer risks and noncancer hazard quotients of individual
compounds.
The total noncancer hazard indices from fish consumption, assuming 100% inorganic arsenic,
range from 0.0009 for summer flounder to 0.004 for bluefish. The total noncancer hazard indices
assuming 10% inorganic arsenic range from 0.0003 for summer flounder to 0.003 for bluefish.
The total noncancer hazard indices assuming 1% inorganic arsenic rangefrom 0.0003 for
summer flounder to 0.003 for bluefish.
The total cancer risks from fish consumption, assuming 100% inorganic arsenic, range from
2E-04 for bluefish and tautog to 6E-04 for black sea bass. The majority of the risk for all species
is due to arsenic, however, total PCBs also contribute to the risk from consumption of bluefish.
The total cancer risks assuming 10% inorganic arsenic range from 4E-05 for summer flounder
and tautog to 1E-04 for bluefish. The majority of the cancer risk from consumption of bluefish is
attributable to total PCBs. For summer flounder, the risk is due primarily to arsenic, but total
PCBs also contribute. For black sea bass and tautog, the risk is attributable to arsenic and PCBs.
The total cancer risks assuming 1% inorganic arsenic range from 1E-05 for summer flounder to
1E-04 for bluefish. For all species, the majority of the risk is due to total PCBs, but arsenic also
contributes to risk from consumption of summer flounder.
References:
Deshpande, A.D., A.F.J. Draxler, V.S. Zdanowicz, M.E. Schrock, A.J. Paulson, T.W. Finneran,
B.L. Sharack, K. Corbo, L. Arlen, E.A. Leimburg, B.W. Dockum, R.A. Pikanowski, B. May, and
L.B. Rosman. 2000. Contaminant Levels in Muscle of Four Species of Recreational Fish from
the New York Bight Apex. U.S. Department of Commerce National Oceanic and Atmospheric
-------
Administration (NOAA) Technical Memorandum NMFS-NE-157. National Marine Fisheries
Service, Northeast Region, Northeast Fisheries Science Center, Woods Hole, Massachusetts.
Donohue, J.M. and C.O. Abernathy. 1999. Exposure to Inorganic Arsenic from Fish and
Shellfish. Arsenic Exposure and Health Effects. W.R. Chappell, C.O. Abernathy and R.L.
Calderon (Editors). Elsevier Science B.V.
Neff, J.M. 1997. Ecotoxicology of Arsenic in the Marine Environment. Environmental
Toxicology and Chemistry, Vol. 16, No. 5, pp. 917-927.
USEPA Region 2. 2001. Draft Scientific Peer Review Package and Charge: Proposed Changes to
the Bioaccumulation Testing Evaluation Framework for Assessing the Suitability of Dredged
Material to be Placed at the Historic Area Remediation Site (HARS).
-------
Attachment 1
Average Muscle Tissue Concentrations for Fish from the New York Bight Apex
Compounds (mg/kg)
Bluefish
Average
Concentration
Summer Flounder
Average
Concentration
Black Sea Bass
Average
Concentration
Tautog
Average
Concentration
Arsenic
0.49
1.72
3.65
1.02
Cadmium
0.165
0.12
0.145
0.109
Chromium
0.442
0.141
0.408
0.190
Copper
0.453
0.259
0.423
0.339
Lead
0.288
0.155
0.252
0.173
Mercury
0.1021
0.0358
0.0504
0.0811
Nickel
0.160
0.132
0.184
0.138
Silver
0.0411
0.0221
0.0353
0.0271
Zinc
11.4
3.69
4.83
4.28
Total PCBs*
0.624
0.041
0.136
0.098
alpha Chlordane
0.026
MDL but the mean was
-------
Attachment 2
Exposure Factors for Risk Calculatibns
Exposure
Factors
Fish Ingestion Rate
7.2 g/day
Conversion Factor
0.001 kg/g
Exposure Frequency
365 days/year
Exposure Duration
70 years
Body Weight
70 kg
Averaging Time
25,550 days
Reference:
All exposure assumptions are consistent
with USEPA Region 2 (2001).
Page 1 of 1
-------
Attachment 3
Toxicity Factors for Risk Calculations
Cancer
• Potency
Reference
Factor
Dose
Compounds
(kg-day/mg)
(mg/kg-day)
Arsenic
1.5
0.3
Cadmium
1
Chromium
3
Copper
37.1
Lead
1
Mercury
0.1
Nickel
20
Silver
5
Zinc
300
Total PCBs
2
0.02
alpha Chlordane
0.35
0.05
Total DDT
0.34
0.5
Acenaphthene
60
Benz(a)anthracene
Note:
All toxicity factors are from Table 5 in USEPA Region 2 (2001).
Page 1 of 1
-------
Attachment 4
Summary of Noncancer Hazard from Fish Consumption
Noncancer Hazard Quotients
Bluefish
Summer Flounder
Black Sea Bass
Tairtog
100% Inorganic
Arsenic
10% Inorganic
Arsenic
1% Inorganic
Arsenic
100% Inorganic
Arsenic
10% Inorganic
Arsenic
1% Inorganic
Arsenic
100% Inorganic
Arsenic
10% Inorganic
Arsenic
1% Inorganic
Arsenic
100% Inorganic
Arsenic
10% Inorganic
Arsenic
1% Inorganic
Arsenic
Arsenic
Cadmium
Chromium
Copper
Lead
Mercury
Nickel
Silver
Zinc
2E-04
2E-05
2E-05
1E-06
3E-05
1E-04
8E-07
8E-07
4E-06
2E-05
2E-05
2E-05
1E-06
3E-05
1E-04
8E-07
8E-07
4E-06
2E-06
2E-05
2E-05
1E-06
3E-05
1E-04
8E-07
8E-07
4E-06
6E-04
1E-05
5E-06
7E-07
2E-05
4E-05
7E-07
5E-07
1E-06
6E-05
1E-05
5E-06
7E-07
2E-05
4E-05
7E-07
5E-07
1E-06
6E-06
1E-05
5E-06
7E-07
2E-05
4E-05
7E-07
5E-07
1E-06
1E-03
1E-05
1E-05
1E-06
3E-05
5E-05
9E-07
7E-07
2E-06
1E-04
1E-05
1E-05
1E-06
3E-05
5E-05
9E-07
7E-07
2E-06
1E-05
1E-05
1E-05
1E-06
3E-05
5E-05
9E-07
7E-07
2E-06
3E-04
1E-05
7E-06
9E-07
2E-05
8E-05
7E-07
6E-07
1E-06
3E-05
1E-05
7E-06
9E-07
2E-05
8E-05
7E-07
6E-07
1E-06
3E-06
1E-05
7E-06
9E-07
2E-05
8E-05
7E-07
6E-07
1E-06
Total PCBs
3E-03
3E-03
3E-03
2E-04
2E-04
2E-04
7E-04
7E-04
7E-04
5E-04
5E-04
5E-04
alpha Chlordane
Total DDT
5E-05
3E-05
5E-05
3E-05
5E-05
3E-05
-------
Attac 15
Summary of Cancer Risk from Fish Consumption
Compounds
Cancer Risks
Bluefish
Summer Flounder
Black Sea Bass
Tautog
100% Inorganic
Arsenic
10% Inorganic
Arsenic
1% Inorganic
Arsenic
100% inorganic
Arsenic
10% Inorganic
Arsenic
1% Inorganic
Arsenic
100% Inorganic
Arsenic
10% Inorganic
Arsenic
1% Inorganic
Arsenic
100% Inorganic
Arsenic
10% Inorganic
Arsenic
1% Inorganic
Arsenic
Arsenic
8E-05
8E-06
8E-07
3E-04
3E-05
3E-06
6E-04
6E-05
6E-06
2E-04
2E-05
2E-0S
Cadmium
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
Chromium
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
Copper
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
Lead
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
Mercury
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
Nickel
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
Silver
NA
NA
m
NA
NA
NA
NA
NA
NA
NA
NA
NA
Zinc
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
Total PCBs
1E-04
1E-D4
1E-04
8E-06
6E-06
8E-06
3E-05
3E-Q5
3E-Q5
2E-05
2E-05
2E-05
alpha Chlordane
9E-07
9E-07
9E-07
-------
Proposed Army Corps of
Engineers (USACE) Statements of Work
(SOW) for Additional Data Collection
-------
SOW#l
Harbor Sediment Survev/Bioaccumulation Analysis for Proposed Additional
Contaminants of Concern
The goal of this task is to establish the presence of contaminants of concern (COCs) at
areas in the harbor region where navigational dredging will be conducted. A number of
compounds have been proposed as additions to the list of required analytes to be
measured for all HARS placement proposals. A survey of harbor sediments has not been
done to determine the appropriateness of this requirement. Sampling will focus on
specific target areas within the harbor that will be sites for navigational channel dredging,
and possibly at private applicant sites. In addition, sampling will occur at dry-dock areas
in the proximity of federal channels where organotins in sediment will also be analyzed.
The current analytical methods approved by the USEPA for coplanar PCBs, alkylated
PAHs, and the organotins in various media (water, sediment, tissue) will be determined.
Next, it will be necessary to determine which laboratories have acceptably applied these
methods in sediment and tissue analyses and at what detection limits; and how many of
these laboratories should be engaged to analyze the sampled sediments.
If contaminants from the proposed list are found at detectable concentrations in the
channel sediments (and assuming that the sediment concentrations could reasonably
bioaccumulate to levels of concern), the next tier in this task will be to conduct a pilot
study involving bioaccumulation tests to determine whether these exposures result in
bioaccumulation levels of concern.
As noted above, the proposed list of contaminants includes the alkylated polycyclic
aromatic hydrocarbons (PAHs), the butyltins (tetra-, tri-, di- and mono-), and several
coplanar polychlorinated biphenyls (PCBs). If there is reason to believe that the mono-
and di-butyl tins behave less like organic compounds and that their quantitation is thereby
imprecise, it should be determined if the butyl tins should be used independently in
determining effects, or if they should be summed to compare with a guidance value.
Thus, if the initial sediment screen reveals these contaminants in sediment, it will be
necessary to verify whether the relevant analytical methods can be used for the kinds of
tissue samples typically available in the Tier 3 bioaccumulation tests. Is the present
dioxin method EPA 8290 (EPA SW846 update, 1994) for sediment and tissue analyses
suitable for the coplanar PCBs? Is the method for analysis of alklylated PAHS the same
as for parent PAHs, EPA 8270B (SW 846 update, 1994)? The DLs in ppb wet weight
for TTBT, TBT, DBT, and MBT currently achieved by Battelle, for example, are
respectively: 1.234, 1.442,0.881, and 0.354 (A1 Uhler/Battelle Labs, personal
communication); are these appropriate DL's given the use of these data in this program?
These DLs appear to be significantly below the levels of organotins associated with
adverse bio-effects.
-------
• Subtask 1: Verify detection limits for the various contaminants. Once the analytical
methods and laboratories' ability to achieve the method requirements for the COCs
have been agreed upon, the remaining subtasks can be performed.
• Subtask 2: (1) Choose specific target areas from suggested list or other suggested
locations; (2) Identify potential dry-dock source areas.
• Subtask 3: Obtain sediment samples based on current shoaling pattern (suggested:
ten sample locations, 4 cores per location to project depth in the channels; 3 cores to 6
feet per dry-dock area for organotins, with 6 one -foot vertical slices per core). This
will yield the required volume of sediment for all sediment and bioaccumulation tests
should all be implemented.
• Subtask,4: Analyze sediment samples for target analytes. Determine if sediment
concentrations of the COCs warrant further tissue analysis. Determine vertical
distribution of organotins from 6 one-foot sections at dry-dock areas.
• Subtask 5: Run bioaccumulation tests and analyze tissue samples for target analytes.
Subtask 1: Identify acceptable EPA approved methods for the analysis of all of the COCs
- present methods should be identified and reviewed. Guidance values and "elevated
contaminant levels" should be defined as they relate to the COCs being investigated.
Subtask 2: Identify target areas — The second step in this task is to specify the areas
proposed for navigational dredging or private applicant areas within the harbor and
delineating these areas on navigational charts and maps. Navigationally relevant areas
should be identified for sampling where there is the greatest likelihood for finding the
specified analytes. In addition, where possible the location of dry-dock areas as sources
of organotin pollution, should affect the choice of adjacent or nearby federal channels
and/or applicant sites. Possible areas include:
Federal Channels
• Flushing Bay
• Hudson River
• East River
• Buttermilk
• Newark Bay (Elizabeth Pierhead Channel)
• Red Hook Rats Anchorage
• Perth Amboy Anchorage
Dry-dock Areas
• Morris Canal, Jersey City
• Brooklyn Navy Yard, Red Hook, Brooklyn
• Union Dry Dock, Hoboken, NJ
• Caddell Dry Dock, Staten Island, NY
-------
• MOTBY, Bayonne, NJ
• Port Newark/Port Elizabeth
Subtask 3: Obtain and analyze sediment samples — Once the areas have been identified,
the next step is to determine the number of samples and sampling depths at each location.
The sampling program should be designed to provide confidence that the area proposed
for dredging has been adequately characterized. Following the collection of sediment
samples, these samples will be analyzed for the target analyte list, which includes the
alkylated PAHs; four organotins; and 3 coplanar PCBs.
Subtask 4: Conduct bioaccumulation tests for sediment samples with elevated
concentrations of target analytes - If subtask 3 reveals the presence of the target analytes
in sediments from particular areas, this task will involve conducting standard Tier 3
bioaccumulation tests on a subset of the sediment samples. These tests will be conducted
in accordance with US Army Corps of Engineers/US EPA guidance.
Subtask 5: Analyze tissue samples for target analytes - The final subtask in the pilot
tissue study is to determine the feasibility and limits to quantifying the target analytes in
the tissues of bioaccumulation test species. If time and budget permit, efforts may be
made to have the tissues analyzed by more than one contract laboratory to define the
availability and costs for these analyses.
-------
SOW #2
Determination Of Spatial Ranges for Fish Found at the HARS (Site Use Factor)
The goal of this task is to quantitatively evaluate the home ranges of the predominant fish species
found at HARS in order to refine the site and area use factors in the food web and risk models.
The home range is defined as the area over which an animal normally travels and can, for some
species, be very well defined. Acoustic telemetry and mark- or tag-recapture techniques, together
with a computer analysis of the data, can be used to determine home range parameters.
The objectives of this study are to:
• Select Target Fish
• Select An Appropriate Method for Fish Tracking
• Design and Execute the Fish Tagging Study
• Develop Area Use Factors
Subtask 2a: Selection of Target Fish
Human health and ecological risk assessors will contribute to the discussion of criteria for the
selection of target species because the study design and results must meet the requirements of a
risk based evaluation at the HARS site. The human health risk assessors will provide
information on those characteristics and parameters of the consumed species that will provide a
realistic and simultaneously health protective estimate of human health risk. Such factors may
include (among others):
• That the fish be eaten by humans;
• The size of the fish (so that they are of edible size);
• The lipid content of the target species (insofar as this may affect partitioning of some
contaminants);
• The likely preparation and cooking method (as this may affect contaminant
concentrations and organs consumed);
• The human populations likely to consume the species (as this will affect the
parameterization of exposure scenarios);
-------
• The frequency of consumption of the species (as this will affect the continuity of
exposure);
• Penetration into the population (i.e. how wide among the general population is the
exposure);
• Handling of the species prior to distribution or consumption (insofar as this may
affect the form or amount of contaminants).
Ecological risk assessors will work with fisheries biologists to develop selection criteria that
assure that the target species also represents those species most likely exposed to local
conditions. Such criteria may include (although not be limited to):
• That fish have limited home ranges (insofar as this will affect the proportion of
exposure that is locally derived);
• That fish are top-level predators (insofar as this will maximize bioaccumulation for
those compounds that are subject to it);
• That the fish are in a predominantly sediment-driven food web (insofar as this will
tend to maximize local exposure and minimize exposures that occur through the more
temporally variable water column);
• That the fish are resident in the area through most of their life cycle and especially
during critical life stages (insofar as this will minimize spatial variation in exposure
and assure that the populations are exposed throughout their life cycle).
Thus, the fish that are selected for monitoring may include flatfish such as summer or winter
flounder, and perhaps a forage fish such as the sandlance that might be consumed by the
flounder.
Subtask 2b: Select an Appropriate Method for Fish Tagging Study
National Oceanic and Atmospheric Administration Scientists will specify the most efficient
method for conducting a fish tagging study based on (but not limited to):
• Prior experience;
• The criteria developed in subtask 2a for selecting target species;
• The biology of the target species;
• The specific physical oceanographic conditions of the Mid-Atlantic Bight;
-------
• The specific geological conditions of the Mid-Atlantic Bight;
• The specific operational conditions posed in the Mid-Atlantic Bight;
• The sea keeping capability of available platforms;
• The season selected for the study;
• The constraints of the study design (to be developed in subtask 2d);
The most likely method will be acoustic tracking because radio telemetry is not appropriate in
deep or highly conductive water. The Behavioral Ecology Division of the National Oceanic and
Atmospheric Administration (NOAA) Northeast Center at Sandy Hook (headed by Dr. Mary
Fabrizio) has extensive experience conducting radiotelemetric analyses. Their experience
suggests that offshore, open-water telemetry can be very difficult, primarily related to equipment
issues. For example, transmitters can become lost as fish move out of range, or can only weakly
transmit signals due to noise interference. The batteries in the transmitters are also not always
reliable.
A major task for the study design is to develop protocols to overcome these limitations of
equipment and technique. For offshore, high saline environments, acoustic telemetry, using
soundwaves, is used to convey location information for fish. The acoustic tags emit beeps and
fish movements are tracked. Tracking is done via boat using a directional underwater
microphone (hydrophone) or via fixed locations from buoys. The goal is to capture, tag with
acoustic transmitters in the field, and successfully track approximately 10 - 15 fish.
New technologies employing passive tracking with small ultrasonic transmitters and an array of
receivers that triangulate on the tag attached to the subject fish appear promising. Such systems
are expensive (one capable of following bluefish in a small sub-estuary has been estimated by the
Behavioral Ecology Branch at $75k). But, in addition to location, the units report temperature
and depth at the fish making interpretation of fish movements more tractable.
Subtask 2c: Design and Execute the Fish Tagging Study
Researchers at Sandy Hook have commented that the study design should:
• Center on the need for scientifically robust experiments;
• Determine more than simply how long a released fish stays inside the HARS box (for
example, factors need to be measured that are expected to influence behavior - prey,
turbidity, temperature, salinity, etc.);
• Seek to determine the disposition of each tagged fish - did it die, stop just outside the
HARS box, take up residence in a definable area, etc.;
-------
• Employ controls, including establishing an understanding of how the experimental
protocols (including catching, tagging and releasing fish) influence behavior of the
studied species in that area at the time of the study;
• Endeavor to determine whether the fish are part of a transient, temperature-driven
migration, an annually resident school, or another pattern.
This requires a system that is at least capable of following fish into the near-shore environment,
5-10 km beyond the edge of HARS box. In effect, the need appears to be to track fish in a major
portion of the New York Bight apex and place the results in a habitat context.
Subtask 2d: Develop an Area Use Factor
Assuming that 10 - 15 fish can be successfully tracked over some period of days, each fish's
home range will be graphically displayed on a calibrated map to determine home range areas.
These are typically determined using the minimum-area method (Mohr, 1947). Since home
range size is often correlated with the number of observations or the length of time that a fish is
observed, it is useful to derive an observation-area curve (Odum and Kuenzler, 1955; Abies,
1969, Winter, 1977; Mesing and Wicker, 1986). The home range has been adequately defined
for each individual fish at the point on the observation-area curve where each additional
observation produces less than a one-percent increase in area. After plotting the tracking data on
maps, connecting the outermost points to form the most convex polygon defines the home range
for each fish (Winter, 1977).
-------
SOW #3
Angler Survey at the HARS
The survey objectives for the HARS site are to:
- Identify the population(s) of anglers who catch and eat fish from the HARS;
- Determine the amount and frequency of consumption of fish from the HARS by
members of the target population;
- Use the consumption rate information for the target population to refine risk estimates
and target levels for the HARS;
The goal of this task is to provide species-specific fish consumption rates for a
recreational angling population at the HARS. To the extent that there might be
subsistence anglers, these individuals would be represented in a distribution of fish
ingestion rates of which the subsistence angler represents a low frequency, high
consumption receptor.
Following the USEPA's Guidance for Conducting Fish and Wildlife Consumption
Surveys (USEPA, 1998), this task consists of the following subtasks:
• Subtask 3a: Identify the survey objectives and prepare a sample design and
analysis plan, including:
1. Identification of the target population(s) and selection of the sampling strategy for the
survey population(s)
2. Identification of the specific data to be gained from the survey
3. Selection of the analytical/statistical methods to be used once the data are collected
• Subtask 3b: Select the survey approach to be used to obtain the data and design
the survey instrument(s);
• Subtask 3c: Implement the survey; and,
• Subtask 3d: Analyze the results.
Subtask 3a: Identify the survey objectives and prepare a sample design and analysis plan
- The first step in designing a fish consumption survey is to identify the population that
will be surveyed (the "target population"). The target population for this survey are
recreational anglers fishing in the HARS. However, there may be certain subpopulations
of anglers in the HARS that are of concern. For example, subsistence fishers may be at
greater risk than other anglers because they consume more fish; however, it should be
-------
noted that because of the expense involved it is doubtful that the HARS supports_even a
modest population of subsistence fishers. Also, children, women of child-bearing age,
older persons, and persons with preexistinghealth conditions might be at higher risk from
exposure to certain contaminants through fish ingestion. The survey should be designed
to capture information about these "sensitive" subpopulations, although they are not the
focus of the study.
A sampling strategy for surveying the population is selected once the population(s) of
concern have been identified. If the survey population is relatively small, it may be
desirable and feasible to survey the entire population. However, it is more typical to
randomly sample a subset of the population. The results obtained from sampling a subset
of the population are expected to be approximations of the population parameters.
Sources of error in this type of sample estimate are sample selection bias, sampling error,
and reliability, validity and measurement error in the survey response (USEPA, 1998).
The information collected in a fish consumption survey should be targeted to address the
objectives of the study. Before collecting such information, however, the "sampling unit"
for the survey must be defined. For example, the sampling unit may be an individual
consumer, chosen at random from all members of a household. Alternatively, the
sampling unit may be a household, in which case all members of the household should be
questioned, either individually or with one member of the household speaking for all
members of the household.
The researcher may also need to choose the time period for which the respondent will be
asked to recall consumption.
The survey may include information on the following:
• Characteristics of the angler (and possibly of each household member), such
as ethnicity, gender, date of birth, height, weight)
• Number and type of household members (e.g., child or adult, fish consumer or
non-consumer, resident or migrant)
• Purpose for fishing (consumption, sport only: catch and release, etc.)
• Number of animals by species caught per outing
• Size ranges of animals caught (minimum and maximum weights and lengths
by species)
• Length of involvement in fishing activities and consuming self-caught animals
• Type of fish or other aquatic organisms consumed that were caught at the
HARS in New York Harbor
• Amounts (raw wet weight or cooked weight) of self-caught fish (or other
aquatic organisms) eaten per meal/day/week/month for each person in the
household (to determine portion size and frequency of consumption in
meals/week and weeks/year)
• Geographic and seasonal variations in consumption
• Parts of animal consumed (may vary with species)
-------
• Parts of animal used for cooking but not ingested (e.g., boiling of bones or
fish heads)
• How the animals were prepared for eating (e.g., skinned, fillet, steak,
shucked)
• How the animals were cooked (e.g., baked, fried, steamed)
• Whether fish or other aquatic organisms were frozen or preserved and eaten
throughout the year, or eaten only when fresh
The data collected in this fish consumption survey will be used to develop a distribution
of fish consumption rates. This distribution will be used in the risk assessment "for the
HARS and may be incorporated in developing site-specific target levels. Therefore,
survey questions to gather the types of information listed above should be designed to
gather a distribution of data rather than point estimates. Also the sampling size (i.e.
number of sampling units) will depend on the level of precision required for the
estimates.
Subtask 3b: Select the survey approach to be used to obtain the data and design the
survey instrument(s) - Surveys can be conducted via mailed questionnaires, telephone
calls, or through personal interviews. However, on-site personal interviews to conduct
the creel survey may be the best approach. The advantages of this approach include:
• Assessment of site-specific consumption rates by targeting the HARS.
• Identification of specific subpopulations at high risk by obtaining data from
actual anglers at the HARS
• Augmentation of interviewees' responses with first-hand observations of the
respondents, their fishing activity, their catch and the interview sites.
• Minimization of recall bias by providing appropriate visual aids (for species
and portion or meal size) and by basing the survey on the fish caught at the
time of the interview.
• High success rate for completing interviews (minimizing nonresponse bias)
because of personal contact
• Verification of information is comparatively easy, especially if data collected
are based on the actual catch of the day. It is also relatively easy to obtain
sensitive information and to use special techniques such as visual aids and
probing (USEPA, 1998).
The limitations of an on-site personal interview/creel survey include:
• The number and complexity of survey questions must be limited so that
surveys can be performed quickly.
• Interviewers might encounter language barriers
• This approach is costly because it requires the coordination, hiring, training
and supervision of interviewers and field staff for quality control.
• Responses to questions about consumption are hypothetical because
consumption of the catch has not yet occurred and it is unknown how many
fish will be given away and consumed by the friends and family of the angler.
-------
In addition, these responses measure only the intent to consume, which might
not be an accurate representation of the true consumption rate. As a result,
estimates of consumption are conservative, potentially overestimating
consumption for the fishing population (USEPA, 1998).
To solve the problems associated with this type of survey, the questions need to be well
designed to minimize the length of time needed to conduct the survey, but still provide
information that meets the survey objectives. Interviewers who can conduct interviews in
different languages may also be needed. To address the problem of "hypothetical"
consumption rates, personal interviews can be followed up with telephone interviews
asking whether fish caught on the day of the on-site interview had been eaten and how it
had been prepared.
Once a draft questionnaire has been developed, it should be tested by a focus group to
ensure that questions are understood and elicit the desired response.
Subtask 3c: Implement the survey — After selecting a survey approach, the researcher
must define a method to select interviewees. Interviewers may be instructed to question
everyone, only those who have caught fish at the time (nonuniform sampling), or anglers
selected at random (USEPA, 1998).
After identifying the anglers to be interviewed, the researcher should select the
location(s) where interviews will be conducted. Interviews may be conducted at access
points (boat ramps, docks), along the shoreline or on the water from a boat. There are two
distinct types of creel surveys - access point surveys and roving creel surveys. The access
point survey is preferred when entry points into the fishery are relatively few and well
defined. When access to the fishery occurs at many points, the roving method might be
preferred. In a roving creel survey, the interviewer moves from angler to angler and
sometimes from site to site. This type of survey is appropriate in fisheries with many
access points and low fishing pressure (i.e. activity) (USEPA, 1998). It should be noted
that using a roving creel survey does not ensure that anglers using sites are representative
of all anglers using the fishery, because people fishing longer are sampled
disproportionately more than short-term fishers, and the information gathered on effort
and harvest may be biased. Sites should be selected where there will be more respondents
over longer periods of time to avoid having to "search" for respondents at a site or switch
sites in the middle of the study (USEPA, 1992).
Next, the researcher must define when interviews should be conducted. Bias in creel
surveys can be associated with time of year, time of day, and length of the interview
(USEPA, 1998). First, the study period should be defined (e.g., one year, summer
months, other months with most fishing activity occurs). Next, the researcher should
decide whether interviews will be conducted on weekdays, weekends, holidays, opening
days of the fishing season, the beginning of the fishing season for a particular species, or
some other day or combination of days (USEPA, 1992). Then, the researcher must decide
on what times of day interviews should be conducted (e.g., four-hour periods, two-hour
-------
periods, at the beginning and end of the day). Finally, the length of the interview must be
defined, which will be largely dictated by the survey instrument A,long interview period
can limit the number of people that can be interviewed in a specified time period, can
discourage cooperation of the interviewee, and can bias the data towards anglers who fish
longer. Decisions about when to conduct interviews should be made based on
information about fishing patterns for the HARS area from ongoing creel survey
activities, state agencies, local anglers, or other sources of local information.
An important consideration when implementing the survey is that meal size estimates are
subject to considerable error. Questions about meal sizes should be accompanied by
models of a typical fillet meal and be related to the species that the person has caught
(USEPA, 1998). Visual aids help increase the response rate and minimize the level pf
effort and time needed to conduct interviews (USEPA, 1992).
Subtask 3d: Analyze the results -Following successful implementation of the survey, the
results are analyzed to develop fish consumption distributions. These distributions can be
developed empirically or through parametric fitting methods, and appropriate statistics
can be obtained for use in the risk assessment (e.g., mean, 95th percentile, etc.).
-------
-------