United States
Environmental Protection
Agency
Office of Research and
Development
Washington DC 20460
Workshop on the Use of
Available Data and
Methods for Assessing the
Ecological Risks of 2,3,7,8-
Tetrachlorodibenzo-p-
Dioxin to Aquatic Life and
Associated Wildlife
RISKASS(:SSM N FORUM
-------
EPA/630/R-94/002
MAY 1994
WORKSHOP ON THE USE OF AVAILABLE DATA AND METHODS
FOR ASSESSING THE ECOLOGICAL RISKS OF
2^,7,8-TETRACHLORODIBENZO-p-DIOXIN TO AQUATIC
LIFE AND ASSOCIATED WILDLIFE
Risk Assessment Forum
U.S. Environmental Protection Agency
Washington, DC 20460
Printed on Recycled Paper
-------
DISCLAIMER
This document has been reviewed in accordance with U.S. Environmental Protection
Agency policy and approved for publication. Mention of trade names or commercial products
does not constitute endorsement or recommendation for use.
-11-
-------
CONTENTS
ACKNOWLEDGMENTS '. ....... .... vi
FOREWORD vii
1. INTRODUCTION 1
1.1. Background and Workshop Objectives 1
1.2. Workshop Organization and Source Materials 1
1.3. General Summary 6
1.3.1. Ecological Effects and Endpoint Selection 6
1.3.2. Stressor Characterization 6
1.33. Conceptual Model 7
1.4. Identification of Research Needs 8
1.4.1. Toxicological Considerations 8
1.4.2. Fate, Transport, and Exposure Considerations 8
2. SIGNIFICANCE OF WORKSHOP DISCUSSIONS FOR ECOLOGICAL RISK
ASSESSMENT 11
2.1. Ecological Risk Assessments for TCDD 11
2.2. General Ecological Risk Assessment Considerations 11
3. COMMENTS ON ECOLOGICAL EFFECTS AND ENDPOINT SELECTION .... 13
3.1. Issue 1—Focus on Fish Species 13
3.2. Issue 2—TEFs and BSAFs 13
3.3. Issue 3—Use of Tissue Levels vs. Exposure Concentrations 14
3.4. Issue 4—Use of Laboratory Data to Predict Field Effects 14
3.5. Issues 5 and 6—Concentration-Response Curves and Data Extrapolations 15
3.6. Issue 7—Additional Effects Data Not Considered in the Scenario 17
3.7. Assessment and Measurement Endpoints 17
4. COMMENTS ON STRESSOR CHARACTERIZATION 19
4.1. Issue 8—Uncertainties in K^, and ]£„. 19
4.2. Issue 9—Exposure Routes 20
4.3. Issue 10—Fate-and-Transport Models 21
4.4. Issue 11—Additional Exposure Issues 24
4.5. Issue 12—BCFs and BSAFs 24
4.6. Issue 13—Applicability of Lake Ontario BAF Data 25
4.7. Issue 14—Biomagnification 25
4.8. Issue 15—BAF Uncertainties 26
4.9. Key Research Needs in Stressor Characterization 26
-iii-
-------
CONTENTS (continued)
5. COMMENTS ON THE CONCEPTUAL MODEL ., 27,
5.1. Responses to Conceptual Model Issues — 27
5.1.1. Issue 16—Conceptual Model Focus on Fish and
Piscivorous Wildlife ... 27
5.1.2. Issue 17—Linking TCDD Loadings to Tissue Residues 27
5.13. Issue 18—Applicability of BAFs and BSAFs to the
Conceptual Model 29
5.1.4. Issue 19—Temporal Dynamics of TCDD in the Reservoir 29
5.2. Comments on the Conceptual Model 30
6. REFERENCES ........... 39
APPENDICES A-C: PREWORKSHOP MATERIALS
Appendix A. A Preliminary Problem Formulation for a Dioxin Scenario:
Proposed Paper Mill on a Southern Reservoir ... A*l
Appendix B. Three Workshop Exercises (Questions to Panelists) B-l
Appendix C. Panelist Premeeting Comments C-l
APPENDICES D-F: WORKSHOP MATERIALS
Appendix D. Workshop Agenda D-l
Appendix E. Workshop Participants and Final Observer List E-l
Appendix F. Work Group Assignments — F-l
APPENDIX G: REVISED RISK ASSESSMENT CONCEPTUAL MODEL G-l
IV
-------
LIST OF FIGURES
Figure 1. Ecological risk assessment framework ....; ,.> ... 3
Figure 2. Ecological risk assessment framework: problem formulation — ...'..:.... 4
Figure 3. Residues vs. effects .... .:.......... 16
Figure 4. Overall model 31
Figure 5. Developing discharge permit limits based on ecological risk assessment 32
Figure 6. Fate-and-transport diagram 33
Figure 7a. Example fate model outputs—TCDD (log [concentration in
compartment]) vs. probability — ... 34
Figure 7b. Example fate model outputs—TCDD (concentration in compartment) vs.
loading 34
Figure 8. Framework for predicting BSAFs applicable to different aquatic
ecosystems. 36
LIST OF TABLES
Table 1. List of Workshop Participants 2
Table 2. Progressive Levels of Aquatic Chemical Models 23
Table 3. Comparison of BSAFs and Food Chain Models. 35
LIST OF TEXT BOXES
Text Box 1. Framework Report Terminology. 5
Text Box 2, Predicting TCDD Concentrations in the Hypothetical Scenario 28
-v-
-------
ACKNOWLEDGMENTS
This report is based on discussions and presentations at a workshop held in Minneapolis,
MN, on September 14 and 15,1993. Dr. Robert Huggett (Virginia Institute of Marine Science,
The College of William and Mary) chaired the workshop, which included work groups led by Dr.
William Adams (ABC Laboratories), Dr. Charles Menzie (Menzie-Cura & Associates), Dr.
Randall Wentsel (U.S. Army), and Dr. Huggett. Scientists from the U.S. Environmental
Protection Agency's (EPA's) Duluth Environmental Research Laboratory, including Dr. Philip
Cook, Dr. Russell Erickson, Dr. Robert Spehar, Dr. Steven Bradbury, and Dr. Gerald Ankley,
prepared the report that served as the primary background document for the workshop.
Dr. William van der Schalie and Dr. John Gentile of EPA compiled this report using
materials generated at the workshop and assisted the Duluth EPA scientists in preparing the
hypothetical TCDD ecological risk assessment scenario. The workshop was organized with the
assistance of Ms. Helen Murray and Ms. Arlene Levin of Eastern Research Group. Mr. John
Bergin provided final editing of this report.
-VI-
-------
FOREWORD
In April 1991, EPA initiated efforts to conduct a scientific reassessment of the human
health and environmental risks of TCDD and related compounds. The first document to address
ecological effects developed under this reassessment is the Interim Report on Data and Methods
for the Assessment of2,3,7,8-Tetrachlorodibenzo-p-Dioxin (TCDD) Risks to Aquatic Life and
Associated Wildlife (U.S. EPA, 1993a)(the Interim Report), which was originally peer reviewed in
the fall of 1992 arid published in March 1993. The Interim Report compiles and critically reviews
current scientific literature concerning toxicity and exposure data for TCDD ecological risks to
aquatic and wildlife species.
This report summarizes the discussions of a peer panel workshop that evaluated the
utility of the information in the Interim Report for future ecological risk assessments. In addition
to the Interim Report, workshop participants used a hypothetical scenario and the EPA report
Framework for Ecological Risk Assessment (U.S. EPA, 1992) as a context for identifying
significant issues, discussing major uncertainties, and recommending research needs for future
assessments. The focus of the workshop was on the problem formulation, or scoping phase, of
the ecological risk assessment process. The resulting "conceptual model," including comments
and suggestions on the transport, fate, and effects of TCDD and related compounds, should
serve as a valuable resource for risk assessors planning to evaluate the ecological risks of TCDD.
Dorothy E. Patton, Ph.D.
Executive Director and Chair
Risk Assessment Forum
-V11-
-------
In April, 1991, EPA initiated efforts to conduct a scientific reassessment of the human health
and environmental risks of TCDD and related compounds. The first document to address ecological
effects developed under this reassessment is the Interim Report on Data and Methods for the
Assessment of 2,3,7,8-Tetrachlorodibenzo-p-Dioxin (TCDD) Risks to Aquatic Life and Associated
Wildlife (U.S. EPA, 1993a; Interim Report. The Interim Report compiles and critically reviews
current scientific literature concerning toxicity and exposure data for TCDD ecological risks to aquatic
and wildlife species.
This report summarizes the discussions of a peer panel workshop that evaluated the utility of
the information in the Interim Report for future ecological risk assessments. In addition to the Interim
Report, workshop participants used a hypothetical scenario and the EPA report Framework for
Ecological Risk Assessment to identify significant issues, discuss major uncertainties and recommend
research needs for future assessments. The focus of the workshop was on the problem formulation or
scoping phase of the ecological risk assessment process. Participants evaluated a "conceptual model"
from the hypothetical scenario and provided comments and suggestions on the transport, fate, and
ecological effects of TCDD and related compounds.
Vlll
-------
1. INTRODUCTION
1.1. Background and Workshop Objectives
For several years, the U.S. Environmental Protection Agency (EPA) has been evaluating
studies on the health and environmental risks associated with exposure to 2,3,7,8-
tetrachlorodibenzo-p-dioxin (TCDD) and related compounds. Scientists at EPA's Duluth
Environmental Research Laboratory issued a report that summarizes and analyzes scientific
information about the effects of TCDD on aquatic life and associated wildlife. Following a peer
review hi the fall of 1992, this report, entitled Interim Report on Data and Methods for the
Assessment of2,3,7,8-Tetrachlorodibenzo-p-Dioxin Risks to Aquatic Life and Associated Wildlife
(U.S. EPA, 1993a)(the Interim Report) was published in March 1993.
While the Interim Report provides guidance on data and models for evaluating the
exposure and effects of TCDD on aquatic life and wildlife, identifies major uncertainties, and
outlines some risk assessment issues, the report was not intended to constitute a risk assessment
or even to serve as guidance for conducting a risk assessment. Thus to examine how the data
and methods in the Interim Report could be applied in future ecological risk assessments, EPA
held a workshop in Minneapolis, Minnesota, on September 14 and 15,1993. EPA asked the
workshop .peer panel to discuss scientific uncertainties, identify related research needs, and
consider how risk assessments might be planned.
Several important topics were beyond the scope of the workshop and were not addressed.
These included consideration of EPA policy on TCDD, revisions to EPA's ecological risk
framework, formulation of a complete TCDD risk assessment, and issues related to the health
effects of TCDD.
1.2. Workshop Organization and Source Materials
Prior to the workshop, the 20 participants (table 1) received copies of a scenario
describing the development of a "conceptual model" for a hypothetical risk assessment of TCDD
introduction to the Omigoshee Reservoir (appendix A), questions on the major aspects of the
scenario (appendix B), and the EPA report Framework for Ecological Risk Assessment (EPA,
1992)(Framework Report), which proposes terminology and offers a structure for conducting
ecological risk assessments. As described in the Framework Report, a conceptual model is
developed during the "problem formulation" stage of the risk assessment process (figure 1). This
initial planning and scoping stage has several elements, including stressor characterization and
ecological effects and endpoint selection (figure 2). The Framework Report's definitions of key
terms are shown in text box 1 (page 5).
As a starting point for the workshop, EPA scientists extracted several major findings from
the Interim Report and posed questions concerning the relationship of these findings to the risk
assessment approach described in the hypothetical scenario. Each panelist prepared written
premeetihg comments addressing the questions concerning the scenario as well as other issues
identified by the participants. Panelists were asked to comment on the use of the Interim
Report's findings hi the scenario as well as to propose any changes or additional information that
might be necessary for conducting risk assessments. The comments were collated and distributed
-------
Table 1
LIST OF WORKSHOP PARTICIPANTS
Workshop Chair
• Robert Huggett, Virginia Institute of Marine Science, The College of William and
Mary
Workshop Work Group Leaders
• William Adams, ABC Laboratories
• Charles Menzie, Menzie-Cura & Associates
• Randall Wentsel, U.S. Army Edgewood Research, Development, and Engineering
Center
Other Participants
• Nigel Blakefy, Washington State Department of Ecology
• Peter Chapman, EVS Environment Consultants
• Keith Cooper, Rutgers University
• G. Michael DeGraeve, Great Lakes Environment Center, Inc.
• Joseph DePinto, University of New York at Buffalo
• John Giesy, Michigan State University
• Wayne Landis, Western Washington State University
• Derek Muir, Freshwater Institute
• Thomas O'Connor, National Oceanic and Atmospheric Administration
• Robert Pastorok, PTI Environmental Services
• Richard Peterson, University of Wisconsin
• Paul Rodgers, LTI Limno-Tech
» Thomas Sibley, University of Washington
• John Stegeman, Woods Hole Oceanographic Institution
• John Sullivan, Wisconsin Department of Natural Resources
• Bill Williams, Ecological Planning and Toxicology, Inc.
-2-
-------
Discussion
Between the
Risk Assessor
and
Risk Manager
(Planning)
Ecological Risk Assessment
Characterization
of
Exposure
Characterization
of
Ecological
Effects
Discussion Between the
Risk Assessor and Risk Manager
(Results)
Figure 1. Ecological risk assessment framework. The risk assessment process, shown within
the heavy line, includes three phases: problem formulation, analysis, and risk
characterization. Each phase includes the elements of the exposure to and the
effects of stressors. (EPA, 1992)
-3-
-------
PROBLEM FORMULATION
\
ANALYSIS
RlSlt CHARACTERIZATION
Discussion
Between the
Risk Assessor
and
Risk Manager
(Planning)
Ecological
Effects
Potentially
at Risk
Characteristics
Endpolnt
Selection
Assessment
Measurement
ANALYSIS
Figure 2. Ecological risk assessment framework: problem formulation. Problem
formulation, an initial planning and scoping activity, has several elements,
including stressor characterization, ecosystem characterization, and ecological
effects and endpoint selection, all leading to a conceptual model for the risk
assessment (EPA, 1992)
-4-
-------
to each participant prior to
the meeting (appendix C).
At the workshop,
panelists were divided into
three work groups assigned
to consider the three major
aspects of the scenario:
ecological effects and
endpoint selection; stressor
characterization; and the
conceptual model
development (appendices
D-F). Work group leaders
summarized the written
premeeting comments
relevant to their topics hi
plenary session, then led
discussions of issues in
breakout groups. Finally,
breakout group comments
and recommendations were
discussed in plenary
session. This report is
organized around the
material developed by the
three work groups and is
generally presented as
responses to EPA's
questions concerning the
hypothetical scenario
(appendix A). Many
comments and
recommendations were
made, both individually
and jointly, by the
panelists. This document
attempts to convey those
inputs.
Text Box 1. Framework Report Terminology (EPA, 1902)
assessment endpoint—An explicit expression of the
environmental value that is to be protected
conceptual model—The conceptual model describes a series
of working hypotheses of how the stressor might affect
ecological components. The conceptual model also
describes the ecosystem potentially at risk, the
relationship between measurement and assessment
endpoints, and exposure scenarios.
ecological component—Any part of an ecological system,
including individuals, populations, communities, and
the ecosystem itself.
ecological risk assessment—The process that evaluates the
likelihood that adverse ecological effects may occur or
are occurring as a result of exposure to one or more
stressors.
exposure—Co-occurrence of or contact between a stressor
and an ecological component.
exposure scenario—A set of assumptions concerning how an
exposure may take place, including assumptions about
the exposure setting, stressor characteristics, and
activities that may lead to exposure.
measurement endpoint—A measurable ecological
characteristic that is related to the valued
characteristic chosen as the assessment endpoint.
Measurement endpoints are often expressed as the
statistical or arithmetic summaries of the observations
that comprise the measurement.
stressor—Any physical, chemical, or biological entity that can
induce an adverse response.
-5-
-------
1.3. General Summary
As noted above in section 1.2, workshop discussions were organized around major
elements of the problem formulation phase of an ecological risk assessment, including evaluating
ecological effects and stressor characteristics, selecting endpoints, and developing a conceptual
model for the assessment. While panelists felt that the information contained in the Interim
Report would be useful for conducting ecological risk assessments, they identified a number of
areas for which additional research should be conducted.
Ecological Effects and Endpoint Selection
Ecological effects attributable to TCDD are identified in the Interim Report and in the
scenario that was provided to the workshop panelists. Data on these effects are available from
both laboratory and field sources. Data on ecological effects and other information are used to
help select assessment endpoints for the risk assessment. Assessment endpoints may be defined
as explicit expressions of the actual environmental value that is to be protected (EPA, 1992).
Panelists commented on the utility of existing ecological effects data for risk assessments and on
the assessment endpoints proposed for the scenario.
Most panelists felt that the report's focus on early life-stage effects in fish as representing
the most sensitive aquatic effect was appropriate given the present state of scientific knowledge,
in spite of the uncertainties. Some panelists, however, felt that additional data should be
generated on the Ah receptor and its role in TCDD toxicity for a wider range of species than
have been tested. Such data would help evaluate interspecies differences in sensitivity to TCDD.
Given the difficulty in measuring TCDD levels in water, most panelists felt that
evaluating TCDD effects based on tissue levels was appropriate, although linkages between
residue levels and effects and residue levels and TCDD loadings need to be established.
Panelists also discussed whether "single species" data from laboratory tests are
appropriate for solving a complex ecological problem. While most panelists agreed that existing
data on single species could be used to make a reasonable assessment of ecological risk, others
felt that it would be more appropriate to use other types of test data, such as multispecies tests.
Several panelists pointed out that a high probability exists for unanticipated, indirect effects in a
"real world" situation that could not be addressed by typical laboratory tests.
Assessment endpoints selected by the panelists included maintenance of the aquatic
community; maintenance of sport fishery populations; and maintenance of piscivorous wildlife
(i.e., birds, mammals, and reptiles such as turtles or alligators). Panelists made several
suggestions for developing additional data to support measurement endpoints (see section 1.4).
1.3.2. Stressor Characterization
Important characteristics of a stressor include type (chemical or physical), intensity,
duration, frequency, and scale (EPA, 1992). Stressor characteristics are clearly important for
-6-
-------
identifying the ecosystems potentially at risk from the stressor as well as for anticipating likely
ecological effects (figure 2). Panelists reviewed the available data on TCDD and related
compounds and evaluated their applicability in risk assessment.
TCDD is a highly hydrophobic chemical and a member of a broad group of non-ionic
organic chemicals. As such, TCDD will be associated primarily with suspended and bedded
solids in the reservoir described in the scenario. Moreover, the fate and transport of solids is
critical to predicting the distribution of TCDD within the reservoir. The key parameter is the
organic carbon/water partition coefficient (K^) for TCDD, along with other physicochemical
characteristics (e.g., Henry's Law constant, octanol/water partition coefficient [K^]) that
influence partitioning. Since available fate-and-transport models vary widely in their complexity
and resource requirements, a tiered approach was suggested whereby the risk assessor would
start with simple but conservative models and proceed toward more complex (and expensive)
models as far as required by the goals (and resources) of the risk assessment.
Given the high K^. values for TCDD, major routes of exposure will be through contact
with sediments or through ingestion of contaminated food. Both the biota-sediment
accumulation factor (BSAF) and food web approaches were suggested for evaluating exposure.
BSAFs will be most appropriate for lower trophic level species associated with the sediments.
Panelists felt that it may be premature to conclude that biomagnification of TCDD does not
occur among aquatic species, although this is suggested based on available data.
Biomagnification may occur due to food chain transfer of TCDD from aquatic to avian and from
aquatic to mammalian species. Present data, however, do not allow this possibility to be
evaluated.
1.3.3. Conceptual Model
The conceptual model is the culmination of the problem formulation stage of an
ecological risk assessment (figure 2). Ideally, the conceptual model includes a discussion of
potential exposure pathways, effects of the stressor on ecological components, descriptions of the
ecosystem potentially at risk, and identification of endpoints. Exposure scenarios are developed
that provide "... a qualitative description of how the various ecological components co-occur with
or contact the stressor" (EPA, 1992). Panelists evaluated and modified the conceptual model
proposed in the Omigoshee Reservoir scenario (see appendix A).
Panelists indicated that a risk assessment for TCDD must emphasize the importance of
physicochemical fate-and-transport modeling to link exposure to loadings. Application of
equilibrium models would introduce less uncertainty than time-variable models, but the latter
must be used if system response time is critical to the management questions being asked in the
risk assessment. For the hypothetical scenario, models that predict concentrations in bedded
sediments and on suspended sediments make the most sense. Linking sediment concentrations
and possibly water concentrations of TCDD to fish body burdens can be done using either
appropriate factors (e.g., BSAFs) and/or a food chain model. Residue levels, in turn, should be
linked to reproductive and other effects and, if possible, these effects should be expressed in
terms of changes at the population level.
-7-
-------
1.4. Identification of Research Needs
It is important to note that while the panelists recognized the usefulness of the additional
research described below, they also felt that TCDD risk assessments could be conducted at
present using the data and methods described in the Interim Report.
1.4.1. lexicological Considerations
The following research needs were identified by individual panelists and listed in no
particular order:
• Collect Ah receptor information from a wider range of species (e.g., amphibians,
reptiles, invertebrates).
" For avian and mammalian species, evaluate feeding behavior, dietary uptake, and
long-term effects on survival and reproduction as a function of whole-body as well
as tissue-specific residues.
• Conduct complete life cycle testing with fish.
• Investigate effects on marine mammals and ecosystems, aquatic plants, and
detrital communities in oligotrophic lakes.
1.4.2. Fate, Transport, and Exposure Considerations
The following recommendations were considered high priorities (they are not listed in any
particular order):
Improve the K^. estimate for TCDD.
K,,,. is used to estimate sediment/soil partitioning and is a key parameter in
assessing TCDD transport and fate in aquatic and terrestrial ecosystems. The "K^.
estimate could be improved by measuring K^ and predicting K,,,. from the K^, for
TCDD. The methodology currently being used to obtain K^s for sediment
quality criteria should be employed (i.e., the slow-stir equilibrium method).
Attempts to remeasure K^, using standard sediment partitioning experiments will
not improve current estimates. Too many operational variables are extremely
difficult to control.
Standardize methods for lipid and organic carbon measurements.
Several methods are currently being used to measure both tissue lipid content and
organic sediment/soil organic carbon content. It would be fairly easy to review
the methods, run comparisons, and standardize single approaches for lipid and for
organic carbon measurements.
-8-
-------
» Compile a library of bioconcentration factor (BCF), bioaccumulation factor
(BAF), BSAF, and biota-suspended solids accumulation factor (BSSAF) values.
(Of greatest interest would be BSAF and BSSAF values from field data).
The need for a compilation of these values stems from the position that for highly
hydrophobic compounds approaches are being derived to assess risk that are not
dependent on measurements of the specific chemical in the water column. These
methods have been favorably reviewed by the panelists. Methods include the use
of parameters such as the BSAF, BSSAF, and BAF. Issues that have arisen with
the use of these estimators include whether the values can be translated from site
to site (even with carbon and lipid normalization) or whether the values are site
specific. A compilation of these values would allow this important question to be
answered. Universal application of the concepts is desirable.
Another recommendation was considered important but not of the highest priority:
» Spatial and temporal patterns of dietary uptake of contaminated prey need to be
investigated.
Adequate characterization of food webs and an understanding of the transport of
TCDD and similar compounds through various food webs are critical for
ecological risk assessments of polychlorinated dibenzodioxins (PCDDs) and
polychlorinated dibenzofurans (PCDFs).
The last recommendation was considered of a somewhat lower priority, although still
important:
» Standardize the method(s) for measuring K^. for highly lipophilic compounds.
This recommendation is of a lower priority because E^. can be estimated from
and because the problems that have to be solved to standardize an approach for
extreme hydrophobes are enormous.
-9-
-------
-------
2. SIGNIFICANCE OF WORKSHOP DISCUSSIONS FOR ECOLOGICAL RISK
ASSESSMENT
Although the peer panelists focused their attention primarily on assessing ecological risks
associated with TCDD, their discussions also raised points that are more generally relevant to
ecological risk assessment. This section describes some of the more important conclusions for
risk assessors concerned with both TCDD-specific and more general evaluations of ecological
risk.
2.1. Ecological Risk Assessments for TCDD
Peer panelists generally found the information in the Interim Report to be sound.
Nonetheless, they suggested that attention be given to a number of important research needs
(section 1.4) and recommended approaches to using the information in the risk assessment
process. Panelists indicated that methodologies for TCDD risk assessment may be more
generally applicable to other hydrophobic organic chemicals (HOCs). This suggests that it may
be possible to develop a generic aquatic ecological risk assessment model that could be adjusted
for the characteristics of the particular chemicals and ecosystems involved.
The overall approach to developing a sound conceptual model for the ecological risk
assessment involves establishing and evaluating the logical linkages among TCDD sources (in this
case, the paper mill), fate and transport of TCDD, uptake by biota, effects on biota, and the
consequences of these effects at the population level (and higher). Many tools are available for
determining TCDD fate and transport, and their selection in the analysis phase will depend on
the management questions being asked and the resources available. BSAF and food chain
approaches can both be used to help predict residue levels in organisms, which in turn can be
related to effects. Yet, predicting biological effects at the population level and above may be
difficult given data and methodological limitations at this time—an aspect of ecological risk
assessment that is not necessarily unique to TCDD.
It is important to note that the scenario used at this workshop was simplified in the sense
of having only one stressor (TCDD) and one primary source (the paper mill). Interpretation of
data in situations where TCDD is already present are frequently much more complicated because
of multiple sources and the presence of other stressors (i.e., chemical, physical, and biological).
In these cases, the emphasis may be less on prediction of effects and more on establishing
causality between an observed effect (e.g., decline of a sport fishery population) and a presumed
cause(s) (e.g., TCDD, overfishing, habitat destruction).
2.2. General Ecological Risk Assessment Considerations
EPA's Risk Assessment Forum recently published a set of 11 peer-reviewed case studies
that evaluated a wide range of ecological assessments from a risk assessment perspective (EPA,
1993b). A number of common themes concerning the nature of ecological risk assessment
emerged during the development and evaluation of these case studies and many of the "lessons
learned" from the case studies were reflected in the peer panel discussions of the hypothetical
TCDD scenario. Three of these recurring themes are described below.
-11-
-------
Formulating the Problem and Developing the Scope Are Critical First Steps.
Several of the studies reviewed for the case studies report involved difficulties that could
be attributed to inadequate problem formulation. Indeed, the peer panel found shortcomings in
the conceptual model for the TCDD case study and provided many suggestions for revision and
expansion of the conceptual model. The final product of these discussions provides a better basis
for proceeding to the analysis phase of the risk assessment. Clearly, an essential factor for
success in the problem formulation stage of an ecological risk assessment is a sound
understanding of both the stressor and the ecosystem involved.
Defining Assessment and Measurement Endpoints Focuses the Scope of the Risk Assessment.
Selection of the proper assessment endpoints for the TCDD risk assessment generated
considerable discussion among panelists. For example, some felt that a focus on the sport fishery
population was appropriate, while others found this to be too narrow a focus. The presence of
rare or endangered species in the area also could have influenced the selection of assessment
endpoints. Panelists emphasized the importance of selecting assessment endpoints that can be
used for decision-making and that at the same time address ecological concerns.
Clearly Stated Risk Scenarios Help Structure the Assessment.
Risk scenarios developed as part of the conceptual model reflect the risk assessor's
judgment concerning which stressors, ecological components, and pathways are likely to be the
most significant in the risk assessment. Such scenarios are critical since resources frequently
limit the range of possibilities that can be explored and the process of problem formulation must
reduce the risk scenarios to a manageable number. Identification of important TCDD pathways
and species that are sensitive to TCDD were helpful in scenario development for the reservoir.
-12-
-------
3. COMMENTS ON ECOLOGICAL EFFECTS AND ENDPOINT SELECTION
Peer panel members were asked to use information in the Interim Report as source
material to address seven issues concerning ecological effects and endpoint selection raised by
the scenario. The Interim Report focuses on the effects of TCDD on freshwater aquatic
organisms and associated wildlife (Interim Report, chapter 4). Available data on TCDD effects
on fish are provided in section 4.2.1 and are summarized in section 4.23 of the Interim Report,
while effects on aquatic-associated wildlife are provided in section 4.3.1 and summarized in
section 4.3.3.
3.1. Issue 1—Focus on Fish Species
The lack of Ah receptors in some species (Interim Report, section 4.1 [U.S. EPA, 1993aJ)
along with the results of a limited number of laboratory studies suggest that amphibians, invertebrates,
and plants are less sensitive to TCDD than fish, birds, and mammals. Fish appear to be most
sensitive in earfy life stages. Because of this range in sensitivity, productivities offish species were
selected as assessment endpoints for the scenario. Comment on whether this focus on fish species
will result in adequate protection for the rest of the aquatic community in the reservoir from the direct
or indirect effects of TCDD.
Most panelists felt that the report's focus on fish early life stage effects as representing
the most sensitive aquatic effect/species early life stage was appropriate, given the present state
of scientific knowledge, in spite of the uncertainties. One panelist thought that this approach put
too much emphasis on a single-species-type solution to a complex ecological problem. Some
panelists proposed that measurement endpoints should be relevant to population-level effects and
should include both reproductive and developmental measurements. Ideally, mink data would be
used (in addition to available rat data) for estimating mammalian wildlife effects. Also, data on
a wild bird species (including Fl reproductive effects) would be useful. Additionally, panelists
discussed behavioral effects and the need for a complete fish life cycle test.
Panelists discussed the relative merits of empirical and mechanistic data. Some panelists
felt that additional data should be generated on the Ah receptor and its role in TCDD toxicity
for a wider range of species than presently have been tested in order to help evaluate
interspecies differences in sensitivity to TCDD. Others were concerned that there could be a
problem relating molecular-level receptor information back to ecological changes.
3.2. Issue 2—TEFs and BSAFs
Section 4.1 of the Interim Report (and the scenario) describe the use of toxicity equivalency
factors (TEFs) for TCDD-Hke compounds. Section 3.5 of the Interim Report discusses the use of
TCDD biota-sediment accumulation factors (BSAFs) for calculating bioaccumulation equivalency
factors for other related compounds. Comment on the use of these approaches for evaluating the
effects and bioaccumulation of dibenzodioxins and dibenzofurans in the paper mill effluent.
TEFs available for fish early life stage effects are based on rainbow trout mortality data.
The data appear to be valid for chlorinated dioxins, chlorinated furans, and polychlorinated
-13-
-------
biphenyl (PCB) congeners. Additivity is shown where compounds are full agonists, but may not
occur where compounds may be only partial agonists. Compounds have been shown to differ in
this respect between fish, birds, and mammals when in vitro hepatocyte cultures were tested.
While the assumption of additivity has not been adequately addressed for avian and mammalian
species, it is an appropriate assumption given the present state of knowledge. For the future, it
would be desirable to examine TEFs for other endpoints and for species other than rainbow
trout.
Panelists found the use of BSAFs to be a reasonable approach given the present state of-
knowledge, but one limitation is that data are available for only two field sites—Lake Ontario
and the Fox River. While the correlation between these sites was good, panelists felt that BSAF
data from additional field sites should be examined.
3.3. Issue 3—Use of Tissue Levels vs. Exposure Concentrations
Because of difficulties in extrapolating from various laboratory exposure conditions to
observed effects, the Interim Report (section 4.2.3.1) emphasizes using tissue levels of TCDD (rather
than exposure concentrations) to evaluate effects. Comment on the applicability of this approach to
evaluating the risks of TCDD from the paper mill effluent.
Certain panelists pointed out that toxic responses depended on TCDD concentrations at
the target organ sites and that reliance on tissue levels for regulatory purposes was necessary
given the difficulty of measuring TCDD levels in the water. Others recognized that general
changes in TCDD tissue levels may not imply that a specific effect will occur and that calculating
back from tissue levels to permit or discharge levels may be difficult for a risk assessor. In
general, however, the panel was comfortable with using tissue levels to evaluate TCDD effects as
described in the Interim Report.
3.4. Issue 4—Use of Laboratory Data to Predict Field Effects
The Interim Report uses both laboratory and field information to predict levels of TCDD in
fish and wildlife tissues that will cause adverse effects. The scenario proposes to use laboratory test
data at the individual level of organization to predict population changes in fish and wildlife.
Comment on the utility of available laboratory data to predict effects oh field populations and discuss
the associated uncertainties.
Certain panelists contended that available laboratory test data are sufficient to allow
reasonable decisions to be made, while others maintained that because of the potential
importance of subtle population and community effects, other types of tests, such as multiple
species tests, should be developed. Gaining a better understanding of the implications of
laboratory data also requires an understanding of the ecosystem (e.g., keystone species,
population control mechanism).
Several points made concerning extrapolation from laboratory test species to resident
species in the reservoir are listed below:
-14-
-------
• Extrapolation should be performed using all existing data on TCDD effects and
dose-response relationships in species in the same taxa.
• Assessment of the relationship of effects to tissue concentrations should take into
account residues in the whole body and all organs for which data are available*
• Effects to be examined should be of any type, ranging from mortality to cellular
and early molecular changes. Reproductive failure or changes directly linked to
reproductive success would be most relevant.
• Markers of change for which dose-response relationships have been demonstrated,
particularly those having a known mechanistic link to TCDD, may be valuable.
At present, the assessment most often will require extrapolation from model species to the
species of concern. It is important that the assessment consider the most sensitive model species
for which data are available. The lipid content of various organs and the biological variables
influencing mobilization and deposition in organs that are sites of action should be considered.
These would include eggs or embryos in the case of early life stage mortality, and brain, liver, or
gonad in the case of reproductive effects.
In the future, new dose-response relationships and an understanding of the mechanism of
TCDD action will provide additional model systems that could be incorporated into the
assessment. Determination of specific effects from tissue residue levels or body burdens must be
tied to estimated uptake (dietary) for each taxa (figure 3). Model species (or surrogate) data are
used to determine most sensitive endpoints (no observed effect level [NOEL]) for each category
of concern (figure 3). The allowable dietary uptake (or tissue residue, if available) can be back-
calculated to determine maximum allowable media concentrations.
3.5. Issues 5 and 6—Concentration-Response Curves and Data Extrapolations
Issue 5. The Interim Report sites data that indicate a very steep concentration-response
curve for TCDD effects in fish and wildlife. Discuss the implications of this observation for
evaluating ecological effects in the scenario.
Issue 6. The general summary of effect levels for aquatic species and associated wildlife
(Interim Report boxes 1 and 2, section 4) is based on extrapolations from a limited number of test
species and from tests that do not span complete reproductive cycles. Associated uncertainties are
summarized in section 5.1.3. Discuss the utility of these data and uncertainties for evaluating
ecological effects.
Panelists felt that these issues did not require further discussion, since both areas were
adequately addressed during the discussion of other issues.
-15-
-------
BIRDS
Adult Residues
Reproductive failure
Juvenile Residues
Residues contribute
to adult reproductive failure
Egg Residues
Failure to hatch
Dietary
Fish
MAMMALS
Adult Residues
Reproductive failure
Juvenile Residues
Residues contribute
to adult reproductive failure
Fetal Residues
Mortality
Deformed young
Residue Transfer
Adult Residues
Mortality
Behavior effects
Reproductive failure
Larval Residues
Mortality
Behavior effects
Egg Residues
(Maternal Transfer)
Mortality
Residue transfer to larvae
Figure 3. Residue vs. effects. Consumption of contaminated food by birds, mammals, and
fish will increase residues within the organisms, causing a range of effects that
depend on both the residue level and life stage of the organism.
-16-
-------
3.6. Issue 7—Additional Effects Data Not Considered in the Scenario
As discussed in the Interim Report, few data on the effects of TCDD on estuarine and
marine organisms have been reported (section 4.2.1.5), and no data were found in the literature for
TCDD effects on reptiles or marine mammals. Although all the current wildlife toxicity data were
reviewed, an analysis to establish an effects profile for terrestrial organisms was beyond the scope of
the report. Describe other effects data not identified in the Interim Report or the scenario that will
be important for future ecological risk assessments.
Panelists generally agreed that a reasonable risk assessment could be conducted with
existing data. Nevertheless, obtaining additional data in specific areas could strengthen the risk
assessment. A number of data gaps were pointed out along with suggestions for use in future
TCDD ecological risk assessments:
» Ah receptor information from a wider range of species (e.g., amphibians, reptiles,
invertebrates) is needed.
• Residue-effects data for a wide variety of species are needed.
• Aquatic plant effects are needed.
• Complete life cycle testing with fish should be conducted.
• Data relevant to effects on marine mammals are needed.
• Marine ecosystem effects in general are lacking.
• Effects on detrital communities in oligotrophic lakes should be assessed.
Panelists recommended that additional research be directed toward reptiles, specifically
snapping turtles and alligators that may represent top predators in many aquatic ecosystems,
particularly in the southern United States. Given the potential influence of external variables
(e.g., temperature) on sex ratios in reptiles such as turtles, panelists considered development of a
measurement endpoint sensitive to changes in sex ratios to be important.
3.7. Assessment and Measurement Endpoints
Although not specifically raised as an issue area, panelists discussed the suitability of the
assessment and measurement endpoints proposed in the hypothetical scenario. Panelists
suggested three assessment endpoints for the scenario:
• maintenance of the aquatic community;
• maintenance of sport fishery populations; and
• maintenance of piscivorous wildlife (birds and mammals).
-17-
-------
Measurement endpoints were discussed for both "real" situations (using existing data) and
"ideal" situations (where new data could be obtained). Measurement endpoints involving real
data include reproductive effects in fish (early life stage tests), birds (pheasant data), and
mammals (rat data). Some panelists were concerned that there was an inadequate basis for
extrapolation from the available data to the species and effects of concern in the scenario. Some
discussion also took place about the adequacy of TCDD data for the rat and chicken.
Ideally, measurement endpoint data could be expanded using information listed in section
3.6. Additional data needs might include a multispecies test system that could factor in direct
and indirect effects on other species such as benthic invertebrates. Additional single species tests
would include developmental as well as male and female reproductive toxicity tests for fish, birds,
and mammals (e.g., mink). Also, full life cycle tests (including exposure of Fl generation) would
be very useful. Tissue levels of TCDD could be measured in organisms, or eggs, for instance.
Test organisms should include resident species in the reservoir, if at all possible.
-18-
-------
4. COMMENTS ON STRESSOR CHARACTERIZATION
Peer panel members were asked to use information in the Interim Report as source
material to address eight issues concerning stressor characterization raised by the scenario.
Available information on TCDD physicochemical properties and exposure characteristics are
described in chapter 2 of the Interim Report, while bioaccumulation is described in chapter 3.
4.1. Issue 8—Uncertainties in Kow and
The Interim Report (sections 2.1 and 22) indicates that there is considerable uncertainty in
the estimates of parameters including K^ and K^ and the partitioning of TCDD onto organic
matter. This uncertainty results in part from difficulties in analytical measurements of various
fractions of TCDD in water. Since these limitations may affect predictions of TCDD partitioning
and exposure, please address how they should be handled in stressor characterization and conceptual
model development for this scenario.
HOCs are known to partition to solids as a result of their organic carbon content (i.e.,
lipophilic characteristics). Thus, HOCs will generally distribute themselves among a soluble
phase and a paniculate phase. TCDD is among those HOCs that have a relatively high tendency
to partition themselves in the paniculate phase. This characteristic of TCDD results in the fate
and transport of TCDD being tied to the fate and transport of solids in aquatic systems.
Experience with simulating HOCs in aquatic systems has revealed that if one succeeds in
quantifying the fate and transport of solids (including bulk transport, settling, resuspension,
burial, and internal primary production), then a substantial portion of HOC fate and transport
has been defined. This observation is programmatically fortunate since monitoring programs to
support site-specific HOC exposure assessment will be well served by an emphasis on the
measurement of solids and solids transport phenomena, a less-costly approach than a nearly
exclusive emphasis on chemical measurements.
Knowledge of the dominant role that solids dynamics play in determining the fate and
.transport of highly partitioning HOCs has contributed greatly to the ability to develop models
that are not chemical specific, but depend instead on measurable physicochemical characteristics
such as KO,., KOW, and Henry's Law constant. Therefore, it is necessary to continue to develop a
library of these physicochemical characteristics for the HOCs of concern. It may not be
necessary, however, to treat each HOC, homolog, or congener with a chemical-specific paradigm
of fate and transport.
The panelists agreed that there should be a standardization of protocols to be used when
determining both total organic carbon (TOC) and the partitioning to organic carbon (K,,,.). A
lack of standard methods will result in different values based on the method selected. The
panelists also agreed on the need to characterize the type or organic carbon, since from the
Green Bay study there is limited evidence for different K^. values for sediment and algal carbon.
Both laboratory and field validation for the E^. for selected isomers at concentrations ranging
from background to environmental levels are needed. This will answer the question of whether
laboratory-spike-type studies are providing results similar to those from field (or site specific)
situations.
-19-
-------
Once these data are produced, a central database library listing the results should be
created. This information should be screened for compliance with QA/QC criteria. If the data
do not meet such criteria, they could be listed with a footnote noting any irregularities. This
data base also could list field data with K^ and tissue concentrations on a lipid normalized basis.
Such information would allow for the screening of species-specific information in similar
ecosystems. These data could be used by the risk manager to get information for particular
species of concern and could help prevent duplication of field and laboratory studies.
4.2. Issue 9—Exposure Routes
The Interim Report (section 2.4) indicates that most TCDD exposures witt arise from food
consumption and contact with sediments or suspended solids, with the water pathway being less
important Address the implications of this information relative to the exposure routes in the
conceptual model
Given the high K^ values for TCDD and many PCDD/PCDF congeners, the main routes
of exposure for most receptors will be through food ingestion and contact with or ingestion of
sediment or suspended solids. Implications of this for development of the conceptual model and
the risk assessment include:
• Food webs and sediment exposure pathways need to be characterized in detail,
including possible spatial and seasonal variations.
• Assimilation of PCDDs/PCDFs from key food items and sediment needs to be
determined for receptors of concern.
• BSAFs, BSSAFs, and BAFs are important, while BCFs may be negligible (except
for algal species).
• While the Interim Report indicates that BSAFs apply as much to pelagic organisms
as to benthic organisms, workshop panelists felt that the use of an empirical
BSAF/BSSAF-based approach is more applicable for benthic food webs. Panelists
indicated that food web models may be needed for other systems where higher
trophic level fish species depend on pelagic food webs or algal foods.
Adequate characterization of food webs is critical for ecological risk assessments of
PCDDs/PCDFs, especially assessments based on predictive models such as those of Thomann et
al. (1992a). The conceptual model for the hypothetical scenario needs to include key food web
species for each receptor of concern, as well as ingestion of sediment or suspended solids by
some species. Three types of food webs (Paine, 1980) should be considered.
s>
» A connective web, which includes essentially all possible web linkages, should be
considered for any receptors that are linked or for rare, threatened, or
endangered species.
-20-
-------
• A materials-flow web, which emphasizes the relatively important linkages for
transfer of PCDDs/PCDFs, should be the basis for an exposure assessment for all
receptors of concern.
• A functional web, which emphasizes linkages among strongly interacting species
(e.g., keystone predators feeding on competitive dominants), should be considered
to prioritize receptors for detailed consideration in the risk assessment and to
develop risk hypotheses about effects of exposure on community structure.
Trophic compartments hi food webs may be defined on the basis of species, life stage,
sex, etc. (i.e., trophic species rather than simple taxonomic species). (Terrestrial organisms may
also be key prey items in food webs for may aquatic receptors.) For "ideal" assessments, spatial
and seasonal variability in food webs should be included.
If dynamic food chain or food web models are desired, assimilation of PCDDs and
PCDFs in ingested prey and sediments by key receptors may need to be measured in laboratory
experiments. The relative importance of this data gap should be evaluated further through
sensitivity analysis of food web exposure models. Assimilation efficiency will likely vary with prey
species and the nature of the organic carbon.
Empirical determination of BSAFs and BSSAFs from site-specific measurements will be
an appropriate alternative to detailed food web modeling, especially for lower trophic level
species associated with sediments. The nature of organic carbon hi different sediment types will
likely affect BSAF and BSSAF values. Modeling should still play a role for higher trophic level
species dependent on pelagic food webs and for cases where examination of the temporal
disequilibria is a management objective. Since empirical measurements of BAFs are not
available at present, estimates of BAFs will be needed for modeling. BCFs will not be useful
except in the case of algae, which form the base of many aquatic food chains.
4.3. Issue 10—Fate-and-Transport Models
Fate and transport models are beyond the scope of the Interim Report, but are clearfy
critical for risk assessment. In stressor characterization and development of the conceptual model,
they wul be necessary for linking TCDD source loads to concentrations in different compartments of
the reservoir.
a. Comment on the availability of fate and transport data/models suitable for use with
TCDD.
b. Discuss the applicability of available transport models for predating the deposition of
particulate-bound TCDD in the reservoir.
Panelists considered the availability of fate-and-transport models for quantitatively
relating the source loadings of TCDD to TCDD concentrations hi various compartments of an
aquatic system such as the reservoir in the case study. The simple answer was that mass balance
models simulating the fate and transport of HOCs have been developed and applied for a variety
of chemicals in a variety of environmental settings. These models range in their spatial,
-21-
-------
temporal, and kinetic sophistication from relatively simple steady-state, screening-level models
(e.g., Endicott et al., 1991) to more "state-of-the-science" models (e.g., Bierman et al., 1992;
DePinto et al., 1994) that address the problem at a higher resolution of space, time, and kinetic
formulation.
The panelists recognized that no single model is best for all applications. The level of
complexity (or resolution) of a model required for a given application is determined by two basic
factors: the complexity and resolution of the management questions being asked, and the
resources (mainly in terms of data acquisition) available to support the model. (In general,
increased model complexity requires increased data/resources for application.)
The panelists felt that for a typical situation, such as in the Omigoshee Reservoir
scenario, where the questions being asked range from simple to complex, a tiered approach is
recommended. Initially, a simple but conservative calculation would be used, progressing toward
a more complex modeling approach that provides the optimum reliability (i.e., utility) for the
most complex questions and for the resources available. Panelists noted that it is possible to see
a decrease in reliability (model certainty) if a model is more complex than warranted by the
questions posed and resources at hand.
When contemplating the use of models to calculate the chemical exposure in aquatic
systems under various management scenarios, it is essential that the prospective user has a clear
picture of the levels of models that are available, their corresponding features and data
requirements, and the management or investigation questions for which the models can provide
answers or guidance. Table 2 summarizes this information for aquatic models that are capable of
simulating exposure levels for HOCs. Specifically, these models compute environmental
concentration^) resulting from alternative loads of the chemical to the system.
The first two models in the table (level 0 and la) simulate only the total chemical in the
water column, while the remainder of the models simulate HOCs in the paniculate and the
dissolved phase and in the sediment layer (except for level Ib). Therefore, many of these models
supply information regarding the environmental distribution of a chemical loaded to an aquatic
system in four environmental compartments, namely: (1) dissolved phase in the water column;
(2) participate phase in the water column; (3) dissolved phase in the sediment layer; and (4)
paniculate phase in the sediment layer. Beyond these features, the levels of the models reflect
increasing capabilities to simulate HOCs at increasing levels of spatial resolution and as a
function of time. The last level (level 3) represents a highly variable group of models referred to
here as complex models that may address specific additional areas of concern, such as
quantitative characterization of model uncertainty, more complex hydrodynamics (e.g., estuaries
and three dimensional phenomena, such as saline or temperature layers), and more refined
representations of chemical, biological, or population relationships.
The listed model features, data needs, and management answers represent a progression
in characteristics and needs. Several guidance documents have been published by EPA and
others to aid the user in model selection, use of specific aquatic chemical models, collection of
supporting field data, and evaluation of model inputs (e.g., Delos et al., 1984).
The panelists noted that it would be particularly useful if future risk assessment
guidelines included a protocol for applying the tiered approach along with criteria (aspects of
-22-
-------
.
at
H
I
ee -3
a -ts
jrt 13
!
I
£
-23-
-------
complexity of management questions) for governing the progression toward higher resolution
(spatial, temporal, kinetic) modeling approaches.
4.4. Issue 11—Additional Exposure Issues
List some of the major exposure issues not relevant for the paper mill scenario that may be
encountered in future ecological risk assessments (e.g., marine/estuarine, terrestrial),
Panelists identified the following exposure issues (not relevant for the paper mill
scenario) in both premeeting comments and workshop discussions.
• Terrestrial exposure information for wildlife species is generally lacking. Often
information is insufficient regarding relative source strength (e.g., atmospheric,
point and nonpoint) and input to terrestrial ecosystems to perform terrestrial risk
assessments.
» Evaluation of exposure in marine systems, which was not conducted for the
present scenario, is particularly difficult due to the openness of marine systems.
Additional confounding issues include the problem of determining the percentage
of a population that is exposed, the migratory and behavioral patterns of marine
species, the role of recruitment and recovery, time-variable exposures to mobile
species, and spatial heterogeneity resulting from the multiple source inputs
characteristic of estuaries.
» Fish exposure time is rarely well known, often time-variable, and generally
confounded by a lack of understanding of how fish interact with sediment to
accumulate chemical contaminants.
» Information on spatial and temporal patterns of dietary uptake of contaminated
prey that may be particularly important for TCDD and related contaminants is
needed.
• Recovery time is not systematically evaluated, particularly in regard to the impact
of catastrophic or rare meteorological events.
• The paper mill example does not explicitly discuss the role of a mixing zone,
which may be appropriate given the nature of the endpoints.
Panelists suggested that there are substantial exposure issues for both aquatic
(particularly marine/estuarine) and terrestrial systems that are outside the paper mill conceptual
model but that will need to be addressed in other site-specific problem settings.
4.5. Issue 12—BCFs and BSAFs
The Interim Report (sections 3.2-3.5) summarizes available data on TCDD
bioconcentration, bioaccumulation, biomagnification, and biota-sediment accumulation factors from
-24-
-------
laboratory experiments and field measurements. Discuss the applicability of these factors to stressor
characterization for the paper mUl scenario.
Panelists noted that BCFs suffer from two limitations: the need for accurate
measurements of dissolved TCDD concentrations in the water, and the use of data from
laboratory studies that often lack realism under site-specific conditions. Further, the ability to
obtain accurate measures and/or predictions of dissolved TCDD is confounded by uncertainties
associated with K^ measurements. For example, the BCF-K^ relationship departs from linearity
when log K^ exceed 6, which includes the range of K^s for TCDD and related compounds.
Panelists felt that BSAFs should play an important role hi TCDD risk assessment.
BSAFs are likely to have a lower uncertainty, and they may serve as a steady-state, upper-bound
body-burden estimation, particularly for lower trophic levels (e.g., plankton and invertebrates).
Since BSAF values are calculated using concentrations hi muscle, tissue-dose specificity may be
much higher than indicated by typical body burdens. Finally, if BSAFs are to be used
extensively, a library of BSAF values, representing different species, lipid, organic carbon,
sediment types, and ecosystems, will be necessary.
4.6. Issue 13 — Applicability of Lake Ontario BAF Data
The Lake Ontario Upid-normalized bioaccumulation factor for dissolved TCDD (BAFJ) may
be useful as a predictor of residue levels in other systems if the concentration of freely dissolved
TCDD in water (C*) can be estimated accurately (Interim Report section 3.3). Comment on the
applicability of this BAF for the paper mill stressor characterization.
Panelists felt that extrapolation of the Lake Ontario data to other sites would be
appropriate as a Tier I screening level assessment. This is an area where a library of values from
different areas would be particularly useful. In applying this information, ecosystem differences
have to be considered, with a key focus on the dietary habits of consumer organisms. Food chain
length and lake trophic status can influence TCDD residue levels.
4.7. Issue 14 — Biomagnification
The Interim Report (section 3.4) indicates that biomagnification is significant between fish
and fish-eating birds but not between fish and their food. Comment on the biomagnification
pathway relative to stressor characterization and the conceptual model
Based on the properties of TCDD and related compounds, the panel felt that it may be
premature to conclude that biomagnification does not occur among aquatic species (i.e., between
plants and animals, invertebrates and fish, or forage fish and piscivorous fish) despite the
available data that suggest that biomagnification is not occurring. Although the panel found the
available information to support or refute the occurrence of biomagnification to be limited, it
acknowledged — based on existing data — that biomagnification does not appear to be a significant
process within aquatic food webs excluding mammals and birds.
-25-
-------
The panelists felt that biomagnification did appear to occur between fish and piscivorous
mammals and birds. The significance of this phenomenon for risk assessment was thought to
depend, in part, on how risks are estimated. At present, risks to mammals and birds can be
estimated from dietary levels of TCDD and related compounds. The effects data are not based
on the body burdens in these animals and, therefore, the issue of biomagnification from fish body
burdens to mammal and bird body burdens is somewhat moot. If effects data are developed in
the field and laboratory in relation to body burdens, then information on biomagnification
processes would be helpful for relating body burdens and associated effects to dietary levels.
4.8. Issue 15—BAF Uncertainties
Uncertainties associated with bioaccumulation factors are discussed in section 5.1.2 of the
Interim Report Discuss the relevance of these uncertainties to the prediction of TCDD residues in
Omigoshee Reservoir biota.
Previous panel comments on uncertainties about the procedures used to estimate BAFs
apply here as well. The primary concern is the inability to measure freely dissolved TCDD
concentrations in the water phase.
4.9. Key Research Needs in Stressor Characterization
Panelists considered the following areas to be critical topics for future research:
• Improve the K,,,. estimate for TCDD.
• Standardize the methods(s) for measuring KO,. for highly lipophilic compounds.
• Standardize methods for lipid and organic carbon measurements.
" • Compile a library of BCF, BAF, BSAF, and BSSAF values. Of greatest interest
would be BSAF and BSSAF values from field data.
-26-
-------
5. COMMENTS ON THE CONCEPTUAL MODEL
Peer panel members were asked to use information in the Interim Report as source
material to address four issues concerning conceptual model development raised by the scenario.
Different approaches to bioaccumulation are indicated in the scenario arid are discussed in
chapter 3 of the Interim Report; in particular, BSAFs are discussed in section 3.5. In addition to
responding to the individual issues, panelists suggested revisions and additions to the conceptual
model.
5.1. Responses to Conceptual Model Issues
5.1.1. Issue 16—Conceptual Model Focus on Fish and Piscivorous Wildlife
Consistent with the Interim Report, the conceptual model focuses on effects on fish and
wildlife that consume fish. Comment on whether this approach captures the full range of potential
ecological effects for this scenario.
Panelists differed in their perspectives on this topic. Some felt that the conceptual
model's focus on fish and piscivorous wildlife probably captures the most important ecological
effects and is a reasonable attempt to ensure environmental protection by concentrating on
"worst case" situations. Others felt that the degree of protection for non-target components of
the ecosystem would be better understood only through long-term monitoring programs and that
there could be subtle effects on trophic structure. In particular, the possibility of effects on
invertebrates was raised, although the assumption is that these animals do not possess the Ah
receptor.
Some panelists were more pessimistic, maintaining that a wide range of unanticipated
effects are likely to occur because of the present lack of data and information. In this view,
indirect effects have not been adequately considered. Examples might include effects on a forage
fish (e.g., a minnow) or an invertebrate (e.g., a crayfish) upon which other fish or mammalian
species (e.g., the otter) rely for food.
5.1.2. Issue 17—Linking TCDD Loadings to Tissue Residues
The Interim Report emphasizes using tissue residue levels to estimate the adverse effects of
TCDD. To conduct the risk assessment outlined by the conceptual model, however, it will be
necessary to link predicted loadings of TCDD in the paper mill effluent to residues in the organisms
as identified in the assessment endpoints. Discuss the utility of available risk assessment tools for
accomplishing this goal
Analyses are most likely to be data limited, not model limited. Panelists noted that a
range of models is available for estimating exposure fields in sediments and water. No single
model is correct for all situations, but an approach that involves a progression from simple to
more complex models makes the most sense. Panelists gave examples, listed below, of the wide
range of models available.
-27-
-------
• Simple box models
» Fugadty Level m
• Screening models
• Food and Gill
Exchange of Toxic
Substances
(FGETS) model
• Exposure Analysis
Modeling System
(EXAMS)
» Water Analysis
Simulation Program
(WASP4)
• RIVER/FISH
• Green Bay model
• Thomann food
chain-type model
Panelists indicated
that a risk assessment for
TCDD must put much
more emphasis on the
importance of
physicochemical fate-and-
transport modeling to link
exposure to loadings. For
the hypothetical scenario,
models that predict
concentrations in
sediments and on
suspended sediments make
the most sense. Linking
these concentrations and
possibly water
concentrations of TCDD
to fish body burdens can
be done using either appropriate factors (e.g., BSAFs) or a food chain model. One possible
approach is discussed in text box 2. While a model that relies heavily on partitioning coefficients
might be a shaky "house of cards," it may be the best that can be achieved at the present time.
Text Box 2. Predicting TCDD Concentrations in the
Hypothetical Scenario
A two-step modeling approach is needed to predict TCDD
concentrations in fish. The first step is to use a fate-and-transport
model to fink effluent concentrations of TCDD to concentrations in
bed and suspended sediments (also dissolved organic carbon [DOC]
and dissolved) in the depositional zones of the reservoir. The
model could be a simple steady-state model consisting of single
water column and sediment compartments. A more realistic
simulation of the transport and sedimentation of TCDD from the
paper mill effluent would require a multisegmented model run at
either steady state (e.g., constant flow, sedimentation rates) or in
the dynamic mode. A critical parameter in either model is the K^
value. Ideally, K^ values are required for the paper mill effluent,
the suspended solids in the river, and algae in the lake and bed
sediments, because there is evidence that K^ values can vary
depending on the type of carbon (and whether algae/phytoplankton
are in a growth phase). For an initial screening of TCDD
transport/fate, a single K^ value could be used with the steady-state
single water/sediment compartment model.
The second step in the modeling approach would be to
apply a food chain model or BSAFs to predict concentrations in
bed sediments. Because pharmacokinetic parameters for TCDD in
invertebrates are uncertain, a possible approach is to use BSAFs to
predict concentrations in lower food chain organisms, then use a
food chain model to predict concentrations in fish. Pharmacokinetic
parameters for TCDD in fish are readily available, although
information on assimilation and depuration by larger fish is limited.
The food chain model has the advantage of being able to
accommodate multiple age classes and feeding relationships,
whereas the BSAF approach would require a lot of empirical data
to generate the same library" of relationships between fish and
sediment. Given monitoring results from paper mills as well as
studies such as Rassmussen et al. (1990) on lake trout, we know
that the food chain relationships of various fish species and their
age classes will be important in predicting tissue concentrations.
Gobas (1992) essentially used a combined BSAF/food chain
approach for modeling concentrations of PCBs in Lake Ontario.
This approach differs from that used by Thomann et al. (1992b),
which uses phannacokinetics for all trophic levels.
-28-
-------
Panelists suggested examining other similar systems with existing sources to help in
developing empirical relationships. Where possible, incorporation of site-specific (experimental)
data also would help reduce uncertainty in estimates. Use of whole body tissue levels may be
inappropriate for assessing effects, and there may be a need to estimate doses to specific target
organs. In any case, monitoring will play an important role in checking the predictions of the
assessment.
5.1.3. Issue 18—Applicability of BAFs and BSAFs to the Conceptual Model
The Interim Report describes the limited field data that are available for estimating BAFs
and BSAFs. Discuss the applicability of these factors to the Omigoshee Reservoir conceptual model
Panelists felt that BSAFs (and BAFs) provide simple, straightforward models and may be
the best available technique for estimating tissue concentrations. Panelists also felt that it would
be helpful to compare characteristics of the reservoir to areas for which these factors have been
developed.
The utility of the BSAF and BAF approaches depends to a large degree on the accuracy
of organic carbon and lipid measurements. In addition, BSAF and BAF values will be site
specific as well as spatially and temporally variable. Uncertainty could be decreased by making
measurements in a variety of systems and for a variety of environmental conditions.
Panelists suggested using BSAFs for lower trophic levels and perhaps combining this
approach with food chain models for piscivorous fish. Another suggestion was to use the
relationships established for PCB BSAFs as a tool for estimating TCDD BSAFs, given the larger
data sets available for PCBs.
5.1.4. Issue 19—Temporal Dynamics of TCDD in the Reservoir
The temporal dynamics and disequilibrium situations commonly associated with TCDD are
mentioned in the Interim Report (section 2.3). Comment on how these aspects should be
considered in establishing (1) the time course for the build-up of TCDD levels following initiation of
the paper mitt discharge, and (2) the time course for the decrease of TCDD levels and recovery of
biota should the paper mitt cease operation.
Panelists felt that the relevant time course depends on the area under consideration. For
purposes of estimation, steady-state or quasi-steady state conditions should be relied upon with a
temporal constraint being the operational life of the paper mill facility. Attempting to introduce
kinetics would involve so much uncertainty that it would be better to use equilibrium models. If
there are management questions related to system response time, however, then a time-variable
(not steady state) set of models must be used. Analysis of effects could use either the time
profile for exposure or a time-weighted average of calculated chemical concentrations. The
analysis should include a "worst case" scenario.
Important information for the hypothetical scenario includes TCDD loading, reservoir
volume and flushing rate, hydraulic retention time, overflow rate (ratio of mean depth to
-29-
-------
hydraulic retention time), organic carbon levels, and lipid concentrations in fish. The time
course for build-up will depend on hydrodynamics and especially on sediment transport
dynamics. Examination of data from existing facilities (including those that have changed
operations) will assist in determining the likely time course for build-up and decrease of TCDD
levels.
While media contamination may be a straightforward input-output with spatial/temporal
variability due to lake morphology, tissue concentrations will more likely follow bounded chaotic
dynamics since tissue concentrations depend heavily on a variety of biotic and abiotic factors.
5.2. Comments on the Conceptual Model
Panelists modified the proposed conceptual model for the hypothetical scenario in several
ways. Panelists rewrote the text (appendix G) and developed additional figures to provide a
more complete description of the conceptual model. The overall approach involved:
• linking chemical loadings to water and sediment levels;
• relating these exposure levels to residues in aquatic organisms;
» relating organism residue levels to reproductive and other effects (see section 3.4);
and
• relating expected effects to changes at the population level.
The overall approach relating effluent discharge levels, fate-and-transport considerations,
and tissue residues and effects is shown in figure 4. The problem may be viewed as establishing
the effects of concern to the system, then working backwards to determine release levels from
the paper mill that will prevent adverse effects from occurring (figure 5). Further details on
TCDD fate-and-transport relationships are given in figure 6. Models used to evaluate fate and
transport are likely to be data limited; the best available information would have to be used,
whether from experimental data or from the literature. As mentioned earlier (see sections 4.3
and 5.1.2), modeling these relationships can be performed with varying degrees of sophistication.
It may be necessary to divide the reservoir up into several segments (e.g., main channel vs.
tributary arms) to deal with spatial heterogeneity. If a quantitative analysis of uncertainty is
possible, model outputs can be expressed in probabilistic terms (e.g., figures 7a and 7b).
Calculating sediment and water concentrations of PCDDs and PCDFs from allowable
residue levels in fish can be carried out using either an empirical BSAF (Q/C,,,.) or a food chain
model. Table 3 lists some of the considerations involved with these two approaches.
Panelists also discussed how to relate anticipated effects at the individual level to
population level effects. Qualitatively, one could assume that changes in a measurement
endpoint adequately reflect potential changes in the assessment endpoint without further
extrapolation; however, more quantitative approaches also are possible. For example, population
models such as life table models or individual-based models could be used. The main drawback
is the amount of data required to successfully apply these models. Another approach is a "rule
-30-
-------
Population integrity and Stability^
f-*—— Population Modeling
Avian Residue Effects
Fish Residue Effects
I Dietary Uptake
BSAF-Food Chain Model
Fish and Invertebrate
Tissue Residues
I
Food Chain Model
(Higher Trophic Levels)
Empirical BSAFs
(Lower Trophic Levels)
I
Sediment-Water
Concentrations
Mammal Residue Effects
Dietary Uptake
Transport-Fate Model
Photolysis A I
Volatilization
Effluent Discharge
Figure 4. Overall model. The risk assessment needs to establish the linkages between
TCDD loadings in the effluent, TCDD fete and transport within the reservoir,
residue levels in biota, and effect on biota.
Key: BIC = biological carbon; Cw = TCDD concentration hi water; Cdoc = TCDD
concentration associated with dissolved organic carbon; Cpoc = TCDD concentration
associated with paniculate organic carbon; DIS = dissolved; POC = paniculate organic
carbon. .
-31-
-------
Ecological
Risk Assessment
Problem Formulation
Define Ecological
Receptors, Endpoints,
and Chemicals
Home Range
and Diet Analysis
Develop Fish Tissue
Residue NOELs for
Fish and Wildlife
For Each Fish Species
Select Lowest Fish Tissue
NOEL Among Receptors
Empirical
BSAFs
Define Allowable
Fish Tissue Residues
I
i
r
Bioaccumulation
Model
Define Allowable
Sediment Concentrations
Receptor Species A/Life Stage A/
Endpoint A
Receptor Species A/Life Stage B/
Endpoint B
Receptor Species B/Life Stage A/
Endpoint A
Etc.
Additional Policy
Input
Fate/Transport
Model (e.g., WASP4)
Define Allowable
Discharge of PCDDs/PCDFs
Figure 5. Developing discharge permit limits based on ecological risk assessment. (R.A.
Pastorok, FIT, Premeeting Comments. See appendix C.)
-32-
-------
Upper Mixed
Sediment
Figure 6. Fate-and-transport diagram. This figure provides a more detailed view of the
linkages shown in figure 4. Pathways for partitioning of TCDD between water,
sediment, and biota are shown. DOC—dissolved organic carbon;
POC—paniculate organic carbon; BIG—biota.
-33-
-------
TCDD Log
(concentration
in compartment)
'( I I I I I I
-3-2-10 12 3
Probability (Z score)
Figure 7a. Example fate model outputs—TCDD (log [concentration in compartment]) vs.
probability.
TCDD
(concentration
in compartment)
.- 95%
Confidence
Loading
Figure 7b. Example fate model outputs—TCDD (concentration in compartment) vs. loading.
-34-
-------
Table 3
COMPARISON OF BSAFs AND FOOD CHAIN MODELS
Category
BSAF
Food Chain Model
Ecoreceptors
• Fish
• Wildlife food
catfish bass, crappie, bluegill
fish and benthic invertebrates fish—pelagic
Data Types
lipid, TOC
BSAFs by species, age, tissue
wildlife diet
allowable C
population age structure, diet
by age class
lipid, TOC
assimilation efficiency
allowable Q, C^.
Analytical Tools
Uncertainty Analysis
method of standardization
data bases
framework for predicting
BSAFs (figure 8)
joint probability
within-site variance
between-site variance
extent of disequilibrium
nature of TOC, lipid
model structure
software
Monte Carlo
Validation
ground truth with PCBs
ground truth with PCBs
-35-
-------
BSAF SB C|jpid/Coc
R
I
V
E
R
Food
Chain
Model
(Ecosystem)
L
A
K
E
x
^>
-------
of thumb" method. One panelist indicated that EPA uses a 20 percent effect level to distinguish
"acceptable" from "unacceptable" mortality. The population effect of a 20 percent decrement,
however, will clearly depend on the overall fecundity of the population.
-37-
-------
-------
6. REFERENCES ,_
Bierman, VJ., Jr.; DePinto, J.; Young, T.; Rodgers, P.; Martin, S.; Raghunathan, R. (1992)
Development and validation of an integrated exposure model for toxic chemicals in
Green Bay, Lake Michigan. Final report for U.S. EPA Cooperative Agreement. CR-
814855, ERL-Duluth. September.
Cook, P.M.; Kuehl, D.W.; Walker, M.K.; Peterson, R.E. (1991) Bioaccumulation and toxicity of
TCDD and related compounds in aquatic ecosystems. Banbury report 35: Biological
basis for risk assessment of dioxins and related compounds. Cold Spring Harbor
Laboratory Press, pp. 143-167.
DePinto, J.V.; Raghunathan, R.; Sierzenga, P.; Zhang, X.; Bierman, V.; Rodgers, P.; Young, P.
(1994) Recalibration of GBTOX: An integrated exposure model for toxic chemicals in
Green Bay, Lake Michigan. Prepared for U.S. EPA—Large Lakes and Rivers Research
Branch, Grosse He, MI. March.
Endicott, D.D.; Richardson, W.; Parkerton, T.; DiToro, D. (1991) A steady state mass balance
and bioaccumulation model for toxic chemicals in Lake Ontario. Final report to the
Lake Ontario Fate of Toxics Committee, U.S. EPA, ERL-Duluth, LLRS, Grosse lie, MI.
Gobas, F.A.C.P. (1992) Modeling the accumulation and toxicity of organic chemicals in aquatic
food chains. In: Gobas, F.A.C.P. and J.A. McCorquodale, eds., Chemical dynamics in
freshwater ecosystems. Ann Arbor, MI: Lewis Publishers, pp. 129-152.
Paine, R.T. (1980) Food webs: linkage, interaction strength and community infrastructure. J.
Animal Ecol. 49:667-685.
Rassmussen, J.B.; Rowden, J.; Lean, D.R.S.; Carey, J.H. (1990) Food chain structure in Ontario
lakes determines PCB levels in lake trout (SaJveUnus namaycush) and other pelagic fish.
Can. J. Fish. Aquatic Sci. 47:2030-2038.
Thomann, R.V.; Connolly, J.P.; Parkerton, T.F. (1992a) An equilibrium model of organic
chemical distribution in aquatic food chains. Environ. Sci. Technol. 23:699-707.
Thomann, R.V.; Connolly, J.P.; Parkerton, T.F. (1992b) An equilibrium model of organic
chemical accumulation in aquatic food webs with sediment interaction. Environ. Toxicol.
Chem. 11:615-629.
U.S. Environmental Protection Agency. (1984) Technical guidance manual for performing waste
load allocations. Book II: Streams and rivers, Chapter 3: Toxic Substances. Office of
Water Regulations, Washington, DC. EPA/440/84-4/022. June.
U.S. Environmental Protection Agency. (1992) Framework for ecological risk assessment. Risk
Assessment Forum, Washington, DC. EPA/630/R-92/001.
39
-------
tJ.S. Environmental Protection Agency. (1993a) Interim report on data and methods for
assessment of 2,3,7,8-tetrachlorodibenzo-p-dioxin risks to aquatic life and associated
wildlife. Office of Research and Development, Washington, DC. EPA/600/R-93/055.
U.S. Environmental Protection Agency. (1993b) A review of ecological assessment case studies
from a risk assessment perspective. Risk Assessment Forum, Washington, DC.
EPA/630/R-92/005.
40
-------
APPENDICES A-C
PREWORKSHOP MATERIALS
-------
-------
APPENDIX A
A PRELIMINARY PROBLEM FORMULATION FOR A DIOX1N SCENARIO:
PROPOSED PAPER MILL ON A SOUTHERN RESERVOIR
A-l
-------
-------
A PRELIMINARY PROBLEM FORMULATION FOR A DIOXIN SCENARIO:
PROPOSED PAPER MILL ON A SOUTHERN RESERVOIR
Note: EPA scientists have created the attached scenario to supplement the Interim Report for
the workshop exercises on TCDD risk assessment. The scenario provides background
information for a hypothetical reservoir and presents issues in three areas (stressor
characterization, ecological effects and endpoint selection, and the conceptual model),
with each area drawing on information in the Interim Report as a starting point. The
scenario is presented to promote discussion on TCDD ecological risk assessment in
general, EPA is not asking for guidance on how to assess the risk of TCDD discharges
from paper mills.
Background
Introduction. A paper mill is proposed on the Igotchyala River, 5 km upstream
from the Omigoshee Reservoir in the southern United States. The reservoir is an
important recreational area, supporting a large sport fishery and a variety of avian and
mammalian wildlife on public land along its shores. No other industrial development
has ever existed in the area arid, although the adjacent land is mostly forested, no
logging has occurred over the last 30 years. Chemical and biological surveys indicate
that the watershed currently is essentially free from significant stress due to toxic
contaminants.
Because the proposed mill will use chlorine in its bleaching process, the
production and discharge of 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD) and other
chlorinated dibenzodioxins and dibenzofurans is expected. Because low doses of
TCDD can significantly effect egg viability and/or the survival of young fish, mammals,
and birds in laboratory tests, the presence of TCDD and associated organic
contaminants could impact fish and wildlife populations of the reservoir from this
proposed facility.
Risk Management Goals. Under the authority provided by the Clean Water Act
(307a) and federal regulations (49 FR 9016), the state is required to control the
discharge of toxic pollutants to the Nation's surface waters through the NPDES permit
limit process. Because of this mandate, operators of the facility seek to maintain
discharges from the proposed facility to below levels expected to be detrimental to the
fisheries and wildlife of the reservoir. This risk assessment will evaluate the adequacy
of the proposed mill effluent treatment plan and the anticipated chemical discharges in
relation to potential effects on fish and wildlife, based on the concentrations of
chemicals at steady-state exposure conditions expected under average annual inputs
of water and solids to the reservoir. Other stressors and endpoints are subjects of
other assessments and will not be considered here. Results of the assessment will be
used to determine final permit conditions and effluent treatment standards.
A-3
-------
Ecosystem Description. The reservoir (Figure 1) is composed of a broad
central basin with three large arms corresponding to the main channel of the river
(circled 1 on Figure 1) and two major tributaries (circled 2 and 3 on Figure 1). The
shoreline is highly irregular, with numerous coves and inlets of small tributaries. The
average total suspended solids concentration in the main channel and major
tributaries is 20 mg/L, with an organic carbon content of 10%. Each of the major arms
of the reservoir retains about 50% of the sediment load from its respective rivers, with
the remainder reaching the main basin of the reservoir. Sediments range from 2 to
15% organic carbon on a dry weight basis, with an average of 5%.
The reservoir has a substantial warm-water sport-fishery including largemouth
bass, catfish, crappie, and bluegills. The reservoir is moderately productive, with
diverse phytoplankton and littoral vegetation. There are healthy communities of
pelagic and benthic invertebrates.
The shoreline supports a variety of avian wildlife, including many species of
birds that are primarily piscivorous (e.g., herons and egrets; grebes, cormorants, and
various diving ducks; osprey and bald eagle; several species of gulls and terns) and
others that feed heavily on emergent aquatic insects (e.g., various fly-catching
swallows and warblers). Mammals such as the river otter, muskrat, and mink are
found along the shores of some of the coves and tributaries; however, the diet of the
otter and mink are only partly from fish from the reservoir itself.
Stressor Characterization for the Southern Reservoir
The effluent from the proposed paper mill will contain TCDD and other
poiychlorinated dibenzodioxins (PCDDs) and dibenzofurans (PCDFs), which are
formed in paper production which uses chlorine-bleaching. TCDD is highly
hydrophobic and associates strongly with organic matter, distributing primarily into the
sediments, suspended solids, and biota of an aquatic system. This results in low
dissolved concentrations of PCDDs and PCDFs in water. The mill effluent, following
the proposed treatment, is predicted to contain TCDD, 1,2,3,7,8-PeCDD, 2,3,7,8-
TCDF, 1,2,3,7,8-PeCDF and 2,3,4,7,8-PeCDF. Other chemicals that are known to
contribute to toxic effects through an Ah receptor mediated mode of action are not
predicted to occur in the effluent. However, the role of other chemicals, including
certain PCB congeners that are present in most aquatic ecosystems, in modifying the
effect of TCDD through antagonism or synergism is unknown. There are currently no
significant sources of TCDD and related compounds to this reservoir, the TCDD level
in fish being <0.5 pg TCDD/g whole fish. Total PCB levels in fish are 500 ng/g
(presumed to be from atmospheric deposition). There are no data for concentrations
of bioaccumulative organic chemicals in water, sediments, aquatic life (other than fish)
or wildlife associated with the Omigoshee Reservoir.
A-4
-------
The river has no depositional zones that would result in significant loss of the
chemical prior to reaching the reservoir and essentially 100% of the discharged
chemicals will appear as a point source in the main arm of the reservoir (Figure 1).
The discharge of these chemicals will be continuous and is expected to be relatively
constant. Despite the use of expected steady-state conditions for the risk
assessment, the exposure to aquatic life in the reservoir should be rather
heterogeneous, with these chemicals being most concentrated in the main arm of the
reservoir, somewhat less so in the central basin, and even less in the other arms
where uncontaminated tributaries enter (Figure 1). The effect of sediment distribution
and burial on availability are significant aspects of exposure that must first be
evaluated with appropriate exposure models. Additional data and models for fate and
transport of hydrophobic organic chemicals are provided in a recent EPA report on
estimating exposure to dioxin-like compounds (U.S. EPA, 1992b) and the WASP4
model user's manual (Ambrose, 1988). :
A-5
-------
•p
o
a
I
u
i
•
41
S3
35
H H
n?
23
o o
M M
•H-H
14 M
O O
J04*
4J 0
•H
O
n*>
end
• H
HI
•
-I
2
A-6
-------
Ecological Effects and Endooint Selection for the Southern Reservoir
Ecological Effects of TCDD. TCDD has been demonstrated in the laboratory to
be highly toxic to fish and to many warm-blooded vertebrates. Based on lethal doses,
a variety of fish, mink, and gallinaceous birds are especially sensitive. The survival of
early life stages of fish and reproduction in mammals and birds have been shown to
be the most sensitive endpoints, with survival and growth of older organisms being
significantly less sensitive. Other aquatic life (aquatic plants, invertebrates and
amphibians) have been shown to be much more tolerant to TCDD than fish and thus
would not be an endpoint of concern for this risk assessment. Ecological effects of
greatest concern will be the survival of fish fry and the reproductive success of
piscivorous wildlife. Doses of concern to piscivorous wildlife are particularly low
because of biomagnification of TCDD, although they will not necessarily be feeding
exclusively on the most contaminated fish in the reservoir.
The dose/response curves for TCDD and fish are so steep that there is likely
little difference in exposures between no effects on populations and very severe
effects. For both fish and wildlife the most sensitive and most heavily exposed
species appear to be at the top of aquatic food webs. Therefore, this assessment
could largely depend on information on effects to individuals. Because available
toxicity information on early life stage survival is limited to a few species, a major
uncertainty that must be considered is the variability among species and the
extrapolation of available toxicity information to species of interest.
For aquatic associated-wildlife, effects will be based on concentrations in
aquatic organisms in their diet. For some of the wildlife species, especially the
piscivorous mammals, some portion of the diet will be from terrestrial sources or from
aquatic animals in the uncontaminated tributaries. These feeding habits and
movements must be considered in relationship to the expected contamination of food
organisms in order to estimate expected doses. Dose/response relationships for
receptor wildlife species or surrogates can then be applied to assess expected effects
on individuals and then extrapolated as appropriate to expected effects on
populations. As for fish, extrapolations among species of different sensitivities is a
major uncertainty that must be addressed. x
Ecological Effects of Related Compounds. A major consideration of this
assessment is the joint behavior and toxic effects of TCDD and other planar
chlorinated aromatic organic chemicals. Comparative toxicity information is available
for the rainbow trout (see Walker et al., 1992) and for some mammals (e.g., see
DeVito et al., 1993; Safe, 1990). Based on their relative toxicities and concentrations
in paper mill effluent, other chemicals of significant concern are 1,2,3,7,8 PeCDD,
2,3,7,8-TCDF, 1,2,3,7,8-PeCDF, and 2,3,4,7,8-PeCDF. Relative concentrations in the
effluent will not be quantitatively repeated in residues in aquatic organisms due to
differences in chemical fate, transport, bioavailability and bioaccumulation.
A-7
-------
Toxicity equivalence factors (TEFs) appropriate for assessing the toxic potential
of complex mixtures of PCDDs and PCDFs in trout sac fry are available (Walker et al.,
1992). The assessment will have to address the fate and transport of these chemicals
and their expected accumulation in the food chain in addition to their toxic potential
relative to TCDD. The predicted safe tissue concentration of TCDD for each organism
is equal to the total TCDD residue toxicity equivalence concentration (RTEC) of
concern for the organism under the TCDD toxicity equivalence model (Safe, 1990)
which assumes that each chemical's dioxin-like toxicity is additive. Relating the RTEC
to concentrations of chemicals in the effluent is complicated by (1) the influence of
bioaccumulation and chemical fate and transport phenomena on the composition of
the chemical mixture and (2) the choice of appropriate TEFs. Note that there are
considerable uncertainties regarding the choice of TEFs for different endpoints and for
PCB congeners (DeVito et al., 1993). When the chemical composition of an effluent
can be predicted, as in this scenario, the fate/transport and bioaccumulation models
used for TCDD can be used to predict differences in the chemical mixture that
bioaccumulates in comparison to the mixture in the effluent. If a biota-sediment
accumulation factor (BSAF) approach is used, the RTEC can be related to sediment
contaminant mixtures:
RTEC = £. [(C^i (BSAF)i (f2) (TEF) J
where (Coc)| is the organic carbon-normalized concentration of the ith chemical in the
surface sediment and f, is the fraction lipid in the tissue represented by the BSAF.
The chemical fate and transport model chosen to relate concentrations of chemicals in
sediment and water to concentrations in the effluent can be used to determine
changes in chemical mixture composition between sediment and effluent. The final
expression of risk associated with concentrations of these chemicals in the effluent
incorporates all of the above factors in a weighted fashion to represent combined
effects. • :'
Assessment Endpoints. Assessment endpoints of concern to the involved risk
managers are the productivity of largemouth bass, catfish, crappie, and bluegill
populations which are sought by sport anglers and populations of avian and
mammalian wildlife along the shores of the reservoir. As stated above, invertebrate
and plant populations are of less concern because of their demonstrated tolerance to
TCDD in laboratory studies.
Measurement Endpoints. Measurement endpoints that are most relevant to
these assessment endpoints are the effects that TCDD has on reproductive success
(e.g., egg production and viability) and/or larval and offspring survival in laboratory
tests. Because of the uncertainties in establishing the bioavailability of TCDD in
aqueous solutions, measured TCDD concentrations in food or in the test organisms
themselves, as opposed to aqueous TCDD concentrations, are a more useful metric of
expressing and applying dose-response relationships.
A-8
-------
•- Although several studies have been conducted to show reproduction and/or
survival of early life stages are sensitive endpoints for TCDD toxicity, data are
available for only a small number of species. Consequently, there are uncertainties in
extrapolating measurement endpoints from tested species to the species of interest for
the assessment endpoints. As stated previously, there are also few toxicity data
available for these measurement endpoints with regard to other dibenzodioxins or
dibenzofurans.
A-9
-------
Conceptual Model for the Southern Reservoir J>J
...... ' • • •- e
The foundation for the conceptual riiodel is the tissue residue approach
contained in the interim TCDD report. Chemical residues in tissues of sensitive
aquatic organisms exposed to persistent, hydrophobic, lipophilic organic chemicals,
such as those predicted for the paper mill effluent, are the exposure metrics upon
which the estimation of the potential for adverse effects to the organism must be
based. In this case, Figure 2 shows the logical flow of assessment information when
thresholds for adverse ecological effects or fish and wildlife population protection goals
are to be related to safe chemical loadings to the ecosystem. This is a typical
conceptual model for applications of water quality criteria to establishment of effluent
permit conditions for single chemicals except that this model can be expanded to
consider multiple stressors and provides for consideration of populations of multiple
species and their interactions. Models for relating fish populations to chemical dose-
toxic response relationships (Barnthouse et al., 1987) are adaptable to the tissue
residue approach used for TCDD. The boxes in this conceptual model represent
assessment endpoints which are generally quantitative. The arrows between boxes
are specific types of models which are used to interrelate the assessment endpoints.
All the models are reversible, hence the two-way arrows. The conceptual model
applies equally to assessments which seek to determine risks associated with a
known or predicted chemical loading (right to left flow of steps).
Figure 2 shows effects on aquatic organisms linked to exposure levels through
chemical residue-based dosimetry. The same approach could be used for wildlife
assessment, however the toxicity data available for TCDD and related chemicals at
this time relate only to dietary dose. Since the known adverse effects of TCDD and
related chemicals for fish are directly attributable to exposure of the embryo, the
chemical residue levels in eggs is presently the exposure metric of- primary interest. If,
for example, male fertility should be determined to be a more sensitive endpoint,
chemical residue levels in the testes might become an important exposure
assessment endpoint. Care must be taken to insure that appropriate exposure and
bioaccumulation models are chosen for relating each aquatic species' tissue residue
level of concern to chemical concentrations in the water and sediment of the region
they inhabit.
Figure 3 illustrates the pathways for TCDD exposures and bioaccumulation in
Omigoshee Reservoir biota. TCDD exposure to fish and wildlife in natural systems is
expected to be primarily via contaminated food, and effects are often best referenced
to accumulation in food or in the receptor organism itself. Bioaccumulation in aquatic
organisms, and the distribution and bioavailability of TCDD in water and sediments,
will therefore be of central concern in this assessment. Based on predicted
concentrations of chemicals of concern,for aquatic organisms, concentrations in the
sediment, suspended solids, and water in various areas of the reservoir must be
estimated using suitable bioaccumulation models. This can be accomplished via a
A-10
-------
food chain model, such as that of Thomann et al. (1992) shown in Figure 4, or
application of bioaccumulation factors for fish which describe accumulation
relationships without explicitly considering trophic level transfers. The concentration of
chemical predicted for the whole organism can be related to specific tissue
concentrations through lipid normalization or a more specific toxicokinetic model.
Bioaccumulation factors between fish and water are discussed in Chapter 3 of the
interim report. The variability of these factors among different organisms and their
relationship to organic carbon in suspended solids and lipid in organisms are major
uncertainties that must be considered. A third bioaccumulation approach is to
estimate chemical concentrations of concern in the surface sediments of the
organism's habitat by application of measured or estimated biota-sediment
accumulation factors (BSAF; see section 3.5 of the interim TCDD report) to the tissue
residue-toxic response relationship for each species of concern. The BSAF approach
has an advantage of using an accumulation factor which can be directly measured in.
contaminated ecosystems.
A-ll
-------
i
*»
•So
£§•
s
,ss
• O
4*-H
an
O 0
*0
HIM
i
w
CO
A-12
-------
•
I
:
*
•
A-13
-------
I
!
o
I
H
O
-ri
a
I
n «
A-14
-------
References
Ambrose, R.B., T.A. Wood, J.P. Connolly, and R.W. Schanz. 1988. WASP4, a
hydrodynamic and water quality model-model user's manual, and programmer's
guide. Office of Research and Development, US EPA. EPA/600/3-87/039.
Barnthouse, L.W., G.W. Suter II, A.E. Rosen and J.J. Beauchamp. 1987. Estimating
responses of fish populations to toxic contaminants. Environ. Toxicol. Chem. 6:811-
824.
De Vito, M.J., W.E. Maier, J.J. Diliberto and L.S. Birnbaum. 1993. Comparative
ability of. various PCBs, PCDFs, and TCDD to induce cytochrome P450.1A1 and 1A2
activity following 4 weeks of treatment. Fundam. Appl. Toxicol. 20:125-130. '
Safe, S. 1990. Polychlorinated biphenyls (PCBs), dibenzo-p-dioxins (PCDDs),
dibenzofurans (PCDFs), and related compounds: Environmental and mechanistic
considerations which support the development of toxic equivalency factors (TEFs).
Crit. Rev. Toxicol. 21:51-58.
Thomann, R.V., J.P. Connolly and T.F. Parkerton. 1992. An equilibrium model of
organic chemical distribution in aquatic food chains. Environ. Sci. Technol. 23: 699-
707.
U.S. EPA. 1992a. Framework for ecological risk assessment. EPA/630/R-92/001.
Risk Assessment Forum, Washington, D.C.
U.S. EPA. 1992b. Estimating exposure to dioxin-like compounds. EPA/600/6-
88/005B. August 1992 draft. Office of Research and Development, Washington, D.C.
U.S. EPA. 1993. Interim report on data and methods for assessment of 2,3,7,8-
tetrachlorodibenzo-p-dioxin risks to aquatic life and associated wildlife. EPA/600/R-
93/055. Office of Research and Development, Washington, D.C.
Walker, M.K., L.C. Hufnagle, Jr., M.K. Clayton and R.E. Peterson. 1992. An egg
injection method for assessing early life stage mortality of polychlorinated dibenzo-p-
dioxins, dibenzofurans, and biphenyls in rainbow trout (Oncorhynchus mykiss). Aquat.
Toxicol. 22:15-38.
A-15
-------
-------
APPENDIX B
THREE WORKSHOP EXERCISES
(QUESTIONS TO PANELISTS)
B-l
-------
-------
THREE WORKSHOP EXERCISES
The objective of the Minneapolis workshop is to use the Interim Report to
evaluate and develop the conceptual model proposed in the attached scenario. As
defined in EPA's Framework Report, the conceptual model is developed during the
initial "problem formulation" stage of the risk assessment process. Problem
formulation, an initial planning and scoping activity, has several elements, including
stressor characterization, ecosystem characterization, and ecological effects and
endpoint selection, all leading to a conceptual model for the risk assessment.
The workshop agenda includes three group exercises, each designed to use
the attached materials to evaluate the suitability of the approaches described in the
Interim Report for conducting ecological risk assessments for aquatic life and wildlife.
In the first exercise, participants will review the suitability of available ecological effects
information for risk assessment and will discuss endpoint selection, and in the second
exercise, they will evaluate information for characterizing the stressor. in the final
exercise, participants will discuss the feasibility of the risk assessment approach
defined by the proposed conceptual model.
In each exercise, workshop participants are asked to develop approaches for
both "ideal" risk assessments (those with appropriate existing scientific information
and/or abundant resources -- data, time, funding, and expertise), and more "realistic"
risk assessments (where data and/or resources are minimal, but scientifically
acceptable). To focus each exercise, EPA has extracted some major findings and
approaches from the Interim Report and included them in boxes on the following
pages. In their pre-meeting comments, workshop participants are asked to comment
on the use of the Interim Report findings in the scenario as well as to propose any
changes or additional information that may be necessary for conducting risk
assessments.
B-3
-------
Exercise 1. Ecological Effects and Endooint Selection
Peer panel members are asked to use information in the Interim Report
as source material to address the following issues concerning ecological effects
and endpoint selection raised by the scenario. The Interim Report focuses on
the effects of TCDD on freshwater aquatic organisms and associated wildlife
(Interim Report, chapter 4). Available data on TCDD effects on fish are
provided in section 4.2.1 and are summarized in section 4.2.3 of the Interim
Report., while effects on aquatic-associated wildlife are provided in section
4.3.1 and summarized in section 4.3.3. Reviewers should consider using
approaches applicable to both "ideal" and "realistic" risk assessments using the
pulp mill scenario.
Issues for Consideration
1. The lack of Ah receptors in some species (Interim Report, section 4.1)
along with the results of a limited number of laboratory studies suggest
that amphibians, invertebrates, and plants are less sensitive to TCDD
than fish, birds and mammals. For fish, early life stages appear to be
most sensitive. Because of this range in sensitivity, productivities of fish
species were selected as assessment endpoints for the scenario.
Comment on the whether this focus on fish species will result in
adequate protection for the rest of the aquatic community in the reservoir
from the direct or indirect effects of TCDD.
2. Section 4.1 of the Interim Report (and the scenario) describe the use of
toxicity equivalency factors (TEFs) for TCDD-like compounds. Section
3.5 of the Interim Report discusses the use of TCDD biota-sediment
accumulation factors (BSAFs) for calculating bioaccumulation
equivalency factors for other related compounds. Comment on the use
of these approaches for evaluating the effects and bioaccumulation of
dibenzodioxins and dibenzofurans in the paper mill effluent.
3. Because of difficulties in extrapolating from various laboratory exposure
conditions to observed effects, the Interim Report (section 4.2.3.1)
emphasizes using tissue levels of TCDD (rather than exposure
concentrations) to evaluate effects. Comment on the applicability of this
approach to evaluating the risks of TCDD from the pulp mill effluent.
B-4
-------
Exercise 1. Ecological Effects and Endooint Selection (Continued)
Issues for Consideration
4. The Interim Report uses both laboratory and field information to predict
levels of TCDD in fish and wildlife tissues that will cause adverse effects.
The scenario proposes to use laboratory test data at the individual level
of organization to predict population changes in fish and wildlife.
Comment on the utility of available laboratory data to predict effects on
field populations and discuss the associated uncertainties..
5. The Interim Report sites data that indicate a very steep concentration-
response curve for TCDD effects in fish and wildlife. Discuss the
implications of this observation for evaluating ecological effects in the
scenario.
6. The general summary of effect levels for aquatic species and associated
wildlife (Boxes 1 and 2, section 4) is based on extrapolations from a
limited number of test species and from tests that do not span complete
reproductive cycles. Associated uncertainties are summarized in sect/on
5.1.3. Discuss the utility of these data and uncertainties for evaluating
ecological effects.
7. As discussed in the Interim Report, few data on the effects of TCDD on
estuarine and marine organisms have been reported (section 4.2.1.5),
and no data were found in the literature for TCDD effects on reptiles or
marine mammals. Although all the current wildlife toxicity data were
reviewed, an analysis to establish an effects profile for terrestrial
organisms was beyond the scope of the report. Describe other effects
data not identified in the Interim Report or the scenario that will be
important for future ecological risk assessments.
B-5
-------
Exercise 2. Stressor Characterization
Peer panel members are asked to use information in the Interim Report
as source material to address the following issues concerning stressor
characterization raised by the scenario. Available information on the TCDD
physico-chemical properties and exposure characteristics are described in
chapter 2 of the Interim Report, while bioaccumulation is described in chapter 3.
Reviewers should consider approached applicable to both "ideal" and "realistic"
risk assessments using the pulp mill scenario.
Issues for Consideration (Exposure)
8. The Interim Report (sections 2.1 and 2.2) indicates that there is
considerable uncertainty in the estimates of parameters including K^
and Kx and the partitioning of TCDD onto organic matter. This
uncertainty results in part from difficulties in analytical measurements of
various fractions of TCDD in water. Since these limitations may affect
predictions of TCDD partitioning and exposure, please address how they
should be handled in stressor characterization and conceptual model
development for this scenario.
9. The Interim Report (section 2.4) indicates that most TCDD exposures will
arise from food consumption and contact with sediments or suspended
solids, with the water pathway being less important. Address the
implications of this information relative to the exposure routes in the
conceptual model.
10. Fate and transport models are beyond the scope of the Interim Report,
but are clearly critical for risk assessment. In stressor characterization
and the conceptual model, they will be necessary for linking TCDD
source loads to concentrations in different compartments of the reservoir.
a. Comment on the availability of fate and transport data/models
suitable for use with TCDD.
b. Discuss the applicability of available transport models for
predicting the deposition of particulate-bound TCDD in the
reservoir.
11. List some of the major exposure issues not present in the paper mill
scenario that may be encountered in future ecological risk assessments
(e.g., marine/estuarine, terrestrial, etc.).
B-6
-------
Exercise 2. Stressor Characterization (Continued)
Issues for Consideration (Bioaccumulation)
12. The Interim Report (sections 3.2-3,5) summarizes available data on
TCDD bioconcentration, bioaccumulation, biomagnification, and biota-
sediment accumulation factors from laboratory experiments and field
measurements. Discuss the applicability of these factors to stressor
characterization for the paper mill scenario.
13. The Lake Ontario BAP*, may be useful as a predictor of residue, levels in
other systems if 0* can be estimated accurately (Interim Report, section
3.3), Comment on the applicability of this BAF for the paper mill
stressor characterization.
14. The Interim Report (section 3.4) indicates that biomagnification is
significant between fish and fish-eating birds but not between fish and
their food. Comment on the biomagnification pathway relative to
stressor characterization and the conceptual model.
15. Uncertainties associated with bioaccumulation factors are discussed in
section 5.1.2 of the Interim Report. Discuss the relevance of these
uncertainties to the prediction of TCDD residues in Omigoshee Reservoir
biota.
B-7
-------
Exercise 3. Conceptual Model
' Peer panel members are asked to use information in the Interim Report
as source material to address the following issues concerning conceptual
model development raised by the scenario. Different approaches to
bioaccumulation are indicated in thetS&§@afi&:'aMd are discussed in Chapter 3
of the Interim Report. In particular, BSAFs are discussed in se'ction 3.5.
Reviewers should considerlMti^fipffiaWe's^ "ideal" and
"realistic" risk assessments using the pulp mill scenario.
Issues for Consideration
16. Consistent with the Interim Report, the conceptual model focuses on
effects on fish and wildlife that consume fish. Comment on the whether
this approach captures the full range of potential ecological effects for
this scenario.
17. The Interim Report emphasizes using tissue residue levels to estimate
the adverse effects of TCDD. However, to do the risk assessment
outlined by the conceptual model, it will be necessary to link predicted
loadings of TCDD in the paper mill effluent to residues in the organisms
identified in the assessment endpoints. Discuss the utility of available
risk assessment tools for accomplishing this goal.
18. The Interim Report describes the limited field data that are available for
estimating BAF's and BSAF's. Discuss the applicability of these factors
to the Omigoshee Reservoir, conceptual model.
19. The temporal dynamics and disequilibrium situations commonly
associated with TCDD are mentioned in the Interim Report (section 2.3).
Comment on how these aspects should be considered in establishing (1)
the time course for the build-up of TCDD levels following initiation of the
paper mill discharge and (2) the time course for the decrease of TCDD
levels and recovery of biota should the paper mill cease operation.
B-8,;
-------
PF^ ;-^ffTW^V^ $$
-------
-------
United States
Environmental Protection Agency
Workshop on
Ecological Risk Assessment Issues for
2,3,7,8,-Tetrachlorodibenzo-p-Dioxin
Premeeting Comments
Minneapolis, MN
September 14-15,1993
C-3
-------
-------
Table of Contents
SECTION 1 C-7
General Comments
Peter Chapman C-9
Wayne Landis C-13
Thomas O'Connor C-17
SECTION 2 C-23
Exercise 1: Ecological Effects and Endpoint Selection
Randall Wentsel (Workgroup Leader) C-25
Nigel Blakley C-29
Peter Chapman C-33
G. Michael DeGraeve C-43
Wayne Landis C-49
Richard Peterson C-59
Thomas Sibley C-71
John Stegeman C-75
Bill Williams C-81
SECTION 3 C-91
Exercise 2: Stressor Characterization
William Adams (Workgroup Leader) C-93
Keith Cooper C-95
Joseph DePinto C-105
Robert Huggett C-113
Charles Menzie C-117
Derek Muir C-125
Thomas O'Connor C-133
Robert Pastorok C-137
C-5
-------
Table of Contents (continued)
Page
SECTION 4 C-147
Exercise 3: Conceptual Model Development
Charles Menzie (Workgroup Leader) C-149
Peter Chapman C-155
G. Michael DeGraeve C-161
Wayne Landis C-165
Richard Peterson C-177
Thomas Sibley C-181
BiU Williams C-185
Robert Huggett (Workgroup Leader) C-191
Keith Cooper C-195
Joseph DePinto C-201
Derek Muir C-205
Robert Pastorok C-211
SECnON 5 C-221
J.P. Giesy C-223
C-6
-------
Section 1
GENERAL COMMENTS
C-7
-------
-------
Peter Chapman
EVS Environmental Consultants
C-9
-------
-------
PKEMEEUNG COMMENTS:
WORKSHOP ON ECOLOGICAL RISK ASSESSMENT ISSUES FOR
2,3,7,8-Tetrachlorodibenxo-p-dioxin (TCDD)
Prepared By:
Peter M. Chapman
EVS Environment Consultants
195 Pemberton Avenue
North Vancouver, B.C.
Canada V7P 2R4
General Comments
The word "dioxin" is loaded with perception, which will inevitably complicate the scientific search for
facts. This needs to be acknowledged up-front and a commitment made to conduct studies according
to rigorous scientific methodologies, without regard to perceptions or fears associated with the word
"dioxin".
In this regard, it is important to note that there is no record of human death or even life-threatening
illness as a result of TCDD exposure. The only recorded effect, even with volunteer prisoners
exposed to unrealisticalfy high concentrations, or. the Seveso accident (when cows, horses, rabbits,
sheep and chickens died but humans did not despite identical exposures), is chloracne. TCDD is
extraordinarily toxic to some animals, but the label "the most toxic substance known" only applies to
guinea pigs. For instance, while TCDD has an LD50 of 0.6 ug/Kg to guinea pigs, hamsters are over
three orders of magnitude less sensitive, with an LD50 of 3,000 ug/Kg. Clearly there is significant
inter-species variability in toxicity, which is documented for aquatic species in the report.
To be scientifically credible, all studies of TCDD effects must be screened to eliminate those which
do not control for confounding factors, those which include multiple comparisons which produce
positive associations by chance alone, and those which do not establish clear exposure routes. This
is clearly the "ideal" situation. More "realistic" situations may, depending on how far they are from
the "ideal", not be scientifically credible.
C-ll
-------
-------
Wayne Landis
Institute of Environmental
Toxicology and Chemistry
Huxley College of Environmental Studies
C-13
-------
-------
Landis Comments September 1,1993
1
Workshop on the Ecological Risk Assessment Issue for 2,3,7,8-Tetrachlorodibenzo-p-dioxin
(TCDD) Radisson Hotel Metrodome, Minneapolis, MM. September 14-15,1993
Premeeting Comments-Issues for Consideration
Wayne G. Landis
General Comments
The Interim Report is a well written and detailed document reflecting the state of the
current Ecological Risk Assessment working paradigm. It reflects the single species approach,
much like what occurs in human health risk assessment, to a multispecies problem. Dynamics at
the population level or the interactions among the various organisms or guilds of the affected
communities are difficult to extrapolate from the data presented in this report. This fact is not the
fault of the report but perhaps a fault of the lack of a systematic program to ascertain the ecological
effects of TCDD.
I would have liked to seen a better description of the molecular biology and evolutionary
derivation of the Ah receptor. Given a detailed description and some comparative molecular
biology, some of the questions regarding potential senstivities and additive effects would have
been based on a better and more fundamental understanding of TCDD intoxication.
One approach to help facilitate the incorporation of data relevant to the evaluation of
ecological datasets is presented below. Based on the factors determining competitive outcomes
as described by Tilman (Tilman 1982, Landis 1986), the diagram attempts to highlight the factors
that a toxicant impacts in an ecological framework.
On the physiological side, I have found the graphical representation of the physiological
effects of a toxicant produced by the Wildlife Toxicology Group at USEPA Corvallis as a useful
reference framework (Included as an enclosure). Immunological suppression, reproduction and
other functions are delineated.
Another crucial factor that I would like to emphasize is the importance of the behavior of
an organism and the impacts of TCDD intoxication on predator avoidance, mate selection, and
foraging behavior. The outcomes of these alterations in behavior can have dramatic impacts on
population dynamics and extinction rates. Organisms that cease to eat or that show lack of
predatory avoidance are often dead in short order. Changes in behavior may also have dramatic
impacts on survivorship of populations at risk.
Finally, and of course I am biased in this direction, the lack of multispecies toxicity tests
(microcosms, mesocosms, microecosystems, eco-cores) is worrisome. Many of the questions
points below may have been answered by the investigation of specific relationships using these
types of tests. Admittedly, data analysis and extrapolation are difficult, but any more difficult than
C-15
-------
Landis Comments September 1,1993
2
extrapolating from clean water, laboratory cultured, limited time span single species toxicity tests?
As an addendum I have included references and summaries of a variety of multispecies toxicity
tests that may have proven useful in the evaluation of TCDD.
Framework for Evaluating Ecological Effects
Among Interacting Organisms
Competitive Outcomes
and other Species Interactions
Equilibrium Region
ZNGIA,ZNGlB (Zer° net growth
isoclines)
Mortality
Non-specific
mortality
(density independent)
(Resource
, Cg Consumption
Vectors)
Resource Region
Spatial
Variability
Biotic Components of
Resource Base
Temporal
Variability
\
Predation
(Including di
and parasitism)
Historical
Impacts and
Alterations
Toxicant
Figure 1. Resource Competition Theory as a Guide for Analysis of Populations and Ecosystems
C-16
-------
Thomas O'Connor
National Oceanic and Atmospheric Administration
C-17
-------
-------
Comments on "Interim Report on Data and Methods for Assessment of 2,3,7,8-
Tetrachlorodibenzo-p-dioxin risks to Aquatic Life and Associated Wildlife"
Thomas P. O'Connor NOAA N/ORCA21
I wrote these general comments before I realized that I was supposed to respond
to a prescribed scenario. I did not envision the challenge of predicting the
consequences of a future discharge of TCDD to a reservoir. Since some of what I
was specifically asked to address is covered in these comments I will accept the
embarrassment of leaving them in. In effect, I have called my own bluff.
GENERAL COMMENTS
The summary of the Interim Report is Table 5.1. The first column contains TCDD
concentrations in fish that are proposed to constitute risks to fish, terrestrial
mammals and birds. As intended, and I think correctly, this column is central to
the risk assessment. The other columns are calculated concentrations of TCDD in
sediment or water that theoretically correspond to the fish concentrations. I will
express some skepticism on this last point but, even if the calculations could be
done, the risk assessment would still have to be based on TCDD concentrations in
fish.
The calculated TCDD partitioning among organisms, sediment, and water are
based on equilibrium assumptions that are useful in emphasizing the importance of
lipid in organisms, organic carbon on sediment, and the octanol-water coefficient of
TCDD. However, the calculations are only rough approximations. The report is
replete with reasons for calculated partitioning not to conform with field data for
dioxin. For other compounds for which there are much more field data, Bierman
(1990) for example showed how much of a discrepancy exists between measured
and predicted ratios of organic contaminants in sediments and in organisms.
The approximate nature of equilibrium assumptions appears in the EPA proposed
Sediment Quality Criteria for neutral organics. Those criteria are based on the
equilibrium assumption and calculations of interstitial water concentrations. The
1991 version of the "Proposed Technical Basis for Establishing Sediment Quality
Criteria for Nonionic Organic Chemicals Using Equilibrium Partitioning" specifically
points out that it is not technically justified to extend the calculations to body-
burdens in fish. It is unjustified because of the long sequence of equilibrium
assumptions needed to predict body-burdens in fish from concentrations on
sediment.
If TCDD were a risk to the health of invertebrates there would be some incentive for
extrapolating from sediment concentrations to body-burdens in benthic organisms.
However, as the report explains in detail, invertebrates are not endowed with the
Ah receptor necessary to initiate the sequence of reactions through which TCDD
can affect health and reproduction, the concern is for birds, mammals and fish.
C-19
-------
For birds and mammals, fish are the exposure-route, so TCDD concentrations in
water and sediment are only indirectly relevant. For fish, of course, they are
directly relevant but, it appears, that effects can be related to tissue concentrations
instead of to concentrations in abiotic compartments..
To some extent that simplifies the risk assessment because it becomes one of
assessing effects as a function of TCDD body burden in fish. It also provides
insight to the spatial extent of possible risks because there are data plotted in this
report as Figure 5.1 and in EPA (1992) on the distribution of TCDD concentrations
in fish at 388 sites. Most of those site (314) were chosen to maximize chances of
finding high levels of TCDD, but the remaining 74 are probably typical of most
locations in the country. Concentrations in fish as high as 6 pg/g were rare, and
were found near the discharges of paper mills using chlorine and at some industrial
charges. Since 6 pg/g is the second lowest body-burden in Table 5.1, one would
conclude that TCDD poses a threat to fish or to birds only in isolated "hot spots".
While this does not argue against remediation, one could certainly question
whether or not the spatial scales of such areas constitute ecosystems.
On the other hand, the lowest body-burden in Table 5.1 is 0.7 pg/g, a
concentration found even at some of the 74 background sites. Since that fairly
commonly found body-burden poses an hypothesized threat to reproduction
among mink, one would expect feral mink populations to be on the decline. I have
no idea whether that is happening but it seems worth a look. Similarly,
experiments of the reproductive consequences on mink of ingesting 0.1 pg of
TCDD per day seem to be in order.
1 have obviously not considered this risk assessment in an absolutely objective
sense. If the lowest TCDD concentration in fish thought to pose a risk to any
organism is one that is rarely encountered, the assessment becomes very much an
academic exercise. To use round numbers, the first question is whether or not
TCDD body-burdens of 1 pg/g pose a threat to aquatic organisms or their
predators. If there is threat does it constitute an ecological hazard?
I have nothing to add to the Interim Report that would cause me to consider such a
concentration to be a threat to fish or birds. The interesting question concerns
selecting mink reproduction as the critical ecological endpoint. I have no doubt
that reproductive losses are an important endpoint. But I do question selecting the
single most sensitive species. Three factors caused mink to be chosen 1) they
are piscivorous, 2) the lethal TCDD dose to them is slightly less than that to guinea
pigs but ten or more times less than it is for rats, rabbits, mice, or hamsters and 3)
the level of TCDD ingestion shown not to affect Rhesus monkey reproduction was
0.13 pg/g/day and about ten times less than the corresponding value for Sprague-
Dawley rats, it was therefore Hypothesized that the TCDD ingestion rate that would
not cause reproductive damage to minks would be 0.1 pg/day or less.
C-20
-------
This is a testable hypothesis but, assuming that fish body-burdens greater than 1
pg/g are a hazard to mink, are they ah ecological hazard? Here the analysis gets
really sticky so I started to reread the stuff sent to me. Then I found the scenario.
C-21
-------
-------
Section 2
EXERCISE 1
Ecological Effects
and Endpoint Selection
Workgroup Leaden
Randall Wentsel
U.S. Army
C-23
-------
-------
PREMEETING COMMENTS
RANDALL S. WENTSEL, Ph.D.
INTERIM REPORT ON DATA AND METHODS FOR ASSESSMENT
OF 2,3,7,8-TETRACHLORODIBENZO-p-DIOXIN
RISKS TO AQUATIC LIFE AND ASSOCIATED WILDLIFE
I thought the Interim Report was well written and presented
the data in a professional manner. The authors also did a
good job in identifying data gaps and discussing their
significance.
1. a. Realistic
From the data available it appears that the focus on
sensitive fish species will provide adequate protection for
the aquatic community. It has been common in setting water
quality criteria to protect the most sensitive species and
through that protection infer protection of the aquatic
community. Protection of the lake trout population is an
important assessment endpoint in Great Lakes systems. The
bioaccumulation of TCDD in tissues to produce a reduction in
fecundity is an important pathway in the assessment of TCDD
effects in aquatic systems.
Relating protection of fish species to the structure and
function of an aquatic system is somewhat difficult. The
reduction of a key fish species at the top of the food web
will alter the structure of an aquatic system. The issue of
the replacement of a sensitive species by a tolerant species
may result in no functional change in the aquatic system.
However, the loss of a sport fish population would reduce
the value to society of the aquatic system.
fc>- Ideal
In an ideal situation the assessment endpoints would be
at the community and ecosystem levels. The various direct
and indirect pathways of TCDD to effect biota would be
addressed.
2. Because data indicate that planar PCBs and other
compounds act on the same site as TCDD, the use of toxic
equivalents to reflect their additive effects is logical.
Data on bioavailability of these compounds should be
considered when toxic equivalents are calculated. The BSAF
technique is a new application which requires additional
validation before acceptance as a method. However, a
technique is needed to relate sediment TCDD concentrations
to toxic tissue concentrations and the BSAF can perform that
function.
C-25
-------
3. The major accumulation areas, of,TCDD in aquatic systems
are in tissue and sediment. It makes sense to focus on
those areas were measurable amounts of TCDD occur. Because
of its hydrophobia properties TCDD binds with organic
particles suspended in the water column or organic matter in
the sediment. Measuring exposure concentrations for water
are very difficult due to the very low detection limits
required and the dynamic movement of TCDD out of the water
column. The use of models would be difficult to validate
with the data and techniques available. The use of tissue
data to establish no effect, levels is technically valid due
to the relationship of tissue levels to reduced fecundity.
Relating the sediment and tissue concentrations to the
degree of risk for a given system will be necessary.
4. The use of laboratory data to predict effects in the
field has been criticized for over and under estimating the
impact to biota. In an ideal situation using laboratory
data with field validation would be a preferred approach.
For TCDD, the laboratory data identified a sensitive
development phase and generated precise data on tissue
concentration and effects on reproduction. A field study,
under realistic conditions, could not have done this. The
uncertainties will be associated with the bioavailability of
the chemical in the laboratory compared to its
bioavailability in the field. Useful information would
include the distribution of TCDD in fish populations; with
the percentage of the population predicted to be effected at
given TCDD tissue concentrations.
5. Steep concentration response curves indicate a rapid ,
change in effects versus small changes in exposure
concentration. This situation necessitates increased
protection factors for the aquatic system. For the scenario
this would require increase precision in field sampling and
monitoring efforts and/or additional protection factors.
6. Early life stage tests have been shown to produce
results similar to complete reproduction cycle studies. The
authors point out most of the weaknesses of the current
data. Some discussion of the distribution of TCDD in tissue
versus fry survival should be emphasized. Probability
calculations to assess the degree of impact on a fish
population are needed. Uncertainty is addressed in the
report. Toxicity reference values could be used to address
uncertainty when applying the data to other species.
However, I would caution the user in applying uncertainty
factors so that the result of the process is not a "safe"
concentration below background levels of TCDD in the
environment.
7. The data indicate that TCDD is toxic to wildlife at low
levels. This will drive any future ecological risk
assessments. Additional research is needed to insure the
C-26
-------
accuracy of the data and to validate effects in the field.
For the models, parameters shoujd be measured, if possible,
to reduce variability in the estimates. Endangered species
also drive ecological risk assessments. While Bald Eagles
are mentioned in the scenario, no special treatment of then
is recommended. Endangered or threatened species receive
special treatment because the individual must be protected
versus protection of the population for other species.
Future ecological risk assessments will address
Superfund sites contaminated with TCDD. More effort needs
to be spent relatirg the ecological effects to an exposure
source that can be remediated. Research should be conducted
to establish sediment and water criteria. Increased effort
should be made to use biomarkers and relate increased or
decreased levels of them to significant TCDD exposure. This
research should be applicable to a variety of aquatic
systems. Additional data are required on the distribution
of TCDD in the tissue of fish populations. These data will
identify how large a problem elevated TCDD concentration are
to aquatic system. A tiered approach in conducting
ecological risk assessments .should be put forward.
Scientists will rarely, if ever, have all the data they want
to assess risk to ecological systems. Tiers would be
correlated to level of effort or level of concern. Tier 1
would be primarily a paper study using available date. Tier
2 would build on the result for Tier 1 and would address
critical data gaps and reduce uncertainty. Tier 3 would use
more complex methods and higher levels of effort to address
data gaps and quantitate risk. The use of tiers would
focus ecological risk assessments earlier in the process to
address key information that impacts important assessment
endpoints. It would also allow regulators and assessors to
have an agreed upon level of effort in conducting ecological
risk assessments and not result in never ending "financial
black hole" assessments.
16. Certainly higher level ecological effects at the
community, watershed or ecosystem level should be
considered. However, it is doubtful that they would be more
sensitive. Endangered and threatened species also are
important to consider because they must be protected at the
individual level.
17. The tools available are primarily models. However, the
use of several layers of models can produce results that are
not "real world". Models must be validated to field or
laboratory data. Building in conservative assumptions in
model parameters can result in useless results. The use of
distributions instead of single value parameters can reduce
this effect. When using models to support ecological risk
assessment I believe it is best to apply protection levels
(conservative estimates) at the end of the process and not
to build them into models.
C-27
-------
-------
Nigel Blakley
Washington State Department of Ecology
C-29
-------
-------
Comments on Preliminary Problem Formulation for a Dioxin Scenario - Nigel Blakley
1. Although it seems appropriate to expand the risk assessment from TCDD to TCDD-like
compounds using TEFs, I wonder whether the assessment should be further expanded to include
other chlorinated organics likely to occur in the effluent that may also effect fish reproduction in
combination with TCDD-like compounds. The Interim Report concludes (p. 4-35) that "...the
contribution of PCDDs and PCDFs in [chlorine bleached kraft pulp mill effluents] to observed
biochemical and physiological changes in exposed fish populations is unclear at this time", citing
(among others) Munkittrick et al. (1992), whose work suggests that other, possibly water soluble,
chlorinated organics in the effluent may also have adverse effects on fish reproduction.
Although the Problem Formulation enclosure states (p. 1): "Other stressors and endpoints are
subjects of other assessments and will not be considered here", it is not clear how additive or
synergistic effects from multiple stressors in the effluent on the measurement or assessment
endpoints will eventually be integrated.
2. Could TCDD sequestered in fat reserves begin circulating during periods when these reserves are
mobilized? If so, there may be other sensitive stages in the Hie cycle (e.g., during migration or
periods of food scarcity) which apparently have not been studied.
3. The assessment endpoint, "productivity" of various fish populations, seems somewhat vague and
open to differing interpretations. A definition of productivity will be useful if a quantitative
relationship between the measurement and assessment endpoint is needed. The conceptual model
figures could also be improved (e.g., Fig. 2: Community structure and function is not proposed as
an assessment endpoint in the scenario; Fig. 3: ingestion of sediment is not included as a pathway
for direct accumulation of TCDD for fish, mammals or piscivorous birds.).
C-31
-------
-------
Peter Chapman
EVS Environmental Consultants
C-33
-------
-------
EXERCISE 1. ECOLOGICAL EFFECTS AND ENDPOINT SELECTION
Issue 1. Is Focus on Fish Species Appropriate?
Fish, invertebrates and algae are typically used in toxicity testing, and are the focus of most historical
and present methods development studies. Toxicity tests and species currently in use in the United
States are reviewed by Adams (1993) and include the following for acute testing: salmonids (e.g.,
trout) and other fish, invertebrates (in particular daphnids, amphipods, midges, worms, mysids and
other shrimp, mayflies and bivalves), and plants (e.g., algal cultures and duckweed). Chronic toxicity
tests (Adams, 1993) include fish (e.g., minnows), daphnids, mysids, and algal cultures.
The use of fish species for testing is appropriate given their long history of usage. As regards
focussing on fish species, this is appropriate at this time given that this group appears to be the most
sensitive aquatic species to TCDD, and given the utility of "worst case" testing (or assessment) to
assure environmental protection. Given our present state of knowledge, focus on fish species should
result in adequate protection for the rest of the aquatic community in the reservoir from the direct
or indirect effects of TCDD for reasons discussed below. However, note that the simple fact of fish
having Ah receptors is not enough to say they are useful; the report notes that humans also have Ah
receptors, yet as noted above (General Comments), humans appear to be relatively insensitive to
TCDD.
"Worst case" testing is used to reduce uncertainty. It is impossible in science to "prove" that an event
might not occur, under the right combination of (sometimes improbable) circumstances. Bioassay
tests can never "prove" than an event might not occur in the real world, since they are by definition
removed from that environment However, if testing involves "worst case", e.g., most sensitive
organisms and life-stages, most adverse testing conditions, and most likely exposure routes, and there
are no effects, then it is reasonable to conclude that effects are unlikely. At the very least, such
studies indicate low priority for regulatory or other action, which is an important finding given that
there appear to be no lack of high priority issues presently requiring attention.
If, however, "worst case" testing indicates an effect, then there is a good likelihood that laboratory
testing is overestimating what might be occurring or what is occurring in the real environment, and
ideally providing early warning of potential problems before they become critical. Such early warning
cannot be provided by field observational studies which, due to natural background "noise" in the
distribution and abundance of natural populations, can only detect an effect after it has become severe
or even catastrophic.
C-35
-------
Issue 2. Are TEF and BSAF Approaches Appropriate?
This question has major policy implications given that (1) EPA has formally adopted the TEF
procedure for use in all programs (Federal Register, November 07,1989); (2) EPA did so as the lead
in a six-nation project under the auspices of NATO. Technically, I have no problem with assuming
additivity (this is a realistic "worst case", which is supported by various peer-reviewed studies, QSARs,
etc.). I also have no problem with the concept of TEFs. Where I do have problems is with
extrapolations that increase uncertainty. In other words, TEFs need to be directly related to what is
being measured. In this regard, and given known inter-species differences, extrapolation from
mammalian to non-mammalian systems is highly questionable and not presently supportable based
on the evidence I have seen. This is not to say it is wrong, just that we cannot be sure at this point
in time. Does this mean we should not use TEFs? No, because they are a useful tool. However, we
should remember that they are only a tool whose utility may (or may not be) confirmed based on
future research.
As regards BSAF approaches, these seem to be entirely appropriate, at this time, for assessing
bioaccumulation. But (see later comments), bioaccumulation is a phenomenon, not an effect. Thus,
these approaches should not be confused with determinations of effects.
IssueS. Should Tissue Levels be Used to Evaluate Effects?
A bioassay is an assay using a biological system. It involves exposing an organism to a test material
and determining a response. There are two major types of bioassays differentiated by response:
toxicity tests which measure an effect (e.g., acute, sublethal, chronic toxicity) and bioaccumulation
tests which measure a phenomenon (e.g., the uptake of contaminants into tissues). Bioaccumulation
can be from a variety of individual or combined routes, including respiration, ingestion, and direct
contact. The responses from one of these two types of bioassays should never be used to predict the
responses of another.
Bioaccumulation, which can result in bioconcentration, is a consequence of exposure, but cannot be
considered a true response, since there are no data to provide direct correlations between tissue
contaminant concentrations and adverse biological effects. Bioaccumulation is useful as an indicator
of exposure to contaminants. For instance, the U.S. EPA/ACOE (1993) consider such testing to be
an important evaluation of the potential of organisms to bioaccumulate contaminants of concern.
Evaluation of the results of such testing are made by statistically comparing test and reference areas.
In addition, bioaccumulation testing is conducted for human health reasons (are contaminants in
C-36
-------
tissues at levels of concern if humans eat these tissues?), and can be done using both organisms eaten
by humans and those which are not In the latter case, the data are evaluated in light of the
possibilities that certain contaminants can be transferred through food webs, and'that tests with one
species may indicate the potential for accumulation in other species.
Issue 4. How far can Laboratory Test Data be Extrapolated?
Pragmatic environmental monitoring is ultimately directed towards pollution, which is defined as the
presence of contaminants which result in an adverse biological effect Chemistry measures
contaminants, but provides no information on biological effects. Biological effects can only be
determined directly, which is generally easiest, quickest and cheapest using bioassay tests in the
laboratory. Such testing involves controlled experimentation, aimed at producing clear and
reproducible answers. In contrast, studies of resident communities tend to be observational and
cause-effect can never be directly determined, only inferred, by statistical and other methods which
are presently much disputed (e.g., see Dauer, 1993; Smith et al., 1993). Further, resident community
studies, as previously noted, cannot be predictive of effects, only record effects after they have become
severe or even catastrophic.
Proactivity can be achieved by "worst case" testing (i.e., most sensitive species, most sensitive life-
stages, most severe laboratory exposure conditions, likely most toxic and contaminated test samples),
realizing that this is not necessarily "real case". Reactive testing evaluates whether present conditions
can, or have the potential to, affect resident (or analogous) fauna.
For an approximation of "real case", appropriate species and end-points should be used (ideally test
what you are trying to protect, in particular key taxa related to beneficial environmental uses). If
appropriate individual species are protected (e.g., growth, reproduction, survival), it is inferred that
the structure and function of the ecosystem will also be protected. The appropriate combination of
laboratory and field data provides the best means presently available to assess whether effects will or
are occurring, and their potential environmental significance.
There are two basic philosophies involved in the choice of bioassay test organisms. One philosophy
is epitomized by the U.S. EPA (1986) who use standard species which they consider (after lengthy but
not exhaustive testing) to represent the sensitive range of resident species of all ecosystems analyzed.
Such "benchmark" species (U.S. EPA/ACOE, 1993) comprise a substantial data base, represent the
sensitive range of a variety of ecosystems, and provide comparative data on the relative sensitivity of
local test species. I (e.g., Chapman, 1991) and others (e.g., Cairns, 1993) consider that resident species
C-37
-------
data (using species which have not adapted to the contaminant stress being tested for) are more
directly relevant and avoid the major uncertainty of extrapolation between different species
sensitivities. However, this is an area of some debate in the general scientific community.
The inclusion of field data is necessary because laboratory test data only reflect the test conditions,
and conditions in the field can be very different due to, for instance, modifying factors. A modifying
factor is "any characteristic of an organism or the surrounding water (or sediment) which affects
toxiciry" (Sprague, 1985). Modifying factors can act either to increase or decrease the concentration
of a chemical required to produce a biological response, and their impact can vary dramatically
between classes of chemicals and the organisms which are exposed Modifying factors may affect the
distribution, fate, concentration, chemical nature, bioavailability or toxicity of contaminants. They are
divided into the following general, inclusive but not exclusive, abiotic and biotic groupings:
Abiotic Modifying Factors Biotic Modifying Factors
climate species/life stage
temperature sex/reproductive status
oceanography/limnology nutritional/disease status
environmental quality competition/predation
A biotic factor not included above, but which can greatly modify sensitivity in bioassays, is the
capability of organisms to adapt to toxic conditions. Such adaptation can take several forms, including
the alteration of reproductive strategies or the development of acclimation (enhanced tolerance or
resistance after first exposure). Tolerance or resistance are the ability of an organism to exhibit
decreased response to a chemical relative to the response shown on the first occasion. Tolerance
implies that the change is within the normal adaptive range of the organism and can be sustained
indefinitely. Resistance implies that the magnitude of the factor lies outside of the normal range and
that detrimental effects will eventually ensue.
Both tolerance and resistance may vary over an organism's life cycle and among organisms from
different populations depending on their history of exposure. They also vary with a large number of
biotic and abiotic modifying factors.
The advantages and disadvantages of the two types of bioassays, toxicity and bioaccumulation testing,
can be broadly summarized as follows:
C-38
-------
Type of Bioassav
Advantage
Disadvantage
Toxicity Test
holistic, measures
toxicity of all
stressors
does not indicate,
•without further testing,
which stressor(s) are causing
the observed effect(s)
can be relatively
simple and cost-
effective
simple tests can give
environmentally unrealistic
answers
indicates presence
or absence of effect(s)
on the organism(s)
used, by the exposure
tested
all possible organisms,
exposures and end-
points cannot be tested
conducted under
controlled laboratory
conditions
field conditions are much
different than the
laboratory
Bioaccumulation Test indicates whether
the orgamsm(s) used
can accumulate the
contaminants measured,
by the exposure tested
all possible organisms
and exposures cannot
be tested; some toxic
contaminants are
transformed and not
all can be or are
measured
provides quantitative
data (levels accumu-
lated), i.e., bioavai-
lability
measures a
phenomenon, not
an effect
Laboratory bioassay data for specific contaminants provide information on their toxicity and/or
potential to bioaccumulate, which can then be compared to likely organism exposures. Not all non-
C-39
-------
target species can be tested in the laboratory, and extrapolation between species is highly uncertain.
Similarly, prediction of effects under field conditions is not easy, since exposure is variable. However,
laboratory data allow uncertainties about contaminant effects in the "real" environment to be reduced
such that specific hypothesis can be formulated and tested, either through additional laboratory tests,
or field studies, or a combination of the two.
The determination and prioritization of real environmental problems requires effective science which
is relevant to and which will be realistically used in decision-making. In this regard, bioassays are an
essential part of two critical assessments:
1. an after-the-fact evaluation (are environmental conditions better or worse?
why or why not?) and,
2. a before-the-fact indication of whether environmental conditions may or will
change (will environmental conditions become better or worse? why or why not?).
Issue S. Implications of Steep Concentration-Response Curve
As noted by Paracelsus in the sixteenth century (Deichman et al., 1986) - "Only the dose determines
that a thing is not a poison". Although debate is continuing regarding the possibility of some
contaminants having effects at any level, this is not scientifically credible based on what we presently
know. The fact that there is a concentration-response curve for TCDD clearly indicates that we can
determine "safe" NOEL (no observed effect level) or, preferably, NEL (no effect level) concentrations.
The steepness of the curve could indicate that there is less of a gradation of effects for TCDD than
for other contaminants (in other words, it either does or does not have an effect), but may also be
a function of the lack of intermediate concentrations.
Issue 6. Data Utility and Uncertainties
As noted in the report, fish early life stages (in particular salmonids) appear to be the most sensitive
aquatic fauna to TCDD. As such they are, given our present knowledge base, appropriate to use in
"worst case" testing and assessment as detailed previously. It is entirely possible that more sensitive
species exist; it is also possible that no more sensitive species exist However, science must proceed
based on what is known, and based on what we presently know it appears to be entirety appropriate
to use information regarding effects on fish early life-stages for decision-making. Clearly this will
protect the majority of species, and possibly even all species.
C-40
-------
Issue 7. Other Effects Data/Important Scenariofs)
What is being asked here is not dear. If the first question is whether I am aware of other effects data
not included in the report, then the answer is "I am not". The report appears to be quite complete.
If the second question is asking what future information (e.g., research) would be useful, then clearly
information on estuarine and marine organisms, in particular marine mammals would be desirable.
However, direct testing of marine mammals, though scientifically desirable, is unlikely given animal-
rights concerns and activism, and other considerations.
C-41
-------
-------
G. Michael DeGraeve
Great Lakes Environmental Center
C-43
-------
-------
Prepared By: G. M. (Mick) DeGraeve
ECOLOGICAL EFFECTS AND ENDPOINT SELECTION
1. Based upon the currently-available data, focusing on the fish component of the aquatic
community seems reasonable. The available data strongly suggest that fish are the most
sensitive component of the aquatic community; therefore, if the fish are protected,
presumably the other elements of the aquatic community (amphibians, plants, and
invertebrates) will be adequately protected. This approach (using the available data for
those species which have been studied through exposures to TCDD) would be the one
most likely utilized for a "realistic" risk assessment, because such an assessment would
'^^*
be performed using existing information exclusively.
However, if an "ideal" risk assessment were to be performed, the individuals performing
the risk assessment would more than likely want to have TCDD lexicological information
for both the fish species to be protected (largemouth bass, catfish, crappie, and bluegill),
and the vertebrate and invertebrate species that are the forage organisms for the species
to be protected. Additionally, the risk manager may be interested in having TCDD
toxicological data for other species mat play an important role in the ecology of the
reservoir, but mat are not directly related to the food web of the fish species of interest.
2. A considerable amount of effort has been expended developing the TEF and BSAF
concepts. However, there are still uncertainties associated with the application of these
tools for evaluating the risks of TCDD-like compounds to aquatic life. However,
because exposure effects data are currently not available for the majority of these types
of compounds (particularly for the species to be protected in the reservoir), I feel it is
reasonable to use the TEF and BSAF approaches in performing a "realistic" risk
assessment for evaluating the effects and bioaccumulation of dibenzo-dioxins and
dibenzofurans in the hypothetical paper mill effluent.
Under more "ideal" conditions, I feel that the risk manager would want to have actual
data available for some (or all) of the species to be protected for those TCDD-like
compounds that will be present in the effluent. However, it is probably unrealistic to
think that exposure/effects data would be available for the fish species of interest for all
of the TCDD-like compounds to be released. Perhaps a reasonable compromise would
C-45
-------
be to perform the risk assessment using the TEF and BSAF approaches, and then to
validate these approaches by performing laboratory studies exposing 1 or 2 of the species
of concern to 1 or 2 of the compounds of concern.
3. Utilizing TCDD tissue concentration levels (as opposed to anticipated exposure
concentrations) to assess the risks associated with the discharge of the pulp mill effluent
seems like the most effective approach. In order to accomplish this objective, the risk
manager would need to utilize an accepted exposure model (like the one presented in
Figure 4 of the Scenario) to determine the acceptable effluent TCDD concentration, i.e.,
that concentration below which there would not be a detrimental accumulation of TCDD
in largemouth bass, crappie, catfish, and bluegill.
4. In situations where there are limited data to predict environmental consequences, the
level of uncertainty is fairly high. The case for TCDD is no exception. Because there
are TCDD effects data for relatively few individual species (and probably none of those
that could potentially be affected by the pulp mill effluent), extrapolating the existing
data to the pulp mill discharge scenario will result in uncertainties for evaluating the
ecological effects. These uncertainties could be somewhat minimized if the risk manager
could be sure that the existing data cover species that are more sensitive than (or at least
as sensitive as) the species to be protected. However, the relative sensitivities cannot be
determined for certain in this case, because we do not know if the species to be protected
are more or less sensitive than those for which there are TCDD data in the literature.
5. The very steep concentration-response curve for TCDD indicates to me the importance
of making very accurate predictions of the concentration of TCDD and TCDD-like
compounds in the effluent. A factor of two error could result in a population or
community being negatively affected, even though the risk assessment suggested that
there would be no impact. Because of this steep concentration-response curve, it would
probably be wise for the risk manager to apply a safety factor to the effluent discharge
limit that would be protective even if the limit were intermittently exceeded. This type
of situation (steep concentration-response curve) also emphasizes the importance of
accurate TCDD effluent measurements at a low level of detection once the plant is in
operation.
6. These data have limited utility for the pulp mill scenario because there is a limited
amount of data for only a few of the species to be protected in the reservoir. As a
C-46
-------
consequence, the associated degree of uncertainty is fairly high. The •wildlife data are
more uncertain than the aquatic life data, but in both cases a fairly substantial safety
factor would be appropriate to compensate for the high level of uncertainty.
7. To die best of my knowledge, there is no information available on the bacterial
component of the environment, which is very important in understanding the impact
TCDD may have on decomposition in both the aquatic and terrestrial communities.
Also, I am not aware of the existence of substantial data on the effects of TCDD on
primary producers in the aquatic or terrestrial environments. And among the various
fish and terrestrial species that have been evaluated, it is probably not safe to assume that
the limited existing data are representative of the more sensitive species of primary
producers or bacteria.
C-47
-------
-------
Wayne Landis
Institute of Environmental
Toxicology and Chemistry
Huxley College of Environmental Studies
C-49
-------
-------
Landfs Comments September 1, 1993
3
Comments on Workshop Exercises
Exercise 1. Ecological Effects and Endpoint Selection
1. The lack of Ah receptors in some species along with the results of a limited number of
laboratory studies suggest that amphibians, invertebrates and plants are less sensitive to TCDD
than fish birds and mammals. For fish, early life stages appear to be most sensitive. Because of
this range in sensitivity, productivities of fish species were selected as assessment endpoints for
the scenario. Comment on whether this focus on fish species will result in adequate protection for
the rest of the aquatic community in the reservoir from the direct or indirect effects of TCDD.
Comment
It is unlikely that a focus only on the fish species can provide an adequate protection for
the rest of the aquatic community or even the fish. Most of the Interim Report deals with only the
direct effects of TCDD upon the target organisms and little discussion of possible indirect effects
occurs. In seasonal systems slight differences in the timing of reproduction can have severe
repercussions. Several scenarios among many others are possible:
1) Chemicals tend to concentrate at the interfaces, especially organics and other iipophillic
materials. Concentrations higher than in the water column interfere with the pupation or
metamorphosis of the arthropods and other larval forms that important parts of the macrobenthic
assemblage. The subsequent lack of reproduction alters the detrital processing and
subsequently alters the resuspensfon of nutrients and algal production falls. Organisms higher in
the food chain subsequently see a population reduction.
2) Slight alterations in reproductive success can dramatically alter competitive relationships
and thereby change the structure of a community. Many of the sport fish described as the
species of concern tend to specialize on certain components of the invertebrate assemblage. An
alteration of this assemblage could alter the mix of fish, perhaps eliminating some species. As
certain fish are eliminated the invertebrate and producer assemblages are again altered leading to
another change of structure.
3) The myth of the most sensitive species (Caims, 1986). As in most cases, only relatively
few representatives are tested in the laboratory. Entire phyla are routinely omitted (I am as guilty
as the next) while a great deal of concentration is focused on vertebrates of concern in Phylum
Chordata. Given the lack of comparative data, extrapolation from a few invertebrates or algal
species seems unwarranted.
The evolutionary biology of the Ah receptor seems unclear and without an understanding
of the phytogeny of these molecule, generalizations as to sensitivity are difficult to be made. Is
C-51
-------
Landis Comments September 1,1993
4
the protein conserved, how many base pair changes are necessary to increase or decrease
sensitivity, are the Ah receptors a family of proteins, can simple cDNA probes be synthesized for
comparative purposes? Given the planar nature of the compound and the reactive halogen atoms
I would suspect that there may be separate genotoxic (carcinogenic) and physiologic receptors.
4) Focusing on fish, particularly mobile sport fish, may actually underestimate the impact
unless a clear understanding of the role that a particular location plays in the metapopulation
dynamics of the species. Given that most fish produce an excess of fry, migration from population
sources to sinks, sinks due to the elimination of the indigenous population, is highly likely. Two
misleading measurements may occur: 1) body burdens below levels generally assumed to be of
concern can be found in the tissues since the organisms are newcomers to the contaminated
area, and 2); populations may seem to be plentiful due to migration from populations sources. As
the habitat becomes more intolerable or the population source is eliminated through TCDD levels
or habitat alteration, a drastic reduction in resident fish populations may appear and the manager
may took for a proximate cause, however, the populations had been herded to the cliff for a
number of years. Unless a mark recapture type of program is initiated, an accurate estimate of the
population dynamics of the fish would be impossible, and likewise the true impact of the TCDD.
Likewise, organisms that are unable to range as widely as fish can be severely impacted.
Macrophytes, mollusks, protists, and carnivorous invertebrates may subtly affected due to the
continuos exposure to TCDD. Fish are not the only carnivores in a freshwater system. Mollusks
are famous for their ability to accumulate organics to very high levels (Hugget, ASTM Symposium
1993).
5) The important variables that determine the status of an ecosystem also apparently
change over time and circumstance. In the microcosm tests conducted at Western, we have
consistently found that the variables important in identifying the treatment groups change over
the time course of the experiment. This variability has been found for several jet fuel experiments
and with several types of microcosm protocols (Landis et al, in press). Not only do the variables
change in their ability to predict treatment, they change in a bounded stochastic manner. In other
words, while the variable "Phitodina" will likely be important during the latter stages of the
experiment, it is difficult to pick the sampling day or relative ranking of the variable. In field studies
Matthew's et al (1991 a, 1991b) have found similar patterns, in which variables change in
importance in the clustering of field results. A recent study by Dickson et al (1992) also tends to
confirm this trend. While measured impacts on aquatic ecosystems were correlated with
laboratory toxteity tests, the type of impact in the field was not predicted by the laboratory
experiment. In fact, several types of impacts were correlated with the same basic suite of toxicity
tests.
C-52
-------
Landis Comments September 1,1993
5
In summary, it is unlikely that a magic endpoint or measurement can adequately protect an
ecological system, or be used as a vital diagnostic measurement as, lets say, cardiac enzymes in
the bloodstream are indicative or heart attack. Search for such a measurement is probably a
remnant of the ecosystem as super organism and ecosystem health as more than metaphor (see
Suter 1993). Ecosystems are not organisms and do not have health in the sense that an organism
does. It may prove difficult to find one variable that denotes a specific type of impact.
2. Section 4.1 of the Interim Report describe the use of toxicity equivalency factors (TEFs) for
TCDD compounds. Section 3.5 of the Interim Report discusses the use of TCDD biota-sediment
accumulation factors for other related compounds. Comment on the use of these approaches for
evaluating the effects and bioaccumulation of dibenzodioxins and dibenzofurans in the paper mill
effluent.
Comment
The TEFs are a problem until the exact nature and evolutionary history of the Ah receptor
is understood. In essence, it seems that these compounds, that are reactive materials in some
instances, are being treated as if they are narcotics. That is, the modes of action are similar
enough that on a molar basis the materials are biologically equivalent. Are there a variety of Ah
receptors, each slightly different so that different genes are transcribed to mRNA? Hormones and
othertypes of initiator-receptor molecules tend to be very specific and I would not be surprised to
find a family of Ah receptors.
The field data seem to come primarily from Lake Ontario, a quite atypical lake compared to
a southern pond or a northwest glacial lake. How far can these data be extracted to these very
different systems?
The lack of data by which to judge the efficacy of the BSAF methodology is troublesome.
The potential to determine an equilibrium concentration appears to be there, although further
validation is warranted. My major concern is the lack of data on rates of accumulation and the
dynamics and heterogeneity of TCDD uptake. Equilibrium states may rarely be reached by an
entire population and if it is the ecosystem is likely to be heavily contaminated. Given the
importance of population dynamics and interactions in determining the structure of an aquatic
community, information on the dynamics of the toxicant are vital. F. B. Taub (SETAC 92
presentation and personal communication) has demonstrated in a model of an aquatic microcosm
the importance of timing and the subsequent outcome. Differences of a few days in the onset of
mortality produce dramatic differences in outcomes for the algal and daphnid populations. I
suspect that natural systems have even a larger variety of dynamics.
C-53
-------
Landis Comments September*!, 1993
6
3. Because of difficulties in extrapolating from various laboratory exposure conditions to
observed effects, the Interim Report (section 4.2.3.1) emphasizes using tissue level of TCDD
rather than exposure concentrations to evaluate effects. Comment of the applicability of this
approach to evaluating the risks of TCDD from pulp mill effluent.
Comment
Tissue levels are generally better indicators of exposure information than concentration.
Compounds like TCDD and halogenated aromatics require a great deal of time to move into
tissues. However, tissue concentrations in adult organisms would not as good as indicator of
population level impacts as concentrations in yolk and fry.
While tissue concentrations are good indicators of exposure they are more an indication
of total exposure for an organism that has survived in that environment. Hotspots and
concentrations in crucial habitats are not effectively measured. Organisms that have been
exposed to these areas, given the steep dose response curve of TCDD, likely do not survive.
It should not be surprising that the extrapolation of laboratory results to observed effects
should be so difficult with a compound with an estimated Log P of 7.0 Such a material will likely
migrate to interfaces containing high concentrations of organics, sediment, sediment-water
column interface, the surface mfcrolayer (Hardy et al 1982) and the awfuchs. These environments
are difficult to accurately sample, yet they are important to many sensitive life stages and within
reside highly productive aspects of the ecosystem. The chemistry at these interfaces is very
dynamic and temporally and spatially heterogeneous.
Aquatic systems rarely seem to approached in the same manner as terrestrials systems,
that is seeing the lake or stream in the sense of a landscape. The landscapes in an aquatic sense
would incorporate the spatial heterogeneity of the systems. These heterogeneities are crucial in
maintaining the species richness of the planktonfc community as hypothesized by Hutchinson
(1961), Rtoherson et al (1970) and Tilman (1982). Given the biotic importance of the physical
heterogeneity of aquatic systems, the nature of the distribution of toxicant must be equal V
important for estimating ecosystem level effects. Tissue concentrations of fish are generally
moving averages of exposure, tissue burdens of sessile organism coupled with analytical
measurements from interfaces would provide a more accurate picture of the exposure landscape.
4. The Interim Report uses both laboratory and field information to predict levels of TCDD in fish
and wildlife tissues that will cause adverse effects. The scenario proposed to use laboratory test
data at the individual level of organization to predict population changes in fish and wildlife.
Comment on the utility of available laboratory data to predict effects on field populations and
discuss the associated uncertainties.
Comment
C-54
-------
Landis Comments September 1,1993
7
Several books could be written on this topic (and have been). Using data at the individual
level to predict population level effects has been done, but often with a probability of extinction as
the measurement, a fairly dramatic outcome. Some of the uncertainties are:
1) Age Structure. Age structure of the recipient population can have a dramatic influence
upon the subsequent dynamics. Populations may undergo dramatic cycles due solely to the
influence of the initial age structure.
2) Genetic heterogeneity. A much greater diversity of genomes would be expected in a
wild population, although many populations under bottlenecks that make local populations
genetically distinct. The degree of diversity within a population, compared to that population
used in the laboratory, is almost never measured.
3) Intrinsic Dynamics of Population Growth. Populations with high rates of reproduction
can exhibit very different dynamics depending on the initial conditions. May and Oster (1978)
discussed this a some length. Simple non-linear equations (such as Nn+i=N(l +r(l-N/K) can
exhibit dynamics ranging from an asymptotic equilibrium to chaotic (in the mathematical sense)
given appropriate r values. Initial conditions also dramatically alter the outcomes, in the graph
below (Figure 2), the differences in the initial population sizes is only 2, yet when a stress occurs
one population becomes extinct while the other oscillates around a new carrying capacity. Unless
the initial conditions are precisely known, widely varying yet perfectly deterministic outcomes can
occur. These types of dynamics can also become apparent in the spatial distributions of
organisms, with the spatial distributions ranging from regular to contagious (Hassel et al 1991)
5) Stochastic influences. Along with the difficulties of non linear systems, certain aspects
of population biology are stochastic. Chance events do occur, disease is largely unpredictable,
along with storms and other events that have dramatic impacts.
6) Harvesting. Finally, the impacts of harvesting of game fish or commercially important
species is always an important consideration. Even if only certain age-classes are chosen, a
skewing of the age-classes of the population occurs. These dynamics, although the basis of
catch quotas, have proven difficult if not impossible to comprehend so that accurate predictions
can be made.
C-55
-------
Lands Comments September 1,1993
Switch of K from 10,000 to 8,000
Effects of Starting ConcAions on Population
Dynamics'
lowered tt> 8,000 ""
Figure 2, Different Outcomes from slightly different initial conditions, R i$ equal to 2,0, K is initially
10,000 and that is lowered at time 50 to 8,000. One population becomes extinct as the other
cycles around 8,000,
5. The Interim Report sites data that indicate a very steep concentration-response curve for TGDD
effects in fish and wildlife. Discuss the implications of this observation for evaluating ecological
effects in the scenario,
Comment
The steepness of the dose-response curve is a problem. It appears to virtually a step with
no effect to 100 percent effect Essentially, indications at the individual level or even the
population level may be too subtle to give any warning before impending doom. Localized
extinctions or effects could also occur without warning, altering the landscape of that community a
potential cascade of effects.
The advantage to a steep dose-response curve in that intrapopulattonal variance in regard
to the toxicant should be relatively low. This should reduce the uncertainty associated with the
impact of TCDD,
6, The general summary of effect levels for aquatic species and associated wildlife (Boxes 1 and
2, section 4) is based on extrapolations from a limited number of test species and from tests that
C-56
-------
Landis Comments September 1,1993
9
do not span complete reproductive cycles. Associated uncertainties are summarized in section
5.1.3. Discuss the utility of these data and uncertainties for evaluating ecological effects.
Comment
The database is one of the best for a non pesticide organic that I have seen. Most of the
emphasis is on fish and laboratory animals, and I would have enjoyed more of a comparative
molecular toxicotagical approach to ascertain the range of sensitivities. I would suspect that given
the sensitivities of birds, mammals and fish that amphibians and reptiles would show similar
sensitivities. Amphibians, with their early life stages comparable to fish, would be likely to see
similar sensitivity. Reptiles are a difficult extrapolation, since they are a broad group of organisms
in North America. Alligators, lizards and turtles are evolutionary distinctive, and quite separate
from the vertebrates of today. In fact, alligators may actually share more of an evolutionary linkage
with birds that with the lizards.
A primary emphasis of this database are direct effects, the initial disturbance to the system
that drives the follow on effects. Indirect effects are now being realized as more important and
there is a considerable literature detailing these impacts for pesticides and other materials. Given
the test methods, the type and magnitude of the indirect effects are virtually impossible to
accurately predict. . ,
Many of the questions about effects, extrapolation from laboratory tests and the
examination of indirect effects could have been addressed by appropriately designed
multispecies toxicity tests. Multispecies toxicity tests come in a variety of methods using bacteria,
protozoa, a variety of metazoans, including fish, and can be used to test experimentally many of
the assumptions and validate the models that are subsequently used in a risk assessment: Many
of the methods are not experimental and several have histories of round robin testing and
adoption by various laboratories. Examples with short summaries are presented in Appendix!
Multispecies toxicity tests can also examine the dynamics of the entire community using a variety
of tools developed overthe last ten years. These methods range from normalized ecosystem
strain developed by Kersting (1988), the state space of Johnson (I988a, 1988b) to the :
nonmetric clustering and projection techniques developed by Matthews etal (1991, Landis et al
in press). Examining these relationships experimentally allows verification, validation and may turn
up new relationships not easily derived from observatfonalstudies.
7. As discussed in the Interim Report, few data on the effects of TCDD on estuarine and marine
organisms have been reported (section 4.2.1.5), and no data were found n the literature for TCDD
effects on reptiles or marine mammals. Although all the current wildlife toxicity data were
reviewed, an analysis to establish an effects profile for terrestrial organisms was beyond the scope
C-57
-------
Landis Comments September 1,1993
10
of the report. Describe other effects data not identified in the Interim Report or the scenario that
will be important for future ecological risk assessments.
Comment
The lack of data on marine species is an important data gap, especially since pulp and
paper mills are found along the Puget Sound and along the British Columbia coast. Marine
systems, especially those of the Pacific Northwest, are highly productive and important
commercially. Commercially important species include fish, crabs and shellfish. Each of these
commercial species use a variety of different resources.
Comparative data on immune suppression or even enhancement are crucial in estimating
the increased risk of individuals to disease. Immunotogically suppressed populations may be
more susceptible to devastating outbreaks of opportunistic infections.
Behavioral data on nesting patterns, parental care, sensory impediments, and other
factors can also be important in the long term success of populations. These can be crucial with
organisms that migrate to nesting or spawning grounds, rely of behavioral cues to establish mates,
or are crucial in feeding offspring or providing other forms of parental care.
As typical, no data are forthcoming on additive or synergistic effects with other classes of
toxicants. It is unlikely that TCDD and related compounds will be the only toxicant or even the one
in the highest concentration. Are there synergisms When herbicides are added to the mix that
increase the sensitivity of algae and macrophytes ? As the various serene dependent proteins
are inhibited by acetylcholinesterase inhibitors, does the presence of TCDD inhibit turnover and
therefore make the organisms more sensitive to repeated dosing? What role do heavy metals play
in concert with TCDD? Do the genotoxic aspects of TCDD make the organism more or less
susceptiDle to other genotoxic materials, many of which like the aflatoxins, naturally occurring? It
must be recognized that toxicants do not exist alone in the environment and hazard assessments
should recognize the potential interactions.
C-58
-------
Richard Peterson
School of Pharmacy
University of Wisconsin
C-59
-------
-------
PRE-MEETING COMMENTS: RICHARD E. PETERSON
EXERCISE 1; ECOLOGICAL EFFECTS AND ENDPOINT SELECTION
1. If the risk characterization had to be done today,
focusing on early life stage mortality of fish as the
measurement endpoint to use for protection of fish and the
rest of the aquatic community in the reservoir from direct or
indirect effects of TCDD seems reasonable. This is because
early life stage mortality in fish is the most sensitive
endpoint reported for aquatic animals (or plants). Also it is
an ecologically-relevant endpoint that can potentially be
linked by cause-and-effect in population models to predict
declines of fish populations in the environment.
Since literature findings show that aquatic
invertebrates, amphibians and plants are less sensitive to
TCDD toxicity than fish it would be inappropriate to focus on
non-fish species in characterizing the risk to aquatic life in
the pulp mill scenario. Nevertheless a more rigorous screen
for presence of the Ah receptor and for Ah receptor-mediated
toxicity is needed for representative species of invertebrates
that inhabit the reservoir. .This toxicity screen should
include long term chronic toxicity tests that include
measurements of TCDD accumulation in invertebrates. In
separate studies expression of the Ah receptor in
invertebrates should be assessed. Obtaining this information
will reduce uncertainty as to whether the proposed focus on
fish, will also protect invertebrates.
There is less concern about potential adverse effects of
TCDD on aquatic plants, but determining if plants lack Ah
receptors would be useful. Uncertainty about the toxicity of
TCDD in amphibians at all life stages of development should
also be determined by conducting studies in a prototype
C-61
-------
amphibian species that , is commonly exposed to TCDD-like
compounds in inland waters of the United States. Such studies
should focus on reproductive and developmental toxicity
endpoints and include Ah receptor analyses and documentation
of TCDD body burdens associated with toxic endpoints.
2. For predicting population declines in reservoir fish
populations caused by complex mixtures of TCDD-like compounds
in eggs, the fish^specific TEFs determined by Walker et al.
(1991) offer the following advantages: (a) TEFs are based on
a sensitive, ecologically-relevant, Ah receptor-mediated
endpoint in fish - early life stage mortality, (b) TCDD-like
PCDD, PCDF, and PCB congeners act as full agonists in
producing the response, (c) dose response curves of individual
congeners for producing early life stage mortality are
parallel, (d) complex mixtures of, TCDD-like congeners are
assumed to interact in a near additive manner, and (e) fish-
specific TEFs have the potential to be linked by cause-and-
effect in population models that will be developed in 'the
future to predict TCDD-induced declines in feral fish
populations.
Another method for determining TEQs involves extracting
all PCB, PCDD, and PCDF congeners from a biological sample and
testing it's potency relative to that of TCDD in a fish or
mammalian cell culture system. The endpoint used is generally
induction of cytochrome P4501A1—mediated enzyme activity .
However, TEF values obtained with this method tend to be
greater than TEFs based on early life stage mortality in
rainbow trout. In other words, cell culture systems tend to
overestimate the potency of TCDD-like PCDDs, PCDFs, and PCBs
in causing early life stage mortality in fish.
Another problem with using cell culture systems to
determine TEFs is that all PCDD, PCDF and PCB congeners tested
C-62
-------
in these systems do not act as full agonists. That is, the
congeners do not all produce the same maximal level of
cytochrome P4501A1 induction. Yet this is necessary for the
proper determination of EC50 values for TEF determination. If
certain congeners are not able to cause the same maximal
level of induction as TCDD the addition of TEQs contributed by
these congeners is not valid for risk assessment.
For example, assume maximal induction achieved with TCDF
in a cell culture system is 40% of that obtained with TCDD and
an EC50 was calculated for TCDF relative to its own maximal
effect. If that EC50 was used to determine a TEF for TCDF;
TCDF's TEF would be artificially high relative to TCDD as
would TEQs calculated for TCDF using that TEF. Thus, fish-
specific TEFs based on a response where all congeners act as
full agonists is preferred.
Fish^specific TEFs for PCDD and PCDF congeners and for
coplanar PCBs have been published by Walker et al. (1991) and
should be used in the risk characterization process for fish
in the Omigoshee Reservoir. TEFs used by the US EPA for human
health risk assessment of PCDDs and PCDFs are not specific for
early ,1ife stage mortality in fish; it is suggested they not
be used for this purpose. TEFs proposed by Safe (Crit. Rev.
Toxicol,, 1990) for PCBs also should not be used for fish
because they overestimate the potency of PCBs in producing
early life stage mortality. For example, mono-ortho chlorine
substituted analogs of the coplanar PCBs (mono-ortho PCBs) are
weak Ah receptor agonists in mammals but are completely
inactive in causing early life stage mortality in rainbow
trout. ; •';..-..-•
TEFs used to predict adverse effects of TCDD-like
compounds on piscivorous bird and mammal populations should be
different than those used for fish. Ideally, one would like to
C-63
-------
have TEFs based on developmental and reproductive toxicity
endpoints in these wildlife species, but such data does not
exist and will not be forthcoming in the near future. In the
absence of such data, TEFs for PCDDs and PCDFs used by the EPA
for human health risk assessment could be used for ecological
risk characterization of avian and mammalian wildlife. For
coplanar PCBs and mono-ortho PCBs TEFs proposed by Safe (1990)
should be modified, based on findings of De Vito et al.
(1993), to calculate TEQs.
3. A key determinant of whether TCDD-like congeners will
produce toxicity is their concentration at their site of
action in fish, birds and mammals. Exposure concentrations do
not provide this type of information. Concentrations of TCDD
in water, food, soil and sediment are too far removed from the
site of action of TCDD in the body to be as useful for
ecological risk characterization as tissue levels of TCDD. The
most useful laboratory studies for ecological risk
characterization are those that assess TCDD toxicity and TCDD
accumulation in tissues of the same animals in the same
investigation. Furthermore, expressing effects on the basis of
tissue accumulation of TCDD allows for toxic effects observed
in aquatic life and wildlife from TCDD exposures by different
routes in laboratory and field studies to be directly compared
and interpreted more meaningfully than would otherwise be
possible.
4. Since the assessment endpoints proposed for risk
characterization are productivity of fish, avian and mammalian
populations the most meaningful measurement endpoints for TCDD
are its known adverse effects on reproduction in birds and
mammals and early life stage mortality in fish. All of these
endpoints have the potential to be linked to decreases in the
populations of fish, birds and mammals once species-specific
population models are developed and validated in the future.
C-64
-------
In the Interim Report results are presented showing how early
life stage mortality of lake trout in Lake Ontario, caused by
contamination of lake trout eggs by TGDD-like congeners, could
be linked retrospectively to decreased productivity of lake
trout. This example provides strong support for the use of
early life stage mortality as a sensitive, ecologically-
relevant, Ah receptor-mediated measurement endpoint for fish
in the Omigoshee Reservoir. For birds, the .measurement
endpoints are based on TCDD-induced decreases in egg
production and embryo viability which might also be able to be
linked to population declines in piscivorous birds once
population models for these species become available. For
mammals decreases in fertility and litter size also have the
potential of being linked to changes in populations of mink
and otter along the shores of the reservoir.
Nevertheless whenever laboratory data are used to predict
effects on field populations of fish, birds and mammals
uncertainty will exist. In particular there is uncertainty
about the potential modulating influence of non-chemical
stressors encountered by fish, birds and mammals in their
natural environment on measurement endpoints. More
specifically, all laboratory data relating to the measurement
endpoints were obtained in lake trout fry, pheasants and rats
under constant conditions of water temperature, water pH, room
temperature and room humidity and standard laboratory diets
and water were provided ad libitum. Compared to the rigors of
the natural environment for aquatic life and wildlife these
laboratory environment conditions under which the measurement
endpoints were assessed were nonstressful. Thus, it is
uncertain for each measurement endpoint and for each species
of feral animal whether or in what direction naturally
occurring non-chemical stressors that the species encounters
in its natural environment will have in altering its
sensitivity to the reproductive or developmental toxicity of
C-65
-------
TCDD.
A non-chemical stressor that may or may not modulate
early life stage mortality in fish is water temperature. The
dose response assessment for TCDD-induced early life stage
mortality in lake trout was conducted at a colder water
temperature than fish would encounter in the Omigoshee
Reservoir in the southern United States. However, until
studies are conducted to determine if water temperature
modulates TCDD-induced early life stage mortality in fish this
type of uncertainty will be inherent in the risk
characterization.
5. The main implication of the steep TCDD dose response
curve for reproductive/developmental toxicity is that the
response of fish, mammal or bird populations to TCDD-like
congeners in the environment may be "all or none". Using fish
as an example, until a threshold concentration of TCDD in the
eggs is reached none of the embryos of a particular fish
species in the reservoir would be expected to be adversely
affected. However, a three-fold increase in the egg TCDD
concentration (above the threshold for early life stage
mortality) may cause 100% of the fish embryos of that species
to die. This scenario assumes that the slope of the dose
response curve for early life stage mortality for fish in the
reservoir will be as steep as it is for lake trout where the
NOAEL and LD100/ 34 and 104 pg TCDD/g egg, are separated by a
factor of three.
Another implication of the steep dose response curve is
that if any of the fish or piscivorous mammal or bird species
inhabiting the reservoir are more sensitive than the surrogate
species used in laboratory studies, and an extrapolation
factor for species differences in sensitivity to TCDD is not
applied as part of the risk characterization process, then a
C-66
-------
decline in the population of that highly sensitive species may
occur. In this regard there is concern about the failure to
incorporate an extrapolation factor in the risk
characterization for fish. Uncertainty in the fish species
sensitivity distribution for early life stage mortality raises
the very real possibility that egg TCDD concentrations that
provide adequate protection for lake trout (surrogate species)
may not necessarily do the same for largemouth bass, catfish,
crappie or bluegill that inhabit the Omigoshee Reservoir.
6. Reproductive and developmental toxicity of TCDD has
been studied in few of the 10,000 to 15,000 freshwater fish
species and in even less estuarine and marine fish species.
Also it is well known that wide species differences exist
among vertebrates, including fish, in sensitivity to TCDD
toxicity. Given these facts and the uncertainty in the
distribution of fish species sensitivity to TCDD-induced early
life stage mortality it would be prudent to apply an
extrapolation factor for species differences to the NOAEL for
TCDD-induced early life stage mortality in lake trout. In
support of this suggestion findings of Helder (1980; 1982a,b)
suggest that northern pike may be as sensitive or more
sensitive than lake trout to TCDD-induced early life stage
mortality. Furthermore, future research may show adverse
effects on reproduction of adult fish occurring at body
burdens of TCDD that are lower than those that cause early
life stage mortality in lake trout. Yet no extrapolation
factor is applied to cover this possibility either.
Additional uncertainty for TCDD risk characterization in
fish is in not knowing whether TEFs for early life stage
mortality determined in rainbow trout are appropriate for
other freshwater, estuarine and marine fish species. We assume
this to be true, but it has not been verified experimentally.
If bluegill were responsive to early life stage mortality
C-67
-------
caused by the mono-ortho analogs of the coplanar PCBs the
contribution of mono-ortho PCBs to TEQs in bluegill eggs would
be greatly underestimated by using TEFs determined in rainbow
trout (because mono-ortho PCBs do not cause early life stage
mortality in rainbow trout).
If early life stage mortality is used as a measurement
endpoint for ecologic risk characterizations in fish it would
be prudent to document Ah receptor expression throughout early
development in surrogate freshwater, estuarine and marine fish
species. There is rib information on Ah receptor expression at
any time during fish early development in any fish species.
Inasmuch as the risk characterization for TCDD-like congeners
in fish is based on the premise that early life stage
mortality is Ah receptor-mediated obtaining this type of
information would help remove such uncertainty.
Given the limited TCDD database for reproductive toxicity
studies in piscivorous mammals and birds, the laboratory
studies selected (Murray et al., 1979 - rats; Nosek et al.,
1992 - pheasants) to determine fish concentrations of TCDD
associated with risk to mammalian and avian wildlife
populations seem reasonable. Also the extrapolation factors
applied to the results of each study to estimate risk to the
most sensitive species seem appropriate. However, considerable
uncertainty exists with each of these studies. It is due to
several factors such as the one order of magnitude difference
between TCDD treatment levels in each study. Also in
extrapolating from laboratory animals to wildlife, species
differences in the toxicokinetics and toxicodynamics of TCDD
contribute further to the uncertainty. Lastly, it is assumed
in the risk characterization that piscivorous avian and
mammalian wildlife will consume only fish and aquatic
invertebrates from the Omigoshee Reservoir. This seems
unlikely. Certain species like mink will likely consume other
C-68
-------
food items and birds such as bald eagles will undoubtedly
forage over a wider range than the reservoir itself and will
consume food other than aquatic animals obtained from the
reservoir.
7. In the future other reproductive/developmental
endpoints may be shown to be more sensitive to TCDD exposure
than embryo mortality. If sublethal effects of TCDD on pups,
hatchlings or fry (that decrease their ability to survive in
the natural environment) are discovered in the future they may
be more appropriate endpoints than embryo mortality. It is
also possible that future research may reveal adverse effects
of exposure to TCDD-like congeners during early development in
fish and birds (i.e., during sexual differentiation) that
cause adverse effects on reproductive function that are not
manifested until adulthood. In support of this notion Mably
and coworkers (1992) found that in utero and lactational
exposure of male rats to TCDD decreased growth of androgen
sensitive sex organs, inhibited spermatogenesis and feminized
sexual behavior in adulthood. Lastly, there is a paucity of
reproductive and developmental toxicity information on TCDD-
like congeners in marine mammals and none in reptiles. Yet
premature pupping in sea lions contaminated with PCBs (DeLong
et al., 1973) suggests that marine mammals may be sensitive to
TCDD-like congeners and should be included in future
ecological risk characterizations.
Despite the possibility of more sensitive, ecologically-
relevant effects of TCDD being discovered in the future, it is
important to state, at this point in time, that the TCDD
measurement endpoints proposed in the Interim Report for
linkage to population effects in fish, birds and mammals in
the pulp mill scenario are the most appropriate ones to use.
The challenge now is develop and validate species-specific
population dynamic models which will link these laboratory
C-69
-------
effects of TCDD to shifts in the feral populations of fish,
birds and mammals contaminated with tfCDD-like congeners.
C-70
-------
Thomas Sibley
Fisheries Research Institute
University of Washington
C-71
-------
-------
Pre-meeting Comments for U.S. Environmental
Protection Agency
Workshop on Ecolological Risk Assessment Issues for
2,3,7,8-Tetrachlorodibenzo-p-Dioxin
14-15 September 1993
Thomas H. sibley
Fisheries Research Institute (WH-10)
University of Washington
Seattle, WA 98195
Exercise 1. -Ecological Effects and Endpoint Selection
1. In an ideal situation it would be desirable to have
information on the toxicity or biological effects of TCDD to
all the "important" species in the aquatic community being
considered. However, the available data suggest that fish
provide an appropriate focus to protect the rest of the
aquatic resources. I question the use of productivity to
assess TCDD effects. Productivity has a specific meaning to
fisheries biologists and it is sometimes difficult to obtain
sufficient data to calculate productivity. In addition
there tends to be significant interannual variability that
confounds any interpretation. It may be more useful to
consider biomarkers or mortality of juvenile fish.
2. I am not in favor of TEPs. Relative toxicity is often
not equivalent for different species so TEF values are
species specific. Also/ the bioavailability, and
consequently the biological effects, of different chemicals
will depend upon the physico-chemical properties of the
environment and each chemical may be altered differently.
Again, if the selected endpoint was a biomarker(s) the
additive effects of different chemicals would be evaluated
directly. In an ideal situation it would not be necessary
to use BSAF factors because one could measure dissolved
cocentrations of TCDD. Realistically, this is a useful
approach because the chemical concentrations can be measured
accurately enough to compensate for errors introduced by the
normalization procedures,
3. Again, because of the problems of measuring TCDD in
solution, I believe this is an appropriate approach.
4. Obviously, laboratory test data at the individual level
can be used to predict effects on field populations. It is
necessary, however/ to validate those predictions. Despite
C-73
-------
the data that are presented for Lake Ontario it is very
difficult to obtain reliable validations, since high
mortality is normal for most populations in natural systems,
the magnitude of increased mortality from toxicants is hard
to quantify. Specifically, those individuals that are
resistant to the toxicant have an increased probability of
surviving if density-dependent processes are important.
5. Steep concentration-response curves allow one to make
accurate predictions about the effect of a toxicant to a
particular species. However, the concentration-response
curve is species specific and cannot be extrapolated to
other species. If we know the response of sensitive
species, we can establish conservative standards. To
establish appropriate standards for a particular aquatic
community, it is necessary to have concentration-response
curves for the species in that community,
6. It is always the case that we would like to have more
data, and the uncertainty would be reduced somewhat if there
was more data. It appears that the authors have provided an
extensive review of the available information which includes
information on sensitive species. Therefore, it is possible
to establish reasonable limits for fish and wildlife.
7. There is very little information on plants and
invertebrates, only four publications are cited in Table 4.
C-74
-------
John Stegeman
Woods Hole Oceanographic Institution
C-75
-------
-------
Exercise 1. Ecological effects andEndpoint selection
John Stegeman, WHOI
General Comment.
The Interim Report in some respects does not go far enough in the recommendations made
to address important issues. The Issues for Consideration raised here, and drawn from or
based on the Interim Report are important issues. These issues have a common feature: In
each case, there is a deficiency in our basic understanding of the mechanisms underlying the
toxicokinetics and the action of dioxin and other Ah-receptor agonists. The concerns
regarding species extrapolation, dose extrapolation, and inducer structure-activity
relationships will not be adequately nor defensibly resolved until there is an adequate
mechanistic foundation for the effects of TCDD and its relatives. This requires more than
simply identifying the presence or absence of an Ah receptor-like protein or system in selected
taxonomic groups, but eventually establishing exactly how it is that agonists and antagonists
interact with the receptor, which genes are regulated in addition to CYP1A, the mechanisms
by which the genes regulated, and how they are involved in adversely affecting critical cellular
processes.
Issue 1.
If it can be shown that fish are indeed the most sensitive animal group, and that is true for
all species, then the reliance on fish as a surrogate for other species in an ecosystem or for the
ecosystem in general could be adequate. However, there is 1,000-fold plus range in
sensitivity of mammalian species and strains to dioxin toxicity. Evidence suggests that there
are equally extreme differences among fish species. Whether the same will be true of early
life stages is not yet known. If there are species with resistant early stages, then fish
productivity may not provide adequate protection for the system. It is also possible that other
components of some ecosystems may be more sensitive than the resident fish, regardless of
developmental stage. Until mere is an additional foundation of knowledge to explain the
basis underlying species differences and developmental differences in sensitivity, then the
issue of species extrapolation will be unresolved.
Alternately, changes that presage more extensive effects might be monitored in a caged
surrogate species. However, that too would require knowing the precise pathway or function
to be examined. However, it is not clear what level of "protection" is desirable, or necessary,
in a given setting, and how one coupled infer that level from responses in a surrogate species,
C-77
-------
unless the sensitivity of that species was known to be greater than that of all resident species.
Further information on species or group differences in inducer-SAR is required. These
must be empirically established. Further, there is growing evidence that in some systems the
most common basis for determining TEF, i.e., induction of cytochrome P4501A forms, can
be negatively influenced at some doses. The utility of TEFs based on the action of
compounds in mammalian cells we know to be inappropriate for evaluating the sensitivity of
fish. The use of BASF can also be species dependent.
Issue 3.
Obviously, tissue concentrations Internal dose) is most revenant in attempting to establish
the dose-response relationship fora given effect, and for comparing the sensitivity of different
species, etc. Not only the tissue concentration, but also the concentration in a given cell type,
if specific target cells can be identified. However, once having established a relationship
between environmental concentrations and those internal concentrations, and validated that
relationship in the laboratory and the field, then one might use either tissue or environmental
residue concentrations to evaluate risk. The reliance on environmental concentrations would
require a knowledge of the bioavailability from the environmental matrix of greatest concern,
at each site. It would be less arduous to measure the tissue concentrations.
Issue 4.
At present the laboratory data are not, to my knowledge sufficient to properly predict
effects on field populations. On source of uncertainty is the lack of long term studies in the
laboratory. The time of exposure in most environmental settings would be life-time.
Laboratory studies simply have not provided sufficient information of that type.
A second problem is in the lack of suitable laboratory data on the interactive effects of
various compounds. Both of these need to be addressed before any sound predictions can be
made for effects on individuals, much less populations.
Issue S.
The steep dose-response curve implies that a more sensitive site or endpoint than that
examined would be very useful. Selection of such an endpoint would again depend on a
mechanistic understanding of the processes involved in toxicity. It is also possible that the
difficulties of dose extrapolation will be better resolved when the factors influencing; the
C-78
-------
toxicokinetics are better understood. For example, we have long known that the endothelium
is a site of very strong induction of CYP1A, and hypothesize that the endothelial cell CYP1A
could determine the nature of low dose effects and the penetration of highly efficacious but
low mass ligands, like TCDD, into critical target cells. The role of the endothelium must be
understood if we are to put dose extrapolations on a sound footing,
Issue 6.
Risky extrapolations.
Issue 7.
It is important to re-emphasize that a mechanistic basis for effects must be obtained
The types of information that are required but that are not in hand include: Identification of
critical target cells; identification of critical target pathways; an understanding of the
relationships between dioxin and the processes of growth and differentiation at the cellular
level, for estimating the pathways to effect but also for developing more reliable in vitro
approaches to determining TEFs. The basis for AhR involvement in toxic effects must be
established, and the basis for any distinction between AhR and non-AhR mechanisms. The
significance of CYF1A induction as a marker will need to be determined. In general, the
significance of CYP1A induction will depend on 1) the catalytic function(s) of the
protein(s), 2) the relative rates of activation and detoxification of the inducer and other
compounds, 3) inducer avidity for Ah-receptors and hence efficacy in eliciting CYP1A
induction or other gene regulatory changes, and 4) the sites where these events occur.
Though most often studied in liver, induction and attendant changes in extrahepatic sites
may determine the toxicity of inducers.
C-79
-------
-------
Bill Williams
Ecological Planning Toxicology, Inc.
C-81
-------
-------
I have attempted -to address -the listed items in order, but with my
emphasis on avian and terrestrial wildlife, including some generic issues
about food chain exposure assessment and overall concepts of chemical
additivity and effects in the context of pulp mill scenarios. I have
focused on those issue on which I feel I can provide the best commentary.
Exercise 1. Ecological Effects and EndPoint Selection
Issues for Consideration
1. Focus on life stages of fish in aquatic risk scenarios.
Amphibians, invertebrates, and plants are less sensitive to TCDD than
fish, birds and mammals. For fish, early life stages appear to be most
sensitive. Because of this range in sensitivity, productivity's of
fish species were selected as assessment endpoints for the scenario.
Comment on the whether this focus on fish species will result in
adequate protection for the rest of the aquatic community in the
.reservoir from the direct or indirect effects of TCDD.
Comment:
Decreased sensitivity of amphibians and invertebrates to TCDD and other
TCDD-like chemicals suggests that risk assessment approaches focused on
fish early life-stages will provide a conservative protection of other
aquatic species. It is not clear as to the physiological basis for
decreased sensitivity in amphibians and invertebrates, but it is likely
that target tissue (receptors fro TCDD) must be available for the toxic
effects to be manifest. In simple systems, it may be that receptors do
not exist for complex entities are not recognized to any extent, thus
treated the same as inert molecules. I suggest that the rationale to use
early life-stages provides a conservative risk assessment for the aquatic
community.
2. TEPs
Use of toxicity equivalency factors (TEFs) for TCDD-like compounds.
Comment on the use of these approaches for evaluating the effects and
bioaccumulation of dibenzodioxins and dibenzofurans in the paper mill
effluent.
Comment:
Additive Nature of Organochlorine Chemicals, and TEF Concept.
C-83
-------
Numerous articles support the contention that organochlorines probably
<|o not act in a simple additive mode.• There is no reported "generic
effect" of organochlorines other than the fact that they are generally
lipophilic and accumulate primarily in fatty tissue. In most studies
using birds, toxic effects of organochlorines, TCDD for instance,
organochlorines, in general, do not appear to thin egg shells or
adversely affect reproduction. The effect of organochlorines on
reproduction is still generally attributed specifically to individual
chemicals. While numerous studies have attempted to produce an
organochlorine additiviity matrix for reproductive effects, many
organochlorines show no demonstrable negative effects. It is clear that
the organochlorines act in species-specific and chemical-specific ways.
The effects of TCDD on reproduction has recently been evaluated
specifically in birds and fish, and it is still unclear as to the
interaction of these specific chemicals with other organochlorines.
There appears to be a toxic effect of TCDD on early life stages in many
birds, but much work is needed to delineate a clear dose response
relationship. Although it is difficult, at best, to utilize a one-by-
one evaluation of the toxicity of these chemicals, the database is still
too slim to begin to build a complex matrix of combinations of these
chemicals. It is clear that induction of AOX, EROD, and other enzyme
induction is a natural response and clearly indicates exposure to one or
more of these chemicals. It is further clear, however, that exposure is
not an effect. The relationship between enzyme induction and effects
must be the focus of continued researched in this area. Some recent
publications are attempting to provide these causal relationships, but
with limited success. This further supports the contention that while
the AOX and EROD measurement represents a measure of the organochlorine
load, it is not necessarily an indication of the "total additive impact
of organochlorines" in the medium.
Chemicals in the pulp mill effluent exhibit a range of biological
responses; some being primarily time-dependent and some being primarily
dose-dependent. Therefore it is inappropriate to consider the toxicity of
the suite of chemicals as strictly additive. The correct interpretation
is that the toxicity is weighted, and probably less than additive. Given
the current stage of scientific understanding, there is no method that
permits the accurate prediction of toxicity of such complex mixtures based
on -the individual chemical toxicity determinations. It is for this
reason that regulatory entities (eg. EPA, state DEQs) have relied on
bioassay toxicity tests that evaluate the site specific toxicity fo the
medium.
C-84
-------
3. Using •tissue levels of TCDD to evaluate effects— . ,
Comment on the applicability of this approach to evaluating the risks of
TCDD from the pulp mill effluent. ..-..•--
Comment:
Host toxicity data are collected from experimentations with single
chemicals administered in different doses. Unfortunately, the exposure
of aquatic and terrestrial wildlife outside the laboratory is not so
clear-cut. It is difficult, at best, to develop the exposure component
of a risk assessment using the environmental concentration (water or air
or soil) since it is clear that environmental concentration is not
necessarily exposure. Under perfect uptake and utilization, the aquatic
concentration of a chemical may be the most representative of the
possible exposure, but the actual uptake and sequestering of chemical
must be defined in order to properly extrapolate effects. Many
environmental toxicity tests (in the assessment process) are performed
with unknown/undefined mixtures of potentially toxic chemicals.
Attempts to predict toxicity of mixtures or to interpret causality of
individual chemicals in complex mixtures are typically frustrating.
Although tissue concentration may provide some indication of sequestered
chemical, it also suffers from the fact that sequestered chemical is not
necessarily producing toxic effects. It is the best indication of
exposure, however, and provides a more firm foundation for the estimates
of exposure. Some of the dialog regarding AOX or EROD measurements as
indicators of potential toxicity is complicated by the knowledge gap ,
prevailing over how to deal with complex mixtures. Attempts to
correlate the tissue levels or AOX or EROD values with biological
responses observed in various single-chemical toxicity tests or from
effluent toxicity tests are confounded by the variability introduced by
the fact that chemical interactions at the cell level (metabolic and
physiological changes) produces measureable, observable effects, not
just the exposure. The concept of appropriate cellular receptors of
action to produce toxic effects is once again very attractive.
In recent publications, the US Fish and Wildlife Service (Kisler, US
FWS) has recommended that the 2,3,7,8-TCDD limits necessary to "protect"
wildlife should be set at:
o 10-12 ppt in food items (prey) for birds and mammals. This level is
considered protective of terrestrial wildlife.
Similarly, the US FWS has recommended that the 2,3,7,8-TCDD level in
water that is still protective of aquatic wildlife (fish) should be:
o 0.01 ppt = 10 ppq
C-85
-------
The concentration of organochlorine residue in wildlife tissue is a
constantly changing function determined primarily by the concentration
of the chemical in the exposure route (water for aquatic animals and
food for terrestrial animals). It is generally held that the rate of
loss (depuration) is approximately the same as the rate of gain
(uptake). Further, the rate of bioaccumulation/ bioconcentration varies
greatly between species and chemical. The common practice of "back
calculating" water or food concentration (exposure) using tissue
concentration should be thought of as a means of developing an
hypothetical value. The proof of the hypothesis is the actual measured
environmental value. The use of an historic residue value of "a fish
caught in 1985" etc., simply does not provide the needed validation of
the relationship. To be valid, the proper approach is to show that the
tissue concentrations are present in a randomly collected (but site-
specific) sample. Any use of anecdotal information to demonstrate the
back calculation hypothesis is neither scientifically nor statistically
meaningful for 2,3,7,8-TCDD or for any other organochlorine.
Avians show little bioaccumulation of 2,3,7,8-TCDD, and low levels (<20-
25x) of accumulation of other organochlorines. This is thought to be
related to their ability to metabolize many chemicals, either altering
their mode of action or inactivating them. Again, this phenomenon is
generally species-specific.
Because of their high oil content, fish present the worst case for
organochlorine uptake because the lipophilic characteristic of
organochlorines results in their being sequestered in fat. However, it
is also true that these deposits of chemicals are extremely variable and
change in response to changes in the exposure level.
4. Use of laboratory effects on individuals -to predict effects on
populations.
Use laboratory test data at the individual level of organization to
predict population changes in fish and wildlife. Comment on the utility
of available laboratory data to predict effects on field populations and
discuss the associated uncertainties.
Comment:
Attempts to utilize individual (laboratory and field) data to determine
the probable impact of chemicals on populations of non-target wildlife are
generally frought with problems. Although it is not yet practical to
conduct field studies that determine the effects of chemicals on entire
C-86
-------
populations of wildlife, it is likely that efforts will be directed toward
determining the impact of these chemicals on sub-populations or local
populations of wildlife. The use of. individual toxicity data for
predictive models to simulate the effects of chemicals on terrestrial
populations will be impeded by the enormous variety of species and special
conditions associated with sites such as pulp mills and habitat associated
with these sites. It simply is not possible to reliably extrapolate
laboratory toxicity data to construct models that deal appropriately with
all species and all circumstances of interest. Therefore, it would be
useful, before building predictive population models from individual
toxicity data, to determine which input (population) parameters are
critical to the assessment endpoints. Once these assessment parameters
are determined, they can be organized into hierarchical levels of
importance vis a vis their relative impacts on population density or
fitness. Use of such information will eventually enable the construction
of simpler and more generic, predictive terrestrial population
assessments. Important parameters that modify toxicity estimates that are
based solely on tests of individuals include two, partly separable
components: one comprising the purely mechanical descriptors of dynamics
from given demographic parameter values, and the other describing the
modulation of the demographic parameters by environmental factors such as
changes in the physical environment, species interactions, pathogens, and
xenobiotic chemicals. These problems will continue to confound the
problems associated with extrapolations of individual data to populations
in the terrestrial environment.
5. Implications of Slope of Chemical Toxicity:
The Interim Report sites data that indicate a very steep concentration
response curve for TCDD effects in fish and wildlife. Discuss the
implications of this observation for evaluating ecological effects in the
scenario.
Comment:
It is well known in classical toxicology that the toxicity of a chemical
is altered in numerous ways, including temporal and spatial presentation.
Additionally, many chemicals exhibit threshold levels that correlate with
onset of effects and toxicity. A steep slope of toxicity suggests that
the level of error allowed in the risk assessment is lessened and the
assessment should be more conservative. This concept is chemical-
specific, however, and merely suggests that good laboratory data is needed
with more than a trigger number (LCSO or LDSO). A receptor to TCDD that
saturates at a specific level is consistent with the current information
C-87
-------
about the toxioity of this chemical. Special consideration must be given
to the threshold and/or receptor concept in risk assessments for TCDD.
6. Uncertainty from incomplete data sets-
The general summary of effect levels for aquatic species and associated
wildlife is based on extrapolations from a limited number of test species
and from tests that do not span complete reproductive cycles. Discuss the
utility of these data and uncertainties for evaluating ecological effects.
Comment:
It is only appropriate to utilize the available data in a risk assessment,
regardless of the gaps. In the data review process, it is critical to
identify data gaps and uncertainty, in general. It is probable that data
gaps will always exist during any real assessment, demanding the
extrapolation from available information. Current extrapolation
techniques in risk assessments add a safety factor (xlO?) for each level
of uncertainty. These safety factors can be used for data based on
surrogates (sensitivity differences) and for any other extrapolation for
which actual data are not available. Usually, these extrapolations are
overlly conservative and provide a satisfactory margin of safety by
accounting in the negative sense, for these errors in such estimates.
7.0. Few data on esturarine and marine organisms-
No data were found in the literature for TCDD effects on reptiles or
marine mammals. Describe other effects data not identified in the Interim
Report or the scenario that will be important for future ecological risk
assessments. •
Comment:
It is likely that each ecologically realistic risk assessment will be
missing data from numerous components of the system being evaluated. It
is less important to cover each and every species and possible exposure
scenario in a risk assessment than to estimate (measure) effects at some
of the higher organizational levels. Use of top predators as flags to
trigger concern about the total impact of a stress at lower levels. It is
generally believed that the most appropriate sentries of environmental
damage occur at higher levels of organization. Unfortunately, this
approach is flawed in its inability to detect early, but possibly
important, impacts on the understructure of the ecosystem. The most
contentious issue in this approach is that there may be numerous dramatic
impacts of chemicals on species and communities at lowest ecological
C-88
-------
organizational levels that result in irreversible impacts to those animals
affected.
C-89
-------
-------
Section 3
EXERCISE 2
Stressor Characterization
Workgroup Leaden
William Adams
ABC Laboratories
C-91
-------
-------
William J. Adams
ABC Laboratories
WORKSHOP ON ECOLOGICAL RISK ASSESSMENT ISSUES FOR 2,3,7,8- TCDD:
PREMEETEVG COMMENTS
Question #
8. Difficulty in measuring KOW and K^. for TCDD and its implications for stressor
characterization:
Measurement of Kow and K^ for highly hydrophobic chemicals like TCDD is hindered by operational
difficulties. The inability to separate colloids (dissolved organics) from the solute phase in K^.
measurements or to accurately account for the activity of TCDD in water saturated-octanol ha Kow
measurements leads to operationally defined measurements that underestimate the chemical property
of interest. The best approach is to measure and estimate the property of interest a number of
different ways using extreme care and look for convergence of the data. A careful review of the K,,w
data for TCDD suggests that the Log K^, is 7.3-7.4. The inability to accurately measure key
physical/chemical properties for hydrophobic compounds indicates a critical research need.
Approaches for characterizing exposure for TCDD risk assessment might include the following:
1. Assume the Kow for TCDD is Log 7.3 since this appears to be realistic, is hi good
agreement with calculated Log P values, and agrees with the K^ measurement of Jackson.
Current models require the use of a number and this is as good as it is going to get for the
present.
2. Existing theory on the use of K^ for estimating bioconcentration in organism tissue,
predicting K^, etc. is based on extensive data sets which have been developed with organic
chemicals which have Log K^ values hi the range of < 1^6. For this range of K»ws the
relationships are quite good. However, we cannot have the same confidence hi this estimator
(Row) for values above a Log of 6.0. There are several reasons for this which include
analytical difficulties hi measuring extremely low concentrations, micelle formation with
extremely hydrophobic compounds, binding to colloids, and kinetic hindrance of uptake
across membranes of large molecules. The point here is that we have no guarantee that
models which rely on Kow/Koc for estimators provide the same degree of reliability with
extreme hydrophobes as they do for chemicals which are more soluble. These estimators
are the building blocks of the bioaccumulation models we use, yet there is good reason to
question their utility above a Log value of 6.0.
3. Kow is a surrogate for estimating uptake and storage hi fish lipid. A greater effort should be
made to use laboratory and field data to corroborate the Kow measurement. Let the fish be the
octanol.
4. Focus research efforts on developing the approach for estimating TCDD transfer from sediment
to organisms, ie, sediment bioaccumulation factors (sediment/organism transfer coefficients).
A key to remember here is that the hydrophobicity of TCDD is so great that hi natural surface
waters there are sufficient solids and dissolved organics that the amount of TCDD in the free
phase is inconsequential. Sediment concentrations are not always predictors of organism tissue
levels, but when the tissue data are lipid normalized to sediment organic carbon levels, the
sediment concentrations for extreme hydrophobes provide an upper bound estimate of tissue
C-93
-------
sediment concentrations for extreme hydrophobes provide an upper bound estimate of tissue
concentrations. With this in mind one can design risk assessment approaches based on residue
effect levels.
9. Implication of food as a primary exposure route for TCDD:
• Exposure and uptake of freely dissolved TCDD from water as a route of exposure will be
minimal. With a Log P of >7.0 even small amounts of solids and organics will sorb
TCDD: consider the ratio: 10,000,000 to 1.0 (carbon normalized).
• When the primary route of exposure is food-based the focus of the conceptual model has to
shift to provide sufficient information to properly characterize stressors and effects. The
models to assess stressors are different for food exposure than water and are not as well
developed. Food-based residue effects are not well documented in the literature.
Therefore, the amount of data needed to be gathered to characterize stressors and effects will
be considerably greater than for a water based exposure assessment. Additionally,
laboratory and field data will be need to validate transport, uptake, and effects models.
Sediment transport and binding as well as food/prey availability and mobility will be key
considerations.
10. Fate and transport models for TCDD:
• My limited experience with fate and transport models suggest that there are adequate mod
els available for conducting reasonable risk assessments for most chemicals. The key
question I have is whether or not the models have been properly tested with chemicals with
extreme hydrophobicity. Most models are based on a large number of assumptions. These
assumptions should be challenged for chemicals like TCDD.
• The ability of transport models to predict the deposition of particulate bound TCDD will be
very limited if generalized transport models are used. There are hydrogeological models that
can be used, but a good deal of site specific information will be need for the models to
achieve a level of accuracy such that they are reliable for risk assessment purposes.
11. Major exposure issues not presented in the paper mill scenario:
• The conceptual model as it is presented in the scenario is fairly generic and as such is
broad enough to assess many ecosystem types. A detailed assessment of the given
river/reservoir would be more site specific in terms of the exposure characterization and the
populations to be evaluated. Looking beyond the current river/reservoir scenario one would
encounter many different issues in other types of ecosystems. For example, sea grasses and
plant life would be important in estuaries, anadromous salmon fisheries would be importa
nt in the large rivers of the North West and North East. Marine fishes which move into
estuaries to spawn would be important in many areas. Migrating waterfowl and other birds
could be an important part of an assessment for some areas. In the river/reservoir system
the reservoir provides a fairly good removal of TCDD due to deposition. In river systems
without reservoirs stressor characterization would be different and the transport models used
would have to characterize exposure over a much broader area. In estuaries the effects of
tides on the transport of the stressor would have to be characterized.
C-94
-------
Keith Cooper
Health Science Institute
Rutgers University
C-95
-------
-------
K. Cooper Premeeting Comments Exercise 2 & 3
Exercise 2. Stressor Characterization
8. The variable Kow reported for these compounds is due to
variation in temperature, water quality and the concentrations
that were tested. This area needs to be examined in much
greater detail and in a more systematic manner. The studies
should be designed to answer questions which are applicable to
real world situations. There needs to be a better
understanding of the factors which modify the concentrations
that are observed. Even though the scenario that are being
discussed is freshwater the studies should include estuarine
and oceanic salinities. There is a need to better understand
the physical properties of the organic material that these
compounds are associating with. There also needs to be
studies on the particle/particle interaction and the sorption
kinetics at varying concentrations. There also need to be
studies carried out to explain why aged sediments appear to
have less bioavailability than newer deposited sediment.
In the scenario of the pulp mill there would need to be
studies examining the specific process and the different types
of papers being made, as well as the different types of wood
being used. There would also need to be a size distribution
concerning the association of various compounds with different
size particles. The treatment plant for the mill would also
C-97
-------
K. Cooper Premeeting Comments Exercise 2 & 3
play a big role in how much if any of the fine suspended
sediments would reach the river. The pulp mill facility
should conduct studies to determine the bioavailability from
the samples collected from the mill.
In the conceptual model the value of most interest for these
compounds is not the Kow but the K^. This is especially true
in the fact that the paper mill treatment streams will have
very high dissolved organic carbon.
9. The major route of exposure in almost all cases is through
food consumption or through direct contact with either eggs
deposited on contaminated sediment or animals living in or on
the sediment. There needs to be information on the properties
determining the amount of the compound that is biologically
available (either from prey or from sediment) . The prey
species may also effect the amount of the compound that is
available to be absorbed. Studies need to clarify if there is
an effect of dose in a prey species and the percentage that is
absorbed.
In the scenario set out the major route for bioaccumulation is
through the food web for most of the aquatic animals. The
animals which live in or on the sediments may have a very
C-98
-------
K. Cooper Premeeting Comments Exercise 2 & 3
small amount taken up by direct contact, but this would be a
minor route. Filter feeding organisms can remove small
particulates from suspension, but just as with human lungs
there are specific size sorting which will make some animals
exposed to particles which may contain higher concentrations
of the dioxins. Some of the bottom feeding fish species take
in detritus and sediment at the water sediment layer-, which is
high in organic carbon. In the conceptual model there should
also be routes of elimination from the system such as burial
by cleaner sediments, photolyses and some reductive
dehalogenation or oxidative metabolism of the furans.
10. Although I am not a fate and transport modeler I have
dealt with several modelers. The difficulty with these models
is that in many instances they can give you a worse case
scenario in a completely mixed system, but if the system is
very complicated then the models begin to break down. In
order for the model to work there does need to be a large
amount of data on flow, areas and other physical parameters.
The models in the case for dioxin and compounds with similar
high KO,. could assume association with specific sized particles
and these could be modeled in the reservoir. The models could
predict the sedimentation rates above the damn in the scenario
with models that currently exist. Such a model could also
C-99
-------
K. Cooper Premeeting Comments Exercise 2 & 3
predict worse case scenarios in the 50 and 100 year floods.
The models once.they have made a prediction should be tested
to determine if the> model adequately represents the real world
situation (the model must be validated) . This type of
information was generated for Walters Lake, North Carolina for
the Carolina Power and Light which is down stream from a large
paper mill.
11. In the paper mill scenario you have a well defined
boundary and a large reservoir, while in an estuarine
environment this is a much more dynamic system. The effects
of tidal movements could result in depositional areas not seen
in the paper mill scenario. There could also be large areas
that are nursery areas for anadromous fish that are only
exposed for short periods of the year. Some fish species
following spawning over winter in the near shore areas where
these compounds could accumulate. Invertebrate species which
serve as major food sources could accumulate the materials.
The crustacea may be affected by these compounds because of
the juvenile development and molting. Often the estuaries
have large deltas or areas of deposition from upstream. The
requirement for periodic dredging to maintain navigation poses
the questions what to do with contaminated material. Storm
events along the coast often will stir up large areas of
bottom, which suspends more deeply buried contaminants. Many
C-100
-------
K. Cooper Premeeting Comments Exercise 2 & 3
of the estuarine environments have multiple contaminants and
sources for these compounds (atmospheric, chemical plants).
There is also concern for the large mammals which may consume
large quantities of contaminated fish for many years. There
is little or no information on the effects on these animals.
There is data on levels, but biological impact is unclear.
12. In my opinion there is very limited use for the BCFj
values generated for real world scenarios. In any system the
compounds will be associated with the suspended material and
not free in the water. The figure 3-1 makes this point very
well. In the case of the paper mill the average of 5% and a
range from 2-15% would result in a very small portion free in
the water If not zero. The best way to evaluate the potential
BAF is to do it on a site by site basis including in the
scenario food web and ecosystem characteristics. However, the
Lake Ontario study and the comparison with other fish and
areas demonstrates that there is fairly good agreement. The
question about why different animals appear to accumulate at
different levels needs to be better understood since it is not
explained by any of the current models for sediment/water
dissociation or the steady state based on lipid normalized
data. The steady-state biota/sediment accumulation factors
(BSAF) in both laboratory and field experiments are less than
C-101
-------
K. Cooper Premeeting Comments Exercise 2 & 3
1. The variation (Table 3-3) between fish and invertebrates
need to be examined further. The higher levels reported in
the mesocosm work (Rubinstein et al. 1983) may be
artificially high due to the exposure to contaminated
sediments at a constant concentration. There is a large
amount of data that has been generated by the paper mills in
the United States and other industries that can give numbers
for a number of. the species listed in the paper mill scenario.
In most cases there is not good data on sediment levels and
organic carbon content. In some situations it might be
important to know both the surface sediment and suspended
sediment concentrations to evaluate the BSAF.
13. The major concern that might be raised is the fact that
the amount of DOC and the matrix which the dioxins are
associated with may result in very different concentrations
dissolved in the water. Care should be exercised when using
the Lake Ontario TCDD BAPs because of different sources of the
material, which will effect its K^. and greatly alter the BAF.
14. The BMFs for most of the dioxins and dibenzofurans when
measured in the field are lower than what would be expected
from computer models. The reasons for these differences are
C-102
-------
K. Cooper Premeeting Comments Exercise 2•, & .3 ,, , .
not known. In many cases it is stated .that the'differences
are due to. metabolism but in very •• few cases has this been
demonstrated. The variability in the levels- found in-a lake
:area even when normalized to lipid can not .be .explained and
this is an important area for future research.. In the paper
mill scenario there is no indication for levels in crawfish or
other Crustacea which may serve as a source of exposure to
.certain fish species. In. many of these organisms there is a
need for examination of the individual species, in order to
better understand the BMFs. In most of the contaminated sites
the bottom feeding catfish and carp species are generally the
highest. There is a need.for experimentation to-understand
why this is the case. Similar types of studies need to be
carried out on avian species as well.
15.. The section 5.1.2 gives a fairly good summary of the
problems with ,.the uncertainties associated , with
bioaccumulatiqn factors. Even with these uncertainties there
can be a relatively good estimation for the Omigoshee
Reservoir biota. After the estimated levels are determined
than the field testing would either indicate that the model is
reasonably estimating the levels or that for some reason the
estimates are wrong. If the estimates are incorrect than the
site specific approach should be adopted to determine why the
estimates are not consistent with the real, world scenario.
C-103
-------
-------
Joseph DePinto
Department of Civil Engineering
State University of New York at Buffalo
C405
-------
-------
Comments on
"Interim Report on Data and Methods for Assessment of 2,3,7,8-
Tetratchlorodibenzo-p-dioxin Risks to Aquatic Life and Associated Wildlife"
[EPA/600/R-93-055]
by
Joseph V. DePinto, Director
Great Lakes Program
University at Buffalo
Buffalo, NY 14260-4400
ph: (716) 645-2088
Fax: (716) 645-3667
Exercise 2. Stressor Characterization
8. Uncertainty in K^ values
Uncertainty in K^. value used to assess fate, tranport, and exposure of
dioxin can have a major impact on the estimate of available dioxin (freely
dissolved) in a suspension. For example, at 10 mg/1 suspended solids a half-
order of magnitude error in 1^, (e.g., log K^ = 6.5 versus 7.0) will manifest
itself in a factor of three in the relative concentration of freely dissolved
chemical, while the effect on the fraction in the particulate phase will be
negligible. This uncertainty will lead to a large uncertainty in volatilization
fluxes and the rate at which the chemical builds up in the sediments.
Ideally, for the pulp mill scenario a system-specific K,,,, could be measured
along with some understanding of how spatial or temporal variability in
environmental conditions (e.g., sorbent type, characteristics, and concentration)
will affect this parameter. Realistically, the uncertainty in this parameter
inherent in using a non-site-specific value should be carried through the
exposure calculation by employing something like a Monte Carlo modeling
analysis similar to that conducted by Endicott, et al. (1992) in their Lake
Ontario report.
It should also be noted, from studies of other hydrophobic organic
chemicals (such as PCB congeners) in the Great Lakes, that in open lake water
we are very unlikely to measure K^. = K,,w. Studies such as Eadie's (1993) and
J.V. DePinto
C-107
-------
analysis of data from the Green Bay Mass Balance Study suggest that the slope
of a regression line for log K^, versus log K^ has a slope much less than one and
an intercept much larger than zero. Although the exact mechanism is not
known for sure, it has been suggested that sorbents in open lake water,
dominated by algae and algal detritus, behaves differently than soils and
tributary suspended sediments because of the different amount and character
of organic carbon in the solids. Also, some work (Swackhamer, 1993) has
suggested that sorption and desorption kinetics are more significant for algae
than for low /^ tributary solids and that this phenomenon is the cause of the
open lake observations.
Another observation with PCB congener-specific studies (Bierman, et al.
1992) is that binding to operationally-defined DOC is much weaker than to
POC. It seems that KOOC « 10'2 Kpoc for the Green Bay system. Does this
happen with dioxin? Or does dioxin behave more like mirex, which does not
show as big a difference (studies of Hassett at SUNY-CESF)? Again, this
uncertainty may have a very large impact on both exposure and effects
assessments, and it should be carried through all calculations until it is
resolved.
9. Implications of food chain exposure dominance
There are many sources of uncertainty in computing fish exposure
through the food chain, even assuming that the water column and sediment
bioavailable concentration is well known. First, these calculations assume that
the food chain is static (i.e., non-variable forcing function) and that each trophic
level component (like fish of a certain age class or zooplankton) is assumed to
behave according to average metabolic and growth and chemical assimilation
properties. We know that in actual systems, these parameters are highly
variable from site to site and exhibit large spatial and temporal distributions
within a given site. The effect of this variance in food chain bioaccumulation
parameterization remains to be adequately assessed.
J.V. DePinto
C-108
-------
A second question is the observation that the relationship between log
BAF and log K^ deviates (decrease in slope, less dependence on K^) from a
straight line for HOCs with log K,,w * 6.5, which places dioxin in this category.
Any computation of BAF on the basis of K^ should consider that effect.
Another observation, related to the K,,,. for algae in lakes, is that
computation of food chain bioaccumulation in fish is very sensitive to the
chemical partitioning to the base of the food chain (the algae). This value has
been shown to be spatially and temporally variable for PCB congeners in Green
Bay; therefore, its uncertainty in a given site-specific scenario such as the pulp
mill problem must be considered.
10. Availability of fate and transort models
GBTOX, which is the fate and transport mass balance model developed
for the Green Bay Mass Balance Study (GBMBS), is the current "state-of-the-
science" for modeling HOCs in surface waters. This model, which was developed
for PCB congeners would be quite suitable for TCDD, assuming the chemical-
specific properties of TCDD are known. It has been our experience that
accurate sorbent dynamics is the most crucial aspect of a toxic chemical fate and
transport calculation. In that regard GBTOX models both transport and
transformation of three organic carbon based state variables: biotic particulate
carbon (i.e., viable algae), particulate detrital carbon (all particulate matter that
is not viable algae), and "dissolved" organic carbon (passes a filter). Each of
these sorbents has different source, transport, transformation, and fate
pathways, each of which have important ramifications for organic contaminant
exposure pathways and fate. I would recommend that this type of model
become the "state-of-the-art" for aquatic risk assessment.
11. Major issues not covered in paper mill scenario
Following is a list of issues that I feel were not dealt with adequately in
the paper mill scenario:
J.V. DePinto
C-109
-------
• Depending on the hydraulic and, sediment-water interaction
characteristics of the system of interest, achieving steady-state after start
up may take many decades. In particular, building-up dioxin to steady-
state levels in the sediments will be the long-term process. In fact, parts
of the reservoir sediments may essentially never reach steady state.
• The description of sediment dynamics in the reservoir and the main river
channel is not described in nearly enough detail to be able to characterize
the dioxin exposure regime for this system, especially in the sediments.
No deposition and flow-driven resuspension in the river is unrealistic.
We need to specify how much of the reservoir SS is due to autochthonous
algal production.
• We cannot estimate exposure without knowledge of the hydrologic and
morphometric characteristics of the system: flows from each tributary
and plant discharge; hydraulic retention time; volume, mean depth, and
surface area of whole reservoir and each arm.
• We need wind data to estimate air-water exchange rate and we need solar
radiation data to estimate photolysis.
14. Biomagnjfication Pathway
The low lipid-normalized BMF. between lake trout and forage fish may
not only be the result of dioxin metabolism in the lake trout. It may be the
result of differences in chemical assimilation efficiencies for fish versus fish-
eating birds. This parameter is difficult to assess and a source of uncertainty
in exposure models. Also, one should ask the question of what age class of lake
trout were used to determine the value; there is approximately a three year lag
time between initial exposure and achieving some level of steady-state.
JT.V. DePinto
C-110
-------
15. Uncertainties in exposure prediction
Bioaccumulation in aquatic systems is driven by dissolved dioxin. Among
the major uncertainties in governing dissolved dioxin in Lake Ontario are solids
settling velocity, dioxin volatilization rate, and log K^.. There is no particular
reason to believe that this result will be particularly different for the case study.
Of course, there are a variety of uncertainties associated with computation of
BAF that should be added to the list.
J.V. DePinto
C-lll
-------
-------
Robert Huggett
Virginia Institute of Marine Science
College of William and Mary
C-113
-------
-------
R. Huggett
STRESS CHARACTERIZATION
Question 8; Probably the weakest links in the Interim
Report's treatment of the fate, transport and effects of TCDD
in aquatic systems, center around the inability to
analytically determine "dissolved" environmental
concentrations. Due to this limitation, dissolved levels are
derived by via calculations from media with higher
concentrations. Obviously errors are involved with each
calculation due to uncertainties associated with the
coefficients. To be conservative, the assessor may want to
take "worst case " scenarios. However, it should be kept in
mind that due to disequilibria, different types of sediment
organic carbon, different types of dissolved organic matter
and different suspended sediments, the actual concentrations
could be considerably different.
Question 9: The Interim Report's predicted routes of TCDD
exposure as being mainly through food is supported by its very
high Row. A major implication of this when assessing exposure
routes in the conceptual model is that to have accurate
estimates, one must be able to accurately predict, "who eats
what." Many warm and cold water species have terrestrial
organisms as a major part of their diets. As well, many
species shift their diet as various foods become more
abundant. Again, "worst case" scenarios can be used but with
the same caveats as given in Question 8.
Question 10; Both parts "a" and "b" appear to be essentially
the same for very hydrophobia chemicals. The crucial question
centers on how good the models predict deposition and
resuspension of very low density (eg. high organic carbon)
participates and at what temporal and spacial scales.
Flocculation, bioturbation, fecal pellet deposition and
C-115
-------
physical disturbances (eg. periodic scouring as current
velocities change and/or wave action) are all site specific
and most mathematical models do not handle them well.
Question 11; In estuaries/ a TCDD risk assessment would have
to consider the Turbidity Maximum", if one occurs, as well as
the fact that many of its fish and crustaceans are migratory,
thus spending only part of the year in the system.
In terrestrial systems, one must consider many of the
same issues raised in Question 9. The fact that many
terrestrial mammals and birds are omnivorous and migratory,
complicates the situation even more.
Question 12; As previously mentioned, the inability to
analytically quantify TCDD in solution is a serious limitation
in the paper mill scenario (see comments for Question 8). As
discussed in Question 9, diets will not only be different in
different locations but also they will change with time.
These points need to be kept in mind.
Question 13; This question appears to be redundant with parts
of questions 8 & 12 and those answers are appropriate here.
It is apparent that research is needed to lower the detection
limit for dissolved TCDD.
Question 14; The biomagnification pathway in the Stressor
Characterization and the Conceptual Model appear realistic.
Question 15; The Interim Report clearly states that there are
uncertainties associated with TCDD bioaccumulation factors
(BCF's). Steady state BCF's, derived in the laboratory vary
over an order of magnitude. What would the variance be in the
environment where steady state is probably rare? The reasons
for non-steady state have been previously given. Therefore,
residue predictions should be considered tentative at this
time.
C-116
-------
Charles Menzie
Menzie-Gura & Associates, Inc.
C-117
-------
-------
PREMEETING COMMENTS
Charles A. Menzie
Workshop on Ecological Risk Assessment Issues For TCDD
Comments on Exercise 2. Stressor Characterization
8. Uncertainties associated with estimates of Kow and Koc may
lead to uncertain predictions of TCDD partitioning and
exposure. How should this be handled in stressor
characterization and conceptual model development for this
scenario?
Comment
The implications of the uncertainties associated with Kow
and Koc estimates will depend upon the method employed to
represent the exposure field. I suggest that the
distribution of TCDD and related compounds in the reservoir
be characterized by a suspended particulate/sediment
transport model which estimates TCDD distribution based on
estimated suspendd solids and particulate organic carbon
concentrations, and movement and settling of particulates
within the system. Regardless of the uncertainty associated
with the Kow and Koc estimates, most of the TCDD will track
with the particulates and will be distributed in accordance
with sedimentation patterns. Because of the uncertainty
associated with the forms in which TCDD may be present
within the water column (with POC & DOC) and the uncertainty
associated with the Kow and Koc estimates, it may be most
useful to relate exposure to fish to resultant sediment
levels of TCDD.
If models based on Kow and Koc are employed to estimate
exposure, it would be useful to conduct a sensitivity
analysis. It appears the best estimate of the log of these
values is approximately 7.0. Alternative values that might
be considered include 6.0 and 8.0. A sensitivity analysis
would reveal the possible importance of these variations on
the estimate of exposure. As an enhancement, it may be
possible to derive site-specific partitioning and uptake
coefficients using a combination of chemical analytical and
biological methods. This would require obtaining a
representative sample of effluent either from a similar
plant or from a pilot plant.
9. Address the implications that most TCDD exposures will arise
from food consumption and contact with sediments.
Comment
I think the Interim Report makes a good case for the
importance of these exposure routes. The major implication
is that exposure to fish should be related to either
sediment and/or food concentrations of TCDD and related
C-119
-------
compounds. As indicated in Section 3.1 of the Interim
Report, since the BSAF and BSSAF do not vary significantly
with Kow, the great uncertainty existing for the Kow of TCDD
is not incorporated into these bioaccumulation factors.
Limited available information presented in the Interim
Report suggests that the BSAF for warmwater fish species
varies over a relatively small range given all the possible
sources of uncertainty in making estimates. This suggests,
that for predictive purposes, the assessment should be based
on a model that yields an exposure field for sediments. It
may be more difficult to predict the levels of TCDD in prey
organisms and thus difficult to estimate exposure to fish
based on accumulations in their prey.
10. Comment on the availability of fate and transport
data/models suitable for use with TCDD. Discuss the
applicability of available transport models for predicting
the deposition of particulate-bound TCDD in the reservoir.
Comment
This is not my area of expertise. However, recognizing that
all models are wrong but that some may be helpful, I think
it would make sense to consider a range of models from very
simple partitioning or box models to more complex models in
the following order: 1) simple box model for reservoir as a
whole with TCDD partioned between suspended sediments,
water, sediment, and biota under steady state conditions
with a constant source, losses from the system through
advection, and possibly losses through
sedimentation/bioturbation; this could be accomplished with
available simple models including the Level III Fugacity
Model; 2) box model that treats the reservoir as above but
recognizes the different regions of the reservoir; these
regions include the trunk hear the proposed mill, the two
arms into which two other streams discharge, and the broad
basin near the dam; and 3) finite element models such as
Toxiwasp which divide the system into many discrete cells
and track mass flux of suspended sediment and associated
TCDD et al. between them. Such a model would need to
incorporate sedimentation so that exposure concentrations in
the surface sediments could be estimated for the various
cells. Mixing of sediments through bioturbation and
sedimentation would need to be taken into account in order
to derive a surface sediment concentration. Such models are
available bur require estimates of bioturbation and
sedimentation rates. If this is an important issue, it may
be possible to estimate these processes within the reservoir
using a suitable tracer (e.g., lead 210).
It is my understanding that models do exist for estimating
the movement and deposition of particulates in lakes and
reservoirs. These models are applicable to estimating the
distribution of particulate-bound TCDD. A key consideration
in all these models is the degree of sophistication needed
C-120
-------
to reach an "acceptable answer" for risk assessment and
associated risk management purposes. The more sophisticated
models require more information about the system. If an
acceptable answer can be reached using simple models this is
superior to gathering a great deal more information for use
in more sophisticated models.
There are several methods available for assessing transport
of water and particulates. Field measurements should be
considered for assessing the factors that affect these
processes within the system. The kinds of information
sought include: contribution of the three tributaries to the
solids loads to the reservoir, importance of events (e.g.,
storms, spring floods) with regard to loadings and
resuspension, variations in freshwater flows to the
reservoir, circulation and residence times within the
various arms of the reservoir. It would be helpful if there
was a tracer that could be used to estimate the contribution
of solids for the tributary on which the mill is located; it
is unlikely that this is the case unless it is draining a
region that is geologically different from the other
tributaries. Measurements should also be obtained on
particulate formation (primary production) and deposition
within the system. A sediment survey of the reservoir would
provide useful information on deposition patterns and the
overall characteristics of the sediments.
11. List some of the major exposure issues not present in the
paper mill scenario that may be encountered in future
ecological risk assessments.
Comment
There are two major factors to be considered in future
ecological risk assessments regardless of the system:
1. the physical characteristics of the system will
dictate the manner in which the chemical is
distributed and the nature of exposure; these
physical characteristics are unique to different
kinds of aquatic systems (e.g. large lakes vs
small lakes and reservoirs, estuaries vs rivers,
estuary and coastal systems vs offshore areas,
shallow basins vs deep waters, wetlands vs
forrests)
2. the ecology of the ecological receptors will
dictate how they might be exposed; in the present
case study we are examining warm water fish
species that tend to inhabit shallow waters and
nest on or in the sediments; the species all are
typified by limited home ranges during much of the
year; in contrast, anadromous fish species,
species that have planktonic eggs and larvae, or
migratory birds would experience a very different
C-121
-------
exposure regime.
12. Discuss the applicability of bioconcentration,
bioaccumulation, biomagnification, and biota-sediment
accumulation factors to stressor characterization for the
paper mill scenario.
Comment
These are all processes that are recognized to exist.
However, there is considerable uncertainty associated with
estiamting the relative importance of the processes. I
think it would be most useful to identify the most
straightforward method for relating exposure to body burdens
for the fish. Based on my review of the Interim Report,
this would probably involve the use of a range of BSAF
values applied to estimated sediment concentrations. This
is appropriate for this case study because of the ecology of
the target fish species. These species tend to have limited
home ranges and also tend to be closely associated with
sediments both in terms of where they occur in the water
column and also with regard to their food.
13. Comment on applicability of Lake Ontario BAF to the paper
mill stressor characterization.
Comment
The big "if" appears to be if C can be estimated
accurately. I do not know if this is possible and it seems
somewhat hypothetical at this time. I do know that sediment
contact and ingestion of prey items associated with the
sediments may be important and may not be captured by a C
derived for a physically different system, one in which the
water volume to sediment area is probably much greater than
in the reservoir.,
14. Comment on the biomagnification pathway.
Comment
The biomagnification pathway would be important to consider
if the effects information for birds (and mammals) were
related to contaminant burdens in specific target organs.
Based on the Interim Report, data are not adequate for this
purpose and the effects data are based on administered dose.
Thus, at present, the issue of biomagnification pathway is
moot. As data are developed on relationships between body
burdens and effects for birds and mammals it would be
important to consider the role of biomagnification.
However, when that time comes it might be most useful to
relate directly body burdens to dietary doses rather than to
residues in the wildlife. I suggest this because the
information that would be generated in future risk
assessments would initially be body burdens in the fish.
Thus, ultimately, we will be relating effects on wildlife to
body burdens in fish. Focusing on biomagnification does not
C-122
-------
appear to help in this regard. Such information would be
useful as part of developing an effects data base that could
be used to support epidemiological studie.
15. Discuss the relevance of uncertainties in bioaccumulation
factors to prediction of TCDD residues.
Comment
In the face of uncertainty,. I think it would be prudent to
conduct a sensitivity analysis using the ranges and best
estimates on bioaccumulation factors. Because the agency is
in the position of reaching decisions that should be
protective of the environment, I assume that greater weight
would be given to those estimates based on higher BAFs. If
the results of such an analyses were either clearly good or
bad (depending on your point of view) there should be little
argument about the results. However, where results are
equivocal, this should indicate where additional information
could be gathered to reduce the uncertainty in the
estimates.
Consideration could be given to the development of these
factors in laboratory bioassays or mesocosms in which fish
are exposed to representative effluent that has been allowed
to partition among water, sediments, and suspended
sediments.
C-123
-------
-------
Derek Muir
Department of Fisheries and Oceans
C-125
-------
-------
PRE-MEETING COMMENTS FOR THE PEER PANEL WORKSHOP ON 2,3,7,8-TCDD
Comments on Exercise 2. Stressor characterization
8. Problems associated with TCDD phys/chem properties and
measurements of partitioning: There is definitely considerable
uncertainity in the estimates of WS and Kow of TCDD and other
2,3,7,8-substituted congeners. This uncertainty makes application
of chemical fate and food chain models to the pulp and paper
scenario quite difficult unless a lot of site specific measurements
are obtained.
1. Water solubility. Although there are a wide range of WS values
for TCDD, .values in the range, of 200-480 ng/L (Lodge 1989) are the
most consistent with results reported for other PCDD and PCDF
congeners (Friesen et al. 1990; Shiu et al. 1988), Most of the
latter results were obtained with a generator column which is the
best available technique for WS determination of hydrophobia
compounds. The results of Marples et al. and Adams and Elaine did
not use the generator column methodology and therefore should be
used with caution. For the paper milI/reservoir scenario, WS
determined at 20°C would seem most appropriate because water
temperatures are presumably in the 20-30°c range at this site much
of the year. Water temperature information would be required for
the conceptual model along with other basic water chemistry
parameters (TSS, DOC, POC).
2. Octanol-water partition coefficient. The results of Sijm et al.
and Marples et al. are probably as close as we will get to direct
measurement of KQW values for TCDD. All direct methods (including
generator columns with octanol phases) appear to be prone to
emulsion formation. Perhaps an approach involving different ratios
of octanol to water (both presaturated), similar to Lodge's for Koc
determination, would address the emulsion issue. I agree with the
interim report (p. 2-3) that WS is more reliably measured than Kow.
Log Kow can be calulated from Yalkowsky et al.'s relationship. It
is interesting that using a WS of 200 ng/L (0.625 xlO"9 mol/L) the
original estimate of TCDD solubility (Crummett and Stehl, 1975),
gives a log Kow of 6.9 using Equation 2.2 (p.2-3) which is close to
most measured Kows and to the value of 7 (from Burkhard and Kuehl)
used in the interim report.
3. Koc values: Koc is a key parameter in defining the distribution
and exposure concentrations of hydrophobia contaminants. As noted
in the interim report there are uncertainties in Koc values for
TCDD due to difficulties in measuring dissolved concentrations.
Broman's data are probably the best field measurements but his
"dissolved" phase includes a colloidal DOC fraction. The method
used by Lodge and Cook (1989) is perhaps the most elegant approach
to the problem of a non-settling.phase in Koc measurements i.e.
extrapolation to zero solids. Although sparging techniques
(Resendes et al. 1992) offer some promise of direct measurement of
freely dissolved concentrations they are also prone to
interferences from stripping of bacteria and other suspended
particles from the water column by bubble action (Friesen et al.
1993 in press). Setting Koc=Kow=l07 for 2,3,7,8-TCDD and other
2,3,7,8-substituted congeners seems appropriate for generic risk
modelling purposes.
C-127
-------
In the case of the pulp and paper scenario consideration
should be given to actual site specific measurements of Koc of TCDD
using suspended solids from mill effluent, river water and bed
sediments, to confirm the above assumption. Both field and lab
based procedures could be used. Prior to startup of the proposed
mill Koc values could be determined in the laboratory with sediment
and water obtained from the depositional zone in the reservoir,
using the method of Lodge and Cook. During operation of the mill,
field monitoring using methods such as in situ sparging (Resendes
et al. 1992) and high volume sampling/filtration (Broman et al.)
could be employed, with HRGC-MS analysis to determine TCDD in
"dissolved" and particulate phases.
Chemical fate models such as WASP and EXAMS also address pore
water diffusion which may be a significant source or sink of
hydrophobia organics in the water column. For buried TCDD diffusion
may be the only process of movement. The effective diffusivity Deff
can be calculated if the molecular diffusivity in water, Koc and
sediment porosity are available (Formica et al. 1988).
9. Exposure routes In the conceptual model; There is good evidence
from field and laboratory studies that TCDD exposure of fish arises
mainly from food consumption. This implies that the conceptual
model needs to include detailed consideration of the predator-prey
relationships in the river and the reservoir. The schematics (Fig.
3 and 4) representing the Thomann model with links to fish-eating
birds and mammals are an appropriate first step. However to make
accurate predictions of concentrations in each species will require
knowledge of what species are present, their prey, their age
distribution, growth rates, feeding rates, and pharmacokinetic data
for TCDD uptake and depuration. For modelling over an entire year
consideration has to be given to temporal changes in density of
organisms, especially benthic organisms, and resulting diet shifts.
Benthic insects may form a major part of the diet of bottom
feeding and pelagic fish for brief periods in the spring as they
emerge as adults. Emerging insects have been shown to transfer TCDF
to the water column and surrounding terrestrial environment
(Fairchild et al. 1992). In riverine environments, filter feeding
insects may trap suspended solids originating from secondary
treatment ponds of pulp mills, which they subsequently transfer to
fish (Birkholtz et al. 1992). The stable isotope technique used by
Broman et al. to characterize the predator-prey relationships in
the Baltic pelagic food chain is a useful technique to apply to any
detailed site specific study of the proposed pulp mill.
1O. Fate and transport models. Available models include WASP and
EXAMS (USEPA) and RIVER/FISH (from NCASI;Hinton 1991). These three
models have been used to predict fate and distribution of TCDD and
related hydrophobia in lakes and rivers. RIVER/FISH is less
sophisticated in that its current version does not include
sedimentation - it takes a dilution partitioning approach to
estimate exposure concentrations in the water column. Nevertheless
the model has been successful in predicting TCDD levels in fish
near pulp mills. WASP was used by Endicott et al. (1990) to predict
TCDD fate in Lake Ontario. It has also been used along with a
colorful graphics output to predict TCDD/F concentrations in
C-128
-------
sediments of Howe Sound (BC)(Hoiloran 1993). WASP4 would definitely
be the most appropriate model for initial chemical fate modelling
the river-reservior pulp mill scenario because of its flexibility.
Several Canadian modelling efforts for TCDD near pulp mills have
concluded that more sophisticated hydrodynamic and sediment
transport models, than are available in WASP4 should used for large
river systems (Marmorek et al. 1992). The model MOBED (Krishnappan
1981) was recommended for sediment transport in the Eraser River.
12. Applicability of BCF, BAF, BSAF to paper mill scenario:
BCF^: BCFs are strictly lab-based measurements because they involve
exposure via water only. For chemicals with long equilibration
times BCFs must be estimated from the ratio of first order uptake
(jq) and elimination (kg) rate constants. These constants can be used
in food chain accumulation models.
As noted on p. 3-10 there is a wide range of BCF^'s reported.
The large variation is due.to difficulties in measuring freely
available TCDD in the water and to underestimation of k,. Given
this variation it is impossible to recommend a single BCF value to
use in the pulp and paper scenario. But kg' s for TCDD are known
with greater certainty because they are less influenced by initial
exposure conditions. Fish size and lipid content effect kg values
although precise allometric relationships have not been reported
for TCDD or other hydrophobia organics. A range of kg values could
be recommended for food chain modelling of TCDD and related PCDD/Fs
by selecting appropriate values based on organism size and lipid
content. It would be difficult to recommend a range of kt's for
TCDD from the existing literature.
BAF'j: These values can be calculated from field measurements of
TCDD in biota if TCDD concentrations in water can be estimated. The
EPA has done this in initial assessments of, TCDD from pulp mills
(US EPA 1990)and Muir et al. (1992a) have calculated BAF^'s for
fish near Canadian bleached kraft mills. But as with BCF values,
BAFs are dependent on water concentrations (i.e. Cdw or C^) for
which few measurements are available. BAFs are also dependent on
food chain relationships. Lipid normalized BAFs for TCDD in
mountain whitefish/ and in white and longnose suckers, sampled near
pulp mills in Alberta and BC differed consistently by 3 to 4-fold.
This has been explained by feeding differences (Birkholtz et al.
1992; Muir et al. 1992a). The whitefish feed on emerging insects
and filter-feeding insects in the water column while suckers are
bottom feeders. The filter-feeding insects may accumulate high
concentrations of TCDD/F by trapping biosolids from the secondary
treatment ponds of the mills which are subsequently accumulated by
whitefish.
BSAFs: Unlike BCFs and BAFs, BSAFs can generally be calculated for
field and lab studies because concentrations in sediment and biota
are usually measureable. But there are problems of selecting and
sampling the representative sediment particles that benthic
organisms may be feeding on. BSAFs may also vary from site to site
because of disequilibrium between sediment and water - this effect
is especially important in laboratory and field microcosms (Servos
et al. 1992) that are not at steady state and in lakes, such as
Lake Ontario, where past TCDD emissions are buried and not
exchanged readily with the water column. Despite these problems
C-129
-------
BSAFs are perhaps the most consistent and well documented
bioaccumulation parameter available for TCDD. Cook et al. (1990)
found less than a factor of 2 variation in BASFs for pelagic fish
in Lake Ontario. BASFs have also been reported for bottom feeding
marine organisms (Harding and Pomeroy 1990), oligochaetes
(Rubinstein et al. 1990) and for fish near pulp mills (Muir et al.
1992). BSAFs for fish near 17 Canadian pulp mills were generally
greater than those reported by Cook et al. (1990) (geometric mean
0.47, range, 0.14-0.96) but were nevertheless less variable than
BAFs or BCFs calculated for the same sites. Site specific
differences in food chains (e.g. whitefish vs suckers) and
sediments (rivers vs lakes) appear to account for much of the
variation in BSAFs. If some of these differences could be
documented it should be possible to select BSAFs for various
species in the paper mill/reservoir scenario and to derive tissue
concentrations from TCDD concentrations estimated for sediments.
13. Application of the Lake Ontario BAF^t: In principle this BAF
(i.e. 1.9 x 106) could be widely applied"to estimate tissue
concentrations of TCDD assuming that POC and DOC are known because
Cdw could be calculated (i.e. equation 3-6 or Fig. 3-1)from
estimates of C\. But Lake Ontario, an oligotrophic, cold water
lake, may not be the most representative environment from which to
derive BAFs for a mesotrophic, warm water, southern reservoir. Food
chain relationships, species differences in age and lipid, levels
of suspended solids and bed sediment characteristics (10% vs 2% OC)
may simply be too different. Among the species studied by Cook et
al. in Lake Ontario, white perch, had 3-fold higher BSAFs (and BAFs
also) than lake trout because of age differences. Prediction of
bioaccumulation using BSAFs or BSSAFs or food chain models offers a
more robust approach because Cdw does not have to be estimated.
14. Significance of bioma.gnifica.tion pathway: BMFs for TCDD in fish
and their food measured in the field and in the lab are usually
between 1 and 2; far less than for PCBs of similar Kow. As noted in
Sect. 3.4 this is probably due to biotransformation of TCDD by
fish. But invertebrates do not seem to be able to biotransform
PCDD/Fs - hence the wide range of PCDD/Fs (including non-2,3,7,8-
substituted congeners) found in crab hepatopancreas (Harding and
Pomeroy; Norstrom et al. 1991) and filter feeding caddisflies
(Birkholtz et al. 1992). Thus BMFs at the lower end of the food
chain may be >1 although data to support this are limited. If this
is correct then information would have to be available on
bioaccumulation by invertebrates to successfully implement a food
chain bioaccumulation model, as part of the risk assessment for the
pulp mill scenario. This information would could either be kt and k2
values (rare for invertebrates) or BSAFs derived from field and lab
studies.
25. Etocertainities in BAFs: In the paper mill/reservoir scenario,
TCDD concentrations in dissolved and suspended particulate phases
will be calculated using chemical transport/fate models. The models
will presumably give concentrations in surficial sediments assuming
that sedimentation and resuspension rates are known. BAFs and/or
BSAFs, or a food chain model (Fig. 4) could then be employed to
C-130
-------
calculated tissue concentrations. As noted on p. 5-3 there are
uncertainitd.es in BAFs because Cdw and C\ are based on estimates and
are not verified by field studies (except by Broman et al.). Even
if Cdw could be measured in the Omigoshee Reservoir using large
volume samples there would still be debate about what was truly
dissolved because there are problems with all available techniques
(i.e. filtration, sparging). BSAFs also have uncertainities due to
site specific differences in food chains, sediment-water
disequilibrium etc. But BSAFs have the advantage of being
verifiable with existing techniques via sampling and analysis of
biota and sediments. Using BSAFs to estimate tissue concentrations
in lower food chain organisms combined with food chain modelling
(application of pharmacokinetic parameters, growth rates, age and
feeding preferences) to estimate concentrations in forage fish and
piscivorous fish is probably the best approach for the present
scenario.
C-131
-------
-------
Thomas O'Connor
National Oceanic and Atmospheric Administration
C-133
-------
-------
RESPONSE TO EXERCISE 2. Stressor Characterization (questions 8-15)
8. In principle, the TCDD concentration in fish can be calculated by assuming
equilibrium among particles, water, and fish. In fact, that is an extreme assumption
and even having certain values for K^, and K^ would not make the predicted body-
burdens less tenuous. I would assume that all the TCDD from the plant adsorbs
onto particles passing the plant and use empirically defined ratios of body-burdens
to TCDD on particles to predict the body-burdens. The need is for more empirical
observations than for better physical-chemical constants.
If I had to predict body burdens with only the information in hand I would calculate
the steady-state concentration of TCDD on suspended particles and, using all the
attendant adjustments for lipid and TOC, apply the BSAF of 0.3 from the Interim
Report. This is the upper end of the BSAF range and using it errs on the side of
overestimating body-burdens. There are no empirical data for BSSAF
(bioaccumulation relative to suspended solids) so I would be letting the BSAF
values substitute for them. It would not be valid to calculate body-burdens relative
to deposited solids because the TCDD concentration on the floor of the reservoir
will not reach steady state until the upper mixed layer consists entirely of particles
that entered the system after the paper mill began operating. (Prior to that settled
particles will be mixing with and be diluted by preexisting particles)
9. The conceptual model has all the arrows for transfer of TCDD, but the great
advantage of the equilibrium assumption is that pathways are irrelevant. Once the
thermodynamic activity in one phase of the system is known, it is known
everywhere else. This assumption has not been demonstrated to hold but I would
still be more confident relating TCDD body-burdens in fish to TCDD on suspended
solids than I would be trying to define all the partitioning and rate constants
needed to apply the conceptual model.
10. At steady state the flux of dicxin into the system equals the flux out. The flux
out is the discharge of TCDD-laden suspended solids through the dam plus the
flux of TCDD-laden particles to the floor of the reservoir. This latter efflux does not
really remove dioxin. At first it causes exponentially increasing concentration of
dioxin on the sediment bed and eventually that concentration becomes constant
and the deposited layer just becomes thicker. The rate of change of TCDD on the
floor of the reservoir depend on rates of deposition and particle mixing by
bioturbation. Steady-state for TCDD concentration will not be achieved for years.
On the other hand and ignoring resuspension of deposited particles, the TCDD
concentration on suspended particles can be modeled by assuming that all the
particles passing the plant achieve the same TCDD concentration (per g of organic
carbon) and that in toto that is all the TCDD. In terms of a mass balance the
aqueous and biotic TCDD can be ignored. So downstream of the plant the
suspended solid TCDD is known immediately. Now we need to know how much
settling occurs before the TCDD-particles mix with and are diluted by particles from
C-135
-------
other areas. The simple calculation assumes no particle settling upstream of the
reservoir and uniform settling in the reservoir. That leads quickly to a steady-state
TCDD concentration on suspended particles as simply the suspended solid TCDD
concentration at the plant (i.e. mass of dioxin per second/mass of suspended
solids per second) times the fraction of all suspended particles in the reservoir that
come from the river with the paper mill.
The simple assumption is probably good enough because I see no alternative to
assuming that all the fish in the reservoir are uniformly exposed to TCDD. If what
they really have to do is swim up near the plant to get their TCDD dose, the mean
exposure of the mean fish is still going to be what is calculated by assuming a
uniform distribution of suspended TCDD
11. The largest difference between the reservoir scenario and coastal or estuarine
scenarios is that entire populations of organisms will not be exposed to TCDD. In
the reservoir, each individual fish experiences the same TCDD exposure. If there is
a reproductive loss associated with that exposure the effect on the population, in
subsequent years, is calculable given huge assumptions about compensatory
mechanisms. In open systems, all individuals are not exposed and the population
effects of TCDD will diminish in proportion to exposed versus unexposed
individuals
12.1 think I have done this. My obvious penchant for empirical information has led
me to use BSAFs. Their use requires fewer assumptions than alternative methods
for guessing TCDD body-burdens.
13. Use of BAFs assumes equilibrium which is one big assumption but, to
compound this excursion into the nether world, Cwd is strictly a calculated number
never blemished by empirical verification. The BAFs in the Interim Report were
calculated on the basis of measured body-burdens, choices of K^ and K^ and
assumptions of equilibrium. The BASFs in the report have the advantage of at
least being based on measured concentrations in fish and in surface sediment.
14. The conceptual model seems to assume equilibrium among all aqueous
phases and therefore excludes biomagnification between trophic levels. Birds are
not exposed to the same abiotic phases as fish so there is no basis for assuming
equilibrium and lipid-adjusted body-burdens in birds cannot be assumed equal to
their counterparts in fish Calculation of their body burdens requires kinetic data.
Even without rate constants, however, empirical data on body-burdens in birds and
in fish they eat are invaluable.
15. The uncertain assumptions and constants required to convert this proposed
TCDD discharge to TCDD body-burdens are too large to make a worthwhile
prediction. The empirical BSAFs for Lake Ontario can be used in a pinch but it
would be much more preferable to spend the resources needed to get more
empirical relationships on TCDD concentrations in fish, suspended sediment, and
deposited sediment in the vicinity of paper mills.
C-136
-------
Robert Pastorok
PTI Environmental Services
C-137
-------
-------
PREWORKSHOP COMMENTS
Workshop on Ecological Risk Assessment Issues
for 2,3,7,8-Tetrachlorodibenzo-y~Dioxin (TCDD)
Comments by: Robert A. Pastorok
PTI Environmental Services
15375 SE 30th Place
Suite 250
Bellevue, Washington 98007
INTRODUCTION
These technical comments address issues on the assessment of ecological risks associated
with chlorinated dioxins and related compounds. The comments were developed for use
at the U.S. Environmental Protection Agency's Workshop on Ecological Risk Assessment
Issues for 2,3,7,8^etrachlorodibenzo-^ioxin(2,3,7,8-TCDD) to beheld in Minneapolis,
Minnesota, on September 14-15, 1993. The comments address the issues of stressor
characterization and development of a conceptual model described hi materials distributed
prior to the workshop. The issues were based on a hypothetical case study of a pulp mill
wastewater discharge into a river that enters the Omigoshee Reservoir. The stated risk
management goal is to develop final permit conditions and treatment standards that will
maintain chlorinated dioxins and related compounds in discharges at concentrations below
those expected to have detrimental effects on fish and wildlife of the reservoir. In
developing these comments, Figure 1 was prepared to provide an overall framework for
achieving the risk management goal for the Omigoshee Reservoir scenario.
Throughout these comments, the term PCDDs will be used to refer to 2,3,7,8-
tetrachlorodibenzo-p-dioxui and its congeners. The term PCDFs will be used to refer to
2,3,7,8-tetrachlorodibenzofuran and its congeners. Other abbreviations for terms used
in risk assessment models are defined in the preworkshop materials or hi the Interim
C-139
-------
Report on Data and Methods for Assessment of2,3,7,8-Tetrachlorodibenzo-p-dioxin Risks
to Aquatic Life and Associated Wildlife (U.S. EPA 1993).
In the following discussions, each comment is preceded by the statement of the issue
shown in italics. Issues are numbered as in the preworkshop materials. Because issues
concerning ecological effects and endpoint selection are not addressed herein, the
numbering begins with Issue 8.
STRESSOR CHARACTERIZATION ISSUES
Exposure Issues
Issue 8: The Interim Report (Sections 2.1 and 2.2) indicates that there is considerable
uncertainty in the estimates of parameters including Kow and K^ and the partitioning of
TCDD onto organic matter. This uncertainty results in part from difficulties in analytical
measurements of various fractions of TCDD in water. Since these limitations may affect
predictions of TCDD partitioning and exposure, please address how they should be
handled in stressor characterization and conceptual model development for this scenario.
Uncertainties in exposure assessment should be addressed through quantitative analysis
of the sources of variability and error in models and then* input parameters. The
following procedures are recommended for uncertainty analysis:
• Use of a distributional approach (e.g., Monte Carlo) if possible
• Use of an average exposure case and a plausible maximum exposure case
as an alternative to the distributional approach if the final risk models are
deterministic
C-140
-------
• Evaluation of sources of uncertainty to determine data collection and
research needs (e.g., collect data for those variables important in the risk
model when the high uncertainty can be substantially reduced by ob-
taining new data)
• Replicate measurements of key input variables, with simultaneous
quantification of field and analytical variability.
The use of a distributional approach to uncertainty analysis (e.g., Monte Carlo analysis)
in deriving sediment criteria and effluent discharge limits represents a state-of-the-art
approach. Difficulties will likely be encountered hi assigning probability distributions
for all key variables in a risk model. Nevertheless, this approach is preferable to past
approaches that rely on "safety factors" or conservative point estimates for each variable
in the risk model. For example, distributional analysis should be used in estimating
NOAELs hi place of the "safety factor" approach. Use of point estimates involves more
policy decisions during risk calculations than use of a distributional analysis approach.
The choice of a mean and shape of the distribution for variables with high uncertainty
may involve some assumptions, but no more so than the choice of conservative point
estimates. For those variables where the uncertainty is high, sensitivity analysis may be
used to evaluate the effect of the form of the distribution on model outcomes.
Issue 9: The Interim Report (Section 2.4) indicates that most TODD exposures mil arise
from food consumption and contact with sediments or suspended solids, with the water
pathway being less important. Address the implications of this information relative to the
exposure routes in the conceptual model.
The conceptual model for exposure pathways (Figures 3 and 4 of the preworkshop
materials) is relatively complete; however, there are some omissions as well as some
inconsistencies. For example:
C-141
-------
• Although the interim report acknowledges the importance of sediment
ingestion as an exposure pathway for some fish species, this pathway is
not included hi the conceptual model. Sediment ingestion should also be
included for some predatory mammals and birds.
• The conceptual model currently does not address mammals that feed on
freshwater mussels. This could be a serious omission for some reservoir
systems.
• The role of amphibians as food for predatory fish and birds should be
considered.
• Each of the wildlife receptors may receive a portion of their diet from
outside the Omigoshee Reservoir. The "other diet" pathway is indicated
for only mammals in the proposed conceptual model.
Finally, the graphics for the model (Figures 2 and 3 of the preworkshop materials) could
be improved by using bolder arrows to indicate the major exposure pathways of food
ingestion, sediment contact, and sediment ingestion.
The development of a food web exposure model for this risk assessment could benefit
from a more detailed consideration of species dietary patterns and life histories. An
example of a procedure to guide development of the trophic model is shown hi Figure
2. It should be emphasized that "trophic species" need to be considered, not taxonomic
species.
Issue 10: Fate and transport models are beyond the scope of the Interim Report, but are
dearly critical for risk assessment. In stressor characterization and the conceptual
model, they will be necessary for linking TCDD source loads to concentrations in
different compartments of the reservoir.
C-142
-------
• Comment on the availability of fate and transport data/models suitable
for use with TCDD.
• Discuss the applicability of available transport models for predicting the
deposition of paniculate-bound TCDD in the reservoir.
Fate and transport models for chlorinated dioxins and related compounds have been
reviewed by U.S. EPA (1992). Based on their review and work by the U.S. Army Corp
of Engineers for modeling suspended solids and sediments in reservoirs, available models
should be adequate for linking the source load to compartments hi the reservoir system.
Issue 11: List some of the major exposure issues not present in the paper mill scenario
that maybe encountered in future ecological risk assessments (e.g., marine/estuarine,
terrestrial, etc.).
Major issues that may be encountered in other risk assessment scenarios include:
• Model assumptions or algorithms needed to address sources other than
the one for which a permit is being developed. For example, regional
background concentrations, cumulative impacts, and wasteload allocation
may need to be addressed. Also, different sources of PCDDs and
PCDFs will result hi different mixtures of congeners, which may have
slightly different requirements for model development.
• Effects of PCDDs and PCDFs on other species groups for which few
data are available. For example, other assessments may need to address
marine mammals (marine/estuarine systems) or soil biota (terrestrial
systems).
C-143
-------
.•,,. ...• ., .. •. -••
Different sediment types that may be difficult to model (e.g., very low
or very high organic carbon content) or impractical to, sample (e.g.,
flocculent sediments). , .
Mixing zones in marine and estuarine systems.
....' . / " •'; '• ' '•• •" • :.'•'• ',:. ."' "'".''!'
High spatial heterogeneity in terrestrial systems. .,,
Confounding effects of complex mixtures (e.g., presence of petroleum
hydrocarbons or metals).
Bioaccumulation Issues
Issue 12: The Interim Report (Sections 3.2-3.5) summarizes available data on TCDD
bioconcentration, bioaccumulation, biomagniflcation, and biota/sediment accumulation
factors from laboratory experiments and field measurements. Discuss the applicability
of these factors to stressor characterization for the paper mill scenario.
The information on BCFs, BAFs, BMFs, and BSAFs for 2,3,7,8-TCDD summarized in
the interim report is useful for the general stressor characterization and exposure
assessment. For the stated risk management objective related to development of permit
specifications, however, the empirically derived BSAFs may be the primary factors of
interest for estimating bioaccumulation (see comments on Issue 17). BGFs will likely be
less useful than the other factors because of the importance of food-chain accumulation
of PCDDs and PCDFs. Evaluation of BAFs and BMFs may aid in finalizing a list of
receptors. . ' • ' •••' " •• /••.•••'. .•>:••.'•"-* •»' —
C-144
-------
Issue 13: the Lake Ontario BAFf may be useful as a predictor of residue levels in other
systems if C^ can be estimated accurately (Interim Report, Section 3.3). Comment on
the applicability of this BAFfor the paper mill stressor characterization.
The immediate use of BAFs for PCDDs and PCDPs is questionable because of the
problem of quantifying the concentration in lake water (C^). Aside from this potential
problem, the use of the Lake Ontario trout BAF is tenuous for the following reasons:
* Hie Lake Ontario trout BAF was estimated from a model that has not
been validated
» The Lake Ontario trout BAF was estimated for lake trout, which is not
a receptor in the reservoir
• The configuration of the reservoir may lead to significant spatial hetero-
geneities in sediment concentrations of PCDDs and PCDFs that are not
reflected in the concentrations in the water column. Because the sedi-
ment exposure route is important, any disequilibrium between sediment
and water would invalidate the use of a BAF.
Issue 14: The Interim Report (Section 3.4) indicates that biomagnification is significant
between fish and fish-eating birds but not between fish and their food. Comment on the
biomagnification pathway relative to stressor characterization and the conceptual model.
The biomagnification pathway is especially important for any risk assessment for
hydrophobic chemicals such as TCDD and related compounds. The conceptual model
should therefore address biomagnification for key receptors representative of all groups
of higher trophic level species (e.g., bass, mink, heron). It should not be concluded that
biomagnification is not significant between fish and their food items based on the
available data for several reasons. First, the available data are limited in coverage of
C-145
-------
species, selected TCDD and TCDF congeners, and specific exposure conditions (e.g.,
kinds of compounds, duration and magnitude of exposure). The interim report ac-
knowledged that the apparent low biomagnification factor for 2,3,7,8-TCDD in fish could
be due to biotransformation; however, the metabolic products will retain some residual
potency, which may be significant when passed to the next step of a food chain.
Second, the significance of biomagnification will vary substantially among fish species,
depending on their trophic position. Third, the significance of biomagnification at a
given step in a food chain can only be judged relative to the length and species com-
position of the entire chain.
The description of this issue did not mention fish-eating mammals, but they are included
hi the conceptual model (e.g., Figure 3 of the preworkshop materials). Biomagnification
could also be important in the link between freshwater mussels and mammals, such as
mink and raccoon. Biomagnification could play an important role hi the exposure of all
of these trophic groups to PCDDs and PCDFs.
Because of the importance of biomagnification, more information on the trophic structure
of the reservoir food web would have to be developed for this risk assessment.
Characteristics of trophic compartments that are most susceptible to high TCDD
exposures need to be more fully defined. Then, the list of key receptors can be finalized,
including identifying life.stages, age classes, and ecotypes of most concern.
Issue 15: Uncertainties associated with bioaccumulationfactors are discussed in Section
5.1.2 of the Interim Report. Discuss the relevance of these uncertainties to the prediction
of TCDD residues in Omigoshee Reservoir biota.
The uncertainties in BAFs appear to be greater than those in BSAFs. The issues of
analytical difficulties in quantifying C^,, relative bioavailability in laboratory and field
systems, and interspecies extrapolation of the Lake Ontario trout BAF would be
responsible for substantial uncertainties hi the use of BAFs for the reservoir.
C-146
-------
Section 4
EXERCISE 3
Conceptual Model Development
Workgroup Leaders:
Charles Menzie
Menzae-Cura & Associates, Inc.
Robert Huggett
Virginia Institute of Marine Science
College of William and Mary
C-147
-------
-------
Charles Menzie
Menzie-Cura & Associates, Inc.
C-149
-------
-------
Chaiies Menzie, Menzie-Cura & Associates, Inc.
Exercise 3 .
16. Comment on whether the focus on fish and wiuldlife captures
' the full range of potential ecological effects for this
scenario.
Comment
It probably does not capture the full range but probably
captures the most important ones from a risk assessment and
risk management standpoint. Obviously, there could be
effects on the benthic invertebrates, plankton, and aquatic
plants. However, limited data suggest that these are less
sensitive receptors. It would probably be useful to have a
thorough discussion of this, it would also be useful to
review available field data on the presence of benthic
invertebrates and other organisms in TCDD contaminated
sediments. I think there is probably a considerable amount
of field data that could have been 'relied upon in the
Interim Report to consider if TCDD contaminated sediments or
water have affected benthic and planktonic plants and
invertebrates .
17. Discuss the utility of available risk assessment tools for
accomplishing the goal of linking tissue residues to
loadings .
Comment
I do not think this falls into the category of "risk
assessment tools" per se. I think the best approach is to:
1) estimate surface sediment concentrations of TCDD et al.
in different areas of the reservoir associated with steady-
state loadings from. the paper mill; this would be done using
an appropriate model or models; and, 2) relate these
sediment levels to fish body burdens using a best estimate
and range of BSAF values. I think it would be helpful to
examine other similar systems (paper mills on reservoirs or
slow flowing rivers) that exist to develop a more robust set
of estimates for BSAFs. It may also be useful to derive a
BSAF for the complex mixture of dioxins/furans in the
effluent (from a similar plant or pilot plant) in an
experiment bioaccumulation bioassay. This would involve
exposing representative fish species (probably adults) to
sediments which have been exposed to the complex effluent at
various concentrations .
C-151
-------
18. Discuss the applicability of the limited field data
available for estimating BAFs and BSAFs.
Comment
Sediment, water, and food chain models are all subject to
considerable uncertainty. I think the use of BSAFs provides
a simple model which probably has less uncertainty than
models that attempt to represent the complexities of
interactions. I think these simple models are applicable
and perhaps more desrieable than the more complex
approaches. Application of these simple models should be
accompanied by a sensitivity analysis. It would be useful
to compare the characteristics of the sediments and water
column in the reservoir to those sites for which BAF and
BSAF values have already been developed. Key pieces of
information include: suspended sediment, DOC and POC for the
water column and TOG and grain size for the sediments. In
the case of the sediments, information on the benthic biota
would be helpful for comparing the degree to which the
sediments may differ in bioturbation rates.
19. Comment on how temporal dynamics and disequilibrium
situations should be considered in establishing (l) the time
course for build-up and/or decrease of TCDD levels.
Comment
To some degree, the relevant time course depends on the area
under consideration (i.e., localized areas around the mill,
the trunk of the reservoir near the mill, the reservoir as a
whole). The time course for build-up and decrease will
differ for these different spatial scales. For each spatial
scale, the time course selected for build-up should be based
on steady state conditions at which surface sediments are no
longer significantly increasing. A temporal -constraint on
this analysis is the operational lifespan of the facility.
It may be useful to consider conditions after several
selected time periods: e.g., 1, 5, 10, and 20 years. I
agree with the Interim Report with regard to using long
averaging times for estimating body burdens in response to
exposure conditions. The report suggests 1 year. However,
the location and exposure of species could vary seasonally
within the reservoir and if there are pronounced spatial
gradients in exposure concentrations, such seasonal
movements should be taken into account.
C-152
-------
20. Additonal Comment
Exposure and possible risks of TCDD to fish, birds, and
mammals in and around the reservoir can be evaluated at
various spatial scales. At the localized level, there might
be effects on individuals but if these are limited to a
small area, they may not translate to population-level
risks. Because TCDD will not .be homogeneously distributed
throughout the reservoir, it may be useful to estimate the
fraction(s) of the reservoir habitat (aquatic and
terrestrial) that fall within various risk levels. For
species with limited home range, this would provide an
initial estimate of the fraction of the "local population"
that is at risk. This approach would require estimates of
the distribution of TCDD throughout the reservoir,
identification of habitats within and along the reservoir
that would be utilized by selected receptor fish, bird, and
mammal species, information on the general biology of these
species with regard to foraging range and how this range
might vary during the year. A technical and philosophical
issue that arises with this approach is the definition of,
what constitutes the "local population".
C-153
-------
-------
Peter Chapman
EVS Environmental Consultants
C-155
-------
-------
EXERCISES. CONCEPTUAL MODEL DEVELOPMENT
Comments were invited, not required, on this Exercise. Given the time constraints, I have provided
only brief comments to the questions.
I preface these comments by noting that risk assessment is not science, it is simply a technique by
which scientific information can be converted into regulatory action. Mathematical calculations result
in quantitative risk assessment, which provides an appearance of rigour, since numbers are involved,
but which is not science. Rather, this is best described as "trans-science" (Weinberg, 1972), whose
primary function is to attempt to bring order and consistency out of the chaos resulting from public
and other pressures to act, often based on perception rather than reality. Risk assessment is basically
"...an instrument of social compromise, providing numerical answers in the face of vast scientific
uncertainties1' (Gots, 1992).
Conceptual models are an extremely useful way to delimit concerns and provide testable hypotheses
for scientific study. In this regard, bioassays (and other environmental assessment tools) can provide
information regarding three different scenarios: cases where there is clearly an effect, cases where
there is clearly not an effect, and the intermediate (and, unfortunately, predominant) case where there
may or may not be an effect There are four possible responses when "worst case" testing indicates
a potential problem: (1) if clearly required, control/regulate; (2) if the results are unclear and a delay
will not be catastrophic, test and verify; (3) if the issue has low priority compared to other problems,
put available resources to dealing with the highest priority problems, and deal with this issue when
and if it is appropriate to do so; (4) if there is no environmental problem, take no action.
Issue 16. Is the Approach Broad Enough?
This approach does not (and cannot, since we do not know everything there is to know) necessarily
capture the full range of potential ecological effects. However, based on what we do know, it is a
reasonable attempt to ensure environmental protection by focusing on the Vorst case", i.e., fish and
wildlife that consume fish. Worst case exposures would, from the model, appear to be in the main
arm of the reservoir.
C-157
-------
Issue 17. Utility of Available Risk Assessment'Tools
General comments on risk assessments are provided above. Comments on the incompatibility of
tissue residue levels and adverse effects have been provided previously. I am not sure what is actually
being asked here.
Issue 18. Comments on the Omigoshee Reservoir Conceptual Model
Again, I am not sure what is being asked. The model appears to be reasonable conceptually, and as
a basis for assessing the system. But, clearly, data need to be collected. For example, baseline data
on concentrations of bioaccumulative organic chemicals, which are noted as missing, are required.
Such and other information would, I would expect, be collected as part of the EIA.
Issue 19. Consideration of Temporal Dynamics and Disequilibrium
Presumably what is being asked here relates to either or both of: (1) predictions as to "worst case"
accumulation and purging of TCDD from the system, and (2) the timing of environmental monitoring
should the mill proceed. Predictions are only as good as the data they are based on, and in this case
additional site- and situation-specific data are needed. In particular, some experimentation is
required, in the laboratory, using plant bench-scale processes to determine rates of accumulation of
TCDD in sediments and biota. As regards timing of monitoring, this should be based on what is
known (e.g., present data, the EIA, these experimental results), and should initially be fairly
conservative (e.g., frequent, "worst case" philosophy), then amended depending on what is found.
REFERENCES CITED
Adams W. J. (1993) Aquatic toxicology test methods. In: Handbook of Ecotcodcology (eds. D. J.
Hoffman, B. A. Ratner, G. A. Burton & J. Cairns Jr.) Chapter 4, Lewis Publishers, Boca Raton, FL.
Cairns J. Jr. (1993) Environmental science and resource management in the 21st century: scientific
perspective. Environ. ToxicoL Chem. 12, 1321-1329
Chapman P. M. (1991) Environmental quality criteria - what type should we be developing? Environ.
ScL Technol 25, 1352-1359
C-158
-------
Dauer D. M. (1993). Biological criteria, environmental health and estuarine macrobenthic community
structure Mar. Pollut BulL 5, 249-257
Deichmann W. B., Henschler D., Holmstedt B. & Keil G. (1986). What is there that is not a poison?
A study of the Third Defense by Paracelsus Arch. ToxicoL 58, 207-213
Gots R. E. (1992) Toxic Risks - Science, Regulation and Perception. Lewis Publishers, Boca Raton,
Fl. 277 pp.
Smith E. P., Orvos D. R. & Cairns J. Jr. (1993) Impact assessment using the before-after-control-
impact (BACI) model: concerns and comments Can. J. Fish, Aquat. Set. 50, 627-637.
Sprague J. B. (1985) Factors that modify toricity. In: Fundamentals of Aquatic Toxicology (eds., G.
M. Rand & S. R. Petrocelli) pp. 124-163, Hemisphere Publishing Co., NY.
U.S. EPA (1986) Water Quality Criteria for Water 1986. United States Environmental Protection
Agency, Washington, D.C. EPA 440/5-86-001.
U.S. EPA/ACOE (1993) Evaluation of dredged material proposed for discharge in inland and near
coastal waters - testing manual (draft). U.S. Environmental Protection Agency and Army Corps of
Engineers, Washington, D.C.
Weinberg A. M- (1972) Science and trans-science Minerva 10, 209-214
C-159
-------
-------
G. Michael DeGraeve
Great Lakes Environmental
C-161
-------
-------
CONCEPTUAL MODEL
16. Given that fish and fish-eating birds and wildlife are at the top of the food chain in the
area of this reservoir, and these are among the most sensitive of those organisms that
have been evaluated to date, at this time the suggested approach is probably the most
reasonable. However, because we do not fully understand the wide range of potential
effects of TCDD and TCDD-like compounds (bacteria, primary producers, for example),
it is not reasonable to assume that the full range of ecological functions will be protected
by protecting fish and wildlife. The degree of protection for the non-target components
of the aquatic and terrestrial communities will only be understood by long-term
monitoring programs after the mill is in operation.
17. I feel that we have the necessary tools to accomplish this objective. We either know (or
we can determine) the lipid concentrations in the fish and wildlife species to be
protected, and we can determine the organic carbon content of the sediments in the
reservoir. With this information, an acceptable TCDD effluent discharge calculation can
be made for NPDES permitting purposes.
18. Although there are only limited existing field data that are available for estimating BAFs
and BSAFs, it is probably true that the utility of these approaches is controlled to a large
degree by the accuracy of the sediment organic carbon measurements and the lipid
determinations. I feel that as long as these measurements are accurately performed for
the reservoir, BAF and BSAF approaches can be utilized for the conceptual model.
19. Because of the rather constant TCDD discharge anticipated in the mill effluent, it seems
reasonable to expect that the water, sediments, and tissues will reach steady-state after a
period of time; that period of time should be predictable based upon the volume of
effluent being discharged per day, the volume of the reservoir, the organic carbon
content of the sediments, and the lipid concentrations in the fish. Similarly, when the
source of TCDD has been eliminated, the decreases in water, sediment, and tissue
TCDD concentrations are predictable based upon the reservoir flushing rate, the
sediment/water equilibrium kinetics, and the TCDD tissue depuration rate(s).
C-163
-------
-------
Wayne Landis
Institute of Environmental
Toxicology and Chemistry
Huxley College of Environmental Studies
C-165
-------
-------
Landis Comments September 1,1993
11
Exercise 3. Conceptual Model
16. Consistent with the Interim Report, the conceptual model focuses on effects on fish and
wildlife that consume fish. Comment on whether this approach captures the full range of potential
ecological effects for this scenario.
Comment
Of course not, as delineated in many of my prevbus comments. The conceptual model treats
community structure and function as a risk and not as a hazard. In fact, given any alteration at the
population level, a subsequent alteration in the structure of the ecosystem occurs, with indirect
effects likely. Perhaps the risk should be stated as the probability that the alterations in structure
and function of the community exceed that able to support game fish and associated wildlife.
Given the lack of data on the alterations in community structure due to TCDD and related
compounds, the fact that invertebrates are reported to not degrade these compounds, the lack of
data on invertebrate toxicity, and the lack of information of synergisms and antagonisms with other
compounds, a wide range of effects are likely to occur given the southern lake scenario.
Pesticides and herbicides are likely to occur in small amounts, dioxin and related chlorinated
aromatics are likely to occur in small amounts due to forest burning and other sources. All of these
may facilitate impacts at other structural levels within the ecosystem.
Another factor in the assessment of the overall impact of the proposed mill is its location.
Apparently the TCDD will have the opportunity to dose most of the lake, potentially restricting
migration from population sources to area affected by the dioxin. The location will probably accent
the impact due to the release of the toxicant.
I also have major concerns with the Assessment Endpoints and the Measurement
Endpoints.
1) Assessment Endpoints concentrate on the productivity of the bass, catfish, crappie
and bluegill. Driving the system to maximize productivity of each of these species will likely cause
important alterations to the remainder of the system. How are normal population cycles
incorporated into the assessment endpoint description? I would redefine the Assessment
Endpoints as design parameters using the habitat and resource requirements of these species as
the design set. Such a set would incorporate resource availability, habitat, and predator-prey
relationships. Of course, productivity needs to be clearly defined. If one species declines, others
may temporarily increase, altering the top down control on the system. As the structure changes
other resource populations are likely to change, eventually altering long term the fish assemblage
of the lake.
2) Measurement Endpoints are defined as the effects of TCDD on the reproductive
success as defined as egg production and larval survival. Measured or estimated concentrations
in tissues are taken as the best means of estimating theses effects. In both instances there are
C-167
-------
Landis Comments September 1,1993
12
large uncertainties related with both measurement endpoints. Perhaps a more in-depth look at
the lake, assessing the structure and energy flow may be more appropriate. Given the dose
response curves of dbxins and related materials and the variability in tissue concentration
reported in the Interim Report, by the time tissue concentrations have accumulated and are
measured in some individuals, large parts of the affected population have ceased to reproduce.
More sensitive measures that can provide some warning may be more appropriate. Binding of the
Ah receptor, behavioral measurements, or alterations to structural components of the community
may provide better measurement endpoints.
Finally, in regards to the lake system, how much data are really available for this
ecosystem. From the initial conceptual framework, apparently little is known about the current
inputs of toxicants, population dynamics of the fish and invertebrates, and the historical ranges of
these factors. As Katz et al (1987) have recently demonstrated, the past history of a lake is the
best predictor of the future. In fact, for most species other lacks are particularly bad at predicting
future dynamics even when the lakes are as similar as those in the Wisconsin lakes district. You
might never know, but an in-depth understanding of the current status and trends within the lake
may change dramatically the assessment and measurement endpoints.
17. The Interim Report emphasizes using tissue residue levels to estimate the adverse effects of
TCDD. However, to do the risk assessment outlined by the conceptual model, it will be necessary
to link predicted loadings of TCDD in the paper mill effluent to residues in the organisms identified
in the assessment endpoints. Discuss the utility of available risk assessment tools for
accomplishing this goal.
Comment
There are various models, such as FIGHTS and EXAMS that calculate transport and
btoaccumulatfon. These and other simulation models may prove very useful but should be looked
upon with caution. Often the virtual can be looked upon as reality and the system an expert
because of the convenience and lack of validation. Roofs have collapsed and aircraft crashed
because of overlooked aspects in simulation modeling. Experimental determinations would likely
improve the predictability of the models and serve as validation.
Toxicokinetics models may be very useful, especially those developed for anesthetics
and other lipid soluble materials.
18. The Interim Report describes the limited field data that are available for estimating BAFs and
BSAFs. Discuss the applicability of these factors to the Omigoshess Reservoir conceptual model.
Comment
C-168
-------
Landis Comments September!, 1993
13
These BAFs and BSAFs are probably the best shots for estimating eventual tissue
concentrations. As currently designed, however, the lake is assumed to be rather homogenous,
and it is not. Without bathymetry data it is hard to judge, but it is unlikely that the accumulation of
TCDDs will be similar in the four pririeipial basins. Basins 2 and 3 are likely to be shallow, with a
great deal of input from the surrounding watershed. Sediment loading is reported to be high with
a great deal of shoreline complexity: Of course, the complexity of the shorelines make this an
excellent habitat for sport fish and wildlife. The complexity of the system means that dioxin
concentrations are likely to be spatially and temporally heterogeneous. Variation in flow rates from
the various tributaries will also alter the relative TCDD inputs.
I would suggest strongly a adoption of these models with a breakdown into smaller
sections to get a feel for the heterogeneity of the bioaccumulation over space and with different
flow and input rates. From these data a probability distribution may be derived to more accurately
describe these factors for lake.
19. The temporal dynamics and the disequilibrium situations commonly associated with TCDD are
mentioned in the Interim Report (section 2.3). Comment on how these aspects should be
considered in establishing (1) the time course for the build up of TCDD levels following initiation of
the paper mill discharge and (2) the time course for the decrease of TCDD levels and recovery of
biota should the paper mill cease operation.
Comment
Temporal dynamics and disequilibria determine the structure and resulting function of
ecosystems, they are often the stuff of dramatic evolutionary changes, including speciation.
Given the toxicity, long half life and the characteristics of the lake, the time courses for TCDD
buildup and subsequent decrease will complex. The mean trend will likely be well behaved, but
the numbers surrounding the increase and decline could appear stochastic.
In all likelihood, the dynamics will be typical of non-linear systems. Good examples are
trends in weather and in disease. Both exhibit yearly vagaries that are essentially bounded
chaotic systems. In the case of disease, the application of a vaccination program may not show a
decreasing trend in the prevalence of the disease because of the chaotic dynamics, temporally
and spatially. Outbreaks of the disease may still occur, not because of the lack of efficacy of the
vaccine, but because of the intrinsic mathematics of the spread of the agent. The act vaccination
actually puts new bounds of the chaotic dynamics, resulting in an overall decreasing trend.
Persistent materials, coupled with a heterogeneous habitat are also likely to exhibit similar
dynamics as the pulp mill effluent is eliminated.
In contrast, the addition of TCDD will be in two steps, first the contamination of the media,
followed by the bioaccumulation. The first will be more straightforward, rather simple input-output
C-169
-------
Landis Comments September 1,1993
14
with variability due to the morphology of the lake. The tissue concentrations will more likely follow
the bounded chaotic dynamics since tissue concentrations depend heavily upon prey population
productivities, algal dynamics and meterological inputs.
Lastly, decline in the concentration of TCDD in the fish populations does not mean a
return to the original conditions found in the lake. In fact the contrary is probabry likely. I now
believe that recovery to the original state of an ecosystem is not possible. As in the terminology of
the theory of complex systems, ecosystems including their constituent populations retain a
"memory" past events (Nicolis and Prigogine 1989). Because of this memory complex systems
are inreversbte, the opposite of classical Newtonian mechanics. This is not to say that viable fish
populations will not occur, but the structure of the populations and the structure of the supporting
ecosystem will be different. The importance of this in the long term is that the response of the
system to additional stressor events will likely be quite different because of the history of TCDD
contamination. Even another TCDD event will likely have different outcomes. The almost
automatic assumption that stability exists and some sort of recovery to an
original state is unwarranted and perhaps dangerous in that it may lead to an
underestimation of long-term effects.
References
Caims, J., Jr. 1986. The myth of the most sensitive species. Bioscience. 36:670-672.
Dickson, K.L., Waller, W.T., Kennedy, J.H. and Ammann, L.P. (1992) Assessing the relationship
between ambient toxicity and instream biological response. Env. Tox. Chem. 11,1307-1322.
Hassell, M.P.H, Comins, N. and May, R.M. 1991. Spatial structure and chaos in insect population
dynamics. Nature 353:255-258.
Hutchinson, G. E. (1961) Paradox of the plankton. Amer. Nat 95:137-143.
Johnson, A.R. (1988a) Evaluating ecosystem response to toxicant stress: a state space
approach. \nAquaticToxicologyandHazardAssessment: 10th Volume, ASTM STP971
(Adams, W.J., Chapman, G.A. and Landis, W.G., eds) American Society for Testing and Materials,
Philadelphia, pp. 275-285.
Johnson, A.R. (1988b) Diagnostic variables as predictors of ecological risk. Environmental
Management 12, 515-523.
Katz, T.K., Frost, T.M. and Magnuson, J.J. (1987) Inferences from spatial and temporal variability
in ecosystems: Long-term zooplankton data from lakes. Amer. Nat. 129, 830-846.
Kersting, K. (1988) Normalized ecosystem strain in micro-ecosystems using different sets of state
variables. Verh. Internal. Verein. Limnol.23,1641-1646.
Landis, W. G. 1986. Resource competition modeling of the impacts of xenobiotics on biological
communities. Aquatic Toxicology and Environmental Fate: Ninth Volume ASTM STP 921. J. M.
Poston and R. Purdy Eds., American Society for Testing and Materials. Philadelphia, pp 55-72.
C-170
-------
Landis Comments September 1,1993
f5
Landis, W. G., R. A. Matthews, A.J. Markiewicz, N. A. Shough and G. B. Matthews. In press.
Multivariate Analyses of the Impacts of the Turbine Fuel Jet-A Using a Microcosm Toxicity Test. J.
Environ. Sci. Vol 2:2.
Landis, W. G., R. A. Matthews, A. J. Markiewicz and G. B. Matthews. In press. Multivariate Analysis
of the Impacts of the Turbine Fuel JP-4 in a Microcosm Toxicity Test with Implications for the
Evaluation of Ecosystem Dynamics and Risk Assessment. Ecotoxicology.
Matthews, G.B., Matthews, R.A. and Hachmoller, B. (1991 a) Mathematical analysis of temporal and
spatial trends in the benthic macroinvertebrate communities of a small stream. Canadian Journal of
Fisheries and Aquatic Sciences. 48, 2184-2190.
Matthews, R.A., Matthews, G.B. and Ehinger, W.J. {1991 b) Classification and ordination of
limnological data: a comparison of analytical tools. Ecological Modeling. 53,167-187.
May, R.M. and Oster, G.F. (1978) Bifurcations and dynamical complexity in simple ecological
models. Amer. NaM 10, 573-599
Nicolis, G. and I. Prigogine (1989) Exploring Complexity: An Introduction. W. H. Freeman and
Company, New York.
Richerson, P., R. Armstrong and C. R. Goldman. 1970. Contemporaneous disequilibrium, a new
hypothesis to explain the "paradox of the plankton". Proc. Natl. Acad. Sci. l/SA,67:1710-1714.
Suter, G. (1993) A critique of ecosystem health: Concepts and indices. Environ Tox. Chem. in
press.
Tilman, D. 1982. Resource Competition and Community Structure, Princeton University Press,
Princeton, pp. 11-138.
Appendix 1: Examples of Test Systems for the Examination of Impacts at
the Community Level
Effects of Atrazine on Freshwater Microbial Communities
Organisms Multiple species
Test community collection: On polyurethane foam (PF) substrates (6.0 x 5.0
x 4.0 cm)
PF substrate exposure: 14 days
Experimental design
Test type: Multispecies toxicity test
Test vessel size and type: High density polyethylene tubs (35 x 28 x 15 cm)
with an inlet tube to deliver a mixture of diluent
water and toxicant at one end and three drain
holes at the other end
Test volume: 7 L
Toxicant stocks: Primary stock was made by dissolving 200.84
mg of atrazine in 100 mL of methanol and
diluting n distilled water to 1 L
C-171
-------
Landls Comments September 1,1993
16
Secondary stocks were made by diluting the
appropriate amount of primary stock with diluent
water
Both stocks were made once at the start of the
test and again on day 10
Number of replicates: Triplicates of five concentrations of atrazine, a
diluent control, and a solvent (methanol) control
(total of 21 chambers)
Toxicant concentrations: 3,10,30,100, and 300 ug/L (measured at the
start of the test and after 10 and 21 days
Test duration: 21 days
Physical and chemical parameters
Temperature: Uncontrolled; ranged from 13.5°C to 15.0*0
Light intensity: 5000 lux (bulbs located 30 cm above the test
systems)
Photoperiod: 16 h light /8h dark
Type of dilution water: Tap water dechlorinated by passage through
activated charcoal
Hardness: 70 ppm CaCOs
Alkalinity: 45 ppm CaCOs
pHt 8.4
Flow rate: Approximately 7 turnovers of the 7
Clinical Examinations/Endpoints: Species richness, total biomass, protein
concentrations, chlorophyll a concentration
Source: Pratt, J.R., NJ. Bowers, B.R. Niederlehner and J. Cairns, Jr. 1988. Effects of
Atrazine on Freshwater Microbial Communities. Archives of Environmental Contamination
and Toxicology 17:449-457.
Summary of Test Conditions for the Outdoor Aquatic Microcosm Tests to Support
Pesticide Registrations
Organisms Bluegill sunfish (Lepomis macrochirus),
fathead minnow (Promephales promelas),
channel catfish (Ictalurus punctatus), or others
may be present: Phytoplankton, periphyton,
zooplankton, emergent insects, and benthic
macroinvertebrates
Size of organism: Biomass of fish added to the microcosms should
not exceed 2 grams per cubic meter of water
Experimental design
Test type: Microcosm
C-172
-------
Landis Comments September 1,1993
17
Test vessel size and type: Tanks with a surface area of at least 5 m2, a
,'i.^;// ' deptri of rat least 1.25 m, and a volume of at least
6 m3 made of fiberglass or some other inert
...:.-,. material; smaller tanks could be used for special
purposes in studies without fish .-_.
Addition of test material: Allow microcosms to age for approximately 6 to
8 weeks before adding test material
Apply by spraying across water surface, apply
the test material in a soil/water slurry, or apply
test material in a water based stock solution
Sampling: Begins approximately two weeks after the
microcosms are constructed and continues for
two or three months after the last treatment with
test material; frequency depends characteristics
of test substance and on treatment regime
Dosage levels, frequency of test material
addition, and number of replicates per dosage
level are determined based on the objectives of
the study
Physical and chemical parameters
Temperature: Maintained by partially burying tanks in the
ground or immersing in a flat-bottomed pond
Sediment: Obtained from existing pond, containing a
natural benthic community is added to each
microcosm directly on the bottom, in trays or
other containers; sediment should be 5 cm thick
Water: Obtained from healthy, ecologically active pond;
Water level should be set in the beginning and
not allowed to vary more than ±10% throughout
study; if water level falls more than 10%, add
pond water, fresh well water, or rain water; if
water level rises more than 10 %, surplus should
be released and retained
Meteorology: Should be recorded at the study site or records
obtained from a nearby weather station; data
should include air temperature, solar radiation,
precipitation, wind speed and direction, and
relative humidity or evaporation
Photosynthetic Carbon Metabolism of Size-Fractionated Phytoplankton during an Experimental
Bloom in Marine Microcosms
Organisms Phytoplankton
Experimental design
Test type: Marine microcosm
Test vessel size and type: 12 L acid-washed polycarbonate bottles
i - •
Size and number of replicates
C-173
-------
Lands Comments September 1,, 1993 —^ -.,-..,. , ....
18
per sample: 60 mL subsamples removed 4 times during the
experiment to measure 14C incorporation into
macromolecules
Number of test organisms
per chamber: . Phytoplankton samples are fixed in Lugol's
iodine then identified and counted under an
inverted light microscope
Test du ration:__ 9 days
Clinical examinations: Concentrations determined: Nitrate, nitrite,
soluble reactive phosphorus (SRP); determined
four times during the experiment. Samples for
total dissolved phosphorus (TP) and total
dissolved nitrogen (TN) are maintained deep-
frozen and subsequently analyzed after UV-
induced oxidation. Size fractionated chlorophyll
a is measured twice a day
Physical and chemical parameters
Temperature: 15°C
Light intensity: Photon flux density (PFD) of 100 u,E/m2s from
north-light fluorescent tubes
Photoperiod: , 14 h light 710 h dark
Endpoint: Growth
Source: de Madariaga, I. and E. Fernandez. 1990. Photosynthetic Carbon Metabolism of Size-
Fractioned Phytoplankton During an Experimental Bloom in Marine Microcosms. J. Mar. Biol. Ass.
U.K. 70:531-543.
Summary Of Test Conditions for the Standardized Aquatic Microcosm: Fresh Water
Organisms
Type and number of test
organisms per chamber: Algae (added on Day 0 at initial concentration of
103 cells for each algae species):
Anabaena cylindrica,
Ankistrodesmussp.,
ChiamydomonasreinhardiQO,
Chlorella vulgaris,
Lyngbya sp.,
Nitzschia kutzigiana (Diatom 216),
Scenedesmus obliquus,
Selenastrum capricomutum,
Stigeoclonium sp., and
Ulothrix sp.
Animals (added on Day 4 at the initial numbers
indicated in parentheses):
Daphnia magna (16/microcosm),
Hyalella azteca (12/microcosm),
Cypridopsis sp. or Cyprinotus sp.
(ostracod) (6/microcosm),
C-174
-------
Landis Comments September 1,1993
19
Hypotrichs [protozoa] (0.1/mL) (optional),
and Philodina sp. (rotifer) (0.03/mL)
Experimental design
Test type: Multi-species
Test vessel type and size: One-gallon (3.8 L) glass jars are recommended
soft glass is satisfactory if new containers are
used; measurements should be 16.0 cm wide
at the shoulder, 25 cm tall with 10.6 cm
openings.
Medium volume: 500 mi_ added to each container
Number of replicates: 6
Number of concentrations: 4
Reinoculation: Once per week add one drop (Circa 0.05 ml_) to
each microcosm from a mix of the ten species
= 5 x 102 cells of each alga added per
microcosm
Addition of test materials: Add material on Day 7; test material may be
added biweekly or weekly after sampling
Sampling frequency: 2 times each week until end of test
Test duration: 63 days
Physical and chemical parameters
Temperatu re: Incubator or temperature controlled room is
required providing an environment 20 to 25°C
with minimal dimensions of 2.6 by 0.85 by 0.8 m
high
Work surface: Table at least 2.6 by 0.85 m and having a white
or light colored top or covering
Light quality: Warm white light
Light intensity: 80 u.E nrr2 photosynthetically active radiation
s'1 (850 to 1000 fc)
Photoperiod: 12 h light/12 h dark
Microcosm medium: Medium T82MV
Sediment: Composed of silica sand (200 g), ground, crude
chitin (0.5g), and cellulose powder (0.5 g) added
to each container.
pH level: Adjust to pH 7
C-175
-------
Landis Comments September 1,1993
20
Endpoint: DO, pH, enumeration of species, diversity, P/R
ratio, nutrients, available algae, total algae, total
Daphnia etc.
ASTM E 1366-91 (1991) Standard Practice for the standardized aquatic microcosm: fresh water,
Vol 11.04. pp 1017-1051. American Society for Testing and Materials, Philadelphia.
Taub, F.B. (1989) Standardized aquatic microcosms. Environm. Sci. Techno/. 23, 1064-1066.
C-176
-------
Richard Peterson
School of Pharmacy
University of Wisconsin
C-177
-------
-------
PRE-MEETING COMMENTS: RICHARD E. PETERSON
EXERCISE 3. CONCEPTUAL MODEL
16. If TCDD-like congeners reduce the populations of fish
upon which birds and mammals forage in the reservoir it may
indirectly affect their populations. It is not clear how such
an indirect effect of TCDD on bird and mammal populations will
be estimated in the conceptual model.
It is possible that the most sensitive fish species in
the reservoir to TCDD-induced reproductive toxicity or early
life stage mortality is a forage fish such as a species of
minnow. The conceptual model does not consider this
possibility from either a direct effect (TCDD decreasing the
population of minnows) or indirect effect (decrease in the
minnow population decreasing the population of piscivorous
fish) point of view.
For river otters the main source of TCDD-contaminated
food from the reservoir may be crayfish rather than fish.
However, the conceptual model does not consider this
possibility in assessing the exposure of river otters to these
chemical contaminants.
C-179
-------
-------
Thomas Sibley
Fisheries Research Institute
University of Washington
C-181
-------
-------
Pre-meeting Comments for U.S. Environmental
Protection Agency
Workshop on Ecolological Risk Assessment Issues for
2,3,7,8-Tetrachlorodiben2o-p-Dioxin
14-15 September 1993
Thomas H. Sibley
Fisheries Research Institute (WH-10)
University of Washington
Seattle, WA 98195
Exercise 3. Conceptual Model
16. No, focusing on fish and wildlife that consume fish
will not capture the full range of potential ecological
effects. However, it is the most reasonable approach given
the available data and should identify those species that
will be affected first.
17. I think the available tools are more advanced than the
data. The risk is to believe the results of the model and
forget the questionable nature of the data that are input
for the model.
18. I think it is necessary to use these factors because
analytically they are among the more reliable measurements.
However, the appropriate values will depend upon the organic
content of the sediments and the biological species being
considered. It would be more desirable to make predictions
based upon measured concentrations of dissolved TCDD but
that is not feasible for most environments.
19. There are limited data available for these assessments
and most of those data assume equilibrium, it is likely
that few, if any, systems are at equilibrium but the
necessary reaction rates to consider kinetic reactions have
not been obtained yet. In my opinion, attempting to
introduce the kinetics would introduce so much uncertainty
that it is better to utilize the equilibrium models.
C-183
-------
-------
Bill Williams
Ecological Planning Toxicology, Ine,
C-185
-------
-------
Exercise 3. Conceptual Model
Issues for Consideration
16. Fish-eating birds and mammals-
Consistent with the Interim Report/ the conceptual model focuses on
effects on fish and wildlife that consume fish. Comment on the whether
this approach captures the full range of potential ecological effects for
this scenario.
Comment:
The point source of TCDD and other chemicals in the example is manifest in
the water column and sediment, ultimately finding its way up the food
chain to the fish and predators of fish. This is a good approximation of
the potential exposoure and effects of these chemicals.
17. Linkage of environmental concentration, residues and effects-
The Interim Report emphasizes using tissue residue levels to estimate the
adverse effects of TCDD. However, to do the risk assessment outlined by
the conceptual model, it will be necessary to link predicted loading of
TCDD in the paper mill effluent to residues in the organisms identified in
the assessment endpoints. Discuss the utility of available risk
assessment tools for accomplishing this goal.
Comment :
Tissue levels of TCDD must be linked to exposure levels (water column and
sediment) in order to utilize the information in future assessments.
Without the link, each successive risk assessment, or the evaluation of
the success of any mitigation efforts will be lessened by lack of these
important data. It is critical to tie environmental concentration,
exposure level, tissue level, and specific effects to the scenario being
evaluated.
18. The Interim Report describes the limited field data that are available
for estimating BAF's and.BSAF's. Discuss the applicability of these
factors to the Omigoshee Reservoir conceptual model.
Comment:
The concentration of organochlorine residue in wildlife tissue is a
constantly changing function determined primarily by the concentration
C-187
-------
of the chemical in the exposure route (water for aquatic animals,and
food for terrestrial animals). It is.generally held that the rate of
loss (depuration) is approximately the same as the rate of gain
(uptake). Further/ the rate of bioaccumulation/ bioconcentration varies
greatly between species and chemical. The common practice of "back
calculating" water or food concentration (exposure) using tissue
concentration should be thought of as a means of developing an
hypothetical value. The proof of the hypothesis is the actual measured
environmental value. The use of an historic residue value of "a fish
caught in 1985" etc., simply does not provide the needed validation of
the relationship. To be valid, the proper approach is to show that the
tissue concentrations are present in a randomly collected (but site-
specific) sample. Any use of .anecdotal information to demonstrate the
back calculation hypothesis is neither scientifically nor statistically
meaningful for 2,3,7,8-TCDD or for any other organochlorine.
Avians show little bioaccumulation of 2,3,7,8-TCDD, and low levels (<20-
25x) of accumulation of other organochlorines. This is thought to be
related to their ability to metabolize many chemicals, either altering
their mode of action or inactivating them. Again, this phenomenon is
generally species—specific.
Because of their high oil content, fish present the worst case for
organochlorine uptake because the lipophilic characteristic of
organochlorines results in their being sequestered in fat. However, it is
also true that these deposits of chemicals are extremely variable and
change in response to changes in the exposure level.
19. Accumulation and depuration of TCDD-
The temporal dynamics and disequilibrium situations commonly associated
with TCDD are mentioned in the Interim Report (section 2.3). Comment on
how these aspects should be considered in establishing (1) the time course
for the build-up of TCDD levels following initiation of the paper mill
discharge and (2) the time course for the decrease of TCDD levels and
recovery of biota should the paper mill cease operation.
Comment::
The important consideration in this effort would be to provide a rate-
relationship for the bioaccumulation measured in actual exposure studies
and then to track (in spatially and temporally random fish samples) the
reliability of these projections when compared to actual measurements.
The metrics in this example must be pre-determined in clearly defined data
C-188
-------
limit objectives (e.g. if actual data do not fall outside the predicted
natural variation in these measurements, it must be assumed that no
definitive exposure or effect has occured). The concentration of
organochlorine residue in wildlife tissue is a constantly changing
function determined primarily by the concentration of the chemical in the
exposure route (water for aquatic animals and food for terrestrial
animals). It is generally held that the rate of loss (depuration) is
approximately the same as the rate of gain (uptake). Further, the rate of
bioaccumulation/ bioconcentration varies greatly between species and
chemical. The common practice of "back calculating" water or 'food
concentration (exposure) using tissue concentration should be thought of
as a means of developing an hypothetical value. The proof of the
hypothesis is the actual measured environmental value.
C-189
-------
-------
Robert Huggett
Virginia Institute of Marine Science
College of William and Mary
C-191
-------
-------
R. Huggett
CONCEPTUAL MODEL
GENERAL; My first thought is that if a conceptual model
(pages 8 & 9) does not need to be any more detailed than this,
one could develop generic conceptual models based on Row's
since fish, bird or mammal species are not mentioned nor is
much of anything else except that the assumed TCDD mode of
action is via the embryo. Should not more about the biotic
and abiotic components and their interactions relative to TCDD
in the southern reservoir be given? I think it would be
helpful if for no other reason than to help the uninformed
reader who may only see this part.
Question 16; To categorically state that the impacts of TCDD
will only be on fish and fish eating wildlife is risky. While
I think that it is a good assumption, we really do not know
much about benthie worms, etc. relative to TCDD.
Question 17; The discussions given in my comments on stressor
characterization are appropriate here. In effect there will
be a "House of Cards" built of various partitioning
coefficients. It is the best we can do at this time, but it
is still shaky.
Question 18; The same discussions given for the stressor
characterization and question 17 are appropriate here.
Question 19; Please see the answers to questions 9, 10, 12
and 15. In addition, I expect the time course for decrease in
TCDD residues, should the paper mill cease operation, would
mainly be affected by the sedimentation rates. This has been
shown to be the case for other hydrophobia chemicals such as
C-193
-------
kepone in Virginia and DDT and PCB in coastal waters of
California.
C-194
-------
Keith Cooper
Health Science Institute
Rutgers University
C-195
-------
-------
K. Cooper Premeeting Comments Exercise 2 & 3
Exercise 3. Conceptual Model
16. Although most of the data would suggest that
invertebrates respond at much higher concentrations (or
dosages) the studies have mainly dealt with high dose short
term exposures and have not examined life stages that may be
more sensitive. As with the adults from other species the
sensitive stages are those which are undergoing rapid
development and are sensitive to enzymatic alterations. The
assumption that because a animal does not poses a Ah receptor
does not preclude effects being observed. This is an .area of
research that needs to be examined.
17. The only way that a good correlation can be made is by
relating the effects to tissue dose. By establishing
threshold levels for tissue dose then an estimation can be
made for potential effects on the organism* In most of the
studies there has been little work done to examine the
concentrations into other tissues other than the edible muscle
(because of human health concerns). There is a need to relate
the levels in the various tissues to effects in the animal.
In some of the laboratory studies toxicokinetics can give some
indication as to relative levels in the other organs of some
of the fish described in the scenario. There needs to be
evaluations of distribution at different concentrations and
C-197
-------
K. Cooper Premeeting Comments Exercise 2 & 3 '
for different size organisms. In the scenario there needs to
be an estimation of the loading into the reservoir and the
estimated concentrations in the sediment and suspended
sediment. The estimated levels must then be confirmed by
field sampling. There might also be a way to sample fish, and
birds examining the levels obtained from blood samples in a
similar manner that humans are currently examined. These same
animals could be tagged and then monitored at later dates to
establish any change in body burdens. This assumes that there
is a good relationship between serum fat levels and general
body burdens.
18. As stated above there can be a tiered approach using best
estimates and then validation by field sampling.
19. The amount of dioxins which accumulate can be monitored
at specified areas down stream from the plant. The plant
prior to operation normally carry out environmental
assessments which should include background surveys of both
fish and sediment. The establishment of information concerning
the populations of fish, invertebrates and birds should be
characterized. This characterization should continue after
the opening to the plant. The levels which are entering the
river will be determined by the facilities treatment of the
effluent prior to discharge. Part of the problem in early
C-198
-------
K. Cooper Premeeting Comments Exercise 2 & 3
detection of levels is the analytical sensitivities••. These
can be overcome with high volume sampling and the most up to
date cleanup procedures and detection.
There is very good data from a number of paper mills and
state agencies which have undergone modifications to their
process to eliminate TCDD to a large extent. This information
can be obtained for a number of species that inhabit
downstream locations. The tl/2 of elimination for a number of
fish can, be obtained from laboratory studies and field
information. It would appear for bivalve mollusks that the
rate of elimination is related to the amount of water that is
filtered. There is little information on Crustacea for the
rate of eliminatibri of these compounds.
The temporal dynamics and disequilibrium situations will
result in lower expected levels in the tissues if a lake wide
average is used. The movement of fish and other animals into
less or more contaminated areas will effect the amount
estimated in the tissues. The physiological status of lower
vertebrates, and invertebrates varies over the year and with
these variations the amount of material taken up and
eliminated will vary. Because of these temporal changes
sampling should be carried out during different times of the
year to examine the different physiological states.
C-199
-------
-------
Joseph DePinto
Department of Civil Engineering
State University of New York at Buffalo
C-201
-------
-------
Exercise 3. Conceptual Model
16. Adequacy of focusing on fish and wildlife
There is no way that this approach captures the full range of ecological
effects. I am sure that there are subtle effects that effect trophic stucture and
function that are not captured by the conceptual model. The important thing
is that we have focused in on an endpoint that is as sensitive to the stressor as
we can assess within our current sphere of available data and knowledge base.
I think that focusing on fish and fish-eating wildlife has accomlished that
objective.
17. Linking loadings to tissue residues
Most of the previous discussion has focused on this problem, having
indicated the current capabilities in this area and the major uncertainties. The
only additional comment here is that a complete risk assessment, quantitatively
linking material input to effects, must put more emphasis on physical-
chemical fate and transport modeling than was indicated in the scenario
description. This is essential if it is desired to manage risk by imposing loading
restrictions.
IS.Using BAFs and BSAFs
These concepts can provide a reasonable estimate, but the analyst must
recognize that these numbers will be highly site-specific as well as temporally
and spatially variable within a given system. The only way to narrow the
variances (uncertainty) in these values is to continue to measure them in a
variety of systems and for a variety of environmental conditions.
19. Dynamics and disequilibrium conditions
The report recognizes that there may be a time lag in biotic response
(bioaccumulation) to a given exposure. Dynamic bioaccumulation models such
as Connolly's and Thomann's have demonstrated a typical two to three year lag
J.V. DePinto
C-203
-------
between the manifestation of a given exposure concentration in a system and
the response in the top predator fish. However, as discussed above the most
significant time response is associated with the approach to steady-state
following initiation of a loading and the time course for washout following
cessation of the load. Not having the appropriate data for the Omigoshee River
and Reservoir, I cannot give a good estimate of the time scale of these
responses. However, in systems like this there are typically two characteristic
responses: a relatively rapid response related to the response time of the water
column alone, and a typically much slower response related to the slow response
time of the sediments and the associated sediment-water interactions driven by
diffusion and resuspension. In systems like this the rapid response for an HOC
will be on the order of the hydraulic retention time of the water body, whereas
the significance and speed of the slower sediment response will depend on the
degree to which sediment-water interactions are important in the system. A
good estimate of the significance of sediment-water interactions is the value of
the overflow rate (ratio of mean depth to hydraulic retention time) of a system;
the smaller the overflow rate, the more significant a given interfacial transport
process becomes in terms of its impact on the water column. For a system like
the Omigoshee the sediment response time may be on the order of 10-20
hydraulic retention times.
If there are management questions related to system response time, then
a time-variable (not steady-state) chemical fate, transport, and food chain
bioaccumulation model must be used for this portion of the analysis. Then the
analysis of effects could either use the time profile of exposure or a time average
of the calculated chemical concentrations.
I recommend that the full risk assessment for dioxin could benefit greatly
from what has been learned about the fate, transport, and bioaccumulation of
PCB congeners from the Green Bay Mass Balance Study.
J.V. DePinto
C-204
-------
Derek Muir
Department of Fisheries and Oceans
C-205
-------
-------
PRE-MEETING COMMENTS FOR THE PEER PANEL WORKSHOP ON 2,3,7,8-TCDD
Comments on Exercise 3. Conceptual model
17. The availability of tools for estimating tissue residues:
Chemical transport and fate models such as WASP4 and RIVER/FISH
(combined with food chain models) are available and have already
been used to link TCDD concentrations in pulp and paper mill
effluent. Results of these simulations have shown reasonable
agreement between predicted and observed TCDD concentrations in
fish (Hinton 1991; Holloran 1993). These models will calculate TCDD
concentrations in dissolved and suspended particulate phases in the
water column as well as levels in surficial sediments. These
concentrations could be determined within segments of the aquatic
environment i.e. immediately downstream of the mill, the reservoir
and its tributaries. The models can be run in steady state, i.e. no
change of flow, sedimentation rates etc or in a dynamic mode in
which case detailed hydrodynamics and sediment transport are
required. This should not be a problem for the paper mill/reservoir
scenario. A key parameter is Koc value (or Kow from which Koc is
calculated) which determines the extent of partitioning between
water and suspended particles in effluent, river water and bed
sediments. There is much uncertainity as to the best Koc value for
TCDD but 107 could be used as a value for initial modelling
purposes.
Accurate estimates of Koc would best be obtained by lab
measurements using suspended solids from mill effluent, river water
and bed sediments spiked with TCDD (and suitably aged), using the
method of Lodge and Cook. During operation of the mill, field
monitoring using methods such as in situ sparging (Resendes et al.
1992) and high, volume sampling/filtration could be employed, with
HRGC-MS analysis to determine TCDD in "dissolved" and particulate
phases.
Tissue concentrations can be estimated with the Thomann food
chain model using WASP outputs for concentrations in various
compartments and segments of the aquatic system. In this model
assimilation efficiencies (AE)from food for organisms at each
trophic level are key parameters (Thomann 1989) along with k, and k2
values. AE values for TCDD are available for a limited number of
fish species. AE's for TCDF were independent of concentration over
a 100-fold concentration range (Muir et al. 1992) but may be
dependent on feeding rate (Clark et al. 1992). AE's have been
derived for accumulation of a few hydrophobia organics by benthic
invertebrates but information on TCDD assimilation is limited. To
overcome the lack of data for lower food chain organisms the food
chain model of Gobas (1992) assumes equilibrium between
sediment/water and biota (i.e. BSAF or BAF values). This model has
been successfully applied to predict concentrations of TCDD in fish
near pulp mills on the Fra'ser River (Gobas 1993).
18. Applicability of BAFs and BSAFs. BAFs and/or BSAFs, or a food
chain model (Fig. 4) could be employed to calculate TCDD tissue
concentrations in the paper mill/reservoir scenario, assuming that
levels in dissolved, suspended particulate, and sediments are
calculated using chemical transport/fate models. There are
uncertainities in BAFs for TCDD because Cdw and C*w are based on
C-207
-------
estimates and have not verified by field studies. Even if Cdw could
be measured in the Omigoshee Reservoir using large volume samples
there would still be debate about what was truly dissolved because
there are problems with all available techniques (i.e. filtration,
sparging). BSAFs also have uncertainties due to site specific
differences in food chains, sediment-water disequilibrium etc. But
BSAFs have the advantage of being verifiable with existing
techniques via sampling and analysis of biota and sediments. There
is also a reasonably large dataset of BSAFs although there is a
need for a broader selection of data. Laboratory and field derived
BSAFs for oligochaetes are similar for PCBs (Ankley et al. 1992) -
thus lab studies could be used to develop more BSAFs for different
sediment types, and species.
Using BSAFs to estimate tissue concentrations in lower food
chain organisms combined with food chain modelling (application of
pharmacokinetic parameters, growth rates, age and feeding
preferences) to estimate concentrations in forage fish and
piscivorous fish is probably the best approach for the present
scenario.
19. Temporal dynamics and disequilbrium. Using WASP or similar
transport models, with information on the hydrodynamics and
sediment transport in the river and reservoir, it should be
possible to simulate the dispersion of the paper mill effluent and
the deposition of solids on the river bed and in the reservoir.
Although the river is reported to have no depositional zones,
processes such as flocculation of suspended organic particles from
the mill treatment ponds may result in locally high concentrations
of TCDD contaminated particles in the river bed (Krishnappan 1993).
Flushing of bed load in the river immediately downstream of the
mill during high flow events could result in a pulsed rather than
continuous deposition in the deep zones of the reservoir. During
the initial deposition of TCDD surface concentrations will not be
at steady state, disequilibrium between sediments and water (R^)
will be >1. This situation has been simulated in lake enclosures
spiked with radiolabelled 1,3,6,8-TCDD (Servos et al. 1992) and
with 2,3,7,8-TCDF (Muir et al. I992a). In both cases, TCDD and TCDF
were sorbed on particles and BSAFs for benthic invertebrates were
temporality elevated. Disequilibrium between organisms and sediment
(R,,) will also be >1, especially for filter feeding organisms.
During a shutdown phase, or switch to chlorine-free technology,
TCDD in sediments will be buried by less contaminated material.
This is similar to the situation in Lake Ontario studied by Cook et
al. (1990). In this case Rws and R^ <1 because transfer of TCDD to
the water column or to biota from sediment will be limited
kinetically although the fugacity of TCDD is higher in sediment
than water. BSAFs from Lake Ontario would be most applicable to
this scenario.
The time course for the buildup of TCDD levels will be very
much dependent on the hydrodynamics and especially on sediment
transport dynamics. Assuming Koc values of 7 or more, almost all
TCDD would be expected to be deposited in deep sediments. Also
critical to understanding the bioavaliability of the deposited TCDD
is the depth of mixing zone of benthic invertebrates (this can be
determined with 210Pb and other isotopes) because freshly deposited
and buried sediments could be remobilized.
C-208
-------
References not cited in the interim report
Birkholtz, D.A., S. Swanson and J.W. Owens. PCDD, PCDF and EOCL
bioaccumulation in a northern Canadian river system. In:
Proceedings of the 12th Int'l Symposium on Dioxins and related
Compounds. 8:313-314.
Clark, K.E., F.A.P.C. Gobas and D. Mackay. 1990. Model of organic
chemical uptake and clearance by fish from food and water.
Environ. Sci. Technol. 24, 1203-1213.
Fairchild, W.L., D.C.G. Muir, R.S. Currie and A.L. Yarechewski.
1992. Emerging insects as a biotic pathway for movement of
2,3,7,8-tetrachlorodibenzofuran from lake sediments. Environ.
Toxicol. Chem. 11, 867-872.
Formica, S.J., J.A. Baron, L.J. Thibodeaux and K.T. Valsaraj. 1988.
PCB transport into lake sediments. Conceptual model and
laboratory simulation. Environ. Sci. Technol. 22, 1435-1440.
Friesen, K.J., W.L. Fairchild, M.D. Loewen, S.G. Lawrence, M.H.
Holoka and D.C.G. Muir. 1993. Evidence for particle-mediated
transport of 2,3,7,8-tetrachlorodibenzofuran during gas
sparging of natural water. Environ. Contam. Toxicol. In
press.
Gobas, F.A.P.C. 1992. Modelling the accumulation and toxicity of
organic chemicals in aquatic food chains. In: F.A.P.C. gobas
and J.A. McCorquodale Eds. Chemical Dynamics in Freshwater
Ecosystems,Lewis Publishers, Ann Arbor, pp. 129-152.
Gobas, F.A.P.C. 1993. Personal communication. Simon Fraser
University, Burnaby BC
Harding, L.E. and W.M. Pomeroy (1990), Dioxin and furan levels in
sediments, fish and invertebrates from fishery closure areas
of coastal British Columbia. Regional Data Report, Environment
Canada, Environmental Protection, Pacific and Yukon Region,
North Vancouver, B.C. 67 pp + Appendix.
Hinton, S. NCASI/Tuft's University, Boston MA. Presented at the 6th
Colloquium on Pulp & Paper Mill effluents, University of
Toronto, Dec. 10, 1991.
Holloran, M. 1993. Personal communication. Beake Engineering Ltd.,
Richmond BC
Krishnappan, B.G. 1981. Unsteady, nonuniform, mobile boundary flow
model - MOBED. Environment Canada. 107 pp.
Krishnappan, B.G. 1993. Personal communication. Environment Canada,
NWRI, Burlington Ont.
Marmorek, D.R., J. Korman, D.P. Bernard and T. Berry. 1992.
Ecosystem fate and effects of pulp mill effluents in the
Fraser River. Identification of research and monitoring
priorities. Environment Canada, Vancouver BC, 151 pp.
Norstrom, R.J., M. Simon, C. Macdonald, P. Whitehead, J.E. Elliott,
D.C.G. Muir, C. Ford, K. Langelier. 1991. Food chain transfer
and sources of PCDD, PCDFs and PCBs in the Strait of Georgia.
Presented at SETAC, Nov. 1991 #133.
Servos, M.K., D.C.G. Muir and G.R.B. Webster. 1992. Environmental
fate of chlorinated dibenzo-p-dioxins in lake mesocosms. Can.
J. Fish. Aquat. Sci. 49, 722-734.
U.S. EPA (1990), Risk assessment for 2,3,7,8-TCDD and 2,3,7,8-TCDF
contaminated receiving waters from U.S. chlorine bleaching
pulp and paper mills. Report prepared for U.S. Environmental
Protection Agency, Office of Water Regulations and Standards,
Washington, D.C. Aug. 1990
C-209
-------
-------
Robert Pastorok
PTI Environmental Services
C-211
-------
-------
PREWORKSHOP COMMENTS
Workshop on Ecological Risk Assessment Issues
for 2,3,7,8-Tetrachforod/benzo-p-DIoxfn (TCDD)
Comments by: Robert A. Pastorok
PTI Environmental Services
15375 SE 30th Place
Suite 250
Bellevue, Washington 98007
CONCEPTUAL MODEL ISSUES
Issue 16: Consistent with the Interim Report, the conceptual model focuses on
effects on fish and wildlife that consume fish. Comment on whether this
approach captures the full range of potential ecological effects for this scenario.
The focus on fish and wildlife for the quantitative risk assessment is appropriate
for development of criteria. However, effects on plankton, macrophytes, and
benthic macroinvertebrates should be addressed at least qualitatively, if not
quantitatively. The conclusion that these groups are less sensitive to chlorinat-
ed dioxins than are fish and wildlife may be warranted based on the interim
report, but this does not imply that these groups are not themselves at
significant risk. Therefore, secondary effects on fish and wildlife could be
realized through changes in species composition or community structure of
plankton, macrophytes, and macroinvertebrates. The interaction of plankton,
macrophytes, and suspended solids is especially important in reservoirs. The
water-level fluctuation in the reservoir must also be considered because of its
potential effect on the abundance and species composition of macrophyte beds.
The focus on sport fishes alone may not be warranted. Effects on other fish
species could result in indirect effects on the sport fishes of interest.
C-213
-------
Before certain organism groups are dismissed from the conceptual model, they
should be ranked relative to their exposure susceptibility, sensitivity to PCDDs
f^>
and PCDFs, and their importance as food for key receptors, such as predatory
fish, mammals, and birds.
Issue 17: The Interim Report emphasizes using tissue residue levels to estimate
the adverse effects of TCDD. However, to do the risk assessment outlined by
the conceptual model, it will be necessary to link predicted loadings of TCDD
in the paper mill effluent to residues in the organisms identified in the assess-
ment endpoints. Discuss the utility of available risk assessment tools for
accomplishing this goal.
The most important tools needed to link tissue residue levels with discharge
rates for PCDDs and PCDFs are:
• Sediment BSAFs developed from empirical data or a bioacc-
umulation model similar to that of Thomann et al. (1992) to
link sediment concentrations of PCDDs and PCDFs to tissue
residue levels (This model represents a state-of-the-art
approach to estimating BSAFs.)
• A model such as WASP4 to link discharge rates of PCDDs
and PCDFs to allowable sediment concentrations in the
reservoir.
The bioaccumulation model of Thomann et al. (1992) could be applied to
develop permit specifications, which is the stated risk management objective
for the scenario. Allowable tissue residue levels in selected fish species would
be specified from available information on NOAELs for selected receptors and
endpoints (e.g., fish embryo toxicity, fish food in the diet of mink). However,
several important questions remain about application of the approach, including:
C-214
-------
• Which data sets should be used to develop NOAELs?
• Should a quantitative uncertainty analysis (e.g., Monte Carlo
analysis) be applied to the bioaccumuiation model to derive
probability distributions for allowable tissue residue levels
(i.e., the selected NOAELs), BSAF values, and allowable
sediment concentrations of PCDDs and PCDFs?
• How will the model be validated?
• How will spatial variability in parameters of the model be
taken into account during validation and application of the
model?
The issue of spatial variability and spatial analysis of data in the development
of the approach and its application is especially important. Using literature
data, realistic assumptions need to be developed for home ranges and diets of
mammalian and avian wildlife to be addressed in the model. The exposure
factors handbook for ecological receptors that is being developed by EPA's
exposure assessment group may be useful in assigning parameter values.
Proper design of a sampling and analysis program to validate the model for fish
is essential. The model should be validated using data from areas where the
water column contamination is primarily influenced by the sediment compart-
ment rather than direct discharges from multiple sources. During model
validation, the sampling program will need to consider explicitly the fish home
range and the spatial heterogeneity of sediment contamination.
Finally, the model as specified by Thomann et al. (1992) and the conceptual
model for the reservoir scenario appear to ignore direct ingestion of sediment
by fish. This exposure pathway could be especially important for forage fish
such as English sole that feed on benthic invertebrates and incidentally ingest
C-215
-------
sediment. Ecological risk assessments for terrestrial wildlife species have
demonstrated the relative importance of direct soil ingestion compared with
other exposure pathways for persistent contaminants. Similarly, sediment
ingestion may be an important pathway for exposure of fish to contaminants
compared with the water and food routes. If data to quantify sediment
ingestion by fish are not available in the literature, then a focused laboratory or
field study may be needed to support development of the model.
WASP4 is capable of modeling all of the conditions and processes potentially
affecting the reservoir, including multiple types of solids and chemicals, lateral
transport of particulate and dissolved phases, vertical settling of particles,
particle resuspension, horizontal bed load transport, adsorption/desorption in
both the water column and sediment, pore water diffusion, spatially variable
sediment mixing depths, and spatial and temporal variability in both conditions
(e.g., initial chemical concentrations) and processes (e.g., sorption equilibria).
WASP4 can be configured to use a network of compartments (or boxes) that
accurately depicts physical site conditions, and can be applied in one, two, or
three dimensions. Site-specific data on current velocities, particle composition
and settling velocities, sediment accumulation rates, and sediment contaminant
concentrations can be collected during field investigations to support use of the
WASP4 model in a detailed risk assessment. Additional site-specific informa-
tion that would improve model accuracy includes dispersion coefficients.
Because the WASP4 model is essentially a forward-mode fate and transport
model, it does not have the capability to back-calculate allowable source
loadings based on allowable sediment levels determined from the bioaccumula-
tion model and the selected NOAELs for tissue residues in fish. Therefore, the
WASP4 model would have to be run in an iterative fashion to find a convergent
solution.
C-216
-------
issue 18: The Interim Report describes the limited field data that are available
for estimating BAFs and BSAFs. Discuss the applicability of these factor? to
the Omigoshee Reservoir conceptual model.
Application of BAFs for PCDDs and PCDFs is limited by the current inability to
measure PCDDs and PCDFs at low concentrations jn water. Thus, the key
issue in applying the model to develop effluent permit specifications is the use
of BSAFs. The BSAFs for various PCDDs and PCDFs will need to be developed
before complete ecological risk assessments can be conducted. BSAFs could
be developed for each primary food item of the key receptors used to derive
permit specifications. If the apparent constant relationship between the BSAF
for PCBs and the BSAF for 2,3,7,8-TCDD mentioned in the interim report can
be confirmed through additional research, then an empirical BSAF for PCBs
could be determined for the case study reservoir, and the BSAF for TCDD could
be estimated based on a wider database on the ratio of PCB-BSAF to TCD.D-
BSAF from systems with both groups of chemicals. BSAFs for PCDDs and
PCDFs other than 2,3,7,8-TCDD could be derived in a similar manner.
Limitations of the BSAF approach include the inability to evaluate the residues
for any receptor groups that are absent due to TCDD effects and the inability
to measure BSAFs in the system of interest before the discharge is initiated.
Issue 19: The temporal dynamics and disequilibrium situations commonly
associated with TCDD are mentioned in the Interim Report (Section 2.3).
Comment on how these aspects should be considered in establishing 1) the
time course for the buildup of TCDD levels following initiation of the paper mill
discharge and 2) the time course for the decrease of TCDD levels and recovery
of biota should the paper mill cease operation.
It is not clear why the issue of temporary disequilibrium under field conditions
is relevant to the stated risk management objectives for the reservoir scenario.
C-217
-------
The permit specifications for discharge limits on PCDDs and PCDFs should be
developed from steady-state models (or empirical determinations of BSAFs from
data on other systems). These approaches will not apply to the buildup of
PCDDs or natural recovery.
REFERENCES
Thomann, R.V., J.P. Connolly, and T.F. Parkerton. 1992. An equilibrium
model of organic chemical accumulation in aquatic food webs with sediment
interaction. Environ. Toxicol. Chem. 11:615-629.
U.S. EPA. 1992. Estimating exposure to dioxin-like compounds. EPA/600/6-
88/005B. U.S. Environmental Protection Agency, Office of Research and
Development, Washington, DC.
U.S. EPA. 1993. Interim report on data and methods for assessment of
2,3,7,8-tetrachIorodibenzo-p-dioxin risks to aquatic life and associated wildlife.
EPA/600/R-93/055. U.S. Environmental Protection Agency, Office of Research
and Development, Washington, DC.
C-218
-------
Ecological
Risk Assessment
Problem Formulation
Define Ecological
Receptors, Endpoints,
and Chemicals
Home Range
and Diet Analysis
Develop Fish Tissue
Residue NOAELs for
Fish and Wildlife
Receptor Species A/Life Stage A/
EndpointA
Receptor Species A/Life Stage B/
Endpoint B
Receptor Species B/Life Stage A/
EndpointA
— Etc.
For Each Fish Species
Select Lowest Fish Tissue
NOAEL Among Receptors
Empirical
BSAFs
Additional
Policy Input
Define Allowable
Fish Tissue Residues
Bioaccumulation
Model
Define Allowable
Sediment Concentrations
Fate/Transport
Model (e.g., WASP4)
Define Allowable
Discharge of PCDDs/PCDFs
Figure 1 . Development of discharge permit limits for PCDDs/
PCDFs based on ecological risk assessment.
PTI
ENVIRONMENTAL SERVICES
OH090893
C-219
-------
RECEPTORS
Species/Habitat
Inventory
Guild
Analysis
SPECIES ANALYSIS
• Spatial Distribution
• Life History
• Dietary Patterns
• Soil/Water Intake
• Activity Patterns
Data Availability
Assessment
MODELING
4-*
W3$r:'^'>
' f '
Estimate
Site-Specific Risk
CONTAMINANTS
Transport/Fate
Analysis
Chemical Data
Water/Sediment/
Soil/Tissue
Derive Allowable •;
"\ Sediment Concentrations'.
of PCDDs/PCDFs
Figure 2. Application of proposed terrestrial food web
exposure model.
PTI
ENVIRONMENTAL SERVICES
OH090393
C-220
-------
Section 5
J.P. GIESY
C-221
-------
-------
Workshop on the Ecological Risks of
2,3,7,8-Tetrachlorodibenzo-p-dioxin
Minneapolis, Minnesota
Sept. 14-15, 1993
Preconference Comments
J.P. Giesy
Dept. Fisheries & Wildlife
Michigan State University
E. Lansing, Ml, 48824-1222
Tel: (517)353-2000
FAX: (517) 353-1699
C-223
-------
Giesy Briefing Document, Minneapolis, MN Sept 14-15, 1993
I have reviewed the document entitled Interim Report on Data and Methods for
Assessment of 2,3,7,8-Tetrachlordibenzo-p-dioxin Risks to Aquatic Life and
Associated Wildlife. I found this document to be comprehensive and very well
organized and written. I have compared the results of the model calculation of
environmental concentrations associated with ecological risk. While similar, my
determinations of safe concentrations are different from those proposed by the
authors of the document. I have included a discussion of my criteria and calculations
and compared those to both the literature and to the EPA document.
I have also reviewed the document containing the three Workshop exercises
and my comments are also enclosed. I have provided answers to each of the
questions posed by the organizers for this working group.
Issues for Consideration
Question 1.
It is true that plants, invertebrate and amphibians should, based on current
models of the mode of action of TCDD be less sensitive than fish, birds and fur
bearers. Focusing on fish will adequately provide protection of other species if the
accumulation of TCDD into organisms that eat the fish is considered. Based on
information currently available, I believe that the risk assessment should focus on fish,
fish-eating birds (such as eagles) and furbearers (such as mink).
It is not appropriate to assume that the use of survival of fish eggs and fry will
be protective of other organisms. Our most recent research has indicated that
reproduction is not necessarily the most sensitive endpoint for the effects of TCDD
to fish. In fact our work with rainbow trout has demonstrated that the adults were
much more sensitive than effects on fry. Thus, it is important to include long-term
studies which use dietary exposure to determine effects.
Our research has indicated that there is uncertainty in determining
concentrations which can be related to ecological risk and that the concentrations
proposed by the Interim document would not be sufficiently protective of fish and
wildlife from ecological risk.
Here, we describe the current concentrations of TCDD-EQ in tissues of birds of
the Great Lakes Region and discuss their relationship to egg lethality and birth defects.
We also discuss the interactions among some of these A/7-r-active compounds, their
relative toxic potencies, the relative importance of the various congeners and other
possible contributors to the total toxicity of complex mixtures. We also describe and
compare several methods of assessing the toxic potency of complex mixtures to birds
and compare the results of hazard assessments for wildlife with those to protect
human health.
C-224
-------
Giesy Briefing Document, Minneapolis, MN Sept 14-15, 1993
Question 2
Chlorinated Hydrocarbons in the Environment
Many persistent, synthetic, PCH have been released to and are widely distributed in
the environment.'36"401 Some of the PCH are very persistent/35'41' and are
bioconcentrated and biomagnified.'35'39'41"44' Of these, some classes cause adverse
effects at minute concentrations in biota.'45"48' Some of the major groups which are
of environmental significance are the polychlorinated tiibenzo-p-dioxins (PCDDs),
polychlorinated dibenzofurans (PCDFs), and polychlorinated biphenyls (PCBs).(47'48)
Together the PCBs, PCDDs and PCDFs represent 419 individual congeners. Within
this overall class are several subclasses defined by the pattern of substitution of the
chlorine atoms.(39'43'48'49> One subset of these groups of compounds, those that are
both laterally substituted and either non- or mono-ortho substituted can attain a planar
configuration and are referred to as planar or co-planar congeners (p-PCH) of the.(48)
Our discussion will consider only these p-PCH.
Complex Mixtures
It is difficult to understand or predict potential effects of complex environmental
mixtures of PCH on biota because, hot only are there a great number of compounds,
but because their concentrations of the individual components change as a function
of space and time. Thus, the mixture to which organisms are exposed at one time or
at one location may be very different from that to which they are exposed at other
times or locations.'22'50'51' Furthermore, the relative concentrations Of the various p-
PCH congeners is different from trophic level to. trophic level.'52' These differences
are caused by environmental "weathering" and the sorting of compounds, based on
their solubilities, volatilities and rates of degradation. The result is mixtures in the
environment which not only change spatially and temporally, but which are different
from the technical mixtures which were released into the environment. Thus, at this
time, it is impossible to use the results of studies, which have determined the dose-
response relationships of technical mixtures under laboratory conditions to predict
effects in real world wildlife. The study of effects of p-PCH on fish and wildlife was
limited for two decades by the fact that it was impossible to assess the toxicological
implications of constantly changing mixtures and need to monitor for total
concentrations of p-PCH. Recently, greater understanding of the mechanisms of toxic
action of the p-PCHs has made it possible to express the potency of mixtures of p-
PCH to elicit adverse effects relative to one prototype p-PCH. When this approach
has been used, it has been possible to obtain better correlations between observed
effects and 2,3,7,8-tetrachlor-dibenzo-p-dioxin equivalents (TCDD-EQ, sometimes
C-225
-------
Giesy Briefing Document, Minneapolis, MN Sept 14-15, 1993
referred to as TEQ) than could be obtained for single p-PCH congeners of classes of
p-PCH.<48)
Toxlclty of p-PCH
While the suite of effects caused by TCDD may seem diverse and unrelated, they are
thought to all be caused through a common mode of action.<52) The mechanism is
receptor mediated and involves the binding of particular intracellular receptors which
lead to a host of common cellular and subcellular responses and subsequently effects
on the whole animal. The similarities of molecular structure and conformation of the
p-PCH compounds result in similar toxic effects.140'46"48' Their primary toxic effects
are thought to be exerted through a common mode of action.(B3"B5) The most widely
accepted proposed mechanism of action for p-PCHs involves the expression of their
biological potency through a specific cytosolic receptor, the A/I receptor (A/»-r).(B5~57>
This receptor binds p-PCHs, which are approximately 3 X 10 A in size, such as planar
PCBs (p-PCBs), PCDDs and PCDFs, with differing affinities.'48' The resulting receptor-
iigand complex is then translocated to the cell nucleus where it elicits specific changes
in gene expression.(48'56'58'B9) The transformed receptor binds to sequences of DNA
called the dioxin responsive enhancers (ORE). There are a number of ORE in the
genome of most animals, thus a number of genes can be affected.(60> Many of the
observed toxic effects of the p-PCHs are attributable to specific alterations in gene
expression.(48'B3'BB'6V64) The relative toxicity of individual p-PCH is directly
proportional to the strength of binding to the A/?-r and the potential to induce
cytochrome P4501A isozyme activity.'46"48'57'65'66' The p-PCH are generally not
acutely toxic, but cause chronic toxic responses/48'59'67'68' including impaired
reproductive potential offish-eating, water birds. Because the p-PCH are not generally
acutely toxic and accumulate into top predators, such as birds, colonial water birds
have been suggested as useful biological monitors for the accumulation and effects
of p-PCHs in the Great Lakes ecosystem.(24'66'69'70''
The most characteristic of these responses include measurement of mixed
function oxidases. These enzymes are part of the general detoxification response
animals to toxic substances. These enzymatic responses are generalized and lead to,
or are associated with side-effects on many critical substrates used in routine
metabolism. Levels of hormones, vitamins, and by-products of normal cellular
activities are often altered enough to produce a characteristic set of responses now
recognized as symptoms of subchronic or chronic exposures to p-PCH (Table 2).<8)
There are a number of pleiotropic effects of TCDD and p-PCBs on organisms.
These effects can be direct or secondary responses to gene regulation. Some of these
responses can be directly related to the adverse effects observed, while some are
useful as biomarkers, but their relationships to observed adverse effects are less well
characterized. The most subtle and important biological effects of TCDD and the
dioxin-Iike p-PCB congeners on wildlife are their effects on endocrine hormones and
C-226
-------
Giesy Briefing Document, Minneapolis, MN Sept 14-15, 1993
vitamin homeostasis.(71> TCDD also mimics the effects of thyroxine as a key
metamorphosis signal during maturation/621 TCDD has also been shown to down-
regulate the epidermal growth factor receptor/721 which may result in disruption of the
patterns of embryonic development at critical stages.
Altered concentrations of thyroid and steroid hormones and vitamin A are
frequently reported to co-occur with embryonic abnormalities in wildlife populations
exposed to p-PCHs.(8> Individuals from these exposed populations have been
observed to have altered sexual development/71'731 sexual dysfunction as adults and
immune system suppression/56'741 The observations on adult sexual dysfunction are
especially significant since young, which appear to be normal while raised by
intoxicated parents may become reproductively dysfunctional when they mature/75"791
Poor reproductive efficiencies and adventive, opportunistic diseases are characteristic
of the wild animals in these exposed populations of the Great Lakes region/801
Because of these conserved biochemical mechanisms concentrations of TCDD-EQ
correlated with egg lethality or birth defects in populations of colonial, fish-eating,
water birds while concentrations of the total concentrations of PCBs, PCDF and PCDD
do not/22-81'821
Vitamin A (retinol) is important in many functions in animals, such as embryonic
development, vision, maintenance of the dermally derived tissues, immune
competence, hemopoiesis and reproductive functions/83'851 Vitamin A is necessary
for normal embryonic development'861 and, thus, changes in the status of vitamin A
in the plasma or liver may be responsible for the birth defects observed in birds, which
have been exposed to p-PCH. Laboratories studies have determined that both vitamin
A and its storage form in the liver (retinal palmitate) were depleted in birds exposed
to sublethal doses of the dioxin-like, p-PCB congener 77/87> We have observed an
inverse correlation between the concentration of vitamin A in serum and
concentrations of p-PCHs in tissues of birds from the upper Great Lakes/24'691
p-PCH affect concentrations of vitamin A in both the blood and liver of exposed
organisms. These effects are thought to be due to least two processes. In blood,
some of the hydroxylated metabolites of PCBs to the carrier protein transthyretin/841
In the liver, induction of hepatic enzymes such as acyl-CoA:retinol acyltransferase and
uridine diphosphate giucuronysl transferase (UDPGT) is thought to alter the metabolic
pathways involved in the storage and mobilization of vitamin A, and results in the
observed depletion of retinols in the liver/83'851
TCDD is known to have effects on both male and female steroid hormones/87*
891 For instance, TCDD is both estrogenic and antiestrogenic effects, in different
tissues, depending on timing of exposures during development/891 Furthermore,
2,3,7,8-TCDD is a potent thyroxine agonist which may account for its capability to
cause wasting syndrome in homeotherms/67'851 The induction of the mixed function
monooxygenase system can also reduce the concentrations of circulating steroid
hormones, which can have adverse effects on the reproduction of wildlife/251
Thyroid hormone, which is an important regulator of development and
C-227
-------
Giesy Briefing Document, Minneapolis/ MN Sept 14-15, 1993
metabolism can be influenced by exposure to p-PCH.(67'86"89) There are several
possible mechanisms for the observed effects on circulating T3 and T4. First hydroxy-
substituted PCB congeners have been observed to displace thyroxin (T4) from its
carrier protein, Transthyretin (TTR; prealbumin), which results in effects similar to
thyroxin deficiency.(87"89) p-PCHs can induce UDPGT activity in the liver, which then
decreases the concentration of TTR in the blood. Concentrations of TTR are not
determined directly in the plasma, but rather, T4 binding capacity is measured.
Therefore, it is not possible to distinguish which of the two mechanisms may be
causing the observed effects.
There is also evidence that exposure to TCDD can shift normal carbohydrate
dominated metabolism to a fat metabolism, causing afflicted individuals to be unable
to utilize a major source of energy. This effect is thought to be due to inhibition of
the synthesis of the glucose carrier protein by p-PCH.(24/90)
Methyl sulfone metabolites of p-PCH have also been observed to cause adverse
effects. Methyl sulfones accumulate in lung tissue of some marine mammals and are
thought to be responsible for some toxic effects.(81>
In addition to the effects of the pPCBs it is known that the di-ortho-substituted
PCBs, which are not very toxic, due to effects, which are mediated by the A/?-r can
cause adverse effects in wildlife. Specifically, these congeners can inhibit dopamine
synthesis in the brain, which results in behavioral differences in mammals.(92"9B> We
have observed behavioral effects in colonial water birds, which may be caused by
these types of effects/965 but to date, little information is available on this
phenomenon. We feel that it is unlikely that the effects of the di-ortho substituted
PCBs are responsible for the observed birth defects, but may be important in subtle
behavioral shifts.
QSAR
The most potent of the p-PCHs, identified thus far, is TCDD. The relative potency of
other p-PCH congeners to cause biochemical or toxicological effects mediated through
the A/7-r mechanism, is determined by the pattern of substitution of the chlorine
atoms on p-PCH.(48) The most potent congeners are those that have at least four
chlorine atoms in at least two of the lateral (meta and para) positions of both of the
phenyl rings.(4s) The relative potency can be expressed as proportions relative to
TCDD and are referred to as Toxic Equivalency Factors (TEF).<47'48'82'96) The
concentrations of individual p-PCH congeners can therefore be measured for Ah-r-
mediated potency and reported as TCDD-Equivalents (TCDD-EQ).(48'49) The potency
of complex mixtures, which can include several of the more A/>-active congeners in
addition to many less active congeners, can also be expressed relative to TCDD as
TCDD-EQ.(83/94) One method involves the quantification of each p-PCH congener in
a complex mixture from a biological sample. The potency of the mixture is then
C-228
-------
Giesy Briefing Document, Minneapolis, MN Sept 14-15, 1993
calculated by multiplying the concentration of each congener by its TEF value and
summing the products."0'60-42'48'90'93-961
Biochemical Response Assays
There are a number of limitations associated with the determination of TCDD-EQ in
complex mixtures from instrumental analysis and application of TEFs. The large
number of p-PCHs in environmental samples make the chemical analysis time
consuming and expensive. In addition, the most important congeners occur at small,
but lexicologically relevant concentrations that are sometimes difficult to quantify by
routine procedures. The wide range of biological potencies for p-PCHs(48) and
potential synergistic or antagonistic interactions among p-PCH congeners and other
chemicals <46'48'97-100) suggest that an additive model of toxicity is generally adequate
but may not always appropriate.4101' Interactions between or among p-PCH congeners
and other compounds in the mixture may further complicate interpretation of the
toxicological significance of these mixtures to wildlife. In addition, assessment of
possible toxic effects of mixtures on wildlife are complicated by the wide range of
reported TEF values for different species and various toxic endpoints used to set TEF
values.(48) Depending on the test species and chosen endpoints the use of different
TEF values will result in different calculated TCDD-EQ concentrations/102'1031
Furthermore, TEFs are not available for all biological endpoints and most have been
derived for those species amenable to laboratory studies.
The concentrations of TCDD-EQ in complex mixtures of p-PCHs may also be
determined by use of in vitro cell systems in a manner analogous to a chemical
detector.(82) One method which uses H4IIE rat hepatoma cells relies on the fact that
p-PCHs induce specific cytochrome P450-mediated monooxygenase (MO) enzymes
through the A/i-r-mediated mechanism.'82'83'94'104'106' Furthermore, relative induction
of MO activity by individual p-PCHs, as well as by mixtures of these compounds, are
correlated with their toxicity to certain species [Safe, 1986,1987, 1990], particularly
birds.'46-48'76'79' Therefore, induction of cytochrome P450-mediated MO enzymes
integrates the concentration and potency of all of the p-PCH congeners present in
complex, environmental mixtures. This induction of enzymes has been well
characterized in vitro with rat hepatoma cells (H4IIE cells) upon their exposure to p-
PCH-containing extracts of environmental samples/104"1061 This induction measures
the potency of an extract,*22'82'93'94'103'107-109' which can be expressed as TCDD-EQ
by comparing the dose-response of an unknown extract to a standard curve generated
with TCDD.<82)
The H4IIE bioassay method of TCDD-EQ determination has both advantages
and disadvantages relative to the use of chemical analysis and TEF values. The
bioassay is more rapid and less costly than congener-specific chemical analysis. Since
the bioassay is a mechanistically-based determination of an integrated biochemical
C-229
-------
Giesy Briefing Document, Minneapolis, MN Sept 14-15, 1993
response, the results can be expected to be more biologically relevant. The bioassay
integrates possible interactions between p-PCH congeners and compounds of other
chemical classes thereby providing an integrated measure of potency measured at a
cellular site proximate to the site of action. One limitation to the use of bioassay
systems is that culture conditions and the use of various carrier solvents or end-points
may make comparison of results among different research groups difficult.(48'82)
Also, results obtained using the H4IIE bioassay have only been calibrated against
controlled laboratory studies with rodents. This makes it difficult to interpret the
toxicological implications of concentrations of TCDD-EQ observed in wildlife species.
Only limited correlations are demonstrated in field studies of wildlife species'221 and
only a few controlled laboratory studies with individual p-PCH congeners have been
conducted.
Until recently only limited comparative data between the instrumental and
bioassay approaches have been available for complex mixtures of p-PCHs in samples
collected from the environment.1741 More recently we have determined the TCDD-EQ
by both instrumental and H4IIE bioassay analysis of a set of p-PCH-containing extracts
from birds and their eggs collected at Green Bay, Wisconsin, U.S.A.'102'1091 It was
found that the TCDD-EQ determined by the instrumental and bioassay techniques
were positively correlated, but that the use of the results of the instrumental analyses
and application of TEFs in an additive model underestimated the TCDD-EQ measured
in the H4IIE bioassay (Fig. 2). The two possible reasons for the observed differences
were: 1} The mixtures were interacting synergistically; or 2) There were compounds
present that were not quantified yet contributed to the response of the H4IIE cells.
It is likely that compounds other than the PCBs, PCDDs, and PCDFs, were responsible
for a greater concentration of TCDD-EQ in the H4IIE assay than could be accounted
for by an additive model, which considered the compounds measured. This
discrepancy was not caused by the use of inappropriate TEFs or exposure systems,
since the TEFs used were derived by using the same techniques in the H4IIE system.
This difference is thought to be due to the fact that they assay responds to all of the
p-PCH compounds, while we only quantified the PCDD, PCDF and planar PCB
congeners. Concentrations of TCDD-EQ determined by calculation using TEFs or
H4HE bioassay have been correlated with adverse effects in birds.(10'22'67'107)
Relative Contributions of Individual p-PCH
The proportion of TCDD-EQ contributed by PCDD and PCDF congeners in
environmental samples from the Great Lakes region is generally small, frequently less
than 5% in the fish-eating bird species (Figs. 3 & 4). A similarly great relative
importance of pPCBs has been reported in certain marine mammals.(41) This indicates
that marine ecosystems are also strongly influenced by the planar PCBs with toxic
responses mediated through the A/>-receptor mechanism. In contrast to the fish-
C-230
-------
Giesy Briefing Document, Minneapolis, MN Sept 14-15, 1993
eating birds in the Great Lakes, the 'terrestrial' avian species examined in a Green Bay
studied contained lesser concentrations of TCDD-EQ<102) and the relative proportion
of TCDD-EQ, which was contributed by PCDDs and PCDFs in these species ranged
from 3.2% to 71 % and was greater than the fish-eating species (Fig. 4). However,
the absolute concentrations of TCDD-EQ contributed by PCDDs and PCDFs were
similar to the piscivorous species.1102'1091 This suggests that contamination with
PCDD and PCDF can be widespread in ail species. In the fish-eating water birds the
accumulation of TCDD-EQs from PCB congeners due to water and food by forage fish
is an additional trophic level transfer and thus provides greater bioaccumuiation
potential than the terrestrial species. Also, there are known local sources of PCBs,
whereas the sources of PCDD and PCDF may be more distant and due to atmospheric
deposition.
The majority (> 90%) of the TCDD-EQ in the eggs of cormorants and terns in
the Great Lakes was due to p-PCBs,<102) rather than PCDD or PCDF, which accounted
for between 2 and 9% (12-22 ppt) of TCDD-EQ measured in water bird eggs in Lakes
Superior, Huron and Michigan. The primary contributions to the TCDD-EQ were due
to the dioxin-like p-PCBs, especially non-ortho-chlorinated PCB congeners 126
(3,4,5,3'4'- PeCB), 77 (3,4,3',4'-TCB), 169 (3,4,5,3',4',5'-HCB) and mono-ortho-
chlorinated congeners 105 (2,3,4,3/,4'-PeCB) and 118 (2,3',4,4',5-PnCB). The
understanding that dioxin-like bioeffects in fish-eating, colonial water birds are due
largely to pPCBs is an emerging consensus worldwide, except near TCDD point
sources.(41'50)
Currently, much of the discussion Of the safety of consuming fish flesh from
the Great Lakes is centered on the concentration of TCDD-EQ contributed by the
PCDD and PCDF.<84) However, the concentrations of TCDD-EQ from both PCDD and
PCDF is generally in the range of 5 to 10 pptr, wet weight, while concentrations of
TCDD-EQ contributed by the PCB congeners can be as great as 250. pptr.(110)
Interactions among p-PCH and between p-PCH and Other Compounds
There has been much discussion about the possible interaction between and among
individual congeners of p-PCH and between p-PCH classes and other synthetic,
halogenated compounds in extracts of environmental matrices, which contain complex
mixtures of p-PCH.<111> An additive model for the prediction of TCDD-EQ is plausible
and it is unlikely that the use of this model will result in a great deal of error in
predicting the concentrations of TCDD-EQ due to synergisms. While such non-
additive responses could be either greater or less than additive, it seems that the
biochemical effects of p-PCH congeners are simply additive.(48) However, there are
reports of both infra- and supra-additivity between and among individual p-PCH
congeners, complex mixtures of p-PCH and other halogenated
hydrocarbons.*48'67'62'64'112)
When nonadditive responses between individual p-PCHs and complex mixtures
C-231
-------
Giesy Briefing Document, Minneapolis, MN Sept 14-15, 1993
have been observed, they have generally been antagonistic.'75'100'113"11B) This
interaction is probably due to the fact that the less A/j-r-active congeners still have
receptor binding affinities, which when combined with the relatively great
concentrations of these less toxic congeners, make them effective competitors for
binding sites.(114/116) This reduces the probability of the more toxic p-PCH binding to
the A/?-r. However, these less active congeners do not seem to bind with great
enough affinity to be effective inducers of EROD induction or cause any of the other
adverse effects, which are caused by the A/i-r-active congeners.<48) The di-ortho-
substituted PCB congener, 2,2',4,4',5,5' has, under some conditions, been found to
be an effective antagonist for 2,3,7,8-TCDO, but some of the results of the studies
of interactions among congeners are equivocal. For instance, PCB congener 77 and
TCDD caused greater than additive induction of AHH activity in the liver of rainbow
trout at doses which corresponded to 0.13 and 0.5 toxic unit (proportion of dose
required to elicit a given endpoint). However, at greater doses the mixture was less
than additive.1116) Similarly, there was no effect of of PCB congener #153
(2,2',4,4',5,5') on the induction of EROD activity by TCDD.(117) PCB congener 153
was found by the same researcheres to have an antagonistic effect on the potential
of TCDD to induce PROD activity/117' Pretreatment of mammalian cells results in an
increase in the concentration of Ah-r, which could cause a greater response to TCDD,
but this is only observed at submaximal exposures to TCDD.
The evidence seems to support the conclusion that complex mixtures of
halogenated p-PCH congeners in the extracts from the wildlife would be less than
additive (antagonistic) rather than more than additive (synergistic). A number of
compounds, which can bind to the A/?-r, but do not effectively induce the same
momooxygenase (P450IA1) enzyme activity as 2,3,7,8-TCDD, are partial
antagonists.'87"100) The potency of simple combinations and complex mixtures of
PCDD and PCDF to induce EROD activity in in hepatocytes or H4IIE cess indicated
that the effects were due to the simple additive responses to the 2,3,7,8-substituted
congeners.018) Therefore, it is not likely that non-additive interactions between or
among A/?-r-active or inactive congeners would result in differences between the
concentrations of TCDD-EQ derived by instrumental or bioassay methods. In fact, if
there were strong antagonisms between the p-PCH and other components of these
complex mixtures one would expect the TCDD-EQ determined in the bioassay to be
less, not more than those calculated from the additive model, which was the case
when the results of the two methods were compared/102'
Selective Enrichment of p-PCH
When studying a complex mixture, such as p-PCH, if the relative proportions of the
different congeners changes during the bioaccumulation process this change must be
accounted for in the hazard assessment. It is difficult to determine safe exposures if
concentrations observed in the field can not be compared to the results of controlled
C-232
-------
Giesy Briefing Document, Minneapolis/ MN Sept 14-15, 1993
laboratory studies, which have been conducted with the original technical mixtures
of the compounds. Chemical weathering due to differential solubilities, volatilities,
adsorption constants and degradation rates, can result in patterns or relative
concentrations of p-PCH congeners which are different from the technical mixtures
and also different from one location to another.'118) Furthermore, these patterns can
change over time(120) such that the pattern of congeners observed is significantly
different than that in the original technical mixtures, which were released to the
environment.'119) Also, there are changes in the relative pattern of accumulation in
the ecosystem as trophic biomagnification occurs/102'121'122'
In the Great Lakes system, one way to estimate the relative toxic potency of
mixtures is to calculate TCDD-EQ, which measures the total concentration of
congeners from PCDD, PCDF and PCBs and divide this by the total concentration of
PCBs.122'62'102' This relative potency ratio will account for different contributions from
the three major classes of congeners and relate it to that fraction, which is thought
to account for most of the toxic potency, the PCBs.
Selective accumulation of the more toxic PCB congeners can result in a mixture
in the tissues of target animals, which is more toxic than would be predicted from an
estimate of the original Aroclor® mixture.'22' This enrichment of the more toxic, non-
ortho-substituted PCB congeners results in a relative toxic potency of the mixture
which is from four to six times greater than the original technical mixture.*22' The
toxic potencies determined as the ratio of TCDD-EQ detetermined by the H4IIE assay
to total concentration of PCB of extracts of double-crested cormorant eggs were 2.5
to 5.24 (mean = 3.77) times greater than technical mixtures of aroclors®, which were
also measured in the eggs (Fig. 5).<28'81'82'
When a potency ratio was calculated for several of the locations in the Great
Lakes it is found that indeed the ratio of toxic potency varies among locations (Fig.
6). Furthermore, the greatest ratio was observed to occur with the least total
concentration of PCBs. This relationship is most likely for the greater correlation
between adverse effects and TCDD-EQ than with total concentrations of PCBs (see
below). Since the greatest proportion of the TCDD-EQ in birds of the Great Lakes is
contributed by the PCBs there is a general correlation between the concentrations, but
there is sufficient variation in the relative potency that a measure of TCDD-EQ gives
better prediction of the effects of complex mixtures. The relative potencies (EC-50
for EROD induction) of extracts from Green Bay ranged from 6 to 56 pg TCDD-EQ///g
PCB, which indicates that total PCB content of a sample is a poor indicator of the
biological potency of the toxicity mediated through the A/?-receptor, even though the
measured concentrations of TCDD-EQ were correlated with the total concentration of
PCBs.0021 If all of the TCDD-EQ could be attributed to PCB congeners and these
congeners were sorted equally in the environment and assimilated and metabolized
equally there would be no significant difference in relative potency among samples,
except for the contributions of other compounds.
The smallest PCB-normalized potencies observed in these samples were
C-233
-------
Giesy Briefing Document, Minneapolis, MN Sept 14-15, 1993
approximately 10 //g TCDD-EQ/g PCB.(B2) This value is similar to the potencies
observed for technical grade PCB preparations in the H4IIE bioassay system.(22'105)
However, most of the samples in this study had PCB-normalized potencies which were
considerably greater than those of technical grade PCB preparations.122>
The greatest enrichment of TCDD-EQ, relative to technical mixtures of Aroclors
is due to trophic transfers.'62' This is due not only to the presence of non-PCB
congeners, but to the enrichment of Ah-r-active congeners, relative to the total
concentrations of PCBs. The enrichment of specific p-PCH congeners has previously
been demonstrated, in waterbirds.1123"125' This enrichment was greatest for the
2,3,7,8 -substituted congeners which have a relatively great biomagnif ication potential
and are poorly metabolized by most species.(126)
To assess the enrichment of individual PCB congeners within the Great Lakes
ecosystem, the relative proportion of the congeners to total concentration of PCBs in
bird tissues was compared to that of technical AroclorR mixtures and to samples of
fish tissue from lake Michigan (Table 3;<9e'96'110,12?) Tne rejatjve concentrations of
PCB congener 77 (IUPAC) in extracts of bird tissues was the same or less than in the
technical mixtures of Aroclors8 1242 and 1248 but greater than that in AroclorsR
1260. Thus, depending on the relative proportions of these Aroclors" in the original
mixture released to the environment, this congener may have been enriched,
diminished or stayed the same. Similarly, PCB congener 105 could have been
enriched or diminished in the samples, relative to the original technical mixtures
depending on the relative proportions of different Aroclors" making up the complex
mixture. PCB congener 126 was enriched in the extracts of bird tissue regardless of
which of the Aroclor mixtures to which they were exposed. The relative contributions
of these three congeners to the total mass of PCBs were also greater than those in
Chinook salmon from Lake Michigan. It is uncertain whether PCB congener 169 was
enriched. Values for the relative contribution by weight in Aroclor" 1254 ranged from
0.00005 to 0.08. Thus, it is difficult to determine if the value of 0.00141, observed
for bird tissues in our study is an enrichment or diminution (Table 3). The enrichment
of these PCB congeners, however, does not explain the discrepancy between the
measured and predicted TCDD-EQ, based on the H4IIE TEF values, in the tissue
samples. The combined mass contribution of the PCDD and PCDF congeners to the
total concentrations of TCDD-EQ was less than 0.5% of the mass of total PCBs
present in all samples (Fig. 4).
The relative potencies are different among the original technical Aroclor"
mixtures (Table 3; Fig. 5). Aroclor"-1016, which is Aroclor"-1242 with the PCDD,
PCDF and p-PCBs removed contains essentially no detectable TCDD-EQ. The greatest
relative potency is observed in Aroclor"-1248 and the least potency is observed in
AroclorR-1260 (Fig. 5; Table 3). The relative potency in cormorant eggs from the
Great Lakes was more than twice as great as any of the Aroclor" mixtures. Thus, the
observed enrichment could not be caused by any combination of the original technical
Aroclor" mixtures. Furthermore, contributions of PCDD and PCDF were not included
C-234
-------
Giesy Briefing Document, Minneapolis, MN Sept 14-15, 1993
in these extracts of eggs of fish-eating colonial water birds. Thus, we feel that the
greater than predicted potency is primarily due selective enrichment of the A/)-r-active
PCB congeners, but some contribution may be due to the presence of unidentified A/?-
r-active compounds as well as PCDD, and PCDF congeners.(102)
Biotransformation is the key first step in the elimination of P-p-PCH congeners
and differential biotransformation of PCDD and PCDF congeners by cytochrome P450-
mediated pathways has been demonstrated in some species.<128) Both fish and birds
have enzymes that are capable of metabolizing PCB congeners.'123'129* However, the
activities of cytochrome P450-requiring oxygenase enzymes in fish-eating birds are
greater than those in most fishes, but less than those in most mammals.(124'125)
Therefore, birds can be expected to eliminate p-PCHs, but possibly more slowly than
mammals. In general the more substituted congeners tend to be less rapidly
metabolized.'1231 Also, those congeners, which are laterally substituted, such that
they do not have two adjacent, un-substituted carbon atoms are more slowly
metabolized'1231 and tend to accumulate in animais.(41> PCB congeners with vicinal
hydrogen atoms in the ortho and meta positions with more than one o/t/io-chlorine
atom are also resistent to metabolism.1123) The hexa-chloro biphenyl (2,2',4,4'5,5'}
was not metabolized by pigeons, rats or brook trout.(130) More chlorinated congeners
can be metabolized if they have adjacent, unsubstituted carbon atoms, whereas those
with no vicinal unchlorinated carbons were not metabolized and only slowly
excreted.(131"133> This indicates that congeners, which were poorly metabolized would
not be as readily excreted and would tend to be accumulated selectively in tissues,
relative to other congeners, such as the 4,4'-di-chloro and 2,2',5,5'-tetrachloro,
congeners, which were metabolized to more polar hydroy-metabolites, which could
be excreted.
In addition to the effects of metabolism, position in the food chai can affect the
relative concentrations of PCB congeners in the tissues and eggs of birds.(134) This
is due to many factors, but is primarily due to the relative quantities of different prey
items taken by different species of the same species in different locations.
The greater PCB-normalized potency of the p-PCH extracts of avian species
from Green Bay, relative to that of AroclorR indicates that mechanisms of trophic
selection or the presence of unidentified A/?-r-active compounds in the extracts could
be the cause of the elevated relative potency of the p-PCH mixture. There are two
principal ways to correct for changes in potency during a risk assessment: First, each
of the active congeners could be quantified and their individual concentrations
corrected for relative potencies. Alternatively, an application factor or enrichment
factor could be applied to the total concentrations. Since the individual PCB
congeners which express the greatest toxic potency have not been measured
traditionally and are still seldom considered in regulations, if a constant correction
factor can be justified, it could be applied to total concentrations of PCB and correct
for the effects of weathering and enrichment through biomagnification. To investigate
the assumption that an enrichment factor could be applied to total concentrations of
C-235
-------
Glesy Briefing Document, Minneapolis, MN Sept 14-15, 1993
PCBs to correct for different potencies, we calculated water quality criteria (see
section below) then calculated exceedance values for current conditions and compared
the exceedances based on total concentrations of PCBs and those based on TCDO-EQ
the ratio of the water quality criterion to protect double-crested cormorants at seven
locations, based on total concentrations of PCB and TCDD-EQ (Table 3; Fig. 7).
p-PCHs Other than TCDD TCDF and p-PCBs.
A number of compounds, structurally similar to pPCBs, PCDDs and PCDFs, and thus,
should act similarly to these p-PCHs have been observed at concentrations in the
environment that, if they have similar toxicological properties, could be of toxicological
significance (Table 4). Unfortunately, for many of these compounds, there is currently
little known about their toxicological or environmental fate properties. Thus, even if
their concentrations in environmental samples were determined instrumentally, it
would still be difficult to assess their potential to cause significant adverse effects.
It is likely that these compounds do contribute a significant quantity of TCDD-EQ in
environmental samples. When TCDD-EQ are measured in the H4IIE bioassay and
compared to the TCDD-EQ calculated from instrumental analyses for PCDDs, PCDFs
and PCBs and their TEF values measured in the H4IIE bioassay we are unable to
account for 30 to 50% of the TCDD-EQ measured in the bioassay, especially in
samples taken from urbanized and industrialized areas (Fig. 4).
The potency of extracts of peat were found to cause greater induction than
could be accounted for by the concentrations of PCDD and PCDF in the extracts.113B)
This suggests that there are compounds other than the traditionally measured planar
molecules which can be Ah-r-active and thus measured in bioassays of P4501A
activity.
Compounds that could contribute to the total concentrations of TCDD-EQ
measured in the bioassay also include any or all of the following classes of
polychlorinated compounds (Table 4): naphthalenes (PCNs), diphenyl ethers (PCDEs),
diphenyi toluenes (PCDPT), phenoxy anisoles (PCPAs), biphenyl anisoles (PCBAs),
xanthenes (PCXE), xanthones (PCXO), anthracenes (PCAn), fluorenes (PCFIs),
dihydroanthracenes (PCDHAs), diphenyl methanes (PCBMs), phenylxylyiethanes
(PCPXEs), dibenzothiophenes (PCDT), quaterphenyls (PCQs), quaterphenyl ethers
(PCQEs) and biphenylenes (PCBE). In addition to the chlorinated compounds,
brominated and chloro/bromo-substituted analogues of PCDD and PCDF have been
found in the environment"36} and are known to induce ethoxyreurufin-o-deethylase
(EROD) activity in vivo and in vitro.(4Q] In addition to the above mentioned
compounds, there are a number of polychlorinated compounds, which are the
alkylated forms of these same classes. These include polychlorinated-alkyIbiphenyls
(PCAB), alkylnapthalenes (PCAN), alkylphenanthrenes (PCAP) and
alkyldibenzothiophenes (PCADTh).(137) Alkylated analogs of all of these compounds.
C-236
-------
Giesy Briefing Document, Minneapolis, MN Sept 14-15, 1993
including PCDDs and PCDFs are especially prevalent sludges and sediments near paper
mills that use chlorine in the bleaching process."381 Polychiorobibenzyles (PCBB) have
also been observed in the vicinity of pulp mills/139'
The PCNs are known to occur in the environment at great enough
concentrations sufficiently great in some locations that, coupled with their surprisingly
high TEF values, could be of toxicological significance similar to that of the PCDD,
PCDF and pPCBs/101'1401 The TEF values reported by Hanberg (1988) for PCN
ranged from 0.000007 to 0.002. Total concentrations of PCN in fishes from the
Great Lakes have been reported to be as great as 5 mg/kg, wet weight. The TEF for
the most prevalent PCN were 0.002. Thus, PCN could contribute significantly to the
total TCDD-EQ. Chlorinated PCNs have been found in great concentrations in the
effluents and sludge of pulp and paper processes/137'
The PCDEs are byproducts in the manufacture of chlorinated phenols.(129)
Significant concentrations of these compounds have been found in the tissues of
humans, fish and wildlife.<141) PCDEs can be accumulated by fish(142) and induce MFO
activity/143' Furthermore, an unidentified tetrachlorinated PCDE isomer was found at
a concentration of approximately 0.9 mg/kg, wet wt. in the eggs of common terns in
Michigan/1441 Concentrations of tri- and tetra-chloro PCDE as great as 3 mg/kg have
been found in the tissues of common tern chicks/1461 The PCDEs are similar to PCBs
in their dynamics and persistence in animal tissues and induce P-450 type
monooxygenase activities. The EDg, values for several PCDEs are in the range of 15
to 110 umo!/kg.(141) Unlike PCBs the mono- and dichloro-substituted PCDE are also
potent inducers of EROD activity/1461 It has been postulated that the additional
distance between the phenyl moieties reduces the strong effect of ortho-substitution,
which is observed in the PCBs. Thus, PCDEs in the environment may exert a greater
effect on the EROD activity than predicted from QSAR relationships developed for
PCBs. The PCDEs are rarely monitored in environmental samples due to the paucity
of authentic standards and because they occur in the same fraction as the PCDF and
several of the congeners result in the same mass fragments as some of the PCDF/1411
Thus, since their potencies to induce MFO activities are similar to those for PCBs,
PCDEs could account for a significant amount of the induction measured in the H4I1E
assay.
The PCDPT are not used in North America, but are manufactured and widely
used in Europe as substitutes for PCBs, especially as hydraulic fluids in mining and
have been found in the tissues of aquatic organisms/147'1481 The PCDPT, which are
also called diphenyl-methanes, are structurally similar to PCBs and cause similar
effects/1491 PCDPT are potent inducers of EROD activity in mammals and fishes/1481
It is not likely that there are sufficient concentrations of PCDPTs in the tissues of
North American wildlife to contribute significantly to the TCDD-EQ measured in
samples. However, the PCDPT could currently make contributors to the total TCDD-
EQ in samples from some locations in Europe"601 and may contribute to the TCDD-EQ
measured in North America at some time in the future.
C-237
-------
Giesy Briefing Document, Minneapolis, MN Sept 14-15, 1993
The PCPXEs are known to be potent inducers of EROD activity in the livers of
fishes/1511 but we are aware of no reports of these compounds occurring in the
tissues of fish or birds taken from the environment. Thus, neither the hazard nor the
risk presented by these compounds to wildlife can be assessed at this time.
A complete assessment of the possible environmental hazard posed by PCDT
is impossible because little is known about their persistence, bioconcentation, or
toxicological properties. The PCDT are formed during incineration of organic
compounds which contain sulfur and chloro-compounds, such as tires.<152'153) The
PCDT have a relative potency, as measured in the H4IIE assay, similar to that of the
PCDD.{1BB) There are few reports of the occurrence of PCDT in environmental
samples: however, concentrations of the 2,4,6,8-tetra-CDT as great as 8,300 and
1,000 pg/g, wet weight were observed in the tissues of crab and lobster respectively,
from the Elizabeth river, New Jersey. Concentrations of other tetra-chlorinated
congeners were as great as 500 pg/g in the crabs from the same location. Therefore,
it is possible that PCDT could be present locally in sufficiently great concentrations
to contribute to TCDD-EQ in environmental samples, including the tissues and eggs
of birds from the Great Lakes region of North America. The TEF for a synthetic
mixture of PCDTs (2.4 % CI2, 74.6% CI3 22.4% CI4 and 6% CI5) was found to be
0.000425 in the H4IIE assay."08'
PCBE have been little studied, but are structurally similar to the PCDD.(1B6)
PCBE have been found to occur in the environment, generally due to pyrolysis of
PCBs.(51) Few authentic standards are available for the PCBE and the concentrations
in the environment are generally quite small, compared to those of PCBs. In addition
to the chlorinated diphenyl ethers, brominated analogues of these compounds have
also been found to occur in the aquatic environment at concentrations, whih could be
toxicologically relevant.'1 B7)
Both PCQs and PCQEs have been found in environmental samples and have
been demonstrated to be toxic to animals."68' However, they are reported to be
much less toxic than PCDD or PCDF. Both of these classes of compounds were found
in patients who consumed Yusho oil/158' but it is unlikely that either of these classes
of compounds occur in wildlife of the Great Lakes region at concentrations great
enough to be of toxicological significance.
In addition to the diaromatic-type compounds, single ring compounds are known
to induce P450-type monooxygenase activity. For instance hexachlorobenzene (HCB)
induces several monooxygenase activities/63'1 B9) but it is unlikely that it contributes
significantly to the concentrations of TCDD-EQ in wildlife since it does not induce the
P450IA1-type isozyme which is measured as EROD activity. HCB has been detected
in the eggs and tissues of birds in the Great Lakes/160' but the significance of the
concentrations is unknown at this time.
Polybrominated compounds such as PBBs and PBDEs which are used in flame
retardants such as BromkalR, which contains polybrominated diphenyl ethers
(PBDEs)/48' PBBs are known to occur in the environment and can be accumulated
C-238
-------
Giesy Briefing Document, Minneapolis, MN Sept 14-15, 1993
into biota.(1S7) The BromkalR mixture is known to induce EROD activity in vitro, but
the potency for induction is much less than that of the PCDD, PCDF or pPCBs.(101)
Since concentrations of these brominated compounds are not generally measured, it
is impossible to assess the contribution that these compounds may contribute to the
concentrations of TCDD-EQ measured in extracts from biota.
There are a number of polycyclic, aromatic hydrocarbons (PAH) compounds
which are known to bind to the A/j-r. These compounds may contribute to the
toxicity observed in some species. However, it is unlikely that these compounds
would be responsible for the effects observed in eggs, since they are rapidly
metabolized by vertebrates and thus would not be biomagnified into eggs. Also, since
the polar metabolites of these compounds would not be expected to occur in the
extracts which are used in the H4IIE assay, it is unlikely that they are responsible for
the EROD induction which can not be accounted for in comparisons of bioasssay and
instrumental analyses. Because a sufficiently great number of persistent compounds
can interact with the A/?-r, it would be useful to use techniques, such as the H4IIE
bioassay to determine if all of the potential hazardous compounds in complex mixtures
have been accounted for.
When the environmental contamination with A/?-r-active compounds is
discussed, their impacts are often dismissed because the manufacture of PCBs which
contribute the greatest proportion of the known TCDD-EQ in biological samples, has
been discontinued.
It is often concluded that nothing can be done about these compounds that have
already been released to the environment and that the concentrations of these
compounds will eventually decrease to insignificance. While this is partially true, we
believe it is important to remember that a number of other compounds, with similar
ecotoxicological properties are still manufactured, used and released into the
environment at concentrations that could be of ecological ortoxicological significance.
Therefore, the use of these compounds should be as tightly regulated as the other p-
PCHs and their continuing uses reevaluated as more information on their distribution
is gained through monitoring. This is especially important if these other classes of
compounds are introduced to commerce as substitutes for other members of the more
pernicious p-PCH classes.
An implication of these observations for regulators is that discharge limits for
a single p-PCH projecting its toxic effects at a given level of discharge must account
for the additivity with other p-PCHs in the environment to ensure protection of wildlife
as well as humans. We know of no case where this policy has been used in the
regulatory community which has avoided the issues raised by mixtures and
interactions. However, the Great Lakes initiative, developed by the US EPA will take
these factors into account.
The use of BSAF values does not seem to be supported by anything other than
a few empirical ratios. It is difficult to believe that these values will be transportable
from one location to another. If the information necessary to evaluate these ratios on
C-239
-------
Giesy Briefing Document, Minneapolis, MN Sept 14-15, 1993
a site-specific basis is available, they will be unnecessary.
In conclusion, it is my opinion that the complex nature of the exposure of
animals to compounds which work through teh same mode of action we must use
toxic equivalency approaches, but the the endpoints selected and the way in which
they are integrated is very important. The use of these techniques must include a
calibration step, which accounts for the activity at the receptor and for
pharmacological aspects.
Question 3
I think that it is appropriate to use tissue levels for conducting risk assessments for
compounds which are alread in the environment. This approach will obviously not
work for situations where risk assessments are to be conducted for compounds which
have not yet been released to the environment. When possible, it would be useful to
have information on the tissue specific concentrations of TCDD-EQ. We have
developed such values for birds (based on chickens and field observations) fish (based
on a long-term feeding study with rainbow trout) and mink (based on a long-term
feeding study with carp from the Great Lakes).
Question 4
There is little information on the effects of TCDD on wildlife species. Studies are
difficult to reconcile one with the other because animals were exposed in different
ways via different routes, for different periods of time. Here I give a littele information
on the effects of TCDD on birds of the Great Lakes region. We ahve also conducted
studies with fish and mink. This information will soon be published, but for now is
given as briefing documents in Appendix I and Appendix II.
p-PCH Effects on Wildlife in the Great Lakes Region
The greatest weakness of all ecoepidemiological wildlife studies when
considering cause and effect relationships is that all such work is necessarily
correlational and non-experimental. No single chemical cause can be isolated in
wildlife and tested for effects, as it is done in laboratory situations that can control
variables or test each variable singally or in combination. Ideally, one would apply
Koch's postulates to wildlife contamination problems in the search for cause-effect
relationships (Table 5). However, for studies of wildlife under field conditions this
would be difficult. This is true because it is difficult to conduct controlled laboratory
studies with wildlife species and it is difficult to conduct studies with sample sizes
which are large enough to allow sufficient statistical power to test hypotheses about
effects, such as deformities. Also, it is difficult to conduct laboratory studies of
known exposures to the same complex mixtures to which wildlife is exposed under
C-240
-------
Giesy Briefing Document, Minneapolis, MN Sept 14-15, 1993
field conditions. We have conducted several studies with animal models, which have
been fed fishes from the Great Lakes, in an attempt to simulate more natural
exposures. However, the logistics for such studies are difficult. Thus, the task is to
correlate and then compare to other similarly and differently contaminated populations
and species. The best that can be achieved is to consider the weight of evidence and
reconcile effects observed in wild populations with controlled laboratory studies done
with animal models. These studies can be made more powerful by using both in vivo
and in vitro studies of complex mixtures or fractions into which certain p-PCH have
been isolated or enriched. Fox(161> has formalized the six criteria to be used in
determination of the validity of ascribing chemical causes to effects in epidemiological
studies of wildlife populations. These include consistency of observations, strength
of the association, specificity of the association, time sequence, coherence, and the
predictive power of the relationship. On these bases, an informed judgment can be
made that p-PCHs have an influence on populations of wildlife species with a great
degree of certainty.'8'24'69' Here we present what we feel is the strongest information
which supports the hypothesis that the p-PCBs are the most likely cause of the
observed effects on colonial water birds in the Great Lakes region of North America.
All of the symptoms observed in colonial birds of the Great Lakes region are
known to be caused by the p-PCH Thus, even though some of the details of the
mechanisms of action remain undescribed, the widespread common effects of the p-
PCH on wildlife worldwide, and especially in areas of greatest exposure to p-PCH,
suggest largely additive effects of these substances on exposed wild populations.
While these phenomena are not yet studied at the ecosystem or community level, it
seems likely that at least some of the recent decreases in sizes of populations,
extinction's, or shifts in species dominance are at least influenced, if not caused in
their entirety, by p-PCHs. For example, in the Great Lakes region, the smaller tern
species, such as the common (Sterna hirundo], Forster's (S. Forsterii), and black
(Chilodonias niger) were once more abundant in the region. However, in the last 30
years populations of all three species have decreased and now each is listed as a
threatened or endangered species in one or more states. Gulls and cormorants which
have larger body sizes and thus lesser energy requirements per unit body weight are
more tolerant to dioxin-like planar contaminants and their populations have continued
to increase since the concentrations of DDE have decreased to below critical
levels.'67'1071 It is possible that the interspecies dynamics and wildlife populations
balance on the Great Lakes are affected as much by contaminants as other traditional
habitat parameters. The p-PCHs are prime suspects in these phenomena. These
selective effects are subtle and may be difficult to separate from other dynamics of
the ecosystem.
GLEMEDS
C-241
-------
Giesy Briefing Document, Minneapolis, MN Sept 14-15, 1993
Of all of the adverse effects observed to occur in colonial water birds of the North
American Great Lakes region, the most obvious and that which can be most directly
related to survival of individuals and population-level effects are embryo lethality and
developmental deformities. Most of the embryos or chicks, which die during early
development have been observed to also have developmental deformities/8'23-24*107'
particularly abnormalities which are of ectodermal origin.(162) One of the best
documented abnormalities which has been correlated with concentrations of p-PCH
in bird eggs is the crossed-bill syndrome (Fig. 8) in North American cormorants.'241
This suite of conditions found in Great Lakes wildlife has been named the GLEMEDs
syndrome (Great Lakes Embryo Mortality, Edema and Deformity Syndrome)(8); it
mimics chick edema disease caused in offspring of hens exposed to PCDD and PCDF
in their feed.(27) The effects observed in birds are similar to those observed in
mammals which were exposed to these chemicals.
The few available case studies illustrating the effects of these chemicals are the
reports of the Canadian Wildlife Service on studies performed on herring gulls living
on Lake Ontario during the 1970s, particularly between 1974 and 1977. In the late
1960s, anecdotal evidence circulated among field biologists of poor egg hatchability
of Lake Ontario herring gull eggs.(12) Official Canadian Wildlife Service surveys began
in 1971. Hatchabilities of less than 20% were found at some colonies in Lake
Ontario. Productivity was reduced to less than one fledged young per ten nests.
Maintenance of a stable population requires fledging rates in the range of 5-6 chicks
per ten nests per year for this species. Initial examination of herring gull eggs and
eggs of other species in Lake Ontario documented the presence of DDT and PCBs.
However, analytical techniques at the time were insufficient to discriminate among
dioxin, furan and p-PCB congeners or to quantify some of the more toxic p-PCB
congeners. Reliable congener-specific chemistry was a decade away. The
characteristic symptoms in surviving chicks were similar to those of chick edema
disease, which is caused by PCDDs in poultry and had been previously described/27'
Subsequent reanalysis in the late 1980s of a variety of eggs which had been collected
from herring gull colonies in Lake Ontario in the 1971-1976 period and archived
contained 1-3 ppb of actual TCDD. This TCDD is thought to have been discharged
into the Niagara River as a result of herbicide manufacture.1163) This chemically-
caused epizootic in Lake Ontario is probably the best documented example of dioxin-
caused effects on wildlife. During the last decade, the symptoms of chick edema
disease and GLEMEDS have decreased significantly in the herring gull population of
Lake Ontario,'81 but more subtle, biochemical effects persist all species of fish-eating
colonial water bids of Lake Ontario and the other Great Lakes/24'681
Similar studies of nesting Forsters terns, conducted in Green Bay from 1983 to
1988,{10/23/164) and double-crested cormorants and Caspian terns(107) have revealed
a similar suite of biological effects, but implicate different p-PCHs than TCDD as
probable causes. In the Green Bay experience, Kubiak et a/.no] found a variety of
developmental deformities in the embryos and chicks of Forster's terns including
C-242
-------
Giesy Briefing Document, Minneapolis, MN Sept 14-15, 1993
growth deficiencies, deformities and behavioral differences in parental care of eggs
compared to an inland control colony where exposures were significantly less than on
Green Bay. Extrinsic adult behavioral abnormalities of inconsistent incubation led to
a four day longer incubation period than the reference colony. Reciprocal transplant
studies of eggs a similar time delay ascribed to toxic substances in the eggs. That
study suggested widespread, complex contaminant effects on the reproductive cycle,
such as longer incubation times, smaller individuals, and wasting syndrome in those
that did hatch. Similar, but less acute problems, have been observed in double-
crested cormorants and Caspian terns in the upper Green Bay area, where TCOD-EQ
calculated from concentrations of p-PCH and TEFs, from 175 up to over 440 ppt have
been measured.<23/107)
A study of Caspian terns in Saginaw Bay following the disturbance of
sediments in a flood incident has documented concentrations of p-PCB in eggs which
was similar to those observed in Green Bay. In this case between 96 and 98% of the
TCDD-EQ was due to the presence of only four p-PCB congeners. These were, in
order of relative contribution to the total TCDD-EQ, congeners 77, 126 169, and
105.(81'107> The developmental effects observed in Caspian terns at Saginaw Bay
were severe. Caspian tern eggs contained doses equal to the lethal concentration for
95% of white leghorn chicken eggs for p-PCB congeners 77 and 126, plus 10% of
the LD-50 dose of TCDD. The concentrations of five individual p-PCB congeners
reported as TCDD-EQ were more than an order of magnitude greater than that which
has been found to cause heart defects in developing chickens.(165) Thus, it is likely
that the observed concentrations of TCDD-EQ were sufficient to cause the adverse
effects which were observed. Even though the dioxin-iike p-PCB congeners have
relative potencies which are less than that of TCDD, they are hundreds to as much
as thousand-fold more abundant in the environment and thus can have toxic effects
on the wildlife species.
Egg mortality from locations in 5 geographic regions on and one off of the Great
Lakes were found to be directly proportional to the concentration of TCDD-EQ (Fig.
gj (22) -pne current concentrations of both PCBs and TCDD-EQ in cormorant eggs are
greater than the estimated no-effect concentrations for this species. The
concentrations of TCDD-EQ were determined by the H4IIE assay in double-crested
cormorant eggs from several locations in 1986-88.<22> The total concentrations of
TCDD-EQ, as well as the relative potencies, varied among locations (Table 6, Fig. 6),
but was fairly consistant across all of these regions. The death of eggs was found
to be directly proportional to the concentration of TCDD-EQ in the eggs (Fig. 9)(22)
with a significant positive correlation (R2 = 0.703; p< 0.0003). These estimates are
based on actual doses in the eggs, thus there would be less error in predicting these
exceedences than would be expected for ratios, calculated from the WQC, which
include more assumptions. The exceedance range was less than when several species
were compared, but there was an almost 10-fold difference among locations (Table
6).
C-243
-------
Giesy Briefing Document, Minneapolis, MN Sept 14-15, 1993
We investigated the relationships between birth defects observed in both
Caspian terns and double-crested cormorants at a number of locations in the Great
Lakes. For this study we classified and enumerated the number of various types of
deformities in both double-crested cormorants and Caspian terns (Table 7; Fig. 10).
We found rates of abnormalities that ranged from two to 12 per thousand chicks
embryos and chicks examined (Fig. 11). We observed the greatest rate of deformities
in double-crested cormorants to occur in Green Bay (Fig. 11). The greatest
percentage of deformities of Caspian terns occurred in the colony on the contained
disposal facility (CDF) in Saginaw Bay. When the rates of deformities were correlated
with the concentrations of TCDD-EQ, we found very strong, statistically significant
correlations (Fig. 12). While this observation alone does not make a cause and effect
relationship, this correlation is better than that with any of the other contaminants.
Concentrations of TCDD-EQ and the relative potencies of the p-PCH mixtures in
cormorant eggs varied among locations in the Great Lakes (Figs. 13 and 6), but the
greatest concentrations were measured in eggs collected from Green Bay,(81> where
the greatest rates of deformities and poorest survival of eggs were observed.'24'
Wildlife Hazard Assessments for RGBs
Here we provide a proposed method for conducting wildlife hazard assessments, by
using results of field and laboratory studies on target and surrogate domestic species,
along with environmental monitoring and chemical analyses, to determine the
appropriate water quality criteria (WQC) to protect sensitive wildlife species from the
adverse effects of PCBs. We then compare the results of this assessment to an
assessment, which used human cancer and reproductive effects as endpoints. We
based our hazard assessment on PCBs instead of 2,3,7,8-TCDD, because the PCBs
have been found to be the primary source of most of the TCDD-EQ for which
concentrations in biota have been determined. We conducted wildlife hazard
assessments by comparing the threshold for effect to the concentrations of key
contaminants which are currently observed in tissues of fishes that are eaten by
domestic species and eggs of wild birds in the Great Lakes basin. The thresholds for
effects, termed the lowest observable adverse effects level (LOAEL) were derived
from field observations and correlations with concentrations of key toxicants or from
controlled laboratory studies where domestic species of interest were fed known
quantities of contaminants in Great Lakes fishes. We studied four species of wild
birds: herring gulls (Larusargentatus), bald eagles (Haliaeetusleucocephalus), Caspian
terns (Sterna caspfa), double-crested cormorants (Phalacrocorax auritus), and a
domesticated surrogate species, white leghorn chickens (Ga//us ga//us)n73} (Tables 8
& 9). The no observable adverse effects level (NOAEL) values were defined to be
10% of the LOAEL (safety factor of 10; equation 1) to correct for the uncertainty in
determining the NOAEL from the LOAEL. The LOAEL can be reported (Equation 1).
C-244
-------
Giesy Briefing Document, Minneapolis, MN Sept 14-15, 1993
NOAEL = LOAELx 10'1 (Eq. 1)
as the reference dose (Reft = dose in the target tissue), which is defined as a
threshold concentration (mg/kg, wet weight or iipid weight) in a tissue such as that
of egg, which causes a defined adverse effect, such as egg lethality or birth defects
(Table 10). Alternatively, the dose can be given as the daily intake or by making the
appropriate assumptions about food consumption, as the concentration in food (Reff
= dose in food). Here, LOAEL values are reported as either the concentration in bird
eggs or in the food of chickens (Table 10).
The bioconcentration factor (BCF) was used to predict the concentration of p-
PCH in water, which would result in the reference dose in whole fish tissues, which
these animals eat (Table 10). The biomagnification factor (BMP) was used to predict
the concentration in water that would result in the reference dose in eggs (Table 10).
Water quality criteria to protect wildlife were derived by dividing the NOAEL by
the bioconcentration or biomagnification factor (BCF or BMP) for the species of
interest (Table 10).<67) For piscivorous fishes we did not attempt to separate the
proportion of the contaminants observed in the tissues which was derived directly
from the water and that which was obtained from the food. We reasoned that in a
food chain exposed to chemicals in water, a ratio which related the concentration of
the chemical in the fish to that in the water could be derived. This included both
vectors of accumulation, because the prey consumed by the predatory fish would also
be in steady state conditions with the concentrations in the waters. We note that,
in addition to fish, the food web includes other fish-eating wildlife. This fact is usually
ignored in setting WQC and thus the BMF of toxic chemicals can often be
underestimated. We have calculated the reference dose for water (Table 10) by
Equation 2.
Rw = Rf / (BAFw.t) (Eq. 2)
Where:
Rw = Reference dose in water
BAFw.t = Bioconcentration factor from water to fish tissue corrected for
relative potency and biomagnification om one fish to another.
WQC were then developed by applying the appropriate application factors. The only
application factors used in these studies were a lOx factor to estimate the NOAEL
from the LOAEL. This factor was applied for the WQC for each species. The final
WQC for PCBs was corrected for uncertainty in among-species sensitivities by the
application of an additional 10X uncertainty factor. The final WQC for TCDD-EQ was
estimated, based on accumulation by eagles, but not corrected for any uncertainty
factors.
C-245
-------
Giesy Briefing Document, Minneapolis, MN Sept 14-15, 1993
It is difficult to determine WQC because of the uncertainty associated with
estimates both exposure and the dose-response relationships. Estimates of the
factors required to predict the probable accumulation of PCBs or TCDD-EQ from water
are particularly uncertain. First of all, the relative proportion of these compounds that
are freely dissolved in the water or bioavailable, is impossible to know. Furthermore,
the forms of these compounds is continously changing and never at steady state,
which are the conditions under which most predictions are made. We did not use a
correction factor for the bioavailable fraction of PCBs or TCDD-EQ. We made the
conservative assumption that all of the PCB or TCDD-EQ were bioavailable. in reality,
the bioavailable fraction is probably from 1 to 10% of the total. For this reason, there
is probably at least a 10 to 100-fold safety factor in our estimates of NOAEL. Since
it is difficult to know what portion of the PCB or TCDD is available at any given time
and that when the equilibrium between readily available and bound PCB or TCDD-EQ
such that there is a continuous source of these compounds in the bioavailable
fraction, we chose to not correct for this factor. The concentrations of PCBs
measured in the dissolved and paniculate fractions of the Great Lakes (Table 11) were
positively correlated with the total concentrations of PCBs in the eggs of double-
crested cormorants. This indicates that regardless of the relative available fraction the
concentrations in the eggs of fish-eating birds are proportional to the concentrations
of PCBs inthe water.
Furthermore, since the BCF and BMF values used in our assessment were derived
from field observations, ther are more likely to be apparent BCF and BMF values,
which take into account the bioavailable fraction.
The values reported for bioaccumulation factors are influenced by bioavailable
fraction as well as the physiology and food habits of the species for which
bioconcentration or biomagnification factors are to be predicted. Laboratory studies
with fishes, generally do not include the accumulation of p-PCH from food and thus
must be corrected for what would be expected to be accumulated from food, if water
exposures are used to predict the BCF.(179) Better estimates of biomagnification
factors (BMF) are available, based on field observations. For an accurate estimate of
the exposure from consumption of contaminated food, in the absence of observations
of the actual concentrations in food, site-specific estimates of accumulation potentials
are needed. Ideally, the exposure dose, expressed as mass of toxicant, per unit mass
of organisms per unity time should be known. For instance pg of TCDD-EQ/g,
bw/day. For these estimates, in addition to knowledge of the concentrations of the
contaminant of interest in each of the dietary items, one would need to have an
estimate of the proportion of each intern taken in the diet and a conversion value for
accumulation efficience for each compound from each food item. At this point in time
this level of resolution can not be attained. Therefore, we used average values for
accumulation and assumed that the efficiency was the same from all types of food
and that the concentrations in food were uniform. Furthermore, we have assumed
that the diet consists solely of food items with the specified concentration of toxicant
C-246
-------
Giesy Briefing Document, Minneapolis, MN Sept 14-15, 1993
of interest. These assumptions can clearly add uncertainty to the estimates. When
possible, it is more accurate to estimate hazard potential from assessments based on
concentrations of contaminants in the tissues of prey. We have taken this approach
in our hazard assessment, where possible.
The reference doses of complex mixtures to wildlife are difficult to relate to
exposures in food or water. The information on toxicity, reported in the literature is
often based on single exposures and may be via injection. Thus, it is difficult to
determine the effects of longer-term, continuous exposures in food. For this reason,
we have, where possible used the results of feeding studies with great lakes fishes.
These exposures are confounded by the fact that there are complex mixtures of
multiple toxicants in the fishes which could influence the response to PCBs or TCDD-
EQ. We have chosen to base the hazard assessments on reproductive endpoints,
since it is thought that these should be as or more sensitive than effects on adults.
For birds we have, thus, based the hazard assessments on the concentrations of PCBs
or TCDD-EQ in eggs, which have been associated with observable adverse effects.
Alternatively, we have related the effects levels to doses in food, which was fed
during longer-term exposures. The hazard assessment for current conditions are
reported as exceedance values, which are the ratio of current water concentrations
to the WQC (Equation 3).
Exceedance = [Concentration in water]/(WQC) (Eq. 3)
Subsequent to completing the hazard assessment for TCDD-EQ, the concentrations
of PCBs in fish tissues and bird eggs were corrected for the change in relative
potency, due to a greater concentration of TCDD-EQ per unit PCB in the tissues of
fishes and bird eggs than technical PCB mixtures (see section on selective
enrichment). In addition to the total concentration of PCBs, we conducted a hazard
assessment with the concentration of TCDD-EQ.
WQC for PCBs
We calculated the water quality criterion to protect sensitive species of birds from the
adverse effects of PCBs to be approximately 1.0 pg PCB/I (part per quadrillion) (Table
8). In 1986, the year for which the most recent data are available, the total
concentrations of PCBs in all of the Great Lakes ranged from 1 to 10 ng/l (pptr) with
the greatest concentrations occurring in Lakes Michigan, Ontario and Erie (Table 11).
The'concentrations of PCBs in the water of all five of the Great Lakes in 1986
exceeded our proposed WQC for all five species studied (Table 12). The degree to
which the WQC were exceeded varied from a minimum of 11 for herring gulls living
on Lake Superior to a maximum of 1800 for bald eagles living on Lake Michigan (Table
12). The exceedances were the least for all species for Lake Superior and tended to
C-247
-------
Giesy Briefing Document, Minneapolis, MN Sept 14-15, 1993
be the greatest on Lake Michigan, followed by Lakes Erie and Ontario (Table 12). The
exceedence values were calculated for both total PCBs and TCDD-EQ in eggs of
double-crested cormorants from seven locations in the Great Lakes (Table 13). In
^general, the available concentrations of PCB corrected for enrichment by use of the
H4IIE enzyme induction bioassay, predicted to be in water were approximately three
orders of magnitude greater than the NOAEL. The proposed WQC to protect
wildlife from the adverse effect of PCBs range over 5 orders of magnitude (Table 14).
The value of 17 pg/l proposed by the Great Lakes Initiative'180' is approximately a
factor of 10 greater than what we have proposed as a protective value.
WQC for TCDD
There is a range of sensitivity to TCDD and TCDD-like compounds among different
species of birds.(77) Effect concentrations of TCDD-EQ in birds range from
approximately 10 to 500 pg/g, bw.(76'76/17a) Clearly, the chicken is a very sensitive
species. In fact, when concentrations of TCDD-EQ in the eggs of wild fish-eating
birds are compared to the effect concentrations for birds it would be predicted that
none of the wildlife species would be able to successfully reproduce in the Great
Lakes and that for some species the LC-99 would be exceeded. At the other end of
the spectrum, the pheasant.'261 Since no direct studies of the toxicity of TCDD on
wild, fish-eating birds were available, we used the information on the toxicity of TCDD
to chickens to derive a WQC, which should be sufficient to protect all wildlife species
(Table 15). We also used estimates of the no effect concentrations in birds from our
field monitoring studies (Table 13 and 15).
The WQC was calculated from the bioaccumulation factor and the NOAEL for
several species, including the chicken. The product of the BAF used in our hazard
assessment was 4.15x104. There is a great degree of uncertainty in estimating the
accumulation of TCDD from water. In their comprensive assment of TCDD Cook et
a/179) were able to remove some of the uncertainty in estimates of accumulation of
TCDD from water to fish by correcting for organic carbon content of the water and
Hpid content of the animal to which the TCDD was to be accumulated. Unfortunately,
there are few estimates of TCDD concentrations in water or in fish tissue, which
provide the necessary information to make these corrections. The concentration of
TCDD-EQ in forage fish from Lake Huron is approximately 10 pg TCDD-EQ/g, ww,
while that in large lake trout is approximately 350 pg TCDD-EQ/g, ww.(52) Thus, the
BAF from forage fish to preditors is approximately 10X. This is also approximately
the biomagnification factor (within a factor of 2) for persistent, neutral organic
compounds with a molecular weight similar to TCDD. Since fish-eating birds would
be more likely to eat forage fish than the large lake trout and the BAF for lake trout
was used in the calculations, no correction was made for biomagnification from fish
to bird.
C-248
-------
Giesy Briefing Document, Minneapolis, MN Sept 14-15, 1993
The WQC for TCDD-EQ, based on the white leghorn chicken would be 0.04 pg
TCDD-i (Table 15). That for the herring gull, which is known to be more tolerant of
the effects of TCDD, was estimated to be 0.3 pg TCDD/I. This level of protection
takes into account the potential accumulation of TCDD-EQ into higher trophic levels,
For instance, birds which eat larger fish, which have greater concentrations of TCDD-
EQ or other birds would not be adequately protected. Specifically, concentrations of
TCDD-EQ in eagle eggs from Thunder Bay, Lake Huron have been found to be as great
as 2,000 pg TCDD-EQ/g. Since the concentrations of TCDD-EQ in forage fish from
this location contained an average concentration of 10 pg TCDD-EQ/g, ww, the BMP
from fish to eagle egg would be approximately 200. Thus, the WQC would need to
be approximately 200 time less to protect eagles as colonial fish-eating water birds,
if the eagles are approximately as sensitive to the effects of dioxins as are the colonial
birds. To be sure of the protection of eagles we have selected the chicken as a
surrogate, if the WQC to protect chickens is used and a 200 fold BMP is applied, the
WQC to protect eagles would be approximately 0.0002 pg TCDD-EQ/i.
The values determined to represent a "low" risk to avian species by Cook et
al.,(179) were based on the effects on pheasants and made different assumptions
about the degree of availability of the TCDD in water. Even though our assessment
used different assumptions, the results of our anaiyis of colonial fish-eating water
birds were similar to those predicted by the US EPA (Table 16).
Because of the potential uncertainties associated with determining WQC and
because birds accumulate essentially all of their exposure to TCDD-EQ from their
food(187) to assess current conditions we compared the current concentrations of
TCDD-EQ in forage fish from Lake Huron(B2) to the dietary NOAEL for white leghorn
chickens.(173) When this was done it was found that the current exceedance value
is approximately 17 (Table 17). This is a conservative estimate of the current
situation, since larger fish, which could also be taken in the diet contain greater
concentrations of TCDD-EQ.
This exercise seems to support the use of the NOAEL calculated from the
chicken since the exceedance is in the range that one would expect to see subtle
effects on the colonial water birds and more severe effects on birds, such as eagles,
which are of a higher trophic leave!. This is, in fact what is observed: There are
subtle effects on survival of embryos and deformities observed in the colonial birds,
while eagles, which feed on fish from this location fail to reproduce.
Based on our field observations, the value of 6 pg TCDD-EQ/g ww of fish,
which is given as "low risk" by the US EPA hazard assessment'179) is probably
appropriate for the protection of double-crested cormorants, but would not be
sufficient to protect higher trophic leves such as eagles and may not be protective of
some of the smaller, more sensitive species of colonial fish-eating birds. We have
observed a concentration of approximately 10 pg TCDD-EQ/g, ww of forage fish
which is similar to the concentration indicated by the EPA11791 to be a small risk to
birds. In these areas we have observed adverse effects, such a deformities and
C-249
-------
Giesy Briefing Document, Minneapolis, MN Sept 14-15, 1993
embryolethality, but these effects do not seem to be limiting populations of double-
crested cormorants. In these areas wasting syndrome has been observed in Caspian
terns and the Caspian terns can not peproduce normally in some of these areas.(52)
.Thus, based on our field observations and the hazard assessment, we do not feel that
this value would be sufficiently protective of some of the more sensitive species or
those, such as eagles, which are a higher level of the food web.
Water Quality Criteria: Human vs Wildlife Health
PCBs
It is difficult to compare WQC developed for humans and wildlife because humans can
restrict consumption of contaminated food items while wildlife cannot. For this
reason fish consumption advisories that set an allowable quantity of fish that may be
eaten from a particular water body are appropriate for protection of human health but
not for the protection of wildlife.(188"180) Similarly, because individuals consume
different quantities of fish establishment of a WQC to protect all human's health is not
appropriate for everyone. The WQC to protect fish and wildlife from the effects of
TCDD-EQ, which was proposed by the US FWS is 5 pg/l.(6)
The WQC to protect humans from the adverse effects of PCBs, proposed by
the Great Lakes Initiative is 17- and 3- fold greater (Tables 10 and 14} than that
proposed by Swain'191/192) to protect humans from the non-cancerous developmental,
behavioral and cognitive effects of PCBs. The WQC calculated by Swain'191'192'
range from 0.6 to 6.0, with a median of 1.0 pg PCB/I (based on the McCarthy visual
cognition scale) depending on assumptions of exposure.'162'192"199' However, the
WQC proposed to protect human health'180> assumed that the hazard of cancer was
less than that of other non-cancer end points, which can occur in a dose-dependent
fashion at very small intrauterine or intra-egg exposures. For instance, concentrations
of PCBs between 1.5 and 2.5 ug PCB/kg, wet weight in blood of human umbilical
cords has been correlated with subsequent adverse cognitive effects in human
infants.(19B) Greater concentrations of PCBs (7.9-12.9 ug PCB/kg) have been
observed in the blood of women who ate an average of 23.5 Ib (10.6 kg) of Lake
Michigan fish per year and gave birth. The WQC proposed to protect humans'1801 are
similar to the values estimated for the protection of wildlife by Ludwig et a/.'67> These
results indicate that if the most sensitive wildlife species are protected, then humans
will also be protected from the most subtle effects. Thus, the adverse effects of p-
PCH-type compounds on wild species can serve as an early warning system for
potential effects in the human population, but only if WQC for wildlife based on real
effects in wild populations are included in the regulatory process.
The WQC that we propose to protect wildlife is approximately 40-fold less than
that purposed by the Great Lakes Initiative.'1801 Interestingly, the WQC based on both
cancer and non-cancer endpoints for the protection of human health from the effects
C-250
-------
Giesy Briefing Document, Minneapolis, MN Sept 14-15, 1993
of TCDD are greater than those proposed to protect wildlife (Table 18). This is
probably justified, since humans eat less fish in their diet. We feel that the lesser
value should be adopted since this would protect both wildlife and humans. Thus, in
the case of TCDD, it appears that if wildlife are protected so will humans. Thus,
wildlife would be a good sentinal species.
There is on-going controversy over the relative sensitivity of humans to the
effects of p-PCH(200) including both PCBs(207) and PCDDs.*201'204' There is even some
controversy over the potency of TCDD as a mammalian carcinogen.(208'208) This
uncertainty is leading to a reassessment of the reference doses and uncertainty
factors used in human health hazard and risk assessments based on cancer for these
types of compounds. It is difficult to conduct controlled studies on the effects of
chemicals on humans. For that reason, the chronic effects of chemicals on humans
are estimated from short-term, high-level exposures of animal models, which may be
more or less sensitive to the carcinogenicity of p-PCH than humans. A safety factor
of 10 fold is generally added to the assessment process to correct for among-species
differences in sensitivity. This assumes that humans are more sensitive to chronic
effects than are shorter-lived animals.4188-188'207)
Recent epidemiological evidence and the results of studies of the mechanism
and modes of action of p-PCH indicate that humans may be much less sensitive to the
A/?-r-mediated toxic effects of p-PCH than other species'200"202'204-207 This is
particularly true for carcinogenesis.(202"204) The US EPA has stated that, there is
inadequate evidence of carcinogenicity of PCBs in humans".1208) We agree. The
reason for this greater resistance or tolerance is unknown, but may be related to the
interaction of the An-r with the p-PCH or with the DNA.(200) Long-term studies on the
effects on humans exposed to p-PCH from accidental or industrial exposure for over
30 years generally have failed to exhibit rates of cancer greater than expected for the
population.(200> Exposure to p-PCH, such as PCDD, at Seveso, Italy or PCB in
electrical workers have not resulted in measurable increases in the rates of cancers
in these exposed groups. These observations have stimulated a reassessment of the
risk of cancer posed by these compounds to humans(200) and a possible change in the
proposed reference doses for cancer that could lead to a relaxation of proposed
WQC.<202) Currently, the standards for environmental concentrations of PCDD in
Europe and Japan are 170 and 1,700 times less stringent than those in the United
States.(204)
If the environmental standards for p-PCH, based on human exposure and the
cancer end point are relaxed, then the WQC will not be sufficient to protect either
wildlife or humans from the non-cancer adverse effects of these compounds. Many
subtle effects in humans will be ignored. Currently, the Science advisory Board of the
US EPA is reexamining the model that will be used to set environmental standards for
p-PCH.(204) The first phase of the reassessment will be to develop a better model
of human health effects, which is based on the current state of knowledge about the
A/j-r-mediated mechanism of action of the p-PCH.(200> Subsequently, an assessment
C-251
-------
Giesy Briefing Document, Minneapolis, MN Sept 14-15, 1993
of aquatic ecological risk of p-PCH will be conducted. The results of our analysis
indicate that the protection of wildlife species will require more stringent regulations
than for the protection of the humans from cancer. We suggest that the protection
of wildlife populations and humans from subtle non-cancer effects should be given
equal priority to that of protecting human populations from cancer. We have derived
water quality criteria, based on the responses of wildlife rather than cancer in
humans.*67' Our analysis indicates that WQC based on human cancer effects are this
is not adequate to protect sensitive, wildlife species. Additionally, the method for
deriving WQC has profound implications for remediation, litigation and damage
assessments. If wildlife are protected for the most sensitive end points, the most
relevant of which seems to be reproduction, all components of the ecosystem
including human health should be protected adequately.
The proposed WQC are very protective and if these criteria are met we would
not expect to see any adverse effects due to these compounds in wildlife. This does
not mean that if the proposed criteria are exceeded that adverse effects would be
expected to be observed. Also, since the proposed criteria do not consider
bioavaiiability, it would be expected that the actual safe concentration in water could
be as much as 10 to 100 times as great without causing adverse effects. Also, the
proposed criteria assume that the species of interest consume only the identified prey
with a specified BCF or BMP. This too provides some degree of safety. The 14 ng
PCB/I of total PCBs, which was suggested in the 1980 WQC document'209', might
protect some species of wildlife since we have documented effects at a concentration
of less than 2 ng PCB/I in areas of Lake Michigan.
There is a great deal of uncertainty in the risk assessment process (Table 19).
WQC developed as we did here are probably no better than ± 10 to 10OX. For this
reason it does not seem very worthwhile to argue about the exact WQC, especially,
when it will be impossible to directly validate models of concentrations in the water.
For compounds, which are already in the environment such as PCBs and TCDD-EQ,
we advocate monitoring of wildlife species or their diet. For new compounds not
already released into the environment, we advocate a conservative approach to
allowed released of bioaccummulatable compounds.
Water quality criteria (WQC) which are used to establish water pollution
standards and permissible loadings of substances to public waters can be derived by
several methods. These techniques generally involve hazard and risk assessment
procedures. For non-persistent, non-biomagnified compounds and elements, WQC are
derived from the acute and chronic toxicity to aquatic organisms. For persistent
organic compounds which are bioaccumulated and biomagnified, the effects on
organisms higher in the food web must be considered. In the Great Lakes region of
North America, the primary emphasis has been on the potential for adverse effects to
humans who eat fish. The primary endpoint considered in hazard and risk
assessments has been cancer. Other endpoints such as teratogenicity, intellectual
C-252
-------
Giesy Briefing Document, Minneapolis, MN Sept 14-15, 1993
performance, and immune-suppression and reproductive impairment are seldom
considered. By hazard and risk assessments, regulators have endeavored to predict
safe concentrations of potentially toxic chemicals which could be allowed to enter the
environment. Once hazardous materials have entered the environment, these same
criteria can be used as target values to determine if environmental damage has
occurred, and the degree of remediation required to restore an ecosystem. For aquatic
environments, these take the form of water quality criteria (WQC). These WQC
consider many environmental processes, such as dissipation, bioaccumulation,
biomagnification, degradation and dilution. For the most widespread and hazardous,
synthetic, organic chemicals, these regulatory decisions are preoccupied by potential
effects on the health of humans, particularly the risk for additional cancers in the
population. Regulatory actions assume that ecosystems have assimilative capacities
for persistent chemicals'81 and that risk of human cancer is the most relevant and
sensitive endpoint. For that reason, long-term effects of persistent, organic
chemicals, that have the potential to biomagnify and cause chronic, population-level
effects on wildlife have received scant attention. Humans have the option of
restricting consumption of contaminated food. In fact, most fish consumption
advisories are established to recommend a safe quantity of fish to be consumed in
some specified time period by a person of average size, age, and sensitivity. On the
other hand, wildlife cannot avoid contaminated food supplies and thus are at greater
risk.
Cancer in humans may not be the most sensitive endpoint to
measure.'200'210'211) There are physiological and developmental effects in a number
of species, including humans, that are more sensitive, immediate and demonstrative
endpoints than cancer.'8'211) Subtle effects such as growth retardation'162) or altered
development/193"195) immune system suppression and elevated rates of disease,'162'
wasting syndromes,'212'213) and behavioral changes in both adults and juveniles'101
have been observed in human and wildlife populations exposed to synthetic, p-PCHs
such as PCBs. Furthermore, the epidemiologicai evidence supports the contention
that human in the Great Lakes basin are expressing subtle, chronic effects due to the
exposure to PCBs and similar compounds.'191"192*
Not only are there endpoints in humans that are more sensitive than cancer, but
humans may not be the most sensitive species to the effects of p-PCHs such as
PCBs,'214'211) PCDDs, and PCDFs.'206'4* There are many wild species that may be
inherently more sensitive to some classes of contaminants or receive greater
exposures because they do not have the varied diets of humans. When wildlife
species have been considered in the derivation of WQC, generally only direct, acute
effects on aquatic organisms such as death have been considered. Some states in the
Great Lakes region have even based their criteria for wildlife only on the water they
consume, ignoring the much more important route of exposure for p-PCH represented
by consumption of fish. Even though some of the methods for deriving WQC are very
complex and include mechanisms to predict chronic effects and differences in
C-253
-------
Giesy Briefing Document, Minneapolis, MN Sept 14-15, 1993
sensitivities among species, they generally do not include the wildlife species, such
as mink, eagles and other fish-eating water birds, that may eat contaminated aquatic
organisms exclusively. The currently used risk assessment procedures do not account
for exposures to complex mixtures of potentially hazardous compounds, that can
cause subtle, long-term, chronic effects on the dynamics of wildlife populations.
Typically, these agents cause effects on the offspring of contaminated adults through
decreased fecundity or even reproductive failure. These phenomena can affect
wildlife populations directly over longer periods of time.(8) Documented adverse
effects in wildlife populations are consistent with the types and frequencies of effects
observed in laboratory studies of the effects of p-PCH. These effects have been
observed in fish,(215) birds/8-22-210' turtles,(216) and mink(217) and humans."91'192'
The use of non-conventional real-world bioeffects and reproductive endpoints
to establish LOAEL values and NOAEL for persistent, lipophilic, toxic contaminants will
require new testing protocols and data on the reproductive outcomes of serial exposed
generations of human and wildlife populations. Although recommended by several
authors, (65-66) the effectiveness of regulations to protect wildlife, such as birds, have
rarely been assessed by conducting field studies. Regulations have generally been set
based on modelled expected results. The longitudinal studies of behavioral effects of
contaminants on children of parents who have been exposed to PCBs through eating
contaminated Great Lakes fishes is providing some of these data for humans .<193"19e>
Long-term studies of herring gulls in the Great Lakes region have produced a twenty-
year long record of bioeffects which have been linked to chemicals.<169) Similar
studies conducted for shorter periods of time have demonstrated the same types of
effects in other species of colonial, fish-eating water birds/22-68-1611
Question 5
Our research indictes that this assumption is correct and that there is a fairly steep
dose-response curve for TCDD in the species which we have worked. This makes it
difficult to relate tissue concentrations of TCDD to responses (adverse outcomes) at
a particular geographic region. However, as was described above, when these
relationships are examined over greater geographic regions, then relationships can be
observed.
Question 6
As discussed above, we need to use the best estimates available. Our research has
indicated that there are observed adverse effects which can be related to
concentrations of TCDD-EQ in tissues that are currently occurring at exposures which
the Interim document indicates would be of little risk. The use of the pheasant as a
C-254
-------
Giesy Briefing Document, Minneapolis, MN Sept 14-15, 1993
model species to protect birds at the top of the food web is probably inappropriate.
Question 7
It is true that there is limited information on marine organisms, but there is some
limited information of the effects of TCDD and similar compounds on seals. This area
could certainly use more research. See reviews in Ambio, Vol. 21 Number 8 1992
and The Journal of Toxicology and Environmental Health, Vol 33, Number 4. See
information discussed above.
Question 8
It is true that there is a great deal of uncertainty in the estimation of exposure to
TCDD. This is due to several factors. First, it appears that can be adverse effects
form exposure to water at concentrations, which are currently below detection limits.
The available fraction is an important parameter, which has a great impact on the
results of risk assessments and is difficult to predict or measure empirically.
Furthermore, the forms in which TCDD occurs can change overtime. Simply because
a small fraction of the total is not available, does not mean that it can not be
accumulated into target tissues to concentrations which are toxicologically relevant.
The fact that the uncertainties are multiplicative results in a great deal of uncertainty
in the hazard assessments. It is our opinion that the uncertainty caused by
estimations of the exposure are greater than those due to estimates of the dose-
response relationships.
Question 9
This statement is missleading. At the lower trophic levels the greatest exposure is
from water. At trophic levels above 3 or 4 the proportion obtained from the diet is
the greater fraction. For those species, which will have the greatest exposure and
thus are exposed to the greatest risk, the primary route of exposure is through the
diet.
Question 10
There is nothing special about TCDD, which would preclude modeling it's
environmental fate with existing simulation models. Thus there are models available.
it is the parameter estimate to be used in the models, which are lacking. There is little
C-255
-------
Giesy Briefing Document, Minneapolis, MN Sept 14-15, 1993
information on the forms of TCDD in varipous compartments, which could be used to
paramererize models or verify model output.
Question 11
This issue has been discussed above
Question 12
There is definately a lack of information on bioaccumulation and this results in the
greatest uncertainty in the risk assessment procedure. For TCDD we propose that
wildlife can serve as sentinals and tha tmore effort should be given to assessing the
use of integrative models of toxic potency.
Wildlife as Environmental Sentinels
The field of Ecotoxicology and especially Wildlife Toxicology are relatively new fields
of endeavor.(218) However, even in ancient times humans exhibited an awareness of
the condition of birds: Hence the Greek maxim a bad crow lays a bad egg.(2*8] Even
though there are records of human-caused episodes of toxicological effects in
populations of wildlife species from ancient times, only with the wide-spread use of
synthetic organic chemicals since World War II have large-scale chemically-induced
wildlife epizootics occurred. The ability to describe these effects and to understand
the role of toxic chemicals has required developments in a number of scientific fields,
including environmental, analytical chemistry and wildlife biochemical toxicology.
Limited knowledge of the basic biochemistry, physiology and natural histories of
wildlife species has also limited the ability to document and understand the effects of
contaminants on wildlife populations. The analytical tools of these multidisciplinary
fields are providing a comprehensive picture of the effects of trace concentrations of
toxic, synthetic hydrocarbons in wildlife species.(8/69'161) The study of effects of toxic
chemicals on wildlife populations is limited by the complexity of a large number of
species interacting with each other as well as their natural habitat and human-caused
physical changes to their environment along with the effects of synthetic organic
chemicals. However, through the efforts of multidisciplinary research teams of
experts in environmental chemistry, chemodynamics, toxicology, biochemistry,
pathology and ecology rapid progress is being made. The ability to establish the
cause-effect relationships between concentrations of chemicals in complex mixtures
with adverse effects and to be able to predict the fates and effects of these chemicals
in the ecosystem is developing rapidly.(161)
In our efforts to investigate the linkages between certain, synthetic halogenated
C-256
-------
Giesy Briefing Document, Minneapolis, MN Sept 14-15, 1993
chemicals in the Great Lakes and effects on populations of wild birds we have
followed Koch's postulates (Table 5). To date, we have observed adverse effects,
including egg mortality and deformities in chicks. These effects have been correlated
with the concentrations of several compounds, but the strongest correlations are with
the concentrations of TCDD-EQ. The types of effects observed are the same as those
that can be caused by these compounds in laboratory studies of animal models.
Furthermore, the concentrations required to elicit the responses are in the same range
that would be expected from laboratory studies. Thus, we feel as if we have
completed the first four of the postulates. It is likely that the effects observed are due
to the TCDD-EQ contributed by PCDD, PCDF and p-PCBs. This conclusion is further
supported by the fact that when the p-PCH fraction is removed by selective carbon
column chromatography we were able to remove the fraction that caused the
induction in the H4IIE bioassay.
We have not completed the fifth postulate in vivo because it is difficult to
conduct studies of this type in the laboratory with the same species that occur in the
wild. This is because of many logistical problems, such as the fact that an
uncontaminated source of organisms. Evev if uncontaminated wildlife species were
available, our techniques to rear them in the laboratory are not sufficiently well
developed to allow valid laboratory comparisons. Instead, we have fed fish from the
Great Lakes to chickens/1731 The results of these studies indicated that the same
types of effects observed under field conditions could be induced in these species
under laboratory conditions. Finally, relative to studies of the potential for these
compounds to cause deformities, it is difficult to demonstrate statistically significant
effects due to the small rates of deformities observed and the small sample sizes that
are possible in laboratory studies.
A method for conducting wildlife hazard assessments that uses results of field
and laboratory studies on target and surrogate domestic species, along with
environmental monitoring and chemical analyses, to determine the appropriate WQC
to protect sensitive wildlife species from the adverse effects of p-PCHs has been
developed/67' When compared to an assessment, which used human cancer risk
assessments and visual cognition and memory effects as endpoints, the WQC to
protect humans and wildlife were found to be similar in magnitude/1921
When possible it is best to use actual concentrations of PCBs or TCDD-EQ in
tissues of fish or birds or their eggs, as the most proximate measure of exposure. If
this is not possible to make measurements directly on the target organisms, the next
best estimate of exposure is to measure the concentrations of compounds in their
food. Use of concentrations in water is the least accurate estimate of exposure and
thus includes more safety factors to assure protection of all species from any adverse
effects.
Here, we have established that the types of effects observed in wildlife
populations are similar to those observed in mammals, which are used as models of
the potential effects on humans and that environmental criteria to protect wildlife are
C-257
-------
Giesy Briefing Document, Minneapolis, MN Sept 14-15, 1993
similar to those predicted to protect human health. The allowable effects in wildlife
are greater than those in humans, but their exposure is greater, since they can not
restrict their intake. The result is that if wildlife species, such as birds, which are at
the top of the food chain are protected, it is likely that human health in the same
ecosystem will also be protected. We do not imply that if effects are observed in
wildlife, there will necessarily be similar effects in humans. Conversely, if wildlife are
protected form adverse effects due to environmental exposures to contaminants, it
is likely that humans will be protected from the same sources of exposure. However,
it must be remembered that there may be other significant exposures to toxicants,
such as household and occupational exposures. Thus, while wildlife biomanitoring
can not protect humans from all exposures, monitoring their responses can be a useful
procedure. For this reason, wildlife, especially colonial water birds of the Great Lakes
region have been advocated as sentinels of environmental exposures.'2'16/69'213'216)
Question 13
I do not think that the use of the BAF from Lake Ontario is appropriate unless there
is no other information on which to base the risk assessment. It can be used as a
starting point and the uncertainty introduced by using this value assessed with
sensitivity analyses.
Question 14
Our results (see briefing package on the effects of TCDD on rainbow trout) indicates
that the BMF between fish and their food was relatively small (see table on BMF
values in Appendix I). Values for colonial fish-eating water birds seems to be
approximately 30X from food (fish) to egg(E2>. The apparent accumulation factor for
TCDD-EQ from forage fish into eagle eggs is approximately 200. However, this is
probably due, in part to the fact that eagles take birds as well as fish in their diet.
Question 15
The greatest uncertainties in the risk assessment procedure are involved in the
estimation of accumulation from water or sediment. We estimate this uncertainty to
be approximately 10OOX while that from within and among-species sensitivity is only
approximately 10X to 10OX. Even the biomagnification factors from fish to birds is
only a factor of approximately 2X.
C-258
-------
Giesy Briefing Document, Minneapolis, MN Sept 14-15, 1993
Question 16
Yes, this would be an adequate method to assure protection of other species. See
the discussion above of the relationship between WQC based on different species and
endpoints. Based on our research, the WQC to protect rainbow trout from any averse
effects would be approximately 0.02 ppq. Thus, if eagles were used in the hazard
assessment rainbow trout would also be protected, since the value to be protective
of eagles is approximately 10X less. It is difficult to compare the results of our
studies with rainbow trout to those with lake trout, which seem to be mopre
sensitive, but by comparing the concentrations in eggs that cause effects in these
species, it is probably safe to say that the lake trout are approximately 10X more
sensitive than rainbow trout. Thus, the value to protect eagle would also be
protective of lake trout.
Question 17
As discussed above, the models to predict concentrations of TCDD in fish and their
preditors are limited and uncertain. One method to increase the predictability would
be to use semi-permiable monitoring devices (SPMDs) to predict concentrations in fish
before operation of the mill. Subsequent monitoring should focus and fish and/or
SPMDs or a calibrated SPMD.
Question 18
They seem to be the best available, but as is indicated in the briefing document, they
are of limited utility and will result in exposure estimates, which ahve an uncertainty
of at least 1.00X to as much as 1000X.
Question 19
The best that could be done would be to use a loading simulation model coupled to
sedimaent dynamics models. This type of modeling has been used to understand the
sources and sinks of PCBs in the lower Fox River in Wisconsin, but the information
required for the model has required six year and some $10,000,000 to obtain. These
type of modeling efforts are very site-specific. Thus the techniques can be used, but
the data is missing.
C-259
-------
LITERATURE CITED
1. A.P. Gillman, G.A. Fox, D.B. Peakall, S.M. Teeple, T.R.
Carroll and 6.T. Haymes. Reproductive parameters and egg
contaminant levels of Great Lakes herring gulls. J". Wildl.
Manage. 41, 458-468 (1977).
2. D.V. Weseloh, P. Mineau and D.J. Hallett. Organochlorine
contaminants and trends in reproduction in Great Lakes herring
gulls, 1974-1978. Proc 44th N. Amer. Wildlife & Nat. Reso.
Conf. Wildlife Management Inst., Washington, DC. (1979) pp.
543-557.
3. D.V. Weseloh, S.M. Teeple and M. Gilbertson. Double-crested
cormorants of the Great Lakes: Egg laying parameters,
reproductive failure and contaminant residues in eggs, Lake
Huron 1972-1973. Can. J. Zool. 61, 427-436 (1983).
4. D.V. Weseloh, T.W. Custer and B.M. Braune. Organochlorine
contaminants in eggs of common terns from the Canadian Great
Lakes, 1981. Environ. Pollut. 59, 141-160 (1989).
5. R.J. Allan, R.J.; Ball, A.J.; Cairns, V.W.; Fox, G.A.; Gilman,
A.P.; Peakall, A.P.; Piekarz, D.A.; Van Oosdam, J.C.;
Villeneuve, D.C.; Williams, D.T. Toxic Chemicals in the Great
Lakes and Associated Effects. Environ. Canada, Dept.
Fisheries & Oceans, Health & Welfare Canada. Vols I & II,
758p. (1991).
6. R. Eisler. Polychlorinated biphenyl hazards to fish, wildlife
and invertebrates: A synoptic review. USDI, USFWS Contaminant
Hazard Review No. 7. (1986) 72 p.
7. J.P. Ludwig, J.P. Giesy, C.L., W. Bowerman, S. Heaton, R.
Aulerich, S. Bursian, H.J. Auman, P.D. Jones, L.L. Williams,
D.E. Tillitt and M. Gilbertson. A comparison of Water Quality
Criteria in the Great Lakes Basin Based on Human or Wildlife
Health. i7. Great Lakes Res. (in press)
7. R. Eisler. Dioxin hazards to fish, wildlife and invertebrates:
A synoptic review. USDI, USFWS Contaminant Hazard Review No.
8. (1986) 37 p.
8. M. Gilbertson, T.J. Kubiak, J.P. Ludwig and G. Fox. Great
Lakes embryo mortality, edema, and deformities syndrome
(GLEMEDS) in colonial fish-eating birds: Similarity to chick
edema disease. J. Toxicol Environ. Health. 33, 455-520 (1991).
9. M. Gilbertson. Etiology of chick edema disease in herring
gulls in the lower Great Lakes. Chemosphere. 12, 357-370
(1983).
C-260
-------
Giesy Briefing Document, Minneapolis/ MN, Sept. 14-15, 1993
10. T.J. Kubiak, H.J. Harris/ L.M. Smith, T.R. Schwartz, D.L.
Stalling, J.A. Trick, L. Sileo, D.E. Docherty, and T.C.
Erdman. Hicrocontaminants and reproductive impairment of the
Forster's Tern on Green Bay, Lake Michigan-1983. Arch.
Environ. Contam. Toxicol. 18, 706-727 (1989).
11. C.F. Tumasonis, B. Bush and F.D. Baker. PCB levels in egg
yolks associated with embryonic mortality and deformity of
hatched chicks. Arch. Environ. Contamn. Toxicol. 1, 312-324
(1973).
12. M. Gilbertson, R.D. Morriss and R.A. Hunter. Abnormal chicks
and PCB residue levels in eggs of colonial birds on the lower
great lakes (1971-1973). Auk. 93, 435-442 (1976).
13. D.B. Peakall. Accumulation and effects on birds. In: J.S.
Aaid (Ed.) PCBs in the Environment Vol. II, (Chap 3). CRC
Press, Boca Raton, FL. (1986) pp. 31-47.
14. D.B. Peakall. Known effects of pollutants on fish-eating
birds in the Great Lakes of North America. In: N.W. Schmidtke
(Ed.). Toxic Contamination in Large Lakes, Vol. II. Chronic
Effects of Toxic Contaminants in Large Lakes. Lewis
Publishers, Chelsea, MI. (1988) pp. 39-54.
15. D.B. Peakall and G.A. Fox. Toxicological investigations of
pollutant-related effects in Great Lakes Gulls. .Enviroji.
Health. Perspec. 71, 187-193 (1987).
16. J. Struger, and D.V. Weseloh. Great Lakes Caspian terns: Egg
contaminants and biological implications. Colon. Water Birds.
8, 142-149 (1985).
17. J. Struger, D.V. Weseloh, D.J. Hallett and P. Mineau.
Organochlorine contaminants in herring gull eggs from the
Detroit and Niagara Rivers and Saginaw Bay (1978-1982):
Contaminant discriminants. J. Great Lakes Res. 11, 223-230
(1985).
18. D.W. Anderson and J.J. Hickey. Eggshell changes in certain
North American birds. Proc. XVth Int. Ornithol. Cong. pp.
514-540 (1969).
19. J.E. Elliott, R.J. Norstrom and J.A. Keith.
Organochlorinesand eggshell thinning in northern gannets (Sula
bassanus) from eastern Canada. Environ. Pollut. 52, 81-102
(1988).
20. P.C. Baumann and D.M. Whittle. The status of selected organics
in the Laurentian Great Lakes: An overview of DDT, PCBs,
C-261
-------
Giesy Briefing Document, Minneapolis/ MN, Sept. 14-15, 1993
dioxins, furans and aromatic hydrocarbons. Aquatic Toxicol.
11, 241-257 (1988).
21. I.M. Price and D.V. Weseloh. Increased numbers and
productivity of double-crested cormorants Phalacrocorax
auritus) on Lake Ontario. Canad. Field. Natural. 100, 474-482
(1986).
22. D.E. Tillitt, G.T. Ankley, J.P. Giesy, J.P. Ludwig, H. Kurita-
Matsuba, D.V. Weseloh, P.S. Ross, C. Bishop, L. Sileo, K.L.
Stroxnberg, J. Larson, and T.J. Kubiak. Polychlorinated
biphenyls resideues and egg mortality in double-crested
cormorants from the Great Lakes. Environ. Toxicol. Chem. 11,
1281-1288 (1992).
23. N. Yamashita, S. Tanabe, J.P. Ludwig, H. Kurita, M.E. Ludwig,
and R Tatsukawa. Embryonic Abnormalities and Organochlorine
Contamination in Double-Crested Cormorants ( Phalacrocorax
auritus ) and Caspian Terns (Hydroprogne caspia) from the
Upper Great Lakes, Collected in 1988. Environ. Pollut. 79,
163-173 (1992).
24. U. Selestrom, B. Jansson, A. Kierkegaard, C. de Wit, T. Odsjo
and M. Olson. Polybrominated diphenyl ethers (PBDE) in
biological Samples from the Swedish environment. Chemosphere
26f 1703-1718 (1993).
24. G.A. Fox, D.V. Weseloh, T.J. Kubiak and T.C. Erdman.
Reproductive outcomes in colonial fish-eating birds: A
biomarker for developmental toxicants in Great Lakes food
chains. J. Great Lakes Res. 17, 153-157 (1991).
25. P.V. Hodson, M. McWhirther, K. Ralph, B. Gray, D. Thivierge,
J.H. Carey amd M.C. Levesgue. Effects of bleached kraft mill
effluent on fish in the St. Maurice river, Quebec. Environ.
Toxicol. Chem. 11, 1635-1651 (1992).
26. J.A. Nosek, J.R. Sullivan, T.E. Amundson, S.R. Craven, L.M.
Miller, A.G. Fitspatrick, M.E. Cook and R.E. Peterson.
Embryotoxicity of 2,3,7,8-tetrachlorodibenzo-p-dioxin in ring-
necked pheasants. Environ. Toxicol. Chem. 12:, 1215-1222
(1992).
27. G.R. Higgenbotham, A. Huang, D. Firestone, J. Verrett, J.
Reese and A.D. Campbell. Chemical and toxicological
evaluations of isolated and synthetic chloro-derivatives of
dibenzo-p-dioxin. Nature. 220, 702-703 (1986).
28. J.P. Ludwig, SERE Group Unpublished data.
C-262
-------
Giesy Briefing Document/ Minneapolis, MN, Sept. 14-15, 1993
29. D.E. Tillitt. Characterization studies of the H4IIE Bloassay
for Assessment of Planar Halogenated Hydrocarbons in Fish and
Wildlife. Ph.D. Dissertation, Michigan State University, E.
Lansing (1989) p. 152.
30. M.J. Verrett. US Food & Drug Administration Memorandum to D.
Firestone, dated 8 June (1976).
31. D.S. Henschel, Vo, M.T. Hehn, B.M., Steeves, J.D. The
embryotoxicity of 2,3,7,8-Tetrachlorodibenzo-p-dioxin on avian
embryos: 1. LD50 studies. Toxicol. Appl. Pharmacol.(In press)
32. D.H. White, J.T. Seginak and D.J. Hoffman. Dioxins and furans
linked to reproductive impairment in Arkansas wood ducks. J.
Wildlife Mangt. (in press)
33. D.S. Henshel, Hehn, B.M., Vo, M.T., Steeves, J.D. A short-
term test for dioxin teratogenicity using chicken embryos. In
Environmental Toxicology and Risk Assessment: Vol 2 ASTM STP
1173, J.W. Gorsuch, F.J. Dwyer, C.G. Ingesoll and T.W. LaPoint
(Eds.) Amer. Soc. Material. Test., Philadelphia, (in press)
34. A. Poland and E. Glover. Comparison of 2,3,7,8-
Tetrachlordibenzo-p-dioxin , a potent inducer of aryl
hydrocarbon hydroxylase, with 3-methylcholanthrene. Mol.
Pharmacol. 10, 349-359 (1974).
35. S. Tanabe, N. Kannan, M. Fukushima, T. Okamoto, T. Wakimoto,
and R. Tatsukawa. Persistent organochlorines in Japanese
coastal waters: An introspective summary from a far east
developed nation. Mar. Pollut. Bull. 20, 344-352 (1989).
36. R.J. Norstrom. Bioaccumulation of polychlorinated biphenyls
in Canadian wildlife. In: Hazards, Decontamination and
Replacement of PCBs. Plenum Publishing. (1987) pp. 1-16.
37. R.J. Norstrom, M. Simon, D.C.G. Muir and R.E. Schweinsburg.
Organochlorine contaminants in arctic marine food chains:
Identification, geographical distribution and temporal trends
in polar bears. Environ. Sci. Technol. 22, 1063-1071 (1988).
38. P. De Voogt and U.A.Th. Brinkman. Production, properties and
usage of polychlorinated biphenyls. In: Kimbrough and Jensen
(Eds.) Halogenated Biphenyls, Terphenyls, Napthalenes,
Dibenzodioxins and Related Products. Elsevier Science.
Publishers B.V. (1989) pp. 3-45.
39. P. De Voogt, J.W.M. Wegener, J.C. Klamer, G.A. Van Zijl, and
H. Govers. Prediction of environmental fate and effects of
heteroatomic polycyclic aromatics by QSARs: The position of n-
C-263
-------
Giesy Briefing Document, Minneapolis, MN, Sept. 14-15, 1993
octanol/water partition coefficients. Biomed. Environ. Sci.
1, 194-209 (1988).
40. P. De Voogt, D.E. Wells, L. Reutergardh, and U.A.Th. Brinkman.
Biological activity, determination and occurrence of planar,
mono- and di-ortho FCBs. Intern. J. Environ. Chem. 40, 1-46
(1990).
41. S. Tanabe, N. Kannan, A. Subramanian, S. Watanabe, and R.
Tatsukawa. Highly toxic coplanar PCBs: occurrence, source,
persistency and toxic implications to wildlife and humans.
Environ. Pollut. 47, 147-163 (1987).
43. L.S. Birnbaum. The role of structure in the disposition of
halogenated aromatic xenobiotics. Environ. Health
Perspect. 61, 11-20, (1985).
44. D.C.6. Muir, R.J. Norstrom, and M. Simon. Organochlorine
contaminants in arctic marine food chains: Accumulation of
specific polychlorinated biphenyls and chlordane-related
compounds. Environ. Sci. Technol. 22, 1071-1079 (1988).
45. A. Parkinson and S. Safe. Mammalian biologic and toxic effects
of PCBs. In: S. Safe and O. Hutzinger (Eds.) Polychlorinated
Biphenyls (PCBs): Mammalian and Environmental Toxicology.
Springer-Verlag, Berlin. (1989) pp 49-75.
46. S. Tanabe, N. Kannan, T. Wakimoto, R. Tatsukawa, T. Okamoto,
and Y. Masuda. Isomer-specific determination and toxic
evaluation of potentially hazardous coplanar PCBs,
dibenzofurans and dioxins in the tissues of "Yusho" PCB
poisoning victim and in the causal oil. Toxicol. Environ.
Chem. 24, 215-231 (1989).
46. S. Safe, S. Bandiera, T. Sawyer, B. Zmudzka, 6. Mason, M.
Romkes, M.A. Denomme, J. Sparling, A.B. Okey, and T. Fujita.
Effects of structure on binding to the 2,3,7,8-TCDD receptor
protein and AHH induction-halogenated biphenyls. Environ.
Health Perspect. 61, 21-33 (1985).
47. S. Safe. Determination of 2,3,7,8-TCDD toxic equivalent
factors (TEFs): Support for the use of the in vitro AHH
induction assay. Chemosphere 16, 791-802 (1987).
48. S. Safe Polychlorinated biphenyls (PCBs), dibenzo-p-dioxins
(PCDDs), dibenzofurans (PCDFs), and related compounds:
Environmental and mechanistic considerations which support the
development of toxic equivalency factors (TEFs). Crit. Rev.
Toxicol. 21:51-88 (1990).
C-264
-------
Giesy Briefing Document, Minneapolis, MN, Sept. 14-15, 1993
49. V.A. McFarland and J.U. Clarke. Environmental occurrence,
abundance, and potential toxicity of polychlorinated biphenyl
congeners: Considerations for a congener-specific analysis.
Environ. Health Perspect. 81, 225-239 (1989).
50. L.M. Smith, Schwartz, T. R., Feltz, K., and Kubiak, T. J.
Determination and occurrence of AHH-active polychlorinated
biphenyls, 2,3,7,8 TCDD and 2,3,7,8 TCDF in Lake Michigan
sediment and biota: the question of their relative
toxicological significance. Chemosphere 21, 1063-1085
(1990) .
51. L.M. Smith, D.L. Stalling and L. Johnson. Determination of
part—per—trillion levels of polychlorinated dibenzofurans and
dioxins in environmental samples. Anal. Chem. 56, 1830-1842
(1984).
52. P.D. Jones, 6.T. Ankley, D.A. Best, R. Crawford, N. Krishnan,
J.P. Giesy, T.J. Kubiak, J.P. Ludwig, J.L. News ted, D.E.
Tillitt and D.A. Verbrugge. Biomagnification of Bioassay-
Derived 2,3,7,8-Tetrachlor-dibenzo-p-dioxin Equivalents.
Chemosphere. 26, 1203-1212 (1993).
52. J.P. Whitlock. The regulation of gene expression by 2,3,7,8-
tetrachlordibenzo-p-dioxin. Pharmacol. Rev. 39, 147-161
(1987) .
53. J.A. Goldstein, P. Linko, J.N. Huckins and D.L. Stalling.
Structure-activity relationships of chlorinated benzenes as
indicators of different forms of cytochrome P450 in rat liver.
Chem. B±ol. Interacts. 41, 131-139 (1982).
54. J.A. Goldstein. structure-activity relationships for the
biochemical effects and the relationship to toxicity. In:
Kimbrough, R.D. and A.A. Jensen (eds.) Halogenated Biphenyls,
Terphenyls, Napthalenes, Dibenzodioxins and Related Products
1st ed.. Vol. 4-Topics in Environmental Health. Elsevier,
Amsterdam New York, London. (1980) pp. 151-190.
55. A. Poland and C. Knutson. 2,3,7,8-tetrachlorodibenzo-p-dioxin
and related halogenated aromatic hydrocarbons: Examination of
the Mechanism of toxicity. Ann. Rev. Pharmacol. Toxicol. 22,
517-554 (1982).
56. D.W. Nebert. The Ah locus: Genetic differences in toxicity,
cancer, mutation, and birth defects. Crit. Rev. Toxicol. 20,
153-174 (1990).
57. J.P. Landers and N.J. Bunce. The Ah-receptor and the mechanism
C-265
-------
Giesy Briefing Document/ Minneapolis, UN, Sept. 14-15, 1993
of dioxin toxicity. Biochem. J. 276, 273-287(1991).
58. M.S. Denison, P.A. Bank and E.F. Yao. DNA sequence-specific
binding of transformed Ah receptor to a dioxin responsive
enhancer: Looks arn't everything. Chemosphere. 25, 33-36
(1992).
59. F.J. Gonzalez, R.H. Tukey, and D.W. Nebert. Structural gene
products of the Ah locus. Transcriptional regulation of
cytochrome Pl-450 and P3-450 mRNA levels by 3-
methylcholanthrine. Mol. Pharmacol. 26, 117-121 (1984).
60. M.S. Denison, J.M. Fisher, and J.P. Whit lock. Protein-DNA
interactions at recognition sites for the dioxin-Ah receptor
complex. J. Biol. Chem. 264, 16478-16482 (1989).
61. R.H. Pratt. Receptor-dependent mechanisms of glucocorticoid
and dioxin-induced cleft plate. Environ. Health Perspect. 61,
35-40 (1985).
62. J.B. Silkworth, D.S. Cutler, and 6. Sack. Immuno-toxicity of
2,3,7,8-tetrachlorodibenzo-p-dioxin in a complex environmental
mixture from the Love Canal. Fundam. Appl. Toxicol. 12, 303-
312 (1989).
62. J.D. McKinney, J. Fawkes, S. Jordan, K. Chae, S. Oatley, R.E.
Coleman, and W. Briner. 2,3,7,8-tetrachlorodibenzo-p-dioxin
(TCDD) as a potent and persistent thyroxine agonist: A
mechanistic model for toxicity based on molecular reactivity.
.Environ. Health Perspect. 61, 41-53 (1985).
64. J.B. Silkworth, D.S. Cutler, P.W. O'Keefe and T. Lipniskas.
Potentiation and antagonism of 2,3,7,8-Tetrachlor-p-dibenzo-
dioxin effects in a complex environmental mixture. Toxicol.
Appl. Pharmacol. 119, 236-247 (1993).
65. C.F. Mason, C.F. Biology of Freshwater Pollution. Second
edition. John Wiley and Sons, New York. (1991)
66. C.F. Mason, G.T. Sawyer, B. Keys, S. Bandiera, M. Romkes, J.
Piskorska-Pliszcynska, B. Zmudzka and S, Safe.
Polychlorinated dibenzofurans (PCDFs): Correlation between in
vivo and in vitro structure-activity relationships.
Toxicology. 37, 1-12 (1985).
67. E.E. McConnell. Acute and chronic toxicity and carcinogenesis
in animals. In Halogenated Biphenyls Terphenyls, Napthalenes,
Dibenzo-p-dioxins and Related Products. Kimbrough and Jensen
(eds.) Elsevier Science Publishers B.V. (1989) pp. 161-193.
C-266
-------
Giesy Briefing Document, Minneapolis, MN, Sept. 14-15, 1993
68. R.E. Morrissey and B.A. Schwetz. Reproductive and
developmental toxicity in animals. In Halogenated
B±phenyls,Terphenyls, Napthalenes, Dibenzodioxins and Related
Products R, A. Kimbrough and D. Jensen (eds.) Elsevier
Science Publishers B.V. (1989) pp. 195-225.
68. J.P. Ludwig, M.E. Ludwig, and H.J. Auman. Uptake of chemicals
from Great Lakes fish by double-crested cormorants and herring
gull chicks. Abstracts of the papers given at the Cause-
Effects Linkages II Symposium. September 27-28, 1991.
Traverse City, Michigan. (1991) pp. 23-34.
69. G.A. Fox, M. Gilbertson, A.P. Gillman and T.J. Kubiak. A
rationale for the use of colonial, fish-eating birds to
monitor the presence of developmental toxicants in Great Lakes
fish. J. Gret Lakes Res. 17, 151-152 (1991).
70. G.A. Fox. Eggshell Quality: It's Ecological and
Physiological Significance in a DDT-Contaminated Common Tern
Population, ffils. Bull. 88, 459-477 (1976).
70. K. Stromberg, US Fish and Wildlife Service, GReen Bay
Wisconsin, Unpublished Data.
71. T. Colborn and C. Clement. Chemically-Induced Alterations in
Sexual Development: The Wildlife\Human Connection. Vol. XXI.
Advances in Modern Environmental Toxicology. Princeton
Scientific Publishing Co,, Princeton, NY. (1992), 403 p.
72. J.L. Newsted and J.P. Giesy. Effects of 2,3,7,8-
Tetrachlordibenzo-p-dioxin (TCDD) on epidermanl growth factor
binding and protein kinase activity in the RTH149 rainbow
trout hepatoma cell line. Aquatic Tocicol. 23, 119-135
(1992).
73. D.M. Fry and T.K. Toon. DDT-induced feminization of gull
embryos. Sci. 213, 922-924 (1981).
74. A.P. Brouwer. Interference of 2,3,7,8-Tetrachlorbiphgenyl in
vitamin A (retinols) metabolism: Possible implications for
toxicity and carcinogenicity of polyhalogenated aromatic
hydrocarbons. Radiobiological Inst., Health Res. TNO,
Rijswijk, The Netherlands. (1987), 243 p.
75. B. Brunstrom and J. Lund. Differences between chick and turkey
embryos in sensitivity to 3,3/,4,4'-Tetrachlor-biphenyl and in
concentration/affinity of the hepatic receptor for 2,3,7,8-
Tetrachloro-p-dioxin. Comp. Biochem. 91C, 507-512 (1988).
76. B. Brunstrom and L. Andersson. Toxicity and 7-ethoxyresorufin
C-267
-------
Giesy Briefing Document/ Minneapolis, HN, Sept. 14-15, 1993
0-deethylase-inducing potency of coplanar polychlorinated
biphenyls (PCBs) in chick embryos. Arch. Toxicol. 62, 263-266
(1988).
77. B. Brunstrom. Sensitivity of embryos from duck, goose,
herring gull and various chicken breeds to 3,3',4,4'-
Tetrachlorobiphenyl. Poultry Sci. 67, 52-57, (1987).
78. B. Brunstrom. Mono-ortho-chlorinated chlorobiphenyls: toxicity
and induction of 7-ethoxyresorufin O-deethylase (EROD)
activity in chick embryos. Arch. Toxicol. 64, 188-192
(1990).
79. B. Brunstrom. Toxicity and EROD-Inducing potency of
polychlorinated biphenyls (PCBs) and polycyclic aromatic
hydrocarbons (PAHs) in avian embryos. Comp. Biochem. Toxicol.
100C, 241-243 (1991).
80. P. Beland, Deguise, S., Girard, C., Lagase, A., Martineau, D.,
Michaud, R., Muir, D., Norstrom, R., Pelletier, E., and
Shugart, L. 1993. Toxic compounds, health and reproductive
effects in St. Lawrence Beluga Whales. J. Gt. Lks. Res. (in
press).
81. D.E. Tillitt, J.P. Giesy, and G.T. Ankley. Characterization
of the H4IIE rat hepatoma cell bioassay as a tool for
assessing toxic potency of planar halogenated hydrocarbons in
environmental samples. Env. Sci. Technol. 25, 87-92 (1989).
82. D.E. Tillitt, G.T. Ankley, D.A. Verbrugge, J.P. Giesy, J.P.
Ludwig, and T.J. Kubiak. H4IIE rat hepatoma cell bioassay-
derived 2,3,7,8-tetrachloro-p-dioxin equivalents in colonial,
fish-eating water birds from the Great Lakes. Arch. Environ.
TOXicol. 21, 91-101 (1991).
83. A. Brouwer. Inhibition of thyroid hormone transport in plasma
of rats by polychlorinated biphenyls. In Biological
monitoring of exposure and the response at the subcellular
level to toxic substances. Arch. Toxicol. Suppl. 13, 440-445
(1989).
84. A.P. Brouwer. Binding of a metabolite of 3,4,3',4',
tetrachlorobiphenyl to transthyretin reduces serum vitamin A
transport by inhibiting the formation of the protein complex
carrying both retinol and thyroxine. Toxicol. Appl.
Pharmacol. 85, 301-312 (1989).
85. A.P. Brouwer, J. H. Reijnders, and J. H. Koeman.
Polychlorinated biphenyl ( PCB ) contaminated fish induces
vutamin A and thyroid hormone deficiency in the common seal
C-268
-------
Giesy Briefing Document, Minneapolis, MN, Sept. 14-15, 1993
(Phoca vitulina). Aguat. Toxicol. 15, 99-106 (1989).
86. M.H. Zile. Vitamin A homeostasis endangered by environmental
pollutants. Proc. Soc. Environ. Biol. Med. 201, 141-153.
(1992).
87. T.A. Mably, R.W. Moore and R.E. Peterson. In Utero and
lactational exposure of male rats to 2,3,7,8-
Tetrachlorodibenzo-p-dioxin: 1. Effects on androgenic status.
Toxicol. Appl. Pharmacol. 114, 97-107 (1992).
87. P.A. Spear, and Moon, T. W. Low dietary iodine and thyroid
anomalies in ring doves, Streptopelia risoria, exposed to
3,3',4,4' tetrachlorobiphenyl. Arch. Environ. Contam.
Toxicol. 14, 547-553 (1985).
88. T.A. Mably, R.W. Moore, R. W. Goy and R.E. Peterson. In Utero
and lactational exposure of male rats to 2,3,7,8-
Tetrachlorodibenzo-p-dioxin: 2. Effects on sexual behavior and
the regulation of lutenizing hormone secretion in adulthood.
Toxicol. Appl. Pharmacol. 114, 108-117 (1992).
89. T.A. Mably, D.L. Bjerke, R.W. Moore, A. Gendron-Fitzpatrick
and R.E. Peterson. In Utero and lactational exposure of male
rats to 2,3,7,8-Tetrachlorodibenzo-p-dioxin: 2. Effects on
sexual behavior and the regulation of lutenizing hormone
secretion in adulthood. Toxicol. Appl. Pharmacol. 114, 118-
126 (1992).
90. E. Enan, P.C.C. Liu and F. Matsumura. 2,3,7,8-Tetrachlor-
dibenzo-p-Dioxin causes reduction of glucose transporting
activities in the plasma membranes of adipose tissue and
pancreas from the guinea pig. J. Biological Chemistry. 267,
19785-19791 (1992).
91. S. Jensen and B. Jansson. Anthropogenic substances in seal
from the Baltic: Methyl sulphone metabolites of PCb and DDT.
Ambio 5, 257-260 (1976).
92. R.F. Seegal, B. Bush and K.O. Brosch. Sub-chronic exposure of
the adult rat to Aroclor* 1254 yields regionally-specific
changes in central dopaminergic function. Neurotox. 12, 55-66
(1991).
92. R.F. Seegal, B. Bush and K.O. Brosch. Comparison of effects
of Aroclors 1016 and 1260 on non-human primate catecholamine
function. Toxicol. 66, 145-163 (1991).
92. J.A. van Zorge, J.H. van Wijnen, R. M.C. Theelen, K. Olie amd
C-269
-------
Giesy Briefing Document, Minneapolis, MN, Sept. 14-15, 1993
M. van den Berg. Assessment of the toxicity of mixtures of
halogenated dibenzo-pdioxins and dibenzofurans by use of
toxicity equivalency factors (TEF). Chemosphere 19 f 1881-1895
(1989).
93. R.F. Seegal, B. Bush and W. Shain. Lightly chlorinated ortho-
substituted PCS congeners decrease dopamine in nonhuman
primate brain and in tissue culture. Toxicol. Applied
Pharmacol. 106, 136-144 (1990).
93. T. Zacharewski, T. Harris, S. Safe, H. Thoma, and O.
Hutzinger. Applications of the in vitro aryl hydrocarbon
hydroxylase induction assay for determining "2,3,7,8-
tetrachlorodibenzo-p-dioxin equivalents". Pyrolyzed brominated
flame retardants. Toxicology. 51, 177-189 (1988).
94. T. Zacharewski, L. Safe, S., Safe, B. Chittim, D. DeVault, K.
Wiberg, P.A. Berquist, and C. Rappe. Comparative analysis "of
polychlorinated dibenzo-p-dioxin and dibenzofuran congeners in
Great Lakes fish extracts by gas chromatography-mass
spectrometry and in vitro enzyme induction activities.
Environ. Sci. Technol. 23, 730-735 (1989).
95. N. Kannan, S. Tanabe, T. Wakimoto, and R. Tatsukawa. Coplanar
polychlorinated biphenyls in Aroclor and Kanechlor mixtures.
J. Off. Assoc. Anal. Chem. 70, 451-454 (1987).
95. W. Shain, B. Bush and R. Seegal. Neurotoxicity of
polychlorinated biphenyls: Structure-activity relationship
ofindividual congeners. Toxicol. Appl. Pharmacol. ill, 33-42
(1991).
96. M. Mora, H.J. Auman, J.P. Ludwig, J.P. Giesy, O.A. Verbrugge,
and M.E. Ludwig. PCBs and chlorinated insecticides in plasma
of Caspian terns: Relationships with age, productivity and
colony site tenacity. Arch. Environ. Toxicol. Contamn. 24,
320-331 (1992).
96. N. Kannan, S. Tanabe, and R. Tatsukawa. Toxic potential of
non-ortho and mono-ortho coplanar PCBs in commercial PCB
preparations: 2,3,7,8-T4 CDD toxicity equivalence factors
approach. Bull. Env. Contam. Toxicol. 41, 267-276 (1988).
97. R. Bannister, D. Davis, T. Zacharewski, I Tizard, and S, Safe.
Aroclor 1254 as a 2,3,7,8-tetrachlorodibenzo-p-dioxin
antagonist: Effects on enzyme induction and immunology.
Toxicology. ,46, 29-42 (1987).
98. R. Stahlmann, T. Schultz- Schalge, M. Korte, C. Renschler, I.
Baumann-Wilschke and D. Neubert. Chemosphere. 25:1207-1214
C-270
-------
Giesy Briefing Document, Minneapolis, MN, Sept. 14-15, 1993
(1992).
99. Bannister, R., M. Kelly and S. Safe. Synergistic interactions
of 2,3,7,8TCDD and 2,2',4,4',5,5'-hexachlorobiphenyl in
C57BL/GJ and OBA/2J mice: Role of the Ah receptor. Toxicol.
44, 159-169 (1987).
100. Bannister, R., L. Biegal, D. Davis, B. Astroff and S. Safe.
6-methyl-1.3.8-trichlorodibenzofuran (MCDF) as a 2,3,7,8-
tetrachlorodibenzo-p-dioxin antagonist in C57BL/6 mice.
Toxicology. 54, 139-154 (1989).
101. A. Hanberg, M. Stahlberg, A. Georgellis, C. deWit and V.6.
Ahlborg. Swedish Dioxin Survey: Evaluation of the H4IIE
Bioassay for Screening for Dioxin-Like Enzyme Induction.
Pharm. Toxicol. 69, 442-449 (1991).
102. P.D. Jones, J.P. Giesy, J.L. News ted, A. A. Verbrugge, D.L.
Beaver, G.T. Ankley, D.E. Tillitt and K.B. Lodge. 2,3,7,8-
Tetrachlordibenzo-p-dioxin equivalents in tissues of birds at
Green Bay, Wisconsin, USA. Arch. Environ. Toxicol. Safety. 24,
345-354 (1993).
103. D.E. Tillitt T. J. Kubiak, G. T. Ankley and J. P. Giesy.
Dioxin-like toxic potency in Forster's Tern eggs from Green
Bay, Lake Michigan, North America. Chemosphere. 26, 2079-2084
(1993).
108. D.E. Tillitt and J.P* Giesy, Michigan State University, E.
Lnsing, Unpublished Data.
104. J.A. Bradlaw and J.L. Casterline. Induction of enzymes in
cell cultures: a rapid screen for the detection of planar
chlorinated organic compounds. J". Assoc. Off. Anal. Chen. 62,
904-916 (1979).
105. T.W. Sawyer, A.D. Vatcher, and S. Safe. Comparative aryl
hydrocarbon hydroxylase induction activities of commercial
PCBs in Wistar rats and rat hepatoma H-4-II-E cells in
culture. Chemosphere. 13, 695-701 (1984).
106. J.L. Casterline, J.A. Bradlaw, B.J. Puma and Y. Ku.
Screening of fresh water fish extracts for enzyme-inducing
substances by an aryl hydrocarbon hydroxylase induction
bioassay technique. J. Assoc. Off. Anal. Chen. 66, 1136-1139
(1983).
107. G.T. Ankley, D.E. Tillitt, and J.P. Giesy. Maternal transfer
of bio-active polychlorinated aromatic hydrocarbons in
spawning Chinook salmon (Oncorhynchus tschavrytscha). Mar.
C-271
-------
Giesy Briefing Document, Minneapolis, MN, Sept. 14-15, 1993
Environ. Res. 28, 231-234 (1989).
107. J.P Ludwig, Auman, H. J., Kurita-Matsuba,H., Ludwig, M. E.,
Campbell, L.M., Giesy, J.P., Tillitt, D. E., Jones, P.,
Yamashita, N., Tanabe, S., and Tatsukawa, R. Caspian Tern
Reproduction in the Saginav Bay Ecosystem following a 100-year
Flood event. J. Gt. Lks. Res. 19, 96-108 (1993).
108. G.T. Ankley, D.E. Tillitt, J.P. Giesy, P.D. Jones, and D.A.
Verbrugge. Bioassay-derived 2,3,7,8-tetrachlorodibenzo-p-
dioxin equivalents (TCDD-EQ) in the flesh and eggs of lake
Michigan chinook salmon and possible implications for
reproduction. Can. J. Fish. Aquat. Sci. 48, 1685-1690 (1991).
109. G.T. Ankley, G. T. Niemi, K. B. Lodge, H. J. Harris, O. L.
Beaver, D. E. Tillitt, T. R. Schwartz, J. P. Giesy, P. D.
Jones and C. Hagley. Uptake of planar polychlorinated
biphenyls and 2,3,7,8-substituted polychlorinated
dibenzofurans and dibenzo-p-dioxins by birds nesting in the
Lower Fox River/Green Bay, Wisconsin. Arch. Environ. Contamn.
Toxicol. 24, 332-344 (1993).
110. L.L Williams, J.P. Giesy, N. DeGalan, D.A. Verbrugge, D.E.
Tillitt and G.T. Ankley. Prediction of concentrations of
2,3,7,8-TCDD equivalents (TCDD-EQ) in trimmed, Chinook salmon
filets from Lake Michigan from total Concentrations of PCBs
and fish size. Environ. Sci. Technol. 26, 1151-1159 (1992).
112. L.S. Birnbaum Distribution and excretion of 2,3,7,8-
Tetrachlordibenzo-p-dioxin in congenic strains of mice which
differ at the Ah locus. Drug. Metb. Dispos. 14, 34-40
(1986).
113. J. Haake, S. Safe, K. Mayura and T.D. Phillips. Aroclor 1254
as an antagonist of the teratogenicity of 2,3,7,8-
Tetrachlorodibenzo-p-dioxin. Toxicol. Lett. 38, 299-306
(1987).
114. B.T. Astroff, Zacharewski, S. Safe, M.P. Arlatto, A.
Parkinson. 6-Methyl-l,3,8-trichlorodibenzofuran as a 2,3,7,8-
Tetrachlorodibenzo-p-dioxin antagonist: Inhibition of the
induction of rat cytochrome P-450 isozymes and related
monooxygenase activities. Mol. Pharmacol. 33, 231-236 (1988).
115. L.M. Biegel, Harris, D. Davis, R. Rosengren, L. Safe, and S.
Safe. 2,2/,3,3'5,5'-Hexachlorobiphenyl as a 2,3,7,8-
Tetrachlordibenzo-p-dioxin antagonist in C57BL/6J mice.
Toxicol. Appl. Pharmacol. 97, 561-571, (1989).
116. D.M. Janz and C.D. Metcalf. Nonadditive interactions of
C-272
-------
Giesy Briefing Document, Minneapolis, MN, Sept. 14-15, 1993
mixtures of 2,3,7,8-TCDD and 3,3', 4,4' -tetrachlorobiphenyl. on
aryl hydrocarbon hydroxylase induction in rainbow trout
(Oncorhynchus mykiss). Chemosphere. 23, 467-472 (1991).
117. J. van der Kolk, A.P.J.M. van Birgelen, H Poiger and C.
Schlatter. Interactions of 2,2',4,4',5,5'-hexachlorobiphenyl
and 2,3,7,8-tetrachlorodibenzo-p-dioxin in a subchronic
feeding studyin the rat. Chemoshpere. 25, 2023-2027 (1992).
118. D. Schrenk, H.-P. Lipp, T. Wiesmiiller, H. Hagenmaier and K. W.
Bock. Assessment of biological activities of mixtures of
polychlorinated dibenzo-p-dioxins: Comparison between defined
mixtures and their constituents. Arch. Toxicol. 65, 114-118
(1990).
119. T.R. Schwartz and D.L. Stalling. Chemometric comparison of
polychlorinated biphenyl residues and toxicologically active
polychlorinated biphenyl congeners in the eggs of Forster's
terns (Sterna foster!). Arch. .Environ. Contamn. Toxicol. 20,
183-199 (1991).
120. B.G. Oliver and A. Niimi. Trophodynamic analysis of
polchlorinated biphenyl congeners and other chlorinated
hydrocarbons in the Lake Ontario ecosystem. Environ. Sci.
Technol. 22, 388-397 (1988).
121. B.G. Oliver, M.N. Charlton and R.W. Durham. Distribution,
redistribution, and geochemistry of polychlorinated biphenyl
congeners and other chlorinated hydrocarbons in lake Ontario
sediments. .Environ. Sci. Technol. 23, 200-208 (1989).
122. J.F.Brown. Metabolic alterations of PCB residues in aquatic
fauna: Distributions of cytochrome P4501A- and P4502B-like
activities. Mar. Environ. Res. 34, 261-266 (1992).
123. J.P. Boon, F. Eijgenraam, J.M. Everaats and J.C. Duinker. A
structure-activity relationship (SAR) approach towards
metabolism of PCBs in marine animals from different trophic
levels. Mari. Environ. Res. 27, 159-176 (1989).
124. J.T. Borlakoglu, J.P.G. Wilkins, and C.H. Walker.
Polychlorinated biphenyls in fish-eating sea birds- Molecular
features and metabolic interpretations. Marine Environ. Res.
24, 15-19 (1988).
125. C.H. Walker. Persistent pollutants in fish-eating sea birds-
bioaccumulation, metabolism and effects. Aqruat. Toxicol. 17,
293-324 (1990).
126. D. Broman, C. Naf, C. Rolff, Y. Zebuhr, B. Fry and J. Hobbie.
C-273
-------
Giesy Briefing Document, Minneapolis, MN, Sept. 14-15, 1993
Using ratios of stable nitrogen isotopes to estimate
bioaccumulation and f lux of polychlorinated dibenzo-p-dioxins
(PCDDs) in two food chains from the northern Baltic. Environ.
Toxicol. Chem. 11, 331-345 (1992).
127. R.J. Pruell, R.D. Bowen, S.J. Fluck, J.A. LiVosi, D.J. Cobb
and J.L. Lake. PCB congeners in American lobster, Homarus
americanus and winter flounder, Pseudopleuronectes americanus,
from New Bedford Harbor, Massachusetts. USEPA Environmental
Assessment Group Report/ Washington, DC (1988).
128. D.T.H.M. Sijm, A.L. Yarechewski, D.C.G. Muir, G.R.B. Webster,
W. Seinen, and A. Opperhuizen. Biotransformation and tissue
distribution of 1,2,3,7-tetrachlorodmibenzo-p-dioxin,
1,2,3,4,7-pentachlorodibenzo-p-dioxin and 2,3,4,7,8-
pentachlorodibenzofuran in rainbow trout. Chemosphere. 21,
845-866 (1990).
129. G. Sundstrom and O. Hutzinger. The synthesis of chlorinated
diphenyl ethers. Chemosphere. 5, 305-312 (1976).
130. O. Hutzinger, D.M. Nash, S. Safe, A.S.W. DeFreitas, R.J.
Norstrom, D.J. Wildish, and V. Zitko. Polychlorinated
biphenyls: Metabolic behavior of pure isomers in pigeons, rats
and brook trout. Sci. 178, 312-314 (1972).
131. J.C. Gage and S. Holm. The influence of moleckular structure
on the retention and excretion of polychlorinated biphenyls by
the mouse. Toxicol. Appl. Pharmacol. 36, 555-560 (1976).
132. H.B. Matthews and M.W. Anderson. Effect of chlorination on
the distribution and excretion of polychlorinated biphenyls.
Drug. MetaJbol. Disp. 3, 371-380 (1975).
133. H.B. Mattews and Tuey. The effect of chlorine position on the
distribution and excretion of four hexachlorobiphenyl isomers.
Toxicol. Appl. Pharmacol. 53, 377-388 (1980).
134. S. Focardi, C. Leonzio and C. Fossi. Variations in
polychlorinated biphenyl congener composition in eggs of
Mediterenian water birds in relation to their position in the
food chain. Environ. Pollut. 52, 243-255 (1988).
135. P. Kopponen, J. Tarhanen, J. Ruuskanen, R. TSrrSnen and S.
Karenlampi. Peat induces cytochrome P4501A1 in Hepa-1 cell
line, comparison with fly ashes from combustion of peat coal,
heavy fuel oil and hazardous waste. Chemosphere. 26, 1499-
1506 (1993).
136. P. Haglund, K.E. Egeback, and B. Jansson. Analysis of
C-274
-------
Giesy Briefing Document, Minneapolis, MN, Sept. 14-15, 1993
polybrominated dioxins and furans in vehicle exhaust.
Chemosphere 17, 2129-2140 (1988).
137. J. Koistenen and T. Nevalainen. Identification and level
estimation of aromatic coelutes of polychlorinated dibenzo-p-
dioxins and dibenzofurans in pulp mill products and wastes.
Environ. Sci. Technol. 26, 2499-2507 (1920)
138. H.R. Buser, L.O Kjeller, S.E. Swanson and C. Rappe. Methyl-,
polymethyl and alkylpolychlorodibenzofurans identified in pulp
mill-sludge and sediments. Environ. Sci. Technol. 20, 404-
408 (1989).
139. T. Nevalainen and J. Koistinen. Model compound synthesis for
the structure determination of new unknown planar aromatic
compounds originating from pulp mill. Chemosphere. 23, 1581-
1589 (1991).
140. U. Jarnberg, L. Asplund, C. de Wit, A.-K. Grafstrom, P.
Haglund, B. Jansson, K. Lexen, M. Strandell, M. Olson and B.
Johnson. Polychlorinated biphenyls and polychlorinated
naphthalenes in Swedish sediment and biota: Levels, patterns
and time trend. .Environ. Sci. Technol. 27, 1364-1374 (1993).
141. M. Becker, T. Phillips and S. Safe. Polychlorinated diphenyl
ethers: A review. Toxicol. Environ. Chem. 33, 189-200,
(1991).
142. Y.C. Chui, R.F. Addison and F.C.P. Law. Acute toxicity and
toxicokinetics of chlorinated diphenyl ethers in trout.
Xenobiotica. 20, 489-499 (1990).
143. Y.C. Chui, M.M. Hansell, R.F. Addison and F.C.P. Law. Effects
of chlorinated diphenyl ethers on the mixed-function oxidases
and ultrastructure of rat and trout liver. Xenobiotica. 20,
489-294 (1985).
144. C.J. Stafford. Halogenated diphenyl ethers identified in avian
tissues and eggs by GC/MS. Chemosphere 12, 1487-1495 (1983).
145. T.W. Custer, C.M. Bunck and C.J. Stafford. Organochlorine
concentrations in pre-fledging common terns at three Rhode
Island colonies. Colon. ffaterbirds. 8, 150-153 (1985).
146. L. Howie, R. Dickerson, D. Davis and S. Safe.
Immunosupressive and monooxygenase induction activities of
polychlorinated diphenyl ether congeners in C57BL/6N mice:
Quantitative structure-activity relationships. Toxicol. Appl.
Pharm. 105, 254-263 (1990).
C-275
-------
Giesy Briefing Document, Minneapolis/ MN, Sept. 14-15, 1993
147. A.J. Murk, J.H.J. van den Berg, J.H. Koeman and A. Brouwer.
The toxicity of tetrachlorobenzyltoluenes (Ugilec* 141) and
polychlorinatedbiphenyls (Aroclor* 1254 and PCB-77) compared
in Ah-responsive and Ah-non-responsive mice. Environ. Pollut.
72, 46-67 (1991).
148. R.F. Addison, R.F., H.E. ZinJc, D.E.Willis and J.J. Wrench.
Induction of hepatic mixed function oxidase activity in trout
(Salvelinus fontinalis) by Aroclor 1254 and some aromatic
hydrocarbon replacements. Toxicol. Appl. Pharm. 63, 166-172
(1982).
149. H. Friege, W. Stock, J. Alberti, A. Poppe, I. Juhnke, J. Knie
and W. Schiller. Environmental behavior of polychlorinated
mono-methly-substituted diphenyl-methanes (Me-PCDMs) in
comparison with polychlorinated biphenyls (PCBs) II:
Environmental residues and aquatic toxicity. Chemosphere 18,
1367-1378 (1989).
150. P.G. Wester and F. van der Vallc. Tetrachlorobenzyltoluenes in
eel from the Netherlands. Bull. Environ. Contamn Toxicol. 45,
69-73 (1990).
151. R.F. Addison, P.D. Hansen, H.-J. Pluta and D.E. Willis.
Effects of Ugilec-141, PCS substitute based on
tetrachlorobenzyltoluenes, on hepatic mono-oxygenase
indxuction in estuarine fish. Mar. Environ. Res. 31, 137-
144.(1991).
152. H.R. Buser and C. Rappe. Determination of
Polychlorodibenzothiophenes, the sulfur analogues of
polychlorodibenzofurans, using various gas
chromatographic/mass spectrometric techniques. Anal. Chem.
63, 1210-1217 (1991).
153. H.R. Buser, I.S. Dolezal, M. Wolfensberger and C. Rappe.
Polychlorodibenzo-thiophenes, the sulfur analogues of the
polychlorodibenzofurans identified in incineration samples.
Environ. Sci. Technol. 25, 1637-1643 (1991).
154. .D.R. Hilker, K.M. Aldous, R.M. Smith, P.W. O'Keefe, J.F.
Gierthy, J. Jurusik, S.W. Ribbons, D. Spink and R.J. Parillo.
Detection of sulfur analog of 2,3,7,8-TCDD in the environment.
Chemosphere 14, 1275-1284 (1985).
156. E.R. Barnhart, D.G. Patterson, J.A.H. MacBride, L.R.
Alexander, C.A. Alley and W.E. Turner. Polychlorinated
biphenylene production for quantitative reference material.
In: C. Rappe, G. Choudhary and L.H. Keith (Eds.) Chlorinated
C-276
-------
Giesy Briefing Document, Minneapolis, MN, Sept. 14-15, 1993
Dioxins and Dibenzofurans in Perspective. Lewis Publishers,
Chelsea Michigan. (1986), pp. 501-510.
158. T. Hashimoto and H. Miyata. Differences between Yusho and
other kinds of poisonings involving only PCBs. In: J.W. Waid
(Ed.) Vol.3 PCBs in the Environment. CRC Press, Boca Raton,
Florida (1987). pp. 1-26.
159. K. Lundgren and C. Rappe. Detection of aIkylated
polychlorodibenzo furans and alkylated polychlorodibenzo-p-
dioxins by tandem mass spectrometry for the analysis of
crustacean samples. Chemosphere. 23, 1591-1604 (1991).
160. G.J. Niemi, T.E Davis, G.D. Veith amd B. Vieux.
Organochlorine chemical residues in herring gulls, ring-billed
gulls and common terns of western lake Superior. Arch.
Environ. Contamn. Toxicol 15, 313-320 (1986).
161. G.A. Fox. Practical Causal Inference for Ecoepidemiologists.
J. Toxicol. Environ. Health. 33, 359-373 (1991).
162. W.J. Rogan, B.C. Gladen, K.L. Hung, L.L. Koong, L.Y. Shin,
J.S. Taylor, Y.C., U.D. Yang, N.B. Ragan, and C.C. Hsu.
Congenital poisoning by polychlorinated biphenyls and their
contaminants in Taiwan. Sci. 241, 334-336 (1988).
163. M. Gilbertson. The Niagara Labyrinth - The human ecology of
producing organochlorine chemicals. Can J. Fish. Aguat. Sci.
42, 1681-1692 (1985).
164. D.J. Hoffman, B.A. Rattner, L. Sileo, D. Docherty, and T.J.
Kubiak. Embryo-toxicity, teratogenicity, and aryl hydrocarbon
hydroxylase activity in Forster's Terns on Green Bay, Lake.
Michigan. Environ. Res. 42, 176-184 (1987).
165. M.O. Cheung, Gilbert, E. F., and Peterson, R. E.
Cardiovascular teratogenicity of 2,3,7,8 tatrachlorodibenze-p-
dioxin in the chick embryo. Toxicol. Appl. Pharmacol. 61,
197-204 (1981) .
166. T.R. Schwartz and T.J. Kubiak, US Fish and Wildlife Service,
unpublished data.
167. S. Tanabe, Ehime University, Matsuga, Japan, Unpublished Data.
168. J.P. Giesy. Michigan State University, Unpublished data
169. D.V. Weseloh, Bishop, C. A., Norstrom, R. J., and Fox, G. A.
Monitoring levels and effects of contaminants in herring gull
eggs on the Great Lakes 1974 - 1990. Abstracts of the Cause-
C-277
-------
Giesy Briefing Document/ Minneapolis, MN, Sept. 14-15/ 1993
Effects Linkanges II Symposium at Traverse City, MI September
27-28, 1991. Michigan Audubon Society. (1990) pp.29-31.
171. J.P. Ludwig and H. Kurita. ERA, Ann Arbor, Unpublished data.
172. W.W. Bowerman, D.A. Best, E.D. Evans, S. Postupalsky, M.S.
Martel, K. Kozie, R.L. Welch, R.H. Schell, K.F. Darling, J.c.
Rogers, T.J. Kubiak, D.E. Tillitt, T.R. Schwartz,P.D. Jones,
and J.P. Giesy. PCB Conentration in Plasma of nesting Bald
Eagles from the Great Lakes Basin, North America. In: O.
Hutzinger and H. Fiedler (Eds), . Organohalogen Compounds,
Vol 1: Toxicology, Environment Food Exposure-Risk.
Ecoinforma Press, Bayreuth, Germany. (1991), pp. 203-206.
173. C.L. Summer. An Avian Ecosystem Health Indicator: The
Reproductive Effects Induced by Feeding Great Lakes Fish to
White Leghorn Laying Hens. MS Thesis, Michigan State
University. (1992) I17p.
174. M. Gilbertson, International Joint Commission, Windsor, Ont.
175. R.J. Stevens and M.A. Nielson. Inter- and intra-lake
distributions of trace organic contaminants in surface waters
of the Great Lakes. J. Great Lakes Res. 15, 377-393 (1989).
176. J.E. Baker and S.J. Eisenreich. PCBs and PAHs as tracers of
particulate dynamics in large lakes. J. Great Lakes Res.
15, 84-103 (1989).
177. W.M. Strachan and S.J. Eisenrich. Mass balance of toxic
chemicals in the Great Lakes: The role of atmospheric
deposition. In: Appendix I from the Workshop on the
Estimation of Atmospheric Loadings of Toxic Chemicals to the
Great Lakes Basin, October 29-31, 1986, Scarborough Ontario.
International Joint Commission. Windsor, Ontario. (1988)
178. D.M. Swakhammer and D.E. Armstrong. Estimation of the
atmospheric and non-atmospheric contributions and losses of
PCBs for Lake Michigan on the basis of sediment records of
remote lakes. Environ. Sci. Tech. 20, 879-883 (1986).
179. P.M. Cook, R.J. Erickson, R.L. Spehar, S.P. Bradbury and G.T.
Ankley. Interum report on the assessment of 2,3,7,8-
Tetrachlordibenzo-pdioxin risk to aquatic life and associated
wildlife. US Environemtal Protection Agency, Washington, D.C.
(1993)
180. US EPA. Great Lakes Water Quality Initiative: Procedure for
Deriving Criteria for the Protection of Wildlife. US EPA
C-278
-------
Giesy Briefing Document, Minneapolis, MN, Sept. 14-15, 1993
Office of Water Regulations and Standards. Federal .Register,
March, Washington, D.C. (1993)
181. Anon. Proposed Water Quality Guidance for the Great Lakes.
USEPA, Federal ^Register March, 45, 79339, Washinton, D.C.
(1985), (1993) 308p.
182. Michigan Department of Natural Resources, Water Resources
Commission. Guidelines for Part 4. Water Quality Standards
(Rule 323.1057 (1985).
183. Michigan Department of Natural Resources, Water Resources
Commission.
183. US EPA. Maximum Contaminant Level, Safe Drinking Water Act
(1976).
184. International Joint Commission. Great Lakes Water Quality
Agreement of 1978 Revised. An IJC Agreement with Annexes and
Terms of Reference, between the U.S. and Canada, signed at
Ottowa, November 22 and Phosphorus Load Rduction Supplement
Signed October 16, 1993 as Ammended by Protocol Signed
November 18, 1987. (1989).
186. US EPA Federal Register 45, 5831 (1989).
187. C.A.F.M. Romijn, R. Luttik, D. van de Meent, W. Slooff and
J.H. Canton. Ecotoxicol. Environ. Safety. 26, 61-85 (1993).
188. C. Cox, A. Vaillancourt and A. F. Johnson. A Method for
Determining the Intake of Various Contaminants Through the
Consumption of Ontario Sport Fish. Environment Ontario.
(1989).
189. P.A. Cunningham, J.M. Me Carthy and D. Zeitlin. .Results of
the 1989 census of the state fish/shell fish consumption
advisory programs. Report to US EPA Office of Water
Regulations and Standards, Washington, DC. (1989)
190. M.L. Dour son and J.M. Clark. Fish consumption advisories:
Toward a unified, Scientifically credible approach. Regulat.
Toxicol. Pharm. 12, 161-178 (1990).
191. W.R. Swain. Effects of organochlorine chemicals on the
reproductive outcome of humans who consumed contaminated Great
Lakes fish: An epidemiologic consideration. J. Toxicol.
Environ. Health. 33, 587-639 (1991).
192. W.R. Swain. A review of research on the effects of human
C-279
-------
Giesy Briefing Document/ Minneapolis, MN, Sept. 14-15, 1993
exposure to orally ingested polychlorinated biphenyls:
Reference dose and exposure assessment. Abstracts of the
papers given at the Cause-Effects Linkages II Symposium.
September 27-28, 1991. Traverse City, Michigan. (1991) pp.
32-33.
193. J.L. Jacobson, S.W. Jacobson, H.E.B. Humphrey. Effects of
exposure to PCBs and related compounds on growth and activity
in children. Neurotox. Teratotol. 12, 319-326 (1990).
194. J.L. Jacobson, S.W. Jacobson, H.E.B. Humphrey. Effects of in
utero exposure to polychlorinated biphenyls and related
contaminants on cognitive functioning in young children. J.
Pediatrics. 113, 38-45 (1990).
195. J.L. Jacobson, J.L., S.W. Jacobson, H.E.B. Humphrey. Follow-
up on children from the Michigan fish-eaters cohort study:
Performance at age four. J. Gt. Lks. Res.
196. T. Colborn. Epidemiology of Great Lakes bald eagles. J.
Environ. Toxicol. Health. 33, 395-453 (1991).
196. T. Colborn. Nontraditional evaluation of risk from fish
contaminants. Report for the committee on Evaluation of
the Safety of Fishery Products, Food and Nutrition Board,
Institute of Medicine. National Academy of Science,
Washington, DC. (1991)
198. T. Colborn. Background paper-An overview of the toxic
substances and their effects in the Great Lakes basin
ecosystem. International Joint Commission, pp 1-24, (1989).
199. H.A. Tilson, J.L. Jacobson, and W.J. Rogan. Polychlorinated
biphenyls and the developing nervous system: cross-species
comparisons. Neurotoxicol. Teratol. 12, 239-248 (1990).
200. D.J. Hanson. Dioxin toxicity: New studies prompt debate,
regulatory action. Chem. Engen. News. August, 12, 7-14
(1991).
201. C. Gorman. The double take on Dioxin. Time. August, 26, 42
(1991).
202. Anon. Reassessment of Dioxin based on new science.
Chemecology. November: 1-2 (1991).
203. Anon. Dioxin re-examined. The Economist. March, (1991) pp 87.
204. Anon. Science Scope, 10-25-1991. (1991), p 507.
C-280
-------
Giesy Briefing Document, Minneapolis, MN, Sept. 14-15, 1993
205. R.E. Keenan, R.J. Wanning, A.H. Parsons and D.J. Paustenbach.
A Re-evaluation of the Tumor Histopathology of Kociba et
al.(1978) Using 1990 Criteria: Implications for the Risk
Assessment of 2,3,7,8-TCDD Using the Linearized Computer
Multistage Model. In: O. Hutzinger and H. Fiedler (Eds),
Organohalogen Compounds, Vol 1: Toxicology, Environment Food
Exposure-Risk. Ecoinforma Press, Bayreuth, Germany. (1991).
pp. 549-554.
205. Anon. Tier I Wildlife Criteria for the Great Lakes Water
Quality Initiative. Working Group for Wildlife, Wisconsin
Deparment of Natural Resources. Madison, WI, October 18,
1991.
206. M.W. Layard D.J. Paustenbach, R. J. Wenningand R.E. Keenan.
Risk Assessment of 2,3,7,8-TCDD Using a Biologically-Based
Cancer Model: A re-evaluation of thre Kociba et al. (1978)
Bioassay, In: O. Hutzinger and H. Fiedler (Eds),
Organohalogen Compounds, Vol 1: Toxicology, Environment Food
Exposure-Risk. Ecoinforma Press, Bayreuth, Germany. (1991) pp.
549-554.
207. P.H. Abelson. Excessive fear of PCBs. Sci. 261. (1991).
207. US EPA. Assessing human health risks from chemically
contaminated fish and shellfish. Report # EPA-503/8-89-002.
Office of Water Regulations and Standards (WH-552),
Washington, DC. (1989)
208. US EPA. Federal Register. January 30, (1991).
209. US EPA. Ambient water quality criteria for Polychlorinated
Biphenyls. Report No. EPA 440/5-5-80-068, US Environmental
Protection Agency, Office of Water Regulations and Standards,
Div. Washington, D.C. (1990).
210. T. Colborn. Epidemiology of Great Lakes Eagles. J. Toxicol.
Enviorn. Health. 33, 395-453 (1991).
211. T. Colborn, Bern, H.A, P. Blair, S. Brasseur, G.R. Cunha, W.
Davis, K.D. Dohler, G. Fox, M. Fry, E. Gray, R. Green, M.
Hines, T.J. Kubiak, J. McLachlan, J.P. Meyers, R.E. Peterson,
P.J.H. Reyners, A. Sota, G. van der Kraak, F. Vom Saal, and
P.Whitten. Chemically induced alterations in sexual
development: The wildlife/human connection. Elsever Applied
Science Publishers (UK) (1992).
211. B. Leece, M.A. Dinomme, R. Yowner, A. Li and J. Landers.
Nonadditive interactive effects of polychlorinated biphenyl
congeners in rats: Role of the 2.3.7.8-Tetrachlorodibenzo-p-
C-281
-------
Giesy Briefing Document, Minneapolis, MN, Sept. 14-15, 1993
dioxin receptor. Can. J. Physiol. Pharmacol. 65, 1908-1912
(1987) .
212. H.J. Harris. Persistent toxic substances and birds and
mammals in the Great Lakes. Chapter 29 In Toxic contaminants
and Ecosystem Health: a Great Lakes Focus._ Ed. Marlene S.
Evans John Wiley & Sons, New York. (1988) pp. 557-559.
213. T. Clark, K. Clark, S. Patterson, D. Mackay and R. Norstrom.
Wildlife monitoring, modeling and fugacity. Environ. Sci.
Technol. 22, 120-127 (1988).
214. J.F. Brown, R.W. Lawton, M.R. Ross and J. Feingold. In: 0.
Hutzinger and H. Fiedler (Eds) , Organohalogen Compounds, Vol
1: Toxicology, Environment Food Exposure-Risk. Ecoinforma
Press, Bayreuth, Germany, (1991), pp. 283-286..
215. P. Mineau, G.A. Fox, R.J. Norstrom, D.V. Weseloh, D.J. Hallet
and J.A. Ellenton. In: J.O. Nriagu and M.S. Simons (Eds.)
Toxic contaminants in the Great Lakes. John Wiley and Sons,
New York. (1989) pp. 426-452.
217. C.D. Wren. Cause-Effect Linkages Between Chemicals and
populations of Mink (Mustella vison) and Otter (Lutra
canadensis) in the Great Lakes Basin. J. Toxicol Environ.
Health. 33, 549-585 (1991).
218. C.A. Bishop, R.J. Brooks, J.H. Carey, P. Ng, R.J. Norstrom,
D.R.S. Lean. The Case for a Cause-Effect Linkage Between
Environmental Contamination and Development in Eggs of the
Common Snapping Turtle (Chelydra s. serpentina) from Ontario,
Canada. J". Toxicol. Environ. Health 33, 521-547 (1991).
218. D.J. Hoffman, B.A. Rattner and R.J. Hall. Wildlife
Toxicology. Environ. Sci. Technol. 24, 276-283 (1991).
219. L.P. Burkhard, D.E. Armstrong, and A.W. Andren. Partitioning
behavior of polychlorinated biphenyls. Chemosphere 14,
1703-1716 (1985).
C-282
-------
Table 1. Potential causes of deformities observed in embryos and
chicks of the Great Lakes region.
Nutritional deficienciea due to changes to prey base
New generation peatieidea
Continued effects of traditional contaminants
Disease, such as viral infections
Genetic inbreeding
Old, persistent Pesticides
Unidentified chemicals, pulp and paper industry
C-283
-------
Table 2. Effect* observed in birds of the Great Lakes region.
Eggshell thinning
Deformities
Tumors
Behavioral Changes
Immune suppression
Edema
Cardiovascular hemorrhage
Hormonal changes
Enzyme induction P4501A1 and P4502B1
Metabolic changes. Wasting syndrome
Depletion of vitamin A
Porphyria
C-284
-------
s
0
S1
I
5
in
o
(OS
0
a,
u
co
0
I
3 •''
1 1
I*)
o
o
o
o
1 1 10
I 185
| |
||
O CN
o
00
s
01
to
ro
00
§
PJ
•*
w 0i ro
«* 5f O
• • •
o in o
GO
in
i i
i i
s
8 §"§« 8
"o € O»N co •
9 (ajSCNCMCSCNtNCSCM
QCQ O«H«H»HtHfS»H«H
GB -H
E4Q OOOOOOO
•OCAJUOUUUOO
t4-HAOOOOOO ~
***! JS ^3 ^4 Vl l»l $«l Vl Jj
DQ O ri!iC0Ci4!'C'4i
C-285
-------
Table 4. Compounds, which may, based on
experimental evidence or structure
could be expected to have the
potential to cause adverse effects
through the Ah-r mediated mechanism
of action.
Polycyclic Aromatic Hydrocarbons
Polyehlorinated Biphenyls
Polychlorinated Dibenzo-p-Dioxins
Polyehlorinated Dibenzo Furana
Polychlorinated Napthalenes
Polyehlorinated Dipnenyltoluenes
Polychlorinated Diphenylethers
Polychlorinated Anisoles
Polychlorinated Phenoxy Anisoles
Polychlorinated Xanthenes
Polychlornated Xanthones
Polychlorinated Anthracenes
Polychlorinated Fluorenes
Polychlorinated DihydroAnthracenes
Polychlorinated DephenylMethanes
Polychlorinated Phenylxylylethanes
Polychlorinated Dibenzothiophenes
Polychlorinated Quaterphenyls
Polychlorinated Quaterphenyl Ethers
Polyehlorinated Biphenylenes
Polybrominated Diphenyl Ethers
Polychlorinated Azoanthracenes
C-286
-------
Table 5. Koch's Postulate*.
1) Observe Effect(s)'
2) Correlate to Cofactor*
3) Isolate Suspected Causative Agent*
4) Identify Suspected Causative Agent*
5) Introduct Suspected Causative Agent and
Elicit Effect
"indicates that this postulate has been completed.
G-287
-------
8.
§
so
I
H
1
I
toO
5 '
S)g
e ^
i*
O
BS
-M
"U
S
S «
10*
*
Location
-S
CO
t«
fa
p
g fL
fc
5* gi
*1
s
-
0>
••*
5?
**
f
0)
et
en
en
co
in
en
CO
§
9.
t-
H
O
f
"J
in
iJ
1
r-
in
en
rH
in
f
co
CO
f
a.
in
vo
al
11
i
CO
CO
•
en
(^
1
CM
^
CO
a
rH O
0) 0
is
§
CM
in
cs
rH
VO
CN
f
5
§
a.
VO
4
X
i
CM
in
CN
*
in
CM
£•
in
rH
^
•^
in
St. Martins
Lk. Huron
1
en
CM
P-
rH
,
5J-
?i
•
aquamenon IB.
Lk. Superior
&
s
S
CO
in
00
§
ih
co
f
°i
O
.. Winnepegosia
Manitoba
»J
i
..
to
-
t
CO
1
!
Mean potency
C-288
-------
I
f-4
|| Crossed BJ
^j
ft
1 Clubed Fo
a
0
•H
a
i
B
i
«M
b
1
0
§
H Asc!tes/Ed
5
|| Eye Deform!
a
*>
£
4
i
B
1
B
•*<
| GaatroBchi
C-289
-------
s
O O
>f O
rHtH
O O
cus
CH
0 a
0>H
HIM
4J M
8°
o w
M Pi
o -* m
6} O
A »X
Oi-l «J
•
n o 41
o A ts
> fto
Oft M
^ A O
CO
0
I
Q
O
c
H
IM
o
G3 ^^ *
gffi r^ ^)
OX >
WH °
41
O
At
Adverse Ef ft
Species /Location
0)
f^.
H
O
•
1?
egg lethali
w
0)
i
bald eagles
reat Lakes Territorj
o
ff>
vo
H
O
•
to
W
0)
•H
$6
bryonic defor
egg lethali
E
0)
herring gulls
Great Lakes
*wT*
CO
ca
•
(0
0
•H
4J
•H
•i-4 O
egg lethal:
embryonic def
.p
a"
bryonic defor
greater than
§
^rt
s*
(Q >
c-o
0) 3
i> Ij
U
-------
u
o
-H
I
O
IH
•H
H
•8
H
•s
(3
IH
0
o
a
I
0
n
i
o
1
I
.2 w
OI-P **
H • &
g-g ^
CJ S ^^
§ g
S-
Endpoint
t
a
o
o
o
*
V
o
o
0
es
Ot
1
-Field<174>
r^
H
§.
•H
I
o>
CN
O
o^
a1
»*+
1
o
b
1
o
u
IV4
1
Double-cree
0
o
o
^^
1
s
s
§
s
o
e
Vl
1
a
o
in
i
sa
-Field
e
1
3
1
u
m
in
1
A
3
-Field<168)
1
H
&
|
i
I
1
•d
•-t
fa
1
i
2?
«
Double-cre
to
o»
1
c
4*
3
s
a
o
5
!
i
^
0)
•P
.*
8
*O
S
|| Double-cre
P-
CM
g
fa
S
*+.
In
a
M
8
•a
$
|| Double-crei
o
rH
4J
H
1
sa
§
1
c
3
e
Jjj
•ft
o
i
in
u>
0
o
O
Itt
•0
M
at
1
•H
0)
a
Q)
H
u
c
^
o
o
f
'«
J
1
~4
g
JJ
•s
H
in
ft
S
1
r~4
1
2
0
o
•d
S
|| Double-cre
VO
n
4J
o
0)
1
of
•8
*M
_
M
2
s
hicken-Lab
u
e
Q
0
3
C-291
-------
I
41
I
O
4>
I
§
B<
It
O
I
fH
9
O
5
•a
0
8
o
*
£
§
u
5
A
•O
S;
S^
o-q
*H>-4
s
e
I
tH ^
3-sIi
tJl
£
..
9
8
fc"
n
iference dove
[i or LOAELxlO'1)
?CB/kg wet wt.)
Kl«
K *-
|J^
e
•rl
lj
•H
E^
o
*
rt
•0
o|
•«* ~"
fa
o
3
•QW
^ "^
81
A
o
en
o *^
*!
rt "
fa
tH
in
•-4
•H"
B
-------
•
\0
00
ffk
iH
a
•H
Lakes
4>
a
1
IH
O
M
M
0
41
a
9
A
ca
IH
O
H
a
.3
it
Id
M
it
1
O
•
v<
•H
0
H
•§
H
h
0
Str
*4
•gS
S -
•H
S^
•H fl
«H *~
B
culate
-»t_r
•S^.
If
/*»
•si
Is
§ ^
M H
l»
0 &
^^
Location
1
r«-
*o
m
<3
I
i
1
in
O
Lk. Superior
1
*ob
•
^
e
c
u>
•
o
1
<*
f-t
Lk. Michigan
|
"JH
en
«
9
O
1
*
O
1
t*
o
2
s
d
§
"ob
i*
m
+
t-H
1
•
O
1
r»
O
3
M
W
3
^J
•^
^41
^
*
iH
1
o
1
«o
o
Lk. Ontario
C-293
-------
m
1
•"8
• o
ss
5
A O
2 u
41
a °
$ >
o *
O H
s
u
N
ft
O
H
1
l-l
.d
u
a
•
M.
0
u
a
«!•
s
1
0
ft
s
0
M
3
U)
I-)
o
CM
H
rt
O
<•
PI
Superior
CO
t^
o
fH
,_{
CO
in
o
o
CO
«H
Michigan
r~
CN
e'-
en
o
IN
O
cn
«o
Huron
o
to
«H
CO
in
**
o
o
«*
rH
5
n
H
rH
U>
en
CO
U)
«h
o
o
«*
H
Ontario
h
&
m
s .
h a
0
«W -H
O -P
at
H h
(3 +>
0 «
•*4 0
•P U
8g
•P U
g.
O -H
a J3
0 4>
t|
° 0
H4>
•» _
•P O
O M
4> 2
•
§8.
B »?
<8 «
1) ..
• 8
•Q ^
8 5>
a fl
-H
• "P
0 •-<
5 9
H 1
« '
3 *
• •9
5 H
a *H
g«H
•o M
« 5
D-H
M
**i
~ -p
o B
tl . 1
A • b
Co o
0 g g
1j8 S
».
A en *o
•S « •
^* h H P -H
80 2 C M
> g 0 0
*« &-PU
"1 1^5
g- 3§"§
2 »> a « o
|0 fi°S
|H |, n H
2 H o H 0
jS- J»«
C-294
-------
I
I
8
u
•-I
i
3
0
I
.
n
«4
0
«-t
I
ledance
g
H
t»
to
O
0
U
a
0
*e
0
8
£
Location
o
-------
H
I
0
•H
«•
«
4»
0 a
•H H
H *•»
« tr
3 £)
M
o
Endpoint
o
to
0
4J
rt
Cn
ancer-Drinkin
o
to
c
(0
9
K
O
CO
0)
4J
(0
&
c
-H
ncer-Nondrink
0)
U
(0
c
g
"
a\
§
•H
4J
a
E
9
W
I
•H
0)
a
c
(0
§
K
O
rg
Human Health
in
O
H
X
in
i-t
0)
4J
(0
&
ealth-Drinkin
K
C
(0
B
X
*0
H
X
•
H
Wildlife
wo
H
X
0
«
H
Wildlife
H
Wildlife
C-296
-------
4* e
h p
sl
ii
•tt
hi
JSS
ttst
4»4>
I M M
:i:
O
4> O*M
!
1
>..••$
ill
m
Water Quality
Lterion
-------
Table 16. Water quality criteria and fish quality proposed
by Cook et al.c to protect humans and
wildlife from adverse effects of TCDD-EQ in the
Great Lakes.
Risk
High*
Low
Fish
(pg TCDD/g, ww)
6
60
Water
(pg TCDD/1)
POC=0.2 mg/1
0.07
0.7
POC=1.0 mg/1
0.35
3.d
1POC= organic carbon content of water
Table 17. Dietary KOAEL, based on TCDD-EQ in fish fed to
white, leghorn chickens".
Dietary NOAEL
-
0.6
Concentrations of
TCDD-EQ in Forage Fish
10
Exceedance
16.7
C-298
-------
o
o
6
o
o
1
S.I
SLJ ^
« M g
4* O **
1 1
Endpoint
o
00
0
X'
0
8
rt
!£
C
n Health-Cancer :Drihki
g
§
S3
o
CO
1
0
X
o
8
4J
(d
»
O1
C
•H
u
Health-Noncancer : Dr in!
c
(0
5
33
o
CO
CM
1
O
X
o
8
4J
efl
£
tr
c
•H
u
Health-Cancer : Nondr in}
g
s
5
K
o
CO
1
0
X
o
water
Ci
c
•*
ealth-Noncancer : Nondr i
K
C
(0
E
35
vo
00
CM
1
o
X
n
Human Health
vo
CO
CM
I
O
X
Human Health
o
CO
1
0
X
in
CO
Wildlife
C-299
-------
0
4»
O
M
«
Ci
Jtt
«O •
94*
4» a
ti 0
•H m
o a
o a
D 0
« a
a a
41 10
M a
0-H
ra
I
$
a
*fi
«t
4>
Li
0
I
"M
O
o
Cn
B
S
Factor
X
o
H
in
0)
•H
O
0)
fi
(0
o1
IH
1 NOAEL
X
o
H
£
wl
s Sensitiv:
0)
•H
O
0)
ft
CO
X
10
id Content
5
»a
X
OJ
rH
2
3
o
-H
s
1
X
o
0
H
O
H
§
•H
4J
O
1
fa
Q)
H
•S
(0
rH
•H
I
X
rt
fa
g
x
o
0
H
1
0
rH
fa
£
C-300
-------
Giesy Briefing Document, Minneapolis, MN Sept 14-15, 1993
Figure 1. 2,3,7,8-TCDD Avian Toxicological Effects Thresholds. The hatched and
solid bars represent the relative range of concentrations (pg/g or ng/Kg)
of TCDD-EQ and 2,3,7,8-TCDD, respectively, that are currently found in
the eggs of colonial fish-eating water birds of the Great Lakes. The
arrows indicate various toxicity thresholds for avian species which have
been tested in laboratory or field studies. The references are given by
number.
C-301
-------
2,3,7,8-TCDI) Avian Toxicological Effects Thresholds
10,000
bo
a
W
Q
Q
Q
Q
U
fr
00
*t
t^
to"
1,000
100
L: I TCUD-EQ
r\ 2,3,7,8-TCDD
2200: LD50 Plicnsnnl Embryo (26)
1000: Mil00 tt'l-C Eniliryn (27)
750: LD50 CIT Embryo (28)
460: LD50 DCC Embryo (29)
312: EUSO EROD Pheasant Embryo (26)
147: LD50 VVLC Embryo (30)
115: LD50 WLC Embryo (31)
20: LOAEL Rcprod. WD (32)
10-20: LOAEL Tcrata. VVLC Embryo (30)
10: LOAEL Tcrata. VVLC Embryo (33)
10: LOAEL AIIH VVLC Embryo (34)
Figure 1
C-302
-------
Giesy Briefing Document, Minneapolis, MN Sept 14-15, 1993
Figure 2. Concentrations of calculated TCDD-EQ as a function of H4IIE bioassay-
derived TCDD-EQ. TCDD-EQ were calculated using TEF values using
either Tillitt(108) (Squares) or Tillitt(108) and Safe(48) TEF values (stars).
Bioassay TCDD-EQ values were determined using the H4IIE bioassay as
described in the text. The dashed line indicates equality between
calculated and bioassay equivalents. The upper darker solid line is the
linear regression for TCDD-EQ derived from Tillit-Safe TEFs while the
lower solid line represents the linear regression for the TCDD-EQ derived
with Tillitt TEFs (Reprinted from Jones et a/.(102> with permission).
C-303
-------
en
eft
a
O
LI
6
a
z
•§
3
100 20O 300 400
H4IIE Bioassay TCDD-EQ (pg/g)
500
Figure 2
C-304
-------
Giesy Briefing Document, Minneapolis, MN Sept 14-15, 1993
Figure 3. Relative proportions of total TCDD-EQ predicted from instrumental analyses
and TEF (EROD induction) values, which are contributed by the pPCB
and 2,3,7,8-TCDD. Samples are of 23-day old Caspian tern eggs. TEFs
used were those of Safe(48).
C-305
-------
TCDD-EQ
23-Day Caspian Tern Egg:;-1988
Location
Beaver Is. L.M.
CDF-Saginaw L.H.
Gull Is. Green Bay
20 40 60 8U 100
% TCDD-EQ as 2.3,7.8,-TCDD
3,4,5.3'.4'-PcCB IIE.1 3.4,3'. 4
',] 2,3.7.8-TCDD
120
Additive Potency trom GC-MS Analyses
Based on EROD Relative Potency
Figure 3
C-306
-------
Giesy Briefing Document, Minneapolis, MN Sept 14-15, 1993
Figure 4. Relative contribution of specific p-PCH congeners to total concentrations
of TCDD-EQ. TCDD-EQs were calculated, using Safe(48) TEFs, for
individual p-PCH congeners. The contribution of each congener to the
total concentration of TCDD-EQ is expressed as a percentage of total
calculated TCDD-EQ concentration. Due to the small contributions of
equivalents by PCDD and PCDF congeners the TCDD-EQ contributions
of all dioxin and furan congeners were combined (D&F). Values
represent species means. Abbreviations are; F. Tern; Forster's tern; C.
Tern, common tern; R.W.B.Bird, red winged black bird; T. Swallow, tree
swallow. The segments for each bar graph are in the same order from
left to right as shown in the legend (Reprinted from Jones ef a/.(102) with
permission).
C-307
-------
F. Tern
C. Tern
R.W.B. Bird
Tree Swallow
0
20 40 60 00 100
% Contribution of Each Cunijener
Figure 4
C-308
-------
Giesy Briefing Document, Minneapolis, MN Sept 14-1§, 1903
Figure 5. Relative potencies of technical Aroclor* mixtures and extracts of double-
crested cormorant eggs. The relative potency is the ratio of the
concentration of bioassay-derivedTCDD-EQ (ng/kg) to total concentratio
of PCBs (mg/kg). The resulting ratio has units of mg/kg (ppm) (Reprinted
from Tillitt et a/.,(22).
C-309
-------
CORMORANT EGGS
AROCLOR 1242
AROCLOR1248
AROCLOR1254
AROCLOR1260
O 5 10 15 20 25 30 35
TCDD-EQ/PCB RATIO (pg/ug)
Figure 5
C-310
-------
Gfesy Briefing Document, Minneapolis, MN Sept 14-15, 1993
Figure 6. Relative potencies and total concentrations of PCBs in Caspian tern eggs
from three locations on the Great Lakes.
-------
RELATIONSHIP BETWEEN TCDD-EQ AND
TOTAL PCBS IN CASPIAN TERN L'GGS-1988
Beaver Is. L,M.
Gravely Is. G.B.
CDF-Saginaw L.M.
400 300 200 100 0 2 4 6 8 10 12 14
H'J TOTAL PCS (ppm) IID RATIO (ppm)
TCDD-EQ calculated Irom GC-MS and EROD
Induction potencies. Additive model.
Figure 6
C-312
-------
Giesy Briefing Document, Minneapolis, MN Sept 14-15, 1993
Figure 7. Exceedance values for the effects of total PCBs as a function of the
exceedance values in bioassay-derived TCDD-EQ for double-crested
cormorant eggs from seven locations in 1988 (reprinted from Ludwig et
a/.(67)
C-313
-------
PCB-Exc Vs. TCDD-EQ-Exc
Double-crested cormorants
25
PCB Exceedance
PCB-Exc = 0.186x(TCDD-EQ- Exc) + 4.65
20
40 60
TCDD-EQ Exceedance
80
100
Figure 7
C-314
-------
Giesy Briefing Document, Minneapolis, MN Sept 14-15, 1993
Figure 8. Deformed bill of double-crested cormorant from Green Bay Wisconsin (Photo
by J.P. Giesy, 1991).
C-315
-------
jr M-
Figure 8
C-316
-------
Giesy Briefing Document, Minneapolis, MN Sept 14-15, 1993
Figure 9. Egg mortality of double-crested cormorants from 12 locations on and one
location off of the Great Lakes as a function of TCDD-EQ (Reprinted with
permission from Tillitt efa/.{22)
C-317
-------
60
50
40
O 30
O
O 20
HI
10
DOUBLE-CRESTED CORMORANT
INDIVIDUAL COLONIES / 1986, 1987, 1 'J80
- Y= 0.067(X) +13.1 (I2 = O.703, p = O.OOO3)
100 200 30O
DIOXIN EQUIVALENTS (psj/g)
Figure 9
C-318
-------
Giesy Briefing Document, Minneapolis, MN Sept 14-l5f 1993
Figure 10. Occurances (per 1,000 chicks examined) of different types of deformities
in the embryos and chicks of double-crested cormorants in Green Bay,
Lake Michigan.
Cr319
-------
DEFORMITIES IN CORMORANTS
GREEN BAY-1986-1989
Rate/1,000 (4891 Examined)
Diseased Eyes
Ascites-Edema
Deformity Type
Crossbill
Gastroschisis
II Appendages
I I Oelormed Eyes
Figure 10
C-320
-------
Giesy Briefing Document, Minneapolis, MN Sept 14-15, 1993
Figure 11. Rates of deformities in double-crested cormorants and Caspian terns from
five regions of the Great Lakes. The number of embryos and chicks
examined at each location is given above each bar.
C-321
-------
RATES OF DEFORMITIES
CORMORANTS 1986-1989
Abnormality Rate/1,000 Hatched Chicks
4155
2656
."•90
3222
2935
Cormorants
r.,.-.J ...... . I'
Caspian terns
Region
Green Bay
L. Superior
N.L.Michigan LI) N.L.Huron
CDF-Saginaw L.H.
Number oi Chicks Examined Above Bar
Figure 11
C-322
-------
Giesy Briefing Document, Minneapolis, MN Sept 14-15, 1993
Figure 12. Rates of deformities of double-crested cormorants and Caspian terns in the
Great Lakes, as a function of TCDD-EQ.
C-323
-------
Deformities Vs. Concentrations
Cormorants and Caspian Terns-1988
Deformities per 1000 hatched Chicks
Caspian lerns
200 400 600 800 1000
Concentration (ng TCDD/kg, wet weight)
1200
Figure 12
C-324
-------
Giesy Briefing Document, Minneapolis, MN Sept 14-15, 1993
Figure 13. Concentrations of TCDD-EQ in the eggs of double-crested cormorants from
eight locations in the Great Lakes.
C-325
-------
TCDD-EQ in Cormorants
Means of 1986, 1988, 1989 by Location
Location
N. Green Bay
Beaver Is.
St. Martin's
Thunder Bay
N.L. Huron
Georgian Bay
Tahquamenon Is.
Apostle Is.
100 200 300 400
Mean Concentration (ng/ky, wet weight)
Mean Across Yoar
500
Figure 13
C-326
-------
Giesy Briefing Document, Minneapolis, MN Sept 14-15, 1993
Appendix I
Briefing Documents on the Effects of TCDD on Rainbow Trout
C-327
-------
CO v
UJ g
OBJECTIV
To Determ
Q
Q
O
>T
00
^
-
CO
•V
CM
"5
Q
c
£
1
O
O
*""
tion in Tissues
^i^^~^
o
o
u*
CO
Q
oi
JO
"5
•o
I
o
"5
>""™
CO
co"
CO
"5
^ji
-------
O)
c
(/> {
Q.
••
s
Q co
UJ £
ISS i
PA
^JH Vr
••• •«•
C
CJ
O)
O
CO
CO
12
N^~
E
CO
horhyncu
o
c
O
3
O
0
a:
Females
•
CO
CO
CO
o
0
g>
i
D)
O
LO
CO
CO
CO
Q
o
LO
CO
h. -
•o
0
CO
o
CL
X
LU
CO
LZ
LO
CO
• •
£
*co
HM
£
0 CO
E >*
£ CO
CO Q
o
i» O
o
£ CM
0
CO ~
o o
LO
o -
45 0
o
f «v
Am T^
.2
*" 0
^ in
O)
c
"E
CL
CO
VV
CO
o
.c
"co
0
•o
c
o
QL
•u
"o
"o
O
ti\
VI
iZ
CO
o
CO
E
"o
c
o
o
'o
••
^mm
I
CD
composit
'1
"^3
fl)
-C:
CO
CL
CO
CO
0
UJ
CL
E
'i
(0
1
liS
o
°-
CO
CO
00
CM
nz
"c
3
J)
13
JQ
C
CO
CD
LJJ
CO
o
CD
C-32?
^w
0)
5
-------
CO
a
o
X
LJJ
5
mSm
(0
c
Cr
C
O
o
a>
3
o
Q.
X
UJ
OT
0)
O
>
O
O
o
0 >=
-jg -
c c — J
"~ _CO U>
"® •£> ?I
X g>
G) 2 o
5> £ "S
^ . a
j- 5 5
CO O O
IE UL UL
0 *
CO
5 O
o •
^ Li- Q
co *s Q
^ ± CM
^ oi »
.. CM 2
2 --5
0 D) 2
> C o
jrfPVk m^^m ^^
2 =5 a-
£ co E
3 O O
• » •
r.^^0
0)
3
o
_
o
"co
o
a>
0
•D
o
o>
CO
6
a>
Hi
0
0
o
DC
O
"co
JU
CO
Q.
0)
CO
2
0
BCO
"o
c
o
O
0
CO
o>
a>
X
>
CO
0)
5
-------
• mmm^m
X
O
Ti
Q.
6
N
C
Q)
.Q
LU =5
CC O
13 o
C/)Z
o g
X
LU
00
CO
n
CM
CO
•o
o
N
"35
o
.c
**
c
>»
CO
Q
Q
O
.Q
CO
(5
ty checked
ct
.U
^ °
2 3
CLn
w "•
^
08 0s*
O, A
CO
o
Q.
X
(0
(0
O.UJ
2
•o §
•2-1
UJ
^§
> "4=
"£* CO
O) ~
"o
(0
o
6
+-, O
o
2
o
Q.
o
Q
JC
Q
Q
O
o .
o
Q.
0)
O
*"* A
.£ £
•o
o
CD
^F W^
(0
Q O
R 0-
O x
I- LLJ
-o
0)
10
3
(0
O
.12 x
U. LU
CB
•o
0
LL.
(0
\L
CO
o>
o>
(0
0)
5
C-331
-------
C0
Q
O &
X.2
h-Q
LU
*5S •
O
0?
Q.
00
g
-5?
CD
u>
(3
O
Q.
O
o>
±*
O
X
O)
cr
CD
u>
o"
o
o
Q.
o
. .
o
DC
z
o
o
"?
CD
O
o
o
c?
Q.
00
u m
3
a
UJ
2
^2
O)
o-
ffl
S
?
>**
(0
-------
>•
o
o &
O 2>
\- V
a>tt
^^^ ^^
i
0) (0
o .2 £•
o i ®
o. £ E
o S-2 E
S m 0> >«
S Q » -
IJ feBdi ^|* "*™ ^
w "o 5 e- o o
o o S -E > c
ju co -5 u. iu co
o
73
0)
"S
^i Q
S Q
i? O
>» O Q.
U) ^ (0
2. LU i
O CO
to ° °
-2 z Q
X
CO
a>
o>
•^
>
CO
o
•••§
O
-------
(0
ti
O
U.
O
o0 Sag -£
•§ o ^ £- "Sto | ^
8 = C T3^ OOl? 7^
2,*® CD® SIS ^
oSP a" •- S m Q-
iM ii yi *>•*
°gl ^ ^o* ^|
P-cwco-S OF^ w.,..
O Q U. ••- o »- Q
•S1"? £ ~-o -
»•• ^ ^-§* ** 4^ «•
S S o co
jg co c "o — z
CO O C _- CO
(D *^> O ill ^^ i CD
«• ^ir r^ „ ^^ ^^J ^ih ^^
•• i— H"* ^S CD »•*•% &_
~rr- ^IB .^^ ff\ ^r f * " mf*.
O /it 5= w m '— .0 /»t z: S2
QW /•! •-• ^^ CD ^ ^e% t*t -•• \£.
* *• ^^ €8 r-Q^Vl^CO 2
O^cD "SQ^Ss *"
S IO •—
-I i- O
CM co ^r o
C-334
-------
z
o
H
o
••/
o
o
DC
CL
UJ
DC
T3
amine
X
HI
w
o
4-p
o
E
5
5
Q.
•u
0)
o
3
•a
o
a!
(0
o>
o>
LU
H—
o
(0
I_
o
jQ
E
3
Z
•u
0
o
3
o
* §>
tn O>
s> ^
O) H-
UJ O
*o ^*
"3
Q) c
±! o
CO Q
(0
0)
LU
o
*«*
c
0
HM
C
o
O
o
•^^
&»
o
a
O
ri_'2'a
(0
LU
«t-
O
*rf
c
CD
*•»
C
o
O
p
!s-
•
<
>»
•1M
•P^^
HM
9L.
0)
u.
o
"S
•
"55
o
CL
CO
CM
>
4-»
*- 5
"E
o
O)
o
*-
CO
I
LL.
5
"5
0>
O>
(0
O
-------
o
u>
CL O
oc
OC "J
UJ ^
CO
CO
•o
o
10
o
§
c
mo
o
3
C
O
•o
c
0>
CL
O
Q
O
CO
o
Q
o
CO
CO 4-i
o
^ »-
0) CL
>
o
S
T3 ~—
= E
0
CO CM
T- CM
C
o
o
12
E 4JT
•H O
^ •«•
CO Q
CO
O
o
O 00
CM T-
CO
CO
O
J 2
c o
•^H V>/
5 ?
CO
CO
o
O
o
ities
Ac
~ X
C CM
O
= o
•u
c
o .2
c o
CO
O to
°
cc
2
O
CM
o Q-
o cq
E ^
* CO
LO
C-336
o
-------
_
1
(0
cc
o
h-
O
CM
z ^
o «
H^^^^^JB
djT
^^^ 1 1
O ^
LL. CD
O
.
(5
O
^
Muscle
CD CM
0 0
i CD
1 T—
! 0
CM CM
CM OJ
O O
^ CO
"•" d
«J' CO
CM
in
d
CO
T—
6
o>
CO
o
0
q
o
^-
o
rV
o
CO
0
o>
G>
6
c
CO
o
p
T~
o
^
o
0*
> - * •,.
••• • C)' ' • '
. ••' • r" ^
• , .-
o
Q
CO
C-337
-------
(0
DC
O
o
< o
LL O
CM
z
o
LL CO
o
o
m
G>
"co
E
"x
o
Q.
Q.
CO
CO
O
CO
CD
CO
O
O
LL.
UL
m
O
CO
O
Q.
o
u.
CO
h-
sO
0s*
CM
CO
O
•••
O
CO
CD
CO
T- CM
J£
H— O
o co
3
§«
0>®
D>.>
LU LL.
O $
O
LJ.
U.
CO
C-338
CO
10
CM
CO
E
"x
o
0.
0.
o
.E co
o
LL. D.
CD LL CD <
CO
o>
o>
CO
0)
o
-------
<0
§ >
*s ®
(0 > P
*='••= o
c P 2
co g< r
h §^ 1
a °* . I
"- < o S
U. *- c 2
LU w-o K
g"5 §
O -o o g
T • s* o>
0)
CD
111
CQ » * o
s §
aw frt
•a „ — _®
O
- w O •
o: ?
» S L. ' 5 S
iZ h- o IE o g
'a
C-339
-------
CO
z
o
H
DC
1-
LL1
O
•y
EFFECTS COI
- Lethality - 1
^ %
• Hi
i •*•
MMHHl ^^^»
QC
111
CO
CD
O
1-
co
HI
pSs*
o
« J
0)
5
>
0
CO
o
O
"U
o
£
U)
G
G
O
U>
fi.
CO
i-
i
.2
Dietary:
Concentra
§
12
•\
U)
•***.
•u
G
Q
0
h-
0)
0.
Csl
O
O
is
Q
"5
DC
%
a i
S
* U)
^ Q
s s
e 2
2 3
O O
CM 0
^^
1 o> . i
BCO CO J£ £
"S ^ O O
CB .&• CO g
u> =5 =3
o ** *
*
rv*4n
•g
Q.
—
•*
U)
0
O
o
H
U)
a
o
i-
"O
[a
|
3
•s
U)
D
G
O
U)
a,
CO
0
0
CO
o
a>
>
-------
z
o
H
S
DC
O
LL
LU
DC
6
DC
LU
Ij
O
1-
z
o
en
DC
|
O
O
oo
00
O)
•
a
o
c
U)
0)
jQ
JM
"a.
CO
00
in
CO
CO
CO
CO
CO
CM
*5
"75
*v
O
X
c
o
c
LU
•
"o
o
"x
,o
H
•
/j\
V?
"5
3
O
jl
CO
c
.2
ts
.SL
a.
C?i
eft
a.
o
o
0
cT
^
II
o
in
a
^
01
a.
o
o
II
0)
o
0)
c
CO
JC
o
15
BO
U) O
--- o
Q. p
O fli
O x
°- i
m c
ff»
II o
o E
CM c
Q 3
mml CO
C-341
O)
^^^^^
U)
a.
00
•
T.
II
•u
D)
O
0)
U.
.E
LOAEL
0)
a>
00
II
CO
CO
a
•••1
•o
6
0
CM
*2
iff
LU
15
"5
0
"5
o
X
E
Wl
Ui
c
o
c
03
"E
01
\fl
CO
O
u.
8
a>
a>
(5
-------
Ill
o
z
08
*^F
LU
o
00
111 °°
a
•— i
D*
*
H TB
< *
EC ®
ill o
I— H
^^M f*
mmm j^
—i o
os
H
Z
O
(A
DC
<
a.
O
O
CM
(O
I
h*
^t
• •
h*
£
0)
•c
o
^•^
^<
o
.0
X
1
•
c
•SJ|
^
C
Ul
s
ts
^
•o
CD
(0
O
f\
EL
X
LU
>*
LL
Q.
3
E
*
CO
O)
U)
Q_
•••
in
CD
K.
•
II
LO
<5f
Q
>*
(0
T3
00
CM
LU
g
O)
"***-
O)
Q.
LO
CM
II
>*
CO
•D
00
CM
LU
<
O
Z
0>
(0
w^
.a
O
C
5
^
CD
£
>*
•D
2
GO
0)
BC
^
O
•2
CD
_C
C-??45
CO
D)
D)
LU
E
o
IL,
C
§ 0)
o --,
O g
Ui
.E o
a>
CO
0 "°
| c
LU "5
*- C
c o
a 0
o
•— f
C °
••• ••••
a>jc
CO ^
^^ ^ p^^^
CO
o>
o>
(0
o
5
-------
UJ
^
O
z
08
UJ
<
I
.ITERATURE L
»JI
o
r™
O
(/)
oc
2
O
o
^M
0)
r & Peterson, 19
0)
^^
|2
5>
00
CO
CM
d)
CM
• •
CM
_ »
"o
O
1
o
_?*
10
3
*T
^
•3
O
^
^fe
i-Studies: Rainfc
k.
^^
"o
€11
w
,E*
o
o
S"""
U)
UJ
0)
0)
w
CD
CD
LU
c
0)
=3
•
E
CO
"i5
*£,
'E
•^
w
5
O
O *-•
W S
• <0
.
s
"S5
*-
2
CO
c
0
U)
.E
c
Q.
01
O
BX
00
00
^5r
O
**
O
CO
CM
II
O
LO
0
C-lt
•u
0)
r*
Cl
(0
JQ
O
£
o>
^
CO
**
o
CO
t3
O
0)
"O
Q>
Which Contain
W
en
>
TJ
3
CO
U)
.E
^
o
o
u.
0)
•>
' ^
CO
0)
<5
-------
wmam
CO
o
CO
D
•
•••1
O
IMI
O
o
c
0
Id
cs
&-
_J|_J|
•#•*
<_LJ £
"> °
1°
CO £*
r- <*
C •*-»
0 £
•u*
> *
w
J2 •
|j|
"£ 5 5*
C8
« *tj •
*
M ®
s> «
1LJ 0)
cvi
JQ
•V
O)
•o
O
O
o
O)
a.
CM
o
o
II
o
o
*+~*
_J
UJ
<
2
co"
o
« cc
2 uj
Hh*' u^
C ^
a> u
o
c
o
o
CO
ts
_o
U3
o
3
•o
£
.2
HH»
o c
o —
<
O
3
« S3
® o
DC <
o
o
PL
h-
(1)
C
CL
o >>
in
o
•*>->
(9
g
a.
a.
0) >»
o ?=
LU
c
-i
UJ ^
2
-------
z
o
DC
LU
•
r"
DC 1-
03
o
w
t^
— i o
< -a
D £
OB
DC
OC c
in •—
•••
<^HH
LU
§: <
9
LL.
og
ULATION
Based
O
<
O
CO
"co
CO
m
T3
]EL
N**^
<0
CD
O>
•
II
LL.
^
CD
*
• •
CO
c
Assumptio
^r
•
o
II
i •
c
.2
o
2
ul
Available
xO
0^
CO
II
fn
tit
*fw
*o
i^ S
vt
0 ^
t- O
11 °
o S
£ £
^^
O"
Q.
Q.
h-
CM
O
•
O
— !
O)
Q.
CNI
O
•^
f^1
C4
II
O
o
C-345
C
"5
2
**""
•o
o
"o
m j^
CO
BCO
^
^
o
^
* Assumes
CO
o>
o>
•\
•
"ca
i^»
0)
^
o
o
O
*
€0
O
O
(0
O
5
-------
Giesy Briefing Document, Minneapolis, MN Sept 14-15, 1993
Appendix II
Briefing Materials on Effects of TCDD-EQ on Mink
C-346
-------
NJ
^ P
ss
° «<
o
fa a
go
^
KO
gu
O
H H
^C J
"J X
W B
•es xs
w2
gp
O 5«
&- CQ
58 u
« cu
0^
?/>
H
S5
t=;
i
P
o*
a
g
H
o
hH
P
ON
CA
5
C-347
-------
•s
OX)
ft
1
j3
'o
o
S
a
C5
c» .
» ^^^
fa
ffi
S
OS
=
OD
s
fl
cs
ON
i-H
»\
.2
s
C-348
O
-------
en
Assess
£
O
U
fl
o
o B
2 "O
S «
O -4"*
£ C3
68 S
p* ^
£ WD
fl O
^ *3
Q C5
§^|
U ^
^ £
fl C3
^ "S
SM ft
S 8 ff
« « -fl
cn IS tt
2 § 2
ft
S
o
a _
.9 o
__* P*^
73 o
O -73
-^ 0
O) Q.
•PN 2k
K £
M .u
O W
•c
s 1
b &
O)
B-
e
i
jp^j M
O o
Lz3 t2
Qi
a
U £
H ^
00 fl
o
.w
•g-fta
0 w
%
fi ^
I 2
•42 S
C8. >
^ "O
"£<
S g>
fl £
°s
^ 2
DO £
a> ^o
.s^
S
ON
ON
O
C-349
-------
o
fc
tft
A
en
fi
O
»PN
15
••*
a
o
A
c»
t/3
CM
O
g01
8 *2f
s
%t
O
^ 0>
O wa
O
O
• PN
^-A
A
9
V5
e
o
CJ
(O
^M
C5
<4^
fl
a js ^
fi
O
«
IS
S
s
o
£
Q
^
!••' ""in
a
v
a\
&L
O
Q
^
« Jg c» p^l
a b H ""
C-350
»> -^
ss
W
WFI*^
O
-------
tt
u
r i
O
PQ
a-.2
ed
ii
a
! =
v Q
OH
fl S
^^3 *^^^
w ^^^^
"Si 2
1
0
8 S
O
^ ?3
c ~
•s§ s
to M O
O Q
fi
q^ C WD
ej o e
.5 -J2 o
a ^"%
S 2
us °
Q
U .S
OH "
OJ
O
^
C3
53
Q
H
^
o
4>
« ^
o ;s
CD ^3
a
= b i
o tsj a)
H Q
es
G
cd
N
a
CJ
E
"3
c-
S
ON
iH
.2
I
cn
O
C-351
-------
o
a
^ S
o M
fi C
-
TO
o
c«
.o
a
2 -
5e
-
•K
>•
8 "d
w fl
M
•^ M
s fe 8
M TO >-
** S3 *H
S p2 J
a ATS
8 fl
3g.
0>
o>
Q
fi
•PM
43
1
!M
O
fi
C3
Surviva
i
CM
O
X
O>
S
o
u
-^^
3
o
0->
>
•PM
•*^
U
0
"S
o
Ul
a
^
p<
fl
s
^^^ -^^r
ft ft
•^^ ^_j ^^
p»«5 **""! O
«-S o
o ^
<»
=
o
»
^^ ^^» ^^
TO eg ^3
HHHH ^^^^ ^^^^k
^^^^^ ^l \ ^P^^^^
__gi ^^v
W M
P^^ M
ON
«
C^
en
C-352
O
-------
TOTAL PCBs
SAGINAW BAY FISHES
FISH
Total PCBs
(mg/kg)
Walleye 1
Walleye 2
Walleye 3
Walleye 4
Walleye 5
Walleye 6
Alewife
Shad
Carp
Mean-WC
Mean-WOC
0.97
0.87
1.8
2.8
1.2
2.9
1.4
0.5
7.2
2.2
1.5
Giesy & Kubiak, 1993
C-353
-------
C/5
o
<
GO
ce
i~- i
u
,£-
C/2
i o
s
i
S
'68
••-<
-5
C-354
en
o
-------
CALCULATED TCDD-EQs
IN SAGINAW BAY FISHES
FISH
Walleye 1
Walleye 2
Walleye 3
Walleye 4
Walleye 5
Walleye 6
Alewife
Shad
Perch
Carp
Mean-WC
Mean-WOC
Total PCBs
( mg / kg )
1.0
0.9
1 , 8
2.8
1.2
2.9
1.4
0.5
ND
7.2
2.2
1.5
TEQ-A
( ng / kg )
4
13
14
87
14
11
31
11
4
194
38.3
21.0
TEQ-C
( ng / kg )
9
13
16
34
13
41
9
11
11
50
21.0.
17.3
TEQ-C Calculated from H4IIE TEFs & Additive Model
TEQ-A Determined by H4IIE Assay
Giesy & Kubiak, 1993
C-355
-------
CALCULA'J
Kll T(
IN SAGINAW BAY
V
FISH
Walleye 1
Walleye 2
Walleye 3
Walleye 4
Walleye 5
Walleye 6
Alewife
Shad
Perch
Carp
Mean-WC
Mean-WOC
TEQ-A
( ng / kg )
4
13
14
87
14
11
31
11
4
194
38.3
21.0
JDD-EQf
FISHES
TEQ-G
( ng / kg )
9
13
16
34
13
41
9
11
11
50
21.0
17.3
TEQ-C Calculated from H4IIE TEFs & Additive Model
TEQ-A Determined by H4IIE Assay
Giesy & Kubiak, 1993
C-356
-------
QS
PQ
op
^
•W
Qi
Para
so
o -a a
H
W
a
a\
s a
p
^
C3
C-357
^gj k§
3 *
^ >%
—N ?«
«8 .a
•• o
-------
H
C/5
a!
$
I hH
Q
u
H
OS
so
&
cd
a
Q
cc
eg
a;
fi
Q
O
C-358
-------
H
.9
«
03 W)
OJD
ce
WD
O>
H
PD
s
0>
^
^
S
O
c»
ON
00
ithou
C5
J=
C3
C3
CQ
C-359
3=
S
o
Q
P
C
CS
ON
H -2
8?
O
-------
C£
fa
PQ
CU
C3
fl
O
W
fl
O
u
II
fa
Q Q
3 Q Q
o U U
H H
C/2
c:
04D
ns
OJ
3
3
o
,^^^
C3
U
O
^^^^
^
."3
u
"c3
O
fa
HH
TT
B
GJ
s
a
^^^
o
0^
Cfl
ON
CC
•P«
^Q
2
C-360
O
-------
O
r>O
^ H
X
co-
O ..
I:
«
^
^5^
O
. ta
PQ
a;
JM
3
CO
e*
o>
§
fO
^^^»
^^^i
13
^
a
•
o
Q
H
v>
V)
1
a
O1
i
Q
Q
H
^T^
Z^^
^2^J
cu
"cS
1
C-361
cs
•p*
1
B?
-------
Q s
£
Hg
0 §
TO
PQ &
a
es
s
en
QO
C-362
-------
&o
H
U
w
fa
t£
«
tu
£3
as
ss
o |
CM 3
c» ^
S
o
a
P
CO
CO
s-l
O
0
£
OX)
*^»
OX)
5
fl
s
>-^
OX)
s
•N
OX)
OX)
^
-^.
OX)
00
fi
o
CO
o
Q
C3
Q
a>
CO
O
Q
o
CO
O
C-363
O
-------
CO
H
fe
fa
CO
PQ
O
H I
co "«
CB
at
I*
O
op
P-3T
CD
S3
a>
O
w
a
o
U
— *~^^
=
.
OX
s
o
O
•\
OX)
OX)
c
o
0
0
Q
m
ON
rr
OX)
OX)
=
ox
fl
0*>
o
o
o
^
0)
5?
O
CA
r*
c
>
"c
pfi
c<
a
•pi
C
C-364
-------
QC
H
U
W5
2
. . W-g
O o2
O
J
H
h
o
Cs
e*
0£
c
fi
O
c
S
6£
8
O
Q
es
O
•X
0£
O
Q
Of)
DID
^
O
Q
»v
00
DD
S
o
•Q
OS
C3
C-365
O
-------
1/3
H
U
.«
u
ss
B <
§°
GO
M
O
OJ
u
H
bJD
fi
-<
H
6JD
W)
wa
09
0
£
ON
0\
SO
EJ
ON
6JD
6JD
O
S
C
0
ee
-o
8
*S
WD
C
O
Q
es
I
(91)
03
O
Q
cs
Q
i
c«
O
M)
0)
0>
^
fa c
- r
s
HH Q
1
s
O "
o*
a
H
C-366
-------
fe
o
ATION
h-4
D
U
J
d
V
C/5
o
1
Q
^^^^^
S
tt
,/
fe
ag
fl
O
ncentrati
r:
• •
Sx
, «2
^J
W T3
^j -* *
^3 £
O *
z 1
*3
v
^
•^^ P"l
5^.
11 ^
•^k
/— s
T3
V
s-*
9S
•"3
JU
"«
U
« Q»
w y
PM i
- o
^^
s
rr
EC
a
w
i
^^
=
^
c
0
•'S
fi
i
5
^
EC
o
o
^^J
OS
"s
* )
«5
CQ
»>•
i
•o
>
u
o
CJ
•1 ^
« 3
fe
-------
W
co
O
H
H
O
H
o
•M
"eS
"E
ee
B
CA
— CJD
C3 fi
•« S
>»
0)
*
0>
"O
2
IS
i fi
O o
9
9
• O
a
cd
•95-
C-368
-------
5?
PQ
ol
N ffl
53
a
Q
u
H
.2
'S
•s
08
N
08
u
H
09
CO
°°
GO GO GO
i>
«
rt3 I"O
ON
OS
C-369
-------
f
c
9
CO
O
s «
H ^
<
,
IK
£
&
S
^
^s
h"
PS
^
1 -
?§ w
0 ^
^••s
^jfa
^-
=
N ^»
«
PQ
fa ^ ^
oa §
5
<
§
*5jD
1 & «
?^
Q
i 2
Sg
pH
CO
V
y
e
• *•
jg
>£
^
•
• b*
1
O'
TCDD-E
*
M
PH
/— s
S ^1 *$
X-N /— s
o ^ 1 i I
^ ^^
1 1 |
r~~
^a Q&
1 e
;2 a
*a5 5
W •
NH
TT
n
^
"O
*C
a
•
Q
Q
H
C-370
-------
o
HH
H
P
O.22
> C3
^ w
Q *
OWABLE
Sag
£
C3
es
I*
3
o>
u
C
O
U
1!
U
O
o
fl
^>
t-
^
&M
0^
«
II
^
o
-:
II
*l
§
o
z
C-371
"0
A
o
a
=
• •*
X
o
t-
e
s
o 22
** o o
« s-g
c _ «
« C t.
« ° 5
imate
-8
C« O fc
•PN *^« ^U
5 -S ft
H ^ C8
o
V-
o
Q
CO
•-- c
pS3 -S
fl
0
c^ &
o a
% •«
° a
ss
0
o
ON
es
-------
ALLOWABLE DAILY CONSUMPTION
Saginaw Bay Fishes
Based on Total PCBs
FISH
Walleye- 1
WaIleye-2
Walleye-3
Walleye-4
Walleye-5
Walleye- 6
Alewife
Shad
Carp
Perch
Average
ADC
(g/d)
16
18
9
6
13
5
11
31
2
ND
13
% ALLOWABLE
IN DIET
7.4
8.3
4.0
2.6
6.0
2.5
5.1
14.4
1.0
ND
5.8
Daily = Allowable % That Could be in Diet
at The Threshold for Effects
Giesy & Kubiak, 1993
C-372
-------
ALLOWABLE DAILY CONSUMPTION
Saginaw Bay Fishes
Based on TCDD-EQ ( Calc.)
FISH
Walleye- 1
Walleye-2
Walleye-3
Walleye-4
Walleye-5
Walleye-6
Alewife
Shad
Perch
Carp
Average
ADC
(g/d)
6.0
4.0
3.0
1.5
3.8
1.2
5.5
4.5
4.5
1.0
3.5
% ALLOWABLE
IN DIET
2.7
1.8
1 . 4
0.6
1.7
0.6
2.5
2.1
2.1
0.6
1.6
Daily
= Allowable % That Could be in Diet
at The Threshold for Effects
Giesy & Kubiak, 1993
C-373
-------
ALLOWABLE DAILY CONSUMPTION
Saginaw Bay Fishes
Based on TCDD-EQ ( Assay )
FISH
Walleye-1
Walleye-2
Walleye-3
Walleye-4
Walleye-5
Walleye-6
Alewife
Shad
Carp
Perch
Average
ADC
(g/d)
105
32
30
5
30
38
13.5
38
2.2
105
40
% AT JX3WABLE
IN DIET
48
14.7
13.8
2.2
13.8
17
6
17
1
48
18
Daily = Allowable % That Could be in Diet
at The Threshold for Effects
Giesy & Kubiak, 1993
C-374
-------
£
O
Pwo
II
tt £
^2 »PN 0jJ^
Jo* fiu r—
L^^^^^^ ^^^^^^
h.
fl
0
L
u
tt
u
PM
H
O
H
r i
2
GC
^ T 1
fe
3 o
C-375
IT) «
S|
-€8
fO
ON
2
ffi
H^
S "g
H II
-------
w e?
aw
^^
^ h
I
=
H
H
H
O
H
QO
oo
so
QO QO
C-376
w
QO
I
C3
9s
cs
0
§M
H 1
fi
o
HH
g
O1
E3
H
-------
u
(2
>
03
3 w
< <%
'Si b
11
II
^•1
J *
3 ^
?
^c
>
o
p2
3 ^-
1 s
QOJD
< ^
a
H
1
^
?
, ,• i .• .
u ^ ^
. O' ^ S Q^
GO O W ^ . CaJ
H ^ H
^^
J ^ 5
01 • i
«^ • in ^
S • o' ^ ^ o*
H 3 H
v^ ^ ^
N"^ ^
5 ' O «j
^3 ^^ 5
^ fl
fl a
^ \ ^^^~j
^^^g *^^^»
s.
•s
OS
*
ON
'
2
2
2
5
C-377
-------
VI
GO
PQ
O
o
w
oa
U
o
u
a
5
o»
H
u
U
I
o
fl
O
PQ
C-378
O
-------
CC
o>
H
O
03
op
4
Ml
0)
u
C
o
PQ
in
C-379
a;
3
-------
PH
O
IT)
if
OS
oo ON -
/8m)
C-380
en
O>
3
-------
Q
H
is
fl
< g «
»J 3 ?
u
P*
a
c
ffl
C/)
VJ
V©
00
VO
QO
6-
Tf WJ SO
v o « Ss TS «fl
5t* ^* 5? ^ ^ IL
W W W JfT .£• "-
O
rr
.CS .C8 ^C8 .C8 .S .KS
^ • fc & £ £ £
SI
s |
S&"< as
£ S
o Cui ^
C-381
o
A CA
flj Q^
_^*^ ••*
O
-------
CALCULATED TCDD-EQs
PCB IN SAGINAW BAY FISHES
Percent Contribution
FISH
Walleye 1
Walleye 2
Walleye 3
Walleye 4
Walleye 5
Walleye 6
Alewife
Shad
Perch
Carp
Mean
% PCB
58
45
58
45
61
42
69
48
60
43
53
Calculated from H4IIE TEFs & Additive Model
Reported as Percent of Total Calculated TEQs
Giesy & Kubiak, 1993
C-382
-------
CALCULATED TCDD-EQs
PCDD IN SAGINAW BAY FISHES
Percent Contribution
FISH
Walleye 1
Walleye 2
Walleye 3
Walleye 4
Walleye 5
Walleye 6
Alewife
Shad
Perch
Carp
Mean
% PCDD
37
49
33
48
34
48
27
38
33
53
40
:alculated from H4IIE TEFs & Additive Model
.eported as Percent of Total Calculated TEQs
iesy & Kubiak, 1993
C-383
-------
CALCULATED TCDD-EQs
PCDD IN SAGINAW BAY FISHES
Percent Contribution
FISH
Walleye 1
Walleye 2
Walleye 3
Walleye 4
Walleye 5
Walleye 6
Alewife
Shad
Perch
Carp
Mean
% PCDF
5
6
9
7
5
10
4
14
7
4
7
Calculated from H4IIE TEFs & Additive Model
Reported as Percent of Total Calculated TEQs
Giesy & Kubiak, 1993
C-384
-------
~~^\
Q
P Q
H >• =
* ^^^ ^ ^
IS
r ^j r *
II
r^ ™
WD
rv,
gjf
Q ^
U
E
es
1/1
ON
ON
vo
0
00
ON
s
^^5
A \ P*^^ ^F I
I » S
^^^ ^^^
S ca
C-385
3
^«
s
g
O
ON
C5
O
-------
fe
O
O5
85
o
S-i
«
>•
NJ
p*
a
H
P
PQ
s
fc
o
u
^
a
^
p
p
u
H
-------
o
g<
ce
H .
QH
o
c:
"O
O
S
"O
o -^
p
I i§
15
£
'C
o
fi
O
ON
ON
CS
C-387
en
O
-------
o
I-H
H
O o
~
j
J4
a
c:
QO
eo
o
.2
b
•*•*
a
QO
TJ
O
I
S
.2
a
o
J
C-388
O
-------
cc
fc
O
H
§«
8*
*• M
o esa
u
vi
»5 .fi
pS CB
II
K
Q
O
U
H
II
Q
Q
H
*
ns
^
fi
II
il
O «
H "
fl *C
CQ
0
C
a
fl
O
CJ
.9
HH ON
Q Q PQ
u u
PH PH
C-389
O
-------
x
o °
PH "
CJ
^^F
a
PQ
II
O1
Q s
e
o H -a
M
3N
C5 fli
® 0»M
a
o
w
«H T PH
0 Q -
C-390
fr
O
-------
u
o
co
°o
s
^^v
^^^%
o
PQ
WD
g
ON
C-391
-------
C/5
=3
c
o
£fe
ga
H -a
/^ 2
g§
TJ
Ey ej
^P^^^^^ ^^^
^^. cs
^*^
H &
-^ CQ
^w
Fj^ gjj
^^^^^i ^^^^j
0^^m ^^^
B
fc
W
O
O.
W
P
^
Vv^
c»
B
»^ ^O
s S
v^f
^
£
f^ «p*>1
& HJ
CS M
u --a
§
^
^
^%*
5>-
ed
c»
r^K
Cd
cd
s
5
E
2
— ^^^
QM
•^^^^^
O
W
c
u
H
a
•^
C
'DA
C-392
-------
H
O
QO
ON
QO
OD
OJD
QC
fS
u
i
o
U
i
I •
U
O
£
OS
PQ
U
A*
e
o
eg
«
Pk
a
s0'
I*
a-a
Q
Q
O
c
o
PQ
C-393
-------
B
O
u
0
&
2
33
93
fa
ON
1
^ ^
^ ^
ft
CS
CB •=
C-394
-------
PQ
I
o
Ho* t2
j^ ^z
A
oa
i
P9
c:
1
rf
C3
6?
C-395
-------
t/3
PQ
PN
3
a£
3
u
cs
VALUE
WD
c
X
0
— s
C-396
-------
O
r i
PQ
O
H
H U
LJ HH
c/}
o
S
o
.SN
S
OD
«
S
O
S3
P-l 44
a
5 ^
o .2
'D
/^s ^- ^
PQ o **
U « 2
u
X
o H
*
Cfl
o o
2.1
•S Q
•
Q
H U
en
4>
.»-<
O
C-397
-------
o
I-H
cc
O
H
o
O
HH
tt
£
c
03
53 O
•3 **W
CS a*
PQ
Q 2P I
e o>'
Sfi3U
Cd . "i a> JS
0> ^ M ^ ^
p ^ S .
05 fl
c:
o I
05 -< S
•n «
«
« '
O
C-398
-------
ns
o>
2
*3
I -
£ w
W o
q/ PH
fc ^
^ O
x
fl
.2
r ' ^
H H
O»
ed
fl
o
fl
s'B
a.a
»L ^
S? w
^ fi
W3 g
5ft, r^
c: U
o
^ ^
M ptf
o
£
a! v
,* >*^
^ cs
2^5
B »
o
8 r^
« fe ^
v
H ^ H ^ H
68
C5
CA
C-399
O
-------
CC
c>
fl
§
SN
O
£
&H
^ 6s3
Q* S
H pp
"S
«
•
» >» W
•=lf§e'
«*-
9
__
IS
« -e
S 2 5 -a
••S 53 5 1
-
a
<31
o
^
"O :3
o
ON
es
C-400
-------
O
HH
CW
u
fc
OW2
W
u
S
O
O)
S «
*s ^
OD^
T3
Dl
fi
o c
GO
^^
O
"
ce
05
CM
ee
s-<
•S fl o
fl OJD
0
o
U
O Gfi
O E
CQ
-*-*
J
X
&
pfl
x O
Sz
*M 4) ,
0 Jj
X
2 ns
•P^ 5»
U V
V 'O
°- Ji
w 8
- rl
«W
a 85-
•- pa
ce
n s
O 2
Q, S
P"^ «p*
WD
«M c^
0 W
^ ^
S3 S ..
O O -:
•^< ^
1
SM _,
-ta* "O
C «
8l-a
§ §.s
u£§
ON
O
C-401
-------
cc
O
cc
3
U
0 i
U PQ
H
P
CO
a
S
1
a
c«
O
p*
s
•s-s
C3
a
2 -o
&
®
« Q
«
0 «-
5 *> r^ r2
H -Q W fa
ON
C5
wo
V
5
C-402
-------
en
O
C»
3
P
O
U «
l»
0)
OJj
d
C3
rSS *
^ r* P-P<
^
a G>
^^ ^^5
set
v o
15 ft
•s a
pM 4J
o
d
CQ
p^ CO eg S
°2l^
S -2 .s
^ «5 C8
^ 2 a
s ^
Is cs
•PN PHK
00
H PQ
o g *
L- p cd
b
*v d
ee
flS
= =-6
*8
s s >>
1> H 4>
'S "S pd
01
d
OJD
O
^ d
s i
s
u
p^
=
d
0>
.«
E
g
05
na
ON
C5
'08
CA
C-403
-------
fc
O
C/5
S
0 §3
0
a
^ «•
M
W
r.
H
«M
0
.2 -2
§
§ °
-
H
tt 9 i/i
•p* •
ao
cs _ cs ' ^J i5
.3*3 I o ^^
WD - ** -
^ ^
> 'cs
M.2 «
« » fl
"
2
PH ns 5 Q ^ S
^5 oj ^ i—< CQ
^ S3 ^ M ^M
X h9 X U K§ cs
a
o
as
ON
<**
C-404
o
-------
O
s
U
O
0)
o
53
3
S
CM
fl
O
*d
O>
CM
£
"O
P.:
OS
N
03
B
4>
pC
H
ft* ^
S S
1 -2
I 3
Spa
CS
2U
|a
^ a
^
8
2
"3
l
CM
fl
^O
ts
PN
"3
^
^
fl
O
es
o->
^ J -g H
Cft
a
S3 2
w>!3
xi% !•••<
S —
O) ^w
OJD Cd
es «p*
•5 ne
s H
o
PQ
U
c.
13
1/3
I 4S
S o
'BbH
cd
cc ^d
a
3 *
2 «
«£s c>
0£
S
O
U
"«
PM
0
CM
CS
a>
« 2
aj S
O
• P4
wi p2
'% s
04D
Is
«
^*
fl
O
O
S fl
(E^ .p*
S s
I § a
fa s p
fn p2
>^ WD ^*
13 3 3
^i ^N rf,
PQ
V K
£ O
t3 H
1 «
| *
S
O
ptS ns
s 3
sl
.= 0
£ 3> ^
= fl
pS S
« "'
T3
O «i
ON
c*
BP
C-405
-------
CO
o
HH
CO
O a
2
CO
S
CO
£§
-------
to
O
to
z ,
Os=
O
to
V
M
CM
O
Q
Q
a
H
s.s
to
to
^ fl o
£ 2 5r
5 B K
fl ca
§?£
CQ S
O
u
g S S
§.2
fi.e
73 •-< i2
C3
«^
t2 «
p^
S «-- 2 «•
S C 83 M
O '~*
v* fe2
0> S3 pfi S
•S § •- .5
H fi a ^
&
— -2
H fi-1
P-)
^ fl
PH 2
to
4>
01 g S
§ 1
'*» «P^
to
s
o
PBHl ^Z ^*
« "C ^
•V JLrf P4
•*^ S 53
O S «F«.
^^ U
8 O
en
ON
ON
en
V
3
C-407
-------
o
5S
^
o
Overall-V
a «
.2 «
J5 £
1 §-S
g .5 ta-
3 DC
.= o
a
« 15 o
S£ *•
«8 C
.§ * -I -
S-A ^3 SP< ^^
% a> ftj
w s-s S
[Slu
^^^^^ ^^^^^
>»
-" wx
fe s
&
OJ a_,
W5 ^^
cd Cb^
tt o
ell
O
03
of
c» <1
§0
— *—• -^^p
wo js
£^ ^
S3 23 .S
- .5 Q
— i i*
a>
H
Tg F^-l r^r.
A
PH
C
a^
w
fl
o
U
PQ 2
^
c« •
o
eg
o
.03
ON
2
3
2
C-408
0)
• P4
O
-------
z
o
hH
C»
3
U
^o
o
Q
^ O
«8 U
H
a
53' ®
I
Q a
&£
fi
4>
O w
al-
a C3 C;
WW
S
C/3
hi
.O
2 •- = =
a ° °
S « *
o SB es
2
c
00
I =
*> «M
8?
O
,H
CM
C3
4>
^
S
^ fe' L-
H^ M
^ ^
Ct DJD
-T3 Jg S3'
H O hJ
fi-o
.
E
"° 2
OJ ^
w ^
a .a
B"
= ®
o —
ra
es -
a -|
" S
fl fl
O
C-409
-------
O
u
o£
U A
Q
U
H
0
S3
o «0
. a p °
S O
ffl£
o>
-4-»
a
0
o
CJ
41
%
a
g
TJ
a^
£
Is
o
'S
u
a
P
o
^ S3
^ o
^ «iS
«^ gf Q
tn u
.^ cj Cld
dBV ^^^^^
.2 i fe
• o
H ^i «3
Q
o O
.s w y
'• PH
ep
H o
h 5J
OH AH
-2
c;
N
N
3
s
8
03
U
P* o
A^ CM
CQ
N
OS
K
ON
es
C-410
-------
P
H
a
c
o
P
0> fl
PQ *
^ **
" o>
s£
o
i
fl "GO
£«
H ?
S •-
o o 'K
c
IL »
l*«
I
o.S
WD
ce
a
4>
C3
tt
,9)
i&
s
51
^^ r«» P^
0)
o 5 Q
d
.2
-S is
£
I i-2 H
Z Q< -M tr^
S a «
U §
ermine
« S
= e
o
Q S
NM «PN
eg
PQ
U
c
OJ
9
B §
V
PQ ns
o
§•£
o>
pC
H
ne
g s
M «P-<
er
Q S
^•i »PN
ON
C-411
•
O
-------
o>
GT5 S
tt §
gz0
fflS
' a
5O
;o
C/5
C-412
-------
APPENDICES D-F
WORKSHOP MATERIALS
-------
-------
APPENDIX D
WORKSHOP AGENDA
D-l
-------
-------
£
HI
U.S. Environmental Protection Agency
Risk Assessment Forum
Workshop on Ecological Risk Assessment Issues for
2^,7,8-TetracUorodibenzo-p-Dioxin
Radisson Hotel Metrodome
Minneapolis, MN
September 14-15,1993
Agenda
Workshop Chain
Robert Huggett
Virginia Institute of Marine Science
Tuesday, September 14
7:30AM Registration and Onsite Check-In
Eastern Research Group, Inc. (ERG)
PLENARY SESSION
8:30AM Welcome and Introduction
Jack Gentile, U.S. Environmental Protection Agency (U.S. EPA),
Risk Assessment Forum (RAF)
8:45AM Workshop Objectives and Format
Robert Huggett, Virginia Institute of Marine Science
9:OOAM Use of EPA's Framework for Ecological Risk Assessment
William van der Schalie, U.S. EPA, RAF
9:15AM Ecological Effects and Endpoint Selection Issues
Randall Wentsel, U.S. Army
9:30AM Stressor Characterization Issues
William Adams, ABC Laboratories
9:45AM
BREAK
D-3
-------
BREAKOUT GROUPS
10:OOAM Discussion
Exercise 1; Ecological Effects and Endpoint Selection
Randall Wentsel, Workgroup Leader
Exercise 2; Stressor Characterization
William Adams, Workgroup Leader
12:15PM LUNCH
1:30PM Discussion (continued)
Exercise 1; Ecological Effects and Endpoint Selection
Randall Wentsel, Workgroup Leader
Exercise 2; Stressor Characterization
WUUam Adams, Workgroup Leader
3:OOPM
BREAK
3:15PM
4-.OOPM
4:45PM
5:15PM
PLENARY SESSION
Summary Presentation and Discussion
Exercise 1; Ecological Effects and Endpoint Selection
Randall Wentsel, Robert Huggett
Summary Presentation and Discussion
Exercise 2: Stressor Characterization
William Adams, Robert Huggett
Observer Comments
ADJOURN
D-4
-------
Wednesday, September 15
PLENARY SESSION
8:30AM
8:45AM
Conceptual Model Development Issues
Charles Menzie, Menzie-Cura & Associates
BREAKOUT GROUPS
Discussion
Exercise 3; Conceptual Model Development
Charles Menzie, Workgroup Leader
Exercise 3; Conceptual Model Development
Robert Huggett, Workgroup Leader
10:OOAM
BREAK
10:15AM Discussion (continued)
Exercise 3i Conceptual Model Development
Charles Menzie, Workgroup Leader
Exercise 3: Conceptual Model Development
Robert Huggett, Workgroup Leader
12:OONOON LUNCH
1:15PM
1:45PM
2:30PM
3:30PM
PLENARY SESSION
Observer Comments
Summary Presentation and Discussion
Exercise 3; Conceptual Model Development Issues
Charles Menzie, Robert Huggett
Identification of Major Uncertainties and Research Needs
Robert Huggett, Workgroup Leaders
ADJOURN
D-5
-------
-------
APPENDIX E
WORKSHOP PARTICIPANTS AND FINAL OBSERVER LIST
E-l
-------
-------
U.S. Environmental Protection Agency
Risk Assessment Forum
Workshop on Ecological Risk Assessment Issues for
2.3,7,8-Tetrachlorodibenzo-p-Dioxin
Radisson Hotel Metrodoine
Minneapolis, MN
September 14-15,1993
Participants
William Adams
Vice President
Environmental Toxicology
ABC Laboratories
7200 East ABC Lane
Columbia, MO 65205
314.474-8579
Fax: 314-443-9033
Nigel Blakley
Ecologjst
Technical Policy Section
Washington State Department of Ecology
637 Woodland Square Loop, SW
Lacey, WA 98503
206-438-3063
Fax: 206-438-3050
Peter Chapman
Senior Partner
EVS Environment Consultants
195 Pemberton Avenue
North Vancouver, BC V7P2R4
Canada
604-986-4331
Fax: 604-662-8548
Keith Cooper
Associate Professor
Graduate Program in Toxicology
Environmental & Occupation
Health Science Institute
Rutgers University
681 Frelinghuysen Road
Piscataway, NJ 08855-1179
908-932-2230
Fax:908-932-0119
G. Michael DeGraeve
Director
Great Lakes Environmental Center, Inc.
739 Hastings Street
Traverse City, MI 49684
616-941-2230
Fax: 616-941-2240
Joseph DePinto
Professor
Department of Civil Engineering
University of New York at Buffalo
White Road - 207 Jarvis Hall
Buffalo, NY 14260-4400
716-645-2088
Fax: 716-645-3667
E-3
-------
John Giesy
Distinguished Professor
of Fisheries & Wildlife
Michigan State University
13'Natural Resources Building
East Lansing, MI 48824-1222
517-353-2000
Fax: 517-336-1699
Robert Huggett
Professor
Virginia Institute of Marine Science
College of William and Mary
Route 1208
Gloucester Point, VA 23062
804-642-7236
Fax: 804-642-7186
Wayne Landis
Director, Institute of Environmental
Toxicology & Chemistry
Huxley College of Environmental Studies
Western Washington University
516 High Street
Bellingham, WA 98225-9079
206-650-6136
Fax: 206-650-7284
Charles Menzie
President
Menzie-Cura & Associates, Inc.
One Courthouse Lane - Suite 2
Chelmsfbrd, MA 01824
508-453-4300
Fax: 508-453-7260
Derek Muir
Research Scientist
Department of Fisheries and Oceans
Freshwater Institute
501 University Crescent
Winnipeg, MB R3T2N6
Canada
204-983-5168
Fax: 204-984-2403
Thomas O'Connor
Chief, Coastal Monitoring Branch
National Oceanic &
Atmospheric Administration
1035 East West Highway
SSMC4-10th Floor
Silver Spring, MD 20910
301-713-3028
Fax: 301-713-4388
Robert Pastorok
Principal Scientist
PTI Environmental Services
15375 Southeast 30th Place
Suite 250 •
Bellevue,WA 98007-6500
206-643-9803
Fax: 206-643-9827
Richard Peterson
Professor of Pharmacology/Toxicology
School of Pharmacy
University of Wisconsin
425 North Charter Street
Madison, WI 53706
608-263-5453
Fax: 608-265-3316
Paul Rodgers
Vice President/Scientist
LTI Limno-Tech, Inc.
2395 Huron Parkway^
Ann Arbor, MI 48104
313-973-8300
Fax: 313-973-1069
Thomas Sibley
Associate Professor
Fisheries Research Institute
University of Washington
Fisheries Center - Room 104
(WH-10)
Seattle, WA 98195
206-543-4257
Fax: 206-685-7471
E-4
-------
John Stegeman
Senior Scientist ; .
Redfield Laboratory
Woods Hole Oceanographic Institution
266 Woods Hole Road (RM-342)
Woods Hole, MA 02543
508-457-2000
Fax:508-457-2169
John Sullivan
Water Quality Standards Program Supervisor
Wisconsin Department of Natural Resources
101 South Webster Street (WR-2)
Madison, WI 53707
608-267-9753
Fax: 608-267-2800
Randall Wentsel
Team Leader
Edgewood Research,
Development and Engineering Center
U.S.Army
5641 Alley Road (SGBRD-ATL)
Aberdeen Proving Ground, MD 21010-5423
410-671-2129
Fax: 410-671-2081
Bill Williams
Senior Wildlife Toxicologist/Vice President
Ecological Planning & Toxicology, Inc.
5010 Southwest Hout Street
Corvallis,OR 97333-9540
503-752-3707
Fax:503-753-9010
E-5
-------
-------
U.S. Environmental Protection Agency
Risk Assessment Forum
Workshop on Ecological Risk Assessment Issues for
2-3,7,8-Tetrachlorodibenzo-p-Dioxin
Radisson Hotel Metrodome
Minneapolis, MN
September 14-15,1993
Final Observer List
Steven Bradbury
Acting Associate Director for Research
Environmental Research Laboratory
U.S. Environmental Protection Agency
6201 Congdon Boulevard
Duluth, MN 55804
218-720-5527
Fax: 218-720-5703
Patricia Cirone
Environmental Scientist
U.S. Environmental Protection Agency
1200 Sixth Avenue
Seattle, WA 98101
206-553-1597
Philip Cook
Research Chemist
Environmental Research Laboratory
U.S. Environmental Protection Agency
6201 Congdon Boulevard
Duluth, MN 55804
218-720-5553
Fax: 218-720-5539
Thomas Deardorff
Research Scientist
Erling Riis Research Laboratory
International Paper
P.O. Box 2787
Mobile, AL 36652
205-470-3274
Fax: 205-470-3280
Bradford DeVore
Counsellor
Womble, Carlyle, Sandridge & Rice
301 South College Street
3300 One First Union Center
Charlotte, NC 28202-6025
704-331-4941
Fax: 704-331-4955
Russell Erickson
Research Chemist
Environmental Research Laboratory
U.S. Environmental Protection Agency
6201 Congdon Boulevard
Duluth, MN 55804
218-720-5534
Fax: 218-720-5539
Robert Fisher
Director, National Health Effects Program
National Council of the Paper Industry for
Air and Stream Improvement, Inc.
P.O. Box 141020
Gainesville, FL 32614-1020
904-377-4708
Fax: 904-371-6557
E-7
-------
Jack Gentile
Science Coordinator
Risk Assessment Forum
Environmental Research Laboratory
U.S. Environmental Protection Agency
27 Tarzwell Drive
Narragansett, RI 02882
401-782-3015
Fax: 401-782-3030
Maria Gomez-Taylor
Environmental Scientist
U.S. Environmental Protection Agency
401M Street, SW (WH-586)
Washington, DC 20460
202-260-1639
Fax: 202-260-5394
Stewart Holm
Manager, Government Affairs
Science & Health
Georgia Pacific Corporation
1875 Eye Street, NW - Suite 775
Washington, DC 20006
202-659-3600
Fax: 202-223-1398
Timothy lannuzzi
Senior Environmental Scientist
ChemRisk
1685 Congress Street
Stroudwater Crossing
Portland, ME 04102
207-774-0012
Fax: 207-774-8263
Gary Kayajanian
Multinational Business Services, Inc.
11 Dupont Circle
Washington, DC 20036
202-293-5886
Far 202-939-6969
Edward Krasny
Regulatory Specialist
Kimberly-Clarke Corporation
1400 Holcomb Bridge Road
Building 200 - 2nd Floor
Roswell, GA 30076
404-587-8785
Fax: 404-587-7093
Allan Kremer
Compliance Engineer
Hennepin Energy Resource Company
505 Sixth Avenue, N
Minneapolis, MN 55405
612-333-7303
Fax: 612-333-7347
Cynthia Nolt
Biologist
U.S. Environmental Protection Agency
401 M Street, SW (WH-586)
Washington, DC 20460
202-260-1940
Fax: 202-260-5394
Susan Norton
Environmental Scientist
Office of Health &
Environmental Assessment
U.S. Environmental Protection Agency
401 M Street, SW (RD-689)
Washington, DC 20460
202-260-1722
Fax: 202-260-3955
J.W. Owens
Principal Scientist
Human & Environmental Safety
The Procter & Gamble Company
6100 Center Hill Avenue - Room P2E08
Cincinnati, OH 45224
513-634-7151
Fax: 513-634-7364
E-8
-------
Dorothy Patton
Executive Director and Chair
Risk Assessment Forum
U.S. Environmental Protection Agency
401 M Street, SW (RD-672)
Washington, DC 20460
202-260-6743
Fax: 202-260-3955
Susan Snider
Environmental Specialist
American Forest & Paper Association
1250 Connecticut Avenue, NW - 2nd Floor
Washington, DC 20036
202-463-2589
Fax:202-463-2423
William van der Schalie
Science Coordinator
Risk Assessment Forum
U.S. Environmental Protection Agency
401 M Street, SW (RD-672)
Washington, DC 20460
202-260-4191
Fax: 202-260-3955 .
Dwain Winters
Director, Dioxin Policy Project
Office of Policy, Planning and Evaluation
U.S. Environmental Protection Agency
401 M Street, SW (LE-132L)
Washington, DC 20460
202-260-8558
Fax: 202-260-8392
William Wood
Associate Director
Risk Assessment Forum
U.S. Environmental Protection Agency
401 M Street, SW (RD-672)
Washington, DC 20460
202-260-6743
Fax: 202-260-3955
Maurice Zeeman
Chief, Environmental Effects Branch
Health and Environmental
Review Division
U.S. Environmental Protection Agency
401 M Street, SW (TS-796)
Washington, DC 20460
202-260-1237
Fax: 202-260-1236
E-9
-------
-------
APPENDIX F
WORK GROUP ASSIGNMENTS
F-l
-------
-------
U.S. Environmental Protection Agency
Workshop on Ecological Risk Assessment Issues for
2,3,7,8-Tetrachlorodibenzo-p-Dioxin
September 14-15,1993
WORKGROUP ASSIGNMENTS
Tuesday. September 14.1993
Exercise 1: Ecological Effects
and Endpoint Selection
Workgroup Leaden
Randall Wentsel
U.S.Army
Nigel Blakley
Washington State Department of Ecology
Peter Chapman
EVS Environmental Consultants
G. Michael DeGraeve
Great Lakes Environmental Center
Wayne Landis
Institute of Environmental
Toxicology and Chemistry
Huxley College of Environmental Studies
Richard Peterson
School of Pharmacy
University of Wisconsin
Thomas Sibley
Fisheries Research Institute
University of Washington
John Stegeman
Woods Hole Oceanographic Institution
John Sullivan
Wisconsin Department of Natural Resources
Bill Williams
Ecological Planning Toxicology, Inc.
Exercise 2: Stressor Characterization
Workgroup Leaden
William Adams
ABC Laboratories
Keith Cooper
Health Science Institute
Rutgers University
Joseph DePinto
Department of Civil Engineering
State University of New York at Buffalo
John Giesy, Jr.
Department of Fisheries and Wildlife
Michigan State University
Robert Huggett
Virginia Institute of Marine Science
College of William and Mary
Charles Menzie
Menzie-Cura & Associates, Inc.
Derek Muir
Department of Fisheries and Oceans
Thomas O'Connor
National Oceanic and Atmospheric
Administration
Robert Pastorok
PIT Environmental Services
Paul Rodgers
LTI, LimnoTech, Inc.
Wednesday. September 15.1993
Exercise 3: Conceptual Model Development
The Conceptual Model Development group will be divided into two subgroups as above, with
Robert Huggett and Charles Menzie as workgroup leaders.
F-3
-------
-------
APPENDIX G
REVISED RISK ASSESSMENT CONCEPTUAL MODEL
G-l
-------
-------
Revised Conceptual Model for the Southern Reservoir
Note: See appendix A for the original conceptual model. Comments on the conceptual model
figures in appendix A are provided in the text boxes. A revised diagram of the overall
conceptual model is shown hi figure 4 of the main report
Hie foundation for the conceptual model is the tissue residue approach contained in the
TCDD Interim Report. Concentrations of TCDD and related chemicals (PCDDs, PCDFs,
coplanar PCBs, and monoorthochlorine-substituted analogues of die coplanar PCBs) that act by
an Ah receptor mechanism in eggs of fish and in fish consumed by piscivorous birds and
mammals are presently the exposure metrics upon which the estimation of the potential for
adverse effects to the organism must be
based. In this case, figure 2 (appendix
A; see text box) shows the logical flow
of assessment information when
thresholds for adverse ecological effects Flgure 2
on fish and wildlife population
protection goals are to be related to safe
chemical loadings to the ecosystem. A
novel feature of this model is the
between boxes are specific types of
models that are used to interrelate the
Conceptual Model Comments:
This figure represents an ideal that
may not be attainable, hi that our ability to
do population- and ecosystem-level
transformation of reproductive effects to evaluations « quite hmited The chaUenge
populations and community level ?f ^s "?****tO ldf ^SL
impacts. The boxes in this conceptual loa^ m ^effluent to evels in food,
model include endpoints that are !"•"» ^d sedunent to ^esidue? "f effects
generally quantitative. The arrows m aquatic organisms and associa ed
6 } M wildhfe. The level of sophistication
required hi the analysis will depend on the
assessment endpoints. All the models J?01*5? «!*?*? ** specific questions
are reversible; hence the two-way **"* a*c4 A tHnd «HP"«* 1S
advisable—starting with simple methods
and assumptions and proceeding to more
complex analyses (e.g., those requiring
reservoir segmentation) only if necessary.
arrows. The conceptual model applies
equally to assessments that seek to
determine risks associated with a known
or predicted chemical loading (right to
left flow of steps). This model may be
used, following further review and
validation, hi the establishment of acceptable residue levels that form the basis for translation to
permit conditions and associated effluent treatment standards. Within the conceptual model,
exposure characterization is performed with the assistance of appropriate exposure models.
Uncertainty associated with these models is related to the uncertainty associated with
measurements of physicochemical properties, lipid and organic carbon relationships, and spatial-
temporal heterogeneity of chemical concentrations in the ecosystem.
Figure 2 shows effects on ecological systems linked to exposure levels through chemical
residues. While the food chain is recognized as the primary route of exposure for consumers,
uptake of TCDD-like chemicals directly from the water by primary producers (e.g., algae) is
important. Since the known adverse effects of TCDD and related chemicals for fish are directly
attributable to exposure of the embryo, the chemical residue levels hi eggs are at present the
exposure metric of primary interest. The exposure metric of interest for piscivorous avian and
G-3
-------
mammalian wildlife is the concentration of TCDD and related chemicals in fish and other prey
species consumed by these animals that causes reproductive tenacity. If more sensitive endpoints
are found, they may be included. Care must be taken to ensure that appropriate exposure
models are chosen for each aquatic and wildlife species of concern.
Atmospheric deposition of PCBs into Omigoshee Reservoir resulted in fish having
background total PCB concentrations of 500 ng/g. The background levels of TCDD-like PCBs
could result in fish inhabiting this reservoir being at greater risk of approaching an adverse body
burden. The background body burden of TCDD-like PCBs would decrease the amount of
TCDD-like PCDDs and PCDFs that could be released into the reservoir from the paper mill.
Based on the occurrence of background levels of PCBs, site-specific BSAF factors
could be determined. This would allow for pre-paper mill site information for modeling the
reservoir.
Conceptual Model Comments:
Figures 3, 4 (Appendix A)
Panelists felt figure 3 was adequate
as a general descriptive tool but not for
indicating the relative importance of
different pathways. It was suggested that
the benthic invertebrates be divided into
two groups to better show links to birds,
fish, and mammals. Spatial heterogeneity
in the distribution of TCDD in the
reservoir (main channel vs. side arms) will
be especially important if there is parallel
heterogeneity in the distribution of aquatic
species of concern.
Panelists suggested that figure 4
should include crayfish and mussels. It
was suggested that it would be important
to know the dietary proportion of different
organisms for species of concern.
Panelists developed additional
figures for the conceptual model, as
discussed in section 5 of this report.
Figure 3 (appendix A; see text
box) illustrates the patfiways for TCDD
exposures and bioaccumulation in
Omigoshee Reservoir biota. Selection of
which boxes and arrows require inclusion
in the assessment is a reflection of the
management question being addressed
and what supporting data are likely to be
available. Thus TCDD exposure to fish
and wildlife in natural systems is
expected to be primarily via
contaminated food, and effects are often
best referenced to concentration in food
or in the receptor organism itself. Thus
TCDD residues in aquatic organisms,
and the distribution and bioavailability of
TCDD in water and sediments, will be of
central concern in this assessment.
Concentrations and their spatial-
temporal heterogeneity in sediments,
suspended solids, and water should be
estimated using suitable fate-and-
transport models. Tissue residue levels
can be estimated via transport models
and either BSAF, BSSAF, or food chain
models, such as shown in figure 4
(appendix A; see text box). The
concentration of chemical predicted for the whole organism can be related to specific tissue
concentrations through lipid normalization or a more specific toxicokinetic model. The
variability of tissue residues among different organisms and their relationship to organic carbon
in suspended solids and lipid in organisms are major uncertainties that must be considered. A
third bioaccumulation approach is to estimate chemical concentrations of concern in the surface
G-4
-------
sediments of the organism's habitat by application of measured or estimated biota-sediment
accumulation factors (BSAF; see section 3.5 of the Interim Report) to the tissue residue-toxic
response relationship for each species of concern. The BSAF approach has an advantage of
using an accumulation factor that can be directly measured in contaminated ecosystems.
G-5
•& U.S. GOVERNMENT PRINTING OFFICE: 1994 — 550-001 /00163
-------
|