JJZDA United States
Environmental Protection Agency
REGION/ORD WORKSHOP ON
AQUATIC LIFE CRITERIA
SUMMARY REPORT
December 4-7, 2001
Seattle, WA
-------
US Environmental Protection Agency
Region/ORD Workshop on Aquatic Life Criteria December 4-7, 2001
THIS PAGE INTENTIONALLY LEFT BLANK
11
-------
US Environmental Protection Agency
Region/ORD Workshop on Aquatic Life Criteria December 4-7, 2001
TABLE OF CONTENTS
FOREWORD ix
EXECUTIVE SUMMARY xi
WORKSHOPS, MODELS, SOFTWARE, AND WORKGROUPS ASSOCIATED WITH THE
DEVELOPMENT OF AQUATIC LIFE CRITERIA xiii
WORKSHOP SESSION SUMMARIES ... .. 1
US EPA Ollice of ReKaicli and Development
111
-------
US Environmental Protection Agency
Region/ORD Workshop on Aquatic Life Criteria December 4-7, 2001
PLENARY SESSION: OVERVIEW OF AQUATIC LIFE CRITERIA 1
Toxic Chemicals 1
Programmatic Overview of Science - Charles Delos (OW/OST) 1
Water Quality Toxics: Short- and Long-Term Needs - Debra L. Denton (Region 9) 4
Habitat 6
Impaired Habitat: A Water Program Retrospective/Perspective - Douglas J. Norton
(OW/OWOW) 6
Strengthening the Use of Aquatic Habitat Indicators in the Clean Water Act -
Steve Bauer (Pocket Water, Inc. -Idaho) 8
The ORD/NHEERL Approach to Habitat Alteration Research -
James H. Power (ORD/NHEERL) 10
Sediments 12
Suspended and Embedded Sediments: Status Report and Update from the Office of Water
- Susan K. Jackson (OW/OST) 12
Suspended Solids and Sediments Risk Management Research -
Christopher T. Nietch (ORD/NRMRL) 15
Nutrients 17
USEPA National Nutrient Criteria Program Approach to Reference Condition
Development-George Gibson (OW/OST) 17
Nutrient Criteria: Challenges Facing Regions and States - Danielle Tillman (Region 5) 19
Biocriteria 21
National Framework for Tiered Aquatic Life Uses in State and Tribal Water Quality
Standards - Update on Guidance Development - Susan K. Jackson (OW/OST) ..21
Biological Assessments in Region 10 - Approaches, Application and Research Needs -
Gretchen Hayslip (Region 10) 23
IV
-------
US Environmental Protection Agency
Region/ORD Workshop on Aquatic Life Criteria December 4-7, 2001
BIOCRITERIA AND NUTRIENTS SESSION 25
Establishing Multi-Use Reference Sites for Biological & Nutrient Criteria Development -
Don Huggins (University of Kansas) 25
Reference Condition for Biological Integrity - Phil Larsen (ORD/NHEERL) 27
The Use of Reference Condition in Support of Surface Water Assessments and Criteria
Development in Ohio -
Chris O. Yoder (Midwest Biodiversity Institute, Columbus, OH) 29
Use of Reference Sites and Conditions in the Development of Nutrient Criteria -
George Gibson (OW/OST) 33
Developing Nutrient Criteria Using Multi-Metric Indices: A case study in the Mid-
Atlantic - John Hutchens (ORD/NERL) 35
NHEERL National Nutrients Research Implementation Plan -
Emile Lores (ORD/NHEERL) 37
Aquatic Life Use (ALUS) Concept of Reference Sites - Susan K. Jackson (OW/OST) ..38
CONCURRENT BREAKOUT SESSIONS 41
BREAKOUT SESSION I: Multi-Use Reference Sites 41
BREAKOUT SESSION II: Charting a Statistical Course for Navigating the Murky Waters
of Bioindicator Development 45
BREAKOUT SESSION III: Aquatic Life Use Support (ALUS) 49
The Biological Condition Gradient- Susan P. Davies 50
Progression of Ecological Degradation in Mid-Atlantic Streams -
Lester Yuan and Susan Norton 51
Numeric Biocriteria - Rick Hafele 53
Idaho Stream Classification compared to ALUS - Cyndi Grafe 56
TOXIC CHEMICALS SESSION 59
Risk-Based Criteria - Russ Erickson (ORD/NHEERL) 59
-------
US Environmental Protection Agency
Region/ORD Workshop on Aquatic Life Criteria December 4-7, 2001
Discussion of Proposed Guidelines Revisions - Charles Delos (OW/OST) 62
ESA Consultation on Toxic Pollutant Criteria - K.M. Kubena (Region 10) 67
Data Quality, New Information, and Interagency Research Coordination -
Chris Tatara and Tracy Collier (National Marine Fisheries Service) 69
Emerging ESA Issues - Steven Schwarzbach (U.S. Fish and Wildlife Service) 71
Surrogate Species in Assessing Contaminant Risk for Endangered Fishes -
Foster Mayer (ORD/NHEERL) 74
Predicting the Toxicity of Metals to Aquatic Organisms: The Biotic Ligand Model -
Charles Delos (OW/OST) 76
Dietary Metals: How Important Are They? -Russ Erickson 78
Numerical (Criteria) for Sediment-Associated Chemicals -
David R. Mount (ORD/NHEERL) 82
Comparing WQC to Site-Specific Ecological Risk Assessment Results at R9 $fund Sites -
Ned Black and Clarence Callahan (Region 9) 86
Persistent Bioaccumulative Toxicants - Philip M. Cook (ORD/NHEERL) 89
Toxic Chemicals Session: Assessing Risks to Wildlife - Rick Bennett (ORD/NHEERL) 93
Derivation of New Jersey-Specific Wildlife Values as Surface Water Quality Criteria for:
PCBs, DDT, and Mercury -
Dana Thomas and Dan Russell (U.S. Fish and Wildlife Service) 95
NHEERL Wildlife Research Demonstrations Project: Methods to Assess Risks to
Piscivorous Bird Populations - Rick Bennett (ORD/NHEERL) 99
VI
-------
US Environmental Protection Agency
Region/ORD Workshop on Aquatic Life Criteria December 4-7, 2001
APPENDIX A: AGENDA AND LIST OF POSTERS A-l
APPENDIXB: LIST OF PARTICIPANTS B-l
APPENDIX C: SLIDES FROM PRESENTATIONS C-l
APPENDIXD: PLENARY FLIP CHART NOTES D-l
APPENDIX E: WORKSHOP PARTICIPANT EVALUATION SUMMARY E-1
vn
-------
US Environmental Protection Agency
Region/ORD Workshop on Aquatic Life Criteria December 4-7, 2001
THIS PAGE INTENTIONALLY LEFT BLANK
Vlll
-------
US Environmental Protection Agency
Region/ORD Workshop on Aquatic Life Criteria December 4-7, 2001
FOREWORD
The ORD/Regional Training Workshop on Aquatic Life Criteria was the eighth in a series of
Regional Science Topic Workshops sponsored by the Office of Science Policy (OSP) in the
Office of Research and Development (ORD) at the United States Environmental Protection
Agency (EPA). Other workshops in this series included:
Asthma: The Regional Science Issues
Communicating Science: Waves of the Future Info Fair
Fully Integrated Environmental Location Decision Support (FIELDS)
Non-Indigenous Species
Pesticides
Endocrine Disruptors
Emerging Issues Associated with Aquatic Environmental Pathogens
The objectives of the Regional Science Topic Workshops are to: 1) establish a better cross-
Agency understanding of the science applicable to specific region-selected human health and/or
ecological topics, and 2) develop a network of EPA scientists who will continue to exchange
information on these science topics as the Agency moves forward in planning education,
research, and risk management programs.
Each year, EPA regions identify priority science topics on which to conduct workshops. The
workshops address the science issues of greatest interest to the regions on the selected topic area.
Each workshop is planned and conducted by a team of regional, ORD, and interested program
office scientists, is led by one or more Regional Science Liaisons (RSLs) to ORD and is
facilitated by a regional chairperson. Participants maintain the cross-Agency science networks
they establish at the workshops through planned post-workshop projects and activities such as
the identification of collaborative research opportunities, the creation of information sharing
mechanisms (e.g., interactive web sites), and the development of science fact sheets for regional
use.
For additional information on a specific workshop or on the Regional Science Topic Workshop
series in general, contact David Klauder in ORD's Office of Science Policy (202-564-6496).
IX
-------
US Environmental Protection Agency
Region/ORD Workshop on Aquatic Life Criteria December 4-7, 2001
THIS PAGE INTENTIONALLY LEFT BLANK
-------
US Environmental Protection Agency
Region/ORD Workshop on Aquatic Life Criteria December 4-7, 2001
EXECUTIVE SUMMARY
The ORD/Regional Training Workshop on Aquatic Life Criteria' was held on December 4 -
December 7, 2001, in Seattle, Washington. The workshop was chaired by Patricia Cirone
(Region 10) with support from John Helvig (Region 7's Regional Science Liaison).
The workshop was organized into three sessions:
/. Plenary: Overview of Aquatic Life Criteria
II. Biocriteria and Nutrients
III. Toxic Chemicals
Scientists from EPA (Regions and Office of Research and Development), US Fish and Wildlife
Service, National Marine Fisheries Service, and states presented the status of scientific methods
to derive aquatic life criteria as well as the limitations of these current criteria for setting
standards. The need for new and better science came up as a frequent item for discussion, as did
questions and concerns on the implementation and regulatory feasibility of some new advances.
General Overview
The Plenary session was co-chaired by Bob Spehar (ORD/NFtEERL) and Patricia Cirone
(Region 10). This session provided brief overviews of the science dealing with major water
quality research areas including: toxic chemicals, habitat, sediments, nutrients, and biocriteria.
The Regional presenters described their short term and long term problems with applying
scientific research to regional decision making. Scientists from the EPA Office of Water
discussed the status of guidance for determining criteria. The ORD presenters described the
current and future scientific research for developing new or improved aquatic life and aquatic
dependent wildlife criteria.
Habitat and clean sediments were only discussed in the plenary session. It is clear from the
discussion in the plenary session that the concept of criteria has been expanded to include all
elements of aquatic ecosystems. However, it is not clear how to move from concept to
application. The speakers presented an excellent case for the importance of habitat parameters in
determining the quality of the nations waters. They also presented an excellent overview of the
Aquatic life criteria = reference concentrations for the protection of organisms that rely on
aquatic ecosystems to sustain life; exposure is through direct contact with water or ingestion of
prey items that inhabit aquatic ecosystems.
XI
-------
US Environmental Protection Agency
Region/ORD Workshop on Aquatic Life Criteria December 4-7, 2001
scientific advances. The EPA Total Maximum Daily Load (TMDL) guidance documents on
nutrients and sediments are a good example of the types of documents which are useful to
Regions, states, and tribes.
The plenary session was followed by two days of in-depth discussions of biological, nutrient,
and toxic chemical criteria. During these sessions scientists from other federal and state
agencies were invited to present information and join in the workshop discussions. The goal of
the in-depth discussions was to prepare an outline of the concerns and considerations in
establishing aquatic life criteria.
The Biocriteria and Nutrients session was co-chaired by Gary Welker (Region 7) and Susan
Cormier (ORD/NERL). This session first provided brief overviews of the science for
determining reference sites for development of biocriteria and nutrient criteria and then dealt
specifically, in breakout sessions, with approaches for establishing multi-use reference sites,
development of statistical methods for bioindicators, and the development of methods for
establishing an aquatic life use support system.
The conclusions and recommendations of the groups were that the use of reference sites is the
preferred method for aquatic life criteria. The groups also agreed that biocriteria are important.
However, the methods for implementing the desired procedures for establishing reference sites
are difficult. More guidance is needed on how to apply these research and guidance procedures
to regional decision making.
The Toxic Chemicals session was co-chaired by Rick Bennett (ORD/NHEERL) and Lisa
Macchio (Region 10). This session provided overviews of the science regarding current aquatic
life criteria guidelines, emerging issues involving the Endangered Species Act (ESA),
extrapolation techniques for assessing risk of species using limited data, risk-based water quality
criteria (WQC), and issues involving the risks of both inorganic and persistent bioaccumulative
chemicals to aquatic life and aquatic dependent wildlife.
The participants in the toxic chemical session concluded that there was not enough time to
discuss all the important issues. However, this workshop provided an opportunity to establish
partnerships and exchange information. The group also encouraged EPA headquarters to move
some of the research conclusions into guidance which could be implemented by states and
regions. In particular, the wildlife discussion was too limited to reach any conclusions regarding
the status of the scientific approaches.
Regions requested implementation guidance to assist states in selecting test species and methods
for permits limiting toxic chemicals. There needs to be some interim guidance on sediment
criteria while methods are being developed. It is unclear where OW and ORD are in their
combined efforts.
-------
US Environmental Protection Agency
Region/ORD Workshop on Aquatic Life Criteria December 4-7, 2001
Summary of Discussion
Discussions during the three sessions included questions and answers regarding the presentations
on current and future work involving aquatic life and wildlife criteria, as well as new EPA
regulations or improved approaches for biological assessments. The need for new and better
science to develop improved criteria came up often in the discussions, as did questions and
concerns about the implementation and regulatory feasibility of some of the new approaches
being developed. Many participants expressed a wish to have similar workshops on an annual
basis because the workshop provided an informal way of maintaining contact with individuals in
similar fields and situations, as well as a way of keeping informed of new developments.
The group discussed the inconsistency in applying criteria across the regions. The gap between
science and implementation is huge. Regions need some type of national guidance to transfer
known scientific methods to applications of criteria.
Regions need to be included in the joint Office of Water and ORD research strategy meetings.
Regions also need to be included in the discussions with other agencies (USFWS, NMFS) so that
there is a clear understanding across all levels of agency scientists regarding implementation of
science knowledge to criteria development.
There is a need for ORD to provide short term technical advice to states, tribes, and regions.
Hopefully, the contacts made at this workshop will continue to be available as technical
advisors.
There was much discussion about the new process for developing criteria by following the risk
paradigm. This process has yet to be evaluated by regions or states. Although the existing
criteria fit well within the risk paradigm, there are aspects of criteria development which need to
be improved e.g., time variability, dose response curves.
Due to lack of data there is a need to develop criteria based on best professional judgement.
Because of the subjectivity of best professional judgment, participants at the meeting were
encouraged to engage in dialogue with their scientists in all agencies.
The scientists from the US Fish and Wildlife Service, National Marine Fisheries, and states were
particularly thankful for an opportunity to interact with EPA scientists. The discussions and
exchange of information helped to form the basis for improved working relationships for
different resource agencies with similar missions to understand the limitations that each faces
when attempting to derive criteria for agency decision making.
Due to the breadth of topics in this workshop, posters and computer software demonstrations
were on display throughout the meeting. In addition, the participants were asked to prepare a
XI
-------
US Environmental Protection Agency
Region/ORD Workshop on Aquatic Life Criteria December 4-7, 2001
short description of any tools (workshops, software, sampling & analytical techniques, etc.) for
developing aquatic life criteria. One objective of the workshop was to put these tools into a
matrix which could be shared by all meeting participants.
xn
-------
US Environmental Protection Agency
Region/ORD Workshop on Aquatic Life Criteria December 4-7, 2001
WORKSHOPS, MODELS, SOFTWARE, AND WORKGROUPS ASSOCIATED WITH THE
DEVELOPMENT OF AQUATIC LIFE CRITERIA
Strategic Planning and Research Coordination
Contact: Bob Spehar (ORD/NHEERL/MED-Duluth)
The Office of Water and ORD have conducted joint exercises that have outlined the long-
term strategy and implementation plans for research in the area of WQC. These include the
Strategic Planning and Research Coordination (SPRC) document, Aquatic Stressors: Framework
and Implementation Plan for Effects Research, and the Water Quality Standards and Criteria
Strategy (WQSCS). The SPRC document is currently being finalized and has been developed as
a joint effort between OW and ORD to establish cross program contacts for a joint research
planning process and to initiate cross program working teams in the areas of: toxic chemicals,
modeling (TMDLs), ecological assessments and restoration, nutrients, critical aquatic habitats,
and microbial and pathogen contamination. The Aquatic Stressors document, developed by
ORD/NHEERL, is based in part on the SPRC discussions and outlines the long-term effects
research for developing new or improved criteria for habitat alteration, nutrients, suspended and
bedded sediments, and toxic chemicals. This document also delineates the research that will
help develop diagnostic tools for a decision support system for resource managers. The WQSCS
document is currently being developed by OW to provide a customer-focused, long-term
strategic vision and direction for the water quality standards and criteria program. This strategy
is intended to be fully integrated with the needs of the people and programs depending upon
standards or criteria as environmental endpoints. This strategy will set near-term (2-3 years) and
long-term (5-7 years) priorities for national water quality standards and criteria for use by both
headquarters and the Regions to meet regulatory mandates.
Ecological Risk Assessment Support Center (ERASC)
Contact: Mike Kravitz (ORD/NCEA-Cincinnati)
http ://intranet. epa. gov/ncear/erasc/index.htm
Habitat
Contact: Doug Norton, (OW/OWOW)
Habitat Cluster (1992)
Watershed Restoration Action Strategies
Watershed Assessment, Tracking, and Environmental Results (WATERS) database
http ://www. epa. gov/waters/
Xlll
-------
US Environmental Protection Agency
Region/ORD Workshop on Aquatic Life Criteria December 4-7, 2001
TMDL tracking system:
http://oaspub.epa. gov/watrs/national_rept. control?p_cycle= 1998
CALM guidance: http://www.epa.gov/owow/monitoring/calm.html
Integrated listing guidance: http://www.epa.gov/owow/tmdl/2002wqma.pdf
Contact: Patti Tyler (ORD/RSL, Region 8)
Critical Ecosystems ORD/EPA workshop to be held in Denver in June 2002
Contact: Steve Ralph, USEPA, Region 10
Bauer, Stephen B. and S.C. Ralph. 1999. Aquatic habitat indicators and their
application to water quality objectives within the Clean Water Act. EPA-910-R-99-
014: http://www.epa.gov/regionlO/ or
http ://www. pocketwater. com/documents/ahi. pdf
Contact: Bill Swietlik (OW) and Chris Zabawa (OW)
National Sediment Workgroup in 1998
Draft Technical Framework to Support the Development of Water Quality Criteria
for Clean Sediment
Protocols for Developing Sediment TMDLs by OWOW
Nutrients
Contact: George Gibson (EPA/OW/OST)
Technical guidance manuals the program has published or is drafting for Streams &
Rivers, Lakes & Reservoirs, Estuaries & Coastal Marine Waters, and wetlands.
National Strategy for the Development of Regional Nutrient Criteria (July 1998)
A map of the continental United States (U.S.) illustrating Draft Aggregations of Level
III Ecoregions for the National Nutrient Strategy (ORD/NHEERL, Corvallis)
Regional Technical Advisory Groups (RTAGS)
EPA nutrient criteria web site
Contact: Emile Lores (ORD/NHEERL)
Submerged aquatic vegetation model
Biocriteria
Contact: Susan Jackson (OW/OW)
ALUS: Aquatic Life Use Support
Contact: Susan Cormier (ORD/ Cincinnati)
xiv
-------
US Environmental Protection Agency
Region/ORD Workshop on Aquatic Life Criteria December 4-7, 2001
ORD/NHEERL and OW/OST guidance document on data analysis methods used for
biocriteria development
Empirical statistical models - River Invertebrate Prediction and Classification System
(RIVPACS) and its derivative, AusRivAS (Australian Rivers Assessment System) are
empirical (statistical) models that predict the aquatic macroinvertebrate fauna that
would be expected to occur at a site in the absence of environmental stress.
Contact: Jeanne DiFranco (ieanne.l.difranco@state.me.us)
Dave.l.courtemanch@state.me.us
Biomonitoring retrospective, 15 year summary for Maine Rivers and Streams
Contact: Maggie Dutch mdut461 @ecy.wa. gov
Estuarine Bioindicator Development
Puget Sound Ambient Monitoring Program (PSAMP)
Contact: Laura Gabanski (OW/OWOW) and Teri Laidlaw (Region 8)
National Biocriteria Workshop - Training Clinic - December 2002 - Region 8
Contact: Frank McCormick (ORD/NERL-Cincinnati)
Biocriteria development
Ohio River Valley Water Sanitation Commission - www.orsanco.org
Kentucky Division of Water (Mike Compton)
Indiana Department of Environmental Management
Pennsylvania Fish & Boat Commission (Rick Spears)
West Virginia Department of Natural Resources (Dan Cancotta)
US Fish & Wildlife Service, Bloomington, Indiana (Tom Simon)
US Forest Service Southern Region (Mark Hudy, Mel Warren)
Contact: Lester Yuan (ORD/NCEA)
Progression of Ecological Degradation model
Toxic Chemical Session
Contact: Russ Erickson (ORD/NHEERL)
Ecological Risk Assessment Guidelines - http://www.epa.gov/ncea/ecorsk.htm
Contact: Chris Tatara and Tracy Collier (NMFS)
Research papers on toxic chemicals http://research.nwfsc.noaa.gov/ec/ecotox
Contact: Brian Thompson (OW/OST)
Endangered Species Act and Clean Water Act National Memoranda of Agreement
xv
-------
US Environmental Protection Agency
Region/ORD Workshop on Aquatic Life Criteria December 4-7, 2001
Contact: Foster Mayer (ORD/NHEERL)
Description of the Interspecies Correlation Estimation Software (ICE)
Interspecies correlations are developed using Model II least squares methodology for
linear regression (both variables are independent and subject to measurement error).
Slopes and intercepts are derived from the equation, logX2 = a + t^logXj), where Xt
equals the actual toxicity value for a surrogate test species (e.g., rainbow trout) and
X2 equals the estimated toxicity value for another species (e.g., listed species).
Species with paired tests for >5 chemicals are the minimum requirement for inclusion
in each analysis. When either one of the paired values include more than one EC or
LC50 value, the geometric mean is used. Levels of statistical significance of slopes in
all analyses are P<0.05 or <0.01. The Interspecies Correlation Estimation (ICE)
software is based on a Windowsฎ platform and includes two estimates of
uncertainty: 1) 95% CL for each individual model, and 2) 95% CL for uncertainty
due to surrogacy, based on pooled error mean squares for each species.
Contact: Phil Cook (ORD/NHEERL) and Patricia Cirone (EPA/Region 10)
Framework for the application of toxicity equivalence methodology to ecological risk
assessment for PCDDS, PCDFs, and dioxin like PCBs, in progress
Contact: Rick Bennett (ORD/NHEERL)
Great Lakes Water Quality Initiative (GLWQI).
Contact: Sue Norton (ORD/NCEA)
Report on empirical models for screening sediment contamination - Sometime 2002
xvi
-------
US Environmental Protection Agency
Region/ORD Workshop on Aquatic Life Criteria December 4-7, 2001
Welcome: Janis Hastings (Region 10) and William Farland (ORD)
Workshop Goals/Logistics: John Helvig (Region 7)
Workshop Structure: Patricia Cirone (Region 10), Meeting Facilitator
PLEASE NOTE: Slides from the Workshop presentations are available at:
http://intranet.epa.gov/ospintra/regsci/aquatic.htm
PLENARY SESSION: OVERVIEW OF AQUATIC LIFE CRITERIA
Co-chairs: Bob Spehar (ORD/NHEERL) and Patricia Cirone (Region 10)
This session consisted of brief overviews of the current science approach(es), scientific
application by the states and regions, and program office guidance. Opening remarks were made
by Patricia Cirone, who introduced Janis Hastings (Region 10) and William Farland (ORD).
Farland acknowledged the Regional Science Program, including funding for Regional Science
Liaison positions, and the Regional Applied Research Effort (RARE) which provides research
dollars to support short-term projects.
Toxic Chemicals
Programmatic Overview of Science - Charles Delos (OW/OST)
Methods for deriving an aquatic life criterion were described, in addition to aquatic life
methodology and the associated scientific issues. The intent of the methodology is to protect a
very high percentage of species. The methodology produces two criteria values: acute and
chronic; and protects an assemblage of species, including tested "important" species. There are
two criteria because ORD, in the late 1970s, was working primarily with two types of tests:
acute, and chronic. The methodology does not consider interaction between species. Effects on
different life stages are mixed together and treated as equivalent, as are effects involving
different endpoints (survival, growth, and reproduction).
Scientific issues related to aquatic life methodology were listed. When applied to the time-
variable exposures that occur in the real world, the mixing together of effects involving different
endpoints and life stages comes into question. Population modeling suggests that effects on
survival, growth, and reproduction are not equivalent, and should not be treated as such.
However, the work the program started about ten years ago to address this issue became mired in
modeling complexity and did not reach completion due to competing work priorities.
-------
US Environmental Protection Agency
Region/ORD Workshop on Aquatic Life Criteria December 4-7, 2001
Currently, there is no standard procedure for deriving the duration (the averaging period to be
applied to the criterion) and the allowable frequency for exceeding the criterion. These criteria
components are supposed to address the time variability that occurs in the real world. In spite of
being the least supportable facets of the criteria, duration and frequency have been the most
difficult components of the criteria to change because there are no agreed upon approaches for
deriving their values. Among the states the implementation procedures involving the criteria
averaging period and the allowable exceedance frequency provisions are not consistent.
Before the project slowed to a standstill, much of the Guidelines revision effort was applied
resolving time variability issues. The allowable exceedance frequency can be addressed via
population modeling considerations. Population modeling, however, leads into other interesting
issues, for example, regarding spatial variability and the importance of the size of the affected
area. It may be noted that ecological risk assessment within the Superfund program does address
the size of the affected area. Water quality criteria, on the other hand, do not generally address
spatial area.
Other issues raised in recent years relate to bioavailability, other toxic endpoints, aquatic dietary
exposure, and toxicant additivity. It was suggested that perhaps metals toxicity research is using
resources to an extent greater than the problem.
Two issues were further discussed, exposure through food chain contamination, and additivity of
effects from different toxicants.
The concern about exposure through food chain contamination stems from differences between
exposure in chronic toxicity tests and exposure in the real world. In chronic toxicity tests, test
organisms are fed uncontaminated food. In the real world, if the contamination affects a large
enough area for enough time, then the entire food chain may become contaminated, yielding an
additional pathway of exposure not included in the toxicity tests underlying the criteria. Some
have argued that metals criteria should be expressed as total recoverable metal to account for
this. Reverting to total recoverable criteria, however, does not correctly account for food
contamination. Rather it is necessary to recognize that the dietary concentration is linked to the
dissolved concentration through a partition coefficient or bioaccumulation factor. Toxicity is not
determined by the total Mg/L, but by the dissolved //g/L, after adjusting the criterion for the
additional exposure route through diet.
With regard to additivity of toxicity of several metals, an example was shown using New York
Harbor data. At the New York site, assuming additivity of acute toxicity, the total toxicity of six
metals was, on average, only ten percent greater than the largest single contributor. In this large
water body, receiving a wide variety of municipal and industrial contaminant inputs from a
heavily populated area, additivity was thus not a significant issue.
-------
US Environmental Protection Agency
Region/ORD Workshop on Aquatic Life Criteria December 4-7, 2001
Questions and Comments
Question: Is there any way to obtain acute-chronic ratios with the same species?
Response: Yes, it is done for the same species from the same lab and the same water.
Question: How do aquatic life criteria relate to biological criteria?
Response: [After some discussion] ... we really don't know how they relate.
Question: Can we predict bioavailability and synergy?
Response: Bioavailability, yes through chemical modeling using the Biotic Ligand Model, so
named because it treats the fish gill as another chemical ligand in the water. Metal
synergy, no we can't do that.
-------
US Environmental Protection Agency
Region/ORD Workshop on Aquatic Life Criteria December 4-7, 2001
Water Quality Toxics: Short- and Long-Term Needs - Debra L Denton (Region 9)
This talk covered the USEPA Regions list of short-term (within 3 years) and long-term (3-7
years) water quality toxics needs, our challenges, and take home points. A brief overview of the
1985 Guidelines (Guidelines for Deriving Numerical National Water Quality Criteria for the
Protection of Aquatic Organisms and Their Uses) was presented. The following are the EPA
Regions short- and long-term needs for the water quality toxics program.
Short-term Needs
Water Quality Criteria (WQC) are needed for the National Pollutant Discharge Elimination
System (NPDES) permitting and the Total Maximum Daily Load (TMDL) programs. A numeric
target needs to be identified to achieve the Water Quality Standards (WQS) which would be
equivalent to a numeric WQS. An indicator needs to be selected which is applicable to the water
body and local conditions. The scientific and technical validity are factors to consider. More
WQC need to be established for chemicals such as; diazinon, atrazine, pyrethroids, and total
mercury. There are different approaches (e.g., approaches to derive the criteria numbers) among
EPA programs which need to be resolved. The same test species and methods should be used
across the Agency to derive criteria. Approved 304(a) criteria for toxicity are needed. The
appropriateness of the Technical Support Document (TSD) numeric benchmarks (0.3 TUa and
1.0 TUc) needs to be evaluated, as well as the default acute-chronic ratio [(ACR) = 10)]. The
additive, synergistic, and antagonistic effects of various chemical interactions need to be
addressed. Understanding the toxic mode of action can be used to understand potential chemical
interactions, such as diazinon plus the pyrethroid, esfenvalerate (more than additive). Statistical
models and endpoints may need to be re-evaluated and enhanced, (e.g., the alternative point
estimate models and confidence intervals; the hypothesis testing approach; and bioequivalence
testing). With the advancement of enhanced statistical models, should confidence intervals and
different effect levels (LC10 for survival) be evaluated?
Long-term Needs
Threatened and endangered (T&E) species need to be considered by either using cold water
species to protect for warm water T&E fish, or by improving the overall test power. USFWS's
consultation must be on the criteria development approach, not the individual chemical numbers.
Additional species (e.g., amphibians) need to be included. Endpoints need to be sensitive,
precise, and at environmentally and ecologically relevant levels. Additional sublethal endpoints
need to be evaluated, such as swimming performance. Swimming, as a measure of performance
in fishes, is a key factor in linking an organism's phenotypic character with its environmental
resources for the overall reproductive output and survival of the individual and the population.
-------
US Environmental Protection Agency
Region/ORD Workshop on Aquatic Life Criteria December 4-7, 2001
Conclusion
WQS are needed to implement the permitting and TMDL programs. Clients' needs must be
considered; the USEPA Regions have both short-term (within 3 years) and long-term (3-7 years)
needs. EPA needs to examine lessons learned, successes, and technical transfer from programs
such as permitting and whole effluent toxicity to be used in the TMDL program. We need good
science plus the regulatory tools hand in hand. Training, technical transfer and feedback loop
mechanisms are needed. Maintaining open communication among ORD, OW, and the Regions
is of key importance.
Questions and Comments
Question: Strategically, what do you do when you need to select a numeric target for the TMDL
program?
Response: Based on the mandate from the Clean Water Act, I follow the 1985 water quality
guidelines to derive criteria.
-------
US Environmental Protection Agency
Region/ORD Workshop on Aquatic Life Criteria December 4-7, 2001
Habitat
Impaired Habitat: A Water Program Retrospective/Perspective - Douglas J
Norton (OW/OWOW)
A brief overview of the EPA habitat history was presented. While the EPA mission and
statutory goals broadly support habitat protection, the specific regulatory tools are limited. In
1992, the Habitat Cluster took on this issue and developed a strategy (unpublished). Six key
action areas were recommended: 1) improve the use of regulatory authorities, 2) focus on non-
regulatory programs, 3) improve the science base, 4) provide better habitat information
management, 5) form effective partnerships, and 6) set risk-based priorities. This presentation
revisited four of the six 1992 recommendations and, as an evaluation of our progress since then,
focused on the current TMDL program, new data and guidance, and future steps.
A framework for restoring impaired waters was outlined. In short, it consists of determining the
maximum load (TMDLs) and allocating the load reductions among point sources (PS) and
nonpoint sources (NFS). Point sources are controlled via NPDES permits and nonpoint sources
are managed via partnerships, grants, and voluntary programs.
Regarding the first Habitat Cluster recommendation, how well does the TMDL process fit
habitat protection and restoration? Habitats are being improved by some TMDLs, partly as
TMDLs use a watershed approach. The TMDL approach ("allocation-based" formula) is not a
good fit for some habitat impairments, e.g., invasive species and riparian habitat loss. NFS
pollution is the greatest threat to habitats, but EPA has little regulatory power relative to NFS
pollution. The top 12 impairments (some pollutants, some not) were listed from the 1998
Impaired Waters Lists [303(d)]. States are going beyond the Federal regulations by listing
reports of "habitat alteration" specifically. It is unclear what the impairment is for most of these
reports. The regulatory "fit" is not optimal; however, there appears to be potential for creative
use and state initiative to go farther for habitats.
With regard to the second Habitat Cluster recommendation, much about TMDLs is non-
regulatory due to lack of NFS authority. The watershed approach has benefitted habitat
(Watershed Restoration Action Strategies [WRAS]). The incentive approach, such as in the
CWA Section 319, has undergone enormous growth. Section 319, originally best management
practices (BMPs) for pollution control, has become an incubator for restoration and the
watershed approach, and now for TMDLs for NFS pollution.
Habitat Cluster recommendation four is the area of greatest advances in information technology
such as the TMDL tracking system; a National Geographic Information System (GIS)
-------
US Environmental Protection Agency
Region/ORD Workshop on Aquatic Life Criteria December 4-7, 2001
incorporating land cover, surface water, and other data; landscape ecology monitoring applied to
watershed ecosystems; and Watershed Assessment, Tracking & Environmental Results
(WATERS) database links to the afore mentioned. Relevant web site addresses, including two
new guidance documents, were provided:
TMDL Tracking System:
http://oaspub.epa.gov/waters/national_rept.control?p_cycle=1998
WATERS database: http://www.epa.gov/waters/
CALM guidance: http://www.epa.gov/owow/monitoring/calm.html
Integrated listing guidance: http://www.epa.gov/owow/tmdl/2002wqma.pdf
Habitat Cluster Recommendation Six looks to the future and considers priority setting. Two key
points of priority setting are: 1) target what is recoverable, and 2) focus on critical areas.
Targeting "fixable waters" produces results. There is greater recovery potential when working
on the just moderately impaired vs. the "hopeless." Less impaired sites often have better/faster
recovery and high biodiversity/habitat dividends. Critical areas, not just habitats, are of key
importance to the overall sustainability of the Nation's aquatic ecosystems and watersheds.
Critical area protection should be applied at the site scale (e.g., coral reefs, riparian zones,
wetlands, headwaters, T&E species habitats), and at the regional scale as a network. A regional
science workshop on critical ecosystems will be held in June 2002.
-------
US Environmental Protection Agency
Region/ORD Workshop on Aquatic Life Criteria December 4-7, 2001
Strengthening the Use of Aquatic Habitat Indicators in the Clean Water Act-
Steve Bauer (Pocket Water, Inc. - Idaho)
A dominant natural resource issue in the Pacific Northwest is reversing the decline of native
salmonids, attributed in part to the loss of freshwater fluvial habitats. The regulatory aspects of
the Clean Water Act (CWA) have the potential to assist in recovery efforts if habitat evaluation
and assessment are integrated into water quality programs in an effective way. Although
fisheries professionals have evaluated aquatic habitats for a long time, the transferability of these
methods to a regulatory program needs to be assessed before habitat variables can be used to
direct pollution control programs such as state water quality standards or Total Maximum Daily
Loads (TMDLs).
Commonly measured variables for freshwater streams, encompassing flow regime, habitat space,
channel structure, substrate quality, stream-bank and riparian condition, were evaluated based on
four primary characteristics: 1) relevance to the environmental endpoint, specifically salmonid
fish 2) applicability to the landscape and stream network in which they are used, 3)
responsiveness to human-caused stressors, and 4) the degree of measurement reliability and
precision. Integrating these indicators into the CWA is hampered by the state of the science as
well as the existing framework for water quality standards. There is general agreement on the
habitat requirements of salmonid fishes and the effect of non-point source activities on these
habitats. There is less certainty on quantifying the biological effect and on the reliability of
habitat assessment techniques, both of which are required elements for application to the CWA.
These obstacles can be overcome by applying the principles of landscape and stream network
classification to indicator development, identifying and quantifying reference area conditions at
the regional level, calibrating relevant indicators to specific locales, and developing systematic
monitoring procedures that meet rigorous data quality objectives.
Examples of aquatic habitat indicators were presented for the Salmon River Basin in central
Idaho and the Tongas National Forest in southeastern Alaska. These examples are provided as
Appendices in the EPA document cited below.
This presentation is based on the publication completed for EPA Region 10, Seattle, Washington
and an article published in Fisheries. The citations and availability are listed below:
Bauer, Stephen B. and Stephen C. Ralph. 1999. Aquatic habitat indicators and their application
to water quality objectives within the Clean Water Act. EPA-910-R-99-014. US Environmental
Protection Agency, Region 10, Seattle, Wa. Available at the EPA Region 10 library website:
http://www.epa.gov/region 10/ publications, or http://www.pocketwater.com/documents/ahi.pdf
-------
US Environmental Protection Agency
Region/ORD Workshop on Aquatic Life Criteria December 4-7, 2001
Bauer, Stephen B. and Stephen C. Ralph. 2001. Strengthening the use of aquatic habitat
indicators in Clean Water Act Programs. Fisheries, vol. 26, no. 6, 14-25
Questions and Comments
Question: Is habitat an integrator?
Response: I say, no, in the sense that biota is an integrator vs. habitat.
Comment: I think habitat is really important. Aquatic biota is the endpoint.
-------
US Environmental Protection Agency
Region/ORD Workshop on Aquatic Life Criteria December 4-7, 2001
The ORD/NHEERL Approach to Habitat Alteration Research - James H Power
(ORD/NHEERL)
The National Health and Environmental Effects Research Laboratory (NHEERL) began by
developing an Aquatic Stressors Framework (GPRA Goal 2). Framework objectives were listed
in a memo from the NHEERL Director. Five NHEERL workgroups were formed to address: 1)
habitat alteration, 2) suspended and bedded sediments, 3) nutrients, 4) toxics, and 5) diagnostics.
Members of the Habitat Alteration Workgroup were listed: there are two members from each of
the four Ecology Divisions within NHEERL. The workgroup problem statement was presented.
The focus will be on habitat, because habitat alteration may cause a water body to not meet its
designated use. EPA has mandated that research focus on fish, shellfish, and wildlife, and that it
address the life-support functions of habitat.
Annual Planning Goals (APGs) for FY02-FY08 were summarized. The APGs are intended to
scope out the problem, because in some cases the multi-year plan implies a shift of research
focus. The emphasis on fish, shellfish, and wildlife required an initial identification of species
endpoints. Selected species endpoints were listed for each Ecology Division. Goals were
presented for four research areas: 1) Coastal Vegetated Habitat research, 2) Shoreline, Lake, and
Estuary-scale Habitat research, 3) Landscape-scale Habitat research (for salmon and native fish
habitat), and 4) Landscape-scale Habitat research (for mercury risk to common loons and other
piscivorous birds) (see slides 11-14). NHEERL evaluated gaps in the research and recognized
that they cannot address all the gaps. A limited list of some of the gaps identified was provided
(see slide 15). There are currently 27.3 FTEs in NHEERL devoted to Altered Habitat research.
This is not a very large number of resources, which is why the research focus is on species
endpoints. In regard to EPA Plan Reviews, NHEERL expects to be collaborating with other
Federal and state agencies, as well as non-governmental organizations; however, their travel
budget is limited.
The next steps for the Aquatic Stressors Plan include: peer-review of the Plan by external
scientists; development/formalization of contributions by the Ecology Divisions; development of
cross-divisional proposals and Quality Assurance Project Plans (QAPPs) following Plan
guidance and incorporation of internal and external review comments; peer-review of research
proposals; and implementation of the Plan.
Questions and Comments
Question: The development of these plans historically are EPA problems; did you integrate with
the program offices?
10
-------
US Environmental Protection Agency
Region/ORD Workshop on Aquatic Life Criteria December 4-7, 2001
Response: Not to the extent I would have done internally for speed although several of us on the
planning committee had the opportunity to participate in the joint Office of Water -
ORD Science Planning and Research Coordination meeting while we were preparing
our document.
11
-------
US Environmental Protection Agency
Region/ORD Workshop on Aquatic Life Criteria December 4-7, 2001
Sediments
Suspended and Embedded Sediments: Status Report and Update from the
Office of Water - Susan K. Jackson (OW/OST)
A status report and update was presented on EPA activities related to suspended and embedded
sediments from the perspective of the Office of Water (OW). For additional information,
contact either Bill Swietlik (OW) or Chris Zabawa (OW), the two primary scientists working on
sediments in the Office of Water. Suspended and embedded solids refer to: "clean sediments",
total suspended solids (TSS), excessive or too little sedimentation, bed deposits, transported
bedloads, and water column turbidity or cloudiness. Typically, OW does not mean sediments
that have been contaminated with chemicals, thereby making them toxic. EPA has been working
on chemical criteria for sediments to address this problem for some time. Suspended and
embedded sediments include: particulate organic and inorganic matter that suspends in or is
carried by the water, and/or accumulates in a loose, unconsolidated form on the bottom of
natural water bodies.
Suspended and embedded sediments problems refer to imbalances in sediment in aquatic
ecosystems. Too much sedimentation is a common, major water quality problem. As with other
water body features such as nutrients, dissolved oxygen, flow, alkalinity, carbon, etc., all water
bodies need and depend on certain natural levels and fluctuations of sediment. Too little
sediment can be as much of a problem as too much. Therefore, one of the basic objectives in
managing suspended and embedded sediments in water bodies is to ensure that a water body is
getting the right amount of sediment given the natural characteristics and management objectives
for the water body. To achieve this, a better understanding is needed of the natural or optimum
regime of sediment a water body should have, and of the effects variations from this level will
have on an intended use of that water body.
Some of the more common impacts of imbalances in suspended and embedded sediments were
listed (see slide 6). The complexity of the issue was highlighted. Protecting aquatic life and
wildlife, enhancing recreation, and protecting drinking water sources may all have different
optimum suspended and embedded sediments regimes at which these different water body uses
will be protected. In addition, natural variability exists in the suspended and embedded sediment
regimes different water bodies have based on natural classification schemes.
Sediment is the single largest stressor identified as the cause of impairment for 303(d) listed
waters. On a water body basis, using data from the 305(b) reports also submitted to EPA by the
states, the most highly impacted water bodies, in priority order are: rivers and streams, followed
by lakes, reservoirs, ponds, and estuaries. States and tribes are faced with controlling excessive
12
-------
US Environmental Protection Agency
Region/ORD Workshop on Aquatic Life Criteria December 4-7, 2001
sediment in water bodies without having the necessary tools to do so, including good water
quality criteria, monitoring methods, modeling techniques and control measures. There is a
disconnect between the immediate needs facing states and tribes meeting court ordered TMDLs,
and the future development of tools. EPA's programmatic solution is to provide what is needed,
as soon as possible, starting with the highest priority needs. Collaboration is needed between the
regions, states, and tribes to achieve that end.
Existing EPA water quality criteria were developed in 1976 and are listed in EPA's Quality
Criteria for Water, known as the Gold Book (1986). Very few states are using this value in their
TMDL effort, and are instead using some form of a turbidity level (e.g., nephelometric turbidity
units) or TSS as criteria in their standards. Of highest priority is the development of effective
water quality criteria by EPA (OWOW, OST, ORD/NHEERL, NCEA, NERL, NRMRL) along
with other Federal partners (USD A-Agricultural Research Service). Most of the research
necessary for improved water quality criteria development will be conducted by ORD-NHEERL.
Past and present suspended and embedded sediments efforts were listed. Highlights include the
formation of the National Sediment Workgroup in 1998, the development of the Draft Technical
Framework to Support the Development of Water Quality Criteria for Clean Sediment, the
development of the Strategic Planning and Research Coordination (SPRC) Action Plan listing
all identified research needs, publication of a guidance manual on Protocols for Developing
Sediment TMDLs by OWOW, and preparation for conducting research necessary for developing
better water quality criteria for suspended and embedded sediment under the Aquatic Stressors
Research Program (ORD/NHEERL). ORD/NHEERL has prepared a research strategy to
address five key research questions in preparation for developing suspended and embedded
sediment criteria (see slide 13). Additional programmatic work under consideration by OW
includes: developing a sediment web site for information exchange, providing technical
assistance to the states/tribes, holding a National sediment workshop, publishing guidance on
monitoring methods for sediment, guidance on permitting to control sediment, developing a
database on control effectiveness (point sources and BMPs), and holding informational
meetings.
Questions and Comments
Question: When will we have a narrative standard? We need to develop a collaborative effort
between regions, states, ORD and OW.
Response: Sediments is an area that ORD is focusing on.
Comment: Comment and discussion on the need for a workshop on managing an interim
approach while sediment criteria are being developed. Additionally, there was
discussion on the benefit of having a regional lead identified and charged with the
task of coordinating OW program and ORD research efforts with the regions on
13
-------
US Environmental Protection Agency
Region/ORD Workshop on Aquatic Life Criteria December 4-7, 2001
sediment criteria. A regional program expert on detail or a dedicated FTE modeled
after the Wetlands Program/ORD coordinator based in the Corvallis Laboratory were
suggested.
14
-------
US Environmental Protection Agency
Region/ORD Workshop on Aquatic Life Criteria December 4-7, 2001
Suspended Solids and Sediments Risk Management Research - Christopher T
Nietch (ORD/NRMRL)
Traditionally, sediment management and assessment has been approached from specific land-use
contexts. For example, much more attention has been given to erosion processes in agricultural
lands compared to suburban and urban lands. As a result, many of the management measures
used are not well tested in different land use classes. At issue is linking the anthropogenic
causes and ecological effects of sediment stress. Conceptual linkages between anthropogenic
sources of sediment stress and ecosystem effects were shown in a flow diagram, as well as
conceptual linkages among anthropogenic causes of sediment stress and aquatic ecosystem
health.
Two primary processes of concern are hillslope erosion and channel erosion. Channel erosion
has historically been studied within a geomorphic context and theoretically is related to balances
between sediment supply and discharge. Presently, channel erosion is addressed within the
context of habitat disturbance. It is necessary to manage both flow and hillslope sediment
supply, which may change disproportionately in space and over time as a watershed undergoes
land-use change. Controlling hillslope erosion may not solve the problem, and may in fact
exacerbate it, if receiving waters are inadvertently starved of their natural sediment source. In
these instances, the channel erodes to fill the unused sediment carrying capacity of the flow.
The framework was presented for ORD-wide sediment work and the NRMRL approach to
conducting the type of risk management research necessary to make scientific decisions with
respect to the control of sediment stress at the watershed scale. Quantitative steps involved in
making risk management decisions and assessing if they are effective include: 1) allocating
sources (or lack thereof), and determining how they behave relevant to the sediment reference;
2) determining the Best Management Practice (BMP); 3) linking source allocation and BMP
over multiple scales for Decision Support System (DSS); and 4) evaluating the decision with
effective monitoring programs.
Research called for relies heavily on modeling. The relative accuracy and reliability of models
of diffuse pollution was shown in a graph. NRMRL feels this is justified, based on the low
levels of uncertainty for models of flow and sediment compared to other aquatic ecosystem
stressors. Several entities within ORD are involved in source allocation modeling because of the
direct crosswalk to sediment TMDL preparation and regulation. Specifically with respect to
sediments, research that focuses on source allocation models needs to address the following:
Hillslope sediment yield models,
Hydrologic load models,
15
-------
US Environmental Protection Agency
Region/ORD Workshop on Aquatic Life Criteria December 4-7, 2001
In-channel sediment fate and transport models, and
Developing packages that link the loading and receiving water models in space and
time to allocate sources in mixed land use watersheds.
A figure showing a prototype watershed of mixed land use was used to demonstrate the spatial
complexity involved in developing appropriate load allocation models. The complexity of
sediment fate and transport models depends on the nature of the simulated waterbody (e.g., a one
dimensional model for a 1st or 2nd order stream may not be appropriate for a large and deep body
of water such as a reservoir). Concurrently, much effort is being placed toward providing
scientifically sound structural and non-structural BMP alternatives. Examples of BMP
alternatives with recommended designs are ponds and wetlands. Other types of BMPs that are
given considerable attention are buffers, dry detention, and bioretention. Recently, dry detention
ponds have been reported as not working well for reducing SSAS. Low impact development
options / low-structural BMPs include grassed swales, silt fences, porous pavement, elimination
of curbs and gutters, and disconnection of roof top runoff. A diagram was presented depicting
the processes that need consideration in a BMP performance model for sediment removal (see
slide 14). Land-use, treatment, development, and acquisition predicates that BMPs be
implemented on the field scale, which may vary between one and several hundred acres.
BMP performance modeling can be used to make sediment management decisions at the field
scale. First a problem is identified, then a model is used to guide design and predict the results
in an iterative process. This is the same approach put forth by the USDA to guide stream
restoration projects. Two fundamental questions that need to be addressed at larger scales to
make the BMP approach useful for the protection or restoration of aquatic ecosystem health are:
Where to place an individual BMP for the most water quality benefit?, and
How to model the cumulative effects of BMPs at a larger scale?
These questions are addressed within the context of a watershed-scale decision support system.
Major issues and requisites for the decision support system (DSS) were presented along with a
conceptual diagram of a simulation model for sediment management. An example of what the
output of such a system may look like was illustrated. Monitoring drives source allocation
model development and is also essential for determining the effectiveness of the management
decisions in the post-implementation phase. With effective monitoring programs in place, each
time a management decision is made, the result becomes a case study that tests the utility of the
DSS. A flow diagram was described illustrating the framework for using a decision support
system in an adaptive management context from integration of source allocation and BMP
performance models to the DSS to the evaluation of that decision through effective monitoring
programs.
16
-------
US Environmental Protection Agency
Region/ORD Workshop on Aquatic Life Criteria December 4-7, 2001
Nutrients
USEPA National Nutrient Criteria Program Approach to Reference
Condition Development - George Gibson (OW/OST)
Eutrophication of surface waters in the United States has been recognized as a long standing
problem. Much of the Nation's waters do not adequately support aquatic life because of excess
nutrients. Nutrient levels vary from region to region due to geographical variations in geology
and soil type. EPA is publishing recommended National nutrient criteria for fourteen nutrient
ecoregions by water body type including lakes and reservoirs, streams and rivers, estuarine and
coastal marine waters, and wetlands. The criteria can be used for: identification of problems,
standards development, regulatory assessments, project planning and evaluation, and
determination of resource status and trends.
Illustration Of Nutrient Criteria Development Using Lakes As An Example
A critical part of the nutrient criteria process is establishing lake reference conditions for the
physical lake classes within each ecoregion. Reference conditions for each of these classes are
the quantitative descriptions of lake conditions used as a benchmark for comparison purposes.
Realistically, reference conditions represent the least impacted conditions, or what is considered
to be the most attainable approximation of natural conditions; land use practices and atmospheric
pollution have so altered the landscape and quality of water resources that truly undisturbed
lakes are rarely available. When establishing reference conditions, identify those lakes of a
given physical class which are least impacted by human behavior, then measure and assume
nutrient levels represent normal levels for a natural setting for lakes of that type. Reference
lakes must be representative of a region and their conditions should represent the best range of
minimally impacted conditions that can be expected of similar lakes within a region.
An outline was provided of the format of the technical guidance manuals the program has
published or is drafting for Streams & Rivers, Lakes & Reservoirs, Estuaries & Coastal Marine
Waters, and wetlands. Preparation of these documents relied on the National Strategy for the
Development of Regional Nutrient Criteria document (released July 1998), in addition to
information developed from a 1995 nutrient technical conference of experts held by the program
as well as from earlier experience developing the national biocriteria program and consultations
with the USEPA Science Advisory Board. It is expected that the technical manuals for the
development of nutrient criteria will be updated approximately every five years. The ORD
National Health and Environmental Effects Research Laboratory (NHEERL) in Corvallis,
Oregon has created a map of the continental United States (U.S.) illustrating Draft Aggregations
17
-------
US Environmental Protection Agency
Region/ORD Workshop on Aquatic Life Criteria December 4-7, 2001
of Level III Ecoregions for the National Nutrient Strategy, which was presented to illustrate the
initial regionalization approach for nutrient criteria development.
Regionalized reference condition based criteria development is a key aspect of cultural over-
enrichment management for water quality protection and management. Elements of criteria
development were listed, and are described in the National Nutrient Strategy and the technical
guidance manuals. It was noted that initial data used for the freshwater systems criteria
development dated back approximately ten years (1990 to present), and did not draw on
STORET database randomly; the data were first screened for the removal of reports on
obviously degraded water quality . The headquarters nutrient team and Regional Technical
Advisory Groups (RTAGs) evaluated total phosphorus, total nitrogen, nitrite and nitrate, total
Kjeldahl nitrogen, chlorophyll a, dissolved oxygen, and Secchi depth, to determine regional
reference condition values (see also the EPA nutrient criteria web site).
A regional reference condition can be selected using approaches. In both instances, the intention
is to select an optimal reference condition value from the distribution of an available set of lake
data for a given physical class of lakes or reservoirs. The preferred approach is to select the
upper quartile from the distribution of measured variables of known reference lakes. The other
approach is to select the lower quartile from all lakes in the class or from a random sample
distribution of all lakes in the class. A diagram illustrating the two approaches was presented
using total phosphorus as the example together with independent demonstration studies by states
and agencies reflecting the apparent similarity of the values derived from this approach,
however, further investigation is required [See also "Use of Reference Sites and Conditions in
the Development of Nutrient Criteria" below]. The framework used for developing nutrient
criteria was presented as a flow diagram.
EPA expects that the states and tribes will make a good faith effort to adopt and incorporate the
nutrient criteria recommendations into water quality standards. Consequently, States and tribes
will have flexibility in the development of nutrient criteria in the following areas: measured
variables, use of EPA technical guidance, designated uses, antidegradation, reference condition
versus criterion, the role of outliers, and use attainability assessments and site specific
evaluations.
18
-------
US Environmental Protection Agency
Region/ORD Workshop on Aquatic Life Criteria December 4-7, 2001
Nutrient Criteria: Challenges Facing Regions and States - Danielle Tillman
(Region 5)
Options for Criteria Development
The regions and states have been given a lot of flexibility in developing nutrient criteria. The
three options for states and authorized tribes are the following: 1) to adopt EPA's 304(a) criteria
recommendations; 2) to develop criteria to reflect local conditions and protect specific
designated uses, using EPA's guidance (recommended option); and 3) to develop criteria
protective of designated uses using other scientifically defensible methods and appropriate data.
Most states are not considering using option 1. In Region 5, all states are pursuing some form of
option 2, either refining the reference condition (option 2a) or establishing effect thresholds
(option 2b) or developing some combination of the two.
Ongoing Data Collection and Analyses
The New York State Department of Environmental Conservation, the Wisconsin Department of
Natural Resources (and the U.S. Geological Survey), and Indiana University are among those
groups collecting and analyzing additional data to support nutrient criteria development. They
each have ongoing projects aimed at identifying effect thresholds appropriate for protecting
designated uses.
State Questions and Needs
States are looking for answers to science, policy, and resource questions such as how aquatic
ecosystems function, and what parameter values are expected in minimally-impacted sites. EPA
is trying to reassure states that there is flexibility; however, states are still leery about coming up
with a number dramatically different from existing, published EPA numbers. States want clear
and timely answers to policy issues, and they are also seeking additional resources. States need
continued access to experts for study design and data analysis advice. They are looking to EPA
to help them address water bodies that are shared by multiple jurisdictions, and water bodies that
have, so far, been "falling through the cracks."
Conclusions
States and tribes are interested in refining EPA's nutrient criteria; many want to use effect-
threshold information in their criteria development. State, tribal, Federal, and academic groups
are collecting and analyzing data to address some of the questions, but research gaps remain.
EPA research should coordinate closely with and integrate ongoing efforts to ensure that
19
-------
US Environmental Protection Agency
Region/ORD Workshop on Aquatic Life Criteria December 4-7, 2001
resources are targeted as effectively as possible. Finally, EPA scientists can contribute both by
conducting research and by providing technical advice (e.g., via Regional Technical Assistance
Groups).
20
-------
US Environmental Protection Agency
Region/ORD Workshop on Aquatic Life Criteria December 4-7, 2001
Biocriteria
National Framework for Tiered Aquatic Life Uses in State and Tribal Water
Quality Standards - Update on Guidance Development - Susan K Jackson
(OW/OST)
Since the early 1990s, states and tribes have used biological assessments to refine their water
quality standards; this is a long-term goal involving milestones and incremental steps. An EPA
workgroup is in place: 1) to develop guidance on the use of biological assessments and criteria to
refine designated aquatic life uses in water quality standards; 2) to propose how to apply existing
state and tribal water quality standard programs; 3) to identify pitfalls and barriers to
implementation; and 4) to solve problems and propose solutions. Technical guidance documents
have been produced for biocriteria, but not for the uses part of the standard. It is now a program
priority to focus on the "uses" because of issues of National consistency, as well as the need for
better public communication and engagement in decision-making.
The workgroup is developing a conceptual model (framework) predicting the biological response
to increasing human disturbance. This biological condition gradient should: 1) encompass the
entire range of possible conditions (natural and departure from natural all the way to "dead");
and 2) include scientifically defensible (and ecologically meaningful) increments of change, or
benchmarks, along the gradient. The conceptual framework will be used to support
establishment of thresholds for different levels of protection for aquatic life.
The designation of aquatic life uses is a public process, which takes into consideration the
existing condition of a waterbody, the potential to achieve higher water quality (restoration
potential), and social and economic factors. By using biological assessments to provide more
accurate and direct descriptions of aquatic life goals, public dialogue and decision-making are
enhanced. The workgroup is developing a tiered aquatic life use conceptual model of draft
biological tiers; the draft biological condition gradient, illustrating the general concept of tiers,
was presented along with thumbnail definitions of the six tiers. The workgroup is defining
specific attributes for the tiers, and working with the regions to test.
The purpose of the National framework is to provide an ecologically-based model for the
communication of biological conditions. The descriptive, incremental gradient should improve
communication about where goals may be set, and enable a better, more accurate understanding
of current conditions. Key points to emphasize: the framework is conceptual, the number of
tiers will be determined by states and tribes (the six tiers are meant to provide a highly resolved
gradient to assist states and tribes to think about how to characterize the condition and goals for
21
-------
US Environmental Protection Agency
Region/ORD Workshop on Aquatic Life Criteria December 4-7, 2001
their waters), "best fit" of data approach is recommended (the list of attributes should not be
considered a checklist), and the framework is applicable to different methods.
A third workgroup meeting is planned for next spring. The current draft will be used as a
discussion document to engage a broader audience, including internal EPA discussions,
additional states, and other Federal agencies and stakeholders. Proposed breakout session topics
and organization were listed.
22
-------
US Environmental Protection Agency
Region/ORD Workshop on Aquatic Life Criteria December 4-7, 2001
Biological Assessments in Region 10 - Approaches, Application and Research
Needs - Gretchen Hay slip (Region 10)
A brief summary was presented of the use of biological assessments in Region 10, including
approaches, and application and research needs. All states in Region 10 use bioassessments in
their water quality management programs. In addition, some tribes use bioassessments in their
water management programs. Most states have some form of "tiered" aquatic life uses and most
states have some sort of narrative biocriteria. No state or tribe in Region 10 has numeric
biocriteria, although Oregon is very close. Idaho's Waterbody Assessment Guidance was
provided as an example of bioassessment use in water quality management programs.
Additional examples of aquatic life uses were listed for Idaho, Oregon, and Alaska. Numeric
biocriteria development in Oregon was outlined. Issues currently being addressed are reference
site selection and use, and beneficial use categories. Biological assessment research needs in
Region 10 include: application of existing tools to the diverse conditions in Alaska, stressor
indicator indexes (e.g., sediment, nutrients and metals), a small set of quantitative CWA habitat
indicators (that are repeatable, reliable and widely applicable), and research for other waterbody
types (e.g., large rivers, lakes, and wetlands).
23
-------
US Environmental Protection Agency
Region/ORD Workshop on Aquatic Life Criteria December 4-7, 2001
THIS PAGE INTENTIONALLY LEFT BLANK
24
-------
US Environmental Protection Agency
Region/ORD Workshop on Aquatic Life Criteria December 4-7, 2001
BIOCRITERIA AND NUTRIENTS SESSION
Establishing Multi-Use Reference Sites for Biological & Nutrient Criteria
Development - Don Huggins (University of Kansas)
Outcomes of the recent USEPA Region 7 Central Plains Biocriteria Workgroup, and Nutrient
Regional Technical Advisory Group (RTAG) meeting on establishing multi-use reference sites
for biological and nutrient criteria development were presented. The Central Plains within
Region 7 is a diverse area ranging from the Sand Hills in the northeast to the Western High
Plains of Kansas. Many of the streams in the Central Plains have sandy bottoms, and can
become more diverse where there are man-made activities present. Reference condition criteria
were described for given ecoregions in the Central Plains. The Workgroup identified factors for
designation of reference sites, then selected eleven core factor groups that would be accepted
across the region. The group wanted to work toward a rule-based approach.
Characteristics of several core factors under reference conditions were described in more detail
including: instream habitat, riparian habitat, land use / land cover (site-specific), and biological
metrics (not a stand-alone factor). The Workgroup applied the use of reference sites / systems to
define biological and nutrient criteria values to Region 7. Reference sites are proposed for use in
identifying both biological and nutrient conditions that would lead to determination of
biological and nutrient criteria values. Broadly defined and quantified reference conditions
should identify high quality sites or systems that possess minimally altered physical, chemical
and biological states. Reference sites or systems exhibiting high quality biological systems
should be indicative of acceptable and above average water and habitat quality. Approximately
120 reference lakes were identified by regional biologists and experts. The tri-section method
was applied using chlorophyll a values as biological indicators to select potential reference lakes.
Comparisons of total phosphorus, total nitrogen and chlorophyll a for all lakes versus reference
lakes and for all lakes except Sand Hill lakes were illustrated. All values were expressed in
|ig/L. The Sand Hill lakes were not included in the data since they represent a geographically
unique population of least impacted lakes in the region. Based on these results the group was
able to propose benchmarks for lakes and reservoirs. Another example of a joint reference use
approach involved assessment of a total of 787 streams. Reference streams reflected both good
fish IBI scores and low nutrient levels but there was a slight drop in nutrient levels when the fish
IBI was selected as the single reference value of concern. Overall reference conditions were
thought to be the best measure of health rather than a single factor score, such as low phosphorus
or highest fish IBI.
25
-------
US Environmental Protection Agency
Region/ORD Workshop on Aquatic Life Criteria December 4-7, 2001
Questions and Comments
Question: How do you establish a reference site for large rivers (e.g., the Snake River)?
Response: In Region 7, we stay with wadable streams because we are not sure how to go about it
for large rivers.
Comment: The Ohio River Valley has been developing methods, and have established criteria.
They have used the minimal impact fish assembled - difficult to model with 165 fish
species. Minimal distances from discharges is what they shoot for.
26
-------
US Environmental Protection Agency
Region/ORD Workshop on Aquatic Life Criteria December 4-7, 2001
Reference Condition for Biological Integrity - Phil Larsen (ORD/NHEERL)
Larsen explained that our basic interest, regardless of time in history, is: what effect does human
activity have on aquatic ecosystems. One way of getting at this is to characterize conditions
under no or minimal human disturbance, and use that as a reference condition against which to
make future comparisons. Context can be time-independent. The general sense is that the term
"natural" connotes the condition in the absence of human disturbance, even though arguments
have been advanced that humans are a part of nature, therefore natural always implies human
effects. The use of "natural" as a conceptual tool to stand for no human disturbance / effects is
useful to frame discussions. At present, no place on earth can be considered without human
disturbance given the broad dispersion of human derived pollutants. Consequently the phrase
"minimal human disturbance" is useful for approximating a natural condition.
The reference condition for biological integrity was evaluated and characterized for regions
where minimally disturbed reference sites still exist, as well as regions where they do not. Some
issues currently being addressed are: how human activities affect aquatic biota, "natural" versus
minimum human activity, and the status of the reference condition for biointegrity under
minimal contemporary human disturbance. The need for reference conditions was described for
the Environmental Monitoring Assessment Program (EMAP), biocriteria, and watershed
management. Natural variability produces a range or distribution of reference condition scores,
i.e., reference variability, and can be accounted for through associations (e.g., ecoregion,
waterbody size, elevation, and gradient). The concept of reference variability was illustrated in
several graphs. Particularly challenging is the ability to characterize the reference condition in
regions where minimally disturbed reference sites no longer exist. In these regions, a
combination of methods are implemented such as historical reconstruction from times with
minimal stress, use of best ecological judgement (including models), restoration experiments,
and inference from data distributions. A graph was shown modeling reference condition in
heavily disturbed regions using fish data to extrapolate to the Y-axis.
The distinction between "minimal" versus "least disturbed" was described. "Minimal" implies
an absolute (e.g., sites should meet a set of criteria), while "least" implies relative disturbance
(e.g., any region contains a set of sites that are least disturbed, even if there are no minimally
disturbed sites). Of key importance is the selection of reference sites so the outcome is
repeatable. Two sets of criteria were independently applied to sites in an EMAP pilot study in
Region 3 (Appalachia). The resultant description of reference sites was the same in one
subregion but differed in a second subregion demonstrating the need for care in the selection of
criteria by which to identify reference sites. Reference sites should be representative of the
natural gradients in the region of interest, e.g., elevation, latitude, longitude, and stream gradient.
Slides depicting the sample population and reference site gradients were presented to illustrate
27
-------
US Environmental Protection Agency
Region/ORD Workshop on Aquatic Life Criteria December 4-7, 2001
whether reference sites were in fact representative of the target population. Challenges
regarding reference conditions include: assuring that the outcome of characterizing a reference
condition is repeatable; assuring that the reference condition is representative of the resource of
interest; standardizing reference site selection criteria; accounting for natural variability (how to
and how much); identifying alternatives to reference sites, especially for regions with few or no
minimally disturbed sites; and using appropriate classification scales.
Questions and Responses
Question: [Question regarding use of historical reconstruction methods to characterize the
reference condition]
Response: In the Northwest, the best we have available are land surveys dating back 150 years.
In Region 7, there are some good historical data.
28
-------
US Environmental Protection Agency
Region/ORD Workshop on Aquatic Life Criteria December 4-7, 2001
The Use of Reference Condition in Support of Surface Water Assessments and
Criteria Development in Ohio - Chris O. Yoder (Midwest Biodiversity Institute,
Columbus, OH)
[Slide 1] The Midwest Biodiversity Institute, Inc., incorporated in 1997, is an umbrella
organization currently consisting of 124 institutions and corporations and its primary affiliate,
the Ohio Biological Survey. The membership of MBI and OBS, encompassing 12 states and the
Province of Ontario, overlap to a great degree, with institutions and corporations from outside of
Ohio joining MBI and becoming affiliates of OBS. MBI is a 501(c)(3) nonprofit corporation,
which is governed by a Board of Trustees. An Advisory Board, comprised of designated
representatives from member institutions, provides input to the Board of Trustees. The goals
and objectives of MBI differ somewhat from OBS in that an additional emphasis is to leverage
field-trained taxonomists and ecologists back into college and university departments and
curators into museum positions. MBI also seeks funding to provide undergraduate and graduate
students with the financial means to pursue degrees in field-oriented, organismal biology.
Projects undertaken by MBI, which fall under the same goals and objectives as OBS, tend to be
outside of Ohio and multi-state in nature. In other regards, objectives, and philosophy, MBI
parallels OBS except in its greater geographic extent. MBI is presently executing a cooperative
agreement with U.S. EPA in support of biological criteria implementation and environmental
indicators development. MBI has also provided expertise via consulting agreements to a number
of organizations.
[Slide 2] Establishing reference condition is done by selecting and sampling reference sites
which are defined as a collection of sites within a homogenous regional area which represent the
best attainable conditions ("least impacted") for all waters with similar physical dimensions and
attributes for that particular region. [Slide 3] Reference sites differ from the traditional control
site concept in that multiple sites define reference condition within a regional context. A control
site is a single site usually located on or adjacent to the waterbody under study that represents the
best or most appropriate condition for that waterbody whether it is impaired or unimpaired.
Reference sites will always function as control sites, but some control sites may be unsuitable for
determining reference condition. Reference condition is updated every decade by the regular
and continued monitoring of reference sites. [Slide 4] Guidelines for the selection of reference
sites are needed to ensure that reference condition reflects least impacted conditions. These
guidelines are presently based on qualitative guidelines. Keeping the selection process apart
from the data analysis process is an important procedural issue as inappropriate mingling of the
two processes may introduce unintentional bias into the development of reference condition.
[Slide 5] Misperceptions of reference condition include the notion that reference sites reflect a
pristine or unimpacted condition, free from the effects of human disturbance. In reality, few if
any reference sites in the conterminous U.S. reflect a truly pristine condition with most
29
-------
US Environmental Protection Agency
Region/ORD Workshop on Aquatic Life Criteria December 4-7, 2001
representing substantially altered conditions, the results of decades of human settlement and
resource exploitation. This leaves us with the significant challenge of setting realistically
attainable endpoints that meet CWA goals (biological integrity, propagation, fishable/swimable).
[Slide 6] Biological criteria are narrative and numerical ratings based on the numbers and kinds
of aquatic organisms (i.e., the assemblage) that inhabit a sampling site. [Slide 7] Biocriteria are
indexed to the reference assemblage within a geographic region and with respect to the size
properties of the aquatic ecotype. [Slide 8] As such, biocriteria represent a calibrated
assessment tool that fosters an organized goal setting process to reconcile human impacts and
guide restoration efforts. [Slide 9] The calibration and derivation of the Ohio EPA Index of
Biotic Integrity (IBI) was described in six steps: 1) selection and sampling of reference sites, 2)
calibration of IB I metrics, 3) a calibrated IBI modified for Ohio waters, 4) establish ecoregional
patterns and expectations, 5) codify numeric biocriteria in the Ohio WQS), and 6) numeric
biocriteria are used in assessments. [Slide 10] The calibration of IBI metrics was detailed and
includes the calibration vector (e.g., stream size), determination of the 95% "maximum" richness
lines, and the trisection procedure for determining IBI metric scoring. This method reduces the
influence of substantially altered sites and keys on the upper 5% of reference condition. [Slide
11] The effect of increased numbers of reference site samples on IBI biocriteria derivation was
depicted to illustrate the minimum number of reference sites that are needed to adequately
represent the central tendency of reference within a region.
[Slides 12 and 13] A descriptive biological condition axis provides a common basis for
communication about a gradient of biological degradation and tiers around which stratified
designated uses can be set. [Slides 13 and 14] The position of designated aquatic life uses in
Ohio along the biological conditions axis was shown. National application of the biological
condition axis will provide for the following:
Translate different analytical approaches to a common yardstick;
Provide a common basis to derive bioassessments from different methods;
Enable consistent reporting on biological condition;
Communicate ecological relevance of sampling results to the public; and
Enable management within smaller increments of biological change, i.e., to protect for
the best attainable conditions and document incremental improvements as restoration
occurs.
[Slide 15] The reference dataset anchors the biological condition axis in the range of least
impacted reference condition. This step merges the calibration of the bioassessment mechanisms
(i.e., IBI) with the descriptive biological condition axis and ultimately tiered designated aquatic
life uses. [Slide 16] This was illustrated with the actual set of data used to establish the tiered
designated uses for the Exceptional Warmwater Habitat, Warmwater Habitat, Modified
Warmwater Habitat, and Limited Resource Water uses. [Slides 17 and 18] The resulting
30
-------
US Environmental Protection Agency
Region/ORD Workshop on Aquatic Life Criteria December 4-7, 2001
biocriteria are organized by aquatic organism group (fish and macroinvertebrates) and
biological index (IBI. Miwb, and ICI), and are further stratified by use designation, site type
(headwater, wading, boatable), and ecoregion. In all, more than 40 sets of numeric biocriteria
exist in the Ohio WQS.
Bioassessments are used in a variety of ways to describe watershed health and diagnose
impairments. [Slide 19] Using biocriteria to describe changes in river segments through time
and before and after the imposition of pollution control programs was demonstrated with results
from the Scioto River. This particular example shows the positive response of the fish
assemblage to 15 years of water quality based pollution controls. [Slide 20] More recently,
biocriteria are being used in support of TMDL development to portray the spatial extent and
distribution of quality throughout watersheds. The Stillwater and Wabash River basin examples
show the influence of agricultural land use and riparian management practices in addition to the
recent concentration of large Confined Animal Feeding Operations (CAFOs) in parts of both
watersheds. [Slide 21] Bioassessments and attendant chemical and physical assessment tools
demonstrate the complex interaction between habitat and nutrient dynamics in Ohio watersheds.
The cumulative effects of habitat degradation and its associated consequences are frequently
exported into downstream reaches resulting in reduced assimilative capacity and lower quality
biological assemblages, the latter of which may contribute to an impairment of the designated
use as measured by the biocriteria. The example shown is the correspondence of headwater
stream habitat with total phosphorus. The reduced ability of degraded headwater segments to
process and sequester nutrients was illustrated in an inverse relationship between total P and
habitat quality as measured by the Qualitative Habitat Evaluation Index (QHEI).
[Slide 22] The preceding sequence of calibrating, deriving, and using biocriteria illustrates the
regional reference site approach in which consideration of inter-regional and intra-regional
factors are blended in an appropriate fashion to derive regionally relevant and attainable
biological criteria. Examples of the further use and value of biocriteria include the validation
and evaluation of chemical water quality criteria for ammonia-N [Slide 23] and metals [Slide
24]. A similar correspondence with physical habitat quality was also described [Slide 25]. This
has lead to the tiering of chemical water quality criteria and thresholds for a variety of chemical
parameters.
The principles of adequate monitoring and assessment are essential to the successful
implementation and use of biological assessments and criteria [Slide 27].
Questions and Comments
Question: Reference conditions - did you ever adjust for natural conditions (e.g., wildfires)?
Response: We have had extremes that include serious droughts and wet weather periods, but
this did not seriously alter reference condition at least impacted sites.
-------
US Environmental Protection Agency
Region/ORD Workshop on Aquatic Life Criteria December 4-7, 2001
Comment: We are talking about a broad, independent cluster of ecoregions (a stratifying
layer). Do we calibrate on a large uniform basis?
Response: We definitely need to calibrate across a broad enough spatial region to encompass
the full range of reference potential. If we calibrated only by ecoregion, I think
we would have a less descriptive and distinct biological condition axis.
Comment: We don't have a good definition in the landscape of what indicators are and how
they are being used.
Response: The critical vector against which biological condition and response is positioned
is human disturbance. Our problem is that in many cases our measures of human
disturbance are first order approximations at best and we can only compare them
one or two at a time. The best measure is biological response as it captures the
sum total of human stressors in the proper sequence.
32
-------
US Environmental Protection Agency
Region/ORD Workshop on Aquatic Life Criteria December 4-7, 2001
Use of Reference Sites and Conditions in the Development of Nutrient Criteria
- George Gibson (OW/OST)
EPA's National Nutrient Criteria Program approach to reference condition development was
presented. Two aspects of the approach discussed were: 1) the initial experience with
freshwater lakes and streams; and 2) the planned approach for coastal waters. The area of
aggregate nutrient Ecoregion 7 and the stations within that ecoregion were illustrated. Sources
of data include STORET data and data from the states and academic institutions in Ecoregion 7
(NY, PA, MI, WI, MN, IN, OH, and VT including the New York State Department of
Environmental Conservation, and the Lake Champlain Monitoring Project. Data were collected
over the period of 1990-1999 during all seasons. The majority of the data (80-90%) was from
STORET. The Nutrient Criteria Program was aware of the limitations and quality of the
STORET data and made an effort to "clean up" the data. This was done by removing
duplications; reviewing remark codes and deleting irrelevant and likely poor data; contacting
sources of data to determine sampling and analytical methods used; and verifying the locations
of stations and water body names.
Coordination was sought across EPA regions and among states through the Regional Technical
Assistance Groups (RTAGs) in each EPA Region. The lower quartile of STORET derived data
was presented for Ecoregion 7 lakes, ponds, and reservoirs for all seasons. Data parameters
include: concentration of chlorophyll-a measured with a fluorometer, concentration of
chlorophyll-a as measured by spectrophotometry, dissolved oxygen, nitrate, nitrite, total
nitrogen, total Kjeldahl nitrogen, total phosphorus, and Secchi data. Statistical analysis of
Ecoregion 7 data shows:
Percentiles for the entire Ecoregion and subecoregions correlate reasonably with state
data on reference lakes (e.g., MN),
Variability is high when all data are lumped,
There are differences between subecoregions,
There are seasonal differences, and
It is likely that variability will be reduced when we classify lakes according to physical
characteristics.
It was felt that the study was data poor in estuary and coastal data, and efforts are being made to
develop methods to collect more data in these areas. In addition, it was noted that the Minnesota
Lakes guidelines were developed independently and concurrently with the drafting of lakes and
reservoirs technical guidance and the results were comparable.
33
-------
US Environmental Protection Agency
Region/ORD Workshop on Aquatic Life Criteria December 4-7, 2001
Datasets evaluated were presented for New England Lakes (Matt Liebman, December 2000),
Minnesota Lakes (Steve Heiskary, April 2000), New York State Lakes (Scott Kishbaugh,
September 2000), Tennessee Streams (Greg Denton, October 2000), U.S. Geological Survey
(Richard Smith, May 2001), Delaware Streams and Ponds (John Davis, March 2001), and
Delaware Estuaries (Kent Price and Brian Glazer, June 2001).
An example of an initial coastal marine monitoring project conducted off the mid-Atlantic coast
extending from just north of Delaware Bay and south of the mouth of the Chesapeake Bay to the
area of Kitty Hawk, NC was illustrated in several figures showing the location of stations, as
well as results for a two-year summer nutrient survey. The variables addressed include: TN, TP,
Chlorophyll-a, and Secchi depth. The influence of data from estuaries can be seen in the graphs
and potential reference condition values are identified. An illustration of a stratified random grid
sample design for reference condition cells as related to an estuarine or other significant
discharge was provided. States can assist by providing data and needed monitoring support,
especially for riverine or estuarine and similar discharge sites. Contact information was listed
for EPA's National Nutrient Criteria Program Regional Nutrient Coordinators, and headquarters
personnel.
34
-------
US Environmental Protection Agency
Region/ORD Workshop on Aquatic Life Criteria December 4-7, 2001
Developing Nutrient Criteria Using Multi-Metric Indices: A case study in the
Mid-Atlantic - John Hutchens (ORD/NERL)
Development of nutrient criteria is of key importance for several factors. Nutrients are one of
the top three stressors to streams, impacting both human and ecological health. In addition,
upstream inputs can lead to problems for downstream receiving waters. Criteria development
was illustrated in a flow chart beginning with the identification of goals to monitoring and
reassessing criteria ranges. Nutrient criteria can be established by using frequency distributions
of all streams or reference streams, predictive relationships, and published thresholds or criteria.
A method of selecting reference values was illustrated for total phosphorus concentration (//g/L)
using percentiles from reference streams and total stream populations. Reference streams
selected include the 25th percentile of frequency distribution for all streams and the 75th
percentile of frequency distribution of reference streams using best professional judgement
(targeted), meeting criteria, and Indices of Biotic Integrity (IBIs). The benefits of using IBIs
were outlined.
The objectives of this Mid-Atlantic Highlands case study were to determine the total phosphorus
criteria using IBIs based on different assemblages of stream biota, and to compare criteria using
confidence intervals. Stream sites (1st - 3rd order) in New York, Pennsylvania, West Virginia,
Virginia and Maryland were sampled from 1993 to 1995 using an index period from April to
June. IBIs for macroinvertebrates and fish resulted in four of seven and four of nine metrics,
respectively, sensitive to nutrient enrichment. Specific percentiles from frequency distributions
were determined. Confidence intervals were then computed around the percentiles using
resampling techniques (i.e., bootstrapping). Ecoregion stream classification by Omernik (1994)
and Woods et al. (1996) was lumped to highlands versus lowlands. Minimally impacted sites
were compared. A comparison of total phosphorus criteria from "good" streams was presented
from various data sets such as: EMAP (all streams); macro-invertebrate biotic integrity index
(MBII) ("good" streams); Fish IBI ("good" streams); Nutrient Ecoregion XI; and Ohio EPA
Western Allegheny Plateau (OH-EPA WAP). Total phosphorus criteria data sets from "fair"
streams were compared as well. Using one nutrient criterion value means a stream is either
acceptable or impaired. An additional criterion set in the IBI "fair" category provides more
options for managers, while still meeting a certain level of biotic integrity.
Conclusions
This "simple" method to set nutrient criteria based on IBIs can be linked to designated
use classification;
The method can be applied to other stressors to set criteria;
35
-------
US Environmental Protection Agency
Region/ORD Workshop on Aquatic Life Criteria December 4-7, 2001
EMAP data can be used to extrapolate to population of 1 st to 3rd order Mid-Atlantic
streams;
The criteria based on the "all streams" method was more protective than the "IBI-
reference" method; and
Total Phosphorus criteria based on invertebrates and fish were not significantly different,
although invertebrates tended to be lower.
A summary of future directions was highlighted. Some proposed steps were to conduct
experiments to "fine tune" criteria in specific locations, and to examine the relationship between
local criteria and downstream conditions.
Questions and Comments
Question: Four out of seven metrics responded in the macro invertebrate IBI and four out of
nine metrics responded in the fish IBI - was it a positive or negative response?
Response: It was a mixed result, depending on the individual metric.
Comment: You can add nutrients to a system and increase its diversity, but that does not explain
why phosphorus levels increase in a stream that is naturally phosphorus-limited. In
the mid-western streams, if you increase the phosphorus, you increase the number of
some types offish. If nitrogen is increased, the result is a negative response. There
isn't necessarily a toxic threshold response with nutrients (nutrients are not
necessarily toxic). Excessive nutrients may be bad, but can also be beneficial.
Comment: Some states get funding by selling fishing licenses, and want to have high levels of
nutrients [in streams and lakes] because of the increase in the numbers offish.
36
-------
US Environmental Protection Agency
Region/ORD Workshop on Aquatic Life Criteria December 4-7, 2001
NHEERL National Nutrients Research Implementation Plan - Emile Lores
(ORD/NHEERL)
The National Health and Environmental Effects Research Laboratory's (NHEERL) National
Nutrients Research Implementation Plan provides the scientific basis for developing and
supporting nutrient criteria in the Nation's waters by defining nutrient load-ecological response
relationships. Nutrient issues were described as human activities changing the distribution and
movement of major nutrients across landscapes and waterbodies, in particular the increasing
nutrient loads to receiving waters. While EPA is developing guidelines, mandated by the Clean
Water Act, for setting nutrient criteria protective of aquatic systems, many scientific
uncertainties exist, and it is difficult to predict when a given nutrient concentration or loading
will have measurable adverse effects.
Research priorities for waterbodies include coastal receiving waters, streams and wetland, and
watersheds of which there are thirty-two (32) Full Time Equivalents (FTEs) focused primarily
on coastal receiving waters. Three critical endpoints of concern are: 1) low dissolved oxygen
(DO); 2) the loss of submerged aquatic vegetation (SAV); and 3) shifts in food webs leading to a
loss of commercially important fisheries and reduced aquatic biodiversity. NHEERL will be
looking at network analysis to see how systems process nutrients. Conceptual nutrient load-
response relationships were described for each of the three endpoints. Classification schemes
must be developed that group waterbodies by: susceptibility to low DO; importance of SAV in
ecosystem function and susceptibility to SAV loss; food web dynamics; biodiversity;
susceptibility to harmful or nuisance algal blooms; and fisheries loss. Currently, there are plans
for development of empirical and box models. The generic and SAV approaches to model
development were described and illustrated as flow charts. NHEERL would like to build a
model that combines both empirical and numeric modeling, i.e., to develop and validate the SAV
model. In addition, NHEERL plans to feed into and go beyond the Office of Water's reference
conditions. FY02 task activities were listed for DO, food web, and SAV-related issues.
37
-------
US Environmental Protection Agency
Region/ORD Workshop on Aquatic Life Criteria December 4-7, 2001
Aquatic Life Use (ALUS) Concept of Reference Sites - Susan K Jackson
(OW/OST)
The objective of the tiered aquatic life uses framework is to identify a common pattern of
biological response to human disturbance. The framework may be used to help establish
reference conditions, including clarifying the difference between minimally disturbed and least
disturbed reference conditions and identify reference conditions according to a region's
ecological potential.
A decision tree for establishing regional reference conditions and designating aquatic life uses
was presented; three scenarios for establishing reference conditions were included. Scenario 1
reflects a natural/undisturbed or minimally disturbed condition where the best existing
conditions are "natural" and should exhibit biological integrity. Scenario 2 portrays the least
disturbed condition; the best existing conditions are impacted, but the Clean Water Act (CWA)
101 (a) protection and propagation goal is met. Scenario 3 represents restoration targets where
the best existing conditions are impacted and the CWA 101 (a) protection and propagation goal
is not met. In some places there may be insufficient data to establish regional reference
conditions; however, aquatic life uses should still be designated based on restoration targets and
other applicable information. These uses should be periodically reviewed. Steps involved to
designate appropriate aquatic life uses were outlined for each scenario, and hypothetical
examples were provided.
Questions and Comments
Comment: Concern was expressed that the tiered aquatic life use model could result in
downgrading waters.
Response: We are aware of this concern and will address this in implementation guidelines. The
guidelines will recommend a comprehensive monitoring program and adequate data
for the appropriate classification of a waterbody.
Question: Are you trying to address antidegradation?
Response: Greater specificity in classification of waters supported by data should result in more
effective implementation of anti degradation provisions in State and tribal water
quality standards.
Question: How does this model overlay with the T& E Species Act?
Response: The biological assessments on which aquatic life use classification and biological
criteria are based measure attributes of different assemblages of the aquatic
community (e.g. benthic invertebrates, fish, periphyton, amphibians etc). Ambient
38
-------
US Environmental Protection Agency
Region/ORD Workshop on Aquatic Life Criteria December 4-7, 2001
habitat and chemical assessments are an integral component of the biological
assessment. These assessments are measures of the health of an aquatic community -
not one specific species. However, these assessments are " information rich" - and
detailed information on specific T&E species can be a component of the biological
assessment.
Question: How do you go back up the gradient without eliminating humans?
Response: The intent is to emphasize incremental, achievable goals for restoration of the
ecological potential of a waterbody or segment. Humans are an integral part of the
ecosystem - and there are ways to minimize to the extent possible degradatory
impacts on aquatic systems. For example, riparian buffers and stormwater retention
ponds are best management practices that can be implemented. It is useful to
articulate the full range of biological response to increasing human disturbance to
provide the public with a better understanding and context for existing conditions.
This should contribute to a more informed public process for decision making and
setting goals for waterbodies.
Comment: Back to cause and effect. From a biological standpoint, the chemical composition of
water, organic enrichment and sediment are associated with land use. In criteria
development, the challenge is how to technically go from a reference condition and
the designated use (goal) to something that actually protects the use from the stressors
to the system.
39
-------
US Environmental Protection Agency
Region/ORD Workshop on Aquatic Life Criteria December 4-7, 2001
THIS PAGE INTENTIONALLY LEFT BLANK
40
-------
US Environmental Protection Agency
Region/ORD Workshop on Aquatic Life Criteria December 4-7, 2001
CONCURRENT BREAKOUT SESSIONS
BREAKOUT SESSION I: Multi-Use Reference Sites
Facilitator: Bobbye Smith, USEPA/Region 9
Co-chairs: Don Huggins, University of Kansas
Gary Welker, USEP A/Region 7
George Gibson, USEPA/OW/OST
Objectives of Session
USEPA Region 7 is attempting to establish reference sites and is struggling with the definition of
a reference site that all states in the region can agree upon. A repeatable method is needed for
determining reference sites. USEPA Region 7 and Central Plains Biocriteria Workgroup,
formed in 1998, are working to develop reference conditions and site selection guidelines for
streams.
In this session, participants developed a working definition of a multi-use reference site for
wadeable streams; identified core factors (de minimus list) that should be considered when
selecting multi-use reference sites for wadeable streams; and arrived at a group consensus on the
definition of each core factor for the selection of multi-use reference sites.
Facilitated Discussions
A facilitated discussion was held to develop an operational definition of "reference site" for
wadeable streams. The term "reference condition" in relation to wadeable streams was first
defined as a cumulative distribution of variables (parameters) from reference sites. Reference
sites need to be selected by some common physical parameters (e.g., size, water temperature,
homogeneity, or geographic location). Examples of similar, recent efforts by the Biocriteria
Workgroup in Region 7 were given. Some participants proposed using reference conditions as a
method for raising the "standards bar," while other suggested "keeping the bar in place" and
applying resources where "things can be fixed."
A list of 38 factors to consider in the selection of reference sites for wadeable streams, developed
by the Biocriteria Workgroup, was presented for discussion. Participants discussed the various
factors, elected to lump some of them together, voted on which factors were a priority, and
reduced the list to eight core factors.
41
-------
US Environmental Protection Agency
Region/ORD Workshop on Aquatic Life Criteria December 4-7, 2001
The group then discussed and developed a definition for the core factor "representativeness"
based on the definition developed by the Region 7 Biocriteria Workgroup. A definition was also
developed for the core factor "biotic diversity and biomass." Discussion followed on the ability
to attain absolutes in the Ecological Condition Gradient. Some participants argued that
"absolute" never happens, while others disagreed. In addition, the presence or absence of
threatened and endangered species was discussed in terms of defining a reference site.
The question was raised: "what do we do if there are no sites in the upper left hand corner [of
ecological condition gradient - highest historical biological integrity]?" A suggestion was made
to rank the core factors on a scale from one to ten, and then identify how much row crop, pasture
and remaining forest exist. The Regional Environmental Monitoring and Assessment Program
(REMAP) should provide benchmark measures. In agricultural ecological regimes, it is
important to look at historical reconstruction information and measure the percentage of change
in land use/land cover from historical data to present day. The TMDL program can address
when criteria are not met, but may not meet biotic integrity. It would be resource-intensive for a
state to implement this type of approach.
Summary of Participant Comments
The sharing of information and communication between states in a region is of key importance.
Some states do not have reference sites or performance criteria. For Region 7, the "people
process" was the most difficult part of the process, and it was much easier to begin discussion
between states with a strawman on the table at a neutral party facility. There is a need to go
beyond geopolitical boundaries to establish reference sites. An evaluation of methods used for
determining reference sites is also needed.
ORD science needs included:
Assumptions behind "random sampling";
Resources;
Use existing state sites;
QA / QC for data sampling;
Refined / practical modes for extrapolating;
How many different types of systems are there; and
Bring people together.
Conclusions - Reporting Out
The above discussions were summarized in the report out to the larger group. Developed
definitions and identified core factors for reference sites were listed as follows:
42
-------
US Environmental Protection Agency
Region/ORD Workshop on Aquatic Life Criteria December 4-7, 2001
Reference Site for a Defined Environmental System
A location in the landscape that is representative of natural conditions (minimum anthropogenic
impacts) for that system and can be used as a benchmark to address multiple resource
management objectives.
Core Factors for Reference Sites
1. Land use/Land cover broad-scale
2. Land use/Land cover site-specific
3. Altered hydrologic regime
4. Biotic diversity and biomass
5. Physical and chemical parameters
6. Representativeness
7. Habitat, instream
8. Habitat, riparian
Representativeness
Reference sites should represent the range of biological, physical, and chemical
conditions of the ecoregion.
These sites should be minimally disturbed by anthropogenic activities; and
A sufficient number of sites should be selected to adequately represent the ecosystem
type and capture the natural variability among specific types.
Biotic Diversity and Biomass
Biotic diversity is consistent with both historical assemblages (where available) and
current distributions:
ป Presence of rare/unique communities;
* Limited number of exotics;
* Temporal variations considered;
ป Few native species lost; and
ป Presence of threatened or endangered species.
Take into account stream classification and size;
Migration possibilities should be taken into consideration: dams, reservoirs and drainage
divides can prevent recolonization of reaches.;
No nonindigenous species (no measureable effect on trophic structure);
Species composition;
Species diversity;
43
-------
US Environmental Protection Agency
Region/ORD Workshop on Aquatic Life Criteria December 4-7, 2001
Trophic structure;
Biomass;
Departure from native assemblages; and
Presence of sensitive species.
44
-------
US Environmental Protection Agency
Region/ORD Workshop on Aquatic Life Criteria December 4-7, 2001
BREAKOUT SESSION II: Charting a Statistical Course for Navigating the
Murky Waters of Bioindicator Development
Facilitators: Susan Cormier, USEPA/ORD
David Pfieffer, USEPA/Region 5
Presenters: Jeroen Gerritsen, Tetra Tech, Inc.
Michael Paul, Tetra Tech, Inc.
Objectives of Session
USEPA ORD/NHEERL and USEPA Office of Water/OST are developing a guidance document
on the data analysis methods used for biocriteria development. The guidance is intended to
provide states and tribes with the statistical and other analytical and interpretation tools needed
to develop biocriteria. At this point, the focus of the document is on methods used for analyzing
data from stream bioassessment in the United States.
In this session, the proposed contents of the guidance document were presented as a workshop to
potential users. The objectives were to teach the audience the methods and their application
(technology transfer) and to obtain feedback and responses from potential users on the utility,
completeness, and organization of the proposed contents of the document. Three analysis
methods were covered in the session:
1. Multi-Metric Index of Biotic Integrity (IBI);
2. Empirical statistical models (RIVPACS); and
3. Biocriteria discriminant analysis.
For each method, an introductory overview was presented, followed by example calculations and
hands-on demonstration of a single case study, using a single database provided by Wyoming.
Following each overview and demonstration, the group held a discussion of the method to elicit
comments and opinions on how best to present the methods, and on the needs of users.
Multi-Metric Index of Biotic Integrity (IBI)
A central premise of biological assessment is comparison of the biological resources of a water
body to an expected reference condition. Impairment of the water body is judged by its
departure from the expected condition. This approach presumes that the purpose of management
is to prevent, identify, and subsequently repair anthropogenic damage to natural resources.
45
-------
US Environmental Protection Agency
Region/ORD Workshop on Aquatic Life Criteria December 4-7, 2001
Biological assessment of waterbodies depends on our ability to define, measure, and compare an
assessment endpoint between similar systems.
Data analysis for the multimetric approach includes the classification of reference sites into
natural divisions so that appropriate comparisons can be made within classes; calculation of
metrics (biological attributes) to characterize biological condition; selection of metrics that
respond to stressors; and standardizing and aggregating responsive metrics into a dimensionless
index of biological condition.
Empirical statistical models (RIVPACS)
River Invertebrate Prediction and Classification System (RIVPACS) and its derivative,
AusRivAS (Australian Rivers Assessment System) are empirical (statistical) models that predict
the aquatic macroinvertebrate fauna that would be expected to occur at a site in the absence of
environmental stress. A comparison of the invertebrates predicted to occur at the test sites with
those actually collected provides a measure of biological impairment at the tested sites. The
predicted taxa list also provides a "target" invertebrate community to measure the success of any
remediation measures taken to rectify identified impacts.
Model-building begins with the classification of reference sites, as in the multi-metric
development. Following classification, a multivariate discriminant model is developed to predict
(assign) class membership of assessment sites. Actual assessment is done by comparing the taxa
found at a site to those expected to occur at the site. Expected taxa are determined from the taxa
found at the reference sites. Reference conditions are made site-specific by the discriminant
model, which may place a site intermediate among the reference classes, based on the site's
physical characteristics.
Biocriteria discriminant analysis
This approach develops assessment directly from statutory designated aquatic life uses. Aquatic
life uses are determined from the state's regulations. Professional biologists identify a set of
calibration sites meeting each of the aquatic life uses, as well as a set of sites that are degraded
(non-attainment) and do not meet any of the designated uses. Multivariate discriminant analysis
is used to develop one or more models that will discriminate the use classes from one another
using biological attribute variables (metrics). The resultant discriminant models quantify the
best professional judgment of the biologists into a statistical model that can then be applied
uniformly by anyone to assess sites and determine whether the sites are meeting their designated
aquatic life uses.
46
-------
US Environmental Protection Agency
Region/ORD Workshop on Aquatic Life Criteria December 4-7, 2001
Summary of Participant Comments
Participants' comments and discussions on the proposed guidance document were wide-ranging
and diverse. A critical issue was the definition of the audience for the document, and its use.
Some suggested alternatives included:
A complete, step-by-step "cookbook" for development of biological indices;
User-friendly introductory material to familiarize state managers with the analytical
methods so they could make an informed selection of consultants or employees to
develop an index for them;
Introduction to the concepts of biological assessment and biocriteria; and
Web-based training modules and FAQs (frequently asked questions).
Suggestions on content included:
General conceptual and theory overview with a conceptual review for each specific
method;
One or more detailed case studies to illustrate the methods;
Discussion of the advantages and disadvantages of each method, and guidance, or a step-
by-step key to determine which method is optimal for a state's particular objectives and
situation; and
Research to examine performance of the alternative methods under different conditions,
and to identify ways to integrate the methods to use the advantages of each.
There was considerable discussion on the identity of the end users, and why states and tribes
have not used guidance documents in the past to proceed with developing their biocriteria.
Several participants pointed out that a guidance document, no matter how good, is in itself not
enough to spur development of biocriteria. Nearly all successful state development has required
financial and technical support from EPA, in the form of grants and technical support from
experts to lead workshops, teach methods, and analyze data from identifying and selecting
reference sites to final data analysis. A barrier analysis was suggested to determine why many
states have not implemented biocriteria.
47
-------
US Environmental Protection Agency
Region/ORD Workshop on Aquatic Life Criteria December 4-7, 2001
Conclusions - Reporting Out
The above discussions were summarized in the report out to the larger group. In order to
improve EPA's technical assistance to the states, and to improve the scientific basis of
assessment, the session concluded with several recommendations:
Support research to combine the advantages of the multiple alternative approaches;
Identify the relationships of biocriteria to toxicity testing, chemical monitoring and
assessment, and habitat assessment;
Develop firm relationships between bioassessment, biocriteria, and Aquatic Life Use
(ALUS); and
Provide support to states in getting started on biocriteria development and in application
of biocriteria to regulations.
The session developed a partial consensus on tools to advance biocriteria:
Training courses and tapes;
Web-based training;
Frequently asked questions;
Software and technical support (including personal contact); and
Guidance document to include:
ป Basic concepts;
* Sufficient information to create "informed consumers" for technical support;
* Reference to more detailed documents; and
ป Detailed case study as appendix.
48
-------
US Environmental Protection Agency
Region/ORD Workshop on Aquatic Life Criteria December 4-7, 2001
BREAKOUT SESSION III: Aquatic Life Use Support (ALUS)
Facilitators: Susan Jackson, OW/OST
Sue Norton, ORD/NCEA
Gretchen Hay slip, Region 10
Susan Davies, State of Maine
Presenters: Susan Davies, State of Maine
Lester Yuan and Susan Norton, ORC/NCEA
Rick Hafele, Oregon Department of Environmental Quality
Cyndi Grafe, Idaho Department of Environmental Quality
Objectives of Session
Susan Jackson opened the Aquatic Life Use Support breakout session and stated the objectives
of the session:
1. To road-test the draft biological condition gradient;
2. To identify the relevant scientific issues and research questions and plans; and
3. To discuss issues relating to program implementation and communication.
Specifically, the session was intended to identify some of the practical issues with standards and
identify the benefits and potential pitfalls of this tool. Input from participants outside of EPA's
water program would be especially valuable in determining the potential impact on various
programs. Implications for resource protection and restoration could also result from the
session.
A brief introduction session followed, where each participant stated their name and affiliation, as
well as an issue they wished to bring up for discussion or consideration during the session. A list
of these issues has been included in the flip chart notes (Appendix D).
Presentations were made by representatives from three states (Maine, Oregon, and Idaho),
describing the way in which ALUS could be applied to each state's existing use classification.
49
-------
US Environmental Protection Agency
Region/ORD Workshop on Aquatic Life Criteria December 4-7, 2001
The Biological Condition Gradient - Susan P. Davies
The United States Clean Water Act offers no explicit definitions to facilitate implementation of
the Act's long-term goal to "restore and maintain...biological integrity of the Nation's waters,"
nor the interim goal to provide for "the protection and propagation offish." Even so, many
states across the Nation are collecting biological data and attempting to develop biological
criteria to address pressing management needs.
Operative definitions that are independent of any of the various assessment methodologies in use
are needed to provide for uniform interpretation of stages of biological degradation across the
country. A narrative, six-tier gradient of biological condition categories is proposed. The
Biological Condition Gradient attempts to: be consistent with ecological theory; accommodate
commonly observed biological responses to common stressor gradients; and to be consistent
with a common and broadly agreed upon understanding of "good" versus "bad" biological
conditions.
The dialogue among scientists that is required to describe these operational tiers entails an
heuristic problem-solving process of constructing, then refining, descriptions that are successive
approximations of imperfectly understood biological responses to stress. This process serves to
clarify the extent of scientific consensus, and the current level of certainty associated with
various elements of ecological theory.
50
-------
US Environmental Protection Agency
Region/ORD Workshop on Aquatic Life Criteria December 4-7, 2001
Progression of Ecological Degradation in Mid-Atlantic Streams - Lester Yuan
and Susan Norton
The presentation described a project undertaken to develop a better understanding of how the
composition of a stream community changes when various types of anthropogenic stress are
increased. The intent of the study was to:
1. Develop a method to compare the relative severity of ecological effects across different
stressors; and
2. Provide analytical tools to help define aquatic life use tiers.
Data was collected from 726 sites - mostly first to third order streams - during summer low flow
conditions, from 1993 to 1998. Data included the recording of physical habitat and water
chemistry information, as well as the collection of benthic macroinvertebrates and fish. Stressor-
response curves were generated using that data; examples were presented showing the plots of
relative abundance over pH for Plecoptera and Ephemeroptera (slide 4). In this case, the two
insect groups showed markedly different responses to pH, illustrating the point that the curves
would need to be re-scaled in order to be combined into a community response curve.
Metrics were scaled by the means and variances expected under unimpaired conditions.
Reference sites were included in the plots - the criteria for their selection are listed on slide 6.
The means and standard deviations (of each metric) for reference streams of varying sizes were
computed using regression techniques. A mathematical formula was developed for scaling
biological responses.
Stressors considered included:
Acidification (indicated by pH);
Nutrient enrichment (indicated by total phosphorus concentration);
Habitat degradation (indicated by RBP habitat score); and
Human disturbance (indicated by chloride ion concentration).
To isolate the effects of acidification, samples were selected that satisfied all reference criteria
except for the criteria for ANC. For the other three stressors, multivariate Loess surfaces were
computed to control for the effects of multiple stressors.
Scaled responses of metrics were then plotted on the same graphs with plots of several insect
taxa, showing varying responses depending on the stressor. The pH curve shows acidification to
have no effect on the number of tolerant taxa. Decreased pH, however, did result in an increase
51
-------
US Environmental Protection Agency
Region/ORD Workshop on Aquatic Life Criteria December 4-7, 2001
in Plecoptera, and a decrease in Ephemeroptera and total taxa richness. The nutrient enrichment
plot shows increased phosphorus levels to result in increased numbers of tolerant taxa, and in
total richness, although Ephemeroptera and Plecoptera decreased in abundance. Increased
habitat degradation (indicated by low RBP [Rapid Bioassesment Protocol] habitat scores) were
correlated with decreases in the abundance of Ephemeroptera, Plecoptera, and total taxa
richness, but an increase in the abundance of tolerant taxa. Finally, the response to increased
chloride concentration (indicating increased human disturbance) showed somewhat decreased
relative abundances of Ephemeroptera and Plecoptera.
This type of data could be used in defining aquatic life use tiers; a schematic representation of
the aquatic life use tiers in comparison with the scaled response plots for acidification was
presented in slide 13. Tiers three (3) and four (4), for instance, can be distinguished by a defined
level of decline in the relative abundances of sensitive and rare taxa (e.g., Ephemeroptera). Tiers
four (4) and five (5) can similarly be differentiated by a specified decline in total taxa richness.
Conclusions
Standardizing biological response allows the display of multiple metrics in a common
framework. The relative sensitivities of biological metrics change with changes in stressors.
Data displayed as scaled stressor-response curves can lend support to establishing aquatic life
use tiers.
52
-------
US Environmental Protection Agency
Region/ORD Workshop on Aquatic Life Criteria December 4-7, 2001
Numeric Biocriteria - Rick Hafele
The state of Oregon currently has a narrative biocriteria standard which states that, "... waters of
the State shall be of sufficient quality to support aquatic species without detrimental changes in
the resident biological communities." There are twelve definitions that further describe the
narrative statement (see full text of narrative standard at end of this discussion). Furthermore,
Oregon has additional standards, intended to protect beneficial uses for aquatic life. These
beneficial uses include salmonid passage, spawning, and rearing, and protection offish and
aquatic life.
The development of numeric biocriteria is underway by the Department of Environmental
Quality during the current triennial review process. This will allow for technical input, review,
and public comment.
Characterizing the reference condition is one of the first steps in developing these standards.
Reference sites are selected using Geographic Information Systems (GIS) data on surrounding
land use, human uses, population and roads density and other similar factors. The sites are then
categorized based on stream size, elevation, latitude and ecoregion. Final selection is based on
characteristics such as low human population density and road density followed by ground-
truthing. Data collection on all sites selected includes habitat assessment and water chemistry
parameters, and fish and macroinvertebrate assemblages.
Multi-metric (such as the Index of Biological Integrity [IBI]) and multivariant techniques are
used to analyze the data from reference sites, as well as test sites, and determine the degree of
change from the reference condition. Conspicuously absent taxa are also considered in this
assessment, as specific taxa can often point to specific stressors. Metrics such as the Hilsenhoff
biotic index are also used for the purposes of stressor identification.
Analysis of the data led to the realization that not all the reference sites chosen represented ideal
conditions. A classification system was devised dividing the reference sites into three
categories:
1. Class A: Ideal watershed and stream condition; a watershed with virtually no human
disturbance;
2. Class B: Good watershed and stream condition; some limited human disturbance, and/or
best management practices (BMPs) are well-implemented;
3. Class C: Marginal watershed and stream condition. Human disturbance is present, but the
site is the best one available. Replace if better quality reference sites are located.
53
-------
US Environmental Protection Agency
Region/ORD Workshop on Aquatic Life Criteria December 4-7, 2001
Tiered uses can be matched directly to these three types of reference sites: Class A reference
sites represent tiers one (1) and two (2). Classes B and C represent tiers three (3) and four (4),
respectively. Matching aquatic life use tiers to classes can draw attention to the fact that, in
some cases, the best available reference site is still only a mediocre quality stream. This is a
point that must be made carefully, in order to avoid the tendency to make the current "best
available" condition the goal for all other waters in the area. In some cases, however, that is the
only option available. As the condition of the reference sites improves over time, expectations
should be raised for other streams/sites.
Oregon's Narrative Biocriteria Standard
Waters of the State shall be of sufficient quality to support aquatic species without detrimental
changes in the resident biological communities.
Definitions
"Aquatic species" means any plants or animals that live at least part of their life cycle in
waters of the State.
"Biological criteria" means numerical values or narrative expressions that describe the
biological integrity of aquatic communities inhabiting waters of a given designated
aquatic life use.
"Designated beneficial use" means the purpose or benefit to be derived from a water
body, as designated by the Water Resources Department or the Commission.
"Indigenous" means supported in a reach of water or known to have been supported
according to historical records compiled by State and Federal agencies or published
scientific literature.
"Resident biological community" means aquatic life expected to exist in a particular
habitat where water quality standards for a specific ecoregion, basin, or water body are
met. This shall be established by accepted biomonitoring techniques.
"Without detrimental changes in the resident biological community" means no loss of
ecological integrity when compared to natural conditions at an appropriate reference site
or region.
"Ecological integrity" means the summation of chemical, physical and biological
integrity capable of supporting and maintaining a balanced, integrated adaptive
54
-------
US Environmental Protection Agency
Region/ORD Workshop on Aquatic Life Criteria December 4-7, 2001
community of organisms having a species composition, diversity, and functional
organization comparable to that of the natural habitat of the region.
"Appropriate reference site or region" means a site on the same water body, or within the
same basin or ecoregion that has similar habitat conditions and represents the water
quality and biological community attainable within the areas of concern.
55
-------
US Environmental Protection Agency
Region/ORD Workshop on Aquatic Life Criteria December 4-7, 2001
Idaho Stream Classification compared to ALUS - Cyndi Grafe
There are no numerical biocriteria standards in the state of Idaho The water quality standards
state, however, that streams must maintain a healthy, balanced biological community as part of
supporting beneficial uses and cite specifically the Idaho Water Body Assessment Guidance
(WBAG). The WBAG uses several ecologically based multi-metric indices to assess stream
condition. The numeric scoring associated with these indices is, by definition, relative to the
reference condition. Consequently, Idaho can improve or change the biological assessment in the
WBAG without having to initiate formal rulemaking.
Idaho uses multiple lines of evidence or a minimum of two indices. Presently, for streams, Idaho
uses macroinvertebrates, fish, and habitat information. When determining beneficial use support,
each index receives a condition rating based on its index score relative to reference condition.
The condition ratings are integrated by taking an average. The condition ratings were determined
based on an analysis of discrimination efficiencies and Type I/II errors. The result is three
categories of condition which could match with the aquatic life use tiers. For example, the
macroinvertebrates use:
1. Streams with scores above the 25th percentile receive a condition rating of 3;
2. Streams with scores between the 10th and 25th percentile receive a condition rating of 2;
3. Streams with scores between the minimum of reference condition and 10th percentile
receive a condition rating of 1; and
4. Streams below the minimum of reference condition are in violation of the minimum
threshold for a single assemblage.
These scores support the development of the 305(b) report, supplement subbasin assessments (a
section of the Idaho TMDLS) and determine which waters will be listed on the 303(d) list
(Impaired Waters List).
This approach captures information on stressors through analyzing biological indicators. Idaho
uses this in combination with assessing existing chemical numerical criteria (e.g., for dissolved
oxygen).
These condition ratings fit relatively well into the ALUS model, but, some flexibility will be
necessary to refine this system as more data is obtained, and the reference condition is fine-
tuned. In addition we should be able to refer to a process such as WBAG, rather than having to
go through rule-making to classify all state waters. Lastly, as the program stands now, we could
not have the six categories in ALUS, but could incorporate the model into the three existing
categories.
56
-------
US Environmental Protection Agency
Region/ORD Workshop on Aquatic Life Criteria December 4-7, 2001
[A comment was made to clarify that coming up with six tiers is not necessary; showing where
Idaho's three existing tiers fit into the ALUS diagram is sufficient.]
Summary of Participant Comments
Participants' comments were documented throughout the breakout session on flip charts; the
complete list can be found in Appendix D: Flip Chart Notes.
Some of the main topics of discussion included the potential link between ALUS and aquatic life
water quality criteria, and how it relates to the Endangered Species Act. Concern that it might
result in downgrading quality, particularly when moving to implementation; questions regarding
how the model will help; how it can be implemented into different programs; and what it will
look like in the permitting process.
Conclusions - Reporting Out
The presentations and discussions were summarized in the report out to the entire group. The
ALUS concept was shown to be applicable and consistent scientifically. Discussions focused on
how to adapt the rule-making process to refine uses. Some options were proposed by the states
(Oregon and Idaho presentations). Collaboration between the regions and headquarters will be
key in working through the implementation issues and determining if this approach is valid, and
of benefit to the states. In addition, flexibility with the model will be necessary if it is to be
adapted for use by the states (e.g., Idaho can use the model if it limits the number of tiers to three
instead of six and can reference a process in its water quality standards).
Research needs (ORD) and opportunities for collaboration were also identified and are listed in
the report out; the list of these needs can be found in Appendix D - Flip Chart Notes.
57
-------
US Environmental Protection Agency
Region/ORD Workshop on Aquatic Life Criteria December 4-7, 2001
THIS PAGE INTENTIONALLY LEFT BLANK
58
-------
US Environmental Protection Agency
Region/ORD Workshop on Aquatic Life Criteria December 4-7, 2001
TOXIC CHEMICALS SESSION
The Toxic Chemicals session was co-chaired by Lisa Macchio (Region 10) and Rick Bennett
(ORD/NHEERL). Macchio opened the session and identified its main objective as sharing of
information. Several speakers who work in the area of developing aquatic life criteria
participated in the session, and time was scheduled for discussion following each talk. Bennett
explained that speakers were chosen after many discussions attempting to narrow down the
number of topics, since not all relevant issues could be covered in the span of this workshop.
Speakers gave the essence of the issues in brief presentations, then engaged in discussion and
answered participants' questions.
Risk-Based Criteria - Russ Erickson (ORD/NHEERL)
Erickson began by presenting the basic definition of risk - probability of adverse effect - and
how that definition is modified in terms of ecological risk assessment and probabilistic risk
assessment:
Ecological Risk Assessment: The process evaluating the likelihood that adverse ecological
effects may occur or are occurring as a result of exposure to one or more stressors (FERA 1992).
Probabilistic Risk Assessment: A risk assessment that uses probability distributions to
characterize variability of or uncertainty in risk estimates (RAGS 1999).
The current water quality criteria are risk-based, with criteria concentrations defining levels of
effects, while return frequency defines the probability of occurrence. However, they are based
on a weak definition of risk. The critical maximum concentrations are based on toxicity to
aquatic organisms: the value is derived by taking the concentration at the 5th percentile of the
LC50 curve and dividing by two. This value was based on analysis, but provides no information
on any non-lethal effects that might occur at that concentration, or about the percent of mortality
at lower concentrations.
A schematic diagram was presented (slide 6) to illustrate the steps taken in defining risk.
Problem formulation considers the properties of: the ecosystem and the toxic(s); community and
population dynamics; exposure sources and pathways; ecological effects; and critical species and
endpoints. A conceptual model is developed and becomes the basis for designing exposure
analysis and effects analysis studies to determine the exposure profile and response profile.
These can in turn be used along with risk estimation and uncertainty analysis to complete risk
characterization and description. This process was followed when water quality guidelines were
59
-------
US Environmental Protection Agency
Region/ORD Workshop on Aquatic Life Criteria December 4-7, 2001
first being developed, so a lot of the risk characterization information already exists. Some parts
of the process, however, should become more sophisticated and based on data-driven analyses.
Plots of exposure concentration over time were presented as examples, showing greater
variability in concentrations measured in the "real-world " as opposed to those used in a typical
toxicity test (slides 7, 8). In this situation, the results of the toxicity test would not necessarily
reflect real-world conditions. The test exposure levels need to be adjusted to derive the criterion.
Current criteria are based on 96-hour LC50 tests - taking into account only what happens
following 96 hours of exposure. In reality, there are other effects to consider: for example, as
exposure time goes on, lower concentrations can kill most organisms. Currently all organisms
and conditions are treated in the same way, based on the one-hour average used to determine the
criterion. The exposure concentration is adjusted to meet the criterion (slides 11, 12).
Using detailed risk curves, however, information can be gained on the concentrations resulting in
lower percentages of mortality (e.g., five or ten percent). Models can help determine the
probability of reaching these concentrations (slides 17, 18). Time series plots can also give
information on toxicity effects occurring before 96 hours, such as those that might occur with
"fast toxicity" compounds.
Obtaining accurate field data on concentration over time can also have implications on
determining compliance: if compliance is determined by specific values, one extreme value
taken would result in an exceedence. By looking at the data from a broader perspective, risk can
be better defined, and compliance issues improved, perhaps by using mean concentration values.
Questions and Comments
Question: In terms of concept analysis, those curves [slide 8] did not look very different
from each other. Would they be different in the real world?
Response: There will not necessarily be large differences - they may only differ by a factor
of two or three, or even less. This is only one of the aspects of toxicity data, so
small changes in several aspects could add up to a significant change in the total.
Question: Did you use real toxicity data, and have you thought about new toxicity tests to
fill the existing gaps?
Response: Regarding the toxicity data, there is not always enough data available; also, some
toxicity tests are not designed to be able to make intermediate observations. We
have supported some efforts to test models in the past, but we do not, as a lab,
have the time that it takes to conduct broad toxicity testing.
60
-------
US Environmental Protection Agency
Region/ORD Workshop on Aquatic Life Criteria
December 4-7, 2001
Comment: I concur with using more of the data, specifically, different endpoints and lower
mortality percentiles. Have you thought about how to apply this to situations
where there is a mixture of chemicals?
Response: We have only dealt sporadically with multiple chemical effects. Trying to
combine chemicals' effects leads to some complexity, but it can conceivably be
done.
Question: How would this new approach affect implementation, specifically 303(d) listings
and total maximum daily loads (TMDLs)?
Response: We have worked through some of this, and it goes back to efforts ten years ago.
The simplest way to deal with the permitting process is if you are only worried
about one species; and, if there is good toxicity data, criteria can be set. This
value could be tied into an amount of discharge allowed, using a calculation to set
compliance limits to meet a certain level of risk. Compliance can also be
monitored on an average basis, or on a percentile basis. In general, if we can
express a certain concentration, it can be tied in to the permitting process.
Question: How would you look at chronic toxicity exposures?
Response: We have looked at many models for chronic toxicity and found this somewhat
more difficult to deal with than acute toxicity. Toxic effects differ at different
stages, such as critical reproduction and development stages. This also involves
the issue of multiple chemicals. There are some modeling approaches that can be
taken, but acute-to-chronic ratios are something we are attempting to move away
from or handle in a different way.
Question: If you are seeing effects on an endangered species, would you modify criteria to
take into account sub-lethal effects? Also, how would you modify this approach
to include bioaccumulative effects?
Response: For endangered species, we might look at a lower mortality percentage with an
uncertainty factor added. The criteria do not currently handle bioaccumulative
chemicals. Theoretically, these can be handled in the same way using the model.
Issues of multiple routes of exposure have to be considered as well.
Question: When you develop criteria, do you monitor for exceedences using ambient data?
Response: Ambient and permit compliance monitoring would be put on a better footing with
this approach. For example, at present, if we take ten measurements and find one
percent exceedence, we cannot be clear on what this means. Better criteria have
the potential for making the process more meaningful.
61
-------
US Environmental Protection Agency
Region/ORD Workshop on Aquatic Life Criteria December 4-7, 2001
Discussion of Proposed Guidelines Revisions - Charles Delos (OW/OST)
Pursuant to Section 304 of the Clean Water Act, EPA from time to time publishes numerical
water quality criteria intended to protect the beneficial uses of ambient waters. Since 1980 all
criteria for toxic pollutants, whether for the protection of aquatic life or human health, have been
derived following written procedures that are termed "Guidelines." These Guidelines indicate
the amount and type of data that should be obtained, and the procedures to be used for
interpreting and using the data to derive water quality criteria.
EPA is currently working on revisions to these Guidelines. This presentation dealt specifically
with proposed revisions to the aquatic life criteria guidelines. Because the revisions being
considered built from principles set forth in the 1985 Guidelines, it is useful to have some
understanding of how those guidelines are ordinarily applied: (1) Acute toxicity test data must
be available for species from a minimum of eight diverse taxonomic groups. The diversity of
tested species is intended to assure protection of various components of an aquatic ecosystem.
(2) The Final Acute Value (FAV) is derived by extrapolation or interpolation to a hypothetical
genus more sensitive than 95% of all tested genera. The FAV, which represents an LC50 or EC50,
is divided by two in order to obtain an acute criterion protective of nearly all individuals in such
a genus. (3) Chronic toxicity test data (longer-term survival, growth, or reproduction) must be
available for at least three taxa. Most often the chronic criterion is set by determining an
appropriate acute-chronic ratio (the ratio of acutely toxic concentrations to the chronically toxic
concentrations) and applying that ratio to the acute value of the hypothetical genus more
sensitive than 95% of all tested genera. (4) When necessary, the acute and/or chronic criterion
may be lowered to protect recreationally or commercially important species. (5) When
evaluating time-variable ambient concentrations, 1-hour average concentrations are considered
to be appropriate for comparison with the acute criterion, and 4-day averages with the chronic
criterion. (6) The allowable frequency for exceeding a criterion is set at once every three years,
on the average.
Much of the current work effort has focused on improving the applicability of the criteria to
time-variable concentrations, currently handled through the criteria averaging period and the
allowable frequency. In addition, significant effort has been directed at reducing potential
uncertainties in chronic effects levels. It now appears that some significant changes in the
framework perspective may have potential for addressing a number of issues.
Kinetic-based Modeling of Toxicity
The kinetic-based model of toxicity is intended to consider the speed at which effects appear in
different individuals and at different concentrations. As such, data should include survival
62
-------
US Environmental Protection Agency
Region/ORD Workshop on Aquatic Life Criteria December 4-7, 2001
counts taken at various times, rather than at 96 hours, as in the current Guidelines. Such analysis
would yield, for each species: (a) an effect concentration normalized to a standard duration of
exposure; (b) a rate coefficient, reflecting how quickly the toxicant acts on the organisms; and,
(c) a variability parameter, reflecting the range in the time-to-death of different individuals. This
approach evaluates the frequency of lethality in any long series of time-variable concentrations,
and can be used to calculate toxic stress for a group of organisms (formula presented). A
parallel approach can be applied to sub-lethal stresses, by taking sensitivity differences between
lethal and sub-lethal effects into account in the kinetic-based approach.
Assessing the Impact of Toxic Events
Establishing a measure of impact can be approached in various ways; to gauge the full impact
that a particular time series of concentrations would have on the exposed population of an
aquatic species, it was proposed that impact be defined as the resulting absence of individuals.
Any approach selected for obtaining the information should recognize, in a quantitative way,
that: (a) sub-lethal effects ultimately affect the number of organisms present, and (b) it takes
time to replace organisms that are lost due to lethal or sub-lethal toxic stresses. Applying the
analysis to a species exposed to a long series of time-variable toxicant concentrations will yield
the overall degree of impairment of the species, and can be integrated over time.
Derivation of a Criterion
The following example was used to illustrate the approach used to determine the impact on a
single species: for a series of toxicant concentrations, the percent lethality is predicted using the
kinetic-based toxicity model, and the percent of individuals absent plotted over time to show the
population deficit. To portray the effect of variable toxicant concentrations on an aquatic
community, this analysis would be applied to a diverse assemblage of tested species, evaluating
lethal and sub-lethal impacts on each species. This modeling approach is intended to
characterize real-world time series in terms of statistical properties. To evaluate impairment of
the assemblage of tested taxa, a very long time series would be generated, such that it had the
characteristics of a real-world time series. The impairment caused by a very long time series
with a given median concentration could be evaluated, and would yield a particular degree of
impairment stemming from concentrations having a particular median. The results from other
time series would indicate the relationship between median concentration and impairment of the
tested assemblage, and can be illustrated on a curve of impairment.
A decision would then need to be made on the appropriate level of protection. In the proposed
approach this level of protection would be expressed in terms of overall impairment, but can be
narrowed down to a single, critical species where necessary. Having selected the target level of
protection, one could then determine the associated criterion from the impairment curve. This
would yield a single criterion, consisting of a criteria concentration and an allowable frequency
63
-------
US Environmental Protection Agency
Region/ORD Workshop on Aquatic Life Criteria December 4-7, 2001
for exceedences, the combination of which is protective of both lethal and sublethal effects.
Built into the derivation of such a criterion is a recognition of the relative difference in severity
of lethal versus sublethal effects.
Conclusion
The framework under consideration appears to have the potential for advancing the derivation of
aquatic life criteria, particularly with regard to the handling of time-variable pollutant
concentrations, and lethal versus sublethal stresses. Much of this may be possible through
improved use of data rather than by increasing the data base requirements. Furthermore, by
explicitly incorporating whole-assemblage and long time-frame considerations, the framework
might help provide a better link between chemical-specific criteria and field-biological
measures, and might contribute to improvements in ecological risk assessment.
Questions and Comments
Question: How do the input from toxicity tests and the input from this approach compare?
Response: We are interested in lethal versus sub-lethal ratios, and in integrating different times
of exposure. This approach does not handle the effects on different life stages well.
We need to characterize different sensitivities among the life stages of an organism,
but need additional data in order to do this.
Question: What about applying this to the real world? For example, there might be refugia for
recruitment that are not taken into account by the model.
Response: We were interested in addressing this by considering the spatial extent of the
contaminated area. This could incorporate the idea that this is a real ecosystem, with
possibilities for immigration and emigration.
Question: While the method described is better, it is not very realistic.
Response: It is not, and that is why it refers to an assemblage, rather than a community of
organisms. We cannot assume or account for interactions between organisms.
Question: The sensitivity of some of the organisms (such as the ones listed in the graph) is
known. How would you deal with organisms that might not be on that list, and
whose tolerance is not known?
Response: We cannot really apply this to T&E species. Those would be considered separately,
especially in cases where immigration is not possible.
Comment: There is an example of a case in California where all invertebrates were heavily
impacted; an endangered fish in the same stream eats both zooplankton and clams.
Food web relationships are another aspect of community interactions that are not
64
-------
US Environmental Protection Agency
Region/ORD Workshop on Aquatic Life Criteria December 4-7, 2001
always known, or taken into account, increasing the possibility that we might miss
something.
Response: That is a really good point. We are not able to cover everything when using models.
Question: Were these proposed changes intended to be applied on the National level?
Response: That was the initial idea, but this might be too big a change. We might use this
process just to get the allowable frequency, which currently has a very weak technical
basis.
Question: Are we still protecting at the organism level?
Response: I am not sure that we have ever protected at that level.
Comment: It would be essential to collaborate on pertinent ESA work with counterparts in Fish
and Wildlife, and to consult all methodologies rather than only chemical methods.
Response: Informal consultation with those agencies is already taking place.
Question: How far are we from being ready to use these models?
Response: This particular model got more complicated and is almost beyond what we could
handle. In addition, people got re-directed to other projects. We are now back to the
"old" way, of not using time variability.
Comment: ORD remains interested in this project and we are trying to start it again. Efforts
continue to develop these new population methods, as are some efforts on
bioaccumulative chemicals. Moving this project along has been slow because it is a
big change from the current way. However, it can be considered from the viewpoint
that putting together all this information is not an accurate reflection of the situation,
but a useful synthesis of information that can be used for regulation and related to
biocriteria - not necessarily to make accurate predictions, but support the process.
Comment: Natural history characteristics of aquatic communities may be something that is
needed, as is early involvement (simultaneously with chemical tests of toxicity).
Question: What are you looking forward to from this workshop? Are you hoping to start some
engagement with the regions?
Response: The reason the Guidelines revision effort stopped was insufficient resources - not
because of the technical aspects or lack of interest. The project could start again if
ORD and OW had the resources.
Comment: ORD will tailor resources to more developments, but try to get more out of the
directions we are working in. If there are collaborative efforts to help this project
along, it could be picked up again.
65
-------
US Environmental Protection Agency
Region/ORD Workshop on Aquatic Life Criteria December 4-7, 2001
Comment: It seems that it would be a straightforward exercise to develop a matrix, by modeling
species and populations, and figure out the most sensitive life stages.
Response: We do not do much of our own testing, but I agree that the optimum way of doing
this is to start with modeling, and find out from there what testing is needed.
Question: There are some real needs in criteria development; some of the things in Russ
[Erickson's] presentation can be enacted. However, some of the things brought up
seem to be for long-term needs. What can we do now to translate information into
better permit conditions? How do we present this to management to get criteria and
implementation?
Response: The main problem with this is that it is too big a project. It is easier to complete
small projects with distinct endpoints, because they deliver a product.
Comment: The planning process should be addressing regional needs. If something will help the
regions, but might take ten years to complete, it should be funded anyway, and in the
interim regions will need something to work from.
Response: ORD does have a planning process, and regional representatives are on the Research
Coordination Teams to communicate regional priorities.
Comment: EPA and Fisheries are collaborating to some extent on addressing these issues,
through ESA. This can grow to a technical assistance dialogue.
Response: Such consultation should be on the methodologies, rather than the numbers.
Response: We are doing both; we do have to consult the numbers, since they are in the
[Endangered Species] Act.
Comment: Regions are told to voice their concern to several different places / agencies; yet they
still seem to not be heard.
66
-------
US Environmental Protection Agency
Region/ORD Workshop on Aquatic Life Criteria December 4-7, 2001
ESA Consultation on Toxic Pollutant Criteria - KM Kubena (Region 10)
Issues considered during a consultation of Toxic Pollutant Criteria as they relate to the
Endangered Species Act (ESA) included concerns about how well studies translate to the real
world, gaps in what existing criteria account for, and disagreement with methodology, among
others.
Concerns with applying experimental studies to real-world situations were raised. Surrogate
species, often used in experiments to represent effects on threatened and endangered (T&E)
species, vary in their ability to accurately reflect T&E species' responses. Test methodologies
(e.g., flow-through versus static) can over- or underestimate toxicity. Studies based on
individual endpoints, such as LC50, do not give information on the dose-response curve; for
example, there may be a steep descent below the LC50. New information and new technologies
can allow previously unreachable concentrations and previously unknown endpoints to be
detected.
Current criteria exist for toxic pollutants, but do not account for dietary routes of exposure, or
the interactions of chemical mixtures; the effects on critical species and prey; and the effects of
bioaccumulation, including maternal transfer. In addition, neither increased susceptibility to
pathogens, nor lowered thresholds to toxics due to metabolic stressors were accounted for in
derivations of the existing criteria.
Disagreements involving the methodology used include debate on whether to use LC50 versus
LC10, Lowest Observed Effect Concentrations (LOEC), or Incipient Lethal Endpoints (ILL) in
the development of criteria. Differences of opinion also exist on methodology regarding
speciation of chemicals, the effects of other water quality parameters on toxicity, and the use of
dissolved metals criteria. Discrepancies were also encountered in the enforcement of criteria and
the level of protection deemed acceptable, such as protecting at the level of the individual when
dealing with T&E species.
Other topics brought up included accounting for the dangers of banned chemicals; dealing with a
lack of information; limiting analysis to priority pollutants and, banking on the protection
afforded by human health criteria.
Examples were presented of two states' current criteria. The Idaho Standards include 124 toxic
pollutants - recently narrowed down to 22 and seven conventional pollutants. Some uncertainty
issues were addressed in a biological assessment, and the National Marine Fisheries Service
(NMFS) is currently drafting a biological opinion. In California, the Toxics Rule primarily
67
-------
US Environmental Protection Agency
Region/ORD Workshop on Aquatic Life Criteria December 4-7, 2001
addresses selenium (Se), mercury (Hg), and penta-chlorophenol (PCP), and their effects on 40
species. Agreement at the highest organization levels was reached for resolution of this Rule.
Recommendations and suggestions included:
Regional scientists need to know the state of the science and research efforts being
undertaken nationally;
Best professional judgement decisions need to be made together (i.e., uncertainty
factors); and
Continued cooperation between agencies and within agencies is necessary (i.e., National
and regional Memoranda of Agreement [MOA]).
68
-------
US Environmental Protection Agency
Region/ORD Workshop on Aquatic Life Criteria December 4-7, 2001
Data Quality, New Information, and Interagency Research Coordination -
Chris Tatara and Tracy Collier (National Marine Fisheries Service)
New information was presented on sub-lethal endpoints, the additive neurotoxicity of
organophosphate and carbamate insecticides, and the impacts of copper on salmonid fishes,
along with implications on criteria development and interagency research coordination. The
Endangered Species Act (ESA) requires decisions to be made using the "best scientific and
commercial data available." The U.S. Fish and Wildlife Service/National Marine Fisheries
Service (FWS/NMFS) information standards policy was presented.
Data were presented suggesting that short-term, sublethal exposures to environmental
contaminants can interfere with the sensory biology of chinook salmon in ways that might
negatively impact their survival or migratory success. The specific details of the study can be
found in the following scientific publication: Scholz, N.L., Truelove, N., French, B., Berejikian,
B., Quinn, T, Casillas, E., and Collier, T.K. (2000). Diazinon disrupts antipredator and homing
behaviors in chinook salmon (Oncorhynchus tshawytscha). Canadian Journal of Fisheries and
Aquatic Sciences, 57:1911-1918. The potential sublethal impacts of contaminants on olfactory
function in salmon is of particular concern because olfaction and olfactory-mediated behaviors
are critically important for the survival and reproductive success of anadromous species.
New data were presented showing the interactive effects of pesticide mixtures on
acetylcholinesterase extracted from the chinook olfactory nervous system. Threatened and
endangered species of Pacific salmon are commonly exposed to complex mixtures of current use
pesticides. The new study demonstrated that the inhibitory effects of organophosphate and
carbamate insecticides are additive when these chemicals are presented as mixtures. The results
are currently being prepared for publication.
Finally, a sensitive technique for measuring sublethal neurotoxicity in salmon was presented. In
the example given, short-term exposures to copper were shown to inhibit the responsiveness of
the peripheral olfactory nervous system to natural odorants. In vivo electrophysiological
recordings from the salmon olfactory epithelium are very sensitive and reproducible indicators
of sublethal neurotoxic injury in fish. Thirty-minute exposures to copper (at nominal
concentrations in the low ppb range) significantly reduced the sensitivity of primary olfactory
receptor neurons to chemical cues. Presumably, sublethal copper exposures in salmon habitat
could interfere with behavioral patterns (e.g. home stream migration) that are critical for the
viability of natural populations.
The experiments summarized above make a strong case for using physiological and behavioral
effects as endpoints in criteria development. In addition, the additive effects of chemicals that
69
-------
US Environmental Protection Agency
Region/ORD Workshop on Aquatic Life Criteria December 4-7, 2001
share a common mechanism of action should be considered. Criteria that are set as single-
chemical concentrations may not be protective of aquatic organisms that are exposed to
mixtures. The averaging period, currently determined by policy, should be data-driven and
include experiments using sub-lethal endpoints. Exposure concentrations should also be
revisited, as new data is showing significant effects at measured environmental concentrations.
Increased interagency collaboration would expedite resolution of ESA issues in criteria
development. Communication between EPA and NMFS can ensure that experimental designs
meet the data quality objectives for inclusion in criteria development, and that sufficient data is
submitted during pesticide registration to derive Aquatic Life Criteria (ALC) for pesticides when
they are registered. Three areas of contaminant research at the Northwest Fisheries Science
Center with implications for enhanced criteria development were identified:
Integrate the total exposure of the animals (e.g., through diet, water, sediment);
Utilize appropriate endpoints which cover the life history of the species (e.g., olfaction
for salmon); and
Consider cumulative exposure of multiple contaminants, as well as contaminants plus
additional stresses.
Critical body residue approaches were presented (slide 17) for determining the direct and
indirect effect thresholds of poly chlorinated biphenyls (PCBs) and tributyl tin (TBT),
respectively, on ESA listed salmon. A segmented regression approach was presented to
determine the direct effect threshold of poly cyclic aromatic hydrocarbons (PAHs) on groundfish.
The research papers on PCBs, TBT, and PAHs, can be found at the following web site:
http://research.nwfsc.noaa.gov/ec/ecotox
70
-------
US Environmental Protection Agency
Region/ORD Workshop on Aquatic Life Criteria December 4-7, 2001
Emerging ESA Issues - Steven Schwarzbach (U.S. Fish and Wildlife Service)
Criteria based on the concentration of chemicals in fish tissue are under development for some
bioaccumulative elements (selenium, cadmium, polychlorinated biphenyls [PCBs]), and a fish
tissue criterion has already been established for mercury. Whether these criteria will be effective
in protecting threatened and endangered (T&E) species will depend on several factors. Criteria
should be designed to take into account dietary intake to endangered biota other than fish. Both
statistical and mechanistic approaches should be considered, particularly to understand site-
specific variability. Special cases need to be considered, such as fishless waters with benthic
invertebrate pathways that are significant in dietary exposure. Implementation issues should
include measurements of concentrations in fish and other matrices, pathways, non-fish
endpoints, transformations and bioaccumulation factors. A mercury criterion has already been
established by EPA, designed to protect humans eating fish, based on estimates of Nationwide
average consumption rates.
Several studies were listed that have examined the effects of dietary mercury on fish eating
birds, along with the values found in each study to produce effects in certain species of birds
(slide 4). Endangered bird species which are potentially at risk from mercury in freshwater and
estuarine environments in California include the bald eagle, California least tern, California
clapper rail, snowy plover and the marbled murrelet. Studies are under way to determine how
much mercury each might be exposed to.
Data was presented showing the amounts of mercury measured in eggs of several species of
birds collected from San Francisco Bay. One of these species, the clapper rail, is not a fish
eating bird, but it forages when mercury methylation is taking place in the Bay. The egg
methylated mercury hatchability benchmark has been determined from previous studies for
pheasants and mallards and the clapper rail has been found to be equally as sensitive as the
pheasant, with a "safe" egg concentration of 0.5 ppm. Concentrations higher than 1 ppm were
measured in some clapper rail eggs from San Francisco Bay.
A fish tissue criterion for selenium is also under development, designed to protect fish, but other
aquatic dependent wildlife may be considered. In California, the Sacramento splittail and giant
garter snake are the two species at greatest risk from selenium and should be considered in
implementation. EPA has convened a peer review group to make recommendations regarding
species sensitivity, which tissue to use, warm versus cold water fish and lotic versus lentic
ecosystems. One of the recommendations made was to use smaller fish and measure whole-
body selenium. Some studies exist that have measured whole-body selenium in different fish
species (slide 11), and some recommendations of safe tissue concentrations have been made by
different investigators (slide 12).
71
-------
US Environmental Protection Agency
Region/ORD Workshop on Aquatic Life Criteria
December 4-7, 2001
Data was presented in graphs combining the water and whole-body concentrations of selenium
from 1991 to 2000 for fish and invertebrates (slides 13, 14). Measurements were off-set in time,
so that a measurement of water concentration would have occurred one week before its
corresponding measurement offish body concentration. A graph was also presented showing the
relationship between the average selenium concentration in fish tissue and in water (slide 16).
Selenium concentrations in the tissues of several species offish were also correlated with two
different food webs (clams and Crustacea). Water concentration of selenium was measured at
0.2 ppm - however, fish dependent on benthic food webs all had higher concentrations in their
tissues due to recycling of selenium through trophic transfer. A food web selenium model of the
Bay-Delta was presented to illustrate transfer through trophic relationships (slide 18). Specific
food webs may result in different rates of selenium accumulation: white sturgeons feeding on
benthic clams - who in turn feed on crustaceans- had higher accumulations of selenium than fish
who feed only on (non-benthic) Crustacea.
Some of the growth effects of selenium on birds were illustrated at the end of the presentation;
two avocet chicks of the same age, but noticeably different in size were shown (slide 20), the
smaller one of which had been exposed to selenium while in the egg.
Questions and Comments
Question: Has any work been done on determining [mercury] methylation rates?
Response: Not yet, although there is a proposal to do this in some tidal wetlands. There are
concerns that if we create wetlands without resolving the mercury issues, we
might create a lot more places where methylation can occur. Channels in
wetlands, where very high concentrations have been measured, are also an
important habitat for some species of birds.
Comment: Organisms cannot be protected by basing criteria on LC50 values. Comparing that
to human health rules (cancer rates threshold, for example) reveals the extent of
the discrepancy. As the previous presentation (on salmon and diazinon) made
apparent, there may be mechanisms of effects that we have not thought of, at sub-
lethal levels. Linking lethality to other observable effects is needed, and this
connectivity will be critical in planning future research.
Response: The LC50 values are intended to demonstrate lethal effects - not necessarily to be
used for setting protective criteria. The emphasis is on other endpoints.
Response: There is no lethality paradigm; lethality is not what is driving criteria
development.
Question: The O.Sppm fish-tissue mercury concentration criterion was based on a National
average offish consumption - is this number a problem for wildlife?
72
-------
US Environmental Protection Agency
Region/ORD Workshop on Aquatic Life Criteria December 4-7, 2001
Response: The state has to determine that and adjust the number where necessary. It should
be lowered, in some cases.
Comment: The National Average is better than it used to be, but still not protective of all
individuals. However, we should not be engaging in a discussion on human
health - there is a different meeting scheduled on that subject. It seems that there
is a lot of data on effects, however I am concerned about the sharing of data,
especially collaboration across agencies.
Response: There is some concern among scientists about releasing draft data, but
collaboration is still very important.
Response: ESA maybe be a vehicle for crosswalking and collaboration.
73
-------
US Environmental Protection Agency
Region/ORD Workshop on Aquatic Life Criteria December 4-7, 2001
Surrogate Species in Assessing Contaminant Risk for Endangered Fishes -
Foster Mayer (ORD/NHEERL)
The presentation addressed the issues of sensitivity to chemicals, in particular whether
endangered species are more susceptible to toxicants than non-endangered forms. The
possibility of using surrogate species to determine the sensitivity of endangered species was
examined.
Several species offish and one bufonid species were tested (slide 6) in a static acute toxicity
experiment, with mortality observations taken at intervals between 6 and 96 hours of exposure.
Five toxic chemicals were tested to try to represent the majority of modes of action: carbaryl,
copper, 4-nonylphenol, penta-chlorophenol, and permethrin. The known effects of these
chemicals are listed on slide 22. With some of the species tested, there were not enough
individuals available for replicates - due to mortality - and only one run of the test was
conducted.
General tolerance trends were presented for the five chemicals (slides 24-28), plotting the
species tested in order of most tolerant to least tolerant. These results were combined to produce
an overall rank of species sensitivity using all chemicals tested (slide 29). A system was then
developed to estimate species sensitivity using a simple linear regression model (slide 30). A
short demonstration of the model was performed at the end of the presentation, using rainbow
trout to predict the sensitivity of another species. Interspecies correlations - derived from the
model - were presented using fathead minnows, rainbow trout, and sheepshead minnows (slides
31,32 and 33, respectively). Comparisons showed that using rainbow trout and fathead
minnows as surrogates gave the best results as far as predicting the sensitivity of other species
(slide 34).
Conclusions
Endangered species do not appear to be any more sensitive to xenobiotic chemicals than
other species;
The question is not one of sensitivity to xenobiotic chemicals, but one of vulnerability of
endangered populations to additional environmental insults;
Surrogate test species are representative of endangered species;
Interspecies correlations are reliable estimators of acute toxicity to endangered species;
correlations are best within a family;
A factor of 0.5x the rainbow trout geometric mean LC50 also provides an estimated LC50
for endangered fishes;
74
-------
US Environmental Protection Agency
Region/ORD Workshop on Aquatic Life Criteria
December 4-7, 2001
Present approaches to establishing National water quality criteria appear to be protective
of endangered aquatic species; and
State recalculation of National water quality criteria by eliminating certain species from
the data set (e.g., rainbow trout) may leave endangered species inadequately protected.
Questions and Comments
Comment: One issue with PCB is that there is a commercial and a purified form, and that it
affects early life stages in salmonids.
Response: This model does not address that issue.
Question: Do you think the correlation between the two species of minnows could have
been due to external effects, rather than their taxonomic relation?
Response: Probably not, because they still show strong similarities when using a larger
database.
Question: Hard water was used for these experiments; there might be a different hardness
slope for some species that is not reflected in the criteria. Did you choose the
high hardness level to improve correlation?
Response: No, the higher hardness was chosen to help reduce stress in species not normally
cultured or tested under laboratory conditions. The hardness slope would not
matter since the estimated result would be based on the water hardness of the
surrogate species test value entered into the model.
75
-------
US Environmental Protection Agency
Region/ORD Workshop on Aquatic Life Criteria December 4-7, 2001
Predicting the Toxicity of Metals to Aquatic Organisms: The Biotic Ligand
Model - Charles Delos (OW/OST)
Ligands are any chemical structures that can bind with another chemical or metal. Both organic
and inorganic ligands are found in water. Organic ligands include dissolved organic carbon
(DOC) and humic and fulvic acids - from the breakdown of animal and plant matter; inorganic
ligands vary in importance, depending on the metal in question. Ligands can also be biological,
one example being the gill membranes offish. Copper, and possibly other metals, bind to
chloride cells on the gill membrane, where copper can interact with specific enzymes that
regulate sodium levels in the organism. Sodium uptake is inhibited, resulting in ionic imbalance
and the eventual death of the organism. Silver binding to chloride cells has a similar effect, as
another antagonist for sodium binding.
In deriving criteria for metals, a problem has been their interactions with other water quality
parameters; these can cause the same concentration of metal to exert varying toxicity depending
on other water quality values. For example, copper's key controlling water quality constituents
are pH, calcium, carbonate and DOC. The conventional approach in dealing with this problem
has been to derive a regression relationship between total hardness and effect concentration.
Hardness is the measure of total calcium and magnesium concentration (predominantly). It can
be correlated with pH and alkalinity, and is used in this case as a surrogate parameter.
Conventional criteria derived in this manner can be over-protective if DOC concentration is
high, or under-protective at low pH. In addition, the water effect ratio (WER) must be calculated
to obtain the effect concentration for each site.
The Biotic Ligand Model
The Biotic Ligand Model (BLM) was developed as an alternative, and consists of three main
types of interactions: metal-inorganic ligand; metal-organic matter, and biotic ligand
interactions. A schematic diagram was presented showing these three potential interactions
(slide 17). A dose-response curve was presented plotting 120-hour rainbow trout percent
mortality versus short-term, 24-hour gill copper, using two different tests at different DOC
concentrations. The curve can be used to predict a gill LC50 concentration. A fundamental
assumption of the BLM is that the net metal accumulation that is associated with a fixed effect,
such as 50% mortality, is uniquely defined, and therefore independent of site water chemistry. If
this is the case, one would only need to predict the dissolved metal concentration that will result
in this critical metal accumulation to predict the corresponding LC50. This is precisely what the
BLM does.
76
-------
US Environmental Protection Agency
Region/ORD Workshop on Aquatic Life Criteria
December 4-7, 2001
A plot was presented showing a summary of predicted LC50 for copper using five species of
invertebrates, compared with measured LC50 concentrations. The plot (slide 19) shows that the
BLM is able to accurately predict concentrations in these invertebrates. A similar comparison
plot (slide 10) was presented for silver LC50 concentrations in rainbow trout, fathead minnows
and Daphnia. The BLM does not currently account for species sensitivity; however,
mechanistically-based approaches for doing so are being developed, and may eventually lead to
further refinements of the model.
Implementation Issues
The BLM exhibits time variability related to input and output, since hardness, pH, alkalinity and
DOC can vary over time. Each sample generates a different criterion, and different copper
concentration, requiring a comparison of the sample's concentration to the criterion calculated
for that sample. This is not a new implementation issue: for example, current metal criteria are
hardness-dependent, and the ammonia criterion is pH and temperature-dependent. Another
limitation is that the design conditions do not define the allowable frequency of exceedence,
which would have to be determined. Implementation is complicated by the number of degrees
of freedom in the model. Working with any particular data set, the resulting site-specific
criterion will be affected by what the analyst chooses to consider or not to consider.
Questions and Comments
Question: Is chloride a parameter which affects silver?
Response: Some early toxicity experiments with minnows did not show an effect, but later
work with trout does. It could be that silver forms complexes that are somehow
able to pass through membranes to some extent.
Comment: Chloride should definitely be a consideration in fresh water; for marine water,
ingestion, rather than contact with the gills, might be a pathway.
Response: The parameters are also flexible, and can be changed if something is found to
have an effect.
Question: Were the binding affinities of different metals considered?
Response: Yes, each metal requires its own development in the BLM.
Question: How would this model deal with chemicals which exist in multiple valence states,
such as arsenic?
Response: Nobody has talked about either arsenic or chromium in the context of the BLM.
The model functions with metals where the +1 or +2 ion is important. We have
not explored the possibility of expanding it yet.
77
-------
US Environmental Protection Agency
Region/ORD Workshop on Aquatic Life Criteria December 4-7, 2001
Dietary Metals: How Important Are They? - Russ Erickson
Dietary routes of metals exposure have been of importance for some time, especially in regard to
ESA issues. They can contribute to toxicity - chronic toxicity in particular - and add to the
exchange between organisms and the ambient water. Toxicity experiments usually face the
problem of the food eventually contaminating the initially "clean" water, if only to a small
extent. Current criteria are consequently based on studies where the food route of exposure is
under-represented. Diet can nonetheless be the dominant route of metal accumulation, and cause
effects on survival, growth and reproduction.
Although not independent of water borne exposure, dietary exposure to metals can result in
different disposition and effects. In naturally contaminated bivalves, for example, 95% of
accumulated metals occur in the digestive gland, where they form mineral deposits. These
metals remain in the digestive tract and cannot have systemic effects. By contrast, metals taken
up through the water are circulated and can have more effects than those accumulated through
the diet. Measuring whole-body metal concentration would not, in this case, give an accurate
picture of exposure; more information would be gained by studies designed to examine dietary
and water borne uptake separately.
Some studies have been conducted which considered dietary routes of exposure, particularly
copper exposure of rainbow trout, both through natural contamination and salt-amended diets
(slides 4,5). The ratio of copper concentrations in food and water in natural systems is generally
5,000-10,000, and water effects concentrations correspond to about 100|ig Cu/g in food, well
below the threshold identified for dietary effects (500|ig Cu/g dry food).
Survival data was also presented for the amphipod Hyallela azteca exposed to lead through diet,
and a drop in survival was observed, although it was not statistically significant. Effects on
reproduction, however (measured as neonates per survivor) showed significantly lowered
reproduction rates in those individuals exposed to lead through their diet (slide 7).
A similar study examined the relative importance of water and dietary routes for silver effects in
copepods. Individuals were exposed separately to food and water contaminated with silver. No
effects on survival were observed in individuals exposed to up to 5nmol/L of silver in water, but
significant effects are evident at 2nmol/g concentrations in food. The pattern is similar when the
body burden of silver was measured instead of survival (slide 8). Another study, however, using
a closely related organism, found no effects caused by the dietary route of exposure. Finally a
study where naturally contaminated Lumbriculus (earthworms) were fed to fish, resulted in few
effects, although a similar study by a another investigator did show noticeable effects.
78
-------
US Environmental Protection Agency
Region/ORD Workshop on Aquatic Life Criteria
December 4-7, 2001
Due to these differences in study outcomes, some of the data has become questionable, but still
cannot be overlooked. This area is still unresolved and ambiguous, but it can be said with
certainty that many organisms will be grossly under-protected if the dietary route of exposure is
not considered when setting criteria.
Questions and Comments
Question: What was the difference between the control and the lead-contaminated diet [in
the amphipod study]?
Response: The same formulation of food was used (rabbit food) but for the lead diet it was
equilibrated in a chamber that had received the test concentration of lead.
Various lead concentrations were used in the test. The control food may have
been run through the same procedure, but without the lead.
Question: Did you look at all the total recoverable metals, rather than just dissolved metals?
Response: It would be comparing apples and oranges, even though suspended metals, like
suspended solids can have some effects. Even though doing this would add a
small level of protection, there is no real, direct relationship between the two.
Question: If diet were an important pathway for exposure for an organism, how would you
address that with criteria?
Response: I don't have a good answer to that, but one option is to use an additive model.
You would correlate additive effects and design criteria to correct for it (using
data for separate dietary, water and combined to come up with a relative toxicity
value of some kind).
Comment: If the goal is to lower the criteria value, and the diet component is also influenced
by sediments, you may want to de-regulate that component.
Response: You would need an exposure rate calculation. Most contaminated sediments were
contaminated because of high water concentrations which no longer exist. Part of
the process is to ask the question: "If we have criteria based only on water
concentration, without considering the incidental contamination of food, will we
allow that in the field?" If so, how much does the food get contaminated from the
water? A partition coefficient between food and water is needed, and it may need
to be site-specific. This process would take into account distribution from water
to food, to sediment, etc. A problem with this is that partition coefficients vary
with metal concentrations.
Question: What biota would you use?
Response: Rainbow trout, most likely juvenile growth tests assuming the worst case. Also,
we have looked at the literature on macroinvertebrates, and are considering
studies looking at invertebrate metal concentrations relative to water metal
concentrations. This will not, however, apply to contaminated sediments.
79
-------
US Environmental Protection Agency
Region/ORD Workshop on Aquatic Life Criteria
December 4-7, 2001
Question: Since these are dynamic systems, have you considered going to different systems,
with large populations, to compare populations and test reproductive success in
the field?
Response: The problem with field data is that there are very few areas that are contaminated
to this degree with silver and other metals in sediments. It would certainly be
useful to run such a study if we can find appropriate sites; I am not aware of
anyone who has done that yet, however.
Question: Are you aware of any similar effects as a result of bioaccumulation - which is
very dependent on diet?
Response: Yes, there are effects of that kind.
Question: Do you have any information on algae?
Response: Periphyton tends to be in the same magnitude as macroinvertebrates in regard to
metal contamination. Some invertebrates store accumulated metals in less
reactive compounds - such as metal granules - which are not absorbed.
Question: Did you consider detritus-based food routes?
Response: Yes, the food used could emulate detritus or algae, or another organism.
Comment: It was mentioned already in a previous presentation that endangered species are
not uniquely sensitive to toxics.
Response: That was the case for fish species, but there are many other organisms for which
we don't have that kind of information. It would be dangerous to lump all
endangered species into one category.
Comment: That information should hold for closely related groups - they should respond to
toxins in similar ways, whether or not they are listed as endangered.
Comment: Species with similar biological requirements should also respond the same way
and exhibit the same effects to toxins.
Response: Right now, most of the information we have is on taxonomic relationships, we do
not know as much about all species' biological requirements.
Comment: Endangered species did not become endangered because of chemical exposure.
Even if only fish have been studied, we can still understand the concepts behind
that study. Most likely, endangered species should not be any more sensitive to
chemicals than non-endangered species.
Response: We don't know for sure the reasons why they became endangered, chemical
exposure may play a large role.
Response: There may be exceptions, but, in general, chemical exposure was not the reason
for the populations' decline.
80
-------
US Environmental Protection Agency
Region/ORD Workshop on Aquatic Life Criteria
December 4-7, 2001
Question: Is there any information suggesting that endangered species are especially
sensitive? I am not aware of any specific cases.
Response: Selenium has played a role in the decline of the razorback sucker (a fish), but we
do not know that it is especially sensitive to it, compared to other related fish.
Question: What data do we have that indicates that is not the case [i.e., endangered species
being particularly sensitive to toxic chemicals]?
Response: There is evidence that exposure of a population to chemical stress reduces its
genetic diversity, which might ultimately have an effect.
Comment: For example, certain genotypes of mosquitofish are tolerant of acute toxicity,
while others are more tolerant of chronic toxicity. That variation is needed to
ensure a healthy population.
Comment: If guidance cannot be developed on the biotic ligand model, what would the
consensus be on setting criteria?
Response: We could put out criteria and see what the states do with them, or tell all the
states that they must do something; both of these are ways we could approach
implementation guidance.
Question: Could ORD interpret all the data on dietary exposure?
Response: It is difficult to see what actions could be taken right now on that, since there are
so many contradictory studies, many using the same organism. More work is
needed to sort out that data, more resources need to be put in. As far as risk based
criteria, there is an understanding within ORD of the need to do something better
with the biotic ligand model guidance.
Comment: We are already linking criteria with other ecosystems endpoints, and I hope they
will also be linked at the community level. We have been successful in
measuring this in the field, and hope future work will include all scales of
biological communities, rather than being limited to populations.
Response: Improving water quality criteria so that they can be protective of communities is
worthwhile. Some critical processes that are important in assigning risk cannot
be incorporated in the grand analysis, however we could do a more limited
synthesis of the data.
Comment: Continuing to study new situations and creating new tools is key; the information
on dietary exposure is a good example of that.
Response: If we are talking strictly about dietary routes of exposure, we are not yet at the
point of being able to make even a good guess. Because of all the contradictions
in the data I would not now be comfortable including it in criteria - even if some
of the data suggests dietary exposure should be a concern.
81
-------
US Environmental Protection Agency
Region/ORD Workshop on Aquatic Life Criteria December 4-7, 2001
Numerical (Criteria) for Sediment-Associated Chemicals - David R. Mount
(ORD/NHEERL)
Toxic chemicals in sediments can affect organisms that live directly in sediments, as well as
other species in the same ecosystem through contamination of the water column and food chain.
This talk focuses on the former. In the early 1990s, EPA's Office of Research and Development
and Office of Water (OW) began work on the derivation of sediment quality criteria - later
changed to sediment quality guidelines.
Toxicity of sediment contaminants to benthic organisms can vary with the characteristics of the
sediment. This is well demonstrated by studies published by Adams et al., in which midges were
exposed to three different sediments, each spiked with a range of kepone concentrations.
Survival as a function of sediment kepone concentration varied significantly among the
sediments (slide 2). However, if one plots the survival data on the basis of kepone concentration
in interstitial water, survival responses were similar (slide 3). The idea that the toxicity of
chemicals in sediment is proportional to their chemical activity in interstitial water is captured by
the theory of equilibrium partitioning (EqP; slide 4). In sediment, biota can be exposed both
directly through the sediment, and through interstitial water. EqP proposes that the toxicity of
chemicals in sediment can be represented by assuming an equilibrium among the organism, the
interstitial water, and the sediment.
Bioavailability of non-polar organic chemicals, such as kepone, is related to the organic carbon
content of the sediment. In the kepone experiment shown previously, the different responses
observed in each sediment (see slide 2) were related to the differing organic carbon content or
the sediments. If one replots these data as concentration of kepone per gram organic carbon
(OC), we see that much of the variability in response is explained (slide 5). The organic carbon
partition coefficient (Koc, slide 6) can be used to relate the concentration of chemical in the
sediment to that in interstitial water. Using this relationship, one can predict the concentration in
sediment that will cause toxicity based on the concentration that causes toxicity in water (slide
7). The draft sediment guidelines EPA has developed are based on the level of protection
offered by ambient water quality criteria. The sediment guideline concentration is calculated by
inserting the final chronic value (from the water quality criterion) for the chemical as the
interstitial water concentration associated with effects (slide 8). A plot of organic carbon
normalization in sediments spiked with six chemicals showed good predictions in a wide range
of sediments (slide 9). The generic calculation formula for a sediment quality guideline for a
non-ionic organic, using equilibrium partitioning, was also presented (slide 10).
Equilibrium partitioning provides a mechanistic basis for management decisions and can be
adapted easily to new chemicals and to expanded partitioning models. Toxicity of mixtures can
82
-------
US Environmental Protection Agency
Region/ORD Workshop on Aquatic Life Criteria December 4-7, 2001
also be incorporated into the model, and the level of protection altered to meet site-specific
objectives.
Misconceptions about the equilibrium partitioning theory include observations that it can only be
applied to a limited number of chemicals, that it erroneously treats all organic carbon as
identical, and considers interstitial water as the only route of uptake. While specific guideline
documents have been prepared for only a few chemicals, the concept can be applied to any non-
ionic organic compound. Organic carbon is, indeed, currently treated in the same way by the
model; this is a first order approximation, but is very effective in reducing unwanted variability.
In addition, the model can be adjusted to take into account additional information on the specific
type of organic carbon in question, when such information is available. Finally, the conceptual
model of chemical exposure shows a tendency toward equilibrium whether exposure is from
interstitial water or other sources. The fact that chemical concentration in interstitial water is
used as an index of bioavailability does not mean that it is the only route of uptake as is
sometimes suggested.
The ingestion pathway of exposure was also considered and equations presented for the
calculation of water and sediment exposure for sediment-ingesting infauna. Graphs of survival
over sediment kepone concentration were presented, noting that if ingestion were driving
toxicity, the mortality curves plotted over concentration should be similar, regardless of the
percentage of organic carbon in the sediment.
In regards to the importance of EqP guideline values in management decisions - whether they
should be screening, intermediate, or definitive assessment tools - the weight of the evidence
should be proportional to the weight of the decision, and the magnitude of exceedence and level
of protection should be considered. As a general rule, exceedences for organics are intermediate.
Expensive remedies should not be undertaken without verifying that the sediments are
"behaving." The status of the documents for endrin, dieldrin, metals, and polycyclic aromatic
hydrocarbon (PAH) mixtures were reported, along with contact for further information and
copies (slide 18).
Some data on the toxicity of metals in sediments were also presented, specifically mortality
curves with dry weight metal concentrations in sediment for five metals (slide 19). When these
results were normalized for interstitial water, no toxicity was evident when metals were below
toxic concentrations (slide 20). Metals used were: cadmium, copper, nickel, lead, and zinc. In
order to normalize the effects of metals in sediment, the acid volatile sulfide (AVS) and
simultaneously extracted metal (SEM) need to be determined. AVS is produced by bacterial
breakdown of organic material and describes the amount of sulfides in the sediment. SEM
represents the metals extracted during the AVS procedure, and should be less than the "total
metals." Subtracting AVS from SEM yields a number that can be used in ranking sediments.
The effectiveness of normalizing data in this way was demonstrated by plotting the same
83
-------
US Environmental Protection Agency
Region/ORD Workshop on Aquatic Life Criteria December 4-7, 2001
mortality data over total nickel concentration (ug/g), and over SEM-AVS (slides 23, 24). This
relationship can be used to derive a solid-phase criterion, and an interstitial water criterion can
also be derived using the final chronic value (from water quality criteria). If either criterion is
met, sediments should not be toxic due to metals.
While SEM-AVS can predict lack of toxicity, as indicated above, that value divided by fraction
organic carbon (Foc) can be used to predict toxicity. Mortality plots, over SEM-AVS/FOC, can be
used to determine which concentrations, expressed in SEM-AVS/FOC will result in 95%
mortality (slide 28). Equilibrium partitioning theory was compared to empirically-derived
guidelines, which typically use large databases of field-collected information. Such guidelines
do not account for the known effects on bioavailability.
Equilibrium partitioning can also be applied to a mixture problem: PAHs were used as an
example, and values were initially developed for single PAHs. These chemicals do not,
however, exist in the field as single compounds, and PAH toxicity to benthic organisms appears
to be additive. The approach was therefore altered to account for multiple PAHs. Data were
presented of the percent mortality of amphipods exposed to one PAH compound from
contaminated sediments (slide 32). When these results are plotted over the organic carbon-
normalized concentrations of 12 PAH compounds, a criterion can be derived which will take into
account the additive effects of a mixture of chemicals.
Questions and Comments
Question: How would you deal with patchy distribution of sediments?
Response: One assumption behind this theory is that small organisms will spend most of
their time in a small part of a large mosaic of sediment types. Many benthic
organisms do not really "travel" appreciably.
Question: Have you looked at other endpoints [besides mortality] in terms of the additivity
associated with PAHs?
Response: The current approach is based on narcosis, and does consider things other than
lethality to the extent that those sublethal endpoints are induced through the
narcosis mode of action. Other endpoints could be included if we had an alternate
toxicological model. The question is: "If a sediment is clean enough to pass this
test, will it be clean enough for fish?" Understanding the effects of sediment
PAHs on fish also requires dealing with metabolism, which we ignore for
invertebrates.
Question: Is there a specific set of PAHs that was used?
Response: All PAHs that are present contributed to narcosis, but we do not measure every
one of them. "Total PAH" refers to several different commonly measured
84
-------
US Environmental Protection Agency
Region/ORD Workshop on Aquatic Life Criteria December 4-7, 2001
subsets. The specifics are included in the PAH document, which also includes an
uncertainty analysis based on which subset is used. Ideally, however, a site-
specific correction factor should be developed for each case.
Question: Can this approach be used for mercury?
Response: No, it does not apply to methyl mercury. Inorganic mercury does behave in this
way, however it is not normally a concern.
Question: For the PAH data, was a correction factor applied to convert from the thirteen
compounds to total PAH?
Response: No, it was not.
85
-------
US Environmental Protection Agency
Region/ORD Workshop on Aquatic Life Criteria December 4-7, 2001
Comparing WQC to Site-Specific Ecological Risk Assessment Results at R9
$fund Sites - Ned Black and Clarence Callahan (Region 9)
Several case studies were presented to illustrate risk assessment and risk management of
Superfund sites in region 9. The eight-step ecological risk assessment process for Superfund
was outlined, consisting of screening and problem formulation, study design, field sampling
design verification, site investigation and data analysis, followed by risk characterization and
risk management (slides 6,7,8). Hazard quotients are calculated by dividing the on-site
concentration of a contaminant by a literature-based estimate of the toxicity to a particular
receptor. A hazard quotient value greater than one indicates the potential for ecological harm to
the receptor. Some of the toxicity benchmarks used are ambient water quality criteria, sediment
effects range, industry soil screening levels and local toxicity reference values.
In Mansfield Canyon, Arizona, EPA was asked to check for presence of arsenic in small pools
from old mine shafts and adits. Arsenic levels found were well below the concentrations
protective of aquatic life. Since an endangered species of bats lives in the region, however,
water quality criteria were compared to the observed levels of arsenic; arsenic concentrations
were above those criteria. Further studies revealed that the endangered bats did not use these
pools for drinking water, as the pools were too small for the bats, who skim water while in flight.
The possibility that bats may feed on insects which grow in the pools was considered but not
pursued, as EPA involvement ended at that time.
A risk assessment was undertaken for creeks in McClellan Air Force Base near Sacramento,
California. Water samples from the creeks had detections for cadmium and PCBs - with the
PCBs stemming from a recently begun removal action. Cadmium levels were well below
chronic criteria, and PCB levels approaching, but still below such criteria. Sediments from the
same reach were also tested and found to be toxic to the amphipod Hyallela. Sediment toxicity
in this case, however, could not be correlated to any contaminant gradient. Work is continuing
to identify the toxicity drivers.
Mather Air Force Base, also near Sacramento, California, was considering a removal of
sediments from creeks contaminated with chlorinated pesticides. EPA was consulted regarding
how much sediment should be removed. The regional water board's recommendations were
based on ambient water quality criteria for maximum concentration of dieldrin. Sediments were
treated as landfill, and the water board converted the instant maximum of 2.5|ig/L to a sediment
concentration of 2.5|ig/kg. The Air Force did not accept this suggestion, citing a chironomid
assay showing the no observable adverse effect concentration (NOAEC) to be 7.5|ig/kg - an
opinion initially shared by EPA. Further consultation between EPA and the California
Department of Fish and Game, however, resulted in an agreement that NOAEC should be
86
-------
US Environmental Protection Agency
Region/ORD Workshop on Aquatic Life Criteria December 4-7, 2001
3.3|ig/kg. Once EPA and the California Department of Fish and Game presented a united
opinion, the Air Force and water board agreed to the 3.3|ig/kg concentration.
Leviathan Mine is another Superfund site in California, where contamination of a nearby stream
was a result of evaporation ponds which had overflowed on twelve of the fifteen years following
their construction. The responsible parties agreed to try lime neutralization of the ponds
followed by the release of treated water to drain ponds. The short field season in the area made a
quick decision necessary on what the effluent limitations should be. Chronic fresh-water criteria
concentrations for arsenic, copper and nickel were used, and the ponds have been drained
successfully in two out of the past three years.
In a risk characterization performed at the McCormick and Baxter Superfund site, NOAECs
were calculated for sediment pore water from chronic criteria concentrations using equilibrium
partitioning (for various PAHs). The concern in this case was that values derived from water
quality criteria would be too low for any Superfund site. NOAECs derived from these criteria,
however, were higher than those from any other derivation - although the slough would still
present a risk, regardless of the NOAECs.
The final case presented concerned an extensive ecological risk assessment in Lauritzen
Channel, near San Francisco Bay, contaminated with dichloro-diphenyl-trichloroethane (DDT).
The risk assessment involved equilibrium partitioning, invertebrate bioassays, site tissue
analysis, and laboratory bioaccumulation tests. Remediation goals were set as a result for total
DDT and dieldrin concentrations in the water column, although subsequent monitoring has
shown the removal action was not completely successful.
Conclusions
Agreements between ecological risk assessment results and water quality criteria is
variable;
Water concentrations are often not measured at Superfund sites; sediments are measured
instead;
Modeling from sediments to water works better than was expected.
Questions and Comments
Comment: It has become obvious as we talk about setting criteria that these criteria have
many different uses across the board. Superfund is an example, as those sites are
impacted in different ways than watersheds. In developing new criteria, I hope
that there will be consideration of their uses in different programs.
87
-------
US Environmental Protection Agency
Region/ORD Workshop on Aquatic Life Criteria December 4-7, 2001
Comment: Superfund used to be good at taking National standards and coming up with a
number appropriate to a site.
Response: We often have site-specific data, and they do not always agree with the National
standards - however they do sometimes agree more than would be expected.
Comment: The threshold numbers for sediments could be used to help focus investigations,
and to make risk-based recommendations where large monetary decisions need to
be made.
88
-------
US Environmental Protection Agency
Region/ORD Workshop on Aquatic Life Criteria December 4-7, 2001
Persistent Bioaccumulative Toxicants ~ Philip M Cook (ORD/NHEERL)
A conceptual model was provided to show how the ecological risk assessment framework can be
incorporated into water and sediment quality criteria development for chemicals whose effects
are related to residues in tissues in the organisms at risk or their diet. The multidirectional
flowchart links human health, fish and wildlife populations, and ecological community
conditions to the potential loadings of toxicants into an environment through a series of data
types and the models which link them. An important example is the use of bioaccumulation
factors and food chain models to determine concentrations of bioaccumulative chemicals in
water and sediment that correspond to residues in organisms associated with toxicity.
The persistence characteristic of persistent bioaccumulative toxicants (PBTs) was noted to
influence:
PBT accumulation in sediments and food webs;
Chemical transport over long distances and between ecosystems;
Reductions in the magnitude of temporal variations in exposures;
Increased probability of exposures to vulnerable life stages, populations and
communities; and
Long time frames required for natural attenuation.
Potential effects of PBT exposure were illustrated with a salmonid fry example, which showed
cranial-facial malformations, hemorrhaging, yolk sac edema, etc. as a result of exposure to
PCB126. The mortality rate of embryonic fish was shown to increase dramatically with an
increase in concentration of PCB126. These observations are the same as those reported for
2,3,7,8-tetrachlorodibenzo-p-dioxin except that the dose required for PCB 126 is approximately
two hundred times greater.
Cook reviewed the process by which bioaccumulation factors (BAFs and BSAFs) are calculated
for water and sediments. The results of an eighty percent drop in chemical loading to an
ecosystem on relative concentrations in water and sediments was reviewed, showing a very
quick associative drop in water, while the response in sediments was gradual over the course of
many years.
Cook displayed graphs which illustrated how critical residue values for effects of a PBT on trout
and bald eagles relate to specific concentrations in the sediments and water in different
ecosystems. The disequilibrium between sediments and overlying water was explained, where
differences in sediment and water concentrations are calculated, and the structure of the food
chain of an ecosystem is taken into consideration when determining the acceptable levels of
89
-------
US Environmental Protection Agency
Region/ORD Workshop on Aquatic Life Criteria December 4-7, 2001
water and sediment concentrations. A food chain that relies more on benthic versus pelagic
sources will result in different acceptable concentrations than one in which the pelagic organisms
factor higher than the benthic organisms. These factors are further influenced by the rates of
metabolism of the PBT in the receptor organisms involved. The five major determinants of
bioaccumulation were listed as:
Hydrophobicity of the chemical (as measured by Kow, the octanol-water partition
coefficient);
Metabolism of the chemical in the food chain and the receptor organism;
n^, the sediment-water concentration quotient;
Trophic level of the receptor organism; and
Fraction of the food web that is benthic versus pelagic.
Equations were presented for the method by which a BAF for a species can be calculated from a
field-measured BSAF and simple hybrid bioaccumulation model which incorporates food chain
model predictions with relative metabolism ratios determined from measured BSAFs. Toxicity
equivalence factors (TEFs) applied to mixtures of dioxin-like chemicals in an organism's tissues
allow calculation of a toxicity equivalence concentration (TEC) that should not exceed a 2,3,7,8-
tetrachlorodibenzo-p-dioxin (TCDD) Residue Quality Criterion (RCQtcdd). The WHO Toxicity
Equivalence Factors for twenty-nine congeners were reviewed as they apply to mammals, fish
and birds. The tiering scheme for the determination of Relative Potency (REP) Values indicates
that Tier 1 values from the endpoint species in question are the best source for Ecological Risk
Assessments.
The history of Lake Ontario lake trout toxicity risks was based on exposure of eggs as expressed
by TECeggs in dated sediment layers derived from concentrations of dioxin-like chemicals and
the respective TEFs. Peak concentrations of TECeggs from the 1950s through the 1970s coincide
with a one hundred percent mortality rate in fry from either native or stocked lake trout over the
same time frame. The reproductive success of Lake Ontario bald eagles was also predicted to
follow a similar historical TCDD toxicity induced mortality pattern but with PCBs making a
greater contribution to the toxicity than for fish.
Cook gave a description of frameworks used to calculate water and sediment criteria:
Framework for calculation of a water quality criterion for the protection of lake trout
from the effects of TCDD and related chemicals;
Sample calculation of a safe (Cd)tcdd for the protection of lake trout from the effects of
TCDD and related chemicals;
Framework for calculation of a sediment quality criterion for the protection of lake trout
from the effects of TCDD and related chemicals; and
90
-------
US Environmental Protection Agency
Region/ORD Workshop on Aquatic Life Criteria
December 4-7, 2001
Sample calculation of a safe (Csoc) for the protection of lake trout from the effects of
TCDD and related chemicals.
A Waste Load Allocation Model using a TMDL TEQ-SQC approach for mixtures of TCDD and
related chemicals was displayed, indicating the flow of relevant factors and calculations between
the models to the acceptable contamination level. Each dioxin-like chemical in the mixture has
to be mass balance modeled in order to relate loadings to specific exposures of organisms that
result in the TEC.
Questions and Responses
Comment: Doug Norton said the other day that we were missing out on restoration
opportunities.
Response: Research on a problem like Lake Ontario lake trout reproduction is a slow,
deliberate process. And even so, this was a particularly convenient investigation
because dioxin exposures were so great. With PBTs it is difficult to make a
prediction based on a hypothesis involving a singular cause and effect and then
validate the prediction with environmental data which must be interpreted from a
changing multi-stressor perspective; it may take a decade.
Question: Do you have riverine systems that you have applied this to?
Response: I have reservations about the applicability of monitoring data which may not be
adequate for measuring bioaccumualtion. The kind and quality of data that needs
to be collected needs to be better defined for assessors.
Question: What are the monitoring needs?
Response: When we collect fish samples, we should collect surface sediment samples that
are related to them. It is also important for any samples to be properly collected.
This would allow a BSAF data base to be formed that would help in providing
cross ecosystem extrapolations.
Question: Where can you get a published copy of the graphs showing temporal response of
the ratio II/Kow?
Response: We are presently submitting a paper for peer review, so we should be able to send
out copies of that paper soon. We are better off if people understand what it is
they are doing when using this model. At times these concepts use are more
likely to publish tables of Pow, food chain loads, etc.
Question: Do you have any plans to do predictions, and then go out and try the model after a
restoration has been done?
91
-------
US Environmental Protection Agency
Region/ORD Workshop on Aquatic Life Criteria December 4-7, 2001
Response: We have limited resources for doing that sort of thing. We are still missing
effects information. There is less than one order of magnitude of uncertainty
when it comes to protecting lake trout. We would need to wait for enough time to
go by before we are able to test our prediction.
Question: Are you still data-driven?
Response: We have been trying to get a supreme data set for Lake Michigan for years now
and are finally nearing completion of a minimum data set.
Question: If people send you data can you help them understand it?
Response: We try to do that all the time.
92
-------
US Environmental Protection Agency
Region/ORD Workshop on Aquatic Life Criteria December 4-7, 2001
Toxic Chemicals Session: Assessing Risks to Wildlife - Rick Bennett
(ORD/NHEERL)
Rick Bennett (NHEERL) provided an overview of the wildlife criteria development project,
which arose from concern that ambient water quality criteria would not be sufficient to protect
the wildlife who feed upon aquatic biota. Wildlife Values (WV) are based upon the highest
concentration within the water column that will not harm wildlife which use the water for
drinking and foraging. The criteria used to determine harm to wildlife are tracked over multiple
generations, and include:
Significant reduction in growth;
Significant reduction in reproduction; and
Significant reduction in viability or usefulness.
Bennet reviewed the equation by which wildlife values are calculated, and provided descriptions
of the variables involved in the equation [Slide 4].
The Great Lakes Water Quality Initiative (GLWQI) has selected several representative species:
otter, mink, bald eagle, herring gull, and belted kingfisher; and several chemicals: DDT,
mercury, PCBs, and 2,3,7,8-TCDD. The model parameterization process was explained, which
included:
Development of guidance for literature review of the most appropriate test dose;
Development of guidance for determining uncertainty factors (UF);
Guidance for determining chemical specific BAFs and BMFs; and
Guidance for determining species-specific data such as body weight and diet.
Bennett reviewed the tiered approach to establishing the GLWQI criterion. Tier I wildlife
criteria establish a geometric mean of wildlife values for the birds and mammals represented,
with the criterion being set as the lower of the two. Uncertainty factors in tier one can range
from one to one hundred. Tier II wildlife criteria establish a geometric mean of representative
species for either birds or mammals, and uncertainty factors can range from one to one thousand,
for a taxonomic group without data. Study duration for tier I is 90 or more days for mammals,
and 70 or more days for birds, while tier II study duration for mammals and birds is 38 or more
days.
93
-------
US Environmental Protection Agency
Region/ORD Workshop on Aquatic Life Criteria December 4-7, 2001
Bennett identified several uncertainties in regards to the establishment of wildlife criteria, which
include:
Limitations of the toxicity database;
Variability in measured BAFs; and
Limitations on life history information.
The mercury Report to Congress included modifications to the GLWQI model. The herring gull
was replaced with the osprey and the common loon, because these species have a wider
distribution. Bennett listed site-specific modifications, the inclusion of tissue-based criteria, and
the development of methods to understand risks of multiple stressors as the next steps in wildlife
criteria development.
94
-------
US Environmental Protection Agency
Region/ORD Workshop on Aquatic Life Criteria December 4-7, 2001
Derivation of New Jersey-Specific Wildlife Values as Surface Water Quality
Criteria for: PCBs, DDT, and Mercury - Dana Thomas and Dan Russell (U.S. Fish
and Wildlife Service)
Dana Thomas (Region 2) and Dan Russell (USFWS) provided an overview of the New Jersey
Wildlife Criteria effort, a cooperative undertaking with participation from EPA, USFWS, and the
New Jersey Department of Environmental Protection. The goal of this effort was to derive New
Jersey-specific surface water quality criteria for the protection of wildlife, using the EPA's Great
Lakes Water Quality Initiative (GLWQI) methodology. The contaminants of concern for this
project were polychlorinated biphenyls (PCBs), dichlorodiphenyltrichloroethane (DDT), and
mercury.
The technical committee conducting the criteria derivation followed the GLWQI's Tier II
methodology, which generates Wildlife Values using only one wildlife taxon. New Jersey
Wildlife Values were generated using data for bald eagles and peregrine falcons, as contaminant
risks to these avian species were the impetus behind this effort. The osprey was used as the
third representative species in the Wildlife Value calculations. Adoption of the Tier II
methodology must provide assurance that the taxonomic class not considered is protected by the
calculated Wildlife Values.
After an extensive literature review, it was determined that the same avian test dose studies used
in the GLWQI to calculate wildlife criteria would be used for the New Jersey Wildlife Values.
Similarly, the same uncertainty factors used in the GLWQI for PCB and DDT calculations were
used for the New Jersey Wildlife Values. For mercury, Wildlife Values were calculated using
the uncertainty factors from the EPA's 1997 Mercury Study Report to Congress (MSRC).
Chemical-specific bioaccumulation factors (BAFs) for all three contaminants were revised from
the original GLWQI by the EPA, based on subsequent methodology re-evaluations. For the New
Jersey Wildlife Value effort, the BAFs for PCBs and DDT were further revised by using New
Jersey-specific dissolved (DOC) and paniculate (POC) organic carbon data. For mercury, BAFs
from the MSRC were used in the calculations.
Various exposure parameters were determined for each representative species, including adult
body weights, water ingestion rates, and food ingestion rates (defined for each trophic level prey
type consumed). For peregrine falcons, the diet composition of piscivorous and non-piscivorous
prey was determined by a study of New Jersey nest site prey remains. The trophic level
composition of osprey diet was determined based on New Jersey's physical geography
(prevalence of shallow lakes and ponds) and observations of prey capture (largemouth bass,
chain pickerel).
95
-------
US Environmental Protection Agency
Region/ORD Workshop on Aquatic Life Criteria
December 4-7, 2001
The GLWQI methodology calculated chemical-specific geometric means for each taxon, Aves
and Mammalia, and used the lower of these values in setting regulatory water quality criteria.
The technical committee also calculated chemical-specific geometric means, using the single
Aves taxon, but determined that the resulting Wildlife Values would not be stringent enough to
protect all wildlife species of concern. For each contaminant evaluated, the calculated geometric
mean was greater than the species-specific value necessary to protect the peregrine falcon. As
the focus of this effort was to develop water quality criteria protective of bald eagles and
peregrine falcons, the technical committee recommended the peregrine falcon-specific Wildlife
Values generated for each of the three contaminants be adopted as regulatory criteria.
The resulting New Jersey Wildlife values are lower than those generated for the GLWQI, and
have been recommended to the New Jersey Department of Environmental Protection for
adoption as regulatory criteria.
GLWQI Wildlife Criteria
New Jersey Wildlife Values
PCBs
120 pg/L
72pg/L
DDTr
11 pg/L
4 pg/L
Mercury
13 00 pg/L
530 pg/L
Questions and Responses
Question: Did you include TEF or TEQ?
Response: No, we did not include an analysis for TEFs or TEQs in our PCB calculations.
We were only using the methodology for total PCBs, not specific congeners or
Arochlors.
Question: Did this include brackish water?
Response: No, this is just for fresh water.
Question: I wanted to first thank the presenters, because there has been concern on whether
the GLWQI was being used by other people to come up with guidelines. Did you
include mussels in the formulation of criteria?
Response: No, we did not include mussels in the formulation of the Wildlife Values. The
technical committee felt that the danger of exposure by mussels to the
contaminants of concern should be less than with the higher trophic level
organisms examined.
96
-------
US Environmental Protection Agency
Region/ORD Workshop on Aquatic Life Criteria
December 4-7, 2001
Question: How long [did the study take]?
Response: The technical committee first convened in April 1999. The final Wildlife values
were determined by the end of 2000, and the final report was completed in
September 2001.
Question: Does New Jersey have the document now?
Response: The document is going to be published in the New Jersey State Register. They
will have until October to provide an answer, but will likely ask for more time.
Response: New Jersey agrees with the number but there are concerns about implementation.
Question: What impact would this have on decisions about hazardous waste, or the
Superfund program?
Response: We haven't given it much thought beyond this first step. The goal was to develop
water quality criteria for these contaminants that would be protective for bald
eagles and peregrine falcons.
Question: Did you ever go through a process where you considered fish as a receptor with
regard to PCB?
Response: No, we focused on raptors - peregrine falcon, bald eagle, and osprey. The effort
began because of the threat to listed species - the peregrine falcon and bald eagle.
Because of the bioaccumulative nature of these contaminants, the technical
committee felt that higher trophic level predators were more at risk than fish.
Question: Uncertainty factor - lowest number is one - should we use a number less than one
for more sensitive species?
Response: This depends on the uncertainty factor referred to (i.e., interspecies, low
observable adverse effect level [LOAEL] to no observable adverse effect level
[NOAEL], subchronic-to-chronic). Generally, these factors don't address
differences in organism sensitivities. But if you wanted to adjust the interspecies
uncertainty factor because your species of concern is more sensitive than the
species used to determine a test dose, then the uncertainty factor should be higher,
not lower.
Question: Where does it account for amphibians and reptiles (other than fish birds and
mammals)?
Response: We did not take amphibians or reptiles into account when developing these
values. The focus was on threats to listed piscivorous birds (peregrine falcons
and bald eagles).
Comment: They were not listed at that time. There are new species that are listed. It is a
dynamic document.
97
-------
US Environmental Protection Agency
Region/ORD Workshop on Aquatic Life Criteria
December 4-7, 2001
Question: Can you use this methodology on organisms other than birds and mammals?
Response: Yes, [we can].
Comment: We need to go back and refine/re-tune the database.
Question: What are the background concentrations in these areas, or in a pristine area?
Response: We don't know, those areas are harder and harder to find in New Jersey.
Question: What are the piscivorous birds in the peregrine falcon's diet?
Response: Wading birds, not exclusively piscivorous birds, occasionally some gull remains.
Comment: Gulls in San Francisco had almost no mercury in their systems. Are the birds
eating KFC? Gulls do tend to scavenge refuse, particularly those living near
urban areas.
Question: How did you come up with criteria in the range of 500 picograms? This is a very
low, almost unmeasurable criterion.
Response: It is very hard to measure and very expensive as well. We expect a lot of
complaints because it costs between one and two thousand dollars per test.
Comment: Because these numbers are so low, we will need to ground truth these. They are
the best numbers we can come up with at this time.
Question: How comfortable are you with your numbers?
Response: We are very comfortable with the numbers, based on the information presently
available. However, we recognize that data gaps exist, and filling these gaps
would allow for further refinement of the values...which would in turn increase
the comfort level.
98
-------
US Environmental Protection Agency
Region/ORD Workshop on Aquatic Life Criteria December 4-7, 2001
NHEERL Wildlife Research Demonstrations Project: Methods to Assess Risks
to Piscivorous Bird Populations - Rick Bennett (ORD/NHEERL)
Rick Bennett (NHEERL) indicated that there are questions which arise from assessments based
upon single value criteria. The Loon Demonstration Project was developed with the objectives
of:
Developing approaches for assessing the risks to wildlife populations from multiple
stressors by integrating information from population biology, toxicology, and landscape
ecology; and
Using demonstration projects to apply these approaches to specific agency issues, while
considering the possibility of extrapolation to other issues.
Bennett indicated that the availability of rich data sets on life histories for loons, from a number
of agencies, universities and private conservation groups was a factor in the criteria for selection
of this project. Planning activities have included an expert seminar and planning meeting, and
the submission of a draft planning document to the aquatic stressors planning committee, and the
upcoming development of work plans through the aquatic stressors planning committee.
Bennett talked about the organization of research activities into the following inter-dependent
steps:
Geographically referenced data collection;
Stressor response modeling;
Population modeling, population genetics; and
Spatial modeling.
Under the conceptual approach, landscape characterization data and exposure data are used to
define the exposure-response and habitat-response stressor effects relationships. This data is fed
into the population models for the creation of spatially-explicit models. Efforts in landscape
characterization have included the development of a data management system based upon
EMAP design. Stressor-response relationship characterizations have focused upon the effects of
mercury on the common loon, and the effects of habitat alterations and environmental factors
upon the common loon. Dose-response effects of mercury on the reproduction of the common
loon will be estimated through the development of physiologically-based toxicokinetic (PBTK)
models to extrapolate effects from laboratory studies with American kestrels. NHEERL is
cooperating with the state and federal agencies and loon monitoring groups to create quantitative
relationships between environmental stressors and the survival and productivity of the common
loon. Research is underway to develop genetic markers for the common loon and other avian
99
-------
US Environmental Protection Agency
Region/ORD Workshop on Aquatic Life Criteria
December 4-7, 2001
species, and will contribute to the creation of simple and more realistic population models in a
spatial context for avian species.
Questions and Responses
Question: How many more resources would it have taken to look at more chemicals and
more bird species in this project?
Response: With the EMAP-based data management system, it would have been relatively
easy to include information layers on other chemicals and species when
compatible information exists. We are trying to do that now without losing sight
of the specific project.
Question: The actual response of toxicity of mercury may not necessarily be affecting the
embryo, it could be maternal effects (reproductive, rather than developmental)?
Response: We are looking at parental effects as well as embryo effects. There are numerous
steps in the life history where effects can originate, we cannot just study one of
these stages.
Response: There are many ways in which mercury can affect animals just in reproduction.
For example, it can affect male brooding behavior - there is no way to study such
an effect in a laboratory setting.
Comment: There was a three-generation duck study that found effects in duckling response
to maternal calling, and effects on flight feather growth - with aerodynamic flight
loss of about ten percent.
Question: Can you see doing work like this from existing environmental data without the
need for modeling?
Response: Many of the species we are interested in cannot be studied in the laboratory
setting, but we can do field studies. Breeding them successfully in the lab is not
feasible because of the inherent stress of captivity.
Comment: We look at parental effects and embryo effects. Juveniles have to feed in lakes
where they hatch. What if the lake is contaminated? A whole suite of
stages/periods in life stages with various factors is involved.
100
-------
US Environmental Protection Agency
Region/ORD Workshop on Aquatic Life Criteria December 4-7, 2001
Appendix A: Agenda and List of Posters
Appendix B: List of Participants
Appendix C: Slides from Presentations
Appendix D: Plenary Flip Chart Notes
Appendix E: Breakout Sessions Flip Chart Notes
Appendix F: Workshop Participant Evaluation Summary
101
-------
US Environmental Protection Agency
Region/ORD Workshop on Aquatic Life Criteria December 4-7, 2001
THIS PAGE INTENTIONALLY LEFT BLANK
102
------- |