EPA
United States
Environmental Protection
Agency
Aquatic Stressors
framework and implementation
plan for effects research
1 iğ .V.
Nati
-------
EPA 600/R-02/074
September 2002
Aquatic Stressors
framework and implementation plan for effects research
U.S. Environmental Protection Agency
Office of Research and Development
National Health and Environmental Effects Research Laboratory
Research Triangle Park, NC 27711
-------
Foreword
The National Health and Environmental Effects Research Laboratory (NHEERL), as part of the
Environmental Protection Agency's (EPA's) Office of Research and Development (ORD), is
responsible for conducting research on the effects of anthropogenic stresses on human and
ecosystem health. This research is intended to address key Agency problems in a timely and
responsive manner. To meet this responsibility, NHEERL is developing research
implementation plans to achieve the following objectives:
Optimize responsiveness of research activities to Agency needs,
Sharpen the focus of research programs where needed,
Provide a forum for engagement of scientific staff on issues and approaches,
Focus on multi-year planning explicitly linked to Agency performance goals, and
Provide a mechanism for prioritizing research.
This approach builds on the ORD planning process that identifies and prioritizes research topics.
Current areas for research include particulate matter, air toxics, drinking water, aquatic stressors,
support to the Food Quality Protection Act, safe communities and ecosystems, the Environmental
Monitoring and Assessment Program, ecological risk assessment, human health risk assessment,
and endocrine disrupters.
This document identifies the scientific problems and the research that will be conducted
concerning aquatic stressors. The ultimate goal of this research is to develop scientifically valid
approaches for protecting the ecological integrity of aquatic ecosystems from multiple aquatic
stressors, in support of EPA's goal to provide clean and safe water. The framework section was
developed by a steering committee composed of representatives from NHEERL Divisions, other
ORD Laboratories and Centers, EPA's Office of Water, and EPA's Regional Offices.
Implementation plans for research on habitat alteration, nutrients, suspended and bedded
sediments, toxic chemicals, and diagnostics were developed by work groups in the Ecology
Divisions. The document is intended to reflect research that will be conducted over the next
several years. As progress is made in achieving these goals, this document will be updated to
address new and remaining water quality challenges.
Lawrence W. Reiter
Director
National Health and Environmental Effects Research Laboratory
11
-------
Abstract
This document describes the framework and research implementation plans for ecological effects
research on aquatic stressors within the National Health and Environmental Effects Laboratory.
The context for the research identified within the framework is the common management goal of
protecting aquatic systems to prevent degradation of habitat, loss of ecosystem function, and
reduced biodiversity. Five main research products for meeting this goal are identified in the
framework: 1) methods to predict biological effects of habitat alteration; 2) population,
community, and ecosystem stressor-response models; 3) diagnostic tools to determine
impairment or causes of impairment to aquatic systems; 4) classification approaches to aid in the
prediction and management of problems; and 5) methods and models to support development of
ecological criteria. The research implementation plans herein focus on the effects of four aquatic
stressors, including habitat alteration, nutrients, suspended and bedded sediments, and toxic
chemicals. This approach is consistent with recent scientific consensus, recognizing that these
stressors have the greatest potential for causing adverse effects to aquatic ecosystems. In
addition, the document outlines research that will develop diagnostic tools for a decision support
system for resource managers. The major goals, the critical path for research, specific research
projects, and a gap analysis are provided for each of the five research implementation plans along
with the time table of research products that will support EPA's Goal 2 (Clean and Safe Water)
research under the Government Performance and Results Act.
in
-------
Table of Contents
Foreword ii
Abstract iii
Acknowledgments viii
Glossary xi
Executive Summary xiv
Section 1.
Introduction 1
Purpose and Scope 1
Programmatic Needs 2
References 2
Section 2.
Research Approach 3
Context for Research 3
Research Process 3
Decision Support System 5
Section 3.
Research Products and Implementation Plans 8
Section 4.
Implementation Plan for Habitat Alteration Research 10
Problem 10
Goals 12
Critical Path 14
Research Projects 17
Gap Analysis 37
References 38
Section 5.
Implementation Plan for Nutrients Research 40
Problem 40
Goals 42
Critical Path 44
Research Proj ects 49
Gap Analysis 62
References 62
IV
-------
Table of Contents (continued)
Section 6.
Implementation Plan for Suspended and Bedded Sediment Research 66
Problem 66
Goals 67
Critical Path 68
Gap Analysis 75
References 75
Section 7.
Implementation Plan for Toxic Chemicals Research 77
Problem 77
Goals 78
Critical Path 79
Research Proj ects 95
Gap Analysis 125
References 130
Section 8.
Implementation Plan for Diagnostics Research 138
Problem 138
Goals 139
Critical Path 141
Research Projects 146
Gap Analysis 165
References 169
Appendix 1 174
Appendix 2 179
-------
Figures
Figure 1. Research process and products for meeting the goal of effective management and
protection of aquatic resources 4
Figure 2. Manager's decision support system to protect and restore aquatic resources using
ORD's research products (SI box is modified from EPA 2000b) 6
Figure 3. Critical path for habitat alteration research (APGs) refer to those listed and described
in the Goals subsection) 15
Figure 4. Components of coastal vegetated habitat with possible pathways for direct and
indirect effects of habitat alteration on fish, shellfish, and wildlife 21
Figure 5. Critical path for research on the development of nutrient response relationships for
coastal receiving waters 45
Figure 6. Conceptual diagram of the feedbacks among data mining, model development, field
monitoring, and experimental hypothesis validation 55
Figure 7. Critical path for suspended and bedded sediments research 69
Figure 8. Ecological risk assessment framework (modified from EPA 1992) 80
Figure 9. Critical path for developing site-specific methodologies for establishing the risks of
toxic chemicals to aquatic life and aquatic dependent wildlife 81
Figure 10. Simple conceptual model for risk assessments of nonbioaccumulative toxicants. . . 84
Figure 11. Conceptual model for risk assessments and criteria development involving
determination of safe loadings of bioaccumulative toxicants to aquatic systems. . . 91
Figure 12. Critical path (flow of APGs) for diagnostics research 142
Figure 13. A logic for characterizing the causes of ecological injuries at specific sites. Modified
from Figure 4-1 in SI document (EPA 2000c) to show potential inputs from aquatic
stressors diagnostics research 144
Figure 14. Relationship between current stages of State/Tribal assessment, TMDL and watershed
restoration planning processes, and proposed combined path 145
Figure 15. Locations of national water-quality assessment study units 159
Figure 16. Conceptual model of cause-and-effect relationships in coastal systems, providing a
framework for a decision support system. See key to model components at base of
figure. Loading terms include atmospheric component 163
VI
-------
Tables
Table 1. Time line for habitat alteration research 16
Table 2. List of candidate species for study in marine and Great Lakes coastal regions 18
Table 3. Preliminary list of factors influencing response to excess nutrient inputs in coastal
receiving waters 46
Table 4. Existing or proposed approaches to classification at regional, watershed, water-body,
and habitat scales 150
Table 5. Examples of methods incorporated in conceptual framework (Figure 16) for decision
support system 164
Table 6. Single aquatic stressors method development covered by other research areas within the
Aquatic Stressors Framework 166
Table 7. FTE resource allocation for diagnostics by year (FW = freshwater, SW = saltwater)
168
vn
-------
Acknowledgments
Authors
The principal authors of Sections 1-3 (framework for aquatic stressor research) are members of
the National Health and Environmental Effects Laboratory's Aquatic Stressors Steering
Committee, which is composed of representatives from the Office of Research and
Development's (ORD) Laboratories and Centers, Office of Water (OW), and Regional offices.
Principal authors for Sections 4-8 (Research Implementation Plans for Habitat Alteration,
Nutrients, Suspended and Bedded Sediments, Toxic Chemicals and Diagnostics) are members of
work groups from NHEERL's four Ecology Divisions: Atlantic Ecology Division, Narragansett,
RI; Gulf Ecology Division, Gulf Breeze, FL; Mid-Continent Ecology Division, Duluth, MN; and
Western Ecology Division, Corvallis, OR (see below).
NHEERL Aquatic Stressors Steering Committee
Mid-Continent Ecology Division (MED)
Bob Spehar (Chair)
Jack Kelly (Co-chair)
Atlantic Ecology Division (AED)
Jonathan Garber/Steve Schimmel
Walter Berry
Gulf Ecology Division (GED)
Bill Walker
Skeet Lores
Western Ecology Division (WED)
Walt Nelson
Spence Peterson
Neurotoxicology Division (NTD)
BobMacPhail
Experimental Toxicology Division (ETD)
Bob Luebke
NHEERL Assistant Laboratory Director
Jennifer Orme Zavaleta/Barbara Walton
National Exposure Research Laboratory/Ecological Exposure Research Division (EERD)
Susan Cormier
Vlll
-------
Acknowledgments (continued)
National Risk Management Research Laboratory/Subsurface Protection and Remediation
Division (SPRD)
Steve Schmelling
National Center for Environmental Assessment/Exposure Analysis and Risk Characterization
Group (EARCG)
Sue Norton
OW/Health and Ecological Criteria Division (HECD)
BillSwietlik
Regions/Region 4
Joel Hansel
NHEERL Aquatic Stressors Work Groups
Habitat Alteration
Cathy Wigand, Giancarlo Cicchetti (AED)
Rich Devereux, John Macauley (GED)
John Brazner, Anett Trebitz (MED)
Bob Lackey (Chair), Jim Wigington, Jim Power (WED)
Nutrients
Jim Latimer (AED)
Sheet Lores (Chair), Rick Greene, MichaelMurrell (GED)
Jo Thompson, John Morrice, Jack Kelly, Russ Kreis (MED)
Pete Eldridge, Robbins Church (WED)
Suspended and Bedded Sediments
Walter Berry (AED)
Bill Walker, Mike Lewis (GED)
Danny Tanner, Mike Sierszen (MED)
Steve Paulsen (Chair), Phil Kaufmann, Mark Johnson (WED)
Toxic Chemicals
Walter Berry, MattMitro, Diane Nacci (AED)
Larry Goodman, Michael Hemmer (GED)
Russ Erickson (Chair), Bob Spehar, Rick Bennett, Phil Cook (MED)
Jennifer Orme-Zavaleta (WED)
Diagnostics
Skip Nelson, Dan Campbell, Rob Burgess (AED)
IX
-------
Acknowledgments (continued)
Virginia Engle, Jan Kurtz (GED)
Naomi Detenbeck (Chair), Larry Burkhard, Bill Richardson (MED)
Jennifer Orme-Zavaleta, Jana Compton (WED)
Reviewers
A preliminary draft of this document was reviewed by staff within the ORD Laboratories and
Centers, OW, and the Regions. A subsequent draft was reviewed in February 2002 by the
following reviewers outside of EPA:
Dr. Brian Bledsoe
Engineering Research Center
Colorado State University
Dr. Scott Dyer
The Procter and Gamble Company
Dr. Ken Rose
Coastal Fisheries Institute
Louisiana State University
Dr. Patricia Chow-Fraser
McMaster University
Dr. Scott Phillips
US Geological Survey
Water Resources Division
Dr. Keith Solomon
Center for Toxicology
University of Guelph
Dr. Jonathan Higgins
The Nature Conservancy
Dr. Paul McCormick
Everglades Program Team
Department of the Interior
Dr. Al Steinman
Annis Water Resources Institute
Lake Michigan Center
Dr. David Rudnick
South Florida Water Management District
Everglades Division
-------
Glossary
ACE Acute-to-Chronic Estimation
ACWI Advisory Committee on Water Information
AED Atlantic Ecology Division
AhR Aryl hydrocarbon Receptor
APG Annual Performance Goal
APM Annual Performance Measure
AQUIRE AQUatic toxicity Information REtrieval
AVS Acid Volatile Sulfide
B AF B i oaccumul ative Accumul ati on F actor
BASINS Better Assessment Science Interacting Point and Nonpoint Sources
BLM Biotic Ligand Model
BO Biological Opinion
BSAF Bi oaccumul ati ve Sediment Accumulation Factor
CADDIS Casual Analysis and Diagnosis Decision Information System
CENR Committee on Environment and Natural Resources
CFR Clark Fork River
CTR California Toxics Rule
CWA Clean Water Act
CWAP Clean Water Action Plan
DEM Digital Evaluation Model
DO Dissolved Oxygen
EC50 Effective Concentration (50%)
ECOTOX Ecological Toxicity Database
EDNA Elevation Derivatives for National Applications
ELS Early Life Stage
EMAP Environmental Monitoring and Assessment Program
EPT Ephemeroptera, Plecoptera, Trichoptera
EqP Equilibrium Partitioning
ESA Endangered Species Act
ESG Equilibrium-partioning Sediment Guideline
FGDC Federal Geographic Data Committee
FTE Full Time Equivalent
FWS Fish and Wildlife Service
GAP Gap Analysis Program
GC/MS Gas Chromatography/Mass Spectrometry
GED Gulf Ecology Division
GIS Geographic Information System
GLEI Great Lakes Environmental Indicators
GPRA Government Performance and Results Act
HAB Harmful Algal Bloom
HAPR Habitat Alteration Population-Response Model
HUC Hydrologic Unit Code
IBI Indices of Biotic Integrity
XI
-------
Glossary (continued)
ICE
Kow
LaMP
LC50
MAHA
MED
NAWQA
NCEA
NED-H
NEP
NERL
NHEERL
NMDS
NMFS
NOAA
NOAEL
NPDES
N/P
NPS
NRC
NRDA
NRMRL
OEI
OPP
OPPTS
ORD
OST
OW
OWOW
PAH
PBTK/TD
PBT
PCB
PCDD
PCDF
PHYTOTOX
REMAP
RFP
SAV
SI
SOLEC
SPARROW
SPRC
Interspecies Correlation Estimation
Octonal-water Partition Coefficient
Lakewide Management Plan
Lethal Concentration (50%)
Mid-Atlantic Highlands Assessment
Mid-continent Ecology Division
National Water Quality Assessment
National Center for Environmental Assessment
National Elevation Dataset (Hydrology)
National Estuarine Program
National Exposure Research Laboratory
National Health and Environmental Effects Research Laboratory
Nonmetric Dimensional Scaling
National Marine Fisheries Service
National Oceanographic and Atmospheric Administration
No Observed Adverse Effect Level
National Pollutant Discharge Elimination System
Nitrogen/Phosphorus
Nonpoint Source
National Research Council
Natural Resources Damage Assessment
National Risk Management Research Laboratory
Office of Environmental Information
Office of Pesticide Programs
Office of Prevention, Pesticides, and Toxic Substances
Office of Research and Development
Office of Science and Technology
Office of Water
Office of Wetlands, Oceans, and Watersheds
Polycyclic Aromatic Hydrocarbon
Physiologically-Based Toxicokinetic/Toxicodynamic
Persistent Bioaccumulative Toxicant
Polychorinated biphenyls
Polychorinated dibenzo dioxins
Polychorinated dibenzo furans
Terrestrial Plant Toxicity Database
Regional EMAP
Request for Proposals
Submerged Aquatic Vegetation
Stressor Identification
State of the Lake Ecosystem Conference
Spatial Referenced Regressions on Watersheds
Strategic Planning Research Coordination
xn
-------
Glossary (continued)
STAR Science to Achieve Results
STORE! STOrage and RETrieval database
TCDD 2,3,7,8-Tetrachlorodibenzo-p-dioxin
TERRETOX Terrestrial Wildlife Toxicity Database
TIE Toxicity Identification Evaluation
TMDL Total Maximum Daily Load
USGS US Geological Survey
UV Ultraviolet
WATERS Watershed Assessment Tracking and Environmental Results System
WED We stern Ecol ogy Di vi si on
WQC Water Quality Criteria
WQS Water Quality Standards
WRS Wildlife Research Strategy
Xlll
-------
Executive Summary
This document describes the framework (Sections 1-3) and implementation plans (Sections 4-8)
for ecological effects research on aquatic stressors within the U.S. Environmental Protection
Agency's (EPA) National Health and Environmental Effects Research Laboratory (NHEERL).
The ultimate goal of this research is to develop scientifically valid approaches for protecting and
restoring the ecological integrity of aquatic ecosystems from the impacts of multiple aquatic
stressors. The immediate focus is to develop and improve assessment methodologies, diagnostic
capabilities, and ecological criteria to guide management options for 1) protection and restoration
and 2) remediation efforts to meet designated uses.
The context for this research is the common management goal of protecting aquatic systems to
prevent degradation of habitat, loss of ecosystem function, and reduced biodiversity. To this end,
environmental managers must be able to: 1) assess the condition of an aquatic resource and
determine the degree of impairment, 2) diagnose the causes of impairment, 3) forecast the effects
of changes in stressor levels, and 4) develop and implement remediation and maintenance
strategies. Meeting the goals of effective management and protection of aquatic resources
requires that multiple research elements be in place to provide the needed tools. This document
provides a means to develop these tools. The research approach presents a generally linear
sequence, although many research elements will be conducted simultaneously.
The research herein focuses on the effects of aquatic stressors, including habitat alteration,
nutrients, suspended and bedded sediments, and toxic chemicals. This approach is consistent
with recent scientific consensus, recognizing that these stressors have the greatest potential for
causing adverse effects to aquatic ecosystems. In the context of this effects research, the
document also provides research that will develop diagnostic tools for a decision support system
for resource managers.
The importance of habitat quality and quantity for maintaining species is indisputable, but
quantifying exactly how species depend on habitats is multi-faceted and complex. Research is
needed to quantitatively link alterations in key habitats to provide the scientific basis to
implement regulations and policies to protect fish, shellfish, and wildlife populations, and the
ecosystems upon which they depend. To quantitatively assess effects over a range of foreseeable
stressor conditions, stressor-response relationships need to be determined. These relationships
provide fundamental information that helps to define response thresholds, or other patterns, and
to improve aquatic life and aquatic dependent wildlife criteria. As stressor-response
relationships are determined, research can be directed towards developing a "diagnostic tree"
approach to list, analyze, and characterize the causes of impairment. EPA's Stressor
Identification Workgroup has developed such an approach.
Another research need is to develop ecosystem classification approaches that allow for
reasonable extrapolations of diagnostic approaches and stressor-response models. Classification
is valuable for grouping ecosystems according to similar criteria and for spatially classifying
ecosystems connected via stressor actions. Since little is known about scale relative to ecosystem
classification, effects research also will provide guidance about the most appropriate scale for
xiv
-------
various ecosystem classification approaches, up to and including the watershed scale. At the
same time, current criteria, methods, and approaches for stressors need to be improved where
major uncertainties exist, or developed for others where information is scarce. Finally, all
aquatic stressor research elements from this process need to be combined to develop management
strategies to protect the ecological integrity of aquatic ecosystems. A series of EPA workshops
and workgroup meetings has identified five main research products for meeting this goal:
Methods to predict biological effects of habitat alteration;
Population, community, and ecosystem stressor-response models;
Diagnostic tools to determine impairment or causes of impairment to aquatic systems;
Classification approaches to aid in the prediction and management of problems; and
Methods and models to support development of ecological criteria.
In some cases, these products are under development, but in others, development has not yet
begun. Sections 4-8 of this document provide the plans for implementing research in each of
NHEERL's priority areas.
Section 4 focuses on quantifying the life support functions of specific habitat and habitat
complexes to predict the biological effects of habitat alteration and/or loss on populations offish,
shellfish, and wildlife. The main goal of this research is to quantify the role of habitat in
maintaining healthy aquatic and aquatic-dependent populations by 1) describing the relationships
between habitat and biota at the appropriate scales to quantify effects on population endpoints
due to habitat alteration and 2) synthesizing the cumulative support function of individual
habitats and aquatic ecosystems, and integrating habitat alteration effects with effects from other
stressors. Necessary elements of this research include providing: suites of species endpoints,
assessment and measurement endpoints and strategies, habitat alteration-population response
relationships, classification schemes, and models for extrapolating data across spacial scales.
Research projects in this plan deal specifically with coastal vegetated habitat; shoreline, lake, and
estuary scale habitat; and landscape scale habitat.
Section 5 centers on understanding the responses of receiving waters to excess nutrients. The
main goal of this research is to define and quantify relationships between nutrient loading and
ecological responses for different aquatic resource types to develop the basis for deriving
numeric nutrient criteria. Principal components of this research include providing conceptual
models for specific assessment endpoints, a state of the science evaluation to develop and
improve nutrient-load responses, classification schemes, standard measurement endpoints, and
nutrient load-response models. Research projects in this plan deal primarily with coastal
receiving waters and the development of nutrient load-response relationships for dissolved
oxygen, submerged aquatic vegetation, and food web and community composition changes.
Section 6 deals with the direct and indirect effects of suspended and bedded sediments in aquatic
ecosystems. The primary goal of this research is to provide and demonstrate the approach for
xv
-------
establishing sediment criteria that support aquatic life in streams/rivers, lakes/reservoirs,
wetlands, and estuaries. NHEERL's effort concerning suspended and bedded sediments has been
redirected since this section was first written. The majority of the work in this research area will
now occur under Goal 8 (EMAP) because EMAP design techniques will be applied to develop
effect thresholds for suspended and bedded sediments in aquatic systems. Some of these
techniques are described generally in the Critical Path subsection of this section. However, at
this time, the effort under aquatic stressors will only include a literature review of suspended and
bedded sediments research. Results from this review will be combined with EMAP approaches
to synthesize and evaluate the state of the science. Once the review has been completed, data
gaps will be identified and additional research will be conducted, if warranted. Additional goals
and research topics are proposed, but will depend on the results of this combined effort.
Section 7 focuses on developing methods to reduce uncertainty and significantly advance current
methods to derive criteria for toxic chemicals. The general goal of this research is to develop
scientifically-defensible methods for better characterizing the risks of toxic chemicals to aquatic
and aquatic-dependent populations and communities. The key elements for improving aquatic
risk assessments and criteria for toxic chemicals include providing methods to: improve criteria
at the individual level based on improved characterization of risks, link individual-level data to
population endpoints, support risk assessments for chemicals with limited data, and evaluate
risks on populations at various spatial scales in the context of other stressors. Research projects
under this plan center around conceptual models that will support the development and
demonstration of frameworks for better assessing the risks of both non-bioaccumulative and
bioaccumulative chemicals.
Section 8 provides an approach for diagnosing the causes of biological impairment linking
watersheds with receiving water bodies to support the TMDL process and other regulatory
programs. The primary goals of this research are to provide: a framework for interpreting cause
and effect relationships, single-stressor diagnostic methods and models to determine the primary
source of biological impairment of aquatic ecosystems, and methods and models to allocate and
forecast causality among multiple stressors for use in restoration and remediation actions. The
principal components of this research area align with the primary goals and will provide: the
scientific foundation and information management scheme for the 303d listing process, and a
classification framework for surface waters, watersheds, and regions; diagnostic methods to
distinguish among major classes of single and multiple aquatic stressors; and diagnostic tools for
forecasting approaches. Specific research projects will be conducted to establish the conceptual
framework to guide implementation of diagnostics, provide case studies to develop and test
methods for both single and multiple stressors, and to establish the structure for a decision
support system.
Over the next six years (2002-2008) the proposed research, integrated across areas, will provide
increasingly sophisticated tools to help resources managers assess ecological conditions,
diagnose impairment and causes of impairment, and forecast the effects of changes in stressor
levels. As short-term to intermediate-term research on aquatic stressors is completed, this
implementation plan will change, so that new or continued research will provide the tools
necessary to identify, assess, and manage aquatic stressors and contaminated sediments to meet
goals under the Government Performance and Results Act.
xvi
-------
Further development of this plan will require continuing interaction between the Office of
Research and Development's (ORD) Laboratories and Centers, as well as with EPA's Program
Offices and Regions, to ensure that the approaches developed are compatible with those for
exposure and risk characterization. Collaboration with the Office of Water, Regions, States, and
Tribes will be essential to ensure that this research directly supports regulatory mandates. In
addition, it will be essential to integrate the research with future grant initiatives (including
EPA's Science to Achieve Results [STAR] program) to ensure that ORD-sponsored research
complements in-house programs.
xvn
-------
Section 1.
Introduction
Purpose and Scope
This document describes the framework and implementation plans for ecological effects research
on aquatic stressors within the National Health and Environmental Effects Research Laboratory
(NHEERL) beginning in 2002. The ultimate goal of the planned research is to develop
scientifically valid approaches for protecting the ecological integrity of aquatic ecosystems from
multiple aquatic stressors. The framework first defines the context and process for conducting
research to reduce the risks aquatic stressors pose to aquatic resources, including aquatic life and
aquatic-dependent wildlife populations and communities. Specific research plans then identify
and describe NHEERL's priority research areas, linkages among the areas, and specific research
projects which include the types of products that can be expected over the next several years.
Although this document outlines the research that will be conducted within these research areas,
it does not prioritize work across research areas.
This document also provides discussion points for use with other organizations regarding
uncertainties in risk assessment techniques, potential collaborative research, and other
interactions.
The scope of the document is defined by NHEERL's research areas under the Government
Performance and Results Act (GPRA), Goal 2, Clean and Safe Water, sub-objective 2.2.3:
"Provide the means to identify, assess, and manage aquatic stressors, including contaminated
sediments." This Goal is one of 10 EPA strategic goals that cover the programmatic needs of the
Agency. NHEERL's research under Goal 2 focuses on the development of stressor-response
models for habitat alteration, nutrients, suspended and bedded sediments, and toxic chemicals;
and on the development of diagnostic methods that are applicable up to and including the
watershed scale.
This document does not include work on some aquatic stressors (e.g., invasive species, microbial
pathogens, and climate change), monitoring under the Environmental Monitoring Assessment
Program (EMAP), many aspects of biocriteria, or some work related to the aquatic stressors
effort, which is covered under other GPRA goals (e.g., molecular and cellular mechanisms of
toxic chemicals). Research on invasive species is being conducted elsewhere as EPA is an active
member of the Inter-Agency Invasive Species Advisory Council, which includes the Fish and
Wildlife Service (FWS), National Oceanographic Atmospheric Administration (NOAA), and
other federal agencies. Monitoring under EMAP and biocriteria research occurs under GPRA
Goal 8 and other research issues are covered under other NHEERL strategic goals (see
http ://www. epa.gov/nheerl/).
Aquatic stressors research will require continuing interaction between the Office of Research and
Development (ORD) Laboratories and Centers, EPA's Program Offices, and Regions, to ensure
that the efforts are not duplicated and that the approaches developed are compatible with those
for exposure and risk characterization.
1
-------
Programmatic Needs
EPA recognizes the need to advance risk assessment knowledge bases and to develop methods
for reducing risks from aquatic stressors. The immediate focus is to develop and improve
ecological criteria and diagnostic capabilities for managers, to help them meet designated uses,
and to develop options for protection and remediation efforts. Three elements provide the
regulatory context for NHEERL's aquatic stressors research:
First, the Clean Water Act (CWA) provides the legislative mandate to restore and
maintain the chemical, physical, and biological integrity of the Nation's waters.
Second, to fulfill this mandate, EPA has established under GPRA Goal 2, sub-objective
2.2.3, requirements for establishing ecological criteria that protect use designations for
the Nation's aquatic resources. Research directed toward this goal will be linked to
Annual Performance Goals (APGs) and Measures (APMs), which NHEERL has
identified for 2002-2008 (see Sections 4-8).
Finally, the Administration's Clean Water Action Plan (CWAP) establishes key actions
focused on watershed, wetland, and stream corridor protection and restoration; nutrient
assessment and criteria development; and development of a contaminated sediment
strategy (EPA 1998).
A response to these mandates requires the development of research approaches and products that
enhance the Agency's capabilities with respect to the management of aquatic resources.
References
EPA. 1998. Clean Water Action Plan. EPA 840-R-98-001.
-------
Section 2.
Research Approach
Context for Research
The common management goal for all aquatic ecosystems is to maintain ecological integrity by
protecting aquatic systems against degradation of habitat, loss of ecosystem functions and
services, and reduced biodiversity. To this end, environmental managers must be able to: 1)
assess the condition of an aquatic resource and determine the degree of impairment, 2) diagnose
the causes of impairment, 3) forecast the effects of changes in stressor levels, and 4) develop and
implement remediation and maintenance strategies. The first step in this process is to assign a
designated use for a water body and then to apply the available chemical and biological criteria
necessary to protect the use. If a resource does not support the designated use, the cause of the
impairment must be diagnosed.
To accomplish these tasks, managers must be able to make proper assessments, know the
appropriate reference conditions against which to compare their assessments, have the diagnostic
tools necessary to ascertain causes, and understand specific aquatic systems well enough to
forecast the effectiveness of potential remediation processes. While other ORD Laboratories and
Centers will make important contributions to GPRA Goal 2, sub-objective 2.2.3, NHEERL will
focus on conducting ecological effects research relative to steps 1-3 above.
The focus on aquatic stressors such as habitat alteration, nutrients, suspended and bedded
sediments, and toxic chemicals is consistent with recent scientific consensus, recognizing that
these undeniably widespread concerns have the potential for tremendous impact on aquatic
ecosystems (e.g., National Research Council [NRC] 1993, Naiman et al. 1995, Vitousek et al.
1997, EPA 1998, NOAA 1999, EPA 2000a). Because of these stressors, an aquatic resource
often fails to meet its designated use. States and Tribes commonly report these stressors as part
of their Section 303(d) lists under the CWA, thus requiring the development of total maximum
daily loads (TMDLs). Therefore, managers need a decision support system to discern the
probable causes of impairment and to identify remediation action that will restore and protect the
resource.
Research Process
Effective management and protection of aquatic resources requires multiple research elements.
The research process for developing these elements and the products of the research are shown in
Figure 1. This process presents a generally linear research sequence, although some research
elements will be conducted simultaneously. State and Tribal management agencies can protect
the ecological integrity of aquatic ecosystems only through appropriate management action, but
NHEERL can help by providing methods and tools for assessing conditions, diagnosing
impairments, and forecasting changes.
-------
Research Process
Research Products
GOAL:
Protection of Ecological Integrity of
Aquatic Ecosystems (including Aquatic
Dependent Wildlife Populations)
6a. Management Strategy Development
Program Offices, State
and Tribal Management
Agencies
6b. Criteria to Support Use Designations
NHEERL
5b. Criteria Development Approaches
4a. Classification for Extrapolation of
Diagnostic Approaches and Stressor
Response Models
3a. Diagnosis of Current
Ecosystem Conditions
4b. Watershed, Ecoregion Classification Approaches
3b. DiagnosticTools
2a. Stressor-Response Relationships: Habitat
Alteration, Nutrients, Suspended and Bedded
Sediments, and Toxic Chemicals
2b. Population, Community, Ecosystem Stressor-Response
Models
la. Quantification of Inherent Properties of Aquatic
and Aquatic Dependent Habitats Key to Support
Fish, Shellfish, and Wildlife Populations
Ib. Methods to Predict Biological Effects of Habitat Alteration
Figure 1. Research process and products for meeting the goal of effective management
and protection of aquatic resources.
Protection of the ecological integrity of aquatic ecosystems must begin with a quantification of
the inherent properties of aquatic and aquatic-dependent habitats which are critical to the support
of important fish, shellfish, and aquatic-dependent wildlife populations (box la). Research in
this area will help to assess the life support functions of habitats and habitat complexes and
provide methods to predict biological effects which result from habitat alteration (box Ib).
EMAP contributes, in part, to these basic needs under GPRA Goal 8. Ecological characterization
and identification of the priority aquatic research elements are key to developing stressor-
response relationships within each of the aquatic stressor areas (box 2a). Stressor-response
relationships are needed to quantitatively assess effects over a range of foreseeable conditions of
the stressor. These relationships provide the fundamental information required to define
response thresholds or other patterns and to improve criteria. Determining stressor-response
relationships also should help define symptoms of a problem and identify diagnostic measures
that can be broadly applied. This research will provide stressor-response models, if needed,
within each of the aquatic stressor areas (box 2b). Eventually, these models must be capable of
dealing with multiple stressor interactions if they are to support the development of approaches
that allow characterization of the ecological condition of aquatic systems relative to a desired
condition. However, initial research will focus on the stressor-response relationships of the
individual stressors, in order to set the stage for the more difficult problem of dealing with
multiple stressors in the longer term.
-------
As stressor-response relationships are determined, research will be directed towards developing
diagnostic approaches (box 3a), which will provide tools (box 3b) for building a decision support
system. Resource managers then can use the system to assess the condition of a water body,
diagnose the causes of any demonstrated impairment, and predict the results of any corrective
actions that might be needed.
Stressor-response relationships can be specific to different classes of systems. Thus, research
also will focus on developing ecosystem classification approaches (boxes 4a,b) that allow for
reasonable extrapolations of diagnostic approaches and stressor-response models. Classifying
ecosystems is valuable for two primary reasons: 1) grouping ecosystems according to similar
criteria and 2) spatially classifying ecosystems that are connected via stressor actions to facilitate
an effective means for managing the consequences of stressors. Since little is known about scale
relative to ecosystem classification, effects research also will provide guidance about the most
appropriate scale for various ecosystem classification approaches, up to and including the
watershed scale.
At the same time, research identified in boxes la-4a will result in methods and approaches for
deriving criteria (boxes 5a,b) for protecting aquatic ecosystems. Existing approaches, based on
laboratory tests, have focused on individual aquatic life and wildlife species. However, much
uncertainty is associated with extrapolating data to predict safe levels for populations and
communities exposed to individual and multiple stressors (physical and/or chemical). Therefore,
we need to improve current criteria, methods, or approaches for some stressors where major
uncertainty exists, or develop them for others where little information is known (see Sections 4-
8). In some cases, research will lead to relatively short-term fixes (1-2 years) to existing
guidance. In others, research conducted over the longer term (3-6 years) will result in methods or
models useful for deriving criteria with associated uncertainties.
All aquatic stressor research elements (boxes 1-5) thus combine to help improve the tools
available to managers for meeting designated uses (boxes 6a,b). It is important to recognize that
NHEERL research on aquatic stressors supports the development of protective criteria, although
actual criteria and management strategies fall beyond the research responsibilities of NHEERL.
A discussion on a decision support system for using the products of this research follows.
Decision Support System
A general approach that a resource manager might follow for managing water bodies is outlined
in the left side of Figure 2. The assessment of the condition of an aquatic resource to support
ecological use designations first requires ecological criteria and a reference condition with which
to compare those criteria. As shown in the right side of Figure 2, GPRA Goals 2 and 8 research
products will support the development of both chemical and biological criteria that can assist
managers in determining if designated uses are met.
If designated uses are not met, managers will require a means of identifying the stressors causing
the impairment. EPA's Stressor Identification (SI) workgroup has developed an example of a
"diagnostic tree" approach for SI that can be used by resource managers (EPA 2000b). NHEERL
research will contribute by determining stressor-response relationships and specific diagnostic
-------
indicators for aquatic stressors and by developing additional decision support tools as needed. In
this approach, the SI process is iterative, usually beginning with retrospective analysis of
available data, and includes the identification of stressors that might be causing the impairment.
C riteria to Supp art U se D e si grations
Chemical-specific (WQC, ESGs)
Whole Effluent
Bio criteria
YES
NO
Stressor I dentifi cation
List Candidate Causes
Analyze Causes
AbnedlfibiML ? Hdrienls ?
SediD.enb.tian. ' Toxic Chanti!- ? Otter ?
Characterize C auses
Management Action Fore cast Effects of Stressor
Change s; R em ediati on
Goal 2 Re search Products: Methods to Predict
Biologcal Effects of Habitat Alteration,
Stressor-Response Models, Diagnostic Tools,
Classification Appro aches, Criteria
Development Methods
Goal 8 EMAP Re search Pro ducts
Methods for Assessing Reference Condition
Goal 2 Re search Products
Stressor-Response Models
Diagnostic Indicators
Goal 8 EMAP Re search Pro ducts
Monitoring and Process Models
Goal 2 Stressor-Response Models
Goal S Restoration and
R em e di ati on Re search Products
Figure 2. Manager's decision support system to protect and restore aquatic resources
using ORD's research products (SI box is modified from EPA 2000b).
Stressor identification consists of three main steps, the core of the SI process: 1) listing candidate
causes of impairment, 2) analyzing these candidate causes, and 3) producing a causal character-
ization. The support system also involves interactions with decision makers and stakeholders to
assist in forecasting the effects of the stressors and in taking remediation action, if needed.
Remediation requires, first, criteria for what is acceptable in a given environment, and second,
the models necessary to link changes in stressors with improvements in the system. Additional
information is provided in Section 8 (Diagnostics) of this document concerning the development
of a framework for a decision support system. NHEERL-generated products from this research
will be combined with exposure models and with restoration and remediation techniques
developed by the National Exposure Research Laboratory (NERL) and National Risk
Management Research Laboratory (NRMRL), respectively, to meet management needs.
-------
References
EPA. 1998. Water quality criteria and standards plan - priorities for the future. EPA 822-R-98-
003.
EPA. 2000a. OW/ORD Strategic Planning Research Coordination workshop document, version
2. January 31.
EPA. 2000b. Stressor identification guidance document. EPA-822-B-00-025. OW/ORD,
Washington, DC, December.
Naiman, R.J., Magnusen, J.J., McKnight, D.M., Stanford, J.A., eds. 1995. The Freshwater
Imperative: a Research Agenda. Island Press, Washington, DC.
NOAA. 1999. National estuarine eutrophication assessment: a summary of conditions, historical
trends, and future outlook. Draft. National Ocean Service, Silver Spring, MD, June.
NRC. 1993. Managing wastewater in coastal urban areas. Washington, DC.
Vitousek, P.M., Aber, J.D., Howarth, R.W., Likens, G.E., Matson, P.A., Schindler, D.W.,
Schlesinger, W.H., Tilman, D.G. 1997. Human alteration of the global nitrogen cycle: sources
and consequences. Ecol. Appl. 7:737-750.
-------
Section 3.
Research Products and Implementation Plans
Five main products from this research have been identified for meeting the goal of protecting
ecological integrity of aquatic systems (Figure 1, boxes lb-5b). These are:
Methods to predict biological effects of habitat alteration;
Population, community, and ecosystem stressor-response models;
Diagnostic tools to determine impairment or causes of impairment to aquatic systems;
Classification approaches to aid in the prediction and management of problems; and
Methods and models to support the development of ecological criteria.
Although research to produce these products might need to be conducted independently at first, it
also will have to be coordinated so that these products can be integrated into a manager's
decision support system. In some cases, these products have been established, but need to be
improved. In others, they will have to be developed. Therefore, research is being planned
around compatible approaches, scales, critical areas, and geographic regions, when possible.
The remainder of this document (Sections 4-8) provides the plans for implementing research in
each of NHEERL's priority areas. Each plan describes the problem the Agency faces, the goals,
critical path for conducting the research, specific research projects to be conducted, a gap
analysis, and references. The research projects within each plan contain the general objectives,
scientific approach, products, benefits of the products to the client, and an estimate of the
workload (in full time equivalents, FTEs) it will take to complete each project.
Currently, the work groups identified at the beginning of this document along with other staff
within NHEERL's Ecology Divisions are detailing the projects outlined in the implementation
plans in Sections 4-8 (Note: these detailed projects plans are not included in the present
document). Each project will describe further coherent portions of the critical paths conceived to
meet the goals stated in each implementation plan and will show how the projects will be
completed to produce the products listed. Cross divisional projects will be conducted by
scientists from more than one Ecological Division, while others will be conducted by staff of a
single Division. Each project will consist of a design narrative (design, approach, and analysis) a
Quality Assurance (QA) Plan, Health and Safety and Environmental Compliance Plan, and an
Animal Care and Use Plan (if required) consistent with the Ecological Divisions' project plan
documentation, and will undergo internal review prior to final approval. Work load (FTEs) and
staff also will be provided in more detail than what is stated in the implementation plans shown
in Sections 4-8.
The completion of research outlined in this document will require continued interaction within
and across NHEERL's Ecological Divisions, between ORD's Laboratories and Centers, and
-------
within EPA's Program Offices and Regions. Collaboration with OW, Regions, States, and
Tribes will be essential to ensure that this research directly supports regulatory mandates. In
addition, it will be essential to integrate the research with present and future extramural
initiatives (including EPA's Science to Achieve Results [STAR] program) to ensure that ORD-
sponsored research complements in-house programs and to fill research gaps that have been
identified within each of the implementation plans. Many of these interactions are outlined in
Sections 4-8; others will be delineated further in the detailed research projects that are now being
developed by NHEERL staff.
-------
Section 4.
Implementation Plan for Habitat Alteration Research
Problem
Significant improvements in some aspects of the North American environment have been
realized over the past several decades, but the continuing increase in human populations and
associated activities has created an array of regulatory and policy challenges (e.g., land-use
changes, hydrologic modification, climate change, altered biological diversity, introduction of
nonnative species, concern about ecological sustainability, and cumulative effects of manmade
chemicals) that defy traditional command/control approaches (EPA 1999). Many anthropogenic
activities exert their influence on biota via effects on habitat, and habitat alteration is arguably
the most important cause of declines in ecological resources in North America (EPA 1990).
Thriving populations offish, shellfish, and wildlife are valued by the public, not only for
commercial, recreational, and aesthetic reasons, but also as tangible and visible surrogates for the
overall condition of the environment. Habitats essential to the well being of these species are
rapidly being affected by a myriad of land-use activities. Habitat alterations have been identified
as a major cause of endangerment for species within the United States. For example, the U.S.
has the most diverse temperate freshwater fish fauna in the world, but 35-40% of its 790 fish
species are imperiled because of poor land use practices, wetland alteration, introductions of
exotic species, and other habitat-altering factors (Warren and Burr 1994, Stein and Flack 1997).
EPA has not traditionally focused its research, regulatory, or policy effort on habitat alteration.
However, a number of factors converge to justify a new EPA emphasis on habitat issues. The
CWA has a goal "to restore and maintain the physical, chemical, and biological integrity of the
Nation's waters," including the "protection and propagation offish, shellfish, and wildlife."
While the chemical integrity of aquatic resources is much improved, physical and biological
integrity remains a concern. Habitat alteration is a common cause for the failure of aquatic
systems to meet designated uses as required by the CWA, and addressing these failures
increasingly requires ameliorating the cumulative impacts of diffuse stressors including nutrient
loading, sedimentation, and altered hydrologic regime. The necessary integrated approach to
environmental protection is perhaps best provided by habitat-based criteria. As required by the
Endangered Species Act (ESA), EPA is increasingly being asked to participate in interagency
species protection and restoration efforts where habitat issues play a key role. Because one of
EPA's core ecological regulatory authorities is the CWA, the species endpoints for which habitat
alteration is of greatest concern are aquatic species (i.e., fish and shellfish), and water- and
wetland-dependent wildlife. By focusing on aquatic ecosystems and habitats supporting species
of combined ecological and societal importance, EPA can advance broad environmental
protection goals while directly addressing issue-driven stakeholder concerns.
EPA's Office of Water (OW) has identified priority ecosystem types for which habitat alteration
research is especially needed. These systems include freshwater and estuarine wetlands, stream
corridors, and marine and Great Lakes coastal zones. Not coincidentally, these resource types are
of considerable importance in sustaining ecologically and societally valuable fish, shellfish, and
10
-------
wildlife populations. More than 50% of U.S. marine fisheries exploit species that are estuarine-
dependent at some life stage, and many estuarine fisheries are in decline due to combined effects
of over fishing, habitat alteration, and pollutants (Houde and Rutherford 1993). Virtually all
Great Lakes fish depend at least indirectly on coastal wetland habitats, where habitat alteration is
an important threat (Whillans 1992). Within these priority ecosystem types, OW has a particular
interest in vegetated habitats. Aquatic vegetation is not only a key habitat for many wetland,
estuarine, and coastal species, but also a key mediator of stressor effects on aquatic biota and a
primary response variable for anthropogenic stressors such as nutrient and sediment loading. The
combination of OW interests, pressures affecting societally and ecologically important species,
and NHEERL research expertise leads to a focus on these endpoints in the this plan.
Assessing the ecological consequences of habitat alteration has been called one of the most
challenging scientific problems and environmental policy issues confronting society (NRC 1997,
Rapport et al. 1998, EPA 1990). The importance of habitat quality and quantity for maintaining
species is indisputable, but quantifying exactly how species depend on habitats is multi-faceted
and complex. Habitat provides a wide array of species life-support functions, ranging from
providing shelter, substrate, and appropriate physiological conditions; to mediating natural
disturbances and anthropogenic stressors; to maintaining food webs by hosting primary and
secondary production. Consequently, habitat alteration can degrade diversity, food-web
structure, ecosystem function, and populations of valued fish, shellfish, and wildlife species via
complex effect pathways. Mobile and migratory species can use multiple habitats to meet
developmental requirements or sustain local populations, and "habitat" for them may refer to a
combination of quantity, quality, extent, and arrangement of different habitat types at a variety of
spatial scales. Many stressors interact with aquatic systems in ways that alter the normal spatial
distribution or mosaic of habitat patches, with important implications for ecosystem function and
dependent fish, shellfish, and wildlife populations. More generally, successful preservation of
biological diversity and ecosystem structure and function requires protection of multiple habitats
within a landscape framework and not merely individual habitats in isolation. For many
important aquatic habitats, there is little quantitative information on the relationship of habitat to
dependent biota; in particular how changes in habitat quality influence the well being offish,
shellfish, and wildlife populations. Finally, broad biogeographic gradients affect the responses of
ecosystems and biota to habitat alteration. For all these reasons, it is a significant research
challenge to quantify the life support functions of specific habitats and habitat complexes in
sufficient detail to predict the biological effects of both incremental and catastrophic habitat
alteration.
NHEERL has initiated a nationwide research program to quantitatively link alterations in key
habitats to fish, shellfish, and wildlife endpoints because habitat alteration is such an important,
pervasive stressor on valued aquatic resources. The research involves all four Ecological
Divisions of NHEERL and spans the coastal resources of the East, West, Gulf states, and the
upper Midwest. NHEERL will enter into partnerships with other management agencies and
research entities as appropriate to further these research goals. The research described in this
plan will help build the scientific basis to implement regulations and policies to protect aquatic
populations and the ecosystems upon which they depend.
11
-------
Goals
The overarching goal of this plan is to provide the scientific basis for assessing the role of
essential habitat in maintaining healthy populations offish, shellfish, and wildlife and the
ecosystems upon which they depend.
Such a scientific understanding is essential in devising habitat protection and restoration
priorities and schemes. The spatial pattern and temporal dynamics of habitats and habitat
conditions within landscapes play significant roles in the long-term viability of aquatic
populations, species assemblages, and ecosystem functions. Ecosystem level responses to
stressors are functions of both the tolerance of individual species and habitats to those stressors,
and the spatial distribution and connectivity of habitats within the landscape. Furthermore,
habitat components themselves play ecosystem roles that mediate the response of species and
assemblages to stressors. Finally, habitat influences populations and assemblages at hierarchical
spatial scales ranging from patches (micro-scale) to entire ecosystems (macro-scale) to
watersheds or regions (landscape scale). Understanding how effects of anthropogenic stressors
are mediated by alteration in habitat quality, abundance, and configuration at various spatial
scales is necessary in order to develop aquatic resource protection criteria and to predict the
resiliency, restorability, and recovery offish and wildlife populations and their supporting
ecosystems.
More specifically, APGs and APMs for the research are:
APG 1 FY02 Provide suites of relevant fish, shellfish, and wildlife species endpoints suitable for
setting regional-scale habitat protection criteria for coastal systems, along with preliminary
reviews of methods, modeling approaches, and available data for relating habitat alteration to
changes in those species.
APM 1A FY02 Listings of the high-priority species offish, shellfish, and aquatic-
dependent wildlife for study in each biogeographic region, and listings of the habitats that
are considered to be critical to each (WED).
APG 2 FY04 Provide models for linking habitat alteration stressors and mercury to the regional
problems of Northeast Loons and to landscape-watershed alterations for Pacific salmon.
APM 2A FY03 Prototype watershed-stream network modeling approach for Pacific
salmon (WED).
APM 2B FY04 Habitat suitability indices to support population models for projecting
relative risks of multiple stressors including toxic chemicals and habitat alteration to
common loons (AED).
APG 3 FY04 (GPRA # 8) Provide demonstration stressor-response relationships and/or models
linking loss and alteration of habitat to selected fish, shellfish, and wildlife endpoints.
APM 3A FY03 Penaeid shrimp dependence on seagrass habitat (GED).
12
-------
APM 3B FY03 Finfish dependence on seagrass and oyster reef habitats (GED).
APM 3C FY04 Report characterizing the relationship between habitat in stream networks
and salmon-native fish for coastal Oregon watersheds (WED).
APM 3D FY04 Report characterizing the relationship between alteration of vegetated
habitats and nekton use of those habitats (AED).
APM 3E FY04 Report characterizing relationships between multiple habitat types and
economically valuable fish at the scale of an estuarine shoreline (AED).
APM 3F FY04 (GPRA #58) Preliminary report characterizing relationships between
abundance, quality, and arrangement of various habitat types and selected biotic
assessment endpoints in coastal systems (WED).
APG 4 FY05 Provide indices of patch, ecosystem, and landscape-scale habitat integrity based on
support for selected fish, shellfish, and wildlife assemblages.
APM 4A FY05 Develop indices of watershed integrity based on land use/land cover and
relationships to fish (WED, MED).
APG 5 FY06 Provide stressor-response relationships and/or models linking loss and alteration of
habitat to selected fish, shellfish, and wildlife endpoints.
APM 5A FY05 Reports characterizing the relationship between landscape-scale habitat
mosaics and native fish by wetland type in the Great Lakes (MED).
APM 5B FY06 Report characterizing relationships between abundance, quality, and
arrangement of various habitat types and selected biotic assessment endpoints in coastal
systems (WED, AED, GED, MED).
APG 6 FY08 Provide suites of habitat alteration-biological response relationships and
generalization/extrapolation schemes suitable for developing broad-scale habitat criteria for
streams and coastal systems, and provide approaches for evaluating combined effects of habitat
alteration and other stressors.
APM 6A FY04 Ecological consequences of marine derived nutrients and nutrient
enrichment for aquatic biota and stream habitat quality, with an emphasis on salmon and
native fish (WED).
APM 6B FY07 Regional models of landscape influence of salmon/native fish in the
Pacific Northwest and native fish in Great Lake coastal wetlands (WED, MED).
APM 6C FY08 Synthesized quantitative species-habitat relationships suitable for
developing regional habitat-based biocriteria for shorelines, lakes, and estuaries (AED,
GED, MED, WED).
13
-------
APM 6D FY08 Interactions between stream nutrients and habitat alteration on water
quality and aquatic life (WED).
Critical Path
Determining the role of essential habitat in maintaining fish, shellfish, and wildlife populations
involves two distinct research components:
1. Describing the relationship between habitat and biota at appropriate spatial scales and with
sufficient detail and resolution to quantify the effects of both incremental and catastrophic habitat
alteration; and
2. Synthesizing the cumulative support function of individual habitats and ecosystems and
integrating habitat alteration effects with effects of other stressors, so that resource protection and
restoration priorities can be evaluated at spatial scales up to and including regions or large
receiving bodies.
Within these two research components, research efforts can be further categorized by spatial
scale, ranging from habitats within ecosystems to entire ecosystems to landscapes and regional
habitat mosaics. A critical path diagram illustrating this two-component research strategy and
the spatial focus of research efforts is given in Figure 3. Table 1 gives a overall time-line for the
overall plan.
Component 1
NUEERL proposes to focus research describing relationships between habitat and biota on
coastal marshes, estuaries, and nearshore environments. Two primary spatial scales will be
considered: the scale of habitat elements, especially vegetated habitat, within marsh and
estuarine systems; and the scale of entire coastal wetland and estuarine ecosystems (i.e., micro
and macro-habitat scale). Synthesizing the cumulative population support functions of coastal
habitats and ecosystems on a regional basis is a long-term goal, but the initial emphasis will be
on quantifying habitat-biota relationships at the ecosystem and within ecosystem scale.
Necessary elements of this habitat research include:
1. Identifying suites of endpoints that provide a nationally comprehensive and comparable basis
for linking alteration of key coastal habitats to species and assemblages of economic or societal
14
-------
a
I
1/5
"5
o
e\
6
o
U
4H
OJ
2
PH
Habitat alteration - biota response
relationships suitable for setting habitat
criteria and protection/restoration priorities
t
Extrapolation across ecosystems requires
classification & regionalization schemes and
datasets [APG5, 6]
t
Habitat integrity indices based on habitat -
biota response relationships [APG4]
t
Field work, data analysis, and modelling to
develop habitat - biota response relationships
& mechanisms [APG3]
t
Review available data & models for Unking
habitat to biotic endpoints [APG1]
Identify species, ecosystems, and habitats of
concern [APG1]
t
t
Project 1
vegetated
aquatic habitat
Project 2
coastal
ecosystems
Component 1: habitat biota
relationships (habitat and
ecosystem scale)
Methods for integrating habitat alteration and
other stressors in a regional or landscape
level effects context
t
Evaluation of the relative & cumulative
effects of stressors including habitat loss on
spatially structured populations [APG5, 6]
t
Landscape scale habitat integrity indices
[APG4]
t
Landscape scale habitat - biota response
relationships & mechanisms [APG3]
Models and approaches for assessing
importance of spatial structure and
connectivity [APG2]
t
t
Project 3
salmonids /
native fishes
Project 4
piscivorous
birds
Component 2: scaling up to the
landscape (reach, network,
watershed, region)
Figure 3. Critical path for habitat alteration research (APGs) refer to those listed and
described in the Goals subsection.
15
-------
Table 1. Time line for habitat alteration research. APGs in the table are abbreviated versions of those listed under the Goals
subsection, and arrows indicate which research components support which APGs. "HAPR" refers to habitat alteration-population
response models.
year
APG
Component 1,
Project 1
vegetated
habitat - biota
relationships
Component 1,
Project 2
ecosystem
scale habitat-
biota
relationships
Component 2,
Project 3
landscape
stressors on
salmonids &
native fishes
Component 2,
Project 4
piscivorous
birds
FY02 FY03 FY04 FY05 FY06 FY07 FY08
APG1: species APG2: spatial APGS: habitat- APG4: habitat APGS, APG6: synthesize habitat
suites & lit modeling biota integrity indices approaches & relationships
review approaches relationships
species endpoint suites -> APG1
species-habitat use eval. -> APG1
HAPR models for species -ğ APGS, and regions -ğ APGS, APG6
field work to develop/validate HAPR models -ğ APGS, APG4
synthesis of HAPR relationships -* APG5,APG6
species endpoint suites -> APG1
species-habitat use eval. -> APG1
HAPR models for species -ğ APGS and regions -ğ APGS, APG6
multi-species models as needed -> APG5,APG6
field work to develop/validate HAPR models -ğ APGS
synthesis of HAPR relationships -* APG5,APG6
landscape habitat classification -> APGS
network model develop. ~^ APG2
landscape scale HAPR models -ğ APGS & integrity indices -ğ APG4
regional landscape influence models -> APGS, APG6
toxicity model/HAPR models -ğ APG2
population/spatial models -> APG2
16
-------
relevance, and identifying data and models available to make these linkages.
2. Devising assessment and measurement endpoints and strategies for the fish, shellfish, and
wildlife populations and habitat elements of concern so that measurements are comparable or
complementary and informative across regions, ecosystems, and habitats.
3. Developing habitat alteration-population response relationships for the species and habitats of
concern capable of quantifying effects of both incremental and catastrophic habitat alteration.
The relationships must deal with both individual habitat components and interactions among
them. Habitat-biota relationships may require intermediate steps to properly describe the linkage
pathways (e.g., to capture the productivity subsidy to fishes from coastal wetland vegetation).
4. Devising regionalization and classification schemes reflecting the range and distribution of
coastal habitats and ecosystems to identify biogeographic expectations and capture ecosystem
constraints and forcing factors.
5. Identifying data sets and approaches for extrapolation, both to specific unstudied systems, and
to make inferences for the population of systems from the suite sampled.
Component 2
NHEERL plans to pursue the influence of human activities on habitat at landscape, regional, and
watershed scales via two conceptually linked studies. One project will examine watershed and
landscape scale habitat issues affecting recovery of Pacific Northwest salmon, and fishes reliant
upon Great Lakes coastal wetlands. A second project will examine the interaction of a suite of
anthropogenic stressors including habitat alteration affecting piscivorous birds (e.g., common
loons) distributed across heterogeneous lake districts. Unlike the wetland/nearshore research
described above, Component 2 research projects emphasize modeling and Geographic
Information Systems (GIS), and explicitly consider the interaction among habitat alteration and
other stressors. Common components of these research projects include:
1. Identifying data sets, approaches, and measurements for characterizing the factors, including
habitat alteration and other stressors that affect selected species or assemblages at large or
hierarchical spatial scales.
2. Developing and comparing approaches, indices, and models for extrapolating from individual
or local habitat-biota relationships to effects on regionally-distributed populations or
metapopulations.
3. Assessing the importance of spatial structure and connectivity of habitats via modeling efforts
at varying spatial scales and resolutions.
Research Projects
This plan is divided into four parallel but closely linked projects. Project 1 (Coastal Vegetated
Habitat Research) addresses societally important endpoints of concern that are affected by
17
-------
alteration of critical habitats, especially vegetated aquatic habitats, within coastal ecosystems.
Project 2 (Shoreline, Lake, and Estuary Scale Habitat Research) also deals with societally
important endpoints, but focuses on those coastal ecosystems where the interactions of multiple
habitats predominantly determine the condition offish, shellfish, and wildlife populations.
Project 3 (Salmon and Native Fish Habitat Research) and project 4 (Multiple Stressor Risks to
Common Loon and Other Piscivorous Bird Populations [cross-listed in Section 7, Toxic
Chemicals, Project B3}) concentrate on providing the scientific basis to protect critically
important endpoints such as wild salmon and migratory wildlife whose populations are at risk
due to, among various factors, large scale changes in their habitats. As described in the Critical
Path, work at the vegetated habitat and multiple habitat scale primarily addresses the need for
developing habitat-biota-response relationships. Work at the watershed/regional/landscape scale
is divided into two projects that are conceptually related but deal with different ecosystems and
biological endpoints. Research will directly address the consequences of habitat alteration for
societally important fish, shellfish, and wildlife species. Table 2 lists those species that, on the
basis of initial assessment, appear as the best candidates for study under projects 1 and 2. During
development of specific research plans by NHEERL Ecological Divisions, a more detailed
Table 2. List of candidate species for study in marine and Great Lakes coastal regions.
Northeast Atlantic Coast
(Atlantic Ecology Division)
Gulf Coast
(Gulf Ecology Division)
Winter flounder
Striped bass, bluefish, and weakfish
Tautog
Bay scallops and lobster
Waterfowl and shorebirds
Penaeid shrimp
Blue crab
Red drum and other sciaenids
Oysters
Waterfowl and shorebirds
Great Lakes
(Mid-Continent Ecology Division)
Northwest Pacific Coast
(Western Ecology Division)
Northern pike
Walleye
Yellow perch
Largemouth and smallmouth bass
Waterfowl and shorebirds
Salmon and trout
Dungeness crab
Pacific herring
Threatened and endangered native fishes
Waterfowl and shorebirds
analysis of the current state of knowledge concerning the species-habitat relationships for those
on this list will be done. Development of research efforts on particular organisms under this plan
will be closely coordinated with other efforts, both within EPA and in other organizations. For
example, NHEERL also has developed a Wildlife Research Strategy (WRS) (EPA 2000), and
the related research initiated here will be integrated with any NHEERL research on semi-aquatic
wildlife.
A systematic approach will be used to evaluate candidates and select species using a consistent
societal value-based scheme. A data table will be constructed for each region listing the species
18
-------
and their associated societal values or characteristics, such as economic value, charismatic value,
stock status, and extent of coastal/estuarine dependency. Species will then be prioritized.
Development of a final list from the top-priority populations will be used to select species from a
range of life histories and ecological niches to best represent biological diversity and ecosystem
function. The final list also will include information that will consider the scientific feasibility of
studying populations, because species with certain life histories may be very difficult or
impractical to study given the available research tools. For some species, research often will
focus on the juvenile stages because juveniles are typically more dependent on specific habitats
and are not subject to direct fishing pressure, but adults will sometimes be the endpoint that best
integrates over the suite of habitat elements being considered. Secondary species endpoints for
each of the above areas are the key forage fishes and invertebrates upon which larger societally
valuable species depend, and these will be considered as well.
Project Title 1: Coastal Vegetated Habitat Research
Project Coordination and Resources (11.15 FTEs: AED-5.05, GED-2.0, MED-1.0, WED-3.1)
Objectives
To quantify the role of aquatic vegetated habitat in providing structure and life support
functions (e.g., food and shelter) to selected and societally important fish, shellfish, and
wildlife populations.
To identify those attributes of habitat within vegetated aquatic systems that are key to
sustaining societally important species, and to further determine the functional
relationships between those attributes and the utilization of that habitat by primary and
secondary (e.g., forage organisms) assessment endpoints.
To integrate the results of habitat-specific research results with other NHEERL habitat
research that focuses on larger scale questions that range from localized among-habitat
differences to the landscape and regional scale.
To provide to other research teams within NHEERL the functional relationships between
aquatic vegetation attributes found to be important to endpoints, and that are also
impacted by stressors.
To transmit the results of this research to resource managers in a format appropriate for
its application in policy and regulatory decisions.
Scientific Approach
The definition of "habitat" invariably combines two concepts: a geographic location; and the
flora and fauna that are regarded as dependent upon, or otherwise functionally associated, in
some way, with that location. Aquatic habitats can range in character from bare sand, sediment,
or rock substrates to areas of submerged or emergent vegetation. This portion of the plan focuses
on aquatic vegetation - the areas of freshwater or estuarine wetlands, marshes, and sea grass beds
19
-------
that, by virtue of their attachment to the substrate, are also associated with location. Areas of
aquatic vegetation represent the smallest, discrete scale that research described in this plan will
concentrate upon and will be studied at the greatest level of detail. Project 2 (Shoreline-, Lake-
and Estuary-Scale Research) extends to larger scales, but at lesser detail, and will consider how
vegetated habitat interacts with other habitat types to determine biotic structure at the ecosystem
scale.
The emphasis on aquatic vegetation in this project is appropriate for several reasons. First,
aquatic vegetation is one of the most widespread and important types of aquatic habitat, in part
because of the exceptional productivity of the plants. Recreationally and commercially important
fish, shellfish, and wildlife, as well as rare and endangered species that are especially valued by
human society, frequently exploit this productivity, either using the vegetation as a direct food
resource, or indirectly, by feeding on smaller forage organisms that rely directly on the
vegetation. Aquatic vegetation also strongly influences local physical and chemical habitat
conditions of significance to fish and shellfish, including substrate type and stability, wave and
current energy, and water quality. The structural complexity of aquatic vegetation provides
shelter and nursery areas for its inhabitants. Overall, research will focus on evaluating the
importance of habitat attributes of vegetated aquatic systems to the assessment endpoints of
interest to society (see Figure 4). Aquatic vegetation is itself a key endpoint of research plans
being formulated under the other aquatic stressors implementation plans, including Nutrients
(Section 5) and Diagnostics (Section 8), and research outlined in this plan will be closely
integrated with those efforts.
Identification and Prioritization of Assessment Endpoints
The assessment endpoints are organisms believed to be dependent on aquatic vegetation and
identified as of societal value, and hence of regulatory and policy importance. Societal
relevance will be the dominant criterion for assessment endpoint selection, but societal relevance
needs to be intersected with ecological relevance and EPA research capabilities and regulatory
mandates. Endpoints most likely will be chosen from the above listing of candidate species.
However, other, intermediate endpoints such as forage fish species may also be required because
the link between societally relevant species and aquatic vegetation may be mediated through
secondary production, water quality, or other functional aspects of aquatic vegetation. During
the development of specific research plans, the researchers participating in this effort will
evaluate critically the suitability and appropriateness of proposed assessment endpoints with
regard to regional EPA concerns, the laboratory capabilities, and the research activities of other
agencies. Additionally, a concerted effort will be made to coordinate this research with that of
other offices within ORD such as NERL and Federal and State resource management agencies,
20
-------
FISH/SHEILFISH WILDLIFE
- abundance
- species composition
- guild structure -% exotics
- p opulati on structure
Direct Effects:
-shelter
- spawning substrate
-physiology limits
Indirect Effects:
- food abundance
- food composition
- competition
FOODWEBS
- productivity
- troptic structure (levels,
pathways, etc.)
Direct Effects:
- habitat for prey or^nisms
- primary pro ducti on
- disturbance regime
HABITAT COMPONENTS VULNERABLE TO ALTERATION
Sediments \A&ter quality Plants Morphology
-area
-bathymetry
-channel vs. backwater
-grain size
-% organic
-DO
-pH
-turbidity
-temperature
-growth form
sreal extent
-biomass
-species
Figure 4. Components of coastal vegetated habitat with possible pathways for direct and indirect
effects of habitat alteration on fish, shellfish, and wildlife.
so as to preclude duplication of effort, allow the differing objectives of the agencies to be
represented (e.g., utilization vs. conservation), and foster synergy. As an example, both the
South Atlantic Fishery Management Council and the Atlantic States Marine Fishery Commission
have explicitly identified submerged aquatic vegetation (S AV) as areas of concern with respect
to fishery resources, and EPA research efforts will be coordinated with those entities. Research
on endangered species, such as Pacific salmonids, will necessitate close collaboration and
coordination with other agencies, because authorization for and logistics of discrete, stand-alone
efforts may be difficult or impossible. Additionally, a number of existing EPA or other Federal
and State databases and software (e.g., those available at http://www.epa.gov/OW/soft.html) may
be useful during the implementation of this research.
Assessment of Key Habitat Elements for Biota of Vegetated Aquatic Systems
Assessing effects of incremental habitat alteration on species requires quantitative understanding
of a given organism's reliance upon vegetated aquatic habitat. For clarity, this step is listed
separately from the step of identifying candidate organisms, but at least a partial assessment of
vegetation dependency will be done concurrently with species identification and prioritization.
This step will integrate data from ongoing or published research, but EPA field efforts for some
species, ecosystems, and habitat types will be required. It is necessary to critically evaluate the
nature and level of the organism's dependence on vegetated habitat, both with respect to life
history stages and the importance of vegetated habitat relative to alternative habitats. For
21
-------
example, there may be a perception that a given organism is dependent on aquatic vegetation
when in fact it is more of a case of co-occurrence between the organism and the vegetation that
share similar physical requirements, rather than a dependence of the organisms on conditions
provided by the vegetation. An example would be migratory fishes, such as Pacific salmonids.
The level to which these fishes utilize estuarine habitat in their migratory cycle is not well
known, and the utilization of estuarine seagrass beds is thought to vary greatly among salmonid
species and stocks. If such vegetation were not present, would salmon simply reduce their
residence time in the estuary and continue their seaward migration sooner than they would
otherwise? It must also be assessed whether the organism is associated with the vegetated
habitat throughout its life, or only occupies the habitat during specific life history phases or
seasons. It should be noted that many species have life history stages that are notoriously
difficult to observe or sample, and it may be difficult to establish whether and how much these
organisms are truly dependent on aquatic vegetation.
Characterization of the Species (Assessment Endpoint)-Habitat Relationships
In order to quantify species-habitat relationships, it is necessary to define the structural and
functional attributes of aquatic vegetation that are to be documented and their alignment with the
hypothesized assessment endpoint requirements. Standardization is necessary to ensure that data
are collected in an agreed upon and consistent manner among the research efforts. There must
also be agreement and coordination among broader research groups: aquatic vegetation is
considered in this document from the perspective of its function as a habitat, but other research
will be evaluating the effect of nutrient loading and other anthropogenic stressors on aquatic
vegetation, and so the data collected for habitat and nutrient work (e.g., see Section 5) should
support one another. Examples of vegetation attributes include species composition, areal extent,
fragmentation or patchiness of the habitat, zonation, productivity, density, growth rates, condition
(or other evaluations of the vegetation's "health"), degree of protection from predators the
vegetation might afford, substrate characteristics, water quality, seasonal alterations to the
habitat, and whether it directly provides food for an assessment endpoint or food for a secondary
endpoint such as forage organisms. Standardization of data is highly desirable, but it must not
neglect regional differences in vegetated habitat. For example, many marine vegetated habitats
are spatially extensive and largely monotypic, and efforts to characterize habitats may focus
largely on vegetation condition and extent. In contrast, vegetated habitat of Great Lakes coastal
wetlands is diverse and variable over relatively small spatial scales, so that research efforts here
will focus on distribution and interspersion of habitat types with the ultimate goal of deriving
habitat evaluation procedures capable of synthesizing habitat components across entire wetlands.
Quantifying the Consequences of Alteration of Vegetated Aquatic Habitats
After identifying important species that depend in some way on aquatic vegetation, and having
defined the suite of vegetation attributes relevant to those organisms, the research will then
quantify the relationships between the two. Quantifying species-vegetated habitat relationships
in sufficient detail to permit evaluation of both catastrophic and incremental habitat alteration
requires developing mechanistic or empirical relationships spanning a range of habitat extent and
characteristics. Insofar as possible, research also should identify key portions of the response
relationships including response thresholds, maximum biological potential, and levels above
22
-------
which factors other than habitat limit populations. Because ecological constraints and co-factors
independent of vegetated habitat will also determine local biological potential, there is likely to
be a family of curves for different ecological conditions; for example, for different salinities,
wave energy regimes, or wetland types. Assigning habitats or systems to the appropriate
response curve will require development and evaluation of classification systems and methods
for determining ecological potential. This work will be done with the objective of contributing to
the larger-scale habitat considerations (described below). Therefore, achieving broad spatial
coverages for species-habitat relationships will require stratifying response curves by ecoregion
or latitude, devising methods for extrapolation outside the range of measured habitat types and
co-factors, and schemes for regionalizing population endpoints when specific populations of
interest have insufficiently broad distributions. The balance among empirical and
mechanistic investigations will depend on the strength of the linkage among the particular
species endpoint and the habitat, and on the spatial extent for which relationships are desired.
For example, situations where the primary concern is a particular species with identifiable
demographic bottlenecks that depend directly on specific, measurable habitat elements may lend
themselves to mechanistic species-habitat relationships. For other situations, where the
dependence on vegetated habitat is more diffuse and specific population bottlenecks cannot be
identified or where a suite of endpoints is applied over large regions, empirical relationships may
be more easily obtained and appropriate.
Products
APM 1A FY02 Listings of the high-priority species offish, shellfish, and aquatic-dependent
wildlife for study in each biogeographic region, and listings of the habitats that are considered to
be critical to each (WED).
APM 3A FY03 Penaeid shrimp dependence on seagrass habitat (GED).
APM 3B FY03 Finfish dependence on seagrass and oyster reef habitats (GED).
APM 3D FY04 Report characterizing the relationship between alteration of vegetated habitats
and nekton use of those habitats (AED).
APM 3F (GPRA #58) FY04 Preliminary report characterizing relationships between abundance,
quality, and arrangement of various habitat types and selected biotic assessment endpoints in
coastal systems (WED).
APM 5B FY06 Report characterizing relationships between abundance, quality, and arrangement
of various habitat types and selected biotic assessment endpoints in coastal systems (WED, AED,
GED, MED).
APM 6C FY08 Synthesized quantitative species-habitat relationships suitable for developing
regional habitat-based biocriteria for shorelines, lakes, and estuaries (AED, GED, MED, WED).
23
-------
Benefits of Products
The Office of Water will be supplied with information on the effects of alteration of aquatic
vegetated habitat on key endpoints (fish, shellfish, and wildlife populations) to support
development of policies protective of these societally important endpoints.
Project Title 2: Shoreline, Lake, and Estuary Scale Habitat Research
Project Coordination and Resources (5.75 FTEs: AED-2.0, GED-0.75, MED-2.0, WED-1.0)
Objectives
Identify the high-priority populations offish, shellfish, and wildlife in each region;
identify the habitats that are critical to these populations; and characterize the
contributions of each habitat to life-support for these populations. Much of this objective
will be accomplished through synthesizing the available literature.
Develop and validate habitat alteration-population response relationships (classified,
quantitative models) for the identified species and habitats in each region at the scale of
the shoreline, lake, or estuary.
Where feasible, develop and validate comprehensive multi-species models to predict
quantitative changes in fish, shellfish, and wildlife resource value that would result from
habitat alteration to a habitat-mapped shoreline, lake, or estuary.
Scientific Approach (Overview)
This subcomponent of NHEERL habitat research will focus on economically valuable and
charismatic coastal species that use multiple habitats. Population responses will be evaluated at
the scale of an aquatic shoreline (including shallow and intertidal habitats through deeper water
habitats) or at the scale of an entire lake, cove, estuary, or subestuary. This approach is needed
because many fisheries and wildlife species depend on several habitats in their life histories and
migratory patterns. Houde and Rutherford (1993) list 21 estuarine-dependent species that make
up more than 50% of all U.S. commercial fisheries landings, exclusive of Alaska pollock. All 21
of these species depend on multiple habitats at one or more stages of their life histories. EPA
needs to plan research that will examine availability and alteration of multiple habitats within
lakes and estuaries.
The primary goal of this work is to produce habitat alteration-population response models for
high-priority populations offish, shellfish, and wildlife. These habitat alteration-population
response models will be designed to fit into spatially explicit risk assessment population models.
A subsequent goal is to produce larger, comprehensive, multi-species models that can integrate
single-species models to predict the total consequences of habitat alteration to a suite of
economically valuable and charismatic species. Where vegetated habitats are involved, this work
will be conducted with strong ties to project 1. These research plans differ in that this project
examines all the major vegetated and unvegetated habitats at larger scales, but delivers a cruder
24
-------
level of detail for the individual habitats compared to work proposed in project 2. Where overlap
occurs, both groups of habitat researchers will work together to convey data and findings, and to
avoid duplication of effort. This collaboration will occur at every step of the scientific approach
described here. Other collaborations within EPA and with different Federal agencies also will be
necessary to accomplish these research goals. These collaborations need to be actively pursued
and nurtured from the onset of this research.
Scientific Approach (Research Steps)
1. Identify suites of species and habitats that are most critical in each region, and characterize
each habitat's contributions towards the survival of these species.
Species for study in each region will be prioritized and selected as described above in the
introduction to the section on Research Projects. For each selected species, key habitats will be
identified. These habitats might be defined on the basis of "bioengineering" species (such as
S AV, or burrowing shrimp), or on the basis of depth and substrate. It will be necessary to
determine how important each "critical" habitat is relative to the other habitats these populations
utilize. To this end, an understanding of how each habitat functions to support these populations
will greatly help the development of habitat alteration-population response models. For estuarine
and marine fishes and shellfish, the National Marine Fisheries Service (NMFS) Essential Fish
Habitat work has already gone through this process, and summaries of the literature are readily
available. Other compiled resources such as the FishBase Dataset, the FWS Species and
Community Profiles, and the NOAA/NMFS Technical Memoranda will also be very useful. In
many cases, substantial aspects of this research step can probably be accomplished by
synthesizing the available literature.
2. Quantify population responses to habitat alteration at the shoreline, lake, or estuary scale for
the identified high-priority species.
Simple validated models will be developed to quantitatively predict changes in societally
valuable populations due to areal loss (either partial or total) of a given habitat within the spatial
mosaic of habitats that constitute an aquatic shoreline or an entire lake or estuarine system. This
is the central focus of this research component. Development of single-species models to show
changes to high-priority populations that result from areal habitat loss will be a worthwhile
stand-alone goal for NHEERL. Once these single-species models have been developed, research
can proceed to research step 3 (below), the development of multi-species models.
The single-species models may be based on quantitative empirical measures of population
density, production, or export in each habitat; or on quantitative estimates of fecundity and
survival. As appropriate (when species life history dictates, or when broad regional coverage is
needed), models also may be based on relative measures of population response. In each region,
EPA researchers will need to determine which approach is best suited to accomplishing the goals
of this research, with an eye towards ultimate application to risk assessment models.
As a first step in model development, researchers will consider the available literature, assemble
existing models and data, and determine appropriate research approaches for each habitat and
25
-------
population. Wherever possible, existing models and data will be used to develop these habitat
alteration-population response models. When this is not possible, researchers will plan the
research needed to develop the models.
This work will be done from the aquatic shoreline scale (including intertidal and shallow habitats
through deep water habitats) to the whole-lake/whole-estuary scale, as appropriate for the
populations in question. The majority of lake research will probably take place in the U.S. Great
Lakes, but smaller lakes are not excluded from consideration. In addition to developing these
habitat alteration-population response models, targeted community-level and process-level work
may be required to achieve the desired results. For example, it may be necessary to quantify key
functions of certain habitats, or to examine how species in habitats of interest are linked to
adjacent habitats. In each situation, proposed research will be designed to support habitat
alteration-population response models.
An important point is that these models need not consider the full life history of each population.
The goal of this work is to provide the scientific basis for resolving environmental protection
policy and regulatory questions, not for policies to enhance fishable biomass, and so models may
focus on specific stages of valuable species (e.g., juveniles). These "population models" may
then actually model a sub-population, a single life history stage, or a specific "bottleneck" in a
population. Otherwise, full-life-history models would need to consider harvest and over fishing,
which is only going to be of incidental policy interest to EPA regulatory staff. The models we
propose are designed to provide the necessary scientific basis to protect habitats by relating
habitat characteristics to population or sub-population endpoints for fish, shellfish, and wildlife.
Spatial habitat mapping also will be a component of this work. The consequences of habitat loss
need to be considered in the context of how much of each key habitat is available in the area of
interest. Habitat mapping a shoreline, lake, estuary, or subestuary will be a vital tool in
application of models. For this reason, spatial habitat mapping also should be a part of the
quantitative research involved in creating these models.
In some cases when developing these models, it will be possible to link the status of a certain
population to alteration of a single (often vegetated) habitat. This is most realistic for species
that are very tightly tied at key life stages to single habitats that act as population "bottlenecks".
This may apply to penaeid shrimp and vegetated marsh and seagrass habitats in the Gulf of
Mexico, to bay scallops and seagrass habitat in the Mid-Atlantic and Northeast, and to other
species and habitats. In these situations, project 1 will be a more appropriate means of
determining population response. Project 1 evaluates in better detail at the single-habitat scale.
Project 2 is intended to be fairly crude at the scale of the individual habitat, looking primarily at
the effects of areal habitat loss on populations within a larger setting; the shoreline, lake, or
estuary. Alternatively, project 2 analyses may look at how habitat mosaics and landscape
patterns affect the identified populations. In order to best meet the goals of the Aquatic Stressors
Framework, projects 1 and 2 will need to work closely together. Approaches, data, and results
will be shared between the two groups to produce the best product and to eliminate duplication of
effort. In most cases, results of the more detailed habitat alteration work conducted under project
1 will be directly incorporated into the larger scale habitat alteration-population response models
described here.
26
-------
For some populations, a single-habitat approach as in project 1 above may not be the most
appropriate model. In particular, many economically valuable fishes are mobile, migrating from
habitat to habitat. The scale which affects these populations most is that of an entire aquatic
shoreline or even an entire lake or estuary, all of which are composed of many interconnected
habitats. The shoreline approach, which considers the entire nearshore mosaic of habitats from
intertidal through shallow subtidal and into deeper habitats, may be the smallest scale that can be
applied meaningfully to mobile, economically valuable fmfishes. Many of these species recruit
into nearshore habitats (marshes, seagrass beds, and mud and sand shallows) for their juvenile
development, then move into deeper habitats as they grow. Shallow waters, whether vegetated or
not, provide juveniles with a significant refuge from predation by larger aquatic predators (Ruiz
etal. 1993). Shallow waters are also very susceptible to habitat destruction. The general life
history pattern of "shallow-to-deeper" is true for many fmfishes and mobile shellfish. Research
at the scale of a shoreline, estuary, or lake is appropriate for species with these life history
patterns. Thus the two projects will work together to deliver habitat alteration-population
response models for high priority species.
Researchers also will need to consider temporal variability in constructing these models. Given
the short time frame for results (~ 8 years) relative to the time frame for cyclical fluctuations of
some aquatic populations, researchers will need to consult historical long-term data sets for the
species of primary concern. Fortunately, the focus on commercially and recreationally important
aquatic species increases the likelihood that long term abundance data will be available.
In order to achieve larger goals, continuing collaboration and coordination among Divisions will
strive towards the goal of establishing comparable and quantitative methodologies. These efforts
will also link closely to different Aquatic Stressors research implementation plans that are
delivering habitat alteration as stressor-response endpoints. Examples are the Nutrients plan
(Section 5) and collaboration with other EPA laboratories, and other Federal and State agencies
also will be required to develop the best possible products. Another important aspect of model
development will involve determination of the data quality required to produce models with
adequate predictive power. Models will be designed to provide valid information EPA needs to
meet its regulatory requirements.
3. Develop comprehensive multi-species models to quantify population responses to habitat
alteration at the shoreline, lake, or estuary scale.
A series of validated habitat alteration-population response relationships for the individual high
priority species at the shoreline-, lake-, or estuary-scale (as described above) is the first-order
goal of these efforts. Development of these single-species models would be a valuable and
sufficient contribution of this project. However, where feasible, this work will be taken a step
further. A subsequent goal of this work is, for each region, to develop comprehensive multi-
species models that can quantify the effects of habitat alteration on the entire suite of
economically valuable and charismatic populations within a shoreline, lake, or estuary. These
comprehensive multi-species models will be constructed initially by combining the species
models described above, taking care to ensure that the sum of these models represents the
complete set of major species.
27
-------
As a whole-system example, consider seagrass habitat. When seagrass habitat is lost, it is not
replaced with a lifeless void, but more typically with macroalgal or unvegetated habitats.
Seagrass loss may cause populations of some valuable species to decrease, but populations of
other valuable species in the estuary may simply move over, or may even increase. The total
impacts of loss of a single habitat to aquatic populations within a lake or estuary must be
assessed first, with consideration of those habitats that will replace the lost habitat; and second,
with consideration of the entire suite of economically valuable fish, shellfish, and wildlife that
will be affected. If only a small number of the important species are considered,
"comprehensive" habitat alteration models will be driven by the initial selection of species, and
may not reflect the true effects of habitat alteration. EPA ultimately wants to evaluate the
societal impacts of habitat alteration. Though ambitious, it is therefore a priority (where
possible) to develop truly comprehensive multi-species models that can examine habitat
alteration and accurately predict a large majority of the impacts to economically valuable and
charismatic species.
These synthesis-oriented comprehensive models might also consider emergent properties that
develop from combinations and arrangements of habitats within a shoreline, lake, or estuary. For
example, a diversity of habitats arranged as a patchwork may or may not support more fish,
shellfish, and wildlife than would uninterrupted expanses of the same habitats, or certain
combinations of habitats, such as marsh and adjacent SAV, maybe of particular value to fish,
shellfish, and wildlife. These models might further consider how certain sentinel species are
linked to the health of the ecosystems upon which they depend. Development of comprehensive
multi-species models will require a synthesis phase (and iteration) to assemble the individual
habitat alteration-population response relationships. The result will be validated comprehensive
models, based on quantitative data, that can predict the total consequences of areal habitat loss to
the great majority of economically valuable and charismatic populations. These comprehensive
models also should be designed to place a quantitative fish/shellfish/wildlife resource value on
shorelines, lakes, and estuaries, based on spatial habitat mapping. This resource value should
reflect aquatic populations, and need not be tied to monetary standards. Multi-species models
should predict quantitative changes in aquatic resource value that would result from habitat
alteration to any mapped shoreline, lake, or estuary. This will allow better assessment and
management of aquatic habitats and resources.
Both the individual models and the comprehensive multi-species models should be designed
around application to a risk-assessment framework. This will allow insertion of the habitat
models into larger spatially explicit risk assessment models that can consider multiple stressors
including habitat alteration, toxic chemicals, and others.
4. Develop classification schemes within each identified habitat or system type where other
important factors (salinity, geomorphology, and tidal energy) will affect the ability of habitats or
systems to support priority populations.
A classification scheme is necessary to allow appropriate application of the single-species and
multi-species habitat alteration-population response models described above. For example,
juvenile summer flounder settle in shallow sandy or muddy mesohaline and polyhaline estuarine
habitats in the Mid-Atlantic, but do not utilize oligohaline areas (Rogers and Van Den Avyle
28
-------
1983). As support for summer flounder populations, Mid-Atlantic shallows should be classified
by salinity. Other examples of classification may include lagoonal systems versus riverine
systems versus embayments, and microtidal systems versus macrotidal systems. Regionalization
will, of necessity, be part of this exercise as well. In some cases, these questions of habitat
classification will be answerable through the literature; in other cases new research efforts may
be required.
Products
APM 1A FY02 Listings of the high-priority species offish, shellfish, and aquatic-dependent
wildlife for study in each biogeographic region, and listings of the habitats that are considered to
be critical to each (WED).
APM 3F (GPRA #58) FY04 Preliminary report characterizing relationships between abundance,
quality, and arrangement of various habitat types and selected biotic assessment endpoints in
coastal systems (WED).
APM 5B FY06 Report characterizing relationships between abundance, quality, and arrangement
of various habitat types and selected biotic assessment endpoints in coastal systems (WED, AED,
GED, MED).
APM 6C FY08 Synthesized quantitative species-habitat relationships suitable for developing
regional habitat-based biocriteria for shorelines, lakes, and estuaries (AED, GED, MED, WED).
Benefits of Products
This work is designed to provide EPA Program Offices and Regions with simple validated
models that can quantify the effects of alteration or loss (either partial or total) of any major lake
or estuarine habitat on populations of economically valuable or charismatic fish, shellfish, and
wildlife. This work will also be of value to other Federal, State, and local managers striving to
protect living aquatic resources from degradation due to habitat alteration.
This work will also link to the load-response efforts proposed in the Nutrients research
implementation plan (Section 5), in that habitat alteration can occur through destruction and
fragmentation, nutrient loadings, or other stressors. The products in the Nutrient plan (Section 5)
regarding load-response models include loss of SAV habitat, loss of benthic habitat due to
hypoxia and anoxia, and increases in macroalgal habitat; but do not focus on fish, shellfish, and
wildlife. The aquatic shoreline, whole-estuary, and whole-lake scale habitat alteration-
population response work proposed here can tie the nutrient habitat endpoints to population
endpoints for economically valuable or charismatic species.
Since these habitat alteration-population response models will be designed to fit within a risk-
assessment framework by quantitatively linking habitat alteration to population response, this
work can also integrate into larger, multi-scalar, spatially explicit, multiple stressor risk
assessment models.
29
-------
Project Title 3. Salmon and Native Fish Habitat Research
Project Coordination and Resources (9.4 FTEs: MED-1.3, WED-8.1).
Introduction
The research described in this project deals with the influence of human activities on aquatic and
aquatic-dependent biota at landscape, watershed, and regional scales. Specifically, it will
examine watershed and landscape scale habitat issues affecting salmon and native fishes in the
Pacific Northwest, and fishes reliant upon Great Lake coastal wetlands.
Objectives
To evaluate and to quantify the influence of human activities at the landscape and
watershed scales on native fish habitat and fish populations, including wild Pacific
salmon and economically and ecologically important Great Lakes fishes.
To evaluate how habitat spatial structure and connectivity of habitat in stream networks,
wetlands, lakes, and estuaries influence native fishes, including wild Pacific salmon and
wetland-dependent fish populations and overall biodiversity.
Scientific Approach
Although many aspects of aquatic habitat-fish population relationships have been studied, many
knowledge gaps exist. Relatively little attention has been focused on the relationships between
landscape structure and fish assemblages, and landscape structure and aquatic habitat.
Population declines of salmon and other native fish accentuate the need for the quantification of
these landscape relationships. In the report, From the Edge: Science to Support Restoration of
Pacific Salmon, the Committee on Environment and Natural Resources (CENR) identified
science needs for Pacific salmon and related species (CENR 2000). CENR indicated that habitat
for salmonids and all native aquatic species, and hence their populations, are strongly influenced
by watershed conditions at a landscape scale. Modeling and decision support tools are required
to incorporate land use change relative to habitat on the extensive spatial scale, and must
incorporate temporal changes (habitats are dynamic).
The research will be conducted in two regions, the Great Lakes and the Pacific Northwest. In the
Great Lakes, research will focus on coastal wetland fish assemblages. There are approximately
200 species offish in the Great Lakes. It is estimated that about 90% of those species are directly
dependent on coastal wetlands for some aspect of their life history. Among those species that are
heavily dependent on coastal wetlands are yellow perch, northern pike, largemouth bass, walleye,
and a number of forage fishes (Jude and Pampas 1992, Brazner 1997). All of these populations
also have relatively important commercial and/or sport fisheries throughout the Great Lakes and
all appear to be in decline. Habitat alteration is thought to be the most important contributor to
these declines, but over-fishing, pollutants, and exotics are also considered important threats
(Whillans 1992). In the Pacific Northwest, research will focus on wild Pacific salmon and native
fish. Many of the anadromous salmonids populations in the Pacific Northwest are in serious
30
-------
decline, and numerous populations are now listed under ESA. Landscape change, water
pollution, introduced predators, fishing, hydro power development, disadvantageous ocean
conditions, and other factors have led to the extinction or decline of many stocks (Bauer and
Ralph 1999, CENR 2000). Research will be developed to allow for comparisons of factors
influencing native fish assemblages in the two regions.
Upland and Riparian Effects on In-stream and Coastal Wetland Condition
In the Pacific Northwest, these efforts will be based on an integrated modeling/field study
approach. An existing model developed by NMFS (1992) simulates coho salmon population
dynamics based on in-stream habitat condition. For this model, in-stream habitat condition was
determined through simple stream reach classification that does not reflect watershed land
use/land cover conditions. If, however, we are to be able to examine how upland management
affects fish dynamics, then it is necessary to understand how in-stream habitat condition is
influenced by the surrounding uplands and riparian areas. Shading by riparian trees, woody
debris supply, non-point source introduction of sediments and nutrients, and landslides are all
examples of important upland processes that can affect in-stream habitat condition and which
could be influenced by upland management actions. Such information also allows us to predict
habitat condition, based on upland characteristics, at locations which have not been sampled.
Besides affecting habitat condition, upland factors can also influence fish mobility. For example,
warm water temperatures or landslides could reduce or completely prevent fish movement
between stream reaches. Another important upland/riparian issue associated with the restoration
of Pacific salmon is the possible need for nutrient additions (i.e., raw or processed salmon
carcasses, and commercially produced organic or inorganic fertilizers) to headwaters (e.g.,
watersheds, lakes, or streams) to compensate for the loss of marine derived nutrients previously
supplied by healthy salmon populations. Determining the ecological effects of surrounding
upland areas on in-stream condition is therefore a critical component of our research.
Technical approaches to examining upland effects on in-stream condition could include field
studies, empirical modeling, and process modeling. Empirical modeling approaches would
develop correlations between upland independent variables and in-stream response variables.
Upland variables could be derived from GIS DATANET, and could be used to represent the
watershed as a whole or the riparian zone in particular. Data for explanatory and response
variables could be obtained through field sampling, other research projects (e.g., Environmental
Monitoring and Assessment Program, EMAP) or agencies, or through the literature. Process
models would relate upland factors to in-stream condition based on specific processes. Examples
include a model that predicts in-stream suspended sediment concentration based on soil
characteristics, slope, upstream load, or a physical model that calculates water temperature based
on shading by trees. Other modeling approaches are also available. We envision linking such
models with a salmon population model to be able to examine the influence of land use changes
on salmon and fish populations.
In the Great Lakes, initial efforts to understand landscape influences on coastal wetland habitat
condition and native fishes will be based on field studies designed to help build quantitative
empirical models that can eventually be used to construct more process-oriented models. The
current knowledge base related to watershed fragmentation effects in Great Lakes coastal
31
-------
wetlands is quite limited. Although watershed fragmentation has been shown to be related to the
structure and function of biotic assemblages in streams from Western Lake Superior (Detenbeck
et al. 2000), fragmentation effects on habitat condition and biota associated with Great Lakes
coastal wetlands have not been well documented. We suspect that this sort of fragmentation-
induced habitat alteration also will cause changes in higher tropic levels in Great Lakes coastal
wetlands.
Some of our endpoints for testing this idea may be yellow perch and northern pike population
abundances, and overall fish biodiversity at river-influenced coastal wetlands having differences
in fragmentation levels in their watersheds. GIS characterization, statistical analysis, and model
development sequences will parallel those planned for stream watersheds in the Pacific
Northwest. We plan to refine our definition of fragmentation to incorporate a variety of land-use
types by using a "land-use equivalency" approach which will allow us to place our wetland sites
along a vulnerability gradient, and provide a better opportunity to link watershed land-use to
habitat and fish response curves for Great Lakes coastal wetlands. Collaborations with
researchers from other institutions will allow us to increase the number of sites where fish data
will be available and expand the vulnerability gradient to include much of the Great Lakes that
would not otherwise have been possible.
Because we think the response to watershed fragmentation by wetland fishes will vary with
wetland type, we will need to test whether a wetland classification system effectively groups
coastal wetlands into similar response classes. Although there is a presumption that coastal
wetland hydro geomorphology influences biota, there is little direct supporting evidence. It is
well known, however, that aquatic community structure of higher tropic levels is influenced by
vegetation structure in coastal wetlands, and vegetation structure appears to be related to
hydrogeomorphology. So, it seems likely that our fish response variables (population and
assemblage level) also will be related to differences in wetland hydrogeomorphology. After
assessing whether this is the case at different coastal wetland types (e.g., open estuary, barred
estuary, barrier beach lagoons, open coastal), we will be better able to extrapolate the
significance of our fragmentation results on a region-wide basis by knowing the distribution of
wetland types across the landscape and the fragmentation levels in their watersheds.
Effects of Network Structure and Connectivity on Fish Movement
Because fish are mobile, they are not limited to nor exclusively influenced by the habitat quality
of a single stream reach. Rather, they move between reaches and may require different habitat
conditions during different life stages. The spatial distribution of habitat condition and the
ability offish to move between reaches are therefore important considerations. For example,
salmonids returning from the ocean attempt to reach the same stream reach in which they were
spawned. Any obstruction in the stream network, which forces them to expend more energy to
return, could affect spawning success. If a barrier completely prevented them from returning to a
particular home reach, then the ability of strays to decolonize new habitat would depend on the
spatial distribution of habitat near the home reach and the occurrence of other obstructions to
movement. Thus any effort aimed at examining watershed management effects on fish
populations needs to consider the effect of the watershed on the spatial structure of the network
(e.g., the distribution of habitat condition) and on the level of connectivity among stream reaches.
32
-------
This work will primarily be conducted in the Pacific Northwest, but the hydrogeomorphic
classification being tested in Great Lakes coastal wetlands is based on hydrologic connectivity of
wetlands to adjacent lake habitats and their watersheds, so many of the same habitat connectivity
issues are relevant to the Great Lakes studies and will be implicitly incorporated into their design.
In the Pacific Northwest, the approach to examining the effects of network structure and
connectivity will be to build a spatially explicit network data structure that includes habitat
quality and connectivity attributes and which can be linked to specific biological response
models. Such a network structure could be used in a number of ways. For example, it might be
desirable to conduct simulations of several specific drainage networks, and to compare results
between basins with high habitat quality and low habitat quality. Alternatively, it might be
desirable to examine the effect of certain watershed characteristics (e.g., slope, catchment area,
stream density) on fish dynamics by systematically varying those characteristics using synthetic
landscapes.
Biological Response of Fish to Habitat and Stream Network
EPA has responsibilities under the CWA to restore and maintain the biological integrity of the
nation's waters. Therefore, it is desirable to understand how activities aimed at managing
salmon would affect other fish species, in particular, native fish. To address these needs, field
research and modeling efforts will be developed to examine how management actions would
affect dynamics of various fish groups. This may include modeling at different levels of
organization. First, species-level models would examine the biological response of particular
species to watershed and network structure. Models would be run separately for salmon and
possibly a few other species representative of different life history strategies. This will allow us
to examine how salmon and fish with different habitat needs respond to a common set of
management actions. Second, the biological response modeling could also include exploratory
assemblage-level modeling. In this case the dynamic behavior being tracked is overall species
richness, rather than abundance of a particular species. This allows us to examine community-
level response to management actions.
Products
APM 2A FY03 Prototype watershed-stream network modeling approach for Pacific salmon
(WED).
APM 3C FY04 Report characterizing the relationship between habitat in stream networks and
salmon-native fish for coastal Oregon watersheds (WED).
APM 6A FY04 Ecological consequences of marine derived nutrients and nutrient enrichment for
aquatic biota and stream habitat quality, with an emphasis on salmon and native fish (WED).
APM 4 A FY05 Develop indices of watershed integrity based on land use/land cover and
relationships to fish (WED, MED).
33
-------
APM 5A FY05 Reports characterizing the relationship between landscape-scale habitat mosaics
and native fish by wetland type in the Great Lakes (MED).
APM 6B FY07 Regional models of landscape influence of salmon/native fish in the Pacific
Northwest and native fish in Great Lake coastal wetlands (WED, MED).
APM 6D FY08 Interactions between stream nutrients and habitat alteration on water quality and
aquatic life (WED).
Benefits of Products
Research will allow explicit evaluation of human activities at landscape and watershed scales on
salmon and native fish. This will be of direct benefit to OW, EPA Regions, and an interagency
effort on salmon restoration.
Project Title 4. Multiple Stressor Risks to Common Loon and Other Piscivorous Bird
Populations (cross-listed in Section 7, Toxic Chemicals, Project B3)
Project Coordination and Resources (1.5 FTEs)
AED-1.5, and 6.9 additional FTEs devoted to mercury-loon research described in Section
7,Toxic Chemicals, project B3.
Introduction
This project examines the interactive effects of multiple stressors, including landscape-level
habitat alteration and mercury, on common loons and other piscivorous bird populations ('loon
project'). This project was developed as a case study implementing NHEERL's WRS (EPA
2000), and demonstrating an integrated approach to large scale, population-landscape-stressor
assessments. There are significant habitat components to this project, including evaluating the
spatial configuration of loon habitat and mercury impacts in the landscape mosaic and the issue
of scaling up from local to regional impact assessments. Because habitat and toxic chemicals
issues are integrally linked within the demonstration project, the project is relevant to issues
within this section, as well those relative to toxic chemicals (Section 7). Therefore, a brief
description of those elements of the loon project related to the assessment of risks of habitat
alterations at multiple geographic scales, appears here as project 4. To avoid redundancy, a
complete description of the project appears in Section 7, project B3.
Two key research areas, defined within the WRS and described below, reflect the need to
consider landscape context and scale in order to achieve the scientific and management goals of
the risk assessment.
1. Research to Advance Techniques for Assessing the Relative Risk of Chemical and Non-
chemical Stressors on Wildlife Populations.
34
-------
Landscape characterization studies, combined with experimental approaches, are required to
better quantify the relative impacts of chemical stressors, habitat alterations, and the introduction
of exotic species on wildlife populations. Associated with this effort is the need to develop and
integrate predictive models so that the outcome of different management scenarios, based on
chemical loading, habitat alterations, exotic species control, and other management options, can
be quantified.
2. Research to Define Appropriate Geographical Regions/Spatial Scales for Wildlife Risk
Assessments.
A significant effort is needed to define scientifically credible spatial scales for wildlife risk
assessments. Habitat requirements for wildlife species associated with aquatic and terrestrial
ecosystems must be established and referenced to regulatory jurisdictions to ensure coordinated
implementation of risk-based decisions. A consensus on current or potential habitat ranges are
needed to identify wildlife species of concern and to evaluate approaches in risk assessments that
consider spatial population structure (EPA 2000).
Within the WRS, three major research objectives have been defined to address these needs. The
third objective defines the focus for the habitat alteration component of the loon project, i.e., the
development of approaches for evaluating relative risks from chemical and non-chemical
stressors on spatially structured wildlife populations across large areas or regions ("geospatial
modeling"). Research described in the loon project will address issues associated with the spatial
and temporal heterogeneity of populations and stressors in real landscapes. This landscape
context provides a basis for understanding and quantifying how spatio-temporally varying
stressors influence the distribution of wildlife populations. Thus, the approaches, models, and
methods developed within this project are designed both to assess risks from multiple stressors
and evaluate the relative effectiveness of alternative management strategies.
Objectives
Specific to the goals for habitat alteration research, the relevant research objective can be
described as the development of approaches for evaluating the relative risks from chemical and
non-chemical stressors on spatially-structured wildlife populations across large areas or regions.
Consistent with this objective and to address the WRS objectives described above, research
activities within the loon demonstration project focus on the development of geospatial modeling
methods to assess the relative impact of heterogeneously distributed stressors, including dietary
methylmercury, habitat degradation, acidification, and human disturbance on populations of the
common loon, which is resident to the northeastern portion of the U.S. and Canada. For this
purpose, research activities will include the develop of methods to identify spatial relationships
among stressors (i.e., correlations in distributions), potential interactions among stressors, and the
relative risks among potential stressors to populations of loons at varying spatial scales.
Scientific Approach
Consistent with the approach described generally for habitat alteration research, there are two
distinct research components for habitat research within the loon project. The first of these
35
-------
addresses the need to describe the relationship between habitat and biota ("habitat/suitability
models") at appropriate spatial scales and with sufficient detail and resolution to quantify the
effects of both incremental and catastrophic habitat alteration. The second step involves
integrating habitat alteration effects with those of other stressors ("integration within a spatial
context"), so that resource protection and restoration priorities can be evaluated at spatial scales
up to and including regions or large receiving bodies.
For this project, the focal species was selected partly to take advantage of existing data that may
permit the development of linkages between habitat and biological fitness. Specifically, fine-
scaled spatially-referenced information on presence and condition of individual loons across
large geographic areas is available through long-term loon monitoring programs that exist in the
upper Midwest and the Northeast. In addition to information from these programs, available
monitoring databases and/or aerial photographs, provide information to characterize habitat
quality. For loons, key habitat characteristics may include the presence of suitable nesting and
brood rearing sites, measures of human disturbance, density or extent of human dwellings and
other activities around lakes, turbidity, and availability of suitable forage fish supplies. This
unique set of spatially-referenced data will permit the development of habitat suitability
approaches and models, relating environment factors and biological condition.
Within the loon project, integration within a spatial context has been approached through the
development and application of spatially-explicit population models that incorporate stressor-
response relationships that will be applied within the spatially-diverse landscape. Specifically,
within this project, spatial models will be used to evaluate how loon life history, spatial
heterogeneity, and interactions among stressors in the landscape drive the relationship among
breeding success on individual lakes and population trends across broad regions. These models
will be used to: 1) define what constitutes a population (within the context of the assessment
question) and how sub-populations interact in a heterogeneous landscape, 2) determine the
appropriate spatial scale for assessment questions, and 3) determine the relative risks presented
by different stressors. This model development would be a primary objective of this
demonstration project.
Products
As defined within this plan, these specific products will be developed:
APM 2B FY04 Habitat suitability indices to support population models for projecting relative
risks of multiple stressors including toxic chemicals and habitat alteration to common loons
(AED).
Also see Section 7 Toxic Chemicals, Project B3 for associated products.
Benefits of Research
This research will allow explicit evaluation of multiple stressors on piscivorous wildlife and lead
to the development of risk-based criteria. This will be of direct benefit to Program Offices,
Regions, and interagency efforts to protect important wildlife species. More broadly, this
36
-------
research will permit an evolution towards a landscape assessment approach for examining critical
environmental problems over larger spatial scales and the assessment of the cumulative risk
resulting from multiple stressors. This approach allows for a more comprehensive perspective
for evaluating the condition of communities, watersheds and ecoregions. A landscape
assessment approach also provides the ability to evaluate the status and trends of a variety of
ecological resources at multiple scales so that relationships of stressors and effects can be
developed to establish conditions which are influencing the impacts on wildlife populations.
Gap Analysis
A. The following research is within the scope of this plan but outside NHEERL's current
manpower, expertise, or sampling capability:
Assessment data required to extrapolate habitat-biota relationships studied by ORD to the
population of all systems for which nutrient, sediment, and biocriteria are required will
necessitate collaboration with other Federal and State agencies, non-governmental
organizations, and academic institutions.
Near shore fish sampling in Great Lakes (e.g., abundance of commercially and
recreationally valuable species) to support multiple habitat research will require
collaboration with entities capable of open-water fish sampling.
Development of habitat alteration-population response models for the southern Atlantic
coast (e.g., Carolinian biogeographic province), the southern Pacific coast (e.g., San
Diego biogeographic province), and Puerto Rico and the U.S. Virgin Islands, especially
for populations of commercially valuable fish and shellfish.
Survey of songbird, hawk, and waterfowl utilization of reference coastal wetlands of high
ecological integrity, and of disturbed wetlands of varying anthropogenic alteration are
currently outside NHEERL expertise and/or manpower. Wetlands of interest for the
surveys are all coastal vegetated habitats such as fresh- and salt-water marshes, SAV, and
emergent aquatic vegetation.
Development of quantitative methods to evaluate the restoration success of the structure
and function of habitats that support populations of commercially valuable fish, shellfish,
and wildlife.
B. The following research is outside the scope of this plan (which focuses primarily on fish,
shellfish, and wildlife population endpoints of concern to society), but relates to other ecological
endpoints that may be also relevant to society:
Understanding of the effects of habitat alteration on fish and other biotic assemblages
(e.g., zoobenthos, macroinvertebrates) from both a structural (including biodiversity) and
functional perspective even where these responses cannot be immediately linked to fish,
shellfish, and wildlife populations of interest.
37
-------
Understanding of the effects of habitat alteration (both within watersheds and their
coastal ecosystems) on the exchange of materials (e.g., dissolved nutrients, particulate
organic matter, organisms) with adjacent waters where these responses cannot be
immediately linked to fish, shellfish, and wildlife populations of interest.
Role of areas of critical habitats along the coast in providing flood and erosion control,
and an understanding of the effects of habitat alteration (both within watersheds and their
coastal ecosystems) on the retention of sediments by coastal ecosystems.
Importance of areas of critical habitats in watersheds adjacent to coastal waters in
providing nutrient filtration (e.g., riparian zones, inland wetlands, forests, salt marshes).
Effect of nonindigenous organisms on the structure and function of wetlands and other
critical habitats.
Mechanistic understanding of nutrient effects on vegetation and tropic structure of critical
habitats.
References
Bauer, S.B., Ralph, S.C. 1999. Aquatic habitat indicators and their application to water quality
objectives within the Clean Water Act. EPA 910-R-99-014. EPA, Region 10.
Brazner, J.C. 1997. Regional, habitat, and human development influences on coastal wetland
and beach fish assemblages in Green Bay, Lake Michigan. J. Great Lakes Res. 23:36-51.
CENR. 2000. From the edge: science to support restoration of Pacific salmon. National Science
and Technology Council.
Detenbeck, N.E., Arthur, J.W., Bertelsen, S.L., Brazner, J.C., Snarski, V.M., Taylor, D.L.,
Thompson, J.A. 2000. Western Lake Superior comparative watershed study.
Environ. Toxicol. Chem. 19:1174-1181.
EPA. 1990. Reducing risk: setting priorities and strategies for environmental protection.
SAB-EC-90-021.26pp.
EPA. 1999. EPA's framework for community-based environmental protection. EPA 237-K-99-
001. Washington, DC. 40 pp.
EPA. 2000. Wildlife Research strategy. EPA. NHEERL/ORD. September.
Houde, E.D., Rutherford, E.S. 1993. Recent trends in estuarine fisheries: predictions offish
production and yield. Estuaries 16:161-176.
Jude, D. J., Pampas, J. 1992. Fish utilization of Great Lakes coastal wetlands. J. Great Lakes
Res. 18:651-672.
38
-------
NMFS. 1992. Quantifying resource loss through habitat degradation: proceedings of the first
NMFS Northeast environmental workshop, March 13-14 1991. NMFS-F/NER-3. Northeast
Regional Operations Office, Gloucester, MA. 137 pp.
NRC. 1997. Building a Foundation for Sound Environmental Decisions. National Academy
Press, Washington, DC. 87 pp.
Rapport, D.J., Costanze, R., Epstein, P.R., Gaudet, C.L., Levins, R. 1998. Ecosystem Health.
Blackwell Science, Maiden, MA. 372 pp.
Rogers, S.G., Van Den Avyle, MJ. 1983. Species profiles: life histories and
environmental requirements for coastal fishes and invertebrates (South Atlantic) summer
flounder. FWS/OBS-82/11.15. U.S. Fish and Wildlife Service, U.S. Army Corps of Engineers,
TREL-82-4. 14pp.
Ruiz, G.M., Hines, A.H., Posey, M.H. 1993. Shallow water as a refuge habitat for fish and
crustaceans in non-vegetated estuaries: an example from Chesapeake Bay. Mar. Ecol.
Prog.Ser. 99:1-16.
Stein, B.A., Flack, S.R. 1997. 1997 species report card: the state of U.S. plants and animals. The
Nature Conservancy, Arlington, VA.
Warren, M.L., Burr, B.M. 1994. Status of freshwater fishes of the United States: overview of an
imperiled fauna. Fisheries 19:6-17'.
Whillans, T.H. 1992. Assessing threats to fishery values of Great Lakes wetlands. In Kusler, J.,
Smardon, R., eds., Wetlands of the Great Lakes: Protection, Restoration, Policies and Status of
the Science, Proceedings of the International Wetland Symposium, May 16-19, 1990, Niagara
Falls, NY, pp. 156-165.
39
-------
Section 5.
Implementation Plan for Nutrients Research
Problem
There is growing evidence that human activities have dramatically changed the amounts,
distribution, and movement of major nutrient elements (nitrogen-N and phosphorus-P) in the
landscape and have increased nutrient loading to receiving waters. Some of these changes affect
use of the nation's aquatic resources, and pose risks to human health and the environment (NRC
2000). EPA is in the process of developing guidelines that States and Tribes must use to set
nutrient criteria for our nation's waters. For waters failing to meet WQS, States and Tribes will
be required to develop TMDLs to eliminate the causes of non-attainment. Our current level of
understanding of aquatic ecosystem function is inadequate to allow us to extrapolate knowledge
of nutrient loading relationships for systems for which we have intensive data to accurately
predict adverse effects on specific systems with more limited data. NHEERL research will
provide the scientific information on load-response relationships that are required to develop
numeric nutrient criteria protective of aquatic life.
This research implementation plan is ambitious. A complete understanding of the effects of
nutrients on aquatic ecosystems will require additional research. The projects listed in this plan
outline the objectives needed to establish the scientific basis for WQC and TMDLs associated
with nutrients in coastal systems (coastal wetlands, embayments, estuaries, and near coastal
waters). The decision to focus on coastal waters is based on the complexities of these systems
and OW's prioritization of research needs by waterbody type. The most important response
categories for study are given below:
Increase in algal production (or carbon supply as defined by Nixon 1995) and/or changes in algal
community composition, which can result in harmful algal blooms (HABs), are principal
causative agents for the three following effects:
1. Low dissolved oxygen (DO) or hypoxia leading to fish kills or loss of shellfish and
degradation of benthic habitats;
2. Loss of naturalSAVhabitat, important to fish and other biota and due to changes in water
clarity, epiphytic growth, or smothering by invasive algae;
3. Shifts in basic food webs leading to loss of commercially important fisheries and overall
aquatic biodiversity.
The following pathways define our current understanding of how nutrients affect each of these
endpoints:
40
-------
Low DO
Increased nutrient loading to an estuary/receiving water stimulates primary production (largely,
though not exclusively, phytoplanktonic) and produces excess carbon biomass that sinks to the
bottom waters/benthos resulting in respiratory oxygen demand that may exceed oxygen supply.
For this endpoint, our objective is to define relationships between nutrient load and factors
affecting DO concentration in sensitive portions of receiving water bodies. As part of this effort,
NHEERL is developing DO requirements to protect indigenous species in various coastal regions
(North and South Atlantic, Gulf of Mexico, and Pacific). Criteria have been developed for the
Atlantic coastline between Cape Cod and Cape Hatteras (EPA 2000). Ultimately, this research
will provide relationships between nutrient inputs and DO concentrations, which will protect
indigenous species in various coastal regions.
SAVLoss
Increased nutrient loading can result in an accumulation of phytoplankton, epiphytic, and
macroalgal biomass/carbon that shades SAV or alters sediment geochemistry and results in loss
of areal coverage. For this endpoint, our objectives are to develop, for the nation's coastal
receiving waters, sufficient understanding of the relationship between SAV loss and nutrient
loading to provide a sound scientific basis for establishment of nutrient criteria that would protect
these important habitats from degradation or loss and aid in restoration efforts. This work will
directly support or interact with the Habitat Alteration (Section 4) and Diagnostics (Section 8)
research implementation plans.
Shifts in Food Webs
Changing nutrient loadings (includes increased loading, changes in loading ratios, and changes in
the mode/timing of delivery) alter species composition of primary producers. Effects of this shift
are transmitted through the food web, altering the consumer-food web dynamics (carbon or
energy flow pathway) in receiving waters. The result is a change in primary producers that does
not support existing food webs (and hence alters the biological integrity of ecosystems) and does
not support commercially important fish and shellfish production. The objective of food web
research is to identify nutrient loading thresholds that cause shifts in primary producers and other
key components of the food web. In addition, we will assess the use of food web structure and
processes to improve our ability to classify systems and to predict differences in response to
nutrients that affect hypoxia and SAV.
This research will require developing classification schemes for each of these endpoints so that
aquatic systems can be grouped according to their expected responses to nutrient loading, to aid
in the process of setting nutrient criteria and TMDLs. NRC (2000) recommends that
classification frameworks be developed that can be generalized to a broader range of features and
processes than the current classification schemes, which are focused on individual features (i.e.,
flushing or light). NHEERL's classification efforts will focus on coastal receiving waters
(including estuaries, near-coastal waters and the Great Lakes). Our classification efforts will
focus on understanding and linking the influence of physical, chemical, and biological factors to
the response to nutrients across the Nation's coastal receiving waters.
41
-------
NHEERL nutrient research efforts will provide nutrient-response relationships for diagnostic
efforts (Section 8) and, once additional critical habitats subject to effects of nutrients are
identified by habitat alteration research (Section 4), we will develop plans for nutrient load-
response research in those critical habitats. To maximize our efforts, we also must link our
efforts with other Laboratories in EPA (e.g., NERL), other Federal agencies (e.g., U.S.
Geological Survey [USGS]), NOAA, and academic institutions collecting data on coastal
systems. We have initiated efforts to provide common data and files across NHEERL's
Divisions and will continue to seek collaborations to strengthen our efforts.
Goals
The primary goal of this research is to provide the scientific basis and load-response relationships
that are required to develop numeric nutrient criteria protective of aquatic life. The focus of this
research is on coastal receiving waters and does not apply directly to the understanding of load-
response relationships for other water bodies (e.g., streams, rivers, lakes, and wetlands).
Therefore, NHEERL research will initially define and quantify relationships between nutrient
loading and ecological responses for coastal aquatic resources. The APGs and associated APMs
for this research are listed below. Note: some APGs and APMs (including those under GPRA)
were established before this document was written. They are listed here along with APGs and
APMs that were developed under the aquatic stressors process.
APG 1 FY02 (GPRA # 030) Provide a strategic approach for developing TMDL-driven
thresholds for protecting ecosystems from nutrients and sedimentation.
APM 1A FY02 (GPRA # 163) Generalized seagrass/rhizosphere model capable of
predicting effects of reduced light, sedimentation, nutrient depletion, and toxic effects of
sulfides (WED).
APM IB FY02 (GPRA # 164) Effects on estuarine submerged aquatic vegetation from
changes in light quantity and quality due to increased levels of suspended solids (GED).
APM 1C FY02 (GPRA # 165) Minimum dissolved oxygen requirements of aquatic
animals in the Gulf of Mexico estuaries as a measure of the effect of nutrient enrichment
(GED).
APM ID FY 02 (GPRA # 166) Effects of nutrient loadings and altered nutrient ratios on
HABs (GED).
APG 2 FY03 Provide the science to support consistent dissolved oxygen criteria for prevention
of hypoxia impacts in all coastal regions of the US.
APM 2 A FY03 Minimum DO requirements for a suite of the important marine organisms
(fish and crustaceans) from the Atlantic, Pacific, and Gulf of Mexico coastal waters of the
U.S. (AED).
42
-------
APG 3 FY03 (GPRA #15) Complete the framework for including dissolved oxygen and other
receiving water thresholds into watershed management for nutrients.
APM 3A (GPRA # 201) Comparison of effects of zooplankton grazing on estuarine
phytoplankton community under differing natural levels of grazing (GED).
APG 4 FY04 Provide first generation protocol to classify eutrophication models for nutrient load
allocation in coastal systems.
APM 4A FY03 Propose classification scheme for predicting sensitivity of coastal
receiving waters to effects of nutrients on DO (MED, AED, GED).
APG 5 FY07 Provide the scientific foundation for establishing site-specific nutrient threshold
criteria to protect estuarine SAV.
APM 5A FY02 Report on structural and functional characteristics of SAV rhizospheric
communities (GED).
APM 5B FY03 Correlation of water quality with SAV change (GED).
APM 5C FY04 Report on environmental requirements of three main species of seagrasses
(WED).
APM 5D FY04 Development of stress-response model for Zostera marina in Pacific
Northwest and validation of stress-response model for Thalassia testudinum (WED).
APM 5E FY04 Development of empirical load-response models for Zostera marina in
NE U.S. (AED).
APM 5F FY05 Development of load-response models for estuaries of Pacific Northwest
and Gulf Coast, and validation of stress-response model for Zostera marina in NE U.S.
(WED, GED, AED).
APM 5GFY05 Propose classification scheme for predicting sensitivity of coastal
receiving waters to the effects of nutrients on SAV (WED, GED, AED).
APM 5H FY06 Report on the empirical and numeric models for SAV (WED, GED,
AED).
APM 51FY07 Report on a classification scheme for grouping coastal receiving waters
based on sensitivity to nutrients (WED, GED, AED).
APG 6 F Y07 Provide scientific foundation for development and application of quantitative
measures of food web attributes that are sensitive to ecological changes associated with nutrient
enrichment.
43
-------
APM 6A FY04 Sensitivity of food web responses to nutrient loading in coastal systems
(GED).
APM 6B F Y05 Propose classification scheme for predicting sensitivity of coastal
receiving waters to effects of nutrients on food web structure (GED).
APM 6C FY06 Report on empirical and numeric models for food webs (WED).
APM 6D FY 06 Report on parameterization of food web models (GED).
APM 6E FY06 Report on classification scheme for grouping coastal or lake receiving
waters based on sensitivity to food web alterations (WED, GED, AED, MED).
Critical Path
The components of the critical path seen in Figure 5 consist of five main steps that are the same
or each of the response endpoints (DO, SAV, Food Webs):
Step 1. Mine and Assess Existing Information.
Evaluate available data and models from the peer-reviewed literature and determine if the data
and models are useful to aid the development and improvement of nutrient load-response models.
This will be a continuous process since new data and models are continuously being developed
by other Federal and academic institutions.
Step 2. Develop Conceptual Models.
Conceptual models, describing how the three assessment endpoints respond to excess nutrients,
will be defined in order to catalog the controlling mechanisms and processes.
Step 3. Develop Classification Scheme.
Included within this step is the development of a classification scheme and the assembly of data
needed to apply a classification scheme across the Atlantic, Gulf, Pacific, and Great Lakes
receiving waters. A tabulation of some of the factors that likely will be used in classifying
receiving waters or in scaling or standardizing loading or response variables can be found in
Tables.
44
-------
Database of existing information
Conceptual models for nutrient load-endpoint responses
Classification scheme for receiving water bo dies/en dp oints
Standard or comparable methods for assessment en dp oints
Nutrient load - water body response models
Figure 5. Critical path for research on the development of nutrient response
relationships for coastal receiving waters.
Step 4. Develop Standard Methods and Procedures.
This task will involve the determination of a set of standard measurement endpoints for the
assessment endpoints. Comparability of measurements and data across Regions/Ecological
Divisions, where possible, will maximize utility of data collection, research, and modeling
efforts. A periodic review of these endpoints and procedures across the Ecological Divisions will
ensure consistency and comparability and maintain focus on our core objectives.
Step 5. Develop Nutrient Loading-Ecological Response Models.
Nutrient Inputs/Loading
In order to develop nutrient loading-response relationships for each of the three endpoints, it is
necessary to estimate nutrient loadings. There are at least three ways that we may estimate
loadings: 1) estimate using watershed models augmented with point source, atmospheric, and
oceanic inputs [i.e., USGS Spatial Referenced Regressions on Watersheds (SPARROW) model,
Smith et al., 1997]; 2) use field-based water column concentration estimates during the
biologically inactive portion of the annual cycle and then back calculate to loadings; and 3) use
45
-------
field-based input studies (e.g. Cohn et al., 1989). SPARROW model based loading estimates are
available from USGS for the larger watersheds throughout the nation and we will test those
estimates in our response models and against field measurements when possible. However, for
smaller systems, SPARROW estimates may not be available and different watershed models or
alternative methods may be required. Where possible, NHEERL will also seek loading data from
NERL to test other loading models and loading-relationships for coastal water bodies;
Table 3. Preliminary list of factors influencing response to excess nutrient inputs in coastal
receiving waters.
Biological Factors
Physical Factors
Chemical Factors
SAV
Food web efficiency
Phytoplankton community
Primary productivity base
(e.g., phytoplankton base vs.
sea grass based)
Grazing type (e.g., benthic
filter feeders, zooplankton)
and grazing intensity
Flushing
Light/Suspended
solids/Water color
Stratification
Depth
Temperature
Volume
Area
Tidal Height
Geomorphology (e.g.,
drowned river valley)
Physical energy (wind, etc.)
Hypsography (area-depth
relationship)
N:P: Si :Fe Ratios
Salinity
Allochthonous C
Denitrification potential
Nutrient form
(organic/inorganic)
however, if this is not possible, we may need to estimate loading in smaller systems by direct
measurement (flow versus concentration).
In addition, since water quality management is frequently based on returning to some historical
loading or reference condition, to be most useful to OW, the load-response relationships should
be based on historical loading to the maximum extent possible.
Response Models
Nutrient loading-response models are the ultimate products of the new research initiative. They
will be produced using two parallel, yet integrated, efforts. Numerical models will be used to
provide refinement of empirically derived loading-response models and to aid in understanding
mechanisms. If our classification schemes are appropriate, they will identify groups of receiving
water that have significantly less variability in nutrient load-endpoint response relationships than
is present among all receiving waters. Improvement in these relationships will provide the test
for our classification schemes (Step 3) and will provide the scientific basis for grouping receiving
waters to simplify the nutrient criteria/TMDL process.
46
-------
The classification schemes and modeling efforts will be endpoint-specific. For each endpoint in
each of the regions and across all coastal systems, we will begin with a data mining effort and
identify and collect data on key parameters and existing models and classification schemes that
will provide the parts necessary to build a sound scientific basis for nutrient criteria for each of
the three endpoints (DO, SAV loss, and food webs). With empirical data and knowledge of the
key parameters that influence the nutrient-response curves, we can identify correlations and
develop models to test hypotheses experimentally in the laboratory and in the field. This will be
an iterative process to some extent until we have adequately characterized the endpoint/nutrient-
response and controlling factor relationships needed to establish numerical nutrient criteria across
the nation's coastal receiving waters. Once this is accomplished we can move on to watershed
scales and/or other water body types.
Complexity in ecosystem models is always dealt with by simplification; all models are
abstractions of real systems. The simplifications range from complete linearization of the system
with regression models to more complex relational models that incorporate a few critical
measurements of the biological, chemical, and physical domain that makes up an estuary. Food
web models make up the more complex non-linear models that include most of the mechanistic
relations and feedbacks within the estuarine biogeochemical system. Relative to the coupled
atmospheric-watershed-estuarine models developed for Chesapeake Bay and Long Island Sound,
the food web models are easy to develop and have much fewer data requirements.
Regression Models
Regression models may be used to develop relationships among variables within a single system
or among multiple systems. Such models use nutrient load to predict other parameters such as
phytoplankton biomass accumulation, primary production, sedimentation, and community
metabolism, SAV loss, and DO-related response. This research will consist of cross-estuary
analysis that focuses on common responses within classes of estuaries. Regression analysis that
compares data among multiple systems has been used successfully for Maryland estuarine
systems (Boynton et al. 1996); and a diverse collection of estuarine, continental-shelf, and open-
ocean systems (Nixon et al. 1996). In the case of 37 side-embayments of Buzzards Bay,
Massachusetts, such regression analysis has been applied to development of WQC and TMDLs
for nitrogen (Costa et al. 1999). Such regressions are expected to have general application to
systems similar to those for which they have been developed. These regressions will quantify
estuary response to nutrient loading, and thus be directly useful in risk assessment and setting of
nutrient criteria. In addition, regression models can be used as an adjunct to some of the
proposed mesocosm and field work. For example, data from mesocosms to determine the effect
of nutrient loading and benthic oxygen consumption on denitrification and nitrous-oxide
production can provide insight into why some estuaries with nutrient sources having a high
nitrogen/phosphorus (N/P) ratio remain nitrogen limited. Simple regression models also can be
used in conservative mixing curves to determine sources and sinks of nutrients over the length of
an estuary. Nitrogen, P, silicate, or other contaminants, when plotted against salinity, provide
estimates of deposition, utilization, and supply of these materials over the length of an estuary.
Vollenweider (1975) pioneered the development of regression models that allow extrapolation of
data among systems using a scaling of biological processes to hydraulic residence time. This
47
-------
method has been modified (Dettmann 2002) to model the fate and concentrations of total
nitrogen in estuaries. Preliminary work in extending model applicability also has shown that
total N concentrations predicted by the model appear to correlate well with peak annual
phytoplankton concentrations and peak macroalgal abundance in estuaries. The overall goal of
this work has been to explain the response of estuaries to nitrogen loading using as few
parameters as possible. The model appears to reasonably describe annual net N, the dependence
of annual denitrification on water residence time, and the annual average concentrations of total
N in estuaries where it has been tested. The results emphasize the importance of water residence
time in determining export, denitrification, and concentrations, and give quantitative expressions
for these dependencies. At present, the model provides annualized results averaged over the
entire estuary, although recent results indicate that it may also have application to seasonal
response as well. The final extended model is expected to have direct applications to the
evaluation of estuary sensitivity to nutrient loading, which will be useful in setting nutrient
criteria. The model may also serve as part of the foundation of a classification system for
estuarine sensitivity to nutrient loading.
Food Web Models
Models of food webs link the food web components to the overall ecosystem through an explicit
quantification of exchanges. This makes it possible to evaluate how changes in the model
components directly effect ecosystem processes. For example, the cycling of carbon and
nutrients directly result from food web interactions in which many species play a role. Within
many ecosystems, species that contribute little biomass still may have a large influence on
nutrient cycling and energy flow, and thus affect the functioning of other species. Examples of
this include bacterial grazers, which can stimulate microbial activity through nutrient recycling,
and algal grazers which stimulate the productivity of submerged macrophytes by providing better
light conditions through the grazing of periphytic algae. Extinction or changes in abundance of
such species can have a disproportionately large influence on ecosystem function. Hence, food
web approaches can be used for analyzing the effects of nutrient stressors on key target species,
on the biological diversity in communities, and on the functioning of ecosystems. In this way,
food webs are the wiring on the circuit board of the ecosystem, spanning different levels of
ecological organization. Food web models allow managers to identify the critical food web
flows within the estuarine ecosystem, for which small changes in an ecosystem component will
cascade through the system and result in large changes in eutrophication, extinction of important
habitats, or changes in the tropic structure of the overall ecosystem (Vezina and Pace 1994).
Food web models now being developed and evaluated in ORD are used to calculate metrics that
define the state of an ecosystem (Ulanowicz 1986, Hagy 2002). These metrics are similar to
diversity and commonness, but are much more sensitive to ecosystem condition than the older
metrics. These new metrics can be calculated directly from the food web models now being
developed at ORD. In the following two examples, we describe how these indices can be used
to develop useful relationships between nutrient loading and trophodynamics:
1. Indices of ecosystem trophic efficiency can be used to quantify how carbon and nutrients
supplied to the estuary are passed thought the food web to the higher tropic levels. If more
nutrients and carbon are moved into higher tropic levels, then the nutrient capacity of the estuary
48
-------
may be increased. Laguna Madre during the 1990s provides an example of an ecosystem with
small nutrient loading that developed symptoms of eutrophication. An inedible phytoplankton
species Aureoumbra lagunensis developed into a brown-tide bloom because most consumers
could not ingest it. Those that could were being controlled by predators. Because of the
mechanism by which this bloom developed, it could not be predicted easily by regression.
However, susceptibility to such blooms could be predicted by food web models.
2. Food web models can be used to calculate the dependency of charismatic and recreational
species on other components of the food web, and how such dependency would be altered by
changes in nutrient loading. This knowledge would arm managers with a early warning system
to detect alterations in an ecosystem that eventually could lead to reductions or decreased
production of important species.
Each of these modeling strategies has different data requirements and makes predictions at
different temporal and spatial scales. We will integrate the three approaches by working
cooperatively to maximize the benefits of the approaches and minimize the limitations. The final
outcome of this research implementation plan will be a set of empirical or numerical models for
classes of coastal waters. These models will be able to accurately describe how increases in
nutrient loading causes changes in hypoxia/anoxia in coastal receiving waters, losses of SAV,
and changes in algal community composition leading to shifts in basic food webs. The models
will provide the scientific basis for the development of nutrient criteria for coastal receiving
waters. We should be able to extrapolate results between similar classes of receiving waters
once a classification scheme incorporating key factors affecting nutrient response for coastal
receiving waters is developed and tested.
NHEERL Ecology Divisions are strategically located in four major coastal systems with AED
along the Atlantic Coast, GED on the Gulf Of Mexico, WED along the Pacific Coast, and MED
on the Great Lakes. Each of these Divisions will focus intensive studies on a local system and
collect pertinent related information from other systems in their general region. All four
Divisions will work to develop a common approach for key parameters and measurements that
are needed across all regions (i.e., what, when, where, and how for nutrient loading, DO,
Chlorophyll a) and maintain that approach through annual reviews. Other measurements (e.g.,
community metrics or depth of oxygen penetration into sediment) may be developed within or
compared across multiple regions as individual Divisional research plans are developed further.
Research Projects
Project Title 1. Development of nutrient load-DO Response Relationships for Coastal
Receiving Waters
Project Coordination and Resources (9.0 FTEs: AED-4.0, GED-3.0, MED-1.0, WED-1.0)
49
-------
Objectives
To define nutrient load-DO response relationships for coastal receiving waters that will be used
by the States and authorized Tribes to aid in the development of nutrient-related WQC and
TMDLs.
Scientific Approach
Low DO in coastal receiving waters is a symptom of eutrophication. Therefore, controlling the
effects of low DO is not accomplished by directly regulating DO but by regulating nutrients and
oxygen demanding wastes.
The ultimate product of this research effort is to provide the scientific basis to develop dissolved
oxygen based nutrient criteria. The critical path to this product is:
Step 1: Mine and assess existing information on DO response to excess nutrients for coastal
receiving waters and minimum oxygen requirements of commercially and ecologically
important organisms.
Step 2: Develop conceptual model of how different systems manifest low DO in response to
excess nutrients.
Step 3: Propose a classification scheme for coastal receiving waters that groups these waters
according to their sensitivity to DO depletion in response to excess nutrients.
Step 4: Develop a common approach across Divisions. Select methods, parameters, and
measurement endpoints for low DO response to excess nutrients so that data and
models are interchangeable across Divisions and regions.
Step 5: Test the proposed classification scheme to provide the scientific basis for development
of nutrient criteria (or TMDLs) based on nutrient load-DO response relationships for
different classes of receiving waters.
Excessive nutrient loading to an estuary/receiving water stimulates primary production (i.e.,
phytoplanktonic, macroalgal). This production, together with allochthonous (labile) organic
carbon, sinks to the bottom waters/benthos resulting in respiratory oxygen demand (mostly
through microbial decomposition) that exceeds oxygen supply. This leads to hypoxia or anoxia
in the bottom waters, which in turn lead to fish avoidance, fish kills, and mortality of sessile
organisms such as worms and shellfish. Low oxygen also changes the oxidative properties of the
sediments such that organic matter accumulates rather than being oxidized. Low oxygen can
decouple nitrification and denitrification processes in the sediments resulting in increased supply
of nutrients to overlying waters. In addition, the death or debilitation of shellfish and other
sensitive sessile organisms can cause changes in the structure and function of benthic habitats.
For example, after hypoxic conditions dissipate, predators return to feed on vulnerable benthic
organisms before they can rebound from the stressed conditions; this changes the energy balance
of the system.
50
-------
Hypoxic/anoxic conditions typically manifest themselves on diel or seasonal time scales during
critical times of the annual cycle. For southern/semi tropical regions, the critical period is
typically from May to October; for northern/temperate regions the period is from June to
September. Seasonal hypoxia generally develops as a consequence of water column
stratification, whereas diel cycles generally occur in non-stratified or partially stratified systems.
Seasonal hypoxia is persistent throughout the critical period whereas diel hypoxia may be cyclic
(regular frequency and duration; e.g., diel, tidal) or episodic (irregular frequency and duration).
DO Criteria
Part of the process of setting nutrient criteria based on DO involves determination of the
minimum DO requirements of aquatic organisms. NHEERL is in the process of providing the
scientific basis for setting minimum DO criteria for ecologically and commercially important
organisms in coastal receiving waters through laboratory exposures. Survival data from
laboratory exposures, using controlled DO concentrations, will provide risk assessment managers
with the basic information needed to set minimum DO protection limits for the Nation's waters.
General Classification Variables
The classification scheme for the DO endpoint will be a common effort across all Divisions. Our
plan is to link and improve existing models of flushing, light limitation, primary production
controls, and oxygen supply dynamics to sort coastal receiving waters into groups of similar
overall relative sensitivity. The number of groups will depend on the range and variability of
estimates. Chapter 6 of Clean Coastal Waters (NRC 2000) lists 12 factors that influence the
susceptibility of coastal receiving waters to nutrient over-enrichment. We will start with this list
and focus on those factors that will be most useful in a classification scheme to determine the
relationship between nutrient loading and DO as an effect of nutrient over-enrichment. The
relative magnitude of atmospheric oxygen entrainment to the bottom waters and respiratory
depletion determines whether a portion of the water column experiences episodes of low DO.
The most important factors that affect entrainment include: density, salinity, and/or temperature
stratification. In turn, these factors are influenced by climate/weather (temperature, wind),
geomorphology (tides), and circulation patterns (tides and freshwater input) of the receiving
waters.
Once the classification scheme with the key parameters influencing the response of DO to
nutrients has been developed, we will test the classification using field data and comparable data
from literature or other institutions and agencies where available. Both empirical and numerical
simulation models will be used to improve our understanding of how these parameters interact
and control the DO response to increased nutrient supply across a wide range of coastal receiving
waters.
Measurement Endpoints
As mentioned above, "low DO" is one of the assessment endpoints that is to be related to excess
nutrients in receiving waters. There are many methods and approaches to determine DO in
coastal receiving waters. Providing sound basis for a national nutrient criteria based on DO will
51
-------
require development of a common approach to be applied and tested across a wide range of
coastal receiving water systems. Each NHEERL Ecology Division will need to contribute to this
process, and to the extent possible identify other sources of similar data from receiving waters in
their region that could be used to improve our understanding of the relationship between nutrient
loading-DO and the key factors that control that relationship.
Products
APM 1C FY02 (GPRA # 165) Minimum dissolved oxygen requirements of aquatic animals in
the Gulf of Mexico estuaries as a measure of the effect of nutrient enrichment (GED).
APM 2 A FY03 Minimum DO requirements for a suite of the important marine organisms (fish
and crustaceans) from the Atlantic, Pacific, and Gulf of Mexico coastal waters of the U.S.
(AED).
APM 4A FY03 Propose classification scheme for predicting sensitivity of coastal receiving
waters to effects of nutrients on DO (MED, AED, GED).
Benefits of Products
The benefits of the products to OW will be a reduction in the uncertainty associated with setting
DO based nutrient criteria and TMDLs for our nation's receiving waters. Minimum DO
requirements of important species will provide a sound basis for setting protective limits for DO
in coastal receiving waters. Development of an improved classification scheme will aid in
setting nutrient criteria in receiving waters where large historical databases are not available. An
improved understanding of the factors affecting nutrient DO-response relationships will provide
water quality managers with better tools to manage nutrient input to our nation's waters.
Project Title 2. Development ofSA VLoss-Nutrient Load Relationships and Factors which
Control SA VResponse to Nutrients
Project Coordination and Resources (9.0 FTEs: AED-1.0, GED-4.0, MED-2.0, WED-2.0)
Objectives
The objective of this research plan is to develop, for the nations's coastal receiving waters,
sufficient understanding of the relationship between SAV loss and nutrient loading (N and P) to
provide a sound scientific basis for establishment of nutrient criteria that would protect these
important habitats from degradation or loss and aid in restoration efforts. To do this, a set of
models will be used to examine how nutrients interact with the physical and biological
components to affect the health of SAV populations. This work will directly support or interact
with the habitat alteration research (Section 4) providing basic information on production and
function of these habitats and diagnostic research (Section 8) by providing nutrient-SAV-
response relationships.
52
-------
Scientific Approach
Research on SAV survival and production has generally focused on a few "key" parameters (e.g.,
light, nutrients), and much is known about the basic light requirements of SAV. Nutrients
primarily affect SAV through their effects on water quality and the associated effects on light
availability caused by increasing algal biomass (pelagic and epiphytic). Light availability is
generally considered to be the major factor related to SAV survival (e.g., Kenworthy and Haunert
1991, Tomasko and Lapointe 1991, Fourqurean and Zieman 1991, Dennison et al. 1993,
Stevenson et al. 1993, Dunton 1994) and depth of distribution (Dennison 1987, Dawes and
Tomasko 1988, Duarte 1991). However, recent work is beginning to suggest that a more holistic
approach may help delineate additional factors and improve our understanding of the relationship
between nutrients and SAV loss (Koch 2001, Kaldy et al. 2002). In addition to understanding the
requirements of SAV, we must also understand the interactions of nutrients and the physical,
chemical, and biological factors that control accumulation of phytoplankton biomass in our
receiving waters. This is similar to research needed to understand the relationship between
nutrients and hypoxia (project 1); however, there are some differences in the factors associated
with accumulation of biomass versus the increase in production associated with hypoxia.
The approach will be consistent with the general critical path stated above. Essentially, the main
components for this research are:
1. Data gathering, literature review, and compilation of scientific literature (including existing
information on SAV models for coastal systems) will focus on SAV-nutrient relationships.
Where existing nutrient-SAV community data are available, this will include the compilation and
statistical analysis of data. Over the last 30 years, a large body of literature has been published
with regard to seagrass and other SAV. These publications should provide strong guidelines for
directing EPA research. As part of this step, we plan to develop a report on the basic
requirements (e.g., light, salinity, sediment characteristics) of three rooted aquatic seagrasses
(Thalassia, Halodule, and Zostera) which together represent a large fraction of the SAV found in
estuarine/marine systems along the coast of the U.S. Our research on SAV loss-nutrient load
relationships will not be limited to these three seagrasses and will include freshwater and
brackish rooted SAV as well.
2. Develop conceptual models of how SAV communities and nutrients interact with various
environmental parameters to result in decreased survival or production in this important habitat.
These models may be regionally specific because of the inherent differences; however, we will
seek to make them as broadly applicable as possible. Our current conceptual model of how
nutrients are involved in SAV loss suggests that increased nutrients leading to increased algal
biomass accumulation is the basis for this pathway. The most likely factors affecting algal
biomass accumulation are flushing rates, grazing rates, turbidity, climatic conditions, and water
depth. We realize that excess nutrients also can cause nitrate toxicity, and increase sediment
sulfide concentrations and/or toxicity.
3. Propose a classification scheme for coastal receiving waters that groups these waters
according to their sensitivity to SAV loss in response to excess nutrients. Evaluate and modify
existing models in order to develop nutrient-SAV loss responses. Classification parameters will
53
-------
certainly include dominant SAV species as well as many of the parameters listed in Clean
Coastal Waters (NRC 2000). Development of a classification scheme for the SAV endpoint will
be a common effort across all Divisions. Our plan is to use knowledge of species-specific
requirements and link existing models or develop models that include the effects of light,
nutrients, and sediment geochemistry on seagrass physiology. We also will link these with
models of water column chlorophyll a-light absorption/attenuation and nutrient-phytoplankton
biomass relationships to provide a basis for setting nutrient criteria for coastal receiving waters
where protection or restoration of SAV is needed.
4. Develop a common approach across Divisions. Select methods, parameters and measurement
endpoints for SAV response to excess nutrients so that data and models are interchangeable
across Divisions and regions. Each Division will provide measures of the key parameters (the
final list of these parameters will be developed and standardized); however, Division-specific
research plans may focus on specific parameters or measurements as part of individual research
projects (e.g., epi-periphyton community metrics, sediment sulfide production/toxicity). This
phase of the plan development will include interaction with seagrass and SAV specialists in other
governmental and academic institutions where possible.
5. The final step of this SAV critical path is a verification that tests the model's prediction and
our classification scheme against the actual endpoint response. Once tested and verified, the
proposed classification scheme and models will provide the scientific basis for development of
nutrient criteria or TMDLs based on nutrient load-SAV response for different classes of
receiving waters.
Modeling Plan
Our modeling approach for the seagrass endpoint is embodied in a three-tiered scheme that
couples model development, field monitoring, and direct experiments to test specific hypotheses
(Figure 6). We plan to use both holistic and mechanistic approaches that integrate the scientific
literature and conceptualize how seagrasses are affected by stressors by using numerical and
empirical models that quantify the production and distribution of seagrass. Characterization of a
suite of parameters (e.g., light, sediment biogeochemistry, nutrients, and exposure during part of
the tidal cycle) and their inclusion in models and classification schemes will allow EPA to
determine which stressors are most important controlling factors in a particular region. Stressors
which will impact SAV vary between regions as a result of variations in climate, industry,
agriculture, and other land use practices. Consequently, it is critical to include a wide variety of
parameters that influence all SAV (seagrass and freshwater aquatics), regardless of region. The
appropriate scientific approach will likewise be an approach combining model development,
field monitoring, and direct experimentation to test specific hypotheses generated by the models.
Built into this SAV modeling plan is a set decision point used to determine if the conceptual
model, the numerical or empirical model, or the field test of the model has been successful. This
54
-------
SAV
Model Development
Data Mining
N Load or proxy, Historical SAV coverage,
Avg Chi a as proxy for phytoplankton,
existing mo dels (I v Chl,Kd vZ by species,
N v SAV Loss by system)
Define Conceptual Model
Develop Numerical Model
To Test Conceptual Model
Develop Empirical Model
To Test Conceptual Model
Ex ample of parameters
required for an existing
biomass model
Total PARtims series, CM i
TSS, Color, Epectralmodel
Kd Historical and current
c overage, Above fe low
ground bioirnsSjP Q/P^^P
vs. I Curve , Nutrimtuptake
(CVN), Temperature, Leaf
loss rat* , Pore water N,
assimilation te dmiqus
Does Mo del Make Sense?
I
Yes
No
Field TestModel
Does Field Test Validate Model?
Nutrient Criteria For SAV
Figure 6. Conceptual diagram of the feedbacks among data mining, model development, field
monitoring, and experimental hypothesis validation.
iterative process assures that at each stage in the critical research path, error in data and models
are sufficiently small that the completed analysis will be accurate enough to make meaningful
prediction of the SAV response to nutrients.
Model Development
Seagrass models are generally a composite of numerical and empirical relationships that provide
a quantitative prediction of seagrass growth or loss. Each of these components has to be tested
individually and in concert with other relationships that make up the model. Although all SAV
models will have components in common, each regional model will be individualized to
incorporate locally important species, the biogeochemistry of the water-column and sediments,
55
-------
and the local physical regime (Table 3). Successful model development will require long-term
continuous data sets using instruments to measure important plant parameters (e.g., spectral
irradiance, temperature, and salinity). It is critical that monitoring be conducted at appropriate
temporal and spatial intervals that are relevant to the organism(s) and system(s) in question.
Development and implementation of any long-term monitoring plan for use in modeling
activities should be a cooperative effort involving the appropriate personnel (e.g., modelers,
biologists, and field technicians).
Field Monitoring
The purpose of the field monitoring portion of this plan is to collect data on the range of
responses and variability that are present in coastal systems and to provide input data to generate
or refine models. Aquatic environments in general, and particularly estuaries, are stochastic
systems that often exhibit large variations along many temporal and spatial scales. Long-term
data sets are required to determine if variability expressed in a system is a consequence of natural
variability (e.g., storm events) or anthropogenic impacts.
Direct Experimentation (Field and/or Mesocosm Studies)
One of the most important features of a model is the ability to develop testable hypotheses that
will provide confidence in the model. In most cases, field experiments would be the preferred
experimental environment; however, it is often difficult if not impossible to control all of the
variables in the field (e.g., water column nutrient concentrations). The ability to replicate
treatments also permits statistical data analysis. Consequently, mesocosm experiments bring
together the best features of field and laboratory experiments offering environmental parameter
control and a natural environment, while facilitating quantitative data collection (for review see
Lalli 1990).
Development of the proposed models would provide testable hypotheses about the influence of
any number of factors, such as the toxicity thresholds of SAV to various water column and
sediment constituents (e.g., nitrate, ammonium, or sulfide concentrations). Another example
would be determination of dessication stress, or sediment and water column anoxia on
photosynthesis, or the interactions controlling the relationships between nutrients and
phytoplankton versus epiphytes versus macroalgae versus seagrass. Physical factors such as the
impact of wave exposure could also be investigated using replicated mesocosm experiments.
Current Activities
AED is developing empirical relationships between nitrogen loading and the areal extent of SAV
normalized to historical SAV habitat extent. MED is investigating the relationship between
nutrient loading and several quantitative attributes of wetland SAV including: % cover, diversity,
relative abundance, and maximum depth of macrophyte growth. GED is in the fourth year of
developing a database of changes on water quality, light availability, and changes in the
deepwater margin of SAV beds. WED is developing models of seagrass growth and production
based on field data including seagrass biomass and production, underwater light, and sediment
biogeochemistry. Other activities that support this SAV plan include GIS mapping of seagrass
56
-------
distribution over multiple annual cycles and the development of spatially explicit models to
examine seagrass bed dynamics (e.g., expansion). Skills and knowledge represented by these
research projects will be integrated across Divisions and will provide the basis for comparison
and testing of approaches outlined above across the nation's SAV communities. Again, to be
most useful to OW, SAV coverage and the loading response relationships should be based on
historical information to the maximum extent possible since water quality management is
frequently based on returning to some historical reference condition (realizing that such data is
often not available).
Products
APM 1AFY02 (GPRA# 163) Generalized seagrass/rhizosphere model capable of predicting
effects of reduced light, sedimentation, nutrient depletion, and toxic effects of sulfides (WED).
APM IB FY02 (GPRA # 164) Effects on estuarine submerged aquatic vegetation from changes
in light quantity and quality due to increased levels of suspended solids (GED).
APM 5A FY02 Report on structural and functional characteristics of SAV rhizospheric
communities (GED).
APM 5B FY03 Correlation of water quality with SAV change (GED).
APM 5C FY04 Report on environmental requirements of three main species of seagrasses
(WED).
APM 5D FY04 Development of stress-response model for Zostera marina in Pacific Northwest
and validation of stress-response model for Thalassia testudinum (WED).
APM 5E FY04 Development of empirical load-response models for Zostera marina in NE U.S.
(AED).
APM 5F FY05 Development of load-response models for estuaries of Pacific Northwest and
Gulf Coast, and validation of stress-response model for Zostera marina in NE U. S. (WED,
GED, AED).
APM 5G FY05 Propose classification scheme for predicting sensitivity of coastal receiving
waters to the effects of nutrients on SAV (WED, GED, AED).
APM 5H FY06 Report on the empirical and numeric models for SAV (WED, GED, AED).
APM 51FY07 Report on a classification scheme for grouping coastal receiving waters based on
sensitivity to nutrients.(WED, GED, AED).
57
-------
Benefits of Products
The benefits of the products will be a reduction in the uncertainty associated with setting SAV
based nutrient criteria and TMDLs for our nation's receiving waters. A compendium of
requirements of important SAV species will provide a convenient reference and a sound basis for
setting protective limits in coastal receiving waters. Development of a classification scheme will
aid in setting nutrient criteria in receiving waters where large historical databases are not
available. An improved understanding of the factors affecting nutrient-SAV loss relationships
will provide water quality managers with better tools to manage nutrient input to our nations
waters while protecting these important habitats.
Project Title 3. Food Web and Community Composition Changes in Response to Nutrient
Loading in Freshwater and Marine Coastal Systems (Estuaries and Coastal Wetlands)
Project Coordination and Resources (14 FTEs: AED-3.0, GED-5.0, MED-3.0, WED-3.0)
Objectives
The primary objective of food web research is to identify nutrient loading thresholds that cause
shifts in primary producers and other key components of the food web. A secondary objective is
to assess the use of food web structure and processes to improve our ability to classify systems
and to predict changes in response to nutrients that affect hypoxia and SAV. Research will
require identification of measurement endpoints that are sensitive to nutrient loading and reliably
forecast adverse effects to assessment endpoints.
Scientific Approach
As with the DO and SAV assessment endpoints for nutrient research, we are concerned that
increased concentrations or changes in ratios or timing of nutrient inputs can adversely affect
populations of ecologically and commercially important organisms. For food webs, nutrient
loading-response relationships are not as well understood as for DO and SAV; however, food
webs may reveal subtle, low threshold responses to nutrient loading that are more sensitive than
DO or SAV endpoints (Livingston 2000). Changes in patterns of energy flow alter habitats and
support systems required by important organisms. Variations in nutrient ratios, concentrations,
or timing of inputs can alter the competitive advantages of primary producers causing the demise
of species on which important consumers depend. Research investigating the relationships
between nutrient loading and food webs will focus on identifying thresholds of nutrient loading
where pathways of energy flow from primary producers to consumers are altered and populations
of commercially and ecologically organisms are adversely affected. NHEERL research will
focus on the effects of nutrients on primary producers and subsequent interactions with pelagic
and benthic communities. Process-oriented research, such as analysis of food web structure via
stable isotopes will provide basic data for development of nutrient load-response relationships,
and will provide insight into biological factors controlling primary production and energy flow in
coastal systems. This research will focus on three critical food web shifts: 1) changes from
benthic to pelagic basis of production (i.e., SAV or periphyton to phytoplankton), 2) shifts in
58
-------
pelagic algal communities from desirable (edible, nutritious) to undesirable (non-edible,
nuisance, HAB species), and 3) shifts in emergent or marsh grass systems.
Major challenges of this research are to identify practical measurement endpoints (response
variables) and relate those responses to assessment endpoints (e.g., high performance liquid
chromatograhy [HPLC] accessory pigments may be useful as measure of changes in the
phytoplankton community leading to or associated with decreased fisheries production). We will
investigate development of indices of tropic status associated with high nutrient conditions, such
as ratios of algal biomass to zooplankton and/or fish, benthic and pelagic community metrics that
reflect algal composition (e.g., percent blue-greens, and relative abundance of centric and
pennate diatoms, algal size distributions). Developing load-response relationships, which reflect
changes in food web structure, will be challenging due to the complex feedback mechanisms
which accompany degraded habitats in coastal systems (Jude and Pampas 1992, Chow-Frazer
1998). Increasing phytoplankton biomass may uncouple primary production from grazing. This
then leads to algal blooms, possibly toxic or noxious algae (HABs) and successions in pelagic
and benthic communities. The causal mechanisms for HABs remain poorly understood; some
have always occurred and are entirely natural. However, other blooms are tied to nutrient
enrichment, thus leading to more frequent and longer lasting blooms as nutrient loading increases
(NRC 2000). Even more uncertainty exists regarding relationships between nutrient loading and
basic changes in food webs supporting productive marine and freshwater ecosystems. In addition
to nutrients, the activity of top consumers can exert strong controls on zooplankton and/or
phytoplankton affecting phytoplankton size, abundance, and production. Biogeochemical
processes resulting from blooms (i.e., enhanced sedimentation and redox changes) can cause
changes in species diversity, size spectrum of organisms, and average tropic level of the
community. The secondary effects of these water-column and sediment changes may be
persistent changes in pelagic and benthic species assemblages and alterations in the nutrient
recycling potential of aquatic habitats. Our approach is to determine nutrient load-response
thresholds for endpoints reflecting shifts in the food web structure or species composition.
The Scientific approach will be consistent with the general critical path (Figure 5). Essentially,
the main components for this research are:
1. Data gathering and literature review of food web information for coastal food web systems
will be done collaboratively across Divisions, where existing nutrient-food web/community data
are available, statistical analysis will be conducted.
2. Conceptual models that include both bottom up effects models and bottom up-top down
community models will be developed. The bottom up model (nutrients to primary production,
Menge 2000) shows influences on aquatic communities, including the formation of some HABs
and ultimately changes in important fish and shellfish populations. The community level model
includes interactions among populations that form the food web and identify critical interactions
that may lead to changes in species diversity, commercial harvest, eutrophication, and designated
use.
3. A classification scheme for coastal receiving waters that groups these waters according to
their sensitivity to food web changes in response to excess nutrients will be provided. Existing
59
-------
food web models will be evaluated and modified in order to develop nutrient-food web responses
incorporating classification schemes.
4. We will develop common approaches across Divisions for parameters related to food web
shifts (e.g., HPLC pigment analysis, stable isotope measurements, community metrics). We will
select common methods, parameters, and measurement endpoints for food web response to
excess nutrients when available, so that data and models are interchangeable across Divisions
and regions, which will provide the basis for comparison and testing of our models, methods, and
classification schemes.
5. Testing of the proposed classification scheme and models will provide the scientific basis for
development of nutrient criteria and/or TMDLs based on nutrient load-food web response for
different classes of receiving waters.
Initial research projects adopted by all four Ecology Divisions, to collect data for identifying
potential assessment endpoints or testing classification schemes and models, will help in the
development of a better understanding of the processes leading to shifts in food webs and help
identify factors and parameters in common with the process. Comparison of results and
application of successful techniques across Divisions will speed identification of practical
assessment endpoints and development of models and understanding of food web-nutrient
relationships. We plan to use stable isotope measurements in systems with various types of
dominant primary producers at the base of the food web (i.e., pelagic phytoplankton, saltmarsh,
benthic algae, and SAV). Indices that we plan to predict include ascendancy and/or flow
diversity. This indice predicts the relative stability of food webs. System with low flow diversity
tend to have large temporal variations in structure, usually at the lower tropic levels, but
sometimes at the mid and high tropic levels. In the long-term, these systems eventually come to
a new stable state that may be more eutrophic than the original food web. Hence, we can predict
what systems are at risk (unstable) even though no model can reliably predict species succession.
Nutrients as a stressor can reduce food web flow diversity and the ensuing systems instability
may result in HABs, macroalgae, and other harmful species. With sufficient data in the various
system types, we can use optimization techniques to correlate changes in flow diversity (stability)
with nutrient load.
WED-Food web models will be developed that track nutrient and carbon flow through aquatic
systems using a combination of stable isotope and population level data. This is a community
level model designed specifically to identify critical food web interactions that lead to changes in
species diversity, commercial harvest, eutrophication, and other factors that could effect the
designated use of a habitat. This stable isotope based model will be used to compare energy flow
in the different types of coastal systems where each of the Divisions are located. Application of
this model across all four Divisions will provide insight into carbon and nutrient processing
across a wide variety of systems.
GED-Will provide stable isotope measurements to test the WED food web model and will focus
on factors controlling shifts in the phytoplankton community structure and size in Escambia Bay,
FL, which is dominated by pico-cyanobacteria blooms in summer. Size of phytoplankton is an
important factor that determines the structuring of both the pelagic and benthic food webs.
60
-------
Picoplankton is associated with a microbial food web while nanoplankton is associated with
zooplankton consumers and commercially and charismatically important higher tropic levels.
This research will provide insight into the chemical and biological factors controlling pelagic
phytoplankton community.
AED-Will also measure stable isotope shifts and will provide a test of the WED food web model
including average food chain length. AED will investigate the effects of nutrients on saltmarsh
species changes and generate methods to relate nutrient loading to chl-a levels using remote
sensing, (i.e., airframe and satellite spectral analyses), including the ability to distinguish between
different types of phytoplankton blooms (diatoms vs. dinoflagellates).
MED-Will initially focus on two of the Great Lakes, Lake Superior, and Lake Michigan.
In addition to using stable isotopes to investigate food web structure in Great Lakes coastal
wetlands, nutrient effects oriented research will be employed to investigate several assumptions:
Algal-zooplankton size distribution in response to nutrient loading,
Algal community responses to N/P ratios in freshwater coastal wetlands, and
Modeling efforts focusing on establishing relationships of nutrient loadings and ambient
concentrations with chlorophyll, DO, N/P ratios, phytoplankton species composition,
food chain productivity, and water column transparency.
In addition, where possible, we will partner and collaborate with other coastal nutrient efforts
(i.e., State of the Lake Ecosystem Conference [SOLEC], GLEI, Atlantic Coast Environmental
Indicators Consortium, Great Lakes Coastal Initiative, States, and Tribes).
Products
APM ID FY 02 (GPRA # 166) Effects of nutrient loadings and altered nutrient ratios on HABs
(GED).
APM 3 (GPRA # 201) Comparison of effects of zooplankton grazing on estuarine phytoplankton
community under differing natural levels of grazing (GED).
APM 6A FY04 Sensitivity of food web responses to nutrient loading in coastal systems (GED).
APM 6B F Y05 Propose classification scheme for predicting sensitivity of coastal receiving
waters to effects of nutrients on food web structure (GED).
APM 6C FY06 Report on empirical and numeric models for food webs (WED).
APM 6D FY 06 Report on parameterization of food web models (GED).
APM 6E FY06 Report on classification scheme for grouping coastal or lake receiving waters
based on sensitivity to food web alterations (WED, GED, AED, MED).
61
-------
These products will include:
FY03 State of the Science report on nutrient food web relationships in Coastal Systems (AED,
GED, MED, WED).
F Y04 Interim report on sensitivity of food web response to nutrient loading in coastal systems
(AED, GED, MED, WED).
FY05 Recommendation on use of food web related endpoints to predict effects of nutrients on
important fish and shellfish populations (AED, GED, MED, WED).
Benefits of Products
This research will provide the basis for setting ecologically relevant nutrient criteria for 305b
reporting and TMDL development that supports the protection of aquatic life as mandated under
CWA. By providing standardized methodology, it will provide guidance to the States and EPA
Regions for developing appropriate monitoring protocols. In addition, a better understanding of
nutrient-food web response relationships will significantly improve our ability to predict
ecosystem response to other nutrient endpoints (DO and SAV loss).
Gap Analysis
In order to focus on what can be accomplished with the available resources, we have chosen the
four coastal regions described above. There are other major coastal regions we are not covering
such as Mid and South Atlantic systems or Southern Pacific coastal waters. In addition, this
research is focused on coastal receiving waters. It does not directly focus on understanding
nutrient response relationships in streams, rivers, lakes, inland wetlands, or headwater seeps, for
which additional research is needed. Longer range plans do include an integrated watershed
approach as resources become available. We are not developing DO criteria for any freshwater
species offish or SAV light requirements for freshwater SAV.
Currently we do not have sufficient understanding of how the biological components of an
ecosystem interact to process nutrients to be able to predict how differences in these components
affect the capacity of an ecosystem to assimilate nutrients. To do this more effectively, we need
to improve our skills in the area of ecosystems ecology and ecosystem modeling. In addition, the
presence of a database manager to establish a central database or linkages between Divisional
databases would significantly improve the transfer and sharing of data to be used and tested in a
variety of approaches (models and classification schemes). It is hoped that NERL will provide
nutrient loadings for receiving water bodies that will be considered by NHEERL; however, if this
is not possible then NFEERL will need to determine or estimate loadings.
References
Boynton, W.R., Hagy, J.D., Murray, L., Stokes, C., Kemp, W.M. 1996. A comparative analysis
of eutrophication patterns in a temperate coastal lagoon. Estuaries 198:408-421.
62
-------
Boynton, W.R., Kemp, W.M. 2000. The influence of river flow and nutrient loads on selected
ecosystem processes. In Hobbie, J., ed., Estuarine Science: a Synthetic Approach to Research
and Practice. Island Press, Washington, DC, pp. 269-298.
Chow-Frazer, P., 1998. A conceptual model to aid restoration of Cootes Paradise Marsh, a
degraded coastal wetland of Lake Ontario, Canada. WetlandEcol. Manag. 6:43-57.
Cohn, T.A. Delong, L.L. Gilroy, EJ. Hirsch, R.M. Wells, O.K. 1989. Estimating constituent
loads. Water Resources Research 25:937-942.
Costa, I.E., Howes, B.L., Janik, D., Aubrey, D., Gunn, E., Giblin, A.E. 1999. Managing
anthropogenic nitrogen inputs to coastal embayments: technical basis and evaluation of a
management strategy adopted for Buzzards Bay, Buzzards Bay Project.
Dawes, C.J., Tomasko, D.A. 1988. Depth distribution of Thalassia testudinum in two meadows
on the west coast of Florida; a difference in effect of light availability. Mar. Ecol. 9:123-130.
Dennison, W.C. 1987. Effects of light on seagrass photosynthesis, growth and depth distribution.
Aquat.Bot. 27:15-26.
Dennison, W.C., Orth, R.J., Moore, K.A., Stevenson, J.C., Carter, V., Kollar, S., Bergstrom,
P.W., Batiuk, R.A. 1993. Assessing water quality with submersed aquatic vegetation. BioScience
43:86-94.
Dettmann, E.H. 2002. Effect of water residence time on annual export and denitrification of
nitrogen in estuaries: a model analysis. Estuaries (in press).
Duarte, C.M. 1991. Seagrass depth limits. Aquat. Bot. 40:363-377.
Dunton, K.H. 1994. Seasonal growth and biomass of the tropical seagrass Halodule wrightii in a
hypersaline subtropical lagoon. Mar. Biol. 120:479-489.
Eldridge, P.M. Jackson, G.A. 1993. Benthic tropic dynamics in California coastal basin and
continental slope communities inferred using inverse analysis. Mar. Ecol. Prog. Ser. 99:115-135.
EPA. 2000. Ambient water quality criteria for dissolved oxygen (saltwater): Cape Cod to Cape
Hatteras. EPA-822-R-00-012. Office of Water, Washington, DC.
Fourqurean, J.W., Zieman, J.C. 1991. Photosynthesis, respiration and whole plant carbon budget
of the seagrass Thalassia testudinum. Mar. Ecol. Prog. Ser. 69:161-170.
Hagy, J.D. 2002. Eutrophication, hypoxia and trophic transfer efficiency in Chesapeake Bay.
Ph.D Dissertation, University of Maryland, College Park, MD.
Jude, D.J., Pampas, J. 1992. Fish utilization of Great Lakes coastal wetlands. J. Great Lakes Res.
18:651-672.
63
-------
Kaldy, I.E., Dunton, K.H., Kowalski, J.L., Lee, K.S. 2002. Evaluation of environmental factors
controlling the success of seagrass revegetation onto dredged material deposits: a case study in
Lower Laguna Madre, Texas. Restor. Ecol. (in revision).
Kenworthy, W.J., Haunert, D.E. 1991. The light requirements of seagrasses: proceedings of a
workshop to examine the capability of water quality criteria, standards and monitoring programs
to protect seagrasses. NMFS-SEFC-287. NOAA technical memorandum.
Koch, E.W. 2001. Beyond light: physical, geological and geochemical parameters as possible
submersed aquatic vegetation habitat requirements. Estuaries (in press).
Lalli, C.M., ed. 1990. Enclosed Experimental Marine Ecosystems: a Review and
Recommendations. Springer-Verlag, New York.
Livingston, R J. 2000. Eutrophication Processes in Coastal Systems: Origins and Succession of
Plankton Blooms and Effects on Secondary Production in Gulf Coastal Estuaries. CRC Press,
Boca Raton, FL. 327 pp.
Menge, B.A. 2000. Top-down and bottom-up community regulation in marine rocky intertidal
habitats. J. Exp. Mar. Biol. Ecol. 250:257-289.
Nixon, S.W. 1995. Coastal marine eutrophication: a definition, social causes, and future concerns
Ophelia 41:199-219.
Nixon, S.W., Ammerman, J.W. , Atkinson, L.P., Berounsky, V.M., Billen, G., Boicourt, W.C.,
Boynton, W.R., Church, T.M., Ditoro, D.M. , Garber, J.H., Giblin, A.E., Jahnke, R.A., Owens,
N.J.P. , Pilson, M.E.Q., Seitzinger, S.P. 1996. The fate of nitrogen and phosphorus at the land-
sea margin of the North Atlantic Ocean. Biogeochemistry 35: 141-180.
NRC. 2000. Clean Coastal Waters: Understanding and Reducing the Effects of Nutrient
Pollution. National Academy Press, Washington, DC.
Smith, R.A. Schwartz, GE. Alexander, R.B. 1997. Regional interpretation of water quality
monitoring data. Water Resources Research 33:2781-2798.
Stevenson, J.C., Staver, L.W., Staver, K.W. 1993. Water quality associated with survival of
submersed aquatic vegetation along an estuarine gradient. Estuaries 16:346-361.
Tomasko, D.A., Lapointe, B.E. 1991. Productivity and biomass oiThalassia testudinum as
related to water column nutrient availability and epiphyte levels: field observations and
experimental studies. Mar. Ecol. Prog. Ser. 75:9-17.
Ulanowicz, R.E. 1986. Growth and Development: Ecosystem Phenomenology. Springer-Verlag,
New York. 203 pp.
64
-------
Vezina, A.F., Pace, M.L. 1994. An inverse model analysis of planktonic food webs in
experimental lakes. Can. J. Fish. Sci. 51:2034-2044.
Vollenweider, R.A. 1975. Input-output models with special reference to the phosphorus loading
concept in limnology. Schweiz. Z. Hydrol. 37:53-84.
65
-------
Section 6.
Implementation Plan for Suspended and Bedded Sediment Research
NHEERL's effort concerning suspended and bedded sediments has been redirected since this
section was first written. The majority of the work in this research area will now occur under
Goal 8 (EMAP) because EMAP design techniques will be applied to develop effect thresholds
for suspended and bedded sediments in aquatic systems. Some of these techniques are described
generally in the Critical Path subsection of this implementation plan. However, at this time, the
effort under aquatic stressors will only include a literature review of suspended and bedded
sediments research. Results from this review will be combined with EMAP approaches to
synthesize and evaluate the state of the science. Once the review has been completed, data gaps
will be identified and additional research will be conducted, if warranted.
Problem
The Office of Water has identified suspended and bedded sediments as one of OW's highest and
most immediate priorities. The priorities within aquatic systems for developing these criteria
were identified as: rivers and streams; followed by lakes, reservoirs, ponds, and estuaries. For
purposes of this document, suspended sediments are those sediments that exert their negative
impact via their suspension in the water column, such as the effect of shading induced by them
on submerged macrophytes. Bedded sediments are those sediments that have their negative
impact when they are actually settled out and on the bottom of the water body of interest such as
fine sediments which smother spawning beds. Research in this section does not deal with
contaminated sediments (those containing toxic chemicals, see Section 7).
In streams and rivers, fine inorganic sediments, especially silts and clays, affect both the habitat
for macroinvertebrates and fish spawning, as well as fish rearing and feeding behavior. Larger
sands and gravels can scour diatoms and cause saltation of invertebrates, whereas suspended
sediment affects the light available for photosynthesizing plants and visual capacity of animals
(Waters 1995). A major problem with suspended sediment in coastal wetlands, estuaries, and
near-shore zones is the decreased light penetration which often causes aquatic macrophytes to be
replaced with algal communities, with resulting changes in both the invertebrate and fish
communities (Chow-Fraser 1998). Increased sedimentation also may functionally shift the fish
community from generalist feeding and spawning guilds to more bottom-oriented, silt tolerant
fishes (Berkman and Rabeni 1987, Muncy et al. 1979).
Thus suspended and bedded sediments are expected to have two major avenues of effect in
aquatic systems: 1) direct effects on biota and 2) direct effects on physical habitat, which result in
indirect effects on biota. Some examples of direct effects on biota include suppression of
submersed macrophytes through reduced light attenuation, changes in benthic algal communities,
and shifts to turbidity-tolerant fish communities. Effects of suspended and bedded sediments on
habitat structure include changes in refugia for biota (e.g., changes in macrophyte communities),
increased fines (and embeddedness) and scouring in streams, aggradation and destabilization of
stream channels, and filling in of wetlands and other receiving waters (Wilcock 1998, Lisle 1982,
Dietrich et al. 1989). Increased turbidity and concomitant changes in light regime could also be
66
-------
considered to be aspects of altered habitat. Indirect effects on biota will occur as the fish,
invertebrates, algae, amphibians, and birds that rely upon aquatic habitat for reproduction,
feeding, and cover are adversely affected by habitat loss or degradation (Platts et al. 1983,
Hawkins et al. 1983, Rinne 1988).
Sea grasses and other SAV are considered "keystone" species in temperate and tropical coastal
areas. These flora have a variety of beneficial attributes including providing food and shelter for
many biotic species. There has been a worldwide decline in sea grasses including dramatic
regional losses in the Gulf of Mexico. The reasons for the decline are unknown but reduction in
light attenuation (quantity and spectral quality) is thought to be a major factor. The presence of
suspended sediments is one factor that can impact water clarity; however, its significance to this
effect and observed sea grass declines is relatively unknown.
Ultimately, resolution of any problems associated with increases in suspended and bedded
sediments will need to address the sources of the sediment. These influxes of sediments, in
general, are associated with increased sediment delivery via soil erosion often caused by changes
in landuse and landcover, and changes in flow regimes that effect in-channel sediment transport
and loading. Recognition of the proximal and distal "causes" of suspended and bedded sediment
problems will affect the classification scheme used and the development of the "expectation" for
natural sediment loads.
Goals
The primary goal for NHEERL's Suspended and Bedded Sediment program is to provide and
demonstrate the approach for establishing sediment criteria that support aquatic life use in
streams/rivers, lakes/reservoirs, wetlands, and estuaries. A necessary first step for achieving this
goal is to assess the current knowledge and to report on the state of the science of this research
area. A specific APG and APM for this assessment follows. Additional goals are proposed, but
will depend on results from both EMAP research and the literature review.
APG 1 FY02 Synthesize state of the science and remaining uncertainties for developing criteria
for suspended and bedded sediments.
APM 1A FY02 Report on the state of the science and progress for developing suspended
and bedded sediment criteria (AED, MED).
Provide summary of biological response profiles for suspended and bedded sediments in marine
and freshwater systems.
Develop models that predict and scale biological responses to suspended solids and sediment
using assessment endpoints that support management decisions.
Develop classification schemes to optimize efficiency in developing suspended and bedded
sediment criteria.
67
-------
Provide the scientific basis for suspended and bedded sediment criteria for marine and freshwater
systems.
Critical Path
A general diagram for NHEERL's research program on suspended and bedded sediments is
shown in Figure 7. An explanation of the figure is given below, including the logical sequence
of information needed to develop the technical basis for deriving a criterion for suspended and
bedded sediments.
The major NHEERL responsibilities in Figure 7, under Goal 2, are to: review the literature and
existing State criteria and develop a general conceptual model of sediment effects, including
verbal descriptions of suspended and bedded sediment criteria that are explicit enough to shape
subsequent quantitative modeling (box 0). Based on regional-scale data (box la,b), develop
ecoregionally-specific models of the effects of suspended and bedded sediments on aquatic
assemblages in various types of aquatic ecosystems (box 2); then explore and confirm
mechanisms of sediment effects on assemblages and ecosystems through experimental research
incorporating controlled conditions and restricted taxa at smaller scales (box Ic). Using these
stressor-response relationships and models as a technical basis, develop an approach for
establishing sediment criteria (box 3), which OW, EPA Regions, and States may use to establish
criteria for suspended and bedded sediment (box 5). Using data from regional-scale surveys
(e.g., EMAP, Regional EMAP [REMAP], National Water Quality Assessment [NAWQA)]) and
more focused watershed studies, examine the stressor-response relationships between natural and
anthropogenic controls and the levels and transport of sediment (box 8). Based on this research
and models of the effects of sediments on aquatic biota (box 2), identify critical thresholds of
anthropogenic disturbance that lead to biologically-relevant sediment responses (box 4). These
ecoregionally-specific thresholds will be useful guidance to the TMDL process carried on by
States (box 6). Finally, review and revise the stressor-response models and recommended
sediment criteria based on feedback from monitoring data and further research (box 7).
The first step in this process is a review of the literature on the biological effects of suspended
and bedded sediments. The Office of Water has already begun this process. NHEERL can
contribute to this effort and expand it to evaluate the literature for useable stressor/response
relationships. Once the available literature has been reviewed, it will be necessary to develop a
conceptual model or framework of the effects of suspended and bedded sediments in aquatic
environments (box 0). This will guide the development of the remainder of our research
activities.
Classification of the expected response of aquatic systems to suspended solids and bedded
sediments is a critical early phase of research. The first step in developing a classification
scheme for sediments and suspended solids is the division of the national landscape into different
eco-regions that are sufficiently fine-scaled to accurately represent the vegetation, climate,
geology, and soils (box la). These factors will influence both the quantity and type of sediment
and suspended solids that will be carried from the landscape into the receiving water body. The
68
-------
OW, NERL, Goal 8
Loading estimates/models
Stressor/response relationships for sediment transport
NHEERL Goal 2
la
Classification scheme for
response of different aquatic
system types
Stressor/response relationships for
aquatic ecosystems and men- biotic \
lb assemblages <
It
Explore/confirm mechanisms of
sediment effects on specific key
populations and ecosystem components
1C under controlled conditions
Review literature/ develop conceptual model
and verbal description of sediment criteria
,/2 Models of effect of suspended/bedded
^ sediments on aquatic ecosystems
T Recommended approach for establishing
sediment criteria
4 Recommended sediment loading criteria -
identify thresholds of sediment response
OW, Regions, States
5
Suspended/bedded
sediment criteria
TMDL Process
Assessment of
Effectiveness
Figure 7. Critical path for suspended and bedded sediments research.
kinds of soils within these landscape delineations are particularly important. Secondly,
fragmentation, storage, and hydrogeomorphic characteristics of streams and rivers need to be
examined and classified at the watershed level as these factors influence the degree of flashiness
of streams and rivers to precipitation, snowfall run-off, and groundwater inputs (Leopold et al.
1964, Morisawa 1968, Mackin 1948). Ultimately these affect the timing of loadings, and
quantity and type of sediment, not only to streams and rivers but to the receiving bodies into
which they empty. The third step is classification by ecosystem type such as streams, rivers,
coastal wetlands, estuaries, or near-shore zones.
A discussion of the approaches anticipated for establishing stressor-response relationships for
streams and rivers serves as a useful guide to the type of research anticipated in box lb. For
bedded sediments in streams and rivers, it is likely that the expected levels of bedded fine
sediment in relatively undisturbed streams are ecoregionally specific, depending upon natural
climatic factors, topography, lithology, soil, and potential natural vegetation. It is also likely that
the intensity of the response of sediment to anthropogenic disturbance will also be dependent
69
-------
upon similar factors. The optimum regional classification to underlie modeling of sediment
expectations and response to anthropogenic disturbance will be based on work described in box
la. Using a combination of empirical data from relatively undisturbed watersheds and models
describing the physics of sediment supply and transport, we will estimate expected levels of
bedded sediment fines and embeddedness in stream and river reaches of specific size, slope, and
location. We will then examine the association between watershed/riparian land use and the
deviation of sediment concentrations from expected values, using survey data and data from
more detailed watershed studies. This effort should include interaction with NERL and NRMRL
to link landuse and landscape processes that may be responsible for delivering sediments. We
will need to take into account the likelihood that, because of natural disturbances (fire,
landslides, in-channel scouring due to instream hydrologic modifications), a certain portion of
the stream or river resource may have fine sediments substantially above or below the mean
expected value for the region. Therefore, the degree of impairment associated with deviations of
sediment from expected values is likely to be expressed in terms of statistical probabilities.
Once the degree of sedimentation is estimated for sample sites, we will examine associations
between biotic assemblages (algae, macroinvertebrates, fish, rooted aquatic plants), and/or key
aquatic species or guilds and deviations of sediment from expected values. In most cases, our
data sets will include sites affected by multiple stressors besides sediment that could potentially
act upon these aquatic biota. In such cases, a regional plot of sediment concentration versus
some biotic assemblage characteristic (e.g., % EPT [Emphemeroptera, Plecoptera, Trichoptera]),
will appear as a wedge-shaped pattern of points, where progressively higher fine sediment
concentrations are sufficient to limit % EPT numbers, but low concentrations do not guarantee
abundant EPT because of other habitat or chemical limitations (Terrell et al. 1996). These
patterns are consistent with a hypothesis that sediment is limiting biota. After demonstration of
a plausible causal mechanism (from detailed experimental studies) and elimination of other
plausible explanations for these observations, we will use these kinds of associational data in a
weight-of-evidence approach to support modeling the effects of bedded sediments on aquatic
biota.
For suspended sediments in streams and rivers, we will focus initially on chronic levels of
suspended sediments, rather than those resulting from episodic events such as those
accompanying storms. Expected natural levels of background suspended solids will be set on the
basis of data from flowing waters in basins relatively undisturbed by human land uses and (in
rivers) historic water clarity data to the extent possible. Regional reference areas could serve this
purpose. Where no relatively undisturbed waters exist, as for large rivers, we will use historic
data or reconstructions offish and/or macroinvertebrate assemblage composition to infer (from
published tolerance information) pre-disturbance suspended sediment characteristics. In an
approach similar to that for bedded sediments, we will examine associations between biotic
and/or key aquatic species or guilds and deviations of sediment from expected values in
appropriate regional settings. As for bedded sediments, we will seek patterns that are consistent
with biotic limitation by suspended sediment in a weight-of-evidence approach to support
modeling the effects of bedded sediments on aquatic biota, supporting this information with
controlled experimentation or literature reference to establish the suspended sediment levels that
cause substantial impairment of assemblages, sensitive guilds, or key species.
70
-------
A conditional probability approach also will be explored to determine possible effects of
suspended sediments on the biotic condition of streams. Data from the ORD/EMAP Mid-
Atlantic Highlands Assessment (MAHA) streams program (EMAP indicators and design),
reported in the MAHA streams report (EPA 2000), will be used in this application. The
approach uses survey data (sites selected with a probability design) and determines the likelihood
of impaired biological conditions for varying threshold levels of exposure or stressor variable(s)
(in this case, some form of suspended sediment concentration, possibly normalized for an
expectation level). The use of survey data permits an unbiased extrapolation of results to the
statistical population that the probability sample was drawn from. For example, the results
would be applicable to all of the wadeable streams in a state if the sample was drawn from all
wadeable streams in the state.
This approach is different from typical association approaches that relate exposure or stressor
conditions with impaired biological conditions, for example, water quality levels associated with
impaired fish communities (fish IBI values less than 3). The approach here "stratifies" the
resource for exceedance of a specified exposure or stressor value and then determines the fraction
of that strata with impaired biological conditions. Since the sites were selected with a probability
design, the fraction of the resource that is impaired is the probability of observing impaired
biological conditions in the resource for exceedance of the threshold value. This stratification is
then done for all values of the exposure or stressor variable. The result is a relationship for
probability of impaired biological conditions for exceedance of the exposure or stressor values.
This result is not a cumulative distribution function of the biological conditions since it relates
the conditions to a threshold level of another variable, and it is more than a simple scatter
diagram of biological condition with the exposure or stressor variable (the resource is
incrementally integrated or summed for second variable).
Issues associated with suspended and bedded sediments may be approached in a slightly different
manner in estuaries. One of the primary research needs is to determine whether sea grass decline
is correlated with the presence of increased suspended sediments. A combination of laboratory
and field work derived under natural (box Ib) and controlled conditions (box Ic) is needed to
derive protective water clarity criteria or to set management goals to maintain existing sea grass
coverage and community composition. This would be accomplished by the collection of
descriptive data (mapping and field data) at a variety of sampling sites. This research would
include monitoring basic characteristics of sea grass communities in reference areas and areas
which historically receive high levels of suspended solids using fixed transects or experimental
plots. Response parameters would include, but not be limited to, photo synthetic activity,
standing crop, root/shoot ratios, epiphytic coverage, blade characteristics, sea grass cover, and
density. Extensive water and sediment quality monitoring would be combined with this effort.
There are a variety of experimental designs available to determine, under controlled conditions,
the effects of suspended and bedded sediments on sea grass and other important submerged
aquatic vegetation. The determination of sensitive species, sensitive response parameters, and
modifying environmental factors are the objectives of these studies. The experimental designs
include exposing different species in laboratory tests to different levels and types of suspended
solids alone and in combination with other factors (such as salinity and nutrients) to determine
effects on light reduction and accompanying effects on biomass, pigment content, and other
71
-------
structural characteristics. Following these tests and the derivation of the necessary information,
mesocosms would be used in the laboratory and field (enclosures) to determine effects on
populations and communities of seagrasses. The results of the single species, population, and
community exposures will need field validation, which box Ib addresses in part. Following
these experiments, mesocosms could be used to determine effects on populations, and
communities of seagrasses and associated biota, particularly relating the role of seagrass beds as
habitat to fish and shellfish populations and communities.
The technical information on stressor/response relationships can be used to generate thresholds
for sediment effects (box 4). These thresholds can be used in the development of criteria (box 4)
for suspended and bedded sediments. These thresholds can serve as input to both the
development of criteria (box 5) and the TMDL process (box 6). These criteria will be
"integrated" criteria, similar to those discussed in the Toxic Chemical Section (Section 7). They
will consider the effects of suspended and bedded sediments in a more holistic manner than the
standard criteria do, taking into account effects on both benthic and water column organisms, and
direct as well as indirect effects on aquatic life use. These criteria also will have to take
magnitude, duration, and frequency of changes in suspended and bedded sediments into account.
The availability of sediment criteria and thresholds for sediment effects will allow for a TMDL
process that is "effects based". The current TMDL methodologies focus much more on exposure
and reduction of exposure, than on effects. Acceptable levels of suspended and bedded sediment
can not presently be based on effects, because the models and stressor-response relationships to
be derived in boxes 1 and 2 are not currently available. All that the NHEERL effort can provide,
at this time, is information on effects on certain organisms and classes of organisms. Resource
managers will have to use that information to make management decisions. At the same time,
we will have to make sure that the data are provided in such a way that is useful in the context of
designated uses. The data will have to be presented in as general a form as possible, as opposed
to just presenting a list of data on individual species for specific magnitudes and durations of
elevated suspended and bedded sediment concentrations. Priority will be given to those species
which are tied to designated uses.
Once TMDLs are produced, their effectiveness will be assessed (box 7) and this will allow
further refinement of the models developed in boxes 0, 1, and 2.
The final piece of the TMDL process is the exposure component. Loading estimates and models
are currently being developed by NERL, OW, and NFEERL (under Goal 8) (box 8). These
loading estimates and models will allow estimation of the changes in suspended and bedded
sediments inputs which might be needed to reach the targets set on the basis of the effects of
suspended and bedded sediments effects (boxes 4 and 5). These loading estimates and models
also provide input to the conceptual model, setting the bounds of bedded sediment and suspended
solids inputs to aquatic systems.
72
-------
Research Projects
Project Title 1. State of Science Review
Project Coordination and Resources (0.4 FTEs: AED-0.3, MED-0.1)
Objectives
To efficiently plan and manage NHEERL research efforts, a necessary first step is to assess the
current knowledge. The objective of this effort is to fulfill the first Goal under this research
implementation plan, which is to report on the state of the science and progress in developing
suspended and bedded sediment criteria.
Scientific Approach
NHEERL personnel will summarize and synthesize current knowledge on quantitative stressor-
response relationships among sedimentation, biota, and habitat, and the efforts to date in
developing sediment criteria. The overall review will be conducted in concert with other ORD
Laboratories, as well as OW, USGS, and the Army Corps of Engineers. Subject areas to be
reviewed will include:
1. A review of relationships between potential classification variables and suspended and bedded
sediments. Classification variables will include ecoregional factors (e.g., vegetation, climate,
geology, and soils); landscape characteristics such as forest fragmentation, water storage and
hydrogeomorphology; and ecosystem type (streams, rivers, reservoirs, coastal wetlands, estuaries
and near-shore zones).
2. A review of the known ecological effects of suspended and bedded sediments. Effects
categories include effects on different biotic assemblages such as microbes, primary producers,
invertebrates, and fish; issues of scale (e.g., effects at population, community, and ecosystem
levels of biological organization); and effects on habitat quality and quantity. Direct effects on
biota will be contrasted with effects on physical habitat which may result in indirect effects on
biota. Mechanistic, experimental approaches toward detecting and analyzing effects will be
compared to large-scale empirical, correlational analyses.
3. The effects of elevated suspended solids (turbidity) and excessive bedded sediments (i.e.,
increased sedimentation). It is expected that the ecological effects of suspended solids differ
from those of sedimented solids. The review will encompass both stressor types. Potential
impacts due to lack of enough suspended and bedded sediment, which can contribute to habitat
loss, also will be investigated.
4. Routes and mechanisms of the delivery of sediments to aquatic ecosystems. A critical step
toward effects prediction and ecosystem protection is to understand the relationships among
characteristics and processes at the landscape scale and the quantity and quality of sediment
delivered to aquatic systems, whether that be terrestrial runoff with increased sediment loads or
hydrologic modifications that result in increases in-inchannel erosion.
73
-------
5. Approaches toward modeling the effects of suspended solids and sediments on aquatic
ecosystems.
6. Progress to date in developing state, National, and international suspended and bedded
sediment criteria.
Products
APG 1 FY02 Synthesize state of the science and remaining uncertainties for developing criteria
for suspended and bedded sediments.
APM 1A FY02 Report on the state of the science and progress for developing suspended
and bedded sediment criteria (AED, MED).
Benefits of Products
An up-to-date summary of the state of the science review of suspended and bedded sediments
information will be provided with recommendations of needed research to develop and validate
suspended and bedded sediment criteria for OW.
Proposed Research Projects
Based on results from EMAP research and the literature review, the following general projects
are proposed assuming resources are available:
Stressor-Response Relationships
Associations between biotic assemblages and/or key aquatic species or guilds and deviations of
sediment from expected values should be examined. In most cases, our data sets will include
sites affected by multiple stressors besides sediment that could potentially act upon these aquatic
biota. After demonstration of a plausible causal mechanism (from detailed experimental studies)
and elimination of other plausible explanations for these observations, we will use these kinds of
associational data in a weight-of-evidence approach to support modeling the effects of suspended
and bedded sediments on aquatic biota.
Thresholds for Sediment Effects
The technical information on stressor/response relationships can be used to generate thresholds
for sediment effects. These thresholds can be used in the development of criteria for suspended
and bedded sediments, and thresholds can serve as input to both the development of criteria and
to the TMDL process. They will consider the effects of suspended and bedded sediments in a
more holistic manner than the standard criteria do, taking into account effects on both benthic
and water column organisms, and direct as well as indirect effects on aquatic life use.
74
-------
Classification
The purpose of the classification research is to develop an effective scheme for defining those
waters for which similar ambient levels of suspended or bedded sediments are expected. We
anticipated the final solution will incorporate information about water body type, geographic
setting, and specific, local hydrologic settings. This proposed classification scheme that results
should be review with or compared to the classification in use (at that time) for establishing
reference conditions for biological criteria.
Gap Analysis
Research into the problems associated with suspended and bedded sediments also is being
carried out by other groups. The Army Corps of Engineers is working on the effects of
resuspension associated with dredging projects, for example (Wilber and Clarke 2001). Goal 8
research (Kaufmann et al. 1999, Kaufmann and Robison 1998, EPA 2000), research supported by
the Office of Wetlands, Oceans, and Watersheds (OWOW) (TMDL framework for clean
sediments), and NERL are currently contributing to the development of loading estimates and
models (Figure 7, box 8). However, the work has barely begun on the effects of sediments to
aquatic systems, at the low levels that may exert long-term, chronic effects. The first step
outlined in the critical path is the review of the state of science. The intent of this gap section is
to describe what work would remain to reach the goals once NHEERL completed the work
outlined. The dilemma is that NHEERL will not be in a position to outline the details of what we
will do until we complete the state of science review. At that point and in concert with Division
management decisions on FTE dedicated to this project, the gaps that will remain can be
identified.
References
Berkman, H.E., Rabeni, C. F. 1987. Effect of siltation on stream fish communities. Environ. Biol.
Fish. 18:285-294.
Chow-Fraser, P. 1998. A conceptual ecological model to aid restoration of Cootes Paradise
Marsh, a degraded coastal wetland of Lake Ontario, Canada. WetlandEcol. Manag. 6:43-57.
Dietrich, W.E., Kirchner, J.W., Ikeda, H., Iseya, F. 1989. Sediment supply and the development
of the coarse surface layer in gravel bed rivers. Nature 340:215-217.
EPA. 2000. Mid-Atlantic highlands streams assessment. EPA/903/R-00/015. U.S. Environmental
Protection Agency. Region 3. Philadelphia, PA. 364 pp.
Hawkins, C.P., Murphy, M.L., Anderson, NJ. 1983. Density offish and salamanders in relation
to riparian canopy and physical habitat in streams of the northwestern United States. Can. J.
Fish. Aquat. Sci. 40:1173-1186.
Kaufmann, P.R., Robison, E.G. 1998. Physical habitat assessment. In Klemm, D.J., Lazorchak,
J.M., eds., Environmental Monitoring and Assessment Program 1994 Pilot Field Operations
75
-------
Manual for Streams. EPA/620/R-94/004. EPA, Environ. Monit. Syst. Lab., Office of Research
and Development, Cincinnati, OH, pp. 6-1 to 6-38.
Kaufmann, P.R., Levine, P., Robison, E.G., Seeliger, C., Peck, D. 1999. Quantifying physical
habitat in wadeable streams. EPA 620/R-99/003. EPA, Environmental Monitoring and
Assessment Program, Corvallis, OR.
Leopold, L.B., Wolman, M.G., Miller, J.P. 1964. Fluvial Processes in Geomorphology. W.H.
Freeman and Company, San Francisco, CA, 522 pp.
Lisle, I.E. 1982. Effects of aggradation and degradation on riffle-pool morphology in natural
gravel channels, northwestern California. Water. Resour. Res. 18:1643-1651.
Mackin, J.H. 1948. Concept of the graded river. Geol. Soc. Am. Bull. 59:463-512.
Morisawa, M. 1968. Streams, Their Dynamics and Morphology. McGraw-Hill, New York. 175
pp.
Muncy, R.J., Atchison, G.J., Bulkley, R.V., Menzel, B.W., Perry, L.G., Summerfelt, R.C. 1979.
Effects of suspended solids and sediment on reproduction and early life of warm water fishes: a
review. EPA 600/3-79-042. EPA, Washington, DC.
Platts, W.S., Megahan, W.F., Minshall, G.W. 1983. Methods for evaluating stream, riparian and
biotic conditions. Gen. Tech. Rep. INT-138, U.S. Forest Service, Intermountain Forest and
Range Experiment Station, Ogden, UT. 70 pp.
Rinne, J. 1988. Effects of livestock grazing exclosure on aquatic macroinvertebrates in a
montane stream, New Mexico. Great Basin Nat. 48:146-153.
Simons, D.B., Senturk, F. 1977. Sediment transport technology. Water Resources Publications,
Fort Collins, CO. 807 pp.
Terrell, J.W., Cade, B.S., Carpenter, J., Thompson, J.M.. 1996. Modeling stream fish habitat
limitations from wedge-shaped patterns of variation in standing stock. Trans. Am. Fish. Soc.
125:104-117.
Waters, T.F. 1995. Sediment in streams: sources, biological effects, and controls. American
Fisheries Society, Bethesda, MD.
Wilber, D.H., D.G. Clarke. 2001. Biological effects of suspended sediments: a review of the
suspended sediment impacts on fish and shellfish with relation to dredging activities in estuaries.
N. Am. J. Fish. Manage. 21:855-875.
Wilcock, P.R. 1988. Two-fraction model of initial sediment motion in gravel-bed rivers. Science
280:410-412.
76
-------
Section 7.
Implementation Plan for Toxic Chemicals Research
Problem
Effective management of toxic chemicals in aquatic ecosystems requires a capability to
quantitatively predict the ecological effects that can be expected from different levels of chemical
contamination of water, sediments, and food chains. Procedures for deriving aquatic life WQC
have existed for many years (EPA 1973, 1980, 1991, 1994, 1995a; Stephan et al. 1985) and have
been useful for managing toxic chemical inputs to aquatic systems. However, these procedures
are based on simplifying assumptions and a relatively narrow framework that limit their use in
fully assessing the risk of a wide range of toxic chemicals to both aquatic life and aquatic-
dependent wildlife. Sediment guidelines developed more recently (EPA 2000a,b,c,d) have many
of the same limitations as WQC. To address some of these concerns, NHEERL has prepared a
draft wildlife research strategy for assessing risks of multiple stressors to populations of
amphibians, birds, and mammals (EPA 2000e).
Criteria derivation and application require extrapolations of toxicological effects observed in the
laboratory to field conditions, which can result in significant uncertainties. Differences in water
characteristics, chemical partitioning, routes of exposure, organism habits, and exposure time-
series can greatly affect the relationship between exposure concentrations and a chemical's
toxicity, and thus affect the applicability of criteria to natural ecosystems. This is particularly
true for PBTs, for which effects often depend on tissue residues accumulated in tissues over long
times as a result of multiple exposure routes. Moreover, criteria often do not address the
combined effects of multiple chemicals and other stressors, and can lack information for
potentially sensitive life stages of test species.
Other uncertainties in criteria arise from the use of organismal-level toxicity to set concentrations
protective of aquatic populations and communities. The relationship of toxic effects on
individual organisms to population responses is not well established. Important taxa and
endpoints can be missing from the sets of tests used to develop criteria and sediment guidelines.
Indirect effects of chemicals on organisms (effect on food sources, competition, predation, and
shelter) generally are not considered.
Criteria are also limited in that they address only specific water concentrations, rather than
complete dose-response relationships, thus limiting how well risks can be characterized.
Seasonal issues and the significance of the spatial extent of exposures are incompletely
addressed. Current criteria procedures also do not include uncertainty analyses or address how
well risk can be assessed in the presence of limited data.
Because of these limitations, efforts are needed to develop methods to better characterize risks to
aquatic life and aquatic-dependent wildlife populations and communities, and to apply these risks
to criteria development. Assessment endpoints should be better defined and an analytical
framework developed for linking available data to a more complete and accurate description of
risks for these endpoints. This assessment framework should describe a range of responses, be
77
-------
tiered to allow some decisions to be made with limited data, and should include analysis of the
uncertainties in estimated risks.
Goals
The general goal of this work is to develop scientifically-defensible methods for better describing
the risks of toxic chemicals to aquatic and aquatic-dependent populations and communities, in
support of improving criteria procedures needed to satisfy GPRA Goal 2, Objective 2. The work
to be discussed here represents only part of the ecological toxicology efforts within NHEERL
and is focused on prospective assessments of chemicals for which WQC exist or are desired. As
such, the proposed research will not address issues which are not closely related to such
assessments and/or are a subject of research under other goals. This includes such areas as
chemicals and endpoints of emerging or potential concern, biological indicators (including DNA
or protein-based probes) for retrospective assessments and diagnostics, and basic investigations
of the cellular/subcellular and physiological mechanisms of toxicological responses.
Further details on research needs and NHEERL research efforts to address those needs are
described in the following subsections. Specific goals of these efforts are summarized in the
following APGs and APMs Note: some APGs and APMs (including those under GPRA) were
established before this document was written. They are listed here along with APGs and APMs
that were developed under the aquatic stressors process.
APG 1 FY02 (GPRA #31) Provide a method for setting risk-based aquatic life criteria for toxic
chemicals which minimizes uncertainties of translating national and site-specific water quality
criteria.
APM 1A FY02 (GRPA # 167) Report on integrated water and sediment quality criteria
methods for assessing site-specific risks of persistent bioaccumulative toxicants to
aquatic species (MED).
APG 2 FY03 Demonstrate methods to set risk-based water quality criteria for toxic substances.
APM 2A FY03 Describe a framework for WQC for nonbioaccumulative chemicals that
more fully describes risk to aquatic organisms (MED).
APG 3 FY05 (GPRA #111) Provide methods for developing WQC based on characterization of
population-level risks of toxic chemicals to aquatic life and aquatic-dependent wildlife.
APM 3 A FY04 (GPRA # 59) Population models that project the relative risks of multiple
stressors (toxic chemicals, habitat alterations) to piscivorous birds (AED, MED).
APG 4 FY06 Provide methods for extrapolating chemical toxicity data across exposure
conditions and across endpoints, life stages, and species, which can support assessment of risks
to aquatic life and aquatic-dependent wildlife for chemicals with limited data.
78
-------
APM 4A FY02 Interspecies correlation estimations (ICEs) for acute toxicity to aquatic
organisms (GED).
APM 4B FY02 Time-concentration-effect models for use in predicting chronic toxicity
from acute toxicity data (GED).
APM 4C FY03 Acute-to-chronic estimation (ACE) user guide and software (GED).
APM 4D FY06 Report evaluating importance of dietary route of exposures to aquatic risk
assessments for metals (MED).
APG 5 FY08 Provide approaches for evaluating the relative and cumulative risks from toxic
chemicals, with respect to risks from nonchemical stressors, on populations of aquatic life and
aquatic-dependent wildlife at various spatial scales.
APM 5 A FY05 Report regarding assessment of risks to aquatic organisms from
combined exposure to polycyclic aromatic hydrocarbon (PAHs) mixtures and ultraviolet
(UV) radiation in natural systems (MED).
APM 5B FY06 Approaches for addressing spatial scale issues in assessing risks of
multiple stressors to wildlife populations in spatially-diverse landscapes (AED, MED).
Critical Path
Defining critical research paths needed to improve aquatic risk assessments and criteria
development for toxic chemicals should start with consideration of the problem formulation that
should be part of any good risk assessment (Figure 8). There needs to be clear definition of the
assessment problem, including the ecological effects (assessment endpoints) and exposure
scenarios of concern, and better conceptual models which define the logical structure of the
assessments. These conceptual models should identify critical toxicological endpoints to be used
in the assessment (i.e., measurement endpoints) and how these are to be related to the assessment
endpoints, based on knowledge of the dynamics of the ecosystem(s) of concern. There also
should be specification of how assessments might be tiered initially, basing evaluations on
limited data to determine whether risks might be significant, and adding data as needed to make
more definitive assessments.
With better definition of the conceptual model, the needs of the other phases of a risk assessment
can be better identified. Methods are needed so that the exposure profiles (Figure 8) can
describe, in sufficient detail, the distribution of the toxicant(s) relative to the biological receptors.
This would include evaluation of the temporal and spatial variability of exposure and the
chemical's speciation and partitioning to the extent needed to determine the distribution of
toxicological responses. Response profiles (Figure 8) need to provide good organism-level
response models and linkages between organism response and population/community responses.
When limited toxicological data are available, methods for extrapolating among species and
endpoints also will be needed. Methods must support risk characterizations which describe a
79
-------
Problem Formulation
Q)
U
=
.2
a.
E
o
Concept
Mode
Critical
Species and
Endpoints
z>
ual j
1 ^
Population &
Community
Dynamics
Ecosystem
Properties
Exposure
Analysis
Toxicological and Ecological
Data
E
Cfl
M
0)
VI
Fate and
Transport
Special ion,
Partitioning
Receptor
Distribution
Organism
Response
Population
Dynamics
Spatially-
distributed
Response
Exposure Profile
(Temporal/SpatiaiyForm Distribution)
Response Profile
(As Function of E^osure Distribution)
a
'E
o
S.
75
S!
_O
O
O
ractei
Integration of Exposure and Response Profiles
Risk E stimation and Uncertainty Analysis
Risk Description
Risk Mar
aqement
Selection of
Level of Risk
Figure 8. Ecological risk assessment framework (modified from EPA 1992).
80
-------
range of effects, including their uncertainties, and sufficiently account for important site
characteristics. The resultant risk descriptions should allow risk managers to assess the
implications of different exposure scenarios and, once acceptable risks are specified, to back-
calculate acceptable loadings (dotted line on Figure 8).
Critical research paths also should recognize the need to implement progressive improvements in
the assessment methodology. Figure 9 shows four steps in such improvement, which largely
parallel the APGs described in the preceding Goals subsection. The first step will demonstrate
how criteria can reflect more comprehensive risk characterizations, based on more complete
descriptions of organism-level dose-response relationships and how they vary with exposure
conditions. The second step will involve the development and application of methods to address
population-level risks. The third step will involve the development of extrapolation methods so
that the risks can be assessed for chemicals or situations for which limited data is available. The
fourth step will include further enhancement of these techniques and emphasize their
incorporation into multistressor assessments at various spatial scales.
Demonstrate methods for improved criteria
at the individual level based on
improved characterization of risks
Develop methods to link individual-level data
to population-level endpoints
Develop methods to support risk assessments
for chemicals with limited data
Develop methods to evaluate risks on
populations at various spatial scales,
in the context of other stressors
Figure 9. Critical path for developing site-specific methodologies for establishing the risks of
toxic chemicals to aquatic life and aquatic dependent wildlife.
81
-------
The rest of this section will further discuss the improvements needed in aquatic risk assessment
methods and identify the research and development efforts needed to realize these improvements.
This will be done in the context of conceptual models which identify the essential components of
assessments and emphasize the link of this research to the risk assessment framework. The
nature of the assessments and the needed research can vary depending on the chemical and
system of concern. One issue of particular importance is the degree to which chemicals
bioaccumulate. Risk assessments for highly bioaccumulative toxicants will differ in some key
features from assessment for chemicals which have low bioaccumulation. For bioaccumulative
toxicants, dietary exposure will be especially important, residue-based dose-response models will
be more important, and bioaccumulation models will be needed for dose determinations.
Toxicokinetics for these chemicals are generally slow, so acute toxicity and short-term temporal
variation in exposures will often, but not always, be of less importance than for
nonbioaccumulative toxicants. Biomagnification makes risk relatively more important for
organisms high in the food chain, and some risk assessments of bioaccumulative chemicals could
concentrate on populations of such organisms, rather than on the broad aquatic community.
Because of these differences, the assessment needs will be discussed separately below for
nonbioaccumulative and bioaccumulative toxicants.
It is true that assessments of these two groups of chemicals will share many common features and
principles, and thus these separate discussions will involve some redundancies. For example,
toxic response for almost all chemicals and endpoints is ultimately related to the amount of
chemical which accumulates at an internal site of action, so a residue-based framework is of
value to nonbioaccumulative chemicals as well as bioaccumulative chemicals. Also, the
consequences to populations and communities of toxic effects on individual organisms will entail
the use of similar tools for both groups of chemicals. Furthermore, whether a chemical is
"bioaccumulative" is a matter of degree, not a discrete category. Nonetheless, the relative
importance of various issues will vary between assessments of chemicals with low versus high
bioaccumulation. Separate discussions are useful in highlighting how assessments will vary, and
in identifying needed work. However, the results of the work proposed below will often
transcend the chemical group and efforts will be made to apply any results as broadly as is
appropriate.
Nonbioaccumulative Toxicants
Nonbioaccumulative toxicants include a wide range of organic and inorganic chemicals that are
of concern in many aquatic risk assessments. Notable examples include heavy cationic metals
(such as copper, zinc, lead, cadmium, and silver), which are significant contaminants in mining
areas and various effluents; and ammonia, which is at high concentrations in sewage effluents, in
some fertilizer runoff, and in areas with high inputs of nitrogen-containing organic compounds.
These chemicals have been documented to have substantial impacts on certain aquatic systems
and are responsible for many instances of noncompliance with WQC. Treatment and
remediation costs associated with meeting criteria for these chemicals can be high. However, as
discussed above, current assessment and criteria methodology entail various uncertainties
regarding the actual risks of these chemicals and thus whether the regulatory controls are
necessary and sufficient.
82
-------
Cationic metals and ammonia will be used as examples, where helpful, in the discussion here and
as candidates for chemical-specific research discussed later. They represent major concerns in
current aquatic risk assessments and have a rich toxicity literature that will support development
of dose-response models, and evaluation of the quality of assessments made with limited data.
However, while the discussion here will in part highlight these chemicals, the issues discussed
are true for many nonbioaccumulative toxicants. General methods developed for any particular
chemical should be directly, or by example, applicable to other nonbioaccumulative toxicants.
Conceptual Model
As discussed above, research needs for development of risk assessment methods need to relate to
the conceptual model for the assessment(s). Figure 10 presents a simple conceptual model that
identifies important elements that need to be considered in conducting and developing methods
for aquatic risk assessments and criteria development. The horizontal series of boxes on the
bottom of the figure represent major stages in linking chemical stressors to ecological responses.
Loadings of toxic chemicals to an aquatic system are distributed throughout the system, resulting
in exposures to biological receptors. These exposures may adversely affect the survival, growth,
and reproduction of individual organisms. Such individual-level effects may be expressed at the
population level as changes in population size, growth rate, and structure. Changes at the
population level may in turn elicit changes in the community. The "stacked-boxes" used to
represent these stages are intended to indicate that assessments often will involve multiple
chemicals and/or biological species.
The arrows linking the lower series of boxes represent mechanistic relations among these stages,
and involve the application of various models identified by the series of ovals in the center of the
figure. The arrows are bi-directional to indicate the fundamental similarities between two modes
of environmental protection: 1) criteria development and application that determine chemical
loadings consistent with protecting desired system values (left-to-right) and 2) assessments that
characterize risks expected from chemical exposures (right-to-left).
The boxes on the top of Figure 10 identify data needs for the models used in the assessment. All
of the models will generally require various information on various physical, chemical, and
biological characteristics of the system. The exposure and toxicity models also will require
information on the nature of the chemical and its toxicity to various biological receptors.
This conceptual model helps to identify and evaluate research needed to improve aquatic risk
assessments and criteria development. Primary attention should be given to the four basic
models in Figure 10. How are current assessments limited by the inability of exposure models to
completely and accurately predict and describe the relationship between chemical loadings and
exposure concentrations? How well do current methods actually describe risks to individual-
level endpoints under exposure conditions expected in natural systems? What is the significance
of organismal-level effects to populations and communities? Another research focus would be
the adequacy of data needed for these models. Can required chemical properties be estimated
well? Does available toxicity data address the endpoints needed in the assessment?
83
-------
Figure 10. Simple conceptual model for risk assessments of nonbioaccumulative toxicants.
84
-------
How well can toxicity data be extrapolated among species and endpoints? Finally, research
efforts should address the need to describe and test a complete methodology for conducting
assessments and developing criteria.
Research Needs
Per the conceptual model presented in Figure 10, fundamental tools of assessments are the
models linking the state variables. Brief descriptions of each model and its research needs are as
follows:
Exposure Model
The exposure model should produce a description of the spatial/temporal distribution and the
speciation of the chemical that satisfies the input needs of the effects models. It must also
describe other physical and chemical properties that can affect lexicological response. Because
NHEERL research primarily addresses effects assessment, only a brief overview of exposure
research needs will be given here. These do, however, represent significant knowledge gaps that
need to be addressed if improved risk-based criteria are to be implemented.
The exposure model must assess physical transport, degradation, speciation, and partitioning of
the chemical. Current implementations of WQC and WQS often use simple models that will not
support more comprehensive characterizations of risks. For physical transport, the primary need
is to adapt and apply better methods and models that are available. In particular, better risk
characterizations need dynamic, two- and three-dimensional models, which can more fully
describe the spectrum of exposures experienced by biological receptors.
Degradation is not an issue for elemental toxicants such as metals, but is of utmost important for
ammonia. Decomposition of nitrogen-containing organic matter produces ammonia, while
nitrification in oxygenated water reduces ammonia concentrations. Nitrification is reduced at
colder temperatures, resulting in higher and more widespread ammonia concentrations during
winter months. More comprehensive characterization of risk will require more complete
modeling of the spatial and seasonal distribution of ammonia. Of particular concern is the
ammonia concentrations in or near the sediment that results from the interaction of organic
matter decomposition, nitrification, and mixing processes.
Speciation is well characterized for ammonia, but current models for metal speciation have
considerable uncertainty, especially regarding complexation by organic matter. There is a need
both to improve current predictive models and also to develop analytical techniques that can
reliably and efficiently measure metal speciation in laboratory tests and natural systems.
Whether by model or measurement, these methods must describe speciation sufficiently for the
interpretation and application of the effects of various chemical parameters on metals toxicity.
Partitioning between sediment and overlying water is of utmost importance to assessing the risk
of both ammonia and metals. Exposure analyses for the application of WQC need to incorporate
the role of sediment as a source or sink of chemical, processes that entail considerable
uncertainty. Currently there are no published EPA guidelines for ammonia in sediment, but
85
-------
Equilibrium-partitioning Sediment Guidelines (ESGs) for metals (EPA 2000a)use acid volatile
sulfide (AVS) and interstitial water to predict biological effects. These methods have been
demonstrated to be very useful in predicting biological effects in laboratory experiments and in a
limited number of field experiments, but few of them have examined the importance of temporal
and spatial variability in exposures within and near sediments, which limits the usefulness of
these equilibrium approaches. Research is needed to characterize how AVS and metals
concentrations at the sediment/water interface vary seasonally and due to hydrological events to
support better characterization of effects.
Toxicity Model
The toxicity model in a risk assessment must integrate exposure information, toxicological data,
interspecies extrapolations, dose-response models, and effects of exposure conditions into
characterizations of risks for various organism-level endpoints, species, and life stages.
Depending on the requirements of the assessment, including the tier level, various research and
development needs exist.
A significant shortcoming in current criteria and many assessments for nonbioaccumulative
toxicants is consideration of toxicity information for only a single level of effect at a set exposure
duration (e.g., 96-hr LC50). Critical to better risk characterization are models that more
comprehensively describe toxic response, including both the magnitude of response and the
effect of exposure time series. Some capabilities for this have long existed. Models that describe
the relationship of effects to concentration or time are standard tools in toxicity, but traditionally
have not combined the effects of both. Furthermore, these models generally assume that
concentrations are constant. Mancini (1983) described a model that would describe the effects of
fluctuating concentrations based on standard toxicity tests, but did not incorporate variable levels
of effects into the model. Subsequent work has tested the utility of this approach with mixed
results and broadened the model to describe variable levels of effects as a function of both
concentration and time. Although this model does involve significant uncertainties, it is well
suited to form the core element in risk assessments and criteria development. It can use toxicity
information from standard tests and describe the level of effects, with uncertainties, expected
from any exposure series. This model can thus be combined with expected spatial and temporal
exposure distributions to produce individual-level risk curves for populations of biological
receptors. This would provide the individual-level risk characterization in Figure 10 needed for
any further risk characterization at the population or community level. Other modeling
approaches, such as proportional hazard and accelerated failure models, also might provide a
basis for better describing toxicological risks and their dependence on exposure time-series.
Based on current information, the critical immediate need is to further develop these techniques,
establish their validity and uncertainty, and describe their application to WQC, which would be a
major part of the first step in Figure 9. This would establish a core framework for improving
criteria, but also would identify knowledge gaps that should be addressed by further research.
Chemical toxicity to aquatic organisms can be influenced markedly by various physicochemical
exposure factors. For example, ammonia toxicity can vary by orders of magnitude due to the
combined effects of pH and temperature and by significant amounts due to DO and certain major
ions. EPA aquatic life criteria for ammonia (EPA 1999) are a function of pH and temperature,
86
-------
but do not fully account for these effects and do not account at all for other factors. An
especially important uncertainty (which is also relevant to other chemicals) is the effect of winter
conditions on chronic ammonia toxicity to life-stages of sensitive organisms present during the
winter. Aquatic life WQC for metals currently are expressed as dissolved metals, with a
correction for hardness in fresh water, reflecting empirical observations of reduced toxicity due
to adsorption by suspended solids and increased hardness. Research has shown, however, that
several factors other than hardness also are important in determining the toxicity of metals in
water, and has identified mechanistic bases for their effects. Development of a model that
incorporates many of these factors, referred to as the biotic ligand model (BLM) is currently
being supported in part by OW. Applications of current versions of this model to criteria entail
substantial uncertainties, especially for chronic toxicity and for certain taxonomic groups.
Significant research in support of this model is underway under the support of OW and various
industry groups. Such efforts will contribute to the third step in Figure 9.
As discussed above regarding exposure research issues, organisms dwelling in or on sediments
can receive chemical exposures with large temporal and spatial variation, due to hydrologic
events disturbing the sediments or seasonal changes affecting partitioning or processing of
chemicals. Decomposition of nitrogen-bearing organic matter in the sediment can cause
substantial gradients between sediment and overlying water. For metals, sediment can be
relatively more contaminated than overlying water due to past inputs, diagenesis of deposited
particles, or seasonal fluctuations in loads to systems, leading also to substantial gradients from
sediment to overlying water that are subject to marked variation. Sediment pore water metal
concentrations in anoxic layers are greatly affected by AVS, which varies with season and depth
and can be oxidized when sediment is resuspended. Even in laboratory tests, sediments and
overlying water are likely at some disequilibrium that varies with time and between tests.
Regulations based on assumptions of equilibrium therefore can involve considerable uncertainty.
The toxicological implications of non-equilibrium conditions (such as resuspension and
seasonality) need to be investigated as part of the extrapolation methods represented in the third
step in Figure 9.
Laboratory tests used for aquatic risk assessments of nonbioaccumulative toxicants typically
include exposures only via chemical dissolved in water, the organisms either not being fed or fed
food that is not contaminated commensurate with the water concentrations. This is not an issue
with ammonia, as ammonia per se would not be a significant contaminant in food, although
catabolism of food is a source of ammonia that organisms must eliminate. However, the
significance of the dietary route of exposure is an important issue for metals assessments.
Aquatic animals certainly can accumulate substantial amounts of metals from their diets, and
diets contaminated with high concentrations of metals can cause adverse effects. However, the
importance of this route of exposure when there is also commensurate amounts of metal
dissolved in the water is uncertain. Some studies have presented evidence that dietary uptake
results in increased uptake and/or risk compared to water exposure alone, while other studies
have suggested the opposite. The uncertainty regarding the significance of dietary uptake to
metal risk is a fundamental barrier to good risk assessments for these chemicals and could have
substantial implications to regulatory programs.
87
-------
A common uncertainty in any aquatic risk assessment is the sensitivity of species and endpoints
not directly tested. In some cases, there is considerable toxicity data, but certain important data
are missing, for example, having acute mortality data, but not chronic reproduction data, for a
critical or sensitive species. Additionally, there are many cases where desired assessments have
available only a limited number of laboratory toxicity data, if any. Therefore, there is a need for
models that address extrapolations across endpoints and/or species, and the uncertainties of these
extrapolations. This relates to the toxicity data needs box in Figure 10 and is a major part of the
third step in Figure 9. A research effort is needed to improve and test extrapolation methods, and
to apply these methods to simulated risk assessments with limited data, providing better
methodology for conducting interspecies extrapolations for nonbioaccumulative toxicants in
general.
A final area of major concern is the joint effects of multiple chemical stressors and the effects of
chemicals in the presence of non-chemical stressors. Much laboratory-based research has been
done on these issues, but has seen little application to EPA criteria, except recently for sediment
toxicity assessments. A principal need is to determine how this past work should be incorporated
into assessments, and to better identify what additional work will be worthwhile.
Population Model
Toxicity tests can provide information on a diverse set of endpoints, but comparing the relative
risk of these endpoints and their significance at the population and community level is difficult.
An important step in better defining the significance of these lexicological effects is to
incorporate them into population model that will translate these effects into some common
"currency" of population dynamics. A critical research need therefore is to develop, test, and
apply population models for a variety of species relevant to assessments of nonbioaccumulative
toxicants. This is represented in the second step in Figure 9. These models will be used not only
to describe the significance of observed and predicted organism-level toxicity, but also to
evaluate the usefulness of toxicity tests and to determine needed changes in the types of tests
conducted.
Stage-structured population models can be used to link individual-level effects to the population
level, such that a set of vital rates defines the dynamics of a population. Such models require
estimates of vital rates for all life stages, which may or may not show effects from a given level
of toxic chemical exposure. A set of vital rates also can define the life history strategy of a
species. Life history strategies vary along a continuum from species with early reproduction,
high fecundity, and short life expectancy (r-selection model) to species with delayed
reproduction, low fecundity, and long life expectancy (K-selection model). Population-level
responses may differ between species because of differences in life history strategies, even
though individuals of different species may show a similar response to a toxic contaminant.
Variation in life history strategies, and therefore variation in population-level responses to toxic
contaminants, also may occur geographically in the same species.
Population models should be constructed for fish, shellfish, and wildlife species exhibiting
different life history strategies (i.e., species exhibiting different sets of vital rates). Population
models generic to groups of species whose life history strategies are similar also can be
88
-------
developed for use at the screening level or tier of risk assessments, to be followed by models
specific to individual species in higher-tiered assessments.
An important issue regarding population models is the type of individual-level effects they need
as inputs to provide useful assessments of population-level effects. The toxicity tests and data
typically available are often limited with regard to the endpoints and life-stages tested, often
leaving significant gaps in the information needed to assess impacts of toxicity on population
dynamics. Population model development needs to identify the critical inputs and be coordinated
with toxicity model research to provide the methodology to make this information available. In
particular, efforts to improve extrapolation of toxicity data should take such needs into
consideration.
Community Model
Unless assessments are simply intended to provide protection to the most sensitive taxonomic
groups, some synthesis is needed to relate expected effects on populations or individuals to
consequences to the aquatic community. As mentioned above, current WQC are premised on the
judgment that occasional low level toxicity to sensitive species constitutes acceptably low risk to
those species and to the aquatic communities to which they belong. This might be a sound
judgment, but the lack of quantification makes it impossible to conclude whether higher
exposures might also pose minimal risk, or what acceptably low risk means.
There is thus a need to better quantify the association between toxic effects and effects on aquatic
communities. This could entail a few lines of efforts. Better meta-analysis is needed of the
variety of community assessments in experimental or natural ecosystems, linking these to a
metric that better describes available toxicity data. More efforts are needed to study responses to
toxic chemicals in complex aquatic ecosystems. Aquatic community models could be developed
that integrate effects at the individual or populations-level. As part of these efforts, there is also a
need to identify individual- and population-level endpoints important for supporting better
community-level assessments, and to coordinate with lexicological research to provide this
necessary information.
Bioaccumulative Toxicants
Toxicity risks associated with PBTs are expected to be advantageously assessed with residue-
based dose response models. This important concept may be used to describe the minimum
degree to which chemicals must be persistent and bioaccumulative in order to fall into this class.
The degree of persistence determines what concentrations in water and sediments are available
for incorporation of the chemical into benthic and pelagic food chains which lead to exposure of
vulnerable organisms. The potential for a chemical to bioaccumulate in an organism is
commonly referenced to concentrations of the chemical which persist in water and sediments.
Thus, bioaccumulation factors (BAFs) and biota sediment accumulation factors (BSAFs), in
accordance with mechanistic food chain models which integrate all routes of exposure, are
extremely important components of the risk assessment methodology for PBTs. For organic
chemicals, hydrophobicity, as measured by the octanol-water partition coefficient (Kow), is the
primary determinant of bioaccumulation potential with metabolism in the food chain as an
89
-------
important factor for decreased bioaccumulation potential. Bioaccumulation for other PBTs, such
as organometallic compounds like methyl mercury or organic chemicals with mechanisms for
bioaccumulation different than hydrophobicity, must be assessed on the basis of the mechanism
for their bioaccumulation.
The first residue-based WQC for PBTs (PCBs, DDT and metabolites, mercury, and 2,3,7,8-
tetrachlorodibenzo-p-dioxin [TCDD]), were developed for the Great Lakes under the Great Lakes
Water Quality Guidance (EPA 1995a). Criteria were developed for risks to human health and
wildlife, but not aquatic life. A similar procedure with BAFs was used to promulgate the
National methodology for deriving human health criteria (EPA 2000f).
Conceptual Model
The Great Lakes Water Quality Guidance (EPA 1995a) for PBTs is consistent with a conceptual
model for risk based criteria development for determination of safe loadings to aquatic systems
or remedial actions in cases where unsafe loadings have created contaminated sediment
problems. This conceptual model (Figure 11) illustrates a number of factors that are relevant to
development of risk assessment methods and water quality criteria for PBTs in aquatic
ecosystems. The double arrows represent models and relationships for transforming and linking
the data and conditions represented by the boxes in the conceptual model. Concepts therein
which provide an initial basis for development of projects for aquatic PBT research are:
1. Wildlife, aquatic life, and human health risk assessment methodologies for PBTs follow
parallel tracks with common elements such as toxicity models, bioaccumulation factors, and
chemical fate and transport models.
2. Residue-based WQC for PBTs should incorporate all of the elements of the risk assessment
paradigm (problem formulation, effects/hazards analysis, exposure analysis, risk characterization,
and risk management). Setting water and sediment quality standards, in accordance with the
PBT conceptual model (Figure 11), requires a left to right logic for acquiring data and
assembling models.
3. Retrospective risk assessments (e.g., what are/were the ecological risks associated with the
mass of chemical X in lake Y?) most often occur in response to a chemical stressor diagnosis and
utilize the models and data in a right to left direction.
4. While prospective and retrospective risk assessments tend to flow from left to right and from
right to left, respectively (Figure 11), in reality, site specific assessments are expected to involve
iterations with reverse flows in data collection and modeling. For example, in a prospective risk
assessment differences in exposure for species in a particular ecosystem may alter assumptions of
the species at risk and require extrapolation of toxicity data (return to effects analysis) to
complete a risk characterization.
90
-------
Demographic
and Social
Impacts
Model
Reproduction
Cancer
Imunotox
Neurotox. etc
Toxicity
Model
Human
Exposures
Food
Consumption
C omrr
Structu
Funo
i
unity
tion
Aquatic
Life
4 0--
i
E cos ys te m
Model
Wildlife
Population
Dynamics
j
i
Popu
Dyna
i
ation
rnics
Population
Model
Population
Model
Reproduction
Age-sp. Mortal
Disease Resist
Orouith
Reproduction
Age-sp. Mortal
Pred. Vulner.
Disease Resist
Growth
Toxicity__
Model
Toxicity
Model
Res id
Aqu
Orga
i
Food
i
Wile
Expos
Resi
ues in
atic
lEtTlS
Bioacc.
Model
BAF/
B3AF
Food
Wa
Sediment
i i
Cons umption
life
ures/
dues
Bioacc. Model
BAF/BSAF
Chemical
Mass
Balance
Models
Chemical
Loading
Risks
Hazards
Exposures
Figure 11. Conceptual model for risk assessments and criteria development involving determination of safe loadings of
bioaccumulative toxicants to aquatic systems.
91
-------
5. To date, PBT risk assessment methods for wildlife and humans have been based on
assessment of dietary exposures involving concentrations in aquatic food organisms, principally
fish and shellfish. Although this results in common BAFs/BSAFs for aquatic life and wildlife,
BAFs/BSAFs specific for wildlife species are measurable and could be modeled to reduce
uncertainties associated with the present dietary exposure methodology.
6. Determination of sensitive species, critical end points for population sustainability, and
residue-based toxicology data are largely generic considerations which can be assembled on a
national basis. However, bioaccumulation, exposure conditions, and chemical mass balance are
more site-specific considerations.
7. The conceptual model applies to deterministic or probabilistic analyses.
8. Toxicity risks are linked to changes in chemical loadings or remedial actions and vice versa,
so that tiered assessments are possible with choice, in the problem formulation, of steady state as
either a specific assessment condition or as a reference condition for time-dependent risk
assessment.
9. Spatially explicit risk assessments require spatially explicit exposure and bioaccumulation
analysis with feedback (right to left) to population model responses through the residue based
toxicology model (Figure 11). Spatially explicit risk analysis for PBTs can be conceptualized, in
accordance with Figure 11, as multiple fish or wildlife tracks based on different exposure levels,
a common toxicity model, and the potential for different population level responses (assuming
discrete populations, or meta-populations exposed within the assessment region).
10. The third dimension in the conceptual model (stacked boxes) may be viewed as the multi-
stressor component, particularly for chemical mixtures. Each chemical has to satisfy the risk
assessment requirements in a parallel manner with a joint toxicity model (such as toxicity
equivalence) linking each chemical's contribution to the net exposure.
Research Needs
For this discussion, the five different paths in Figure 11 (box-arrow-box components) that
involve the generation and use of data to accomplish components of risk assessments will be
examined for research needs. Separate but parallel chains of the five paths emerge for aquatic
and wildlife species from the conceptual model for PBTs. The five components are: 1) the
relationships between chemical loadings to the ecosystem and exposures via food, water, and
sediment; 2) the relationships between concentrations in food, water, and sediment to
concentrations in tissues of organisms at risk; 3) the relationships between concentrations in
organisms and incidence of effects; 4) the relationships between incidence of effects and
population changes; and 5) the relationships between specific population changes and community
structure and function. The chemical loading, fate and transport, and exposure path (mass
balance model) is not a direct NFffiERL research concern but must be considered as a reference
point for integration of effects research with exposure research.
92
-------
Bioaccumulation Model
Much progress has been made in the last decade toward providing a reliable bioaccumulation
prediction capability for aquatic organisms through complementary use of empirical BAFs and
more mechanistic food chain models. Site-specific variations in bioavailability have been largely
reduced through lipid and organic carbon normalization and tropic level determination. Besides
the chemical's hydrophobicity and potential for metabolism in the food chain, we now recognize
the distribution between water and sediments and the relative benthic versus pelagic food web
composition as critical determinants of bioaccumulation. The following important gaps need to
be filled:
1. Rates of metabolism are needed for many PBTs in order to allow accurate predictions of
bioaccumulation with food chain models. These rates are best determined from field data in
order to fit the risk assessment needs.
2. The metabolism rate gap extends to bioaccumulation of PBTs like the PAHs in embryo-larval
stages offish with potential vulnerability to photo-induced toxicity.
3. Very few bioaccumulation data sets are of sufficient quality to validate the uses of BAFs and
BSAFs, especially when extrapolated across species and/or ecosystems.
4. Existing BAFs, BSAFs, and food chain models are based on whole adult organisms and thus
may not be sufficient when dose to early life stages (ELSs) and/or specific tissues must be
evaluated. Early life stage dosimetry-based bioaccumulation factors and PB-TK models are
needed to fill this gap.
5. Comprehensive and compatible BAFs, BSAFs, and food chain models are needed to meet the
requirements of joint action toxicity models such as for TCDD toxicity equivalence or photo-
induced PAH toxicity.
Toxicity Model
As would be expected, existing toxicity data vary greatly in amount and applicability for different
PBTs. The risk assessment paradigm (Figure 8) and the conceptual model for PBTs (Figure 11)
provide contexts for determining toxicological research needs for PBTs. Chemical residue-based
dosimetry is an essential requirement. Ideally, residue dose-response models should relate to the
most sensitive end points, species, and life stages which may result in population declines.
Often, however, the toxicity data available are obtained prior to establishment of specific risk
assessment requirements. This increases the need for development of tools for extrapolation of
effects across species and end points. It also increases the need to plan future toxicology research
so that the data fit into an ecological risk assessment profile for the particular class of PBT. Such
a profile might be based on the mechanism of toxicity, ecological and exposure vulnerability of
species, and expectations for differences in species sensitivity. A trivial example would be the
low benefit to be expected from investigation of a specific receptor mediated toxicity in a class of
aquatic organisms which is known to not possess the receptor. Most PBTs fall in groups based
93
-------
on a common mechanism of toxicity and thus joint toxicity models are an important facet of this
path. The following important gaps need to be filled:
1. Residue-based toxicity data bases need to be advanced and evaluated for applicability to
aquatic ecological risk assessment requirements for PBTs.
2. Is absence of overt mortality, even for early life stages, an adequate effects end point for
preventing population declines caused by PBTs? If not, how do we determine what is adequate?
3. Commonly measured biochemical effect indicators, such as P450 enzyme induction, have
uncertain relationships to organismal, much less population-level, risks.
4. Complex, multi-stressor models, such as required for photo-induced PAH toxicity to fish
during embryo-larval stage of development, need to be developed and applied to determine the
magnitude of ecological risks which are presently highly uncertain.
5. Virtually unexplored are the toxicokinetic and toxicodynamic determinants for interspecies
and inter-effect extrapolations of potency ratios required for PBT mixture toxicity risk
assessment using a toxic units model approach, such as the additive TCDD toxicity equivalence
model.
Population Model
Population models are used to translate organismal responses to toxicity into population changes
which may reflect risk to the population. Thus, life stage specific mortality or chronic effects
which reduce survival may lead to a reduced population or extinction. While population models,
such as the Leslie matrix age-class or individually based models, have been developed, relatively
few applications in retrospective risk assessments for aquatic organisms exposed to PBTs have
been reported. Examples of prospective use of population models for protection of aquatic life
from PBT toxicity are fewer. This is unfortunate because WQC and other forms of risk
assessment for protection of aquatic life from PBT toxicity should have a problem formulation
based on the levels of population protection required. Clearly, the population model path is
important for relating toxicity-induced mortality of individual organisms to population level
changes in an ecosystem. However, a potentially equally important use is for the systematic
definition of species characteristics, life stages, and toxicity effects that are most likely to
determine risks to populations associated with PBTs having a particular mechanism of action.
The following important gaps need to be filled:
1. Population models need to be developed and applied through case studies to explicitly
demonstrate risk assessment requirements for prediction of adverse population impacts, as a
result of PBT toxicity caused reductions in survival of aquatic organisms.
2. Complex mixtures of PBTs are the norm, so interspecies differences in potency, as well as in
bioaccumulation, for individual chemicals in the mixture must be factored into population level
risk predictions.
94
-------
3. Because PBTs tend to distribute widely, if not uniformly, in aquatic habitats, uncertainty
exists for the extent to which spatially explicit population models are requisite for problem
formulations.
4. A national WQC methodology for different classes of PBTs needs definition, through use of a
generic population model, of the species characteristics, life stages, and toxicity effects that are
most likely to determine risks to populations, regardless of site conditions. This information will
provide the basis for determining site-specific model and data requirements for application of the
criteria.
Research Projects
The research and development needs presented in the critical path constitute a very large effort
encompassing integrated assessment methodology, generic assessment components applicable to
many chemicals, and data or model needs for specific chemicals. Formulating a Goal 2 toxic
chemicals research program for NHEERL requires consideration of what can be done with
available resources to produce the most beneficial improvements in assessment methodology
over the next several years, while also considering what research will be done in other goals and
by other parties. The research presented here was selected to provide both general methods
development and reduction of specific uncertainties.
Overview of Projects-Nonbioaccumulative Toxicants
At the most general level, there is a need to describe the overall framework and methods that
should be used to improve aquatic risk characterizations and criteria development/application.
This includes a short-term need, as represented by the initial step in Figure 9, to describe general
approaches for improving assessments for nonbioaccumulative toxicants and the use of currently-
available methods and knowledge, thus providing a starting point for criteria development and
further research. There also is a longer-term need to periodically update the recommended
framework and methodologies as research in various areas produces useful results. Project Nl
(Improved Risk Characterization Methods for Developing Aquatic Life Criteria for
Nonbioaccumulative Toxicants) will address these needs and serve as a focal point for
developing research projects, and applying and integrating their results. Initial products from this
project will address APG 2, but efforts will continue under this project to synthesize results from
other research, describe their application to assessments, and address, in whole or part, other
APGs. This continuing work will be in large part in collaboration with OW efforts to revise their
guidelines for WQC derivation. The nature and time line of additional products will be defined
as these collaborative efforts develop during FY03.
As discussed in the Critical Path subsection, more meaningful aquatic risk assessments require
improved approaches and knowledge in various areas. Prominent among these are models that
address the significance of toxicological responses at the population-level (APG 3). Projects that
we will pursue in this area will only involve bioaccumulative toxicants (see below), but the
results of such work will have relevance to nonbioaccumulative toxicants and will be part of
continuing efforts in project Nl. Proposed research projects involving nonbioaccumulative
toxicants involve the general topic area identified in APG 4 (extrapolation of toxicity among
95
-------
exposure conditions and biological endpoints). This encompasses many important issues, and
three specific areas were selected which address major uncertainties or gaps for criteria, which
are feasible with available resources, and which are not being resolved to our knowledge in other
research programs.
Methods which can extrapolate toxicity data across different endpoints, species, and life-stages
can benefit criteria development and application in two major ways. First, the importance of
endpoints or organisms missing from toxicological data sets used for criteria development can be
estimated (e.g., endangered species). Second, criteria for some chemicals could be developed
from more limited data sets than generally required. Project N2 (Methods for Extrapolating
Chemical Toxicity Data Across Endpoints, Life Stages, and Species Which Can Support
Assessment of Risks to Aquatic Life for Chemicals with Limited Data) will address certain
methods for conducting such extrapolations.
The toxicities of some contaminants are particularly sensitive to exposure conditions, sometimes
varying by orders of magnitude. Even with a better risk assessment framework and other tools,
good risk assessments of such contaminants are not possible without resolution of the effects of
various exposure parameters on toxicity. A particularly noteworthy uncertainty is the
bioavailability of metals. The Office of Water is currently supporting development of BLM,
which addresses the effects of various chemical constituents on metal toxicity and is based in part
on previous work in this area by NHEERL. Because of these efforts and various industry-funded
research projects, further work on water-borne metal bioavailability is not being proposed here.
Rather, work is proposed in areas that represent critical knowledge gaps that are receiving less
attention.
Toxicological responses of organisms in or on sediment are affected by temporal and spatial
variations in chemical concentrations and speciation in the sediment/water boundary zone. This
is true in laboratory test systems and even more so under field conditions, making the
interpretation and application of toxicity data highly uncertain. This area of concern will be
addressed in project N3 (Assessing the Significance of Non-equilibrium Conditions on Aquatic
Guidelines to Better Predict Field Effects). (Note: Because of resource reductions after the initial
preparation of this plan, this project will not be pursued at this time.)
Aquatic life criteria for metals have generally assumed that exposure is predominantly via water,
with dietary exposure being negligible; however, some work has indicated that this assumption is
not true and that risk might be substantially underestimated. Project N4 (Risks of Heavy Metals
to Aquatic Organisms from Dietary Exposures) will address this area of uncertainty for metals
risk assessment and should also provide results and insights that can help address this issue for
other nonbioaccumulative toxicants.
Overview of Projects-Bioaccumulative Toxicants
The residue-based toxicity approach for PBTs embodied in the conceptual model for ecological
risk assessments and criteria development (Figure 11) is the foundation for advancing methods
that will effectively link PBT loadings to aquatic ecosystems, or watersheds to risks for adverse
population changes in aquatic life and wildlife associated with aquatic food webs. Therefore,
96
-------
there is a need to establish a residue-based toxicity framework with associated models. This
framework is intended to improve aquatic ecological risk assessments and criteria
development/application. In so doing, it can also be used to design, conduct, and report results
for a long term PBT aquatic risk research program. The framework, like the framework for non-
bioaccumulative toxicants, should be updated periodically and revised as new research improves
and expands the methods, models, and data available for effectively applying the conceptual
model. These periodic improvements need to be reported to OW and other interested Program
Offices in a context that emphasizes integration into the framework and practical applications of
the resulting risk assessment methodology. Project B1 (Frameworkfor Development and
Application of Population Risk-Based Criteria for Fish and Wildlife Exposed to Persistent
Bioaccumulative Toxicants) serves this need. The initial products under project Bl will address
APG 1. Subsequent improvements to the framework and validation efforts will concentrate on
population level impacts and thus address APG 3. As for project Nl, work in this project will in
large part be in conjunction with OW efforts to update WQC guidelines. The nature and time
line of additional products will be defined as these collaborative efforts develop during FY03.
As discussed in the Research Needs subsection for bioaccumulative toxicants, BAFs and models
are essential for application of the residue-based toxicity approach. Although basic models and
approaches are available for predicting bioaccumulation throughout aquatic food webs, improved
capabilities are needed. Prime examples are the need to incorporate the effects of chemical
metabolism into predictions of whole organism chemical elimination rates, the need to predict
site-specific bioaccumulation with minimum data sets, and the need to extend bioaccumulation
models to allow tissue residue predictions for vertebrates during the embryo and subsequent early
stages of development. Project B2 (Incorporate Chemical Metabolism Rates and Site-specific
Bioavailability into Bioaccumulation Models Structured for Practical Assessments of Risks to
Fish and Wildlife Exposed to PBTs) is intended to advance the state of knowledge and risk
assessment capabilities for all five of the major gaps identified under the Bioaccumulation Model
Path. In so doing, project B2 will ultimately share with project B4 a goal of development of a
model for fish ELS bioaccumulation of chemicals with significant metabolism potential such as
PAHs. Although influencing APG 3 and APG 5, project B2 will primarily address APG 4.
(Note: Because of resource reductions after the initial preparation of this plan, this project will
not be pursued at this time.)
Projects B3 and B4, although focused on specific PBTs, will provide products needed to fill
important gaps associated with both the toxicity model and the population model paths of the
conceptual model. Exposure to methyl mercury arguably creates the most widespread and
intractable PBT risk problem for piscivorous birds. Project B3 (Multiple Stressor Risks to
Common Loon and Other Piscivorous Bird Populations) will provide population models that are
capable of evaluating relative risks of multiple stressors, including habitat alterations, in response
to APG 3. Project B3 will also advance methods for interspecies extrapolation of dose-response
relationships through development of PBTK/TD models in support of APG 4, and extend the
population model to assessment of risks to wildlife from multiple stressors across spatially
diverse landscapes in support of APG 5. Project B4 (Risks to Fish Populations from PAHs in
Natural Systems) inherently addresses complex chemical mixture modeling issues with
complexity added in association with the need to model photo-activation of PAHs in tissues of
organisms, including early life stages offish. Since data will probably always be limited for
97
-------
assessing photo-induced PAH toxicity and thus extrapolations are required to assess risks, project
B4 will contribute to APG 4. Project B4 also will provide the key aquatic life related
contribution to APG 5. The determination of the extent to which populations of aquatic
organisms, including fish during early life stages, are reduced due to photo-induced toxicity
across the broad range of PAH contamination in aquatic ecosystems will require achievement
and application of a highly advanced PBT risk assessment capability. This PBT risk problem
will provide a uniquely challenging test of PBT risk assessment methodologies because, although
concentrations of complex mixtures of PAHs in organisms maybe sufficiently described as a
steady-state condition, the timing of UV exposure required for photo-activation is highly variable
in time and space as well as being subject to habitat conditions and organismal/species behavior.
Project Title Nl. Improved Risk Characterization Methods for Developing Aquatic Life
Criteria for Nonbioaccumulative Toxicants
Project Coordination and Resources (4.0 FTEs: AED-1.5, GED-0.5, MED-2.0)
Objectives
Current WQC incorporate only limited information regarding the magnitude and time-
dependency of the responses of aquatic organisms to toxic chemicals. They address only one
point (the fifth percentile) in the distribution of toxic effects concentrations among tested species.
No uncertainty estimates are made and the importance of untested species and endpoints is not
assessed. Assessments are not made at all in the absence of certain minimum datasets. The
spatial variation of exposure, especially between sediment and water column, is not addressed.
Except for recent efforts regarding sediment assessments, the effects of multiple stressors are
generally not considered nor are the consequences of toxic effects on populations as a result of
exposure of individuals. These limitations result in a weakly defined definition of risk associated
with criteria conditions and an inability to quantify how risk would change with exposures above
or below criteria concentrations.
Methods do exist for more completely addressing many of these issues, albeit with some
uncertainty, and thus providing a more meaningful statement of risk. However, this will require
adoption of more comprehensive risk assessment framework for criteria. A major short term
objective of this project will be to describe a methodological framework for such risk
characterization that could be used to improve criteria derivation. Current and possible
methodologies for the components of this framework will be described, identifying where
improvements to criteria are possible with current knowledge and where research efforts are
needed. This effort will provide APM (2A) under APG 2.
The longer term objective of this project will be to further test, refine, and describe this
framework and its component methodologies, based on results of research conducted in other
projects in this program and elsewhere. In particular, we will address application of methods for
population-level assessments and for extrapolation of toxicological information among endpoints
and exposure conditions (APG 3 and APG 4). Products will consist of reports which synthesize
new information and update descriptions of risk assessment methods relevant to aquatic life
criteria. Much of this work will be pursued in collaboration with OW efforts to update WQC
98
-------
guidelines, and the specific nature and time line of products will be developed as part of this
collaboration.
Scientific Approach
This project will use existing and developing information to evaluate and demonstrate procedures
for more fully characterizing risks of nonbioaccumulative toxicants to aquatic organisms, and
incorporating these risks into aquatic life criteria. These efforts will address a variety of issues,
identified and discussed as follows. For all these issues, initial efforts will describe general
approaches, discuss the capabilities and limitations of current methods, and identify needed
research efforts.
Improved descriptions of risk must start with methods which better describe individual-
level concentration-response relationships than endpoints that address only a single level
of effect under narrow exposure conditions (e.g., 96-hr LC50). A chemical will be
selected (likely candidates: ammonia, copper) for which raw toxicity test data are
available for a large number of tests and diverse aquatic species. Existing methods for
describing the effects of exposure time series on toxicological response (Mancini 1983,
Breck 1988, Erickson et al. 1989, Hickie et al. 1995, Meyer et al. Newman 1995) will be
evaluated, refined, and tested using these data sets. Models and estimation procedures for
describing fixed-duration concentration-response curves also will be evaluated. These
methods will be used to develop procedures for describing risks, and their uncertainties,
for individual-level endpoints as a function of exposure parameters in a manner useful for
criteria development and application.
Efforts also will address better description of species-sensitivity distributions to allow
criteria to address more than just a single level (i.e., the fifth percentile genus) in the
range of available toxicity data. Methods for describing the distribution of toxic
sensitivities among taxa (Kooijman 1987, Wagner and Lokke 1991, Aldenberg and Slob
1993, Baker et al 1994, Solomon et al. 1996, Hall et al. 1998, Newman et al. 2000) will
be reviewed and used to develop an aggregate, continuous measure of risk from the
assemblage of available toxicity data, which can be applied to various exposure
conditions and provide a more quantitative basis for specifying aquatic life criteria.
Uncertainties in this analysis will be described to the extent possible and the effects of
using limited data on the estimated risks and their uncertainties will be evaluated.
The use of population models in the recently developed saltwater DO criteria will be
reviewed and the broader applicability of population models to criteria will be addressed
in initial efforts. Subsequent work will, as appropriate, review advances in methods from
other research in this plan (projects Bl and B3) and elsewhere, and update
recommendations regarding assessment methods for aquatic life criteria.
Correlations of responses among taxa and endpoints will be discussed with regard to how
criteria can address data gaps and limited data. Current work on interspecies
extrapolation in project N2 will be summarized in initial products, and results of
continued work in this project will be included in later efforts.
99
-------
The importance of spatial variation in exposure for assessing risk to aquatic communities
will be evaluated and discussed. This will include the general issue of the spatial extent
of the area that WQC criteria are intended to protect. It will also include consideration of
the effects of spatial and temporal variability at the sediment/water interface (drawing on
results from project N3) and the implications of this to the integration of sediment and
water column criteria.
The effects of physicochemical exposure conditions and of multiple routes of exposure
will be addressed. Initial efforts will include general reviews of the importance of such
factors and current abilities to address these effects. Later efforts will address
developments from ongoing research regarding these effects, including results from
project N4 on effects of dietary metals.
Exposure to multiple chemicals will be addressed. Existing literature which describes the
effects of chemical mixtures will be reviewed, procedures appropriate for WQC
development will be developed, and the significance of joint toxic action to common
contamination scenarios will be examined.
Products
FY03 Journal articles evaluating methods for describing the relationship of toxic responses to
time and concentration and their application to improved expressions of risk for WQC (part of
APM 2A).
Benefits of Products
This work will provide OW with a prototype framework for criteria based on a more accurate and
informative characterization of organismal-level risk, and will provide a more quantitative basis
for deriving and evaluating criteria. This framework will address several of the limitations of
current criteria and support incorporation of tools addressing other limitations, such as
population-level effects and extrapolations among species and toxicity endpoints. Products will
provide a technical basis for making changes to procedures for developing criteria.
Project Title N2. Methods for Extrapolating Chemical Toxicity Data Across Endpoints, Life
Stages, and Species Which Can Support Assessment of Risks to Aquatic Life for Chemicals
with Limited Data
Project Coordination and Resources (1.5 FTEs: GED-1.5)
Objectives
Ambient WQC have enabled the states to develop scientifically defensible standards and thereby
reduce the amount of specific chemicals discharged into our nation's surface waters. However,
environmental managers must frequently perform risk analyses and make decisions regarding
compounds for which WQC and the data necessary to derive them (EPA 1980) do not exist.
Guidance is needed on comparative sensitivity relationships and on limits of extrapolation, and
100
-------
uncertainty for use by managers who must on many occasions, make decisions based on the
limited data that may be available. This is of particular concern for the protection of certain
endangered species which cannot be tested or other species that are not feasible to test.
Most species sensitivity comparisons have been made on individual chemicals, and modes of
action have been used for certain extrapolations for individual species. However, there is little
understanding of the relationships and uncertainty between chemical classes and the sensitivity of
taxonomic groups of species to these chemical classes. A particular mode of action (chemical
class) may pose greater or lesser jeopardy to certain families/populations of organisms than
others, allowing one to better assess what will occur ecologically (diversity). This concept will
be evaluated in extrapolating chemical toxicity data across species and taxonomic groupings.
The present research specifically covers five major modes of action (see second item below), but
will add others from existing data bases, further supporting chemical modes of action/structure-
activity research being addressed byNHEERL/WED.
Effects on native fish and invertebrate populations are important indicators of changes in surface
waters due to human-related impacts such as the damming of rivers, lowering of aquifers,
addition of pollutants, and introduction of non-native species. The planned research to determine
the utility of using surrogate species in hazard evaluations to estimate the potential for toxic
chemicals to affect other aquatic species has the following objectives:
1. Assess the rainbow trout, fathead minnow, and sheepshead minnow as appropriate surrogate
test species for endangered fishes and other species.
2. Determine differences in acute sensitivity to chemicals with differing modes of action
(carbaryl, copper, 4-nonylphenol, pentachlorophenol, and permethrin) between surrogate test
species and selected endangered organisms.
3. Develop interspecies correlations between surrogate test species and endangered fishes and
other aquatic species using 48-h EC50/96-h LC50 data for the above five chemicals.
4. Develop user manual and software for interspecies correlations of acute toxicity data for
aquatic species using data bases from Mayer and Ellersieck (1986), Mayer (1987), OPP, and
AQUatic toxicity Information REetrieval (AQUIRE).
5. Perform acute toxicity tests to fill data gaps identified in item 4 where performance of specific
tests would significantly enhance the number of data sets available for use in developing
interspecies correlations (e.g., additional modes of action in item 2).
6. Enhance utility of ACE (acute-to-chronic endpoint) model to predict chronic toxicity to
endangered and other species on a population basis.
This research will further develop project Nl by addressing the importance of untested species
and endpoints and providing methodology for assessing limited data sets and filling data gaps. In
addition, models for estimating chronic toxicity from acute toxicity data will concentrate on
population-level assessments.
101
-------
Scientific Approach
Existing toxicity data bases, comparative toxicity literature, and methods for extrapolation among
endpoints, life stages, and species of aquatic organisms will be reviewed to identify areas where
improvements can be made in extrapolation based on existing data.
Specific data gaps, particularly those addressing modes of action/structure-activity, will be
identified and tests conducted where performance of laboratory toxicity tests would significantly
enhance the ability to extrapolate among endpoints, life stages, and species. This research can
best be accomplished by collaboration among the Ecology Divisions, because some of the data
gaps will involve freshwater testing and others will involve testing of saltwater species.
Collaboration with other governmental agencies such as the FWS, NOAA, and U.S. Geological
Survey (USGS) is also desirable.
Methods for acute tests will be based on standard methods such as Weber (1993) and ASTM
(1988). Dissolved oxygen, pH, salinity, and temperature will be measured in all treatments on
day 0 through day 4. Most of the effort with endangered species has been completed and any
additional testing of such species will be conducted by USGS/BRD, Columbia Environmental
Research Center, Columbia, MO through an interagency agreement jointly funded by ORD, OW,
and Office of Prevention, Pesticides, and Toxic Substances (OPPTS).
Statistical analyses will be performed on survival data with probit analysis or the Spearman-
Karber program to generate as a minimum, 24-, 48-, 72- and 96-h EC/LC50s. Three-way
analysis of variance is performed using SAS to determine statistical differences between species
LCSOs, among chemicals, and at each time interval.
Interspecies correlations will be conducted using Model n least squares methodology for two
independent variables (Mayer and Ellersieck 1986) on a combined data base from Gulf Breeze,
FL, Columbia, MO (USGS), Office of Pesticide Programs (OPP), and AQUIRE. Slopes and
intercepts are derived from the equation log y = a + b (log x), where x equals 48-h EC50/96-h
LC50 values for the surrogate test species and y equals 48-h EC50/96-h LC50 values for the
endangered or other species. Chemical groupings by mode of action will also be compared to
taxonomic groupings of species to ascertain any differences in vulnerability. A user manual and
software, based on a Windows platform, will be developed to use the data base as a reference
catalog and for user interaction to derive calculated values based on their input data.
For acute-to-chronic predictive models, both classical and non-classical time-to-event approaches
(Crane et al. 2001, Jones 1964) will be used. The appropriate computer language(s) to combine
all three acute-to-chronic models, linear regression, multifactor probit analysis, and accelerated
life testing (Mayer et al. 1994, Lee et al. 1995, Sun et al. 1995), including life table subroutine,
will be determined. The accelerated life testing model will be the population-based estimation
procedure using an underlying life-time distribution/survival probability distribution (e.g.,
exponential, Weibull, extreme value, log-normal, gamma, or log-gamma models). This approach
will define extra response above natural mortality due to chemical exposure. We will develop
appropriate computer commands using compatible language that will allow for: 1) data
input/transfer from pre-established data bases, 2) selective and continuous processing of data
102
-------
through all or selected combinations of the three models, and 3) printouts of graphics. The final
phase is establishment of a user-friendly Windows version of the ACE software. This will be a
joint effort with the University of Missouri, Columbia, MO, through a cooperative agreement
funded by ORD/NHEERL, OW, and OPPTS.
For selected classes of chemicals, the utility of estimating WQC values based on limited data will
be examined by applying extrapolation techniques to subsets of data for existing WQC. This
exercise will provide insights into the accuracy and margin of error that might be expected by
such extrapolations.
Products
FY01 Report on sensitivity of threatened and endangered fishes to contaminants with
comparisons to that of standard surrogate species.
APM 4A FY02 ICEs for acute toxicity to aquatic organisms (GED).
APM 4B FY02 Time-concentration-effect models for use in predicting chronic toxicity from
acute toxicity data (GED).
APM 4C FY03 Acute to chronic estimation (ACE) user guide and software (GED).
FY06 Refinement of ICEs for acute toxicity to aquatic organisms based on new data.
Benefits of Products
Research products developed under this plan will provide managers with extrapolation methods
and guidance to use when making decisions based on limited data sets. Specifically, 1)
interspecies correlations will allow the user to estimate the acute toxicity for a species having no
data from a common test species having acute data and 2) ACE allows prediction of chronic
toxicity using only acute toxicity data. Estimates for both acute and chronic endpoint predictions
include measures of accuracy and uncertainty.
(Note: Because of resource reductions after the initial development of this plan,
the following project will not be pursued at this time.)
Project Title N3. Assessing the Significance of Non-Equilibrium Conditions on Sediment
Guidelines to Better Predict Field Effects
Project Coordination and Resources (3.0 FTEs: AED-3.0)
Objectives
The majority of U.S. EPA's sediment guidelines are based on the assumption that equilibrium
conditions exist (EPA 2000a,b,c,d). This assumption may be one of the largest weaknesses in
these guidelines. In the laboratory, where many guidelines are developed, this assumption is
103
-------
probably correct. However, in the field, where these guidelines are applied for regulatory
applications, equilibrium conditions may not always be present. Examples of non-equilibrium
conditions include seasonal changes in the benthic environment resulting in changes in AVS
levels (Leonard et al. 1993), storm events and dredging operations under which sediments
become suspended into the water column (Karickhoff and Morris 1985a,b, Calvo et al. 1991,
Simpson et al. 1998, Latimer et al. 1999, Bonnet et al. 2000), and the presence of unusual
binding phases in the water column and benthos (e.g., soot carbon) (Gustafsson and Gschwend,
1997). The occurrence of conditions resulting in non-equilibrium may alter how well guidelines
function and cause under-protective situations as well as introducing unacceptable uncertainty.
Non-equilibrium, along with the importance of dietary uptake, is one of the greatest remaining
sources of uncertainty in the development of criteria for non-bioaccumulative compounds.
The over-all objective of this project is to perform research allowing the Agency to better
understand the importance of non-equilibrium conditions and have greater confidence in the use
of aquatic guidelines under field conditions. The research will focus on three predominant areas
which may result in non-equilibrium: seasonality, sediment suspension, and unusual binding
phases. Seasonality affects nearly every environmental setting and therefore its impacts on the
effectiveness of aquatic guidelines must be understood. Sediment suspension occurs under a
variety of environmental situations ranging from storms to dredging, and signifies a potential
source of toxic chemicals to most coastal systems. Unusual binding phases are a recently
recognized source of variability to the application of aquatic guidelines but nonetheless may
represent a significant source of error to existing regulatory values. Another critical component
of this research will be to complete an evaluation of how well current guidelines predict adverse
effects in field sediments. This information will be very useful in designing studies to assess the
effects of non-equilibrium conditions on guidelines development and application. Although this
work focuses on non-equilibrium in the sediment, the issues explored also have consequences for
the water column, via the sediment-water interface, and for other sources of heterogeneity in the
sediment as well.
Scientific Approach
Seasonality
Seasonality is especially important in relation to metals in sediment. Currently there are no
published EPA guidelines for metals in sediment, but ESGs for metals are under development.
These guidelines use AVS and interstitial water to predict biological effects. These methods
have been demonstrated to be very useful in predicting biological effects in laboratory
experiments and in a limited number of field experiments. The current draft guidelines do not
explicitly consider the effects of seasonality on AVS and metal bioavailability. However, it is
known that AVS varies with season and depth as a function of seasonality (Leonard et al. 1993);
therefore, it is possible that metal bioavailability will correspondingly vary. If metal
bioavailability varies significantly over an annual cycle, decisions based on the comparison of
sediment guidelines with measured chemistry taken at one time of year may not be appropriate
for other times of the year. Research in this section will seek to quantify the magnitude of
fluctuation of metal bioavailability as a function of season.
104
-------
Sediment Suspension
As noted above, sediment suspension is a natural and anthropogenic process that results in the
release of contaminants into the water column. Research has demonstrated this geochemical
phenomena (e.g., Latimer et al. 1999, Cantwell et al. in prep.) but little study has investigated the
biological effects of sediment suspension and contaminant release on benthic organisms. This
section of the project will perform studies to assess the magnitude of contaminant release and
effects under realistic environmental scenarios.
Initially, resuspension and benthic flux experiments will be performed under controlled
(laboratory) conditions in order to determine the significance of specific variables in the
remobilization of contaminants from sediments. In this phase, toxicity tests will be performed to
measure adverse effects. Resuspension of contaminated field sediments will be performed at a
number of energy levels and time scales which are representative of estuarine conditions, with
continuous monitoring of chemical changes to the sediment and overlying water. Additional
experiments will take place resuspending sediments and monitoring longer term (1-6 month)
fluxes and chemical changes to the sediment. Once a better understanding of the variables
influencing stressor availability is gained, field-based experiments will be conducted.
Unusual Binding Phases
Currently, ESGs for metals and organic chemicals are based on the concentrations of the binding
phases AVS and organic carbon, respectively, in the benthos. In recent years, it has been
speculated and observed that other binding phases may also influence the bioavailability of
metals and organic chemicals. For example, organic carbon has been speculated to affect metal
bioavailability beyond the influence of AVS (DiToro et al. 2002). Further, while the
bioavailability of organic chemicals like pesticides and PCBs are well predicted using organic
carbon (EPA 2000b, Burgess et al. 2000), the behavior of PAHs vary widely (Gustafsson and
Gschwend 1997). Unusual binding phases like soot carbon have been proposed to explain these
discrepancies (Gustafsson and Gschwend 1997).
To address the significance of unusual binding phases, research will be conducted to assess the
relative importance of these phases as compared to AVS and organic carbon as currently applied
by Agency guidelines. The draft ESGs for both metals and PAHs contain adjustments for
unusual binding phases but these adjustments are crude and based on limited scientific
information (EPA 2000a,d). Consequently, research will be performed to expand and/or improve
the current adjustments factors to ultimately reduce the variability in regulatory guidelines these
unusual binding phases introduce when used in the field.
Field Validation of Aquatic Guidelines
It is prudent to field validate the existing aquatic guidelines as this provides a way to determine
their effectiveness. One efficient way to do this is to use the guidelines along with our increased
understanding of the effects of contaminants in dynamic and variable sediments to predict the
toxicity of sediments collected in large data bases for which concurrent chemistry and toxicity
data are available, such as the EMAP and NOAA Status and Trends data bases. Numerous
105
-------
attempts have been to match toxicity data to various sediment guidelines, but there are no
published reports of a study which used all of the available guidelines to predict and explain
toxicity in the samples.
In this project, the sediment guidelines (i.e., which are based on the equilibrium partitioning
[EqP] model) will be applied using a toxic unit model to determine whether concentrations of
chemicals measured commonly in sediment monitoring programs (cationic metals, PAHs, PCBs
and other non-PAH narcotic chemicals, and pesticides) appear sufficient to explain observed
toxicity (EPA 2000a,b,c,d). Where they are not, it may be inferred either that unmeasured
chemicals (e.g., ammonia) or measured chemicals not included in the toxic unit model are
contributing to toxicity, or that the EqP model does not provide protective guidelines. Part of
this effort will be to improve the EqP model to correct these potential errors.
Products
As a result of this project, a data set will be generated describing the effects of those factors
resulting in non-equilibrium conditions including seasonality, suspended sediments, and unusual
binding phases. This research recognizes the equilibrium assumption of existing guidelines as a
potential weakness which may compromise their utility if not better understood. Further, we will
perform an assessment of how effective current guidelines are for predicting field effects.
FY02 Report to OW on the effectiveness of ESGs in the prediction of amphipod mortality in
sediments.
FY02 Peer-reviewed journal article on the usefulness of EqP in the prediction of amphipod
mortality in sediments.
FY05 Report to OW on the importance of seasonality and resuspension in predicting the
biological effects of contaminants in sediments.
FY06 Report to OW on the importance of unusual binding phases in predicting the biological
effects of contaminants in sediments .
Benefits of Products
Currently, the usefulness of the Agency's sediment guidelines, which are based on EqP is
potentially limited in regard to field application by our poor understanding of the effects of
seasonality, resuspension, and unusual binding phases. For example, AVS has been used to
develop ESGs for metals, but its application is limited. Better understanding of the biological
effects of contaminants under non-equilibrium conditions would greatly increase the applicability
of the guidelines in the field.
The Office of Water has chosen EqP as the method for developing ESGs, but before these
guidelines can be accepted in the scientific and regulatory community, they must receive further
validation in the field. This research is one step in that validation.
106
-------
Project Title N4. Risks of Heavy Metals to Aquatic Organisms from Multiple Exposure
Routes
Project Coordination and Resources (4.0 FTEs: MED-4.0)
Objectives
A fundamental uncertainty with the use of typical laboratory toxicity tests in assessing risks of
nonbioaccumulative toxicants to aquatic organisms is the failure to account for other routes of
chemical exposure which may occur in natural systems. Of particular concern is exposure to
chemicals via food or by incidental ingestion of contaminated non-food solids. For many
nonbioaccumulative toxicants this failure is arguably of little consequence, but for some
chemicals risk might be significantly underestimated using only water exposures. Conventional
wisdom in aquatic toxicology was that water is the primary exposure route for metals (e.g.,
copper, cadmium, nickel, lead, zinc) to fish and other aquatic organisms (with the exception of
metals such as mercury which form significant amounts of bioaccumulative organometallic
species). Although exposure to metals via the diet was known to produce some level of
bioaccumulation, it was not considered to significantly increase risk relative to water-only
exposures, unless the diet was highly contaminated relative to that in equilibrium with water.
Thus, environmental criteria and other toxicity assessments for metals have focused on
waterborne toxicity, but there has been considerable concern whether this is adequate.
Beginning in the early-1990's a series of dietary toxicity studies were conducted (Woodward et
al. 1994, 1995; Farag et al. 1994) that involved feeding young rainbow trout diets prepared from
invertebrates collected from metal-contaminated rivers, primarily the Clark Fork River (CFR) in
Montana. The Clark Fork watershed is highly contaminated with several metals, with copper
being generally considered to be the metal of greatest concern. Results of these studies showed
that fish fed a diet of pellets prepared from metal-enriched invertebrates showed reduced growth
relative to fish fed similar diets prepared from invertebrates from reference areas, or less
contaminated portions of the CFR. A more recent study from the same laboratory (Farag et al.
1999) reports comparable findings for invertebrates from the Coeur d'Alene watershed in Idaho,
where the primary metals of concern are lead and zinc. The authors of these studies conclude
that the metals in these diets are the cause of the toxicity to rainbow trout. However, these
conclusions conflict with previous studies which have not shown such toxicity from dietary
metals. Additionally, Mount et al. (1994) conducted a laboratory study which fed a live diet of
brine shrimp (Artemia) that were cultured at high metal concentrations to produce nauplii that
were high in metal content. These studies did not indicate that dietary metals caused the degree
of effects noted in studies with the field-collected invertebrate diet. The discrepancies among
these studies is attributed by some to differences in the form of the metal in the diet.
For invertebrates, similar uncertainties and conflicting results exist regarding the importance of
the dietary route of exposure. A variety of studies have demonstrated uptake of metals via
ingestion in various molluscs and crustaceans, but the importance of this uptake relative to
uptake via water and its significance to toxic response generally has not been well determined. In
particular, there are questions about differences in the efficacy of metals taken up via different
routes, about the water exposure a particular dietary exposure should be compared to, and about
107
-------
how separate water and dietary exposures relate to combined exposures. However, Hook and
Fisher (2001) have demonstrated reproduction of freshwater cladocerans and marine copepods to
be reduced at much lower silver concentrations when exposure was via food equilibrated to
various water concentrations, as opposed to directly to water at those same concentrations. Other
unpublished work by Hook and Fisher (2000) has indicated the same is true for mercury, zinc,
and cadmium. However, in similar experiments, Kim et al. (2000) have reported that
reproduction of cladocerans was more severely affected by waterborne cadmium than by dietary
cadmium.
The repercussions of this issue in regulatory programs are large and persistent. The significance
of dietary exposure has not only been a major point of contention in the Superfund assessment of
the CFR, but has infiltrated debates on a number of regulatory issues. These include the
adequacy of ambient WQC for metals, to advisability of assessing waterborne metals on the basis
of dissolved (rather than total recoverable) metal, and the adequacy of EPA's proposed ESGs
(formerly Sediment Quality Criteria) for metals. None of these programs currently considers
dietary exposure to metals as part of assessing risk. Yet, at present, the technical debate is mired
in conflicting and insufficient data with no clear resolution of the biological significance of the
dietary pathway, much less a way of quantifying the risk for incorporation into environmental
regulation. This project will address this fundamental uncertainty in the toxicity model for
metals and will contribute to APG 4.
The general objective of this project will be to assess the importance of considering routes of
exposure other than water in aquatic risk assessments and criteria development for cationic
metals. Initial efforts will address the importance of dietary exposure relative to water and will
include 1) review and synthesis of past and ongoing work of other investigators to better define
the state of knowledge regarding the importance of dietary exposure, and 2) targeted experiments
which confirm critical work and address uncertainties.
Scientific Approach
1. Dietary Metals Effects on Fish.
This will include a series of experiments in which juvenile fish (rainbow trout, fathead minnows,
channel catfish) will be fed diets of live, intact invertebrates that have been enriched with metal
in a variety of ways, including through waterborne exposure of the prey to metals and through
rearing prey in metal-contaminated sediment, both field-collected and artificially spiked. The
oligochaete, Lumbriculus variegatus, has a number of attractive features as a prey species,
including ease of mass culture and tolerance of sediment contamination, and we anticipate using
it as a primary prey model. However, we will also conduct limited experiments with other
invertebrates, such as the midge, since some researchers have speculated that the metal
interaction with chitin may be involved in the observed effects. By using sediments collected
from field locations such as the CFR and the Keweenaw Waterway (Michigan), these
experiments can also address responses to real-world mixtures of metals in addition to
laboratory-prepared mixtures. Exposures in initial experiments will be solely through the diet,
minimizing exposure via water by maintaining high flows of uncontaminated water through
exposure tanks and by not allowing excess prey to remain in the tanks. Concentrations both in
108
-------
the diet and the pore water of the sediments used to contaminate the diet will be documented,
allowing comparison of water concentrations that directly cause toxic effects with the water
concentrations needed to result in toxic concentrations in diet. If these experiments demonstrate
significant toxic effects from diet that might substantially increase risk from water alone,
additional experiments will be conducted with metal exposure both via water and diet to further
evaluate the relative importance of these two routes.
2. Dietary Metals Effects on Zooplankton.
This will include experiments in which zooplankton (freshwater cladocerans and saltwater
copepods) are chronically exposed to metals dissolved in water and incorporated into food, both
separately and combined. Methods will build on those of Hook and Fisher (2001) and Kim et al.
(2000) and will address the reasons for the discrepancies between these studies and the relative
importance of these different routes of exposure under conditions expected in natural systems.
Experiments will be designed to supplement and complement related research in progress
elsewhere.
Products
F Y03 Journal article regarding importance of dietary exposure to chronic metal toxicity to
juvenile fish.
FY04 Journal article regarding importance of dietary exposure to chronic metal toxicity to
cladocerans.
APM 4D FY06 Report evaluating importance of dietary route of exposures to aquatic risk
assessments for metals (MED).
Benefits of Products
The existing controversy over dietary exposure to metals is influencing regulatory decision-
making in several Regions and programs. In addition to the Superfund process and Natural
Resources Damage Assessment (NRDA) litigation on the CFR, similar issues have been raised
regarding the assessment of mining impacts on the Coeur d'Alene River in Idaho. With respect
to State WQS, ORD staff have been contacted by State representatives concerned that shifting
standards from total to dissolved will increase metal loadings and thereby increase risk from
metal toxicity via other pathways such as dietary and sediment exposure. In the consultation of
Region 9 and OW by the Department of Interior on the California Toxics Rule (CTR), the issue
of dietary exposure to metals figured prominently in the draft Biological Opinion (BO), with DOI
citing uncertainty regarding dietary uptake as reason for using total, rather than dissolved, metals
as a measure of compliance. Enhancing understanding of combined waterborne and dietary
exposures to metal mixtures will improve the ability of environmental managers in the EPA and
elsewhere to make informed decisions on the assessment of ecological risk from metals in
aquatic systems.
109
-------
Project Title El. Framework for Development and Application of Population Risk-Based
Criteria for Fish and Wildlife Exposed to Persistent Bioaccumulative Toxicants
Project Coordination and Resources (3.0 FTEs: AED-1.5, MED-1.5)
Objectives
The primary goal of this project is to describe and demonstrate a framework for assessing
ecological risks and developing risk-based WQC criteria for fish and wildlife populations
exposed to PBTs, as required by APG1. Currently, national WQC or contaminated sediment
screening levels are not available for protection offish and wildlife exposed to PBTs. However,
ecological risk assessment methodologies, with PBT doses based on residues in tissues of
organisms, have been under development and extensively discussed for at least a decade (Cook
et al. 1992). Recently, scientific experts have favorably reviewed the state of the models and
methods for risks to fish and wildlife populations associated with early life stage toxicity from
bioaccumulated PBTs. For example, the residue-based, additive toxicity equivalence approach
for dioxin-like chemicals that act through an aryl hydrocarbon receptor (AhR) mediated
mechanism of action was examined in great detail in 1998 and found to be ready for application
in ecological risk assessments (EPA 2001). The fundamental purpose of this project is to insure
that appropriate chemical residue-based toxicity data and models are effectively used, in
conjunction with bioaccumulation and population dynamics models, to determine site-specific
water quality conditions required to sustain populations of aquatic organisms and aquatic-
dependent wildlife. A general risk assessment framework and associated methods will be
developed for all PBTs based on the principles contained in EPA's Guidelines for Ecological
Risk Assessment (EPA 1998). Also, the PBT framework and associated methods will allow
probabilistic and spatially explicit representations of population level risks to the extent possible
and beneficial.
Scientific Approach
This project will function as a framework for a continuing development and application of risk-
based water quality criteria for PBTs. The framework will be based on the conceptual model for
bioaccumulative toxicants (Figure 11). Three major groups of PBTs that must be addressed are
halogenated organics, of which the chlorinated organics are preponderant (Carey et al. 1998);
PAHs which are ubiquitous contaminants of aquatic ecosystems with potential for increased
aquatic life exposures in the future associated with increasing rates of fossil fuel production and
combustion (Neff 1979); and organometallic compounds such as methyl mercury which are a
particular concern for wildlife connected to aquatic food webs (EPA 1997).
Initially, the conceptual model for risk assessments and criteria development involving
determination of safe loadings of PBTs to aquatic systems (Figure 11) will be applied to existing
data and models for chlorinated aromatic chemicals that act through an AhR mediated toxicity
mechanism in vertebrates. This class of chemicals includes polychlorinated biphenyls (PCBs),
polychlorinated dibenzo dioxins (PCDDs), and polychlorinated dibenzo furans (PCDFs) for
which residue-based ELS toxicity data and a mixture toxicity model are available. Thus,
ecological risk assessment tools and approaches will be described and demonstrated for risk
110
-------
based criteria development and application, including use in TMDLs designed to protect
populations of sensitive fish and wildlife species. The association of the historical lake trout
population decline in Lake Ontario with exposure of embryos to AhR agonists such as 2,3,7,8-
TCDD (Cook et al. 1997) will be used as a device to examine the applicability of population
models when used in tandem with residue based toxicity data. The extent to which extrapolation
of lake trout risks to other species involves more than species sensitivity to TCDD and degree of
exposure/bioaccumulation will be examined in the context of toxicokinetic, toxicodynamic,
biochemical, and life history factors.
Data gaps and modeling limitations will be identified and further described as research needs for
development of a general risk assessment capability for all PBTs. This analysis will include an
initial conceptual evaluation of the degree to which similarities and differences in PBT properties
and mechanisms of action will define a balance between generic and chemical specific PBT risk
assessment approaches and models. For example, do the properties of persistence and
bioaccumulation, in combination with principles of population dynamics, indicate that effects on
ELS development and survival are invariably risk determining for PBTs in general? If so, what
are the basic PBT toxicity and exposure data and modeling requirements for ecological risk
assessments? Population matrix modeling will be a primary tool for evaluating vulnerability
differences between species based on differences in life stage sensitivities, exposure profiles, and
reproductive strategies.
Other aquatic stressors research projects involving PBTs are expected to contribute periodically
new or improved risk assessment capabilities which can then be integrated into the PBT
framework under this project. Although the continuing development of the framework will
include incorporation of all relevant new data and models, many new capabilities are expected to
become available as the result of completions of APMs. Examples of expected risk assessment
capability advances, listed in association with the APGs and projects, are:
Methods for developing WQC based on characterization of population-level risks of toxic
chemicals to aquatic life and aquatic-dependent wildlife (APG 3):
Rates of metabolism and improved food chain bioaccumulation models for metabolizable
PBTs(projectB2).
Models for assessing relative risks of multiple stressors to avian populations with large
geographic ranges (project B3).
Improved understanding offish early life stage dosimetry including appearance and
extent of metabolism during development (project B2).
Models for extrapolating chemical toxicity data across exposure conditions, endpoints, life
stages, and species (APG 4):
Methods for inter-site extrapolation of BAFs based on site-specific food chains and
bioavailability conditions (project B2).
Ill
-------
PB-TK model/s for inter-species extrapolation of avian exposure and tissue dosimetry
data consistent with PB-TD models (project B3).
Models for bioaccumulation and metabolism of PAHs by fish during ELSs (project B4).
Approaches for evaluating, at different spatial scales, the cumulative risks from toxic chemicals
on populations of aquatic life and aquatic-dependent wildlife relative to risks from nonchemical
stressors (APG 5):
Methods for assessing spatial and temporal distributions of risks (project B3).
Population models that predict the relative risks of multiple stressors, including toxics
and habitat alteration, to piscivorous birds (project B3).
Methods for assessing PAH risks to feral fish populations with emphasis on vulnerability
offish during early life stages to photo-induced PAH toxicity (project B4).
While integrating new capabilities into the PBT framework, this project will include expansion
and generalization of the models and methods across chemicals, phyla, and effects to the extent
possible and scientifically defensible in order to provide site-specific applicability with minimum
data sets. Scopes of products from this project are therefore subject to the degree of success in
research from the other PBT projects. Much of this work will be pursued in collaboration with
OW efforts to update WQC guidelines, and the specific nature and time line of products will be
developed as part of this collaboration.
Products
FY02 Journal article describing a conceptual model for relating risk-based critical residue values
in fish and wildlife to chemical concentrations in sediment and water (with context of site-
specific risk-based WQC and assessments of ecological risks associated with contaminated
sediments) (Part of APM 1A [GPRA # 167]).
APM 1A (GPRA # 167) FY02 Report on integrated water and sediment quality criteria methods
for assessing site-specific risks of persistent bioaccumulative toxicants to aquatic species (MED).
Benefits of Products
This framework will provide guidance for development and application of WQC for protection
offish and wildlife populations on the basis of PBT residue-based toxicity data. The framework
will allow presently available chemical mass balance models to link site-specific chemical
loading, fate, and transport information to toxicity risks, including population level impacts.
This, and the demonstration of capability for risk assessment of complex mixtures of PBTs
having a common mechanism of toxicity, should allow EPA, the States, and Tribes to more
effectively determine where and to what extent loadings of PBTs to aquatic ecosystems pose
unacceptable ecological risks.
112
-------
(Note: Because of resource reductions after the initial development of this plan,
the following project will not be pursued at this time.)
Project Title B2. Incorporate Chemical Metabolism Rates and Site-specific Bioavailability
into Bioaccumulation Models Structured for Practical Assessments of Risks to Fish and
Wildlife Exposed to PBTs
Project Coordination and Resources (4.0 FTEs: MED-4.0)
Objectives
Ecological risk-based criteria for PBTs require a strong capability to relate chemical residue
based dose-toxicity response data to environmental exposure conditions, and thus to chemical
loading limitations for protection of vulnerable populations offish and wildlife. The present
capability is limited by the amount and quality of data available; significant data gaps (e.g., rates
of metabolism, bioaccumulation in ELSs); inconsistencies in approaches used for predicting
bioaccumulation; and uncertainties associated with extrapolation of bioaccumulation data across
species, life stages, and exposure conditions. This research begins with an objective to develop
the first data base designed to provide a comprehensive set of BAFs and BSAFs for organisms in
a complex food web, coupled with bioavailability and metabolism information, so that methods
for extrapolation of measured BAFs and BSAFs to different ecosystems may be developed and
validated. The ultimate objective is to extend these models to allow directly application to ELSs
offish and wildlife. ELSs are often most sensitive to PBTs, have greatest impact on population
maintenance, and therefore are risk determining. Specific objectives, associated with
development and application of the high quality bioaccumulation data base, are:
Provide a comprehensive conceptual model for relating risk-based critical residue values
in fish and wildlife to chemical concentrations in sediment and water.
Develop a high quality bioaccumulation data base for a four trophic level, mixed
benthic/pelagic food web in southern Lake Michigan.
Provide a master set of BAFs, BSAFs, and associated bioavailability data with an
evaluation of procedures for extrapolation to diverse ecosystems and chemical loading
conditions.
Develop guidelines for sampling and analysis of biota, lipid, sediment, water, and organic
carbon to maximize inter-ecosystem extrapolations and comparisons of bioaccumulation
data.
Determine whole organism based rates of metabolism from the high quality
bioaccumulation data base for PCBs, PCDDs, PCDFs, PAHs, and other PBTs (as data are
available).
113
-------
Refine BAFs and models to incorporate metabolism in the food chain and to reduce
uncertainties associated with species, life stages, and effects end points associated with
ecological risks for PBTs.
Provide a bioaccumulation model for complex mixture of PAHs in fish ELSs suitable for
assessment of risks to feral fish populations to photo-activated toxicity under project B4.
Scientific Approach
Bioaccumulation factors, BSAFs, and food chain models incorporate both generic (non-site
specific) and site specific elements of bioaccumulation science. A conceptual model for
integrating high quality bioaccumulation monitoring data with mechanistically based food chain
bioaccumulation models (e.g., Gobas 1993) will be used for advancing and organizing methods
for linking chemical residue-based effects data to chemical concentrations in water and
sediments; and then for risk based WQC development, sediment remediation evaluations, and
aquatic ecological risk assessments in general. These methods are expected to be an evolutionary
advancement of bioaccumulation methods presently incorporated into EPA's Great Lakes water
quality guidance (EPA 1995b), the methodology for deriving ambient WQC for the protection of
human health (EPA 2000f), and the framework for application of toxicity equivalence
methodology for polychlorinated dioxins, furans and biphenyls in ecological risk assessment
(EPA 2002). The research results will be organized to maximize the accuracy for applying
general bioaccumulation data and models to site-specific assessments with no or minimum data
collection on site is required.
1. Field Data Requirements.
In general, models and tools for extrapolation and/or prediction can be developed only when
adequate experimental data are available. One of the major objectives of this research plan is to
develop field data of the appropriate quality and breadth for improving bioaccumulation models
and tools. Breadth of field data includes the completeness of the measurements on all
components of the food web and its surroundings, and the range of properties associated with the
chemicals of interest. Furthermore, these field measurements must be intrinsically connected so
that data from each component are reflective of the conditions sensed or felt by the other
components. Depending on the ecosystem of interest, measurements over time might be
required. In this research effort, field data will be developed for all analytes using capillary
column gas chromatograph/mass spectrometry (GC/MS) techniques with stable isotopes and MS
resolution of 10,000. The use of analytical techniques with these characteristics will reduce
uncertainties and biases associated with field data, and when used for all environmental
components, will provide data of comparable quality among all components of the food web and
its surroundings.
Initially, a high quality bioaccumulation data set for a four trophic level, mixed benthic/pelagic
food web from southern Lake Michigan will be developed for PCBs, PAHs, PCDDs, PCDFs, and
some chlorinated pesticides. This list of analytes might be expanded after the initial analyses for
other PBTs which have useful properties for development of bioaccumulation models and tools
(e.g., chemicals with large molecular weights, > 600 amu, or useful metabolism rates in fish).
114
-------
This effort will include samples for components of the food web and its surroundings (i.e., phyto-
and zoo-plankton, benthic invertebrates, forage fishes, and piscivorous fishes), sediments, and
water column. In addition, ancillary data such as lipid contents and organic carbon contents will
be measured. For the purposes of evaluating extrapolation procedures, existing data sets will be
used and when necessary, high quality measurements will be performed in this investigation.
2. Determination of Metabolism Rates.
Mechanistic models for predicting chemical residues in aquatic food webs historically have
included a rate constant for the metabolic loss of chemical in organisms (Gobas 1993, Thomann
et al. 1992). Although biotransformation of most all compounds occurs in fishes, common
modeling practices to date set the metabolic loss rate to zero because many PBTs are thought to
have rates of metabolism so slow that they can be considered non-metabolizable (although
metabolism rates have not been measured for nearly all the PBTs). In the absence of measured
metabolism rates, the default modeling assumption of "no metabolism" is used in modeling
exercises. This practice of tacitly assuming that metabolism is not important for PBTs has
largely come about because nearly all the modeling exercises have been performed for PCBs, a
class of chemicals that bioaccumulate to a degree that suggests extremely low rates of
metabolism. Consistent with this assumption, the validation efforts to date have shown that the
food web models have excellent predictive ability for PCBs.
This research effort will evaluate the use of field data to infer and deduce information about
metabolic rates for PBTs. The approach for determining rates of metabolism involves the use of
mechanistic food web models together with high quality field data. With these models and the
high quality field data, the models can be solved for the rate of metabolism for the chemical of
interest. In essence, the difference between the model prediction using no metabolism and the
actual field data is accounted for by the metabolic rate loss parameter if the model parameters are
set to accurately predict bioaccumulation for non-metabolized congeners. This research effort
will define for the approach the data quality requirements, and the range of metabolic rates and
bioaccumulation potential for which the methodology will work.
3. Bioaccumulation Models for Fish ELSs.
If metabolism rates can be effectively determined from field data as proposed, the final step in
bioaccumulation model development for risk assessments involving metabolizable PBTs will be
the development of empirical and mechanistic bioaccumulation models for fish during ELSs.
This is a four step process: 1) develop models for bioaccumulation of PBTs (with varying rates
of metabolism) by female fish, 2) develop maternal transfer models (empirical to PB-TK) to
predict bioaccumulation by embryos, 3) develop post-spawning bioaccumulation models
(empirical to mechanistic) to predict uptake and elimination of PBTs (with varying rates of
metabolism) by fish at different stages of development, and 4) determine inter-species
differences in ELS bioaccumulation and strategies for inter-species extrapolation of ELS
dosimetry data. Step 1 is part of the general bioaccumulation model development and step 2 has
a foundation in field data (e.g., Guiney et al. 1996) and PB-TK models developed (e.g., Nichols
et al. 1997). Step 3 is an essential element of research on risks of PAHs to fish ELSs under
project B4. Thus, over time projects B2 and B4 will have an increasing degree of shared research
115
-------
objectives. Step 4 requires further consideration offish ELS dosimetry in the context of the
mechanism of toxicity associated with the PBT risk being assessed. For example, photo-
activated toxicity of PAHs in larval fish will involve accumulation of the PAHs in specific
tissues for which UV light activation is possible.
4. Inter-Ecosystem Extrapolations.
The research approach for improving inter-ecosystem extrapolations and comparisons of
bioaccumulation data will involve the development of appropriate data sets for a variety of
ecosystem classifications and chemical loading scenarios. With this information,
bioaccumulation models and tools can be tested and evaluated. The emphasis, initially, will be
on testing the effectiveness of simpler extrapolation techniques. Extrapolations and comparisons
of data across ecosystems will involve consideration of differences in the sediment-water
chemical concentration ratios, departure from steady-state distribution of chemical, food web
depth (number of trophic levels), and food web composition (benthic-pelagic). The results of
these comparisons will be used to determine the need for more specific and data intensive
modeling approaches to reduce uncertainty to a level sufficient for accuracy consistent with
capability to perform probabilistic risk assessments.
Products
FY02 Journal article describing the conceptual model for relating risk-based critical residue
values in fish and wildlife to chemical concentrations in sediment and water (with context of
risk-based WQC criteria and contaminated sediment based risk assessments).
FY02 Technical support document and site-specific methods for determination of BAFs
associated with the methodology for deriving ambient WQC for the protection of human health
(MED co-authorship with OW and NCEA).
FY02 Journal article on BSAFs for PCBs, PCDDs, PCDFs, and PAHs associated with the Lake
Michigan food web and applicability to other ecosystems.
FY02 Journal article on validation of the BAF methods used for the methodology for deriving
ambient WQC for the protection of human health.
FY03 Journal article on BAFs for PCBs, PCDDs, PCDFs, and PAHs associated with the Lake
Michigan food web and applicability to other ecosystems.
FY03 Journal article on proof of concept for measurement of rates of metabolism of PCBs,
PCDDs, and PCDFs from high quality food web data.
FY04 Journal article on measurement of rates of metabolism of PAHs from high quality food
web data with toxicokinetic model interpretation.
FY04 Report on application of site-specific BAFs and models in conjunction with risk
assessments requiring extrapolations across species, life stages, and effects endpoints.
116
-------
FY05 Journal article or internal report on bioaccumulation of PAHs by fish during early life
stages of development.
FY06 Journal article on model/s (including PB-TK) for prediction of bioaccumulation by fish
during early life stages of development.
Benefits of Products
The overall benefit is anticipated improvements in bioaccumulation measurements, models, and
site-specific applicability, coupled with conceptual clarification of how to incorporate multi-
media exposure relationships (water, sediment, food chain) into application of chemical residue-
based toxicity data for development of WQC/WQS, TMDLs, or management of contaminated
sediments. A specific benefit will be much improved capability, including guidance for sample
collection and analysis requirements, for extrapolating measured BAFs and BSAFs between
ecosystems. The capability to provide species and chemical specific rates of metabolism for food
chain model predictions of bioaccumulation will be a significant improvement since many PBTs
have reduced bioaccumulation in parts of the food web due to metabolism. Practical and
beneficial combinations of empirical and mechanistic modeling methods will be defined to
increase both ease and accuracy of bioaccumulation predictions for site-specific criteria and risk
assessment applications. Also expected is demonstration of the integration of the basic
bioaccumulation models with toxicokinetic models as required for PBT criteria and risk
assessments involving particular modes of action and critical effect end points.
Project Title B3. Multiple Stressor Risks to Common Loon and Other Piscivorous Bird
Populations (cross-listed in Section 4, Habitat Alteration, Project 4)
Project Coordination and Resources (6.9 FTEs: AED-5.4, MED-1.5)
Objectives
This project has been developed as a demonstration and evaluation of the utility of the research
approaches described in NHEERL's WRS (EPA 2000e). The wildlife strategy describes research
approaches that address key research needs, including improved capabilities for cross-species
extrapolation, prediction of population dynamics in spatially-explicit habitats, assessment of the
relative risk of chemical and non-chemical stressors, and definition of appropriate spatial scales
for wildlife risk assessments. This project is one of several that will be developed to evaluate the
approaches and key hypotheses regarding risks to wildlife species that are described in the
wildlife strategy. In addition to its consistency with these overall objectives, this specific project
was identified because it involved minimal data collection activities and addressed a problem of
immediate concern to the Agency.
The overall objective of this demonstration project is to develop the tools and approaches for
assessing the risks of multiple stressors to populations of piscivorous wildlife, leading to the
development of risk-based criteria. Three major research objectives include:
117
-------
Develop approaches for predicting population-level responses to stressors, and identify
the responses at the individual level that have the greatest influence on population-level
responses (APG 3).
Develop mechanistically-based approaches for extrapolating toxicological data across
wildlife species, media, and individual-level response endpoints (APG 4).
Develop approaches for evaluating the relative risks from chemical and non-chemical
stressors on spatially structured wildlife populations across large areas or regions (APG
5).
The research described here attempts to make advances in each of these research areas through a
single demonstration project designed to develop the tools and approaches necessary to conduct a
multi-tiered assessment of the risks of PBTs, (e.g., mercury) to populations of piscivorous birds
in New England and the upper Midwest. In the process of developing the approach and tools for
conducting the risk assessment, we will also develop a framework for establishing wildlife
criteria using piscivorous birds and Hg as the example. In conjunction with project Bl, the
methods and models developed in this demonstration project will be evaluated for their
applicability to other PBTs and other wildlife species. This demonstration project focuses on
issues starting with the exposure of birds to mercury in the fish (and other dietary components)
they consume, rather than focusing on fate and bioaccumulation within the wholly aquatic
portion of the food web. Many of the issues addressed in the development of population and
spatial models in this project should contribute to the development of more generalized
approaches for assessing risks to wildlife species.
It should also be noted that another component of this project involves the assessment of the
interactive effects of landscape-level habitat alteration and mercury on loons. This project also is
described in Section 4 (Habitat Alteration) because there are significant research issues regarding
habitat alterations, including evaluating the spatial configuration of loon habitat and mercury
impacts in the landscape mosaic and the issue of scaling up from local to regional impact
assessments.
Scientific Approach
Mercury contamination remains a high priority issue for several EPA Program Offices and
Regions because of widespread atmospheric deposition and concerns of accumulation through
aquatic food webs. Although there is evidence of reduced productivity in some piscivorous birds
and widespread reports of wildlife tissue mercury concentrations exceeding levels associated
with adverse effects in controlled studies, it is unclear what impact this has on the viability of
populations of piscivorous wildlife. Also, mercury contamination exists within a patchwork of
other co-occurring stressors to wildlife populations, but the relative risks are poorly understood.
Because mercury bioaccumulates in the aquatic food web, this demonstration project focuses on
the risks of mercury to top level predators (piscivorous birds, in particular, common loons [Gavia
immer]), associated with mercury exposure in the environment. Given the heterogeneous
distribution of stressors (e.g., dietary methyl mercury, habitat degradation, acidification, human
disturbance), the project will attempt to identify the spatial relationships among stressors (i.e.,
118
-------
correlations in distributions), the potential interactions among stressor impacts, and the relative
risks among potential stressors to populations of loons at various spatial scales. The research
will secondarily characterize risks for the belted kingfisher (Ceryle alcyon), another piscivorous
bird that is often the focus of ecological risk assessments because its small body size and high
food ingestion rate lead to estimates of high potential risk.
To improve the process of developing risk-based criteria for mercury protective of loons and
other piscivorous birds, advancements in model and method development and data acquisition
are needed in five major areas. First, the landscape of interest needs to be characterized,
including the spatial and temporal distribution of stressors and available habitat for species of
interest. Second, stressor-response relationships are needed, especially for endpoints related to
survival and fecundity rates. Relationships may be developed empirically from field data or
generated in laboratory tests based on representative exposure scenarios. Third, methods for
interspecies extrapolations of stressor-response relationships for mercury are needed. Fourth,
age-class matrix population models incorporating stressor-response relationships are needed for
loons and other piscivorous birds. Fifth, population dynamics need to be assessed across
heterogeneous landscapes where variable stressor levels and habitat qualities influence the
distribution of populations. Information from the five areas is connected through important
feedback loops (e.g., knowledge of population dynamics captured in the matrix models also
informs the selection of landscapes and endpoints for dose-response testing). Research on
landscape characterization, population model development, and determining and extrapolating
dose-response information will proceed in parallel. The level of refinement needed in the models
and acquired data is a function of the degree of uncertainty acceptable for setting criteria, so the
application of tools will explicitly consider the availability and quality of data for various tiers of
risk assessment.
1. Landscape Characterization.
A demonstration project on risks to loons would focus on landscapes in the upper Midwest and
Northeastern United States. Characterization of these landscapes would include:
Collection of mercury residue distribution in fish and water bodies across the landscape.
This would be based primarily on a synthesis of existing monitoring data from Federal,
State, academic, and non-profit organization sources.
Characterization of habitat quality for species of interest across the landscape. For loons,
key habitat characteristics would include presence of suitable nesting and brood rearing
sites, measures of human disturbance, density or extent of human dwellings and other
activities around lakes, turbidity, and the availability of suitable forage fish supplies.
This information would be synthesized from available monitoring databases and/or aerial
photographs.
Collection of data on the abundance and distribution of loons and other piscivorous birds,
and on juvenile production rates by location. This information would be based on long-
term loon monitoring programs that exist in the upper Midwest and the Northeast.
119
-------
2. Stressor-response Relationships.
Although information currently exists on the toxicity of mercury to several bird species, none of
the tests were conducted in such a way that dose-response relationships can be estimated and
none were conducted with carnivorous species. In fact, testing of avian reproduction effects is
rarely done with the intent of defining dose-response relationships for population-related
endpoints. Development of dose-response relationships would include:
A controlled dosing study to determine dose-response relationships for methyl mercury
effects on survival and fecundity of American kestrels (Falco sparverius). Husbandry
methods for breeding piscivorous birds in captivity generally have not been developed,
but the kestrel has a long history of successful breeding under laboratory test conditions.
This study is currently being conducted in cooperation with USGS Patuxent Wildlife
Research Center and NCEA.
Empirical development, from existing field monitoring data, of stressor-response
relationships for measures of productivity for other types of stressors, including habitat
impairment, human disturbance, and lake acidification. It will be more difficult to show a
relationship of these stressors to adult survival.
3. Inter species Extrapolation.
Extrapolation methods are needed for estimating toxicity in untested species from tested species,
from laboratory tests to free-flying wildlife, and across media. Extrapolation of toxicity
information includes:
Development of PBTK/TD models for methyl mercury in kestrels for predicting effects in
loons and other piscivorous birds. Mercury residue information from studies with
kestrels will be used to develop residue-response relationships sensitive to the duration of
exposure. The PBTK/TD model then will be used to predict the movement and effects of
methyl Hg in other species. This model will be developed through a cooperative project
with USGS (Patuxent) and NCEA.
Empirical extrapolation methods based on a synthesis of existing toxicity databases.
These methods estimate a distribution of sensitivity across tested species, but are poor
predictors of where a specific species would fall in that distribution. This work would
build on existing syntheses projects by including new data from the kestrel studies and
other published data.
4. Population Matrix Models.
Age-class matrix models will be used to organize information on the population dynamics of
loons and other piscivorous birds. Development of matrix models would include:
Integration of life history information including adult and juvenile survival, fecundity,
immigration and emigration rates, and density dependent factors into the model
120
-------
framework. This information largely will be synthesized from existing monitoring
databases and literature. A data system for population modeling parameters will be
evaluated with NCEA potentially as a component of the Wildlife Exposure Handbook
and the NCEA-supported Wildlife Canada Exposure Model. Current monitoring efforts
could also be focused to include measurement of this information.
Development of approaches for integrating individual-level responses in order to
extrapolate and predict population-level effects of anthropogenic stressors. This model
development would be a primary objective of this demonstration project.
Identification of those responses at the individual-level that have the greatest influence on
population-level responses (i.e., elasticity analysis). This analysis would be part of this
demonstration project.
Products
FY03 PBTK/TD model for predicting individual effects of chronic mercury exposure to facilitate
cross species extrapolation of toxicity responses.
APM 3 A (GPRA # 59) FY04 Population models that project the relative risks of multiple
stressors (toxic chemicals, habitat alterations) to piscivorous birds (AED, MED).
APM 5B FY06 Approaches for addressing spatial scale issues in assessing risks of multiple
stressors to wildlife populations in spatially-diverse landscapes (AED, MED).
Also see Section 4 Habitat Alteration, Project 4 for associated products.
Benefits of Products
This demonstration project will directly address APG 3 by providing methods for developing
WQC based on characterization of population-level risks of toxic chemicals to aquatic-dependent
wildlife.
Given the paucity of comparative toxi city data across taxonomic groups of wildlife, the
development of a PBTK/TD model for mercury in birds will improve the capability for
extrapolating chemical toxi city data across endpoints, life stages, and species of wildlife (APG
4).
The focus of this demonstration project, understanding stressor risks of individuals in the context
of effects at the level of populations in spatially-diverse landscapes, will provide approaches for
evaluating the relative and cumulative risks from toxic chemicals and non-chemical stressors on
populations of aquatic-dependent wildlife (APG 5).
Project Title E4. Risks to Fish Populations from PAHs in Natural Systems
Project Coordination and Resources (4.0 FTEs: MED-4.0)
121
-------
Objectives
A major uncertainty in assessing the risks of PBTs to aquatic life is whether application of
laboratory lexicological data adequately reflects complex exposure relationships and interactions
important to responses in natural systems. Chemicals in one class of PBTs, PAHs, have been
found in traditional laboratory tests to have relatively low toxicities, due both to a nonspecific
mechanism of action and to reduced bioaccumulation in some organisms because of metabolic
transformations of these chemicals. However, the toxicities of some PAHs to various aquatic
organisms have been demonstrated to be greatly increased (by orders of magnitude) due to
exposure to UV radiation (Bowling et al. 1983; Cody et al. 1984; Kagan et al. 1984, 1985; Oris
and Giesy 1985, 1987; Newsted and Giesy 1987; Hoist and Giesy 1989; Tilghman Hall and Oris
1991; Huang et al. 1993; Buckler et al. 1994; Ankley et al. 1994, 1995, 1997; Boese et al. 1997;
Erickson et al. 1999). General principles of dosimetry for this enhanced toxicity, based on PAH
accumulation and UV intensity, have been described (Newsted and Giesy 1987; Ankley et al.
1995, 1997;Ericksonetal. 1999).
An analysis of fuel contamination of the clear waters of Lake Tahoe concluded that
photoactivated toxicity posed a significant risk to zooplankton (Oris et al. 1998) and current data
suggest that ELS fish in PAH-contaminated littoral zones of the Great Lakes are at risk (Mount et
al. 2001). However, this risk is uncertain due to several factors, some specifically related to
photo-activated toxicity and some of more concern to PBTs in general: 1) most research to date
has used laboratory UV light sources with spectra different from natural sunlight; 2) both the
intensity and spectra of UV light in natural systems vary spatially and temporally, resulting in
receptor organisms receiving widely varying exposures depending on their life habits and the
properties of the system; 3) PAH exposure can also vary widely within natural systems,
especially between sediments and overlying waters, so that PAH accumulation can also depend
on organism attributes and system properties; 4) the accumulation, and thus the effects, of PAHs
can vary between laboratory and natural systems due to food chain influences and maternal
transfer to young organisms; 5) the relative accumulation of and sensitivity to PAHs of different
life stages are poorly known; and 6) research has usually used individual compounds or simple
mixtures of commercially-obtained PAHs, in contrast to complex mixtures of PAHs occurring in
contaminated natural systems. However, current knowledge of a) the general levels of PAH
contamination and the magnitude of UV light in natural systems and b) the sensitivity of many
organisms to photo-activated toxicity in the laboratory indicate a potential for major impacts due
to these interacting factors.
Risks from PAHs will depend on a complex interaction among light, chemical, receptor
organisms, and system characteristics. ELS fish are one group of organisms at potential risk.
ELS fish potentially can have significant PAH accumulation due to maternal transfer to eggs,
exposure of eggs to water and sediment, accumulation after hatching from food and water
(especially when closely linked to sediments), and the absence of metabolic pathways which
limit PAH accumulation in older fish. Provided that the fish behavior or the system attributes
result in significant exposure to light, ELS fish might be particularly susceptible to UV-activated
PAH toxicity because of their small size (i.e., large surface to volume ratios and short penetration
distances) and lack of protective pigmentation and gill coverings. Past research has shown early
life stage fish to be susceptible to photo-activated toxicity (Bowling et al. 1983; Oris and Giesy
122
-------
1985, 1987; Tilghman Hall and Oris 1991; Buckler et al. 1994), but use of this research to assess
risk in natural systems is limited by the factors discussed above.
The overall objective of this research project is to develop more comprehensive and accurate
assessments of the risks of PAHs to early-life-stages offish that address the influence of UV
radiation and exposure relationships in natural systems. To this end, PAH accumulation in ELS
fish will be evaluated both in the laboratory and in natural systems; effects of accumulated PAH
in ELS fish will be evaluated both under laboratory UV light and natural sunlight for both
individual PAHs and PAH mixtures from contaminated systems; and likely risks based on these
observed effects and on fish habits will be estimated for natural systems. This project will not
address the incorporation of effects on ELS fish into fish population models, but will provide
information important to population-level assessment methods such as those developed in
projects B1 and B3 and also will examine the correlation of expected ELS effects with fish
community health indices. Extensions of this work to general PBT assessments will be
addressed in project B1.
Scientific Approach
Assessing the risks of PAHs to ELS fish requires consideration of several factors. First,
environmental PAH concentrations must be characterized, including chemical partitioning
information important to bioavailability. Second, UV radiation exposures must be evaluated
relative to fish behavior and environmental conditions. Third, the accumulation of PAHs must
be estimated as a function offish age and environmental exposure concentrations, including
consideration of maternal transfer, uptake by both egg and fry, multiple routes of exposure, and
age-dependent metabolism. Fourth, good dosimetry relationships are needed which link
mortality and growth to PAH accumulation in the fish, to the varying levels of UV radiation they
receive, and to their age. Fifth, the combined effects of complex mixtures of PAH must be
evaluated, including the effects of those PAHs which are not measured. This project will
develop needed capabilities in these areas and will assess the likely risk of PAHs to ELS fish in
selected natural systems.
1. ELS Accumulation of PAHs.
Meaningful assessment of the effects of PAHs (and other PBTs) on ELS fish requires that
toxicity test results be applied with consideration of the importance of PAH accumulation to
organism response and how accumulation might differ from that in natural systems. Such
consideration should include not only accumulation after hatch from water and food, but also
uptake from water during egg incubation and transfer from the maternal fish to the egg. The
importance of these different routes of exposure depends on how rapidly PAHs accumulate
during different life stages and at what life stages UV exposure is most important. Laboratory
experiments will examine the accumulation of PAHs in ELS fish as a function of when exposure
starts (parent fish, egg, newly hatched fry) to determine if PAHs originating from parents or
taken up by eggs might significantly contribute to risks. Uptake relationships will be monitored
as fish grow to determine at what age and to what extent metabolic transformation of PAHs
becomes important for regulating accumulation. Accumulation also will be measured from ELS
fish collected from and exposed in natural systems to determine its relationship to environmental
123
-------
PAH concentrations and to compare this with laboratory results. Results of these efforts will
contribute to development of ELS fish bioaccumulation models for PBTs under project B2.
2. Dosimetry Relationships for PAH Toxicity to ELS Fish.
Effective assessment of the risk of PAHs to fish populations requires improved understanding of
the relationship of PAH accumulation, UV intensity and spectrum, and fish age to various effects
in ELS fish. Dosimetry relationships will be developed in the laboratory for selected individual
PAHs. Experimental methods will be developed to allow continuous exposure of ELS fish to
various levels of both PAHs and UV radiation. Tests will be conducted to address the effects of
PAH levels, the duration/nature of prior exposure to PAHs, fish age, and UV intensity, duration,
and variability.
A major impediment to PAH risk assessments is the complex mixtures that occur in most PAH-
contaminated natural systems. Even if such mixtures could be completely characterized, the
potencies of most constituents would be unknown and the combined toxicity could be
significantly underestimated. The unaccounted-for toxicity could be addressed with an index
which compares total measured photo-activated toxicity from a complex mixture to the expected
toxicity based on measured accumulations of a subset of mixture constituents for which photo-
activated potencies are known. Such an index could be based on the oligochaete Lumbriculus
variegatus, for which phototoxicity bioassays are relatively simple and precise, and can involve
either water or sediment exposures. The phototoxicity of complex PAH mixtures from selected
sites and of reference PAHs to bothZ. variegatus and ELS fish will be tested to determine the
utility of such an index.
3. Assessment of PAH effects in Natural Systems and the Use of Photo-activated Potency as an
Ecological Indicator.
The effects of PAHs on ELS fish in selected natural systems will be evaluated using floating
experimental platforms containing chambers in which fish can be exposed to site water and to
various levels of UV radiation through selective wavelength screening. Such systems will
provide a direct test of the legitimacy of concerns about photo-activated toxicity for natural levels
of PAHs and UV radiation, and provide data for testing the applicability of laboratory data and
methods to the prediction of these effects.
MED scientists will also work cooperatively with researchers from the Great Lakes Ecological
Indicators (GLEI) project to evaluate photo-activated PAH toxicity potency in sediments as an
ecological indicator. GLEI staff will collect sediments samples from a stratified random set of
coastal wetlands and tributaries to the Great Lakes. Samples will be shipped to MED and
evaluated for photo-activated PAH toxicity potency using the oligochaete L. variegatus as
described above. The measured potency of PAHs in these sediments will be compared to
chemical characterization being conducted by the University of Minnesota cooperators, and
indices of UV transmission in the water column, to estimate relative risk from this pathway at
each site, using the photo-activated PAH toxicity models developed by MED in this and previous
research. Additionally, estimated risk from photo-activated toxicants with be examined relative
124
-------
to measures of biological condition being collected for each site by GLEI staff to evaluate
measured photo-activated PAH potency as an ecological indicator.
Products
FY03 Journal article on bioaccumulation of PAHs in ELS fish.
FY04 Journal article on dosimetry of photoactivated PAH toxicity to ELS fish.
FY05 Journal article on effects of complex PAH mixtures and ambient sunlight to ELS fish in
natural systems.
APM 5 A FY05 Report regarding assessment of risks to aquatic organisms from combined
exposure to PAH mixtures and UV light in natural systems (MED).
Benefits of Products
Research to date has indicated that current PAH risk assessments do not adequately handle such
issues as photo-activated toxicity, incompletely-characterized complex mixtures, and ELS
exposure issues, so that risk might be greatly underestimated. However, there has been no direct
evidence that risk is significant for the combinations of PAH contamination, UV light, and
biological receptors in natural systems. This work will determine whether there should be
concern regarding such toxicity for fish populations in PAH-contaminated areas. If this is the
case, this would greatly impact the development of WQC and sediment guidelines for PAHs, as
well as hazard assessments for other classes of PBTs which might require similar considerations.
This work will also help improve general PBT exposure assessment methodologies, such as fish
ELS accumulation models.
Gap Analysis
The overall goal of this research program is to develop procedures for risk-based criteria for toxic
chemicals. The Critical Path subsection described the research needed to better describe the risks
of toxic chemicals to aquatic life and aquatic-dependent wildlife populations and communities.
The proposed research projects address these needs. Although the research program will expand
and improve the approaches for developing criteria, this gap analysis reviews the needs (e.g.,
models, test methods, toxicity data, methods for assessing temporal and spatial distributions of
exposure, and role of life history on species vulnerability) that will not be completely addressed
by this program, but are required for development of a complete risk assessment capability for
toxic chemicals.
Developing the scientific basis for risk-based criteria will require the overall conceptual
framework and better modeling tools described in the research projects outlined above. While
the focus of this research program is on the development of methods and models, there will be
parameters required by these methods and models that can not be met adequately with existing
data sources alone. New data will be generated for specific questions in a couple of our
demonstration projects; however, a major gap for achieving our overall goal of developing risk-
125
-------
based criteria for chemicals will be the generation of new data for several steps on the critical
path. For example, risk-based criteria will require data on dose-response relationships for
various chemicals exposed to aquatic life and wildlife, but very few existing chronic exposure
toxicity tests generate dose-response relationships, and we are not proposing new testing. Also,
we are developing the basis for PBTK modeling that will be applied to a demonstration project
for extrapolating toxicity estimates to other species, but more will be needed for generating
physiological and metabolic information for broader applications of PBTK models. Finally, our
demonstrations projects will use existing life history information, such as survival and fecundity
rates, for setting parameters in population models, but scientifically-defensible population
models for most aquatic and wildlife species will require new life history data that will not be the
focus of this research program. These needs for new data for various models will be discussed
further below.
Specific gaps are related to key components of the conceptual models for nonbioaccumulative
and bioaccumulative toxicants (Figures 10 and 11). They are organized under the following
headings: bioavailability, dosimetry and bioaccumulation, toxicity, and population models.
Bioavailability
One gap relates to the need to better understand exposure, which is outside ofNHEERL 's
mission:
1. Exposure models are needed that can estimate physical transport and fate of chemicals. This
includes persistence/degradation, partitioning, and especially chemical forms reaching aquatic
organisms. Improved models will be needed for predicting chemical forms of metals and
activities of organic chemicals in water, bulk sediments, and associated pore waters.
Two gaps relate to the needs for extending our modeling approaches to other chemicals and
exposure scenarios:
2. Although this plan will, initially, develop models for ammonia and metals, gaps will remain
for applying these models to other nonbioaccumulating chemicals. Organic chemicals with low
bioaccumulation potential are an important class of chemicals which pose ecological risks
through a variety of mechanisms of toxicity including disruption of endocrine functions. Present
and future research in and outside ofNHEERL on the effects of these chemicals requires models
and data for exposure and bioavailability appropriate for linking the toxicology research results
into criteria development and ecological risk assessment procedures.
3. Seasonal and spatial variability in bioavailabilities of chemicals, coupled with life stage
changes, population movements, and mechanisms for avoidance of exposure, produce
complications for determination of species vulnerabilities. This gap is presently being addressed
only in the proposed research for loons (project B3).
126
-------
Dosimetry and Bioaccumulation
Two gaps relate to the development ofPBTK models for improving understanding of dose-
residue-response relationships and extrapolation among species or life stages:
1. Although the mercury-loon project (B3) will develop a PBTK/TD model in kestrels for use in
estimating the toxicity of mercury to piscivorous birds, the toxicokinetics of mercury are not
representative of other PBTs. This will limit the general applicability of these models to other
chemicals.
2. Another gap for developing avian PBTK models is the paucity of information on avian
physiology and metabolism for setting parameters in the models.
Several gaps relate to improvements needed in bioaccumulation information:
3. A method for determination of rates of metabolism for PBTs is needed in order to allow
accurate predictions of bioaccumulation with aquatic food chain models. If the research
proposed to demonstrate the feasibility of determining rates of metabolism from field data fills
this important risk assessment need, additional studies will be required to provide the metabolism
data for all PBTs of concern in diverse food webs.
4. Existing BAFs, BSAFs, and food chain models are based on whole adult organisms and thus
may not be sufficient when dose to ELSs and/or specific tissues must be evaluated. ELS
dosimetry-based BAFs and PB-TK models are needed to fill this gap. The metabolism rate gap
extends to bioaccumulation of PBTs like the PAHs in embryo-larval stages offish with potential
vulnerability to photo-induced toxicity. Proposed photo-induced PAH toxicity research provides
a beginning for filling this large gap. Additional research will be required to establish a general
ELS toxicity risk assessment capability equal to that available for juvenile or adult organisms
exposed to a wide variety of chemicals.
5. Very few bioaccumulation data sets are of sufficient quality to validate the uses of BAFs and
BSAFs, especially when extrapolated across species and/or ecosystems. The intent of proposed
NHEERL research is to maximize the capability for extrapolation of BAFs and BSAFs for PBTs.
However, consistent data gathering efforts sponsored by Offices interested in the application of
BAFs and BSAFs for criteria development and risk assessment are needed in order to provide
measures of uncertainly involved in such extrapolations as well as bioaccumulation model
predictions performed without calibration (site-specific measurement of BAFs and BSAFs).
6. Comprehensive and toxicity hazard assessment compatible BAFs, BSAFs, and food chain
models are needed to meet the requirements of joint action toxicity models such as for TCDD
toxicity equivalence or photo-induced PAH toxicity.
127
-------
Toxicity
Two gaps relate to the needs for toxicity data:
1. Critical to the idea of "risk-based" criteria is the need to understand the relationship between
stressor intensity and responses affecting survival and reproduction (i.e., dose-response
relationships). Although acute toxicity tests for aquatic life and wildlife usually produce dose-
response relationships, most chronic exposure test and reproduction tests are designed to estimate
effects thresholds (e.g., no observed adverse effect levels [NOAELs]) rather than describe dose-
response relationships. Although some fish reproduction tests have been conducted with
sufficient number and spacing of concentrations that dose-response relationships could be
estimated, almost no avian reproduction testing has been done to quantify dose-response
relationships. This is a large data gap, and OW presently may have no mechanism for generation
of new toxicity data.
2. Because of their hydrophobicity, a majority of PBTs accumulate in the benthos of freshwater
and marine systems. From benthic environments, PBTs transfer to higher tropic levels where
adverse effects to wildlife may occur. Despite the acknowledged effects of PBTs at higher tropic
levels, a gap exists in our knowledge of whether or not this class of chemicals also causes
significantly adverse effects to benthic organisms, populations, and communities.
Bioaccumulation data for benthic organisms from contaminated sites around the country indicate
exposure is occurring but we are not certain of the effects. To insure benthic ecosystems and
resources are fully protected this gap should be addressed.
Other gaps relate to the extrapolability of toxicity data among species and endpoints:
3. Compared to aquatic organisms, the database of chronic exposure tests of effects on avian and
mammalian survival and reproduction is much more limited, with much of the testing of non-
pesticide chemicals done without standardized procedures. Consequently, the database for
making interspecies extrapolations for wildlife is insufficient to significantly improve methods
based on comparative toxicity relationships among species. This research program will be
adding little new data for improving species sensitivity relationships for wildlife. Also, previous
analyses of avian toxicity databases demonstrated no relationship between acute and chronic
toxicity measurements for birds.
4. Virtually unexplored are the TKTD determinants for interspecies and inter-effect
extrapolations of potency ratios required for PBT mixture toxicity risk assessment using a toxic
units model approach, such as the additive TCDD toxicity equivalence model.
Several gaps relate to the need for improved toxicity models and databases:
5. Although models for predicting effects from fluctuating exposures models are proposed for
development for metals and ammonia, a gap exists in this capability for organic chemicals.
6. Residue-based toxicity data bases need to be advanced and evaluated for applicability to
aquatic ecological risk assessment requirements for PBTs.
128
-------
7. Complex, multi-stressor models, such as required for photo-induced PAH toxicity to fish
during embryo-larval stage of development, need to be developed and applied to determine the
magnitude of ecological risks which are presently highly uncertain.
Other important gaps:
8. In some cases, populations of chronically exposed aquatic organisms have demonstrated an
evolved tolerance or genetic resistance to toxicity through chronic exposure. The potentially
enormous and irreversible consequences of rapid evolutionary change suggest the importance of
better understanding, predicting, and managing these anthropogenic impacts on aquatic and
wildlife populations. In addition, technological advances now permit an identification of the
genetic changes that may provide the key to understanding the mechanisms by which populations
and species adapt or become extinct.
9. A gap exists in the development of biological indicators that can lead to diagnosis of
developing toxicity problems in aquatic ecosystems before population impacts are observable.
For example, commonly measured biochemical effects, such as P450 enzyme induction, are
sometimes used as diagnostic indicators of exposure to specific categories of chemicals.
However, these measurements are of limited use because their relationships to organismal, much
less population-level, risks are not well understood.
10. NHEERL toxicology research plans organized outside of Goal 2 should consider criteria and
aquatic ecological risk assessment needs in order to prevent gaps for the utilization of the
research products to meet aquatic stressor data and methods requirements.
Population Models
Several gaps relate to the needs for information for developing population models:
1. The greatest limitation to the application of population matrix models is the paucity of high
quality data on mortality and fecundity rates, and our limited understanding of density-dependent
feedbacks and other ways populations compensate for losses due to stressors. Generalized
population models are helpful for identifying information needs, but scenario-specific population
models often will be limited by the lack of data about the populations of interest. Specific case
studies have been chosen because of their relative wealth of population parameter information
and the potential to identify the types and formats of data required for population models when
applied to other species/region/stressor(s) combinations. The collection of basic life history
information on fish and wildlife species is a gap not addressed by this effort outside of a few
species covered under case studies.
2. Although life history information is being gathered by State and Federal resource agencies and
others, it is not always in an adequate form for use in a population model. In addition to
proposing methods for population modeling, we need to work with resource agencies to
influence the format of life history information being gathered to improve its utility in risk
assessment.
129
-------
3. Complex mixtures of PBTs are the norm, so interspecies differences in potency, as well as in
bioaccumulation, for individual chemicals in the mixture must be factored into population level
risk predictions.
4. Is absence of overt mortality, even for ELSs, an adequate effects end point for preventing
population declines caused by PBTs or non-PBTs? If not, how do we determine what is adequate
for aquatic invertebrates, fish, amphibians, or avian and mammalian wildlife? The presently
proposed toxic chemicals research will fill this fundamental gap only to the extent that specific
chemicals are intensively used as foci for development of models and risk assessment methods.
5. A national WQC methodology for different classes of PBTs needs definition, through use of a
generic population model (or a suite of generic population models) of species characteristics, life
stages, and toxicity effects that are most predictive of risks to populations, regardless of site
conditions. This information will fill a gap which presently limits development of population
level based, chemical-specific criteria, as well as determination of site-specific model and data
requirements for application of the criteria.
Two gaps relate to the need for demonstrating and verifying the usefulness of population models:
6. Population models need to be developed and applied through case studies to explicitly
demonstrate risk assessment requirements for prediction of adverse population impacts as a result
of PBT toxicity-caused reductions in survival of aquatic organisms.
7. The proposed approaches in this research plan rely heavily on the accuracy of population
models. The utility of these approaches must be evaluated through field verifications. More
broadly, it must be better understood how different model types and levels of complexities are
necessary to achieve the desired reduction in uncertainties for specific assessment needs. Only
one project in this program addresses this need.
One gap relates to the importance of spatial scales for aquatic life assessments:
8. Spatially explicit population models are often required for assessment of risks to aquatic-
dependent wildlife populations that function in landscapes that integrate many lakes, wetlands,
and streams. Because PBTs tend to distribute widely, if not uniformly, for long periods of time
in aquatic habitats, uncertainty exists for when, and the extent to which, spatially explicit
population models are requisite for assessment of risks to populations of aquatic organisms.
References
Aldenberg, T., Slob, W. 1993. Confidence limits for hazardous concentrations based on
logistically distributed NOEC toxicity data. Ecotoxicol. Environ. Saf. 25:48-63.
Ankley, G.T., Collyard, S.A., Monson, P.D., Kosian, P. A. 1994. Influence of ultraviolet light on
the toxicity of sediments contaminated with polycyclic aromatic hydrocarbons. Environ. Toxicol.
Chem. 13:1791-1796.
130
-------
Ankley, G.T., Erickson, R.J., Phipps, G.L., Mattson, V.R., Kosian, P.A., Sheedy, B.R., Cox, J.S.
1995. Effects of light intensity on the phototoxicity of fluoranthene to a benthic
macroinvertebrate. Environ. Sci. Technol. 29:2828-2833.
Ankley, G.T., Erickson, R.J., Sheedy, B.R., Kosian, P.A., Mattson, V.R., Cox, J.S. 1997.
Evaluation of models for predicting the phototoxic potency of polycyclic aromatic hydrocarbons.
Aquat. Toxicol. 37:37-50.
ASTM. 1988. Standard practice for conducting acute toxicity tests with fishes,
macroinvertebrates, and amphibians. Designation: E 729-88. In 1988 Annual Book of ASTM
Standards, Vol. 11.04. American Society for Testing and Materials, Philadelphia, PA, pp. 304-
322.
Baker, J.L., Barefoot, A.C., Beasley, L.E., Burns, L., Caulkins, P., Clark, J., Feulner, R.L., Giesy,
J.P., Graney, R.L., Griggs, R., Jacoby, H., Laskowski, D., Maciorowski, A., Hihaich, E., Nelson,
H., Parrish, R., Siefert, R.E., Solomon, K.R., van der Schalie, W. 1994. Aquatic Dialogue
Group: Pesticide Risk Assessment and Mitigation. Society of Environmental Toxicology and
Chemistry, Pensacola, FL.
Boese, B.L., Lamberson, J.O., Swartz, R.C., Ozretich, RJ. 1997. Photoinduced toxicity of
fluoranthene to seven marine benthic crustaceans. Arch. Environ. Contain. Toxicol. 32:389-393.
Bonnet, C., Balbut, M., Ferard, J.-F., Martel, L., Garric, J. 2000. Assessing the potential toxicity
of resuspended sediment. Environ. Toxicol. Chem. 19:1290-1296.
Bowling, J.W., Leversee, G.J., Landrum, P.F., Giesy, J.P. 1983. Acute mortality of anthracene-
contaminated fish exposed to sunlight. Aquat. Toxicol. 3:79-90.
Breck, I.E. 1988. Relationships among models for acute toxicity effects: applications to
fluctuating concentrations. Environ. Toxicol. Chem. 7:775-778.
Buckler, D.R., Kemble, N.E., Echols, K.R., Mount, D.R., Tillitt, D.E. 1994. Photoactivated
toxicity of PAH to endangered fishes and standard laboratory test species. Abstracts, 15th Annual
Meeting, Society of Environmental Toxicology and Chemistry, Denver, CO, October 30-
November 3, 1994.
Burgess, R.M., Ryba, S.A., Cantwell, M.G. 2000. Importance of organic carbon quantity on the
variation of koc in marine sediments. Toxicol. Environ. Chem. 77:9-29.
Calvo, C., Donazzolo, R., Guidi, F., Orio, A. A. 1991. Heavy metal pollution studies by
resuspension experiments in Venice Lagoon. Water Res. 25:1295-1302.
Cantwell, M.G., Burgess, R.M., Kester, D.R. 2002. Mobilization and phase partitioning of metals
from anoxic estuarine sediments during periods of simulated resuspension. (In preparation.)
131
-------
Carey, I, Cook, P., Giesy, 1, Hodson, P., Muir, D., Owens, W., Solomon, K., eds. 1998.
Ecotoxicological Risk Assessment of Chlorinated Organic Chemicals. SET AC Press, Pensacola,
FL. 375 pp.
Cody, T.E., Radike, M.I, Warshawsky, D. 1984. The phototoxicity of benzy[cc]pyrene in the
green alga Selenastrum capricornutum. Environ. Res. 35:122-132.
Cook, P.M.., Carlson, A.R., Lee, H, II. 1992. Tissue residue approach. In Sediment Classification
Methods Compendium, Chapter 7. EPA 823-R-92-006. EPA, Office of Water, Washington, DC.
Cook, P.M.., Zabel, E.W., Peterson, R.E. 1997. The TCDD toxicity equivalence approach for
characterizing risks for early life stage mortality in trout. In Rolland, R., Gilbertson, M.,
Peterson, R., eds., Chemically-Induced Alterations in the Functional Development and
Reproduction of Fishes. SETAC Press, Pensacola, FL, pp. 9-27.
Crane, M., Newman, M.C., Chapman, P.F., and Fenlon J. 2001. Risk Assessment with Time to
Event Models. Lewis Publishers, Boca Raton, FL. 175 pp.
Di Toro, D.M., Hansen, D.J., McGrath, J.A, Berry, W.J. 2002. Predicting the toxicity of metals
in sediments using organic carbon normalized SEM and AVS. (In preparation.)
Erickson, R., Kleiner, C., Fiandt, J., Highland, T. 1989. Feasibility of predicting the effects of
fluctuating concentrations on aquatic organisms and possible application to water quality criteria.
EPA-600/X-89-307. EPA, Duluth, MN.
Erickson, R.J., Ankley, G.T., DeFoe, D.L., Kosian, P.A., Makynen, E.A. 1999. Additive toxicity
of binary mixtures of phototoxic polycyclic aromatic hydrocarbons to the oligochaete
Lumbriculusvariegatus. Toxicol. Appl. Pharmacol. 154:97-105.
EPA. 1973. Water quality criteria 1972. A report of the Committee on Water Quality Criteria,
Environmental Studies Board, National Academy of Sciences, National Academy of
Engineering. EPA-R3-73-033. Washington, DC.
EPA. 1980. Guidelines for deriving water quality criteria for the protection of aquatic life and its
uses. Fed. Reg. 45:79341-79350.
EPA. 1991. Technical support document for water quality-based control. EPA/505/2-90-001.
Office of Water, Washington, DC.
EPA. 1992. Framework for ecological risk assessment. EPA/630/R-92/001. Risk Assessment
Forum, Washington, DC.
EPA. 1994. Interim guidance on determination and use of water-effect ratios for metals. EPA
823-B-94-001. Office of Water, Washington, DC.
EPA. 1995a. Final water quality guidance for the Great Lakes system, final rule. Fed. Reg. 60.
132
-------
EPA. 1995b. Great Lakes water quality initiative technical support document for the procedure to
determine bioaccumulation factors. EPA-820-B-95-005. Office of Water. NTIS PB95187290.
EPA. 1997. Mercury study report to Congress-Volume VI: an ecological assessment for
anthropogenic mercury emissions in the United States. EPA-452/R-97-008. Office of Air Quality
Planning and Standards and Office of Research and Development, Washington, DC.
EPA. 1998. Guidelines for ecological risk assessment. EPA/630/R-95/002F. Risk Assessment
Forum, Washington, DC.
EPA. 1999. 1999 update of ambient water quality criteria for ammonia. Office of Water,
Washington, DC.
EPA. 2000a. Equilibrium partitioning sediment guidelines (ESGs) for the protection of benthic
organisms: metal mixtures (cadmium, copper, lead, nickel, silver, and zinc). EPA-822-R-00-005.
Office of Science and Technology and Office of Research and Development, Washington, DC.
EPA. 2000b. Technical basis for the derivation of equilibrium partitioning sediment guidelines
(ESGs) for the protection of benthic organisms: nonionic organics. EPA-822-R-00-001. Office of
Science and Technology and Office of Research and Development, Washington, DC.
EPA. 2000c. Methods for the derivation of site-specific equilibrium partitioning sediment
guidelines (ESGs) for the protection of benthic organisms: nonionic organics.
EPA-822-R-00-002. Office of Science and Technology and Office of Research and
Development, Washington, DC.
EPA. 2000d. Equilibrium partitioning sediment guidelines (ESGs) for the protection of benthic
organisms: PAH mixtures. Draft report. Office of Science and Technology and Office of
Research and Development, Washington, DC.
EPA. 2000e. Wildlife research strategy. NHEERL/ORD. September.
EPA. 2000f. Methodology for deriving ambient water quality criteria for the protection of human
health. EPA-822-B-00-004. Office of Science and Technology, Washington, DC.
EPA. 2001. Workshop report on the application of 2,3,7,8-TCDD toxicity equivalence factors to
fish and wildlife. Risk Assessment Forum, Washington, DC.
EPA. 2002. Framework for application of the toxicity equivalence methodology for
poly chlorinated dioxins, furans, and biphenyls in ecological risk assessments. Office of Research
and Development, Washington, DC. (in prep.)
Farag, A.M., Boese, C.J., Woodward, D.F., Bergman, H.L. 1994. Physiological changes and
tissue metal accumulation in rainbow trout exposed to foodborne and waterborne metals.
Environ. Toxicol. Chem. 13:2021-2029.
133
-------
Farag, A.M., Woodward, D.F., Brumbaugh, W., Goldstein, J.N., MacConnell, E., Hogstrand, C.
1999. Dietary effects of metals-contaminated invertebrates from the Coeur d'Alene River, Idaho,
on cutthroat trout. Trans. Amer. Fish. Soc. (in press).
Gobas, F.A.P.C. 1993. A model for predicting the bioaccumulation of hydrophobic organic
chemicals in aquatic food-webs: application to Lake Ontario. Ecol. Model. 69:1-17.
Guiney, P.D., Cook, P.M., Casselman, J.M., Fitzsimons, J.D., Simonin, H.A., Zabel, E.W.,
Peterson, R.E. 1996. Assessment of 2,3,7,8-tetrachlorodibenzo-p-dioxin induced sac fry
mortality in lake trout (Salvelinus namaycush) from different regions of the Great Lakes. Can. J.
Fish. Aquat. Sci. 53:2080-2092.
Gustafsson, O., Gschwend, P.M. 1997. Soot as a strong partition medium for poly cyclic aromatic
hydrocarbons in aquatic systems. In Eganhouse, R.P., ed., Molecular Markers in Environmental
Geochemistry. American Chemical Society, Orlando, FL, pp. 365-381.
Hall, L.W. Jr., Scott, M.C., Killen, W.D. 1998. Ecological risk assessment of copper and
cadmium in surface waters of Chesapeake Bay watershed. Environ. Toxicol. Chem. 17:1172-
1189.
Hickie, B.E., McCarty, L.S., Dixon, D.G. 1995. A residue-based toxicokinetic model for pulse-
exposure toxicity in aquatic systems. Environ. Toxicol. Chem. 14:2187-2197.
Hoist, L.L., Giesey, J.P. 1989. Chronic effects of the photoenhanced toxicity of anthracene on
Daphnia magna reproduction. Environ. Toxicol. Chem. 8:933-942.
Hook, S.E., Fisher, N.S. 2000. Sublethal toxicity of metals to marine copepods following dietary
exposure: a consistent pattern emerges. Abstracts, 21st Annual Meeting of Society of
Environmental Toxicology and Chemistry.
Hook, S.E., Fisher, N.S. 2001. Sublethal effects of silver in zooplankton: importance of exposure
pathways and implications for toxicity testing. Environ. Toxicol. Chem. 20:568-574.
Huang, X.D., Dixon, D.G., Greenberg, B.M. 1993. Impacts of UV radiation and
photomodification on the toxicity of PAHs to the higher plant Lemna gibba (duckweed).
Environ. Toxicol. Chem. 12:1067-1077.
Jones, J.R.E. 1964. Fish and River Pollution. Butterworths, London. 203 pp.
Kagan, J., Kagan, P. A., Buhse, H.E. Jr. 1984. Light-dependent toxicity of cc-terthienyl and
anthracene toward late embryonic stages ofRanapipiens. J. Chem. Ecol. 10:1115-1122.
Kagan, J., Kagan, E.D., Kagan, I.A., Kagan, P.A., Quigley, S. 1985. The phototoxicity of non-
carcinogenic polycyclic aromatic hydrocarbons in aquatic organisms. Chemosphere 14:1829-
1834.
134
-------
Karickhoff, S.W., Morris, K.R. 1985a. Sorption dynamics of hydrophobic pollutants in sediment
suspensions. Environ. Toxicol. Chem. 4:469-479.
Karickhoff, S.W., Morris, K.R. 1985b. Impact of tubificid oligochaetes on pollutant transport in
bottom sediments. Environ. Sci. Technol. 19:51-56.
Kim, S.D., Metzler, D.M., Cha, O.K., Allen, H.E. 2000. Is food-borne cadmium more
bioavailable than sorbed/water-borne cadmium? Abstracts, 21st Annual Meeting of Society of
Environmental Toxicology and Chemistry.
Kooijman, S.A.L.M. 1987. A safety factor for LC50 values allowing for differences in sensitivity
among species. Water Res. 21:269-276.
Latimer, J.S., Davis, W.R., Keith, DJ. 1999. Mobilization of PAHs and PCBs from in-place
contaminated marine sediments during simulated resuspension events. Estuar. Coast. Shelf Sci.
49:577-595.
Lee, G., Ellersieck, M.R., Krause, G.F., Mayer, F.L. 1995. Predicting chronic toxicity of
chemicals to fishes from acute toxicity test data: multifactor probit analysis. Environ. Toxicol.
Chem. 14:345-349.
Leonard, E.N., Mattson, V.R., Benoit, D.A., Hoke, R., Ankley, G.T. 1993. Seasonal variation of
acid volatile sulfide in sediment cores from three northeastern Minnesota lakes. Hydrobiologia
271:87-95.
Mancini, J.L. 1983. A method for calculating effects on aquatic organisms of time-varying
concentrations. Water Res. 17:1355-1361.
Mayer, F.L. 1987. Acute toxicity handbook of chemicals to estuarine organisms. EPA/600/8-
87/017. EPA, Gulf Breeze, FL. 274 pp.
Mayer, F.L., Ellersieck, M.R. 1986. Manual of acute toxicity: interpretation and data base for 410
chemicals and 66 species of freshwater animals. Resource Publ. 160. U.S. Fish and Wildlife
Service, Washington, DC. 579 pp.
Mayer, F.L., Krause, G.F., Buckler, D.R., Ellersieck, M.R., Lee, G. 1994. Predicting chronic
lethality of chemicals to fishes from acute toxicity test data: concept and linear regression.
Environ. Toxicol. Chem. 13:671-678.
Meyer, J.S., Gulley, D.D., Goodrich, M.S., Szmania, D.C., Brooks, A.S. 1995. Modeling toxicity
due to intermittent exposure of rainbow trout and common shiners to monochloramine. Environ.
Toxicol. Chem. 14:165-175.
Mount, D.R., Earth, A.K., Garrison, T.D., Barten, K.A., Hockett, J.R. 1994. Dietary and
waterborne exposure of rainbow trout (Oncorhynchus mykiss) to copper, cadmium, lead, and zinc
using a live diet. Environ. Toxicol. Chem. 13:2031-2041.
135
-------
Mount, D.R., Diamond, S.A., Erickson, R.J., Simcik, M.F., Swackhamer, D.L. 2001. Linking
exposure and dosimetry to risk from photo-activated toxicity of PAHs. Presentation, 22nd
Annual Meeting of the Society of Environmental Toxicology and Chemistry, Baltimore,
Maryland.
Neff, J.M. 1979. Polycyclic Aromatic Hydrocarbons in the Aquatic Environment: Sources, Fates,
and Biological Effects. Applied Science Publishers, London.
Newman, M.C. 1995. Quantitative Methods in Aquatic Ecotoxicology. Lewis Publishers, Boca
Raton.
Newman, M.C., Ownby, D.R., Laurent, C.A.M., Powell, D.C., Christensen, T.R.L., Lerberg,
S.B., Anderson, B.-A. 2000. Applying species-sensitivity distributions in ecological risk
assessment: assumptions of distribution type and sufficient numbers of species. Environ. Toxicol.
Chem. 19:508-515.
Newsted, J.L., Giesy, J.P. 1987. Predictive models for photoinduced acute toxicity of polycyclic
aromatic hydrocarbons to Daphnia magna, Strauss (cladocera, Crustacea). Environ. Toxicol.
Chem. 6:445-461.
Nichols, J.W., Jensen, K.M., Tietge, I.E., Johnson, R.D. 1997. A physiologically-based
toxicokinetic model for maternal transfer of 2,3,7,8-tetrachlorodibenzo-p-dioxin in brook trout
(Salvelinusfontinalis). Environ. Toxicol. Chem. 17:2422-2434.
Oris, J.T., Giesy, J.P. 1985. The photoenhanced toxicity of anthracene to juvenile sunfish
(Lepomis spp.). Aquat. Toxicol. 6:133-146.
Oris, J.T., Giesy, J.P. 1987. The photo-induced toxicity of polycyclic aromatic hydrocarbons to
larvae of the fathead minnow (Pimephalespromelas). Chemosphere 16:1395-1404.
Oris, J.T., Hatch, A.C., Weinstein, J.E., Findlay, R.H., McGinn, P.J., Diamond, S.A., Garrett, R.,
Jackson, W., Burton, G.A., Allen, B. 1998. Toxicity of ambient levels of motorized watercraft
emissions to fish and zooplankton in Lake Tahoe, California/Nevada, USA. Presentation, 8th
Annual Meeting of the European Society of Environmental Toxicology and Chemistry,
Bordeaux, France.
Simpson, S.L., Apte, S.C., Batley, G.E. 1998. Effect of short-term resuspension events on trace
metal speciation in polluted anoxic sediments. Environ. Sci. Technol. 32:620-625.
Solomon, K.R., Baker, D.B., Richards, R.P., Dixon, K.R., Klaine, S.J., LaPoint, T.W., Kendall,
R.J., Weiskopf, C.P., Giddings, J.M., Giesy, J.P., Hall, L.W. Jr., Williams, W.M. 1996.
Ecological risk assessment of atrazine in North American surface waters. Environ. Toxicol.
Chem. 15:31-76.
136
-------
Stephan, C.E., Mount, D.I., Hansen, D.J., Gentile, J.H., Chapman, G.A., Brungs, W.E. 1985.
Guidelines for deriving numerical national water quality criteria for the protection of aquatic
organisms and their uses. PB85-227049. NTIS, Springfield, VA.
Sun, K., Krause, G.F., Mayer, F.L., Ellersieck, M.R., Basu, A.P. 1995. Predicting chronic
lethality of chemicals to fishes from acute toxicity data: theory of accelerated life testing.
Environ. Toxicol. Chem. 14:1745-1752.
Thomann, R.V., Connolly, J.P., Parkerton, T.F. 1992. An equilibrium model of organic chemical
accumulation in aquatic food webs with sediment interaction. Environ. Toxicol. Chem. 17:615-
629.
Tilghman Hall, A., Oris, J.T. 1991. Anthracene reduces reproductive potential and is maternally
transferred during long-term exposures to fathead minnow. Aquat. Toxicol. 19:249-264.
Wagner, C., Lokke, H. 1991. Estimation of ecotoxicological protection levels from NOEC
toxicity data. Water Res. 25:1237-1242.
Weber, C.I. 1993. Methods for measuring the acute toxicity of effluents to freshwater and marine
organisms. EPA/600/4-90/027F. EPA, Environmental Monitoring Systems Laboratory,
Cincinnati, OH.
Woodward, D.F., Brumbaugh, W.G., DeLonay, A.J., Little, E.E., Smith, C.E. 1994. Effects on
rainbow trout fry of a metals-contaminated diet of benthic invertebrates from the Clark Fork
River, Montana. Trans. Amer. Fish. Soc. 123:51-62.
Woodward, D.F., Farag, A.M., Bergman, H.L., DeLonay, A.J., Little, E.E., Smith, C.E., Barrows,
F.T. 1995. Metals-contaminated benthic invertebrates in the Clark Fork River, Montana: effects
on age-0 brown trout and rainbow trout. Can. J. Fish. Aquat. Sci. 52:1994-2004.
137
-------
Section 8.
Implementation Plan for Diagnostics Research
Problem
States list surface waters as impaired on 305(b) reports or 303(d) listings based on one or more of
three types of criteria: biological criteria (narrative or numeric), chemical criteria, or physical
attributes (e.g., habitat quality assessments). When impairment is determined based on
biological criteria (26% of impairment decisions), States are faced with the problem of
diagnosing the cause of impairment before plans can be made to reduce the loading of pollutants
through the TMDL process (40 CFR Ch.l, Part 130; U.S. EPA 1991;
http ://www. epa. gov/owow/tmdl/). The nation-wide scope of this problem is enormous;
approximately 21,000 water bodies have been designated as impaired; or 44% of stream or river
miles, 49% of lakes, reservoirs, and ponds; 98% of Great Lakes shoreline waters; and 42% of
estuaries (EPA 2000a).
To improve overall efficiency of the TMDL process and to coordinate remediation activities,
diagnosis of the cause of impairment is needed not only at the scale of individual water bodies,
but also at the watershed scale. Unified Watershed Assessments, as specified in the Clean Water
Action Plan (EPA 1998a) are needed at the watershed scale to identify aquatic systems for
restoration actions (EPA 1999, Federal Register 2000, http://www.epa. gov/owow/uwa).
Overall vision
This plan for diagnostics research provides a comprehensive and integrated approach for problem
solving. It is primarily responsive to EPA's regulatory and management needs, in particular the
need for research related to the TMDL program, but also supports needs defined under
Superfund, National Pollution Discharge Elimination System (NPDES), site remediations, and
other relevant activities (e.g., FWS Natural Resource Damage Assessments). It provides a
conceptual framework to determine current and future research priorities coordinated across
NHEERL, and also discusses implementation in the context of research and expertise provided
by other ORD Laboratories and the broader scientific community. Outputs from diagnostics
research will be incorporated into a series of decision-support systems or modules for use by
clients (EPA Regions, States, Tribes, and Program Offices).
NHEERL's diagnostic research focuses on the need to diagnose causes of biological impairment
within an integrated framework linking watersheds with receiving water bodies to support the
TMDL process and other regulatory programs. All stressors (habitat alteration, nutrients,
suspended and bedded sediments, and toxic chemicals) will be considered under diagnostics
research; however, greater emphasis may be placed on an individual stressor, combinations of
stressors, and/or modes of action according to the prevailing problems or issues of a habitat,
water body, ecosystem, region, or the nation as a whole. The starting point for diagnostic
research is the need to respond to reports of biological impairment, nonattainment of aquatic life
use, and other indications of adverse effects (e.g., toxicity). Initial assessments also can record
evidence of multiple potential causes of impairment and conflicting lines of evidence that might
138
-------
complicate a diagnosis. Thus, the endpoint for the diagnostic process includes both the definition
of the primary causes of impairment as well as the allocation of observed effects among multiple
potential stressors, and the assessment of potential interactive effects among stressors.
To narrow the number of realistic stressors of concern, an approach based on the Toxicity
Identification Evaluation (TIE) procedures will be developed. In the TIE process, toxic
chemicals are first considered in broad classes. As the evaluation proceeds, the focus moves
towards specific chemicals. In this way, large numbers of insignificant chemicals are excluded
from further evaluation. For example in a sediment TIE, sediment may be classified as toxic due
to organic chemicals, then narrowed to pesticides, and finally to DDT. Analogously, in
diagnosing causes of impairment, an approach will be developed which starts with broad stressor
classes (i.e., habitat alteration, nutrients, suspended and bedded sediments and toxic chemicals),
then unimportant stressors will be disregarded, and ultimately, a specific stressor(s) will be
selected as the cause of impairment.
In developing this plan, we considered and evaluated the States' implementation stages from
monitoring through diagnosis to restoration. Implementation stages were then linked to
associated uncertainties, research needs, and desired research products. From these efforts,
APGs, their accompanying APMs, and the critical path for diagnostic research were developed,
and are presented in the next two subsections. The evolution of a combined TMDL/Restoration
Path from the current parallel paths for State/Tribal assessments, TMDL, and watershed
restoration processes is described below.
Goals
There are four primary goals for this diagnostics research:
Provide a framework for interpreting cause-and-effeet relationships, including:
Conceptual ecosystem models based on appropriate mechanisms of action that can
be used to improve the accuracy of impairment decisions;
Conceptual models to define ecosystem and watershed natural conditions and
driving factors to use as a basis to quantify degree of impairment and to set
restoration goals; and
Classification frameworks that explain variation in the response of individuals,
populations, communities, and ecosystems at regional, watershed, water body, and
habitat scales.
Develop single-stressor diagnostic methods and models to determine the primary source
of biological impairment of aquatic ecosystems.
Develop methods and models to allocate causality among multiple stressors and/or to
diagnose interactions among them.
139
-------
Develop methods and models capable of forecasting causality to evaluate the ecological
benefits of source reductions, to investigate stressor interactions, and to assess the gains
and losses realized by various alternatives for restoration and remediation.
Ancillary goals are to improve the state-of-the-science of monitoring and assessment in support
of diagnostic methods, and to provide clients with diagnostic tools in user-friendly interfaces.
Tools with different levels of accuracy and sophistication are needed within the program
depending on cost-benefit ratios of decision making. Accordingly, the tools currently presented
in this research plan range from simple screening tools (watershed classification schemes) to
those of intermediate complexity (e.g., development of diagnostic community-scale indicators) to
those of even greater complexity (e.g., use of linked mass-balance and food web models for Lake
Michigan regional case study). Decision-support systems will be developed to incorporate all of
these features as a format for technical transfer to ORD's clients. Conceptual model
development will provide a general framework for decision-support systems, which will then be
regionalized based on classification systems developed to explain differences in system behavior
(e.g., stressor-response relationships). Tools for diagnosing both single-stressor impacts and
multiple stressor interactions will be piloted using regional case studies. These pilots will then
be incorporated as example applications into decision-support systems. Ultimately, the decision-
support systems will be linked to tools developed by NRMRL, forecasting not only future
impacts based on no action, but also the results of alternative remediation scenarios.
Annual Performance Goals have been derived by defining five implementation stages that the
States must go through between monitoring and diagnosis of the causes of impairment
(Appendix 1). Within each implementation phase, tasks that the States need to perform are
defined, along with their associated uncertainties. These defined tasks are then used to derive
related NHEERL research needs. Finally, research and technical transfer products are linked
with these tasks, uncertainties, and research areas, and research and technical transfer products
are identified as APMs associated with each APG. The time line for implementation of APMs is
shown below, with APMs grouped by APGs.
APG 1 FY03 (GPRA #16) Provide the scientific foundation and information management
scheme for the 303(d) listing process including a classification framework for surface waters,
watersheds, and regions to guide problem formulation.
APM 1A FY02 Conceptual framework for both single and multiple stressors including a
consideration of cross-scale issues (AED, MED).
APM IB FY03 (GPRA # 202) Classification frameworks for geographic regions and at
the watershed, water body and habitat scale (MED, GED).
APG 2 FY05 Provide first generation diagnostic methods, including stressor identification (SI)
methods, for causal linkage of observed major classes of single stressors and biological
indicators to stressors in freshwater and marine systems; scale the methods to States and
watershed organizations.
APM 2A FY03 Guidance on whole sediment TIE procedures (MED, AED).
140
-------
APM 2B F Y05 Application of coastal watershed and estuarine/lacustuary classification
schemes to predict probability of impairment based on Great Lakes and Gulf of Mexico
regional case studies (GED, MED).
APM 2C FY05 Guidance on and user-friendly interfaces for derivation of diagnostic
indicators for individual stressors (MED, AED).
APG 3 FY07 Provide diagnostic methods and technical support documents for determining the
relative significance of multiple stressors in 303(d) listed waters.
APM 3 A FY02 Case studies of multivariate approaches to community data analysis to
apportion cause among stressors. (AED, MED).
APM 3B FY06 Simulation of key stressor interactions with generic ecosystem models
using sensitivity analysis to define the range of stressors and stressor combinations under
which nonadditive interactive effects will occur (MED, AED).
APM 3C FY07 Decision-support system(s), including forecasting of future cause-effect
relationships (MED).
Critical Path
The relationship of the diagnostic APGs to each other and to the research of other ORD
laboratories (NERL, NRMRL) is defined in the critical path diagram (Figure 12). The steps in
the critical path are described as follows:
Step 1. Develop a conceptual framework (APG 1 FY03).
This APG includes the development of conceptual models illustrating stressor-response
relationships for single and multiple stressors and development of appropriate classification
frameworks at the habitat, water body, watershed, and regional scales. Development of
hierarchical classification frameworks involves the determination of which types of habitats,
water bodies, watersheds, and regions are expected to behave similarly in response to a given
level of stressor or loading. Thus, classification helps establish regional, watershed, or habitat-
specific criteria and the range across which model extrapolations (including empirical stressor-
response curves) are appropriate. At this stage, the nature of significant interactions among
stressors also will be defined based on the expected modes of action.
Step 2. Development of single-stressor methods and models (APG 2 FY05).
Methods and models for diagnosis of the predominant source of impairment from single stressors
are needed. Significant input is required at this stage from other research areas (Sections 4-7).
Step 3. Development of multiple-stressor methods and models (APG 3 FY07).
141
-------
Conceptual/
Classification
Framework
(APG 1)
02
Single-s tress or
M e th o d s+M o dels
(APG 2)
Multi-stressor
M eth o d s+M o d el s
(APG 3)
Forecasting
(APG 3)
TMDL
(NERL)
Restoration
(NRML)
Figure 12. Critical path (flow of APGs) for diagnostics research.
c
o
'i
o
u
I
.3
u
o
.ts
c
o
142
-------
Methods and models for single stressors are combined and refined to diagnose multiple sources
of impairment. The latter stage includes the development of tools both to allocate cause among
multiple additive stressors and to diagnose significant interactive effects among stressors. The
first stage in diagnosing significant stressor interactions will involve the use of generic ecosystem
models to perform sensitivity analyses to determine the probability of observing significant
interactions among stressor classes over realistic ranges of loadings or stressor levels (e.g.,
Bartell et al. 1984, Mitsch and Reeder 1991, Hanratty and Stay 1994, EPA 2000b). Ultimately
these tools will be incorporated into a decision-support system.
Step 4. Develop forecasting approaches.
This step builds upon the development of multi-stressor methods and models to include
forecasting techniques to project the response of aquatic ecosystems to load reductions and/or
watershed restoration activities into the future. Forecasting methods will be particularly
important in protecting large, complex, unique resources (e.g., Great Lakes, Gulf of Mexico,
Chesapeake Bay) for which costs of restoration are large, interactions are involved, and lag times
between an event and the eventual system response must be taken into consideration.
Development of forecasting techniques will also allow NHEERL to be proactive in defining
potential shifts in causes of impairment, and to anticipate future threats to the environment.
Activities in this area will be coordinated with NRMRL.
Outputs from this research path will require collaborative efforts with NERL to develop
improved loading models for TMDLs that include components predicting biological responses,
and the development of appropriate exposure metrics to improve monitoring designs. In
addition, classification frameworks and other tools developed here will be coordinated with
research on prioritization of watershed restoration activities, prediction of recovery paths, and
assessment of the success of remediation actions currently under way within NRMRL.
Methods and models developed under diagnostics research also will feed into the diagnostic logic
flow sequence described in the EPA SI document for analysis of data for a weight-of-evidence
approach (EPA 2000c). Potential points of influence on the SI overall process are presented in
Figure 13.
In Figure 14, the research products (APMs) from the critical path are connected to the
State/Tribal implementation stages for both the TMDL and watershed restoration processes for
impaired surface waters, and then merged into an integrated process. The first APG,
development of a conceptual framework, feeds directly into the problem formulation aspect of
the diagnostic process and provides the basis for developing a decision-support system for the
diagnostic process. The second APG connects directly to diagnosing the primary cause of
impairment. The third APG provides research products which support three areas of the
combined TMDL/Watershed Restoration path: Diagnostic and Condition-based Monitoring and
Assessment, Allocation of Causes and Interactions among Multiple Stressors, and Confirmation
of Diagnosis with Uncertainty Evaluation. The fourth goal supports diagnostic model
development for forecasting and predictive approaches for the evaluation of remediation options.
143
-------
Conceptual
Models
Classification
Singfc /
Stressor ,'
Models /
Eliminate Sources
Of Impairment
How Many
Remaining?
More Than One
or Unclear
Causes Subject
to Diagnosis?
Multipk /
Stressor /
Models /
Cause(s)
Diagnosed?
Stressor /
Interactions /
Strength
Of Evidence
Analysis
Effects X No
Real?
A Locate Cause
Among Multiple
Stressors
Sensitivity /
Analysis /
Sufficient
Evidence
Report Results
Figure 13. A logic for characterizing the causes of ecological injuries at specific sites.
Modified from Figure 4-1 in SI document (EPA 2000c) to show potential inputs from aquatic
stressors diagnostics research.
144
-------
Current State.iRegional
Restoration Path
Current StateJRegional
TMDL Path
Monitoring and Assessment
of Condition
Support
to States
State listing process for
i mpai r merit
I
Evaluation of impairment
decision
Unified Watershed
Assessment
303(d) listings
I
TMDLs
Watershed Restoration
Plans
Proposed Combined
TMDLJRestoration Path
Diagnostic and
Condition-based
Monitoring and Assessment
Problem Formulation
Diagnosis of pri mary cause
of i mpairment
Allocation of causes and
interactions among multiple
sires sors
Conf ir mati on of diagnosis
vjth uncertai nty ewal uati on
Forecast! ng f or eval uati on of
remediation options
TMDLs, Renredation
Options for Classes of
Systems
HHEERL Research Products
Provide framework for incorporating diagnostic
element into routine monitoring and assessment
a ctivities s up po rti ng state 303(b) and 3Q3(d) I tstJn g
processes.
Provide conceptuaI frarnework and information
m an age me nt s che me f or p rob lem f o rmu lation i n 303(d)
Iteting process, including classification framework for
wate rs he ds a nd co asta I wet an ds f o r differ ent sea les
APIvte 1A1B, SI guidance documents
decision- tree guiding use of toote
Provide diagnostic methods and modete for each of the
major d asses of aquatic stress ors.
APIvte & guidance from other ASF groups on regionally
specific criteria for toxic chemicals and nontoxics
Provide diagnostic methods and models for multiple
stress ors to all ow a Ho cation of cause and evaluate
stress or i ntera ctio re
Provide methods for confirmation of diagnosis
including monitoring designs for assessment of
restoration success and uncertainty anatjeis
APIvte, SI guidance documents
Develop diagnostic modeIs which incorporate
forecasting or predictke approaches
Figure 14. Relationship between current stages of State/Tribal assessment, TMDL and watershed restoration planning processes, and
proposed combined path.
145
-------
Research Projects
The research projects proposed here establish a conceptual framework to guide implementation
of diagnostics, provide case studies to develop and test methods and models for both single-and
multiple-stressor scenarios, assess the likelihood of multiple stressor interactions, and establish
the structure for a decision-support system. Opportunities for interlaboratory collaboration on
this research are shown in Appendix 2.
Project Title 1. Conceptual Model Development and Information Management Framework
Project Coordination and Resources (6.5 FTEs: FY02: AED, GED, MED [total = 6.0]; FY03:
MED-0.25, AED-0.25 [total = 0.5])
Objectives
The goal of this project is to support the problem formulation stage in diagnostics (EPA 1996).
The main objective is to develop conceptual models describing stressor-response relationships
within ecosystems, including potential interactions among multiple stressors across all scales
relevant to setting a protective TMDL (EPA 1996). These conceptual models will then provide
the basis for creating a national database on nontoxic aquatic stressor-response relationships and
for improving information management systems in support of 303(d) assessment activities.
Scientific Approach
Conceptual model development will focus on the effects of habitat alteration, nutrients,
suspended and bedded sediments, and toxic chemicals on appropriate endpoints (individuals,
populations, communities, ecosystems) across spatial scales (habitats, water body, watershed,
region) relevant to setting a protective TMDL. This research will be coordinated across all of the
Ecology Divisions and across all five research areas (Sections 4-7). Priorities for refining
conceptual models for single and multiple stressors at the habitat scale will be established by
examining the relative frequency of stressor X or stressor combination X; ,X2... with habitat type
Y combinations in the OW 303(d) listing database (EPA 2001,
http://www. epa.gov/owow/tmdl/trcksy s.html). A determination of impairment may be controlled
by different factors as the scale of impact gets larger. Therefore, conceptual models will examine
the cross-scale interactions (habitat <=> water-body <=> watershed <=>region) that must be
understood to determine the appropriate scale at which a protective TMDL must be established.
In addition, the interactions among predominant stressors will be included as appropriate, based
on expected mechanisms of action. Conceptual models will be developed through one or more
cross-divisional workshops.
Two types of information management frameworks will be developed in support of the 303(d)
listing process and diagnosis of the causes of impairments. First, a framework for supplying
existing geospatial information for the problem formulation stage will be established. To
identify potential causes of impairment, this framework will build upon database networks
currently under development or refinement by OW, including geospatial databases incorporated
within the Better Assessment Science Interacting Point and Nonpoint Source (BASINS)
146
-------
modeling support system developed by EPA's OW/Office of Science and Technology (OST)
(http://www.epa.gov/ost/basins/. EPA 1998b), STORAGE and RETRIEVAL database
(STORET), the Water Quality Standards Database, the TMDL tracking database, and Watershed
Assessment Tracking and Environmental Results Systems (WATERS). In addition,
incorporation of toxicity data into EPA's STORET database will be coordinated with this effort.
In support of diagnostic efforts at the watershed scale, NHEERL will continue to collaborate with
EPA's Office of Environmental Information (OEI) and USGS Mapping Division (EROS Data
Center) and Water Division in their work to produce a seamless nationwide geospatial database
of watershed boundaries and associated hydrological derivatives [NED-H (now EDNA,
Elevation Derivatives for Natural Applications), see
http://edcntsl2.cr.usgs.gov/ned-h/index.html. Protocols for deriving watershed boundaries have
been developed through an interagency task force coordinated by the Federal Geographic Data
Committee (FGDC) and Advisory Committee on Water Information (ACWI), thus ensuring
consistency across Federal agencies (see Federal Standards for Delineation of Hydrologic Unit
Boundaries, 06/12/01 Draft; http://www.ftw.nrcs.usda.gov/huc_data.html ). The USGS 8-digit
hydrologic unit codes (HUCs) are being divided into finer units (10- and 12-digit HUCs) that are
generally consistent with watershed boundaries for integrated drainages, and with boundaries for
internal drainages to moderately-sized water bodies. The smallest of these units (12-digit HUCs)
correspond to the scale of watersheds associated with wadeable streams. NHEERL is supporting
development of GIS tools for automated watershed delineation using digital elevation models
(OEMs), and hydrologic correction of existing OEMs to ensure consistency between synthetic
streamlines and mapped hydrography. Thus, attributes coded using the National Hydrography
Database for streams (the successor to EPAs Reach 3 stream files) will be consistent with the
Nationwide Watershed Boundary Dataset under development. Regional case studies (project 3)
will provide an opportunity to use the Nationwide Watershed Boundary Database under
development to demonstrate its usefulness in watershed-scale monitoring designs, assessments,
diagnosis, and management across an integrated series of watershed scales.
Second, a nonpoint-stressor analog to the Ecological Toxicity Database (ECOTOX)
(http ://www.epa.gov/ecotox/. Hunter et al. 1990) will be developed that will contain information
on stressor-response relationships for nontoxics, stratified by an appropriate classification
framework (project 2). ECOTOX is a source for locating single chemical toxicity data for
aquatic life, terrestrial plants and wildlife. ECOTOX integrates three toxicology effects
databases: AQUIRE, terrestrial plants (PHYTOTOX), and terrestrial wildlife (TERRETOX).
Toxicity test results and related testing information for any individual chemical from laboratory
and field aquatic toxicity tests are extracted from the literature and recorded. Lethal, sublethal,
and bioconcentration effects are recorded for freshwater and marine organisms, along with
pertinent information on laboratory or field test conditions. The current database structure could
be adapted readily to store information on stressor-response relationships for non-toxics. These
data would ultimately support the development of regional and/or national criteria for nutrients
and suspended and bedded sediments.
147
-------
Products
APM 1A FY02 Conceptual framework including consideration of both single and multiple
stressors and cross-scale issues (AED, MED).
Database structure to support problem formulation in the diagnostic process (303(d) listings) and
Nation-wide database for stressor-response relationships for non-point source stressors (all
Ecology Divisions, FY03).
Benefit of Products
One of the difficulties in diagnosing the causes of impairment is the lack of an adequate
information framework to support problem formulation. The APMs for this research area
provide a conceptual framework for describing stressor-response relationships and an
information framework for providing geospatial and toxicity data tailored to diagnostic
applications (e.g., methods and models). Development of the geospatial database support system
will be coordinated with the OW/OST because NHEERL will be adding to their BASINS
modeling support system. These APMs will provide State, Regional, and Tribal authorities with
critical and essential tools, which are currently unavailable, for starting the diagnosis process on
an impairment problem. Ultimately, these information management tools can be incorporated
into decision-support systems.
Project Title 2. Classification Framework
Project Coordination and Resources (10.6 FTEs: FY03: AED-0.4, GED-2.6, MED-5.0 [total =
8.0]; FY04: GED-2.6 [total = 2.6])
Objectives
Integrated hierarchical classification schemes will be developed at the scale of habitats, water
bodies, watersheds, and regions to identify systems that are expected to respond similarly to
aquatic stressors (see Sections 4-7). For example, estuaries with longer retention times are more
susceptible to the effects of nutrient loading (Palter and Dettman 1999). The relative impact of
suspended and bedded sediments via sedimentation and physical habitat alteration versus
turbidity also will depend on retention time. Even the effect of toxic chemicals can be expected
to vary systematically depending on physico-chemical characteristics of water bodies and
sediments such as organic carbon, acid-volatile sulfides, suspended solids, and hardness
(Hamelink et al. 1994, Bergman and Dorward-King 1997).
Scientific Approach
The central question that must be answered to determine if a classification system will be useful
for diagnosis is, "does grouping of systems by class simplify the problem of determining the
cause of the observed ecological effects which are equated with an impaired condition of a water
body?" We propose to answer this question by developing classification systems that are keyed
to the different levels of a nested spatial hierarchy that proceeds as follows: habitat, water body,
148
-------
watershed, and region. Some classification schemes already exist for systems at each of the
above levels of organization (Cowardin et al. 1979, Omernik 1987, McKee et al. 1992, Brinson
1993, Maxwell et al. 1995, Frissell et al. 1986, Rosgen 1996, Detenbeck et al. 2000). To be
useful in diagnoses, classification systems must be based on differences in the spectrum of
forcing functions that result in differences in the behavior of systems among classes, (e.g., fluvial
versus lagoonal geomorphology of a water body). The key to the viability of a classification
system at any of these hierarchical levels is that the classes identified behave differently under
the influence of the stressor of concern. Once classes have been identified based on existing or
new classification systems, an initial screening of the stressor-response data at all four levels of
organization will determine if research should proceed further on a specific stressor/class
combination.
Existing classification frameworks and necessary elements of an integrated classification strategy
will be reviewed by a work group consisting of representatives from all Ecology Divisions, from
each of the Aquatic Stressors research areas, and from experts on classification at each of the
scales of interest. Representation will be requested from other ORD Laboratories and Centers,
other Federal agencies, EPA Program Offices, and non-governmental organizations as
appropriate. Logical collaborators on this task include the Landscape Sciences Branch atNERL-
Las Vegas, the Watershed Restoration planning group in NRMRL, OW [(OWOW), Office of
Science and Technology (OST)], USGS (under NAWQA), FWS, NOAA, and the Nature
Conservancy. Recent work on classification approaches within or outside of NUEERL is
summarized in Table 4.
The classification workgroup will work towards the following goals:
Identification of key factors (forcing functions) controlling sensitivity of response to
different classes of toxic and nontoxic (non-point source) Stressors.
Identification of key factors determining sensitivity of response across multiple Stressors
to facilitate development of a comprehensive classification scheme rather than multiple
schemes.
Development of national and regional classification frameworks.
Coordination of opportunities for testing classification strategies in a systematic fashion.
Efforts of this workgroup will be supplemented by the extramural grants STAR program. The
STAR grants program has an existing request for proposals (RFP) on aquatic ecosystem
classification, and might add an RFP on watershed classification strategies in the future.
Alternative strategies for classification will be tested through regional case studies (project 3).
Regional case studies will be based on multiple-scale classification schemes, with coordination
across Divisions to bring together appropriate areas of expertise. In particular, stressor-response
relationships will be compared among regional/watershed/water-body classes.
149
-------
Table 4. Existing or proposed approaches to classification at regional, watershed, water-body, and habitat scales.
Responsible entity
MED
MED
AED
AED
AED
WED
Scale
Regional/watershed
Water body
Water body
Habitat
Habitat/Water body
Regional watershed
Water-body type
Streams
Coastal wetlands
(freshwater, Great
Lakes)
Estuaries
Estuaries
Estuaries
Streams, coastal
wetlands, estuaries
Stressor(s) of interest
Nonpoint-source (NFS)
suspended and bedded
sediments, nutrients,
habitat alteration
(including structure,
flow and thermal
regimes)
Nutrients,
Habitat alteration
Nutrients
Toxic chemicals
Toxic chemicals,
habitat alteration,
nutrients,
suspended and bedded
sediment
Nutrients, habitat
alteration
Analytical approaches
(e.g., a priori vs
posteriori )
a priori
posteriori (initially)
a priori
a priori
Biogeochemical
Multi-variate
classification and
ecosystem models
Mixed
Conceptual basis
Based on hydrologic
thresholds derived from
land- use/land-cover
attributes and regional
soils/climate/
geomorphology
(natural runoff
potential)
Hydrogeomorphic
classes reflecting
different water sources
and retention time
Retention time
Toxic chemicals
will/will not occur
under certain
biological, geological
and chemical
circumstances
Pattern recognition and
energy dynamics
Ecoregional,
hydrogeomorphic,
tie in from EMAP
150
-------
Table 4. Existing or proposed approaches to classification at regional, watershed, water-body, and habitat scales.
Responsible entity
GED
NERL-Las Vegas
USGS-NAWQA
US Forest Service
Scale
Water body
Regional
Regional/watershed
Continental/
regional/
subregional/
waterbody/
habitat
Water-body type
Estuaries
All
Streams, lakes,
estuaries
Stressor(s) of interest
Nutrients
All
Analytical approaches
(e.g., a priori vs
posteriori )
Multivariate models
Multi-variate
classification based on
similarity of natural
landscape attributes
Multi-variate
classification based on
similarity of natural
landscape attributes;
watershed classification
along land-use
gradients
A priori
Conceptual basis
Biological endpoints,
retention time, tidal
frequency,
geomorphology
Pattern recognition
Pattern recognition,
physiographic controls
on pollutant
expo sure/transport
Based on hydrologic,
geologic, landscape
position
151
-------
Table 4. Existing or proposed approaches to classification at regional, watershed, water-body, and habitat scales.
Responsible entity
The Nature
Conservancy and
Aquatic Gap Analysis
Program (GAP)
Scale
Reg ional/subreg ional,
watershed, ecosystem,
stream valley segment,
water body
Water-body type
Streams, lakes,
ecosystems of streams,
lakes, wetlands
Stressor(s) of interest
All
Analytical approaches
(e.g., a priori vs
posteriori )
A priori
Conceptual basis
Based on water body
size, hydrologic
regime, geology, size,
gradient, catchment
characteristics,
connectivity, landscape
network position, biota
Also includes whether
lakes are unconnected,
headwaters, or flow-
through lakes, which
all influence sources,
chemical and
temperature
characteristics, and
residence times of
water
152
-------
Products
APM IB (GPRA # 202) FY03 Classification frameworks for geographic regions, watersheds,
water-body, and habitat scales (MED, GED).
Benefit of Products
A classification framework will provide regional, State, and Tribal regulatory authorities a tool to
collapse the over 40,000 water bodies requiring TMDLs into a more manageable number of
similar units or water body classes. With defined water body classes, a TMDL template for
remediating the impairment could be created which could then be applied to all of the water
bodies within the class with minor adjustments. This would eliminate the need for 40,000
unique TMDLs.
Classification frameworks will also help to regionalize criteria development or definition of
thresholds for impairment. This would improve the applicability of criteria to specific sites or
classes of sites and lower the error rate in identifying impaired or threatened aquatic ecosystems.
In particular, a watershed classification scheme within a regional framework will help to
integrate and coordinate the 303(d) listing process at the watershed scale.
Project Title 3. Diagnostic Tool Development and Application through Regional Case Studies
Project Coordination and Resources (36.55 FTEs: FY03: AED-3.25, GED-1.5, MED-8.4 [total =
13.15]; FY04: AED-7.8, GED-2.5, MED-13.1 [total = 23.4])
Objectives
Develop diagnostic tools for single and multiple stressors;
Develop forecasting models;
Illustrate the application of diagnostic methods, tools and models for single and multiple
stressors, including forecasting models;
Provide input to regional decision-support systems;
Demonstrate how assessment results can be extrapolated across regions, watersheds, and
water bodies, and biological levels of organization; and
Illustrate how stressor-response relationships vary among different classes of systems in a
predictable fashion.
153
-------
Scientific Approach
Case studies are a useful vehicle for developing and testing conceptual models, classification
systems, diagnostic tools and models, and stressor-response relationships. Furthermore, case
studies focused on specific places or issues of interest to the Agency provide an excellent
mechanism to address high priority environmental problems, including the development of
TMDLs. Diagnostic case studies will provide a mechanism for developing, testing, and applying
methods and models for distinguishing among single aquatic stressors and allocating cause
among multiple stressors. Case studies will be performed to incorporate the habitat, ecosystem,
watershed, and regional spatial scales as well as the organismal, population, community, and
ecosystem levels of biological organization.
Case studies will be selected based on the following critical attributes:
Sites will be selected from those already designated as impaired (or threatened) based on
the 305(b) or 303(d) reporting process, representing a range of degrees of impairment, a
range of stressor combinations, multiple stressors with interaction potential, and common
stressor-resource class combinations.
Sites will be selected to represent specific region-watershed-water body classes, such that
results can be extrapolated using a regional or nationwide classification system.
Coastal systems will be selected, to include both watershed(s) and receiving waters.
Methods and information management will be coordinated across case studies.
Additional desired attributes of regional case studies include the following:
Data have already been collected and/or analyzed in time series or at multiple time points.
Studies will involve cross-Agency collaboration.
Sites will have well-organized stakeholder groups [e.g., National Estuarine Reserve
Program, National Estuary Program, Areas of Concern or Lakewide Management Plan
(LaMP) committees for Great Lakes].
Studies will not duplicate assessments of well-studied systems.
Representative models exist.
Research or monitoring at the sites is ongoing.
Sites are in logistical proximity to an Ecology Division.
Each case study will include the following elements:
154
-------
Study sites will be selected along stressor gradients within each class in the classification
framework, to determine if stressor-response relationships vary among these classes.
Stressor gradients will be defined by land-cover/land-use attributes in order to provide the
opportunity to develop simple diagnostic indicators of impairment at the watershed scale.
Gradients representing multiple stressors will be used to test multivariate methods in
order to allocate variation in biological community composition to different
environmental gradients or stressors.
Model development will incorporate both single stressor-response relationships and
multiple stressor interactions to enhance forecasting of potential future impacts or
recovery based on future scenarios.
1. Great Lakes Coastal Watersheds/Wetlands/Nearshore Zones.
Magnitude of watershed loadings and the relative sensitivities of instream and coastal ecosystems
will be predicted and assessed through the application of watershed and coastal wetland
classification schemes in the Lake Michigan basin. A watershed classification scheme based on
hydrologic thresholds of response to forest fragmentation and watershed storage already has been
tested for small coastal watersheds surrounding the western arm of Lake Superior across two
distinct hydrogeomorphic regions (Detenbeck et al. 2000). This classification scheme is being
extended to other regions with different single land-use gradients, and to watersheds with mixed
land-use gradients (Simon 1999, Cincotta 2000). Through collaboration on a West Virginia
REMAP project, watershed classification schemes are being developed for a region with
mountainous terrain, and consider land-use gradients related to agriculture, urban and residential
development, and mining. Through collaboration on a Great Lakes coastal wetlands REMAP
project, watershed classification strategies are being developed for watersheds spanning a range
of sizes (1 to 450 ha) and a mixture of land-uses. In addition, MED scientists have assessed
differences in nutrient dynamics among hydrogeomorphic types of coastal wetlands, based on
expected differences in retention time and relative influence of riverine versus lacustrine inputs.
This approach will be extended to evaluate impacts from suspended and bedded sediment
loading. For toxic chemicals, methodologies will be developed for predicting the parameters and
variables which affect the bioavailability of the chemicals for the different levels of the
classification schemes.
MED will continue to develop a series of empirical stressor-response relationships in the Great
Lakes basin for stressors known to constrain community composition for specific combinations
of taxa and aquatic resource classes: a) temperature, flow, and clean sediments for instream fish
communities; b) suspended and bedded sediments and flow for macroinvertebrate communities;
c) flow, suspended and bedded sediments and nutrients for periphyton communities; d) nutrients
and suspended and bedded sediments for coastal wetland vegetation; and e) habitat alteration and
food-web structure for coastal fish populations and communities (Detenbeck et al. 2000).
Current protocols for development of indices of biotic integrity (IBIs) yield indicators of general
condition of biological communities for State 305b assessments, but do not yet provide
155
-------
information on causes of biological impairment for the State 303d listing process. Analysis of
existing State, REMAP, and EMAP monitoring databases (www.epa.gov/emapA) and aquatic
toxicity databases (AQUIRE, PHYTOTOX; www.epa.gov/ecotox/) with associated exposure
data will be coordinated across Divisions to derive diagnostic indicators to predict cause of
impairment based on aquatic community composition. Examples of multivariate tools that can
be applied to suggest causal hypotheses include nonmetric dimensional scaling (NMDS)
ordinations to identify environmental gradients associated with gradients in community
composition (Beals 1984), indicator analysis (Dufrene and Legendre 1997), discriminant function
analysis relating environmental factors with species presence/absence (Scheller et al. 1998), and
redundancy analysis to partition variance among multiple potential causal factors (Richards et al.
1993).
To address water column toxicity, EPA has developed TIE methods (Mount and Anderson-
Carnahan 1988, 1989; Burkhard and Ankley 1989; Mount 1989; Ankley et al. 1991; Norberg-
King et al. 1991,1992; Durhan et al. 1993; Mount and Norberg-King 1993; Burgess et al. 1996;
Ho et al. 2002), which are a battery of physical/chemical manipulations coupled with toxicity
tests. By determining which physical/chemical manipulation affect toxicity of the samples, the
general characteristics of the causative chemical(s) can be determined. With this knowledge,
appropriate analysis techniques and in some cases, in combination with additional sample
fractionation techniques, are used to obtain a list of the tentative chemical(s) in the sample. With
this information, toxicity tests using the suspected chemical(s) would be performed to establish
the effect level for these chemicals in the water samples of interest and in reference waters using
the TIE organisms. Successful TIEs occur when the concentrations of the suspected chemicals at
the affect endpoint agree among the water samples and reference waters.
For toxicity in sediments, substantial progress has been made to date for a number chemical
classes and manipulations for whole sediments and sediment pore waters (Ankley and
Schubauer-Berigan 1995, Besser et al. 1998, Ho et al. 1999, Leonard et al. 1999, Burgess et al.
2000). With the successful development of solid-phase sediment TIE methods, field validation
of interstitial water and whole sediment TIE methods is needed. After development of the whole
sediment and interstitial waters TIE methods, field validation of the methodologies are required
to determine if the causes of toxicity identified by TIE represent the source of toxicity at the field
site. Field validation will involve the TIE analysis of sediments with impaired benthic
communities from both fresh and marine sites, and ideally, the causes of impairment for these
sediments would not be some other stressor (e.g., suspended and bedded sediment or degraded
habitat). Once a suspected toxicant is identified, field sediments and organisms would be
analyzed. The final step in the validation process would be to reproduce the same community
signature observed in the field, within laboratory-controlled situations by introducing the
suspected toxicant into clean sediments in a mesocosm. The field validation effort will also
allow the evaluation of benthic community signatures and toxicant relationships. If useful
relationships can be developed, a library of chemical stressor-benthic community responses
would be developed to complement relationships derived from toxicity databases above, and this
library would be developed on a water body class scale. Field validation will also permit the
evaluation of toxicant/stressor and biological indices relationships for benthic communities.
Specifically, a collaborative effort between MED and AED will seek to link cause and effect
relationships observed in the laboratory to field effects using micro/mesocosm simulations.
156
-------
A suite of ecosystem response models is being developed for Lake Michigan that link inputs
from tributaries to their associated large receiving water bodies and ecological responses. The
construct is best described as mechanistic, mass balance models, and the primary suite of coupled
and linked models being applied include: atmospheric, meteorological, hydrodynamic, sediment
transport, eutrophication, sorbent dynamics, water quality transport and fate, and food chain
bioaccumulation. The modeling focuses on establishing relationships of contaminant and
nutrient loadings and ambient concentrations with chlorophyll, DO, N/P ratios, phytoplankton
species composition, lower food chain productivity, water column transparency, habitat, fish
consumption advisories, and lake trout egg hatchability. These models improve the predictive
ability for forecasting environmental benefits of specific load reduction scenarios of nutrients and
contaminants, as well as the time to realize those benefits.
2. Shallow Estuarine Systems in the Northeast Atlantic.
A case study is proposed by AED to quantify the endpoint parameters being proposed in the
Habitat Alteration, Nutrients, and Toxic Chemicals implementation plans (see Sections 4, 5, and
7, respectively), integrate the results within the conceptual framework proposed in this Section,
and use this information to test the utility of stressor-response relationships and diagnostic
methods and models under development. The initial research studies will be carried out within
the coastal ecosystems of the Northeastern U.S., particularly the Narragansett Bay and
neighboring coastal systems in Regions 1 and 2 at sites listed for TMDL development.
Two stressors will be emphasized initially: nutrients and toxic chemicals. Several related
projects will examine the effects of a range of nutrient loadings on several different coastal
ecosystems, (e.g., marshes, shallow coves, and small estuaries) through field studies and model
development. Concurrently, another proj ect will study the effects of several classes of toxic
chemicals on organisms, populations, and communities that dwell in critical habitats along a
salinity gradient from fresh to salt water. Each investigation will synthesize data in a manner that
allows us to characterize the contribution of each stressor to adverse ecological effects. For
example, a collaborative effort between MED and AED will seek to link cause and effect
relationships observed in laboratory to field effects using micro/mesocosm simulations. These
studies will be integrated through an ecological model that examines the individual and
interactive effects of nutrients and specific toxic chemicals on important habitats. This model
will be validated by laboratory and field studies of systems where both nutrients and toxic
chemicals are thought to be responsible for observed conditions.
3. Shallow Estuarine Systems in the Gulf of Mexico.
A case study is proposed by GED to quantify the endpoint parameters being proposed in Sections
4, 5, and 7, respectively, integrate the results within the conceptual framework proposed in this
Section, and use this information to test the utility of stressor-response relationships and
diagnostic methods and models under development. The initial research studies will be carried
out within the coastal ecosystems of the Gulf of Mexico, particularly the Pensacola Bay and
neighboring coastal systems in Regions 4 and 6.
157
-------
Again two stressors will be emphasized initially: nutrients and toxic chemicals. The first step in
the approach will be to examine existing 303(d) impairment lists, databases on nutrients and
toxic chemicals, and land use/land cover characteristics for Gulf of Mexico estuaries to delineate
four classes of sites based on observed effects or criteria. The four classes include sites affected
by nutrients only, toxic chemicals only, both nutrients and toxic chemicals, and neither nutrients
nor toxic chemicals. A candidate suite of biological indicators would then be developed that
demonstrates differential sensitivity to either stressors. This would require examination of
historical effects databases and population or community response data as well as receiving input
on single-stressor response models for nutrients and toxic chemical efforts (see Sections 5 and 7).
In collaboration with AED, lab or field tests will be used to validate the sensitivity of these
indicators in each of the four classes of sites. If historical data exists for indicators in the study
areas, tests will confirm the sensitivity of indicators to these two stressors. Multivariate analysis
methods will be applied to allocate variation in the response indicators to differentiate between
nutrient and toxic effects. Modeling approaches will then be used to integrate individual and
interactive effects of nutrients and toxic chemicals on biological indicators. Models would
account for population and community levels of response across the four classes of sites and
along stressor gradients. GED and AED will coordinate the development of models within the
context of the classification framework and diagnostics.
4. Coordination with other Goal 2 Research (WED).
The freshwater habitat alteration group at WED is developing a project to examine the influence
of human activities on native fish habitat at reach, watershed, and landscape scales. This group,
led by Jim Wigington, is developing salmon and native fish assemblage modeling approaches
while concurrently evaluating the interactive influences of flow, temperature, physical habitat,
and nutrients on salmon and native fish. This project is focused on coastal drainages of Oregon
where there is a great opportunity to contribute to the restoration of salmon populations through
cooperative research efforts with State (Oregon Department of Fish and Wildlife, Department of
Environmental Quality) and other Federal agencies (NMFS, U.S. Forest Service).
5. Coordination with GPRA Goal 8 (AED, GED, MED, WED).
Goal 2 activities under Diagnostics will be coordinated with Goal 8 activities in monitoring and
assessment through the EMAP program. There are currently monitoring initiatives underway
through: 1) the Coastal Initiative, examining the condition of marine estuaries; 2) the STAR
grants program to develop indicators for coastal freshwater and marine systems, (including the
GLEI cooperative agreement for indicator development on coastal Great Lakes systems); 3)
Western EMAP (including coastal watersheds in the states of Washington, Oregon, and
California as intensive monitoring sites); and 4) a variety of Regional EMAP projects. The latter
program currently emphasizes watershed-scale approaches to monitoring and assessment. For
example, a recently initiated REMAP project with the state of West Virginia will demonstrate
both the development of a watershed classification system and test thresholds of land-use/land-
cover along gradients of disturbance (related to forestry, agricultural, development, and mining
activity), while at the same time developing fish indices of biotic integrity.
158
-------
LOCATIONS OF NATIONAL WATER-QUALITY
ASSESSMENT STUDY UNITS
6. Coordination with other State, Regional, and Federal agencies (AED, GED, MED, WED).
Development of regional case studies will take advantage of ongoing monitoring and assessment
activities by State, regional, and Federal agencies. The USGS NAWQA program is examining a
nationwide approach to regionalization of watersheds, based on common vulnerabilities to
nonpoint stressors (McMahon and Cuffney 2000 and Carolyn Couch, personal communication).
Within this regional classification
framework, watersheds in NAWQA
study units are being selected along
gradients of urbanization to examine
response of stream ecosystems to
development pressures. Each of the
NHEERL divisions is located in
proximity to NAWQA study units that
are scheduled to adopt this approach
during the near future, with the start of
the second 10-year phase of monitoring
(units in dark blue on Figure 15).
| Began in fiscal year 1991
| Began in fiscal year 1994
| Began in fiscal year 1997
Began in fiscal year 1999
High Plains Reg. Ground Water Study \~\ Not scheduled
In addition, opportunities exist to
coordinate studies with the National
Estuarine Program (NEP), Great Lakes
Areas of Concern, and Great Lakes
LaMPs.
Figure 15. Locations of national water-quality
assessment study units.
Products
1. Single stressor-methods and models.
APM 2A FY03 Guidance document on whole sediment TIEs ( MED, AED).
APM 2C F Y05 Guidance on and user-friendly interface for derivation of diagnostic indicators for
individual stressors (MED, AED).
2. Multiple stressor-methods and models.
APM 3 A FY02 Case studies of multivariate approaches to community data analysis to apportion
cause among stressors (AED, MED).
APM 3B FY06 Methods and models for multiple stressors with case studies (MED, AED).
159
-------
Forecasting Approaches
FY07 Methods and models to support alternative remediation schemes to achieve specific
management goals (MED).
3. Case Studies.
APM 2B F Y05 Application of coastal watershed and estuarine/lacustuary classification schemes
to predict probability of impairment based on Great Lakes and Gulf of Mexico regional case
studies (GED, MED).
Benefit of Products
Diagnostic tool development will produce the single stressor, multiple stressor, and forecasting
methods and models necessary to determine the causes of adverse effects on intact water bodies.
Regional case studies will provide the basis for verifying the efficacy of these diagnostic tools.
Further, regional case studies will provide the basis for development of the guidance listed
above, and allow diagnostic tools to be demonstrated to stakeholders in sites where TMDLs need
to be developed. These studies will enable OW to understand how multiple stressors, such as
nutrients and toxic chemical loadings, affect important habitats separately and in combination for
several types of coastal ecosystems across the U.S. We expect the methods and models
developed here to be generic for specific stressor-ecosystem combinations. Therefore, we predict
that they can be applied in other regions that contain similar stressor-ecosystem combinations.
Classification schemes will allow us to regionalize results and recommendations for TMDLs and
watershed restoration activities. The scientific approach used here is also generic and it could be
applied to develop similar relationships for the ecosystems and stressors that predominate in any
region.
Project Title 4. Generic Models for the Evaluation of Multiple Stressor Interactions
Project Coordination and Resources (1.5-8.0 FTEs: AED, GED, MED, starting in FY04/05,
increase over time)
Objectives
The objective of this research area is to assess the likelihood that synergistic and/or antagonistic
ecological effects will occur from the interactions of multiple stressors. To set priorities for the
development of TMDLs and the restoration of impaired water bodies, it is necessary to
understand how the potential interactions among stressors will affect system recovery once one
of the stressors is reduced. For example, in a turbid coastal wetland, a reduction in suspended
solids loading without an accompanying reduction in P loading from upstream animal feed lots
could unmask a eutrophication problem that was previously not evident due to light limitations
on primary production.
160
-------
Scientific Approach
As a first step, key combinations of stressors for which interactions are expected to occur will be
identified based on mechanisms of action outlined in conceptual models and review of 303(d)
listings for common combinations of stressors. Individual stressor dose-response relationships
and models developed under project 3 will provide a starting point for examining the interactions
of multiple stressors in freshwater and marine ecosystems. The importance of interactive effects
will be evaluated by including the documented pathways of stressor action and interaction in
deterministic dynamic models calibrated with field studies and/or historic data, and then
simulated over many runs to discover the sensitivity of measurement endpoints to changes in one
or more of the stressors (e.g., Bartell et al. 1984, Mitsch and Reeder 1991, Hanratty and Stay
1994, EPA 2000c). A generic model including the impact and interaction paths for the dominant
stressors of interest will be applied for each of three resource classes: streams, lakes and
reservoirs, and estuaries. These generic model frameworks will be developed as a joint product
among the four Ecology Divisions. Sensitivity analysis of these models will serve as a first order
estimator for allocating observed ecological effects among two or more interacting stressors, as
well as a means for evaluating the relative importance of indirect and interactive effects. This
approach is independent of scale requiring only that the stressor- response relationships and
interaction pathways be documented on the scale of interest. Once expected interactive effects
and ranges of interactions are identified, the results of existing case studies and ongoing regional
case studies will be reviewed for evidence of interaction effects. Pending outcomes of simulation
exercises, additional field studies will be performed, combined with carefully crafted laboratory
experiments and physical models calibrated to match loadings and functional properties observed
in the field system.
Products
FY03 Identification of key combinations of stressors expected to interact within conceptual
model (all Ecology Divisions).
APM 3B FY06 Simulation of key stressor interactions with generic ecosystem models using
sensitivity analysis to define the range of stressors and stressor combinations under which
nonadditive interactive effects will occur (MED, AED).
Benefit of Products
Part of the TMDL process involves allocation of the cause of impairment among multiple
stressors. The simplest case possible involves additive effects, which can be predicted from
single stressor models. If effects are synergistic or antagonistic, then the results of reduced
loadings will be more difficult to predict. The proposed products would identify the extent to
which stressor interactions are expected to occur in natural ecosystems and those combinations of
factors which favor their occurrence. This knowledge is critical to allow the States and Tribes to
develop viable restoration and remediation plans for water bodies and watersheds affected by
multiple stressors.
161
-------
Project Title 5. Decision-Support System
Project Coordination and Resources (3.0-6.0 FTEs: AED, GED, MED, starting in FY05/06, plus
programming support)
Objectives
The objective of this research is to develop a decision-support system for diagnosing the causes
of biological impairment at multiple scales. The decision-support system will be based on
conceptual models outlining expected cause-effect relationships involving single and multiple
stressors (project 1). At the most basic level, this interface will consist of a guide to the use of
existing EPA databases, methods, and models. As tools are developed for diagnosis of single or
multiple stressors, these will be incorporated into an expert-system framework. Simple empirical
relationships between exposure and response indicators will be incorporated, as well as more
detailed mechanistically-based models.
Scientific Approach
Phase 1. Identify existing tools, methods, and models available to support establishment of
cause-effect relationships
A conceptual model is proposed to provide a framework for documentation of existing tools to
support diagnosis (Figure 16). Within this conceptual framework, it is possible to develop
methods and models that bypass many of the linkages, for example, relating alterations in land-
use/land-cover within an existing watershed class to projected effects on fish community
integrity. Examples of existing methods and models available to establish or confirm the cause-
effect linkages in Figure 16 are listed in Table 5.
Phase 2. Produce a decision support system design
Existing decision support systems will be investigated to choose an appropriate approach. Based
on the approach chosen, a system framework will be developed and a document produced to
instruct collaborators on the intent and use of the decision-support system and to build
management support for this system through practical demonstrations of its utility.
The technology now exists to produce a computerized, web-based decision support system. The
BASINS system supported by OW consolidates a geospatial framework to support a
comprehensive diagnostic decision-support system. At present, it is a collection of models and
databases organized around a GIS-based computer program. This system will be expanded to
include the information and tools produced under this Aquatic Stressors research effort. "Expert
system" modules will be added that would lead persons in the Program Offices and State
agencies through a guided, rule-based scenario that leads to a diagnosis of the causes of
162
-------
Native Fish Populations fl
Communities
Classification
Regional,
Watershed
Process
Classification
Rlter
EoaBTng "-
:Cori ceritratioh:
QS3S
Food Web Component
Assessment
Eridfjint
Figure 16. Conceptual model of cause-and-effect relationships in coastal systems, providing a framework for a decision support
system. See key to model components at base of figure. Loading terms include atmospheric component.
163
-------
Table 5. Examples of methods incorporated in conceptual framework (Figure 16) for decision support system.
Linkage
Links at reach or water-body
scale between State variables
and biological endpoints
25, 1 => 19
3, 4=>19
Aquatic
Sediments
8 => 9,22
12,13, 15 => 10
11, 12,15,18 => 23
Method/Model
SI Method
Watershed classification
scheme
Fish temperature-flow
matching database system
TIE
TIE
Nutrient effects on
phytoplankton and
zooplankton food webs
Toxicity of contaminated
sediments to plants
Benthic community indicator
of estuarine condition
Reference
US EPA 2000c
Detenbeck et al. 2000
Scheller et al. 1998
Mount and Anderson-
Carnahan 1988, 1989; Mount
1989; Ankleyetal. 1991;
Norberg-King et al.
1991,1992; Durhan et al. 1993;
Mount and Norberg-King
1993; Burgess et al. 1996
Ho et al. (2002)
Lores et al. (2002a,b)
Murrell et al. (2002)
Lewisetal. 2001
Lewis et al. 2000
Engleetal. 1994
Engle& Summers 1999
Division
SI Group
MED
MED
AED (marine)
MED (fresh water)
AED (marine)
MED (fresh water)
GED
GED
GED
164
-------
impairment, the development of a TMDL, and recommendations for restoration of remediation
(e.g., see Reynolds 1999a,b; http://www.fsi.orst.edu/emds/).
Phase 3 . Development and implementation
A team will be formed consisting of research scientists, modelers, computer systems analysts,
and WEB page designers to develop and implement the system. Contact persons from each of
the research teams working in the Aquatic Stressors research program will be assigned to provide
liaison with the decision support team. If appropriate, other team members and contacts would
be provided from other EPA Laboratories, Program Offices, and State agencies. This
development effort would produce the decision support system and a guidance document on its
application.
Phase 4. Application
During the final stages of development, data, models, and information from selected case studies
will be incorporated into the expert system. This would be done in cooperation with the decision
personnel associated with the studies and training would be provided along with a guidance
document and system documentation.
Products
FY04 Framework(s) for decision-support system (all Ecology Divisions).
FY06 Decision-tree guiding application of approaches with case studies (all Ecology Divisions).
APM 3C FY07 Decision-support system(s), including forecasting of future cause-effect
relationships (MED).
Benefit of Products
A series of decision-support systems will provide States and Regions with a common set of tools
for approaching TMDL development and assessing alternative options for watershed restoration.
Existing tools and databases will be made readily available, and gaps in knowledge bases will be
more readily identified. Decision-support systems will also facilitate an integrated approach to
assessment and diagnosis at the regional scale, thus improving the efficiency of the TMDL
process.
Gap Analysis
Gaps by geographic region, biological scale, resource class, andstressor type
The research tasks outlined above will be applied within coastal watersheds bordering on both
freshwater and saltwater receiving water bodies. Regional case studies will be developed within
each of three areas, the Great Lakes, the Atlantic coast, and the Gulf coast. This research area
will depend heavily upon each of the other research areas of this document (Sections 4-7) for
165
-------
development of stressor-response relationships for individual stressors and will coordinate with
these respective research groups in the development of diagnostic methods and models for
individual stressors. The focus of each stressor-specific research area is outlined below in Table
6, with the biological endpoints and spatial scale of investigation noted. Gaps have been
identified relating to the investigation of nutrient effects in freshwater streams and rivers, and the
investigation of clean sediment effects on coastal wetlands and estuaries. Because these are
significant stressors in the EPA Regions, biological scales and resource types of interest, the
diagnostics work will have to be supplemented by work in other research areas or rely on
Table 6. Single aquatic stressors method development covered by other research areas within
the Aquatic Stressors Framework [in italics = endpoints: biological scale (organism,
population, community, ecosystem); spatial scale (habitat, water body, watershed, region)]
Stressor
Nutrients
Suspended and
Bedded Sediments
Habitat Alteration
Toxic Chemicals
Resource class
Fresh Water
Coastal & Lakes
Community,
Ecosystem
Nutrient Group
Population scale,
Habitat Alteration
Group
Population level,
Toxic Chemicals
Group
Fresh Water
Watershed
Habitat, community
Salmon populations,
Habitat Alteration
Group
Population level,
Toxic Chemicals
Group
Marine Coastal
Organismal (SA V)
Population (SAV)
Community,
Ecosystem,
Watershed
Nutrient Group
Organismal (SAV)
Population (SAV)
Shrimp populations,
Habitat Alteration
Group
Organismal,
population,
community,
Toxic Chemicals
Group
existing literature to fill these gaps. In addition, the lack of effects work at the community level
for many resource class/stressor combinations will need to be addressed, as this is the biological
scale at which the States are assessing biological impairment in current monitoring programs.
Other gaps have been identified by geographic region. Expertise available at AED and GED will
be supplemented with expertise in watershed classification and assessment through collaboration
with other research groups (e.g., MED or WED, USGS NAWQA program). Development of
regional case studies or decision-support systems for the Pacific Northwest will require close
166
-------
collaboration with Goal 8 research efforts, and/or other research areas under the Aquatic Stressor
research effort.
Gaps by skill area and level of effort
The level of effort and required areas of expertise for each of the different research tasks
discussed in this document are summarized in Table 7. Overall, the total number of FTEs
estimated for implementation of the Diagnostics plan is consistent with the total number of FTEs
allocated to Diagnostics across all Ecology Divisions. This conclusion is based on two
assumptions:
1. There will be close coordination with other Aquatic Stressors research areas (Sections 4-7) in
the development of conceptual models, classification frameworks, implementation of regional
case studies, and development of decision support systems.
2. The requisite expertise available within the Ecology Divisions will be made available for
diagnostics research by management at the respective Divisions.
Some gaps do exist in the types of expertise available relative to resource needs. Additional
expertise and/or support will be needed for statistics (design issues), futures analysis
(socioeconomics, ecological economics), near coastal modeling for marine systems, ecosystem
simulation modeling, and programming support for decision-support system design and
development. Some of these needs could be met through additional infrastructure support for
existing contracts (statistical support available through AED, ADP support contracts at all
Divisions), while other needs must be factored into long-term hiring plans (ecological
economics, socioeconomics) or development of extramural grants. Some modeling support
needs could be met through coordination with NERL Divisions, Regions, or Program Offices,
but additional support will be needed to develop food web components of models.
167
-------
Table 7. FTE resource allocation for diagnostics by year (FW = freshwater, SW = saltwater).
Research Themes
1
2
3
4
5
6
7
8
Skills needed
Diagnostics 1: Conceptual Model
Conceptual Model & Information
Management Framework
6
6
Resource classes (FW, SW)
ecosystem/watershed scale
all workgroups
empirical/deterministic models
Diagnostics 2: Classification Framework
Classification Framework
2
2
8
4
Resource classes, GIS, cross class &
scale, all workgroups
multivariate stats
Diagnostics 3: Diagnostic Tools and Regional Case Studies
Single-stressor diagnostic methods
Multivariate approach to community
analysis
Whole Sediment TIE
Comparison of forecasting
approaches
2
3
7
2
3
7
8
3
8
4
8
8
8
8
11
11
All workgroups
indicator development
cross-scale [comm. -watershed]
Statistician, community ecol.
resource classes (FW, SW)
Chemist, biologist (FW, SW)
Modelers (chem. &biol.), ecological
economics, social science, cross-
scale
Diagnostics 4: Generic models for the evaluation of multiple stressor interactions
Stressor interaction models
1.5
8
Statistician, modelers, ecology,
toxicology, chemists, resource
classes, cross-scale
all workgroups
Diagnostics 5: Decision-support system
Decision-support system
Total FTE per Fiscal Year
20
20
1.5
22
1.5
25.5
1.5
17.5
1.5
17.5
3
14
3
14
EMAP, statistician, design
all workgroups, NERL
168
-------
References
Ankley, G.T., Schubauer-Berigan, M.K. 1995. Background and overview of current sediment
toxicity identification evaluation procedures. J. Aquat. Ecosyst. Health 4:133-149.
Ankley, G.T., Schubauer-Berrigan, M.K., Dierkes, J.R., Lukasewycz, M.T. 1991. Sediment
toxicity identification evaluations: phase I (characterization), phase II (identification) and phase
IE (confirmation) modifications of effluent procedures. EPA 600/6-91/007. EPA, Environmental
Research Laboratory, Duluth, MN.
Bartell, S.M., Gardner, R.H., O'Neill, R.V. 1984. The fates of aromatics model (FOAM):
description, application, and analysis. Ecol. Model. 22:109-121. Beals, E.W. 1984. Bray-Curtis
ordination: an effective strategy for analysis of multivariate ecological data. Adv. Ecol. Res.
14:1-55.
Beals, E.W. 1984. Bray-Curtis ordination: an effective strategy for analysis of multivariate
ecological data. Adv. Ecol. Res. 14:1-55.
Bergman, H.L., Dorward-King, EJ, ed. 1997. Reassessment of Metals Criteria for Aquatic Life
Protection. SETAC Press, Pensacola, FL.
Besser, J.M., Ingersoll, C.G., Leonard, E.N., Mount, D.R. 1998. Effect of zeolite on toxicity of
ammonia and metals in freshwater sediments: implications for toxicity identification evaluation
(TIE) procedures. Environ. Toxicol. Chem. 17:2310-2317.
Brinson, M.M. 1993. A hydrogeomorphic classification for wetlands. Wetlands Research
Program Technical Report WRP-DE. U.S. Army Corps of Engineers, Washington, DC.
Burgess, R.M., Ho, K.T., Morrison, G.E., Chapman, G., Denton, D.L. 1996. Marine toxicity
identification evaluation (TIE): phase I guidance document. EPA/600/R-96/054. Office of
Research and Development, Washington, DC.
Burgess, R.M., Cantwell, M.G., Ho, K.T., Kuhn, A., Cook, H.F., Pelletier, M.C. 2000.
Development of a toxicity identification evaluation (TIE) procedure for characterizing metal
toxicity in marine sediments. Environ. Toxicol. Chem. 19:982-991.
Burkhard L.P., Ankley, G.T. 1989. Identifying toxicants: NETAC's toxicity-based approach.
Environ. Sci. Technol. 23:1438-1443.
Cincotta, D. 2000. A small watershed characterization, classification, and assessment for West
Virginia utilizing EMAP design and tools. R-EMAP proposal submitted to EPA, Mid-Continent
Ecology Division, Duluth, MN.
Cowardin, L.M., Carter, V., Golet, F.C., LaRoe, E.T. 1979. Classification of wetlands and
deepwater habitats of the United States. FWS/OBS-79/31. U.S. Fish & Wildlife Service,
Washington, DC.
169
-------
Detenbeck, N.E., Batterman, S.L., Brady, V.J., Brazner, J.C., Snarski, V.M., Taylor, D.M.,
Thompson, J.A., Arthur, J.W. 2000. A test of watershed classification systems for ecological risk
assessment. Environ. Toxicol. Chem. 19:1174-81.
Dufrene, M., Legendre, P. 1997. Species assemblages and indicator species: the need for a
flexible asymmetrical approach. Ecol. Monogr. 67:345-366.
Durhan, E.J., Norberg-King, T., Burkhard, L.P. 1993. Methods for aquatic toxicity identification
evaluations: phase II toxicity identification procedures for samples exhibiting acute and chronic
toxicity. EPA/600/R-92/080. Environmental Research Laboratory, Duluth, MN.
Engle, V.D., Summers, J.K. 1999. Refinement, validation, and application of a benthic condition
index for Gulf of Mexico estuaries. Estuaries 22:624-635.
Engle, V.D., Summers, J.K., Gaston, G.R. 1994. A benthic index of environmental condition of
Gulf of Mexico estuaries. Estuaries 17:372-384.
EPA. 1991. Guidance for water-quality-based decisions: the TMDL process. EPA 440-4-91-001.
Office of Water, Washington, DC.
EPA. 1996. Proposed guidelines for ecological risk assessment. EPA-630/R-95-002B.
Washington, DC.
EPA. 1998a. Clean Water Action Plan. EPA 840-R-98-001.
EPA. 1998b. Better assessment science integrating point and nonpoint sources: BASINS version
2.0. EPA-823-B-98-006. Office of Water, Washington, DC.
EPA. 1999. Clean water action plan: restoring and protecting America's waters. EPA-840-R-98-
001. Washington, DC.
EPA. 2000a. National water quality inventory: 1998 report to Congress. EPA-841-F-00-006.
Office of Water, Washington, DC.
EPA. 2000b. Aquatox release 1: a simulation model for aquatic ecosystems. EPA-823-F-00-015.
Office of Water, Washington, DC.
EPA. 2000c. Stressor identification guidance document. EPA-822-B-00-025. Office of Water
and Office of Research and Development, Washington, DC.
EPA. 2001. 1998 TMDL tracking system (version April 2, 2001).
Federal Register. 2000. Unified federal policy for a watershed approach to federal land and
resource management. 62566 Federal Register, Vol. 65, No. 202, Wednesday, October 18, 2000.
170
-------
Frissell, C.A., Liss, W.J., Warren, C.E., Hurley, M.D. 1986. A hierarchical framework for stream
habitat classification: viewing streams in a watershed context. Environ. Manag. 10:199-214.
Hamelink, J.L., Landrum, P.P., Bergman., H.L., Benson, W.H., eds. 1994. Bioavialability:
Physical, Chemical and Biological Interactions. Lewis Publishers, Ann Arbor, MI.
Hanratty, M.P., Stay, F.S. 1994. Field evaluation of the littoral ecosystem risk assessment
model's prediction of the effects of chlorpyrifos. J. Appl. Ecol. 31:439-53.
Ho, K.T., Kuhn, A., Pelletier, M.C., Burgess, R.M., Helmstetter, A. 1999. Use of Ulva lactuca to
distinguish pH dependent toxicants in marine waters and sediments. Environ. Toxicol. Chem.
18:207-212.
Ho, K.T. et al. 2002. Methods for aquatic toxicity identification evaluations of freshwater and
marine sediments. EPA, Office of Research and Development, Atlantic Ecology Division/Mid-
Continent Ecology Division, Narragansett, RI/Duluth, MN. (in prep.)
Hunter, R.S., Niemi, G.J., Pilli, A., Veith, G.D. 1990. Aquatic information and rEtrieval
(AQUIRE) database system. In Pillman, W., ed., Envirotech Vienna 1990: Computer
Applications for Environmental Impact Analysis. International Society for Environmental
Protection, Vienna, Austria, pp. 42-48.
Leonard, E.N., Mount, D.R., Ankley, G.T. 1999. Modification of metal partitioning by addition
of synthetic AVS to freshwater sediments. Environ. Toxicol. Chem. 18:858-864.
Lewis, M.A., Weber, D.E., Stanley, R.S. 2000. Wetland plant seedlings as indicators of near-
coastal sediment quality: interspecific variation. Mar. Environ. Res. 50:535-540.
Lewis, M.A., Weber, D.E., Stanley, R.S., Moore, J.C. 2001. Relevance of rooted vascular plants
as indicators of estuarine sediment quality. Arch. Environ. Contain. Toxicol. 40:25-34.
Lores, E.M., Murrell, M.C., DiDonato, G.T., Stanley, R.S., Snyder, R.A., Sipura, J., Flemer,
D.A. 2002a. Effects of zooplankton grazing on phytoplankton communities in Escambia Bay, FL.
Mar. Ecol. Prog. Ser. 39 pp. (submitted.)
Lores, E.M., Lewis, M.A., Malaeb, Z.A. 2002b. Spatial and temporal variability in zooplankton
community dynamics in three urbanized bayous of the Pensacola Bay system, Florida, USA. Gulf
Mex. Sci. 21 pp. (submitted.)
Maxwell, J.R., Edwards, C.J., Jensen, M.E., Paustian, S.J., Parott, H., Hill, D.M. 1995. A
hierarchical framework of aquatic ecological units in North America (Nearctic Zone). Technical
report, USD A, Forest Service, NC.
McKee, P.M., Batterson, T.R., Dahl, T.E, Glooschenko, V., Jaworski, V., Pearce, J.B., Raphael,
C.N., Whillans, T.H., LaRoe, E.T. 1992. Great Lakes aquatic habitat classification based on
171
-------
wetland classification systems, Chapter 4. In Dieter, W. Busch, N, Sly, P.G., eds., The
Development of an Aquatic Habitat Classification System for Lakes. CRC Press, Ann Arbor, MI.
McMahon, G., Cuffney, T.F. 2000. Quantifying urban intensity in drainage basins for assessing
stream ecological conditions. J. Am. Water Resourc. Assoc. (in press).
Mitsch, W.J., Reeder, B.C. 1991. Modeling nutrient retention of a freshwater coastal wetland:
estimating the roles of primary productivity, sedimentation, resuspension and hydrology. Ecol.
Model. 54:151-187.
Mount, D.I. 1989. Methods for aquatic toxicity identification evaluations: phase HI toxicity
confirmation procedures. EPA/600/3-88/036. EPA, Office of Research and Development,
Environmental Research Laboratory, Duluth, MN.
Mount, D.I., Anderson-Carnahan, L. 1988. Methods for aquatic toxicity identification
evaluations: phase I toxicity characterization procedures. EPA/600/3-88/034. EPA, Office of
Research and Development, Environmental Research Laboratory, Duluth, MN.
Mount, D.I., Anderson-Carnahan, L. 1989. Methods for aquatic toxicity identification
evaluations: phase II toxicity identification procedures. EPA/600/3-88/035. EPA, Office of
Research and Development, Environmental Research Laboratory, Duluth, MN.
Mount, D.I., Norberg-King, T. 1993. Methods for aquatic toxicity identification evaluations:
phase HI toxicity confirmation procedures for samples exhibiting acute and chronic toxicity.
EPA/600/R-92/081. Environmental Research Laboratory, Duluth, MN.
Murrell, M.C., Stanley, R.S., Lores, E.M., DiDonato, G.T., Smith, L.M., Flemer, D.A. 2002.
Evidence that phosphorus limits phytoplankton growth in a Gulf of Mexico estuary: Pensacola
Bay, FL, USA. Mar. Ecol. Prog. Ser. (submitted).
Norberg-King, T., Mount, D., Durhan, E., Ankley, G., Burkhard, L, Amato, J., Lukasewycz, M.
Schubauer-Berigan, M.R., Anderson-Carnahan, L. 1991. Methods for aquatic toxicity
identification evaluations: phase I toxicity characterization procedures, 2nd ed. EPA/600/6-
91/003. Environmental Research Laboratory, Duluth, MN.
Norberg-King, T.J., Mount, D.I., Amato, J.R., Jensen, D.A., Thompson, J.A. 1992. Toxicity
identification evaluation: characterization of chronically toxic effluents, phase I. EPA/600/6-
91/005F. Environmental Research Laboratory, Duluth, MN.
Omernik, J.M. 1987. Ecoregions of the conterminous United States. Ann. Assoc. Am. Geogr.
77:118-125.
Palter, J.B., Dettmann, E.H. 1999. The effects of nitrogen loading and freshwater residence time
on the estuarine ecosystem. 15th Biennial International Conference of the Estuarine Research
Federation, September 25-30, 1999, New Orleans.
172
-------
Reynolds, K.M. 1999a. EMDS users guide (version 2.0): knowledge-based decision support for
ecological assessment. Gen. Tech. Rep. PNW-GTR-470. U.S. Department of Agriculture, Forest
Service, Pacific Northwest Research Station, Portland, OR. 63 pp.
Reynolds, K.M. 1999b. NetWeaver for EMDS version 1.0 user guide: a knowledge base
development system. Gen. Tech. Rep. PNW-GTR-471. U.S. Department of Agriculture, Forest
Service, Pacific Northwest Research Station, Portland, OR. 75 pp.
Richards, C., Host, G.E., Arthur, J.W. 1993. Identification of predominant environmental factors
structuring stream macro invertebrate communities within a large agricultural catchment.
Freshw. Biol. 29:285-294.
Rosgen, D.L. 1996. Applied river morphology. Wildland Hydrology, Pargosa Springs, CO.
Scheller, R.M., Snarski, V.M., Eaton, J.G., Oehlert, G. 1998. An analysis of the influence of
annual thermal variables on the occurrence of warm water fishes. Trans. Am. Fish. Soc. 128:257-
264.
Simon, T.F. 1999. Regional environmental monitoring and assessment (REMAP): evaluation of
the Great Lakes nearshore coastal wetlands: with emphasis on development of watershed
environmental indicators and status. R-EMAP proposal submitted to EPA, Mid-Continent
Ecology Division, Duluth, MN.
173
-------
Appendix 1. Critical path table identifying implementation stages and tasks for States from monitoring through diagnosis, associated uncertainties, research needs,
and research products (blue color refers to products being produced in whole or in part by other research areas).
Stage of Process
(3 03 (b)=>3 03 (d)=>TMDL=>
watershed restoration
Implementation steps
for States
Overall Uncertainties
Associated NHEERL
research needs
Associated research products
and form of tech transfer
Monitoring and Assessment
Part of APG 3: Provide framework
for incorporating diagnostic elements
into routine monitoring and
assessment activities, supporting
streamlining of State 303(b) and
303(d) listing processes (FY02-08)
Apply appropriate
monitoring design
for both assessment
of condition and
diagnosis
Joint monitoring design,
appropriate exposure metrics
for nontoxics1 , based on
appropriate model or
framework
Develop sampling design
(i.e., adequate power) and
monitoring program
indicators list for diagnostic
Guidance on joint monitoring
of stressor and response
endpoints at appropriate
scales2
use
Determine that
impairment exists
- Biological, based
on reference
conditions
- Chemical
- Physical/Habitat
Definition of reference
condition by habitat type3
Undeveloped or incomplete
criteria4
Development of tools for
determining reference
conditions in freshwater/
estuarine/marine benthos and
other ecological systems3
Guidance on classifying
reference conditions
Screening tools
Diagnostic
methods/indicators
Watershed scale
Ecosystem/habitat scale
Community scale
Diagnostic indicator
development
Watershed scale
Ecosystem scale
Community scale
Population scale5
Organism scale5
APM 2C: Guidance on
diagnostic indicators with
user-friendly interface/tools
for derivation (includes
mapped product: watershed
indicators)2
Evaluation of impairment decision
Part of APG 1: Development of
protocols for evaluating impairment
status in support of 303 (d) listing
process (FY02-03)
Application of
formal logic/decision
tree (e.g., field
reconnaissance)
Decision rules
- e.g., effect of stochastic
events
Methods for confirming
existence of threat or
impairment
SI Guidance Document (EPA
2000c)6
Protocol for confirmation of
impairment status in support
of 303(d) listing process
174
-------
Stage of Process
(3 03 (b)=>3 03 (d)=>TMDL=>
watershed restoration
Implementation steps
for States
Overall Uncertainties
Associated NHEERL
research needs
Associated research products
and form of tech transfer
Screening systems
Defining "similar" systems
for ranking of stressors
(Vollenweider-type
approach)
Building ranking
system/indices
Prototype for empirical
approach to ranking stressors
(nutrients)
Problem Formulation
APG 1: Provide conceptual
framework and information
management scheme for problem
formulation in 303(d) listing process,
including classification framework
for ecosystem to watershed to (eco-
regional scales)(FY02-04)
Derive list of
potential stressors
Background and
historic information
on potential stressors
Quality and utility of data
Development of database for
information supporting
diagnostic process (e.g.,
stressor-response data)7
APM 1A: Database structure
for supporting data (e.g.,
BASINS information) and
nation-wide database of
effects of non-toxic stressors
(e.g., nutrients, suspended
and bedded sediments)
Assess scale and
extent of impairment
Cumulative impacts
Approach or scheme for
performing diagnostic
assessments considering
scale, level of organization,
and stressor(s); Methods for
assessing cumulative impacts
SI Guidance Document (EPA
2000c)6
APM IB: Classification
scheme2
APM 1A: Conceptual
framework including cross-
scale issues both single and
multiple stressors (matrix
with multiple stressors)
Decision-tree for identifying
scale/extent
Guidance on methods for
assessing cumulative impacts
175
-------
Stage of Process
(3 03 (b)=>3 03 (d)=>TMDL=>
watershed restoration
Implementation steps
for States
Overall Uncertainties
Associated NHEERL
research needs
Associated research products
and form of tech transfer
Evaluate adequacy of
available stressor-
response
relationships
Missing stressor-response
relationships by
habitat/ecosystem/watershed/
regional class combination
Stressor-response
relationships2
Criteria guidance from other
research areas2
Prioritization scheme
for multiple TMDLs
Criteria for prioritization
Conceptual models
Strategy for linking impacts
across scales
Classification
scheme
Classification scheme
APM IB: Classification
frameworks2
Geographic regions
Watershed scale
Ecosystem/habitat scale
Application of methods and models
to diagnose cause(s) of impairment
APG 2: Provide diagnostic methods
and models for each of the major
classes of aquatic stressors (habitat
alteration, nutrients, suspended and
bedded sediments, and toxic
chemicals). These diagnostic tools
will be based on research that
identifies the causal mechanisms that
contribute to biological impairment
of marine and freshwater systems
(FY02-FY06)
Determine
predominant cause of
impairment
APG 2
Regionally-specific criteria
for nutrients, suspended and
bedded sediments, habitat
alteration, and toxic
chemicals (including
consideration of chronic
effects)
Methods for developing
regionally-specific criteria
for nontoxics8
Methods for assessing
chronic effects of toxic
chemicals9
Development of diagnostic
methods/tools for
ecosystem/community/
population and organism
scale5
Guidance for developing
regionally-specific criteria
for nontoxics8
Guidance for assessing
chronic effects of toxic
chemicals9
APM 2C: Guidance
document on methods with
case studies2
APM 2A: Guidance on whole
sediment TIEs
176
-------
Stage of Process
(3 03 (b)=>3 03 (d)=>TMDL=>
watershed restoration
Implementation steps
for States
Overall Uncertainties
Associated NHEERL
research needs
Associated research products
and form of tech transfer
APG 3: Provide diagnostic methods
and models for multiple stressors, to
allow allocation of cause among
multiple stressors, and evaluate the
effect of interactions among stressors
(FY02-08)
Allocate cause
among multiple
stressors
APG3
Methods to apportion
variance with known
uncertainty
Ecosystem models
Community-level:
multivariate statistics
Population-scale models
Organismal scale:
manipulative/experimental
approaches5
Evaluate effect of
interactions among
multiple stressors
APG3
Methods to evaluate if
interactions exist
Models
Experimental approaches
APM 3A: Case studies of
multivariate approaches to
community data analysis to
apportion cause among
stressors
Identify key combinations of
stressors expected to interact
(conceptual model)
APM 3B: Simulation of key
stressor interactions with
generic ecosystem models to
define range of interactions
Review of existing case
studies for evidence of
interactions
Decision-tree guiding
application of approaches
with case studies
APM 3C: Decision-support
system (long term)
APG 3: Develop diagnostic tools
which incorporate forecasting or
predictive approaches (FY04-08)
Forecasting or
predictive
approaches
APG4
Availability of adequate
validated models with
appropriate resolution and
extrapolation capability
Modeling framework for
watersheds and receiving
waters - forecasting,
interpretation of interactions,
indirect effects, time lags
Methods and models to
support alternative
remediation schemes to
achieve specific management
goals
Case studies
Comparison of approaches
177
-------
Stage of Process
(3 03 (b)=>3 03 (d)=>TMDL=>
watershed restoration
Implementation steps
for States
Overall Uncertainties
Associated NHEERL
research needs
Associated research products
and form of tech transfer
Confirmation of diagnosis and
evaluation of uncertainty
Part of APG 3: Provide methods for
confirmation of diagnosis, including
monitoring designs for assessment of
restoration success and uncertainty
analysis (FY01-08)
Weight-of-evidence
proof
Ranking procedures
Partitioning - single vs
multiple stressors
Quantification of uncertainty
Weighting scheme
Methods to assess
cumulative uncertainty
SI document (EPA 2000c)6
Guidance on weighting
scheme and methods to
assess cumulative uncertainty
Validation of models
Error analysis techniques for
complex models
Quantitative methods for
validating complex models
Test robustness of diagnostic
models over varying spatial,
temporal and biological
scales
Guidance on quantification
methods10
Training on use of simple
models
Determination of costs,
requirements for model
validation10
Confirmation of
diagnosis
Methods for assessing
environmental and/or
economic costs to determine
level of certainty needed
Environmental accounting
methods
Guidance on integration of
economic cost/benefit
analysis with environmental
accounting methods11
Controlled
experiments/
manipulations
Monitoring design for
assessment of restoration
success
Develop approaches for
validating diagnosis (e.g.,
pilot studies)
Guidance on monitoring
designs for watershed
restoration3
Need joint planning/input from NERL on exposure metrics
Joint product with other Aquatic Stressors Workgroups
Predominantly Goal 8
Aquatic Stressors Workgroups (habitat alteration, nutrients, suspended and bedded sediments, toxic chemicals)
Input from Aquatic Strressors Toxic Chemicals Workgroup (populations organismal scale) and Habitat Alteration Workgroup (population scale)
SI Group
Stressor-response data from other Aquatic Stressors Workgroups: database structure involves coordination between NHEERL-Diagnostics and OW for nation-
wide database of non-toxic stressors (e.g., nutrients, suspended and bedded sediments), coordination with NERL, OW, Regions for exposure data inventory
From Aquatic Stressors Nutrients, Suspended and Bedded Sediments Workgroups
From Aquatic Stressors Toxic Chemicals Workgroup
To be incorporated into APM on model products
Integration with groups doing Natural Resource Damage Assessments (Fish and Wildlife Service, NOAA) needed
178
-------
Appendix 2. Opportunities for interlaboratory collaboration.
Office/Entity
NCEA
NERL/NCEA
NERL/NCEA
NERL/NCEA
NCEA
NERL/NCEA
NCEA
NRMRL
OWOW
Project Description
Conceptual Modeling
Case study demonstrating the stressor identification process. This project identifies
the causes of biological impairment in the nation's water bodies.
Decision-support system development for causal evaluation.
Architecture for a decision support system (Casual Analysis and Diagnosis Decision
Information System CADDIS). This project is designed to assist users with
application of the 2000 stressor identification guidance document.
Decision support system(s) including forecasting of future cause-effect
relationships. Under this project research is conducted on approaches for
incorporating ecological risks and economics into a generalized decision-making
framework.
Land-cover/land use attributes and their relationship to cause of impairment.
Forecasting as a natural outgrowth of the application of stressor-response
relationships.
Linking research and diagnostic applications with management applications.
Consolidated Assessment and Listing Methodology (CALM).
APM/APG
FY02
APM
Goal
2(74)
FY03
APM
Goal 8(61)
FY07 APM
Point of Contact
Vic Serveiss
202-564-3251
Sue Norton
202-564-3246
Randy Bruins
513-569-7581
www.epa.gov.
owow/monitoring/calmprint.html .
179
------- |