Ecological Risk Assessment - Part 1 pdf

The emphasis is still on providingclear, scientifically sound, and unbiased technical advice to environmental decision makers.Although other examples are included in this edition, the fo

Risk Assessment

Second Edition

CRC Press is an imprint of the Taylor & Francis Group, an informa business Boca Raton London New York

CRC Press

Taylor & Francis Group

6000 Broken Sound Parkway NW, Suite 300

Boca Raton, FL 33487-2742

© 2007 by Taylor & Francis Group, LLC

CRC Press is an imprint of Taylor & Francis Group, an Informa business

No claim to original U.S Government works

Printed in the United States of America on acid-free paper

10 9 8 7 6 5 4 3 2 1

International Standard Book Number-10: 1-56670-634-3 (Hardcover)

International Standard Book Number-13: 978-1-56670-634-6 (Hardcover)

This book contains information obtained from authentic and highly regarded sources Reprinted material is quoted with permission, and sources are indicated A wide variety of references are listed Reasonable efforts have been made to publish reliable data and information, but the author and the publisher cannot assume responsibility for the validity of all materials or for the consequences of their use

No part of this book may be reprinted, reproduced, transmitted, or utilized in any form by any electronic, mechanical, or other means, now known or hereafter invented, including photocopying, microfilming, and recording, or in any informa- tion storage or retrieval system, without written permission from the publishers.

For permission to photocopy or use material electronically from this work, please access www.copyright.com (http:// www.copyright.com/) or contact the Copyright Clearance Center, Inc (CCC) 222 Rosewood Drive, Danvers, MA 01923, 978-750-8400 CCC is a not-for-profit organization that provides licenses and registration for a variety of users For orga- nizations that have been granted a photocopy license by the CCC, a separate system of payment has been arranged.

Trademark Notice: Product or corporate names may be trademarks or registered trademarks, and are used only for

identification and explanation without intent to infringe.

Library of Congress Cataloging-in-Publication Data

Ecological risk assessment / edited by Glenn W Suter II 2nd ed.

To my parents, Glenn W Suter and Kathleen T Suter

We are products of our heredity and environment, and parents provide all of one and most of the other.

Preface to the Second Edition

The primary purpose of preparing this edition is to provide an update In the 14 years sincethe first edition was published, ecological risk assessment has gone from being a marginalactivity to being a relatively mature practice There are now standard frameworks andguidance documents in the United States and several other countries Ecological risk assess-ment is applied to the regulation of chemicals, the remediation of contaminated sites, theimportation of exotic organisms, the management of watersheds, and other environmentalmanagement problems Courses in ecological risk assessment have been taught at severaluniversities As a result, there is a much larger literature to draw on, including many casestudies This is reflected both in the citation of ecological risk assessments published in theopen literature and in the use of more figures drawn from real assessments Hence, the readerwill notice a greater diversity in the graphical style, resulting from the many sources fromwhich figures have been drawn so as to give a flavor of the diverse practice of ecological riskassessment

The second edition also provides an opportunity for a new organization of the materialthat is more logically consistent In particular, whereas the first edition had separate chaptersfor types of ecological risk assessments (i.e., predictive, retrospective, regional, surveillance,and exotic organisms), this edition presents a unitary process of ecological risk assessmentthat is applicable to various problems, scales, and mandates All risk assessments are aboutthe future consequences of decisions Those that were described in the first edition asretrospective, following EPA terminology, are simply risk assessments that must begin with

an analysis of the current consequences of past actions in order to predict future consequences(Chapter 1)

Since 1992, ecological risk assessment has become sufficiently important to acquire criticsand opponents Some criticisms deal with aspects of the technical practice Ecologicalrisk assessment is often criticized for being based on inadequate data and models, for notaddressing large-scale spatial dynamics, and for using conservatism to compensate for thoseinadequacies (DeMott et al 2004; Landis 2005; Tannenbaum 2005a) Other critics areopposed to ecological risk assessment per se (Pagel and O’Brien 1996; Lackey 1997;O’Brien 2000; Bella 2002) These criticisms arise from a misperception of the nature andpurpose of risk assessment In particular, risk assessment is technical support for decisionmaking under uncertainty, but the critics hold risk assessment responsible for thedecision itself If decision makers listen to fishermen, loggers, chemical manufacturers, orutility companies more than to environmental advocates, critics say it is the fault of riskassessment If risk assessments are limited by regulatory context to considering only onealternative, they say that also is the fault of risk assessment If decisions are based onbalancing of costs and benefits, it is again the fault of risk assessment If the best availablescience does not address all of the important complexities of the system, they say that riskassessors who use that science are to blame Similarly, risk assessors are blamed when holisticproperties, endocrine disruptors, regional properties, or other favorite concerns are notaddressed Some of this criticism arises from an opposition to technology, science, and evenrationality, but more generally it is based on anger that the environment is not beingadequately protected One partial solution is to avoid the phrase ‘‘risk-based decision mak-ing.’’ Environmental decisions are, at best, ‘‘risk-informed.’’ They are based on risk informa-tion plus economic considerations, technical feasibility, public pressures, political pressures,

and the personal biases of the decision makers Another partial solution is to be fastidious inquantifying, or at least describing, uncertainties and limitations of our assessments.

Some things have not changed since the first edition The emphasis is still on providingclear, scientifically sound, and unbiased technical advice to environmental decision makers.Although other examples are included in this edition, the focus is still on risks from chemicals

or chemical mixtures, indicating that most ecological risk assessments are concerned withthese issues

The text is still aimed at practitioners and advanced students with at least a basic ledge of biology, chemistry, mathematics, and statistics It does not assume any familiaritywith ecological risk assessment or risk assessment in general A glossary is provided, becauseterms from risk assessment, ecology, toxicology, and other disciplines are used

know-As with the first edition, I have written most of the book myself in order to provide acommon voice and a common vision of the topic This is a service to the reader as well as anopportunity for me to share my particular vision of what ecological risk assessment is andwhat it could be However, for some major topics, the readers would be ill-served by mymeager expertise Fortunately, Larry Barnthouse, Steve Bartell, and Don Mackay agreed toparticipate in this edition as they did in the first I believe they are the preeminent experts inthe application of population modeling, ecosystem modeling, and chemical transport and fatemodeling, for the assessment of ecotoxicological effects Fortunately, they have similarpragmatic approaches to mine

The preface to the first edition described it as a manifesto The program of that manifestowas that ecological assessors must become more rigorous in their methods and practices inorder to be taken as seriously as human health and engineering risk assessors That program

is no longer needed Ecological risk assessments are at least as rigorous as humanhealth assessments and in some ways, particularly in the use of probabilistic analysis,ecological assessments are more advanced As a result, ecological risks are more often thebasis for environmental regulatory and management decisions However, ecologically drivendecisions are still far less common than health-driven decisions To a certain extent, this isinevitable, because humans are making the decisions based on the concerns of other humans,the public However, we can make progress in protecting the nonhuman environment bygreater integration of ecological risk assessment with concerns for human health and welfare.Hence, the greatest challenge in the coming years is to estimate and communicate ecologicalrisks in a way that makes people care

Glenn SuterCincinnati, Ohio

I gratefully acknowledge the innumerable environmental scientists who contributed to thistext Those who are cited are thereby acknowledged, although you are probably not cited asmuch as you deserve Many of you who are not cited at all deserve citation but must settle forthis apologetic acknowledgment I have heard your talks at meetings, exchanged ideas at yourposters or in the halls, and even read your papers, but have forgotten that you were the source

of those ideas Even more sadly, many of you have done important work and producedimportant ideas that should appear in this text but do not, because I am unaware of them.There are forlorn piles of books, reports, and reprints on the table behind my back as I writethis that I really wanted to read before completing this book, but could not So, if you feelthat I have not given your work the attention it deserves, you are probably right

Parts of this book draw upon material in Ecological Risk Assessment for ContaminatedSites Thanks to Rebecca Efroymson, Brad Sample, and Dan Jones who were coauthors ofthat book

My 7 years with the US Environmental Protection Agency have improved this book bygiving me a deeper understanding of the role of risk assessment in environmental regulation.Thanks to all of my agency colleagues Particular thanks to Susan Cormier and Susan Nortonwho have been wonderful friends, inspiring collaborators, and guardians against sloppythinking

Finally, deep thanks to Linda who, after decades of marriage, has learned to tolerate mylong hours in my study and even helped with the final rush to submit the manuscript

Glenn W Suter IIis science advisor in the US

Environmental Protection Agency’s National

Center for Environmental Assessment,

Cincin-nati, and was formerly a senior research staff

member in the Environmental Sciences

Div-ision, Oak Ridge National Laboratory, United

States He has a PhD in ecology from the

University of California, Davis, and 30 years

of professional experience including 25 years in

ecological risk assessment He is the principal

author of two texts in the field of ecological risk

assessment, editor of two other books, and

author of more than 100 open literature

publi-cations He is associate editor for ecological

risk of Human and Ecological Risk Assessment,

and reviews editor for the Society for

Environ-mental Toxicology and Chemistry (SETAC)

He has served on the International Institute

of Applied Systems Analysis Task Force on

Risk and Policy Analysis, the Board of Directors of SETAC, an expert panel for theCouncil on Environmental Quality, and the editorial boards of Environmental Toxicologyand Chemistry, Environmental Health Perspectives, and Ecological Indicators He isthe recipient of numerous awards and honors; most notably, he is an elected fellow ofthe American Association for the Advancement of Science and he received SETAC’sGlobal Founder’s Award, its highest award for career achievement, and the EPA’s Level 1Scientific and Technical AchievementAward

His research experience includes development

and application of methods for ecological risk

assessment and ecological epidemiology,

de-velopment of soil microcosm and fish toxicity

tests, and environmental monitoring His

workis currently focused on the development

of methods for determining the causes of

biological impairments

Susan M Cormieris a senior science advisor in

the U.S Environmental Protection Agency’s

National Risk Management Research

Labora-tory Dr Cormier received her BA in Zoology

from the University of New Hampshire, her

MA in biology from the University of South

Florida, and her PhD in Biology from Clark

University

Donald Mackay (BSc, PhD (Glasgow)) is

director of the Canadian Environmental

Modelling Centre at Trent University,

Peterbor-ough, Ontario, Canada He graduated in

chem-ical engineering from the university of Glasgow

After working in the petrochemical industry he

joined the University of Toronto, where he is

now Professor Emeritus in the Department of

Chemical Engineering and Applied Chemistry

He has been director of the Canadian

Environ-mental Modelling Centre of Trent University

Ontario since 1995 His primary research

inter-est is the development, application, validation,

and dissemination of mass-balance models

de-scribing the fate of chemicals in the environment

in general, and in a variety of specific

environ-ments These models include descriptions of

bioaccumulation in a variety of organisms,

water-quality models of contaminant fate in lakes, rivers, sewage-treatment plants, and insoils and vegetation He has developed a series of multimedia mass-balance models employingthe fugacity concept that are widely used for assessment of chemical fate in national regions inthe global environment A particular interest is the

transport of persistent organic chemicals to cold

climates such as the Canadian Arctic and their

accumulation and migration in arctic ecosystems

Susan B Nortonis a senior ecologist in the U.S

Environmental Protection Agency’s National

Center for Environmental Assessment Since

joining EPA in 1988, Dr Norton has developed

methods and guidance to better use ecological

knowledge to inform environmental decisions

She was an author of many agency guidance

documents including the 2000 Stressor

Identifi-cation Guidance document, the 1998 Guidelines

for Ecological Risk Assessment, the 1993

Wild-life Exposure Factors Handbook, and the 1992

Framework for Ecological Risk Assessment She

has published numerous articles on ecological

assessment and edited the book Ecological

As-sessment of Aquatic Resources: Linking Science to Decision-Making She is currently astic about making methods and information for causal analysis more available via the WorldWide Web at www.epa.gov=caddis Dr Norton received her BS in plant science from PennState, her MS in natural resources from Cornell University, and her PhD in environmentalbiology from George Mason University

enthusi-Neil Mackay(BSc (Waterloo), DPhil (York)) is a senior research scientist for environmentalmodelling with DuPont (UK) Limited As a member of the DuPont Crop Protection GlobalModelling Team he is active in strategic development and regulatory exposure and riskassessment activities Previous work experience includes employment as a consultant to

both industry and government bodies, primarily

in Europe He was a participant in the European

Commission Health and Consumer Protection

Directorate General and the FOCUS Risk

As-sessment Working Group and is a member of

the UK government expert advisory panel

on veterinary medicines Particular interests

in-clude aquatic risk assessment and use of spatial

tools (GIS and remote sensing methods) to

evaluate risks at various scales (field, catchment

and regional scales) and assessment of long

range transport potential for persistent organic

pollutants (POPs)

Lawrence W Barnthouse is the president of

LWB Environmental Services, Inc and adjunct

associate professor of zoology at Miami

Univer-sity He was formerly a senior research staff

member and group leader in the Environmental

Sciences Division at Oak Ridge National

Laboratory In 1981 he became co-principal

investigator (with Glenn Suter) on EPA’s first

research project on ecological risk assessment

Since that time, he has been active in the

devel-opment and application of ecological risk

assessment methods for EPA, other federal

agencies, state agencies, and private industry

He has chaired workshops on ecological risk

assessment for the National Academy of

Sci-ences and the Society of Environmental

Toxi-cology and Chemistry, and served on the peer

review panels for the EPA’s Framework for

Ecological Risk Assessment and the Guidelines

for Ecological Risk Assessment He continues

to support the development of improved

methods for ecological risk assessment as the

Hazard=Risk Assessment Editor of

Environmen-tal Toxicology and Chemistry and a Founding

Editorial Board Member of Integrated

Environ-mental Assessment and Management

Steven M Bartell is a principal with E2 Consulting Engineers, Inc He is also an adjunctfaculty member in the Department of Ecology and Evolutionary Biology at the University

of Tennessee, Knoxville His education includes a PhD in oceanography and limnology(University of Wisconsin 1978), an MS in botany (University of Wisconsin 1973), and a BA

in biology (Lawrence University 1971) Dr Bartell’s areas of expertise include systemsecology, ecological modeling, ecological risk analysis, risk-based decision analysis, vulner-ability analysis, numerical sensitivity and uncertainty analysis, environmental chemistry, andenvironmental toxicology He works with a variety of public and private sector clients

in diverse projects in ecological risk assessment, environmental analysis, and more recently

in ecological planning and

restor-ation in the context of adaptive

en-vironmental management and

ecological sustainability Bartell

has authored more than 100

peer-reviewed publications He is a

se-nior contributing author on several

books including Ecological

Model-ing in Risk Assessment (2001),

Ecological Risk Assessment

Deci-sion-Support System: A Conceptual

Design (1998), Risk Assessment and

Management Handbook for

Envir-onmental, Health, and Safety

Pro-fessionals (1996), and Ecological

Risk Estimation (1992) He

cur-rently serves on the editorial boards

of Aquatic Toxicology and

Chemo-sphere having served previously on the editorial boards of Human and Ecological RiskAssessment and Ecological Applications Bartell served for 11 years on the USEPA ScienceAdvisory Board, mainly on the Environmental Processes and Effects Committee and reviewcommittees

1.3.5 Remediation and Restoration

1.3.6 Permitting and Managing Land Uses1.3.7 Species Management

Chapter 2 Other Types of Assessments

2.1 Monitoring Status and Trends

2.11 Health Risk Assessment

2.12 Environmental Impact Assessment

3.2.3 Ecological Epidemiology

3.2.4 Causal Chain Framework

3.3 Extended Frameworks

3.4 Iterative Assessment

3.4.1 Screening vs Definitive Assessments

3.4.2 Baseline vs Alternatives Assessments

3.4.3 Iterative Assessment as Adaptive Management3.5 Problem-Specific Frameworks

4.3.1.2 Developing the List

4.3.1.3 Developing Maps and Conceptual Models4.3.2 Analyzing the Evidence

4.3.4 Iteration of Causal Analysis

4.4 Identifying Sources and Management Alternatives

4.5 Risk Assessment in Ecoepidemiology

5.1.3 Variability Uncertainty Dichotomy

5.1.4 Combined Variability and Uncertainty

5.3 Ways to Analyze Probabilities

5.3.1 Frequentist Statistics

5.3.2 Bayesian Statistics

5.3.3 Resampling Statistics

5.3.4 Other Approaches

5.4 Why Use Probabilistic Analyses?

5.4.1 Desire to Ensure Safety

5.4.2 Desire to Avoid Excessive Conservatism

5.4.3 Desire to Acknowledge and Present Uncertainty5.4.4 Need to Estimate a Probabilistic Endpoint

5.4.5 Planning Sampling and Testing

5.4.6 Comparing Hypotheses and Associated Models5.4.7 Aiding Decision Making

5.5.8 Listing and Qualitative Evaluation

5.6 Probability in the Risk Assessment Process

5.6.1 Defining Exposure Distributions

5.6.2 Defining Effects Distributions

5.6.3 Estimating Risk Distributions

5.7 Parameters to Treat as Uncertain

6.4.7 What to do with Multiple Dimensions?

Chapter 7 Modes and Mechanisms of Action

7.1 Chemical Modes and Mechanisms

7.2 Testing for Mechanisms

7.3 Nonchemical Modes and Mechanisms

Chapter 8 Mixed and Multiple Agents

8.1 Chemical Mixtures

8.1.1 Methods Based on Whole Mixtures

8.1.2 Methods Based on Tests of Components

8.1.2.1 Simple Similar Action and Concentration Addition8.1.2.2 Independent Action and Response Addition8.1.2.3 Interactive Action

8.1.2.4 Multiple Chemicals and Multiple Species

8.1.3 Integration of Complex Chemical Mixtures

8.2 Multiple and Diverse Agents

8.2.1 Categorize and Combine Agents

8.2.2 Determine Spatial and Temporal Overlap

8.2.3 Define Effects and Mode of Action

8.2.4 Screen Effects

8.2.5 Simple Additive Effects

8.2.6 Additive Exposures

8.2.7 Mechanistic Models of Combined Effects

8.2.8 Integration of Complex Sets of Agents and ActivitiesChapter 9 Quality Assurance

9.1 Data Quality

9.1.1 Primary Data

9.1.2 Secondary Data

9.1.3 Defaults and Assumptions

9.1.4 Representing Data Quality

Part II Planning and Problem Formulation

Chapter 10 Impetus and Mandate

Chapter 11 Goals and Objectives

Chapter 12 Management Options

Chapter 13 Agents and Sources

13.1 Emissions

13.2 Activities and Programs

13.3 Sources of Causes

13.4 Properties of the Agent

13.5 Sources of Indirect Exposure and Effects

13.6 Screening Sources and Agents

Chapter 14 Environmental Description

Chapter 15 Exposure Scenarios

Chapter 16 Assessment Endpoints

16.1 Assessment Endpoints and Levels of Organization16.2 Generic Assessment Endpoints

16.2.1 Generic Endpoints Based on Policy Judgments16.2.2 Functionally Defined Generic Endpoints

16.2.3 Applying Generic Endpoints

16.3 Making Generic Assessment Endpoints Specific

16.4 Endpoints Based on Objectives Hierarchies

Chapter 17 Conceptual Models

17.1 Uses of Conceptual Models

17.2 Forms of Conceptual Models

17.3 Creating Conceptual Models

17.4 Linkage to Other Conceptual Models

Chapter 18 Analysis Plans

18.1 Choosing Measures of Exposure, Effects,

and Environmental Conditions

18.2 Reference Sites and Reference Information

18.2.1 Information Concerning the Precontamination or

Predisturbance State18.2.2 Model-Derived Information

18.2.3 Information Concerning Other Sites

18.2.4 Information Concerning a Regional Reference18.2.5 Gradients as Reference

18.2.6 Positive Reference Information

18.2.7 Goals as an Alternative to Reference

Part III Analysis of Exposure

Chapter 19 Source Identification and Characterization

19.1 Sources and the Environment

19.2 Unknown Sources

19.3 Summary

Chapter 20 Sampling, Analysis, and Assays

20.1 Sampling and Chemical Analysis of Media

20.2 Sampling and Sample Preparation

20.9 Biota and Biomarkers

21.2.2 Point and Nonpoint Sources

21.2.3 Steady-State and Non-Steady-State Sources

21.3.5 Solutions to the Mass Balance Equation

21.3.6 Complexity, Validity, and Confidence Limits

21.4 Illustration of a Simple Mass Balance Model

21.4.1 The System Being Modeled

21.4.2 Concentration Calculation

21.4.2.1 Chemical Input Rate21.4.2.2 Partitioning between Water, Particles, and Fish21.4.2.3 Outflow in Water

21.4.2.4 Outflow in Particles21.4.2.5 Reaction

21.4.2.6 Deposition to Sediment21.4.2.7 Evaporation

21.4.2.8 Combined Loss Processes21.4.3 Fugacity Calculation

21.4.4 Discussion

21.5 Chemicals of Concern and Models Simulating their Behavior21.5.1 General Multimedia Models

21.5.1.1 Level I21.5.1.2 Level II21.5.1.3 Level III21.5.1.4 Level IV21.5.1.5 Fugacity Models21.5.1.6 CalTOX Model21.5.1.7 Simplebox Model21.5.1.8 Regional, Continental, and Global-Scale Models21.5.2 Models Specific to Environmental Media

21.5.2.1 Plume Models in General21.5.2.2 Atmospheric Models21.5.2.3 Aquatic Models21.5.2.4 Soil Models21.5.2.5 Fish Uptake and Food Chain Models21.5.2.6 Miscellaneous Models

21.5.3 Models Specific to Chemical Classes

21.5.3.1 Agricultural Pesticides21.5.3.2 Veterinary Medicines21.5.3.3 Biocides

21.5.3.4 Metals21.6 Concluding Thoughts on Selecting and Applying Models

Chapter 22 Exposure to Chemicals and Other Agents

22.1 Exposure Models

22.2 Exposure to Chemicals in Surface Water

22.3 Exposure to Chemicals in Sediment

22.4 Exposure to Contaminants in Soil

22.4.1 Chemical Analyses to Estimate Exposure

22.4.1.1 Partial Chemical Extraction and Normalization22.4.1.2 Input Form of the Chemical

24.4.1.3 Chemical Interactions

24.4.1.4 Nonaqueous Phase Liquids

22.4.2 Soil Depth Profile

22.5 Exposure of Terrestrial Plants

22.5.1 Rooting Depth

22.5.2 Rhizosphere

22.5.3 Wetland Plant Exposures

22.5.4 Soil Properties and Exposure of Plants

22.5.5 Plant Interspecies Differences

22.5.6 Plant Exposure in Air

22.6 Exposure of Soil Invertebrates

22.6.1 Depth of Exposure and Ingested Material

22.6.2 Soil Properties and Chemical Interactions

22.7 Exposure of Soil Microbial Communities

22.8.1.4 Spatial Issues in Wildlife Exposure

22.8.1.5 Temporal Issues in Wildlife Exposure

22.8.1.6 Exposure Modifying Factors

22.8.2 Parameters for Estimation of Exposure

22.8.2.1 Body Weight

22.8.2.2 Food and Water Consumption Rates

22.8.2.3 Inhalation Rates

22.8.2.4 Soil and Sediment Consumption

22.8.2.5 Home Range and Territory Size

22.9 Uptake Models

22.9.1 Aquatic Organism Uptake

22.9.1.1 Neutral Organics

22.9.1.2 Ionizing Organic Chemicals

22.9.1.3 Inorganic and Organometalic Chemicals22.9.1.4 Aquatic Plants

22.9.1.5 Aquatic Toxicokinetics

22.9.2 Benthic Invertebrate Uptake

22.9.3 Terrestrial Plant Uptake

22.9.3.1 Soil Uptake22.9.3.2 Empirical Models of Inorganic Chemicals22.9.3.3 Empirical Models for Organic Chemicals22.9.3.4 Surface Contamination

22.9.3.5 Plant Tissue Type22.9.3.6 Mechanistic Models22.9.4 Earthworm Uptake

22.9.5 Terrestrial Arthropod Uptake

22.9.6 Terrestrial Vertebrate Uptake

22.10 Exposure to Petroleum and other Chemical Mixtures22.11 Exposure to Natural Extreme Events

22.12 Exposure to Organisms

22.13 Probability and Exposure Models

22.14 Presenting the Exposure Characterization

Part IV Analysis of Effects

Chapter 23 Exposure–Response Relationships

23.2.1 Thresholds and Benchmarks

23.2.2 Time as Exposure and Response

23.2.3 Combined Concentration and Duration

24.2.4 Oral and Other Wildlife Exposures

24.3 Microcosms and Mesocosms

24.4 Effluent Tests

24.5 Media Tests

24.5.1 Contaminated Water Tests

24.5.2 Contaminated Sediment Tests

24.5.3 Contaminated Soil Tests

24.5.4 Ambient Media Tests with Wildlife

24.6 Field Tests

24.6.1 Aquatic Field Tests

24.6.2 Field Tests of Plants and Soil Organisms

24.6.3 Wildlife Field Tests

24.7 Testing Organisms

24.8 Testing Other Nonchemical Agents

24.9 Summary of Testing

Chapter 25 Biological Surveys

25.1 Aquatic Biological Surveys

25.1.1 Periphyton

25.1.2 Plankton

25.1.3 Fish

25.1.4 Benthic Invertebrates

25.2 Terrestrial Biological Surveys

25.2.1 Soil Biological Surveys

25.2.2 Wildlife Surveys

25.2.3 Terrestrial Plant Surveys

25.3 Physiological, Histological, and Morphological Effects

25.4 Uncertainties in Biological Surveys

25.5 Summary

Chapter 26 Organism-Level Extrapolation Models

26.1 Structure–Activity Relationships

26.1.1 Chemical Domains for SARs

26.1.2 Approaches for SARs

26.1.3 State of SARs

26.2 Effects Extrapolation Approaches

26.2.1 Classification and Selection

26.2.8 Toxicokinetic Modeling for Extrapolation

26.2.9 Multiple and Combined Approaches

26.3 Extrapolations for Particular Biotas

Chapter 27 Population Modeling

27.1 Basic Concepts and Definitions

27.1.1 Population-Level Assessment Endpoints

27.1.2 Implications of Life History for Population-Level

Ecological Risk Assessment

27.1.3 Representation and Propagation of Uncertainty

27.1.4 Density Dependence

27.2 Approaches to Population Analysis

27.2.1 Potential Population Growth Rate

27.2.2 Projection Matrices

27.2.3 Aggregated Models

27.2.4 Metapopulation Models

27.2.5 Individual-Based Models

27.3 Applications to Toxic Chemicals

27.3.1 Quantifying Uncertainties in Individual-to-Population Extrapolations27.3.2 Life History–Based Ecological Risk Assessment

27.3.3 Quantifying Impacts of Chemical Exposures on Risk of Extinction27.3.4 Quantifying Impacts of Chemicals on Metapopulations

27.3.5 Individual-Based Models

27.4 Future of Population Modeling in Ecological Risk Assessment

Chapter 28 Ecosystem Effects Modeling

28.1 An Ecosystem Paradigm

28.2 Ecosystem Risk Assessment

28.2.1 Ecosystem Assessment Endpoints

28.3 Ecosystem Simulation Modeling

28.3.1 Physical Ecosystem Models

28.3.2 Ecosystem Network Analysis

28.5 Innovations in Ecosystem Modeling

28.5.1 Structurally Dynamic Models

28.5.2 Interactive Modeling Platforms

28.5.3 Network-Enabled Ecosystem Models

28.5.4 Ecosystem Animation

28.6 Ecosystem Models, Risk Assessment, and Decision making

28.6.1 Model Results and NOECs

28.6.2 Atrazine Levels of Concern

28.7 Models or Modelers

Part V Risk Characterization

Chapter 29 Criteria and Benchmarks

29.2.5 Test Endpoints with Safety Factors

29.2.6 Distributions of Effects Levels

29.2.7 Equilibrium Partitioning Benchmarks

29.2.8 Averaged Values as Benchmarks

29.2.9 Ecoepidemiological Benchmarks

29.2.10 Summary of Screening Benchmarks

Chapter 30 Integrating Exposure and Exposure–Response

30.1 Quotient Methods

30.2 Exposure is Distributed and Response is Fixed

30.3 Both Exposure and Response are Distributed

30.4 Integrated Simulation Models

30.5 Integration of Sense and Nonsense

30.6 Integration in Space

30.7 Examples

30.7.1 Shrews on a Mercury-Contaminated Site

30.7.2 Egrets and Eagles in South Florida

30.7.3 Egrets and Herons in Hong Kong

30.7.4 Bioaccumulative Contaminants in a Stream

30.7.5 Secondary Poisoning in Hawaii

30.7.6 Atrazine

30.7.7 Warming Subalpine Forests

30.8 Summary

Chapter 31 Screening Characterization

31.1 Screening Chemicals and Other Agents

31.2.1 Screening Chemicals at Sites

31.2.1.1 Screening Against Background

31.2.1.2 Screening Against Detection Limits

31.2.1.3 Screening Against Waste Constituents

31.2.1.4 Screening Against Physical–Chemical Properties31.2.1.5 Screening Against Ecotoxicological Benchmarks31.2.1.6 Screening Species Against Area

31.2.2 Exposure Concentrations for Sites

31.2.3 Screening Media

31.2.4 Screening Receptors

31.2.5 Screening Sites

31.2.6 Data Adequacy and Uncertainties

31.2.7 Presentation of a Site Screening Assessment

31.3 Examples

Chapter 32 Definitive Risk Characterization by Weighing the Evidence

32.1 Weighing Evidence

32.2 Sediment Quality Triad: A Simple and

Clear Inference Method

32.3 Inference to the Best Conclusion at Contaminated Sites

32.3.1.5 Body Burdens of Endpoint Organisms

32.3.2 Ambient Media Toxicity Tests

32.4.1 Characterizing Contaminated Site Risks

32.4.2 Characterizing Contaminated Sediment Risks

32.4.3 Characterizing Wildlife Risks

32.4.4 Characterizing Pesticide Risks

32.4.5 Characterizing Effluent Risks

32.5 Interpretation

Chapter 33 Comparative Risk Characterization

33.1 Methods of Comparative Risk Characterization

33.1.1 Risk Ranking

33.1.2 Risk Classification

33.1.3 Relative Risk Scaling

33.1.4 Relative Risk Estimation

33.1.5 Net Environmental Benefits Analysis

33.1.6 Economic Units

33.1.7 Reporting Comparative Risk

33.2 Comparison and Uncertainty

34.5 Limitations

34.6 Conclusions

Part VI Risk Management

Chapter 35 Reporting and Communicating Ecological Risks

35.1 Reporting Ecological Risks

35.2 Communicating Ecological Risks

Chapter 36 Decision Making and Ecological Risks

36.1 Preventing Exceedence of Standards

36.2 Preventing Adverse Effects

36.8 Miscellaneous and Ad Hoc Considerations

Chapter 37 Integration of Human Health Risk Assessment

37.1 Wildlife as Sentinels

37.2 Integrated Analysis of Human and Ecological Risks

37.2.1 Coherent Expression of Assessment Results

38.2 Ecological Risk and Economics

38.3 Ecological Risk and Ethics

38.4 Ecological Risk, Stakeholder Preferences,

and Public Opinion

38.5 Conclusions

Chapter 39 Monitoring the Results of Risk Management

Part VII The Future of Ecological Risk Assessment

Glossary

References

1 Defining the Field

Risk assessment is the product of a shotgun wedding between science and the law

William Ruckelshaus

‘‘Technical support for decision making under uncertainty’’ is the only definition of riskassessment that describes its many uses As Bernstein (1996) plausibly argues, the use ofrational methods for dealing with the uncertain future in place of prayers, prophecies,traditions, auguries, and hunches is the hallmark of modern culture Risk assessment beganwith the need to calculate odds for gamblers, and subsequently, in seventeenth-centuryEngland and the Netherlands, with the need to determine premiums on annuities and theprobability that a ship sent on a trading voyage would return successfully (Hacking 1975;Bernstein 1996) Most risk assessors are still involved in finance and insurance (Melnikov2003) Risk assessment has since spread to many spheres of human endeavor includingengineering, wildfire management, medicine, and environmental regulation The generaldefinition indicates that two features are common to all of these enterprises: a decision to

be made and uncertainty concerning outcomes

The conventional, objectivist definition of risk is: a combination of the severity (nature andmagnitude) and the probability of effects from a proposed action Severity may be variouslydescribed depending on the situation, e.g., the number of deaths, the reduction in abundance,and the reduction in areal extent Probability may be derived from an estimate of the frequency

of an effect among individuals in an exposed population or a hypothetical frequency of effects ifthe same decisions were made multiple times For example, a risk might be a 0.3 annualfrequency of mass mortalities in an exposed population, or a probability that an effluent willreduce the number of fish species in a lake by as much as 15% Alternatively, risk may bedefined subjectively as a state of mind of an individual making an uncertain decision or of thoseexposed to the consequences of a decision This subjective risk is an important issue whenassessing risks to humans, who are subject to anxiety and dread, but is less relevant to the topic

of this book Note that subjective risk is fundamentally different from the Bayesian, subjectiveinterp retation of probabil ity in estimat es of objective risk (Chapter 5)

The terms ‘‘environmental risk’’ and ‘‘ecological risk’’ can cause confusion because of theirsimilarity In the United States, the term environmental risk has been used to describe risks tohumans due to contaminants in the environment Ecologists subsequently invented the termecological risk to refer to risks to nonhuman organisms, populations, and ecosystems (Barnt-house and Suter 1986) However, the term environmental risk is commonly used in Europe inthe way that ecological risk is used in the United States

The decision to be supported is too often neglected in risk assessment ‘‘It is hard to imaginerisk analysis existing without the need for decisions, without the need for a systematicapproach to aiding those who make decisions’’ (Crawford-Brown 1999) Yet, the most influ-ential guidance for environmental risk assessment (ERA) has stressed the need to isolaterisk assessors from the influence of decision makers in order to avoid bias (NRC 1983)

ERA practice has tended to emphasize the risk assessment process in the abstract without agrounding in a decision-making process For example, due to the peculiarities of Superfundregulations, the guidance for baseline ERA at contaminated sites in the United States does notaddress the consequences of remedial decisions (Sprenger and Charters 1997) This situation ischanging with the realization that the highest-quality assessment is worthless if it does notaddress the needs of the decision maker (National Research Council 1994; The Presidential=Congressional Commission on Risk Assessment and Risk Management 1997) Risk-basedenvironmental decisions generally fall into three categories: should we permit x (e.g., use of anew chemical, release of an effluent, or increased harvest of a resource); what should we doabout x (e.g., remediate, treat, or restore); should we do x, y, or z (e.g., which pest managementmethod poses the least risk)?

Pro bability, the other core concep t in ERA, has also been surprisingl y neglect ed Theprob abilities that charact erize risks may result from variab ility or uncerta inty (Chapter 5).Although quantitative methods for analyzing uncertainty and variability in terms of prob-ability have existed for centuries, most ERAs treat them qualitatively or in nonprobabilisticterms This does not mean that uncertainty and variability are ignored or that, as some havecontended, most current risk assessments are not truly assessments of risk Rather, they areoften dealt with by semiquantitative precautionary practices That is, conservative assump-tions and safety factors have been assumed to provide sufficient safety to avoid the need for

a formal probabilistic analysis However, formal probabilistic analysis of uncertainty isincreasingly common This is because the semiquantitative practices are subject to criticismthat they are insufficiently precautionary, excessively precautionary, or precautionary to anundefined degree

Risk assessment uses science, but is not science in the conventional sense, i.e., it does notseek to develop new theories or general knowledge It rather uses scientific knowledge andtools to generate information that is useful for a specific purpose In this sense, risk assessorsare like engineers, and in fact much of the practice of ERA has been developed by engineers(see, e.g., Haimes 1998) However, contrary to some critics, risk assessment is based predom-inantly on factual information and scientific theory, and is not simply a scientific smokescreen for policy Typically, risk assessments and their components are intensely and publiclyreviewed and are often challenged in court As a result, the use of bad science to justify apreordained decision is likely to be detected in contentious cases

1.1 PREDICTIVE VS RETROSPECTIVE RISK ASSESSMENT

The EPA’s framework and guidelines for ecological risk assessment and the previous edition ofthis text distinguish retrospective from predictive risk assessment This distinction has createdsome confusion, because it is nonsensical to speak of risks of events in the past This texteliminates that distinction and focuses instead on the decision-supporting function of riskassessment Hence, when assessing risks from spills or other past events, we are assessing risksassociated with future consequences of those events They include ongoing toxic effects, thespread of toxic levels of contaminants to other areas, loss of habitat due to failed restoration, andother sequela Even when performing assessments to set monetary damages for past actions, weare not assessing risks of past events For example, during the Exxon Valdez oil spill, a certainnumber of sea otters, bald eagles, and other wildlife were killed The natural resource damageassessment (Section 1.3.8) did not assess risks to those organisms Rather, to the extent that theuncertainties in the damage assessment could be interpreted as risks, they should be interpreted asrisks that the level of monetary damages assessed will be either insufficient or excessive withrespect to the cost of restoration of the populations of those species, making good lost services

of nature, and otherwise remediating ecological and economic injuries

Although all risk assessments are in some sense predictive, it does not mean that tion concerning the past is irrelevant Analyses of such data may be used to help formulate theassessment problem, elucidate trends that may extend into the future, identify causal rela-tionships between agents and injuries, and define a baseline for remediation and restoration.

informa-By analogy to human health epidemiology, analyses of past ecological effects and their causesare termed ecologi cal ep idemiology (C hapter 4) Hence, what the US Environme ntal Protec-tion Agency (US EPA) terms retrospective assessments should be thought of as predictiveassessments of the future consequences of past actions

1.2 RISKS, BENEFITS, AND COSTS

When assessing an action, it may be necessary to consider the risks associated with the action,the potential benefits, and the costs of carrying out the action For example, when consideringwhether to apply a remedial technology to a contaminated site, it is important to considerthe risks of ecological injury from the contaminants and from the remedial action itself(e.g., injuries to benthic communities due to dredging), the benefits of the action (e.g., reducedcontaminant risks to epibenthic fish), and the cost of carrying out the remediation Which ofthese dimensions is formally analyzed and how they are compared depends on the context.Some laws require consideration of costs while others do not allow it Further, there isconsiderable variation in whose benefits, costs, and risks are considered For example, theregistration of pesticides or biocontrol agents in the United States may consider costs andbenefits to farmers if there are no good alternatives to the pesticide in question, but not thecosts to the manufacturer Although the primary focus of regulatory agencies is on the risksfrom new chemicals, effluents, spills, and exotic organisms, the consideration of benefits andcosts as well as ethical concerns and public preferences can provide a more complete basis fordecisio n making (Chapter 36) Althou gh costs to the regulated party can be readily identifi edand relatively easily estimated, the benefits to the environment are always incompletelyidentified and are difficult to quantify Hence, cost–benefit analysis tends to be biased againstenvironmental protection

1.3 DECISIONS TO BE SUPPORTED

The form and content of an ecological risk assessment is determined by the decision to besupported This is true not only because different decisions require different sorts of informa-tion, but also because of the different formal and informal traditions and constraints that havedeveloped in the various decision-making cultures For example, the assessments of newindustrial chemicals in the United States must be performed within 90 days and normallycannot demand data generation On the other hand, assessments of contaminated sites mayrequire years of effort involving expensive field surveys, sample collection, analysis, and testing.1.3.1 PRIORITIZATION OFHAZARDS

In the United States and many other countries, priorities for environmental managementhave been set in a highly inconsistent manner, based primarily on public pressures astranslated by legislators into laws and budgets Because resources for environmental man-agement are limited, it would be desirable to devote them to the highest priority hazardsrather than the ones that were given a strong mandate some decades ago This concept ofusing risk assessment to prioritize hazards is appealing (Grothe et al 1996), but controversial

in practice (Finkel and Golding 1995)

The US EPA and its Science Advisory Board have performed comparative risk assessmentsfor the purpose of prioritization (SAB 1990, 2000; MADEP 2002) The results have beencontroversial and have not greatly influenced the EPA’s regulatory and management prac-tices However, they have led to the development of guidance for comparative assessment and

to the performance of such assessments in most states of the United States (Bobek et al 1995;EPA 1997a; Feldman et al 1999) The performance of such assessments is difficult because ofthe paucity of information and the difficulty of comparing risks to disparate entities andprocesses over large ranges of spatial and temporal scales As a result, expert judgment hasbeen used in the absence of data analysis, as in the case of Harwell et al (1992), and even thathas been largely replaced as a means of prioritization by consensus of representatives ofstakeholders and the public (EPA 1997a) Consensus-based assessments have some potentialbenefits beyond prioritization itself, such as better understanding of environmental issues andpromotion of coordinated action by the participants, but they have not been followedand implemented by prioritized risk management programs (Feldman et al 1999) Prioritiza-tion based on actual estimation of risks must await further development of assessmentmethods and a willingness to devote sufficient resources to the problem

The need to replace expert judgment with technical approaches is illustrated by theconsideration of oil spills in the prioritization of environmental hazards by the US EPA’sScience Advisory Board They gave a low rank to oil spills because of the perception thatecological effects of oil in the marine environment were short term (SAB 1990) However, thatperception seems to be a result of a lack of high-quality long-term monitoring It has beenreported that some detectable effects of the Exxon Valdez spill persisted for at least a decade(Peterson et al 2003)

Beyond the technical difficulties, risk-based prioritization has had little influence in latory agencies, in part because it may be considered illegal or immoral The potentialillegality arises because environmental laws in the United States and most other countriesrequire protection independent of other laws For example, the US EPA cannot decide to stopenforcing the Clean Air Act because resources would be more effectively spent enforcing theClean Water Act In addition, prioritization may assign high priority to hazards for which nolegal authority for action exists The accusations of immorality most commonly arise from theaccusation that technical analysis is used to minimize the legitimate subjective concerns ofcitizens or to ignore risks to small groups with particular exposures (e.g., indigenous peopleconsuming traditional foods) On the other hand, consensus-based prioritization may alsoprovide unequal protection When prioritization is based on a stakeholder process, there is apotential for higher ranking of the risks that concern the most articulate and influentialsegments of the population Because of these issues, the potential benefits of a rationalprioritization process must await a mandate from the highest levels of government toovercome the technical, social, and legal impediments (Davies 1996)

regu-1.3.2 COMPARISON OF ALTERNATIVE ACTIONS

As discussed above, risk assessment is performed to inform decisions concerning tive actions Unfortunately, the range of alternatives is often small and the range of issuesconsidered is often narrow For example, registration of a pesticide often does not includeconsideration of the risks from the alternatives, existing pesticides that may be more persist-ent or toxic, and nonpesticide pest control techniques that have their own potentially severeecological risks Rather, the alternatives are typically restricted to registration, registrationwith restrictions, or rejection of the new pesticide It is clear that the decision-making processcan meet the legal mandate and be rational within its scope, but result in a less than optimumdecision for the environment

alterna-Comparative risk assessment raises two complications First, it is not adequate to estimaterisks; one must also estimate the benefits of the alternative actions A relatively low-riskalternative action may be undesirable, because it produces small benefits or even net decre-ments The comparison of risks and benefits (not just costs and benefits) is important in anycase, but is essential when comparing a set of alternatives Second, comparison of risks ofteninvolves the common units problem That is, if the alternatives involve disparate risks andbenefits, they may not be directly compared by simply quantifying future ecological condi-tions The temporal integration of expected benefits and decrements of each action, expressed

in common units, is termed net environmental benefit analysis (Efroymson et al 2004) If thecomparison must consider the costs of implementation, the net benefits must be monetized toyield a cost–be nefit analys is (Ch apter 36)

Comparative risk assessment is a different way of looking at any of the decisions discussed

in this chapter Although all risk-based decisions involve the comparison of at least twoalternatives (e.g., permit an action or not), a more comparative approach opens up theprocess to a range of potentially desirable alternatives However, it complicates the assess-ment a nd decision -making process App roaches to these issues are discussed in Chapter 34.1.3.3 PERMITTINGRELEASES

Ecological risk assessments have been concerned primarily with two activities: determiningwhether releases of chemicals or other agents should occur and determining how to dealwith the releases that have already occurred Clearly, we should do a better job of the former

to reduce the need for the latter Ecological risk assessments for permitting releases aredistinguished by the type of agent released (i.e., chemicals, effluents and other wastes, andexotic organisms) and by whether the agents are novel or have been permitted before and arebeing reconsidered

1.3.3.1 Chemicals

In the United States, new chemicals are regulated as pesticides under the Federal Insecticide,Fungicide, and Rodenticide Act (FIFRA), as industrial chemicals under the Toxic SubstancesControl Act (TOSCA) or under the Food, Drugs, and Cosmetics Act (under which ecologicalconcerns have received little attention) The difference in ERA under FIFRA and TOSCAserves to illustrate the importance of legal constraints on assessment practices Becausepesticides are designed to be toxic, FIFRA allows the government to require relativelyextensive characterization and testing of new chemicals by the manufacturer and allows thegovernment time to complete the assessment TOSCA does not allow characterization andtesting requirements beyond basic descriptions of the compounds and allows only 90 days forassessment and decision making As a result, assessment of pesticides has been based on afairly elaborate tiered scheme and is moving to a system of probabilistic assessment (Urbanand Cook 1986; Ecological Committee on FIFRA Risk Assessment Methods 1999a,b) Thepesticide industry has responded with its own tiered and probabilistic ecological risk assess-ments of products such as atrazine (Section 32.4.4) In contrast, ERA under TOSCA relies onsmall data sets, quotient methods, and assessment factors of 10, 100, or 1000 (Zeeman 1995;Nabholz et al 1997) Assessments of new chemicals in the European Union and elsewherehave their own testing requirements and schemes, but in general they rely on tiered testingapproaches and simple methods using factors and quotients (RIVM 1996; Royal Commission

on Environmental Pollution 2003) Similar assessment approaches have been developed byresponsible chemical manufacturers to assure that their products ‘‘are safe for the environ-ment’’ (Cowan et al 1995) Methods for ecological assessment of chemicals are rapidly

developing because of advances in science such as computational toxicology and because ofrenewed interest in existing chemicals, particularly the European Community’s REACHregulations (Bradbury et al 2004).

1.3.3.2 Effluents and Wastes

The release of aqueous and gaseous effluents and other waste streams is regulated in the UnitedStates and most other countries through a permitting process The most important of thesefrom an ecological perspective is the permitting of aqueous effluents, by a process known asNational Pollutant Discharge Elimination System (NPDES) in the United States This isaccomplished primarily by specifying that the effluent will not violate water quality standards.Standards include concentrations, durations, and frequencies of exceedence that must not beviolated (Section 2.2) In most states, standards are based on National Ambient Water QualityCriteria published by the EPA (1985) Equivalent criteria and standards are used in othernations (Roux et al 1996; CCME 1999; ANZECC 2000) Alternatively, permits may specifythat the toxicity of the effluent be tested using standard acute or subchronic tests (Section 24.2)

or that the receiving community achieve biological criteria (EPA 1996a; Ohio EPA 1998).1.3.3.3 New Organisms

Organisms may be deliberately imported for horticultural use, biological control, pets, orother purposes Determining whether an importation should be permitted is conceptuallydifficult because organisms are complex and can display unexpected properties or may evolvenew properties Risk assessments in support of the regulation of importation of foreignorganisms may be based on structured expert judgment as in the United States (Orr 2003) ormore objective analyses An example is the assessment of import of shrimp for aquaculture thatmay carry a virus, which is pathogenic to native shrimp (Fairbrother et al 1999) Geneticallyengineered organisms are regulated in the United States as though they are chemicals That is,novel biocontrol agents are regulated by the pesticides office of the EPA, and other novelorganisms are regulated like industrial chemicals by the toxic substances office

1.3.3.4 Items in International Trade

The World Trade Organization (WTO) 1995 Agreement on the Application of Sanitary andPhytosanitary Measures and some regional trade agreements require that a risk assessment

be performed if a nation excludes an item from importation due to risks that it poses tohuman health, animals, or plants The exclusion may be based on the determination that theitem may be toxic, a pathogen, a pest, or otherwise pose an unacceptable risk, or there is asignificant risk that an item may be a carrier of such a hazardous agent Items that have beenexcluded from import range from controversial genetically modified crops to wood and plantproducts that may contain exotic pests The WTO agreement is more demanding than mostlegal bases for risk assessment, in that risks must be expressed in terms of probabilities orlikelihoods, not possibilities (Codex 1997; OIE 2001) New Zealand provides an excellentguide for conducting probabilistic risk assessments for imports of animals, animal products,and associated pathogens and pests (Murray 2002)

1.3.4 LIMITINGLOADING

The regulation of the uses of chemicals and their disposal in effluents or solid wastes fails to

be protective because of the effects of multiple releases from multiple sources One solution is

to define the rate at which an ecosystem may receive a pollutant from all sources without

unacceptable effects For atmospheric deposition, this is referred to as the critical load(Hettelingh and Downing 1991; Holdren et al 1993; Hunsaker et al 1993; Strickland et al.1993) The same term is applied to aqueous pollution (Vollenweider 1976), but for waterquality regulation in the United States, it is referred to as the total maximum daily load(TMDL) (Houck 2002) Setting limits on loading requires defining the resources to beprotected (e.g., water quality, biotic communities, or human health) and the endpoints to

be measured for each (e.g, water quality criteria, benthic invertebrate species richness, and soilpH) A mixture of field measurements and modeling is then used to determine whether a limit

is exceeded, whether a new source will cause exceedence, the relative contributions of existingsources to an exceedence, or the likelihood that a remedial action will result in acceptableloading In some cases, this is relatively straightforward For example, if a persistent andsoluble chemical is released at multiple points in a stream or watershed, simple transport andfate modeling can be used to determine contributions to exceedence of a water qualitycriterion at a downstream point However, other cases are complex and difficult Forexample, NOx deposition in a watershed is difficult to trace back to point and nonpointsources in the atmosphere, and effects on terrestrial and aquatic ecosystems are difficult tomeasure or predict because of the complexity of nitrogen cycling and acidification andeutrophication processes

1.3.5 REMEDIATION ANDRESTORATION

The predominant use of ecological risk assessment in the United States has been to supportthe remediation of contaminated sites under Superfund (Suter et al 2000) A full ecologicalrisk assessment for a contaminated site would consider the risks from the existing contamin-ation (the no action alternative), from the remedial actions themselves, from residual con-tamination, and from subsequent land uses In addition, if ecological restoration activitiesmay be performed after remediation, risks associated with restoration must be considered

In the United States, remedial assessments are performed in two stages: a baseline ment that determines whether the unremediated contamination poses a significant risk and anassessment of remedial alternatives termed the feasibility study Procedural and technicalguidance are available for baseline assessment (Sprenger and Charters 1997) (see also theEPA’s Environmental Response Team and Office of Solid Waste and Emergency Responseweb sites), and these assessments are often well performed and based on ample data Becausethere is a contaminated site to be sampled, surveyed, and tested, the full range of assessmenttechniques is available to the assessor (Suter et al 2000) In contrast, there is little guidance forthe assessment of risks from dredging, soil removal, capping, construction of roads and othersupport facilities, chemical or thermal treatment of media, spills of treatment chemicals, andother remedial activities that pose obvious ecological hazards Remedial decisions tend to focus

assess-on the efficacy of technologies in reducing the cassess-ontaminant risks and assess-on their costs rather thanthe risks from remediation The ecological risks of remedial alternatives are usually given aserious assessment only when, as in the polychlorinated biphenyl (PCB) contamination of theHudson River, New York, the remediation is particularly costly or controversial

Restoration involves recreating to some extent the ecological structure and function of asite disturbed by a remedial action or any other action Hazards associated with restorationinclude erosion and siltation, introduction of exotic species (e.g., as ground covers) withundesirable properties, use of pesticides and fertilizers, and conversion to parks with associ-ated mowing and trampling In addition to these risks from restoration, planted trees may die,instream structures may wash away, and for other reasons restoration activities may fail.Therefore, it may be appropriate to adapt engineering risk assessment techniques to restor-ation projects As with other sorts of risk assessments, assessments of restoration should

compa re the risks of alternati ves For exa mple, after the erupti on of Moun t St Hele ns,parts of the area wer e seeded with exotic herbaceou s plants Although these plants red ucederosion of the ash, they apparently slow ed the reest ablishment of the native fores t Hence, therisks to stream and river ecosyst ems from ero ding ash cou ld be compared with the risks fromdelayed recovery of the nativ e terrestrial communi ty.

1.3.6 PERMITTING AND MANAGING LAND USES

The conversi on of one land use to another, particular ly the co nversio n of land supporti ng anatural communi ty to agric ultural, residen tial, or urban uses, is one of the mo st severeanthrop ogenic hazards that eco systems face How ever, land use co nversion is little regula ted

in the Unite d States, wher e land use permi tting is usually perfor med at the local level ofgovern ment Hence, eco logical risk asses sment is seldom involv ed in land use decisi on Theexcepti on is the major action of the Unit ed State s federa l governm en t, whi ch is subject toenvironm ental impac t asses sment (Sect ion 2.12) Nevertheless, the ecological risk assessmentframework can be applied to land management decisions, even the complex decisions con-cerning land development and water use in South Florida (Harwell 1998)

Land use also involves decision s co ncerning the intens ity of use: How many cattle should

be a llowed to graze on an area of rangel and? How often should a forest be logged ? Thesedecisi ons are the subject of their own well-deve loped asses sment practice s (Davis et a l 2000;Holchek et al 2003) Alt hough they make little use of risk asses sment concepts and termin-ology, to the extent that they have clear qua ntifiabl e goals and analyze unc ertainty to informdecision makers , they are eco logical risk assessments

1.3.7 SPECIES MANAGEMENT

Risk s to specie s or specie s popul ations are estimat ed for the managem ent of resourc e spe ciesand endan gered specie s The techn iques developed for these purposes have been adapted forestimat ing risks to populatio ns from pol lutant chemi cals and intr oduced organisms, and riskasses sment co ncepts have been adap ted to resourc e managem en t (Ch apter 27) (Franci s andShotton 1997) Management of game species, fisheries, timber trees, and other harvestedplants requires assessments to determine harvest levels that do not pose unacceptable risks ofextirpation Probabilistic modeling has been particularly important in fisheries management.Other innovations in the assessment of resource management include modeling populations

in a community or ecosystem and adaptive management (Walters 1986) Management ofthreatened or endangered species is often based on a form of population modeling termedpopulation viability analysis that estimates the time to extinction or the probability ofextinction within a prescribed time period such as the next 50 years (Sjogren-Gulve andEbenhard 2000; Beissinger and McCollough 2002; Keedwell 2004) These two assessmentpractices overlap when harvesting is forcing a population to extinction (Musick 1999).Assessment of species management becomes particularly complicated when the valued species

is potentiall y damagi ng to the ecosyst em or to other valued specie s (Box 1.1) Risk assessmentlends itself to species conservation, and ecological risk assessment for conservation and forpollution regulation have developed in parallel (Burgman et al 1993; Burgman 2005).1.3.8 SETTINGDAMAGES

When ecological injuries occur due to negligence or criminal activities, the responsible partiesmay be required to pay monetary damages These damages are used to restore the damagedecosystems, or, when that is not possible, to acquire and protect other areas In the UnitedStates, natural resource trustees are required to seek damages from polluters under the Clean