风险评价中概念性模型

风险评价资料

ES/ER/TM-186

Guide for Developing

Conceptual Models

for Ecological Risk

Assessments

This document has been approved for release

to the public by the K-25 Site Technical

Information Officer. Date:

风险评价资料

This report has been reproduced directly from the best available copy.

Available to DOE and DOE contractors from the Office of Scientific and

Technical Information, P.O. Box 62, Oak Ridge, TN 37831; prices available

from 615-576-8401 (fax 615-576-2865).

Available to the public from the National Technical Information Service, U.S.

Department of Commerce, 5285 Port Royal Rd., Springfield, VA 22161.

风险评价资料

ES/ER/TM-186

Guide for Developing

Conceptual Models

for Ecological Risk

Assessments

G. W. Suter II

Date Issued—May 1996

Prepared by the

Environmental Restoration Risk Assessment Program

Oak Ridge National Laboratory

Oak Ridge, Tennessee 37831

Prepared for the

U.S. Department of Energy

Office of Environmental Management

under budget and reporting code EW 20

LOCKHEED MARTIN ENERGY SYSTEMS, INC.

managing the

Environmental Management Activities at

Oak Ridge K-25 SitePaducah Gaseous Diffusion Plant

Oak Ridge Y-12 PlantPortsmouth Gaseous Diffusion Plant

Oak Ridge National Laboratory

under contract DE-AC05-84OR21400

for the

U.S. DEPARTMENT OF ENERGY

风险评价资料

PREFACE

This white paper was prepared to present guidance for preparing conceptual models for ecologicalrisk assessments, which are an important component of the Remedial Investigation process. This workwas performed under Work Breakdown Structure 1.4.12.2.3.04.07.02 (Activity Data Sheet 8304).Publication of this document meets an Environmental Restoration Risk Assessment Program milestonefor FY 95. Use of this guidance document will standardize the conceptual models used in ecological riskassessment so that they will be of high quality, useful to the assessment process, and sufficientlyconsistent so that connections between sources of exposure and receptors can be extended acrossoperable units.

风险评价资料

CONTENTS

PREFACE......................................................................iiACRONYMS...................................................................ivEXECUTIVE SUMMARY.........................................................v

1. INTRODUCTION..............................................................1

2. CONCEPTUAL MODEL DEVELOPMENT AND THE RI/FS PROCESS................1

3. CONCEPTUAL MODELS OF CURRENT AND FUTURE RISKS ......................2

4. COMPONENTS OF A CONCEPTUAL MODEL....................................2

4.1 OU TYPES AND DEFAULT CONCEPTUAL MODELS.........................3

4.2 SOURCES...............................................................9

4.3 ROUTES OF TRANSPORT................................................9

4.4 EXPOSURE MEDIA.....................................................10

4.5 ROUTES OF EXPOSURE.................................................10

4.6 RECEPTORS...........................................................11

4.7 INDIRECT EXPOSURE AND EFFECTS.....................................12

4.8 OUTPUT TO OTHER OUs................................................12

5. RELATIONSHIP TO OTHER CONCEPTUAL MODELS............................13

6. CONCEPTUAL MODELS FOR THE FEASIBILITY STUDY.........................13

7. REFERENCES...............................................................13

FIGURES

1. Generic conceptual model for a source OU...........................................

2. Generic conceptual model for an aquatic integrator OU.................................

3. Generic conceptual model for a groundwater integrator OU.............................

4. Generic conceptual model for the terrestrial integrator OU..............................

5. Generic conceptual model for aquatic biota..........................................45678

风险评价资料

ACRONYMS

CERCLA

DOE-OR

DQO

EPA

FFA

OU

RI/FS

WAGComprehensive Environmental Response, Compensation, and Liability Act U.S. Department of Energy's Oak Ridge Operationsdata quality objectiveU.S. Environmental Protection AgencyFederal Facilities Agreementoperable unitremedial investigation/feasibility studywaste area grouping

风险评价资料

EXECUTIVE SUMMARY

Ecological conceptual models are the result of the problem formulation phase of an ecological riskassessment. They may be thought of as a hypothesis concerning the nature of ecological risks at acontaminated site. They include the hypothesized sources of contaminants, routes of transport ofcontaminants, contaminated media, routes of exposure, and endpoint receptors. They are presented inthe form of a flow chart and a descriptive narrative. Conceptual models should be developed in draftform as input to the data quality objectives (DQOs) process, developed and agreed upon during the DQOprocess, and modified as necessary during the remedial investigation (RI) process as new informationchanges the understanding of the site. All screening and baseline ecological risk assessments should useand present the conceptual model.

Conceptual models should be presented for the current case and for any credible future cases thatcould result in increased risk. The same conceptual model can be used for the baseline ecological riskassessment in the RI and for the reductions in risk associated with the remedial actions assessed in thefeasibility study (FS). However, the ecological risks associated with the physical damage caused byremedial actions require separate conceptual models.

Generic conceptual models are presented for four types of operable units (OUs): source OUs,aquatic integrator OUs, groundwater integrator OUs, and terrestrial integrator OUs. The guidancedescribes the use of these generic models to develop site-specific models including how to representsources, routes of transport, contaminated media, routes of exposure, endpoint receptors, and indirecteffects. It also describes how the ecological conceptual model should be integrated with a conceptualmodel for the site and made consistent with the conceptual model of human health risks.

风险评价资料

1. INTRODUCTION

In the U.S. Environmental Protection Agency’s (EPA's) framework for ecological risk assessments,the conceptual model is the output of the problem formulation phase of the assessment (EPA 1992). Itpresents a working hypothesis of how the site contaminants might affect the ecological components ofthe site. It includes descriptions of the source, the receiving environment, and the processes by which thereceptors come to be exposed directly to the contaminants and secondarily to the effects of thecontaminants on other environmental components.

This document presents guidance for the development and presentation of conceptual models forecological risk assessments. It builds on relevant documents from the EPA (1989, 1992 and 1994) andon previous guidance documents developed for the U.S. Department of Energy's Oak Ridge Operations(DOE-OR) (Suter et al. 1995, Barnthouse and Suter 1995). The Approach and strategy for performingecological risk assessments for the U.S. Department of Energy’s Oak Ridge Reservation report(referred to hereafter as Approach and Strategy) (Suter et al. 1995) defines different types ofComprehensive Environmental Response, Compensation, and Liability Act (CERCLA) operable units(OUs) and presents generic conceptual models for each of them. That document, now in its third version,is the product of an extended data quality objectives (DQO) process by the Federal Facilities Agreement(FFA) parties, of revision in response to regulator comments, and of the accumulated experience inecological risk assessment at Oak Ridge, Tennessee; Portsmouth, Ohio; and Paducah, Kentucky.The DQO guidance document for ecological risk assessment presents the results of the genericecological DQO process that was conducted for the Oak Ridge Reservation (ORR) and explains howthey can be used to efficiently conduct DQO processes for individual OUs (Barnthouse and Suter 1995).This conceptual model guidance document should not be used without first reading these more generalguidance documents.

2. CONCEPTUAL MODEL DEVELOPMENT AND THE RI/FS

PROCESS

Conceptual models are developed and used iteratively in the remedial investigation/feasibility study(RI/FS) process. First, following the initial site survey, draft conceptual models should be developed asinput to the DQO process. These models should be inclusive in that they should include all sources,receptor classes, and routes of exposure that are of plausible concern. These preliminary conceptualmodels also serve as the conceptual model for the screening assessment that should be performed tosupport the DQO process.

During the DQO process, the FFA parties, with input from their technical staffs, refine theconceptual model thereby making it more focused. This refinement is created by eliminating (1) receptorsthat are not deemed to be suitable assessment endpoints, (2) routes of exposure that are not credible orimportant, (3) routes of exposure that do not lead to endpoint receptors, and (4) potential sources thatare not deemed credible or important. In addition, the DQO process makes the conceptual model morespecific by identifying particular endpoint species, defining the spatial and temporal scale of theassessment, and other judgments (Barnthouse and Suter 1995). The results of the DQO process arepresented in the conceptual models published in the RI work plan. If a new screening assessment isperformed for the RI work plan or for an interim report of a phased RI, it should be based on thisconceptual model. The conceptual models reappear in the RI/FS, and in most cases, they will be the same

风险评价资料

as in the RI work plan. However, the results of ongoing communications among the FFA parties and theresults of the RIs may result in modification of the conceptual model.

The bases for developing the conceptual models depend on the stage of the assessment and theamount of prior assessment that has been done at that stage.

The first conceptual model is based on qualitative evaluation of existing information and expert

judgment. It should be conservative in the sense that sources, pathways, and receptors should bedeleted only if they are clearly not applicable to the site.

Early in the RI/FS process, preferably before or during the DQO process, a screening assessmentshould be performed using existing data. The results of the screening assessment can be used toeliminate receptors or even an entire medium for which no contaminants present a potentiallysignificant risk.

The participants in the DQO process can apply their professional judgment and managerialauthority to modify the draft conceptual model presented by DOE’s assessment scientists. Forexample, the FFA parties may decide that the results of the screening assessment are not based ondata of sufficient quality and quantity to justify deleting media or receptors. Some receptors maybe eliminated because they are not judged to be sufficiently important or sensitive or not sufficientlyrelated to the remedial decision.

If the RI is conducted in phases, the screening assessment performed at the end of the preliminaryor intermediate stages should be used to modify the conceptual model. Typically, this involvesreducing the model by eliminating components that were shown by the assessment to beunimportant or even not present.

3. CONCEPTUAL MODELS OF CURRENT AND FUTURE

RISKS

The Approach and Strategy report specifies that a separate ecological risk assessment should beperformed if ecological risks could increase in the future. This could occur if contaminant transport hasnot brought wastes to the surface or if succession or other changed ecological conditions could bringmore susceptible species onto the site. In simple cases, this will not require a different conceptual model.For example, if range expansion is hypothesized to bring a more susceptible species to the site in placeof the current representative species for a trophic group (e.g., river otters in place of mink) or a moreprotected species (e.g., bald eagles in place of osprey), the model need not change except to add thefuture endpoint species to the list of current endpoint species. However, other cases such as developmentof a forest ecosystem on a currently bare or closely mowed site will require a separate model for futureconditions.

4. COMPONENTS OF A CONCEPTUAL MODEL

A conceptual model should be presented in both graphic and narrative form. The graphic form maybe pictorial (i.e., with drawings of plants and animals), but pictorial representations are typically costlyto produce and often ambiguous. Therefore, flow charts are generally recommended. The charts should

风险评价资料

of receptors to media, (4) endpoint receptors, and (5) output to other OUs. The narrative should describethe contents of the diagram in sufficient detail to ensure that it can be understood by an educatedlayperson. However, the narrative conceptual model should not duplicate information that is present inother sections of the document in which it occurs, such as the site description.

In addition, the narrative conceptual model should explain the underlying logic of the modelincluding the following.

It should describe the spatial bounds of the assessment and any subdivision of the site into reaches

or other subunits.

If receptors or routes of exposure are omitted due to lack of information or knowledge, thatomission should be acknowledged and included in the analysis of uncertainty.

If receptors or routes of exposure are omitted because of the judgment of the FFA parties to limitthe scope of the assessment to critical pathways and receptors, that judgment should beacknowledged and explained. However, it should not be treated as an uncertainty because it is a riskmanagement decision.

If receptors are representative of a class of receptors, then that relationship should be explained.

The following graphical conventions are used in the flow charts for the generic ecologicalconceptual models. The distinctions among compartments can enhance the readability of the model andthe use of these conventions in conceptual models for individual OUs will result in consistency andcomparability among assessments. Rectangles represent components of the OU, rounded rectanglesrepresent inputs from other OUs, and circles represent components of other OUs that receive input fromthe depicted OU. Rectangles with heavy borders represent receptors that are potential assessmentendpoints.

4.1 OU TYPES AND DEFAULT CONCEPTUAL MODELS

The Approach and Strategy report divides OUs into four classes: source OUs, aquatic integratorOUs, groundwater OUs, and the terrestrial integrator OU which corresponds to the entire reservation.For each of these classes of OUs, this guidance report provides generic conceptual models including aflow diagram of the routes of transport and exposure and generic assessment and measurement endpoints(Figs. 1, 2, 3, and 4). In addition, a generic conceptual model of an aquatic ecosystem is presented whichis an elaboration of the aquatic biota compartment in the conceptual models for source and aquaticintegrator OUs (Fig. 5). Chapter 3 of the Approach and Strategy report describes each of these genericmodels in terms of the compartments that are relevant, the composition of each compartment, inputs tothe compartment, and outputs from the compartment. All developers of conceptual models shoulddetermine which type of OU they are assessing and begin the development process with the genericmodel. The following discussions explain how to modify those models to make them OU-specific.

风险评价资料

风险评价资料

风险评价资料

风险评价资料

风险评价资料

风险评价资料

4.2 SOURCES

All conceptual models for contaminated sites begin with sources. On source OUs, the wastesdeposited in pits, trenches, ponds, tanks, etc., are treated as the ultimate sources. Each distinct type ofultimate source should be identified in a separate box. Types of sources should be distinguished whenthey contain wastes that are distinctly different in form or composition or when the wastes are disposedof in different manners (e.g., ponds versus tanks) or in situations that would result in different modes oftransport (e.g., floodplains versus uplands). Because of the disagreements that have occurred among theFFA parties about whether to treat waste sumps and ponds as sources or as aquatic ecosystems that areincidentally contaminated, it is important to clearly explain the nature of all such bodies including thepurpose for which they were created and their current ecological condition.

Integrator OUs usually have no ultimate sources, but they have as proximate sources thecontaminated inorganic media: surface water, shallow groundwater, sediments, and soils. These may bein the form of fluxes of surface water, groundwater, eroded soil, or suspended sediments and should beidentified in terms of their nature and source. For example, sources to Waste Area Grouping (WAG) 2,the OU that includes McCoy Branch and lower White Oak Creek on the Oak Ridge Reservation, includecontaminated surface water from WAG 1 (upper White Oak Creek) and shallow groundwater from seepsat the toe of WAG 5. In addition, some proximate sources are not associated with any ultimate source.For example, soils may be contaminated by past spills or other actions to which no ultimate source orupstream source is contributing.

4.3 ROUTES OF TRANSPORT

The conceptual model should identify the routes by which contaminants in the sources aretransferred to ambient media to which organisms may be exposed. The specific routes of exposureshould be described. For example, the transport from sources to surface water should be identified asoccurring in leachate emerging at seeps, in leachate mixed with groundwater entering streams throughgaining reaches, by erosion of contaminated soil, etc.

The routes of transport for ecological conceptual models do not normally include deep groundwatertransport because it does not contribute to surface water contamination and because wildlife do not drinkwell water. However, the FFA parties included a generic conceptual model for groundwater (Fig. 3)largely because of concern for the biota of caves that are known to occur on the reservation. Thatconceptual model has not been invoked at any OU to date, but it is important to avoid rejecting that modeof transport without considering the possibility that cave biota may be exposed. This aspect of the ORRis currently uncharacterized.

Except for movement into downstream OUs, these conceptual models do not include fate processesthat remove contaminants from the system (e.g., degradation and sequestration) because these conceptualmodels are intended to illustrate how ecological receptors come to be exposed rather than illustrating thefate of the contaminants. These fate processes may be included in the overall conceptual model of theOU, which is included in Sect. 2 of the RI.

风险评价资料

4.4 EXPOSURE MEDIA

The conceptual model should identify the media that are known to be significantly contaminated,are hypothesized to currently be significantly contaminated, or are predicted to be significantlycontaminated in the future. If possible, significance of contamination should be based on the results ofan assessment that compares screening of measured contaminant concentrations against ecotoxicologicalbenchmarks and background concentrations (Suter 1995). Alternatively, modeled concentrations maybe screened in the same way. In the absence of measured or modeled concentrations, expert judgmentshould be conservatively applied. A medium should be included in the model if any chemical in themedium is retained by the screening process or any chemical is judged to potentially be present atsignificant concentrations.

In some cases, the contaminated medium is also the waste (i.e., the source of the contaminantchemicals). This is true of the coal ash in the filled coal ash pond on Chestnut Ridge OU 2. It would alsobe the case for any waste sumps that are treated as receptor ecosystems rather than as sources. In suchcases, the source box is simply combined with the soil, water, or sediment box.

4.5 ROUTES OF EXPOSURE

The conceptual model should identify the routes of exposure that are assumed to result in uptakeof chemicals from contaminated organic and inorganic media. The number of routes of exposure islimited to those that are deemed to be important for the endpoint receptors. The following points shouldbe considered.

Fish, aquatic invertebrates, and aquatic plants are assumed to be exposed to contaminants in water.

Conventionally, the EPA and most risk assessors have assumed that dietary exposures arenegligible and that is likely to be true for most chemicals. For example, the National Ambient WaterQuality Criteria for Protection of Aquatic Life are based on toxicity tests in which organisms areunfed or fed clean food. This is reasonable given the relatively high rate of exposure of organismsto chemicals in the water that pass their respiratory surfaces and the fact that most chemicals arenot highly bioaccumulative and do not biomagnify.

Dietary exposures should not be routinely included for fish or aquatic invertebrates. Dietaryexposure is important for a few long-lived and biophilic chemicals such as polychlorinatedbiphenyls (PCBs) and dioxins and may be important for a wider variety of chemicals than iscurrently recognized. However, appropriate dietary exposure models are not available for chemicalsin streams such as occur on the ORR, and toxicity information based on dietary exposure isuncommon and poorly standardized. Fish body burdens integrate dietary and direct aqueousexposures, but toxicity information is not standardized or available for exposures to most chemicalsin terms of body burdens. Therefore, dietary exposures should be included only if the assessorshave reason to believe that they are a significant route and have a method for assessing risks dueto that route.

Benthic invertebrates are exposed to sediment pore water and whole sediment. Although the graphicversion of the conceptual model need not depict this distinction, it is important to include in thenarrative. Although EPA’s sediment quality criteria are based on exposure to the aqueous phaseof sediments (i.e., pore water), the evidence is strong that some benthic invertebrates aresignificantly exposed to a variety of chemicals by ingestion of sediment particles. Pore water

风险评价资料

than neutral organic compounds, but pore water may be extracted and analyzed. Therefore, it isimportant to characterize risks due to both modes of exposure.

Wildlife exposure routes usually include ingestion of food, drinking water, and incidental soil

ingestion. Soil ingestion may be excluded for species that have little exposure to soil (e.g., ospreys).Dermal exposure of wildlife should not normally be included. Unlike humans, birds and mammalianwildlife are covered with feathers and fur. These coverings exclude most dermal exposures.However, they create another route of exposure: grooming and preening, which contribute toincidental soil ingestion. Amphibians are likely to experience significant dermal uptake, but neitherexposure models nor toxicity data are available to address this route and receptor for terrestrialexposures. Aqueous dermal exposures for amphibians are equivalent to respiratory exposure of fishin that they are assumed to be due to direct uptake of dissolved chemicals through the respiratoryepithelium, which is the skin.

Respiratory exposure of wildlife is not normally included. Few if any OUs on the ORR havesignificant concentrations of contaminant chemicals in the air. This judgment has been confirmedby the FFA parties who have not called for measurements of atmospheric contamination in the RIsconducted to date.

Plants, soil invertebrates, and soil microbes are assumed to be directly exposed to whole soil.In cases where shallow groundwater is contaminated, plants are exposed to that water.

4.6 RECEPTORS

The receptors presented in the conceptual model should be those that have been proposed to be ordesignated as assessment endpoint receptors (organisms, populations, communities, or ecosystems). TheApproach and Strategy report presents generic assessment endpoints for each type of OU and providesguidance for selecting endpoint receptors for particular sites. This section explains how they should bepresented in the conceptual model.

Ecosystems. Ecosystems are assessment endpoints if the properties to be protected are ecosystemproperties. This is the case for wetlands which are protected for their habitat value to wetland-dependentspecies and their roles in nutrient retention and cycling and hydrologic regulation. If significant areas ofwetlands are present (i.e., areas sufficient to significantly contribute habitat, nutrient cycling, andhydrologic regulation functions to the watershed in which they occur), they should be included in thegraphical model and their size, type, and assumed functional properties defined in the narrative. Acomponent of an ecosystem that is valued for its functional properties rather than its community orpopulation properties may also be considered an ecosystem-level endpoint. Soils, which degrade naturaland anthropogenic organic materials, recycle nutrients, and support plant growth, are identified in theApproach and Strategy report as valued for their functions.

Community. Fishes, benthic macroinvertebrates, soil invertebrates, and upland plants arecommunity level assessment endpoints. That is, the species richness and abundance of the communitiesare the endpoint properties rather than properties of the component populations. Cases wherecomponents of the community such as benthic-feeding fish or trees are believed to differ in theirsusceptibility should be distinguished in the conceptual model. The model should describe each

风险评价资料

operational terms (e.g., all invertebrates collected by a Surber sampler and retained by a 1-mm meshscreen).

Population. Most wildlife are population level assessment endpoints. The endpoint properties areabundance and production of individual populations. The populations used are chosen to represent aparticular trophic group and taxonomic class (i.e., birds and mammals). The conceptual model shouldidentify these receptors both in terms of the species and location of the population (e.g., short-tailedshrews in WAG 2) and the group that they represent (e.g., ground invertebrate feeding mammals). Sometrophic/taxonomic groups will have more than one representative species (e.g., kingfishers and ospreyfor piscivorous birds). Others such as reptiles may have none because of the paucity of toxicologicalinformation concerning those species. The narrative for these receptors should indicate why therepresentative species was chosen and exactly what other species it represents.

4.7 INDIRECT EXPOSURE AND EFFECTS

The generic conceptual models include indirect routes of exposure (i.e., food web transfers) but notindirect effects. An endpoint may be affected indirectly through toxic effects on lower trophic groups,by toxic effects on groups that provide physical habitat, or by other mechanisms. The importance ofexplicitly including indirect effects depends on the nature of the ecological relationship that causes theindirect effect and the relative sensitivity of the groups involved. For example, it is assumed for mostchemicals that aquatic invertebrates and fish are more sensitive than the algal community on which theydepend. Therefore, while that trophic relationship should be acknowledged in the conceptual model, itshould be made clear that the indirect effects on fish and invertebrates of direct toxicity to algae are notincluded (if that is the case). The indirect effect that is most likely to be of concern in aquatic ecosystemsis the reduction in food for fish due to toxic effects on invertebrates. Planktonic crustaceans and benthicinsects are often more sensitive than fishes, and benthic invertebrates are more exposed to contaminatedsediments than are fish.

When indirect effects are included in the conceptual model, it is important to distinguish them fromtransfer of contaminants. The generic conceptual models are based on chemical transport and transfer.If indirect effects are hypothesized instead of or in addition to those relationships, they should bedistinguished by a different sort of arrow (e.g., a dashed line) in the graphical model, and they should beexplained and justified in the narrative.

4.8 OUTPUT TO OTHER OUs

The Approach and Strategy report indicates that all OUs are responsible for characterizing theircontributions of contaminants to other OUs. This responsibility is obvious with respect to hydrologicallytransported contaminants. However, it is also true of the contribution of source and aquatic integratorOUs to the contaminant burden of wide-ranging wildlife species that are associated with the reservation-wide terrestrial integrator OU. Therefore, the conceptual model should show connections to downstreamaquatic integrator OUs and groundwater OUs and connections to the terrestrial integrator OU. The routesof transport in the former case would be identified as dissolved chemicals in surface water orgroundwater flow and transport of chemicals sorbed to suspended particles. In the latter case, the routesof exposure would be consumption of contaminated food, soil, or water on the subject OU by wide-ranging species.

风险评价资料

5. RELATIONSHIP TO OTHER CONCEPTUAL MODELSThe conceptual model for ecological risks must be consistent with the conceptual model for humanhealth risks. That is, it should identify the same contaminant sources, routes of transport ofcontaminants, and contaminated media. However, the routes of exposure and receptors will be different.

Some RIs will have an overall conceptual model for the OU. Such models depict the sources androutes of transport of contaminants. They may emphasize particular physical aspects of the site such assurface flow patterns or the relationship between groundwater transport and geological stratigraphy.They may be in the form of maps showing, for example, the location of streams and seeps relative towastes and drainage patterns. The ecological conceptual models should be consistent with these moregeneral conceptual models and should refer back to them to provide the reader of the ecological riskassessment a context for the ecological conceptual model.

Ideally, the ecological conceptual models should be an extension and elaboration of a genericconceptual model for the site. The generic conceptual model would identify the sources, the routes oftransport of contaminants from the sources, the contaminated media, and the transport of contaminantsout of the OU. The ecological conceptual model as well as the human health conceptual model could thenbe limited to the components that are particular to ecological and health risks: contaminated media,routes of exposure, and receptors.

6. CONCEPTUAL MODELS FOR THE FEASIBILITY STUDYEcological assessments for the FS have not always followed the conventions of ecological riskassessment although the choice of remedial options depends on a balancing of risks and benefits. Thedecision to leave in place mercury in the East Fork Poplar Creek floodplain that presents a risk to shrewsand wrens was based on such a balancing of the benefits of remediation against the risks of remediationto the floodplain ecosystem. One component of the EPA’s ecorisk framework that could potentiallyincrease the clarity and rigor of the assessments in FSs is the presentation of conceptual models. The FSmust consider both the benefits of reducing exposure and the risks of habitat loss associated with eachremedial option. The conceptual models in the RI could serve for the former purpose, but separateconceptual models would be needed for the risks associated with the remedial options. Often it is notclear which of the risks from remediation are included in the FS.

Very generic conceptual models for physical disturbances were developed for terrestrial, wetland,and aquatic ecosystems during the DQO process for ecological risks on the ORR (Suter et al. 1995, Figs.8–10). Because of the great diversity of physical disturbances that could occur during remediation, thesegeneric models will require even more adaptation to specific cases than the generic models forcontaminant risks. However, they do incorporate some hypotheses about often neglected risks fromremediation that were of concern to the FFA parties during the DQO process for ecological riskassessment on the ORR.

7. REFERENCES

Barnthouse, L. W. and G. W. Suter II. 1995. Guide for Developing Data Quality Objectives for

Ecological Risk Assessment at DOE Facilities: white paper, Lockheed Martin Energy Systems

风险评价资料

EPA (U.S. Environmental Protection Agency). 1989. Risk Assessment Guidance for Superfund. Volume

II, Environmental Evaluation Manual, EPA/540/1-89/001, Washington, D.C.

EPA. 1992. Framework for ecological risk assessment, EPA/630/R-92/001, Risk Assessment Forum,

Washington, D.C.

EPA. 1994. Ecological risk assessment issue papers. EPA/600/R-94/009, Washington, D.C.

Suter, G. W. II, B. E. Sample, D. S. Jones, T. L. Ashwood, and J. M. Loar. 1995. Approach and

Strategy for Ecological Risk Assessments for the U.S. Department of Energy’s Oak RidgeReservation: 1995 Revision, ES/ER/TM-33/R2, Lockheed Martin Energy Systems Inc., OakRidge, TN.

本文来源：https://www.bwwdw.com/article/gla1.html