|
Items in cart [0] |
Questions? Call 888-624-8373 |
| Copyright © 2009. National Academy of Sciences. All rights reserved. Terms of Use and Privacy Statement |
Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter.
Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.
OCR for page 309
Issues in Risk Assessment
Appendix F
Breakout Sessions
HAZARD IDENTIFICATION
A. Maki and D. Patton
The hazard identification group examined the case studies in light of the 1983 Red Book paradigm and experience with Environmental Protection Agency (EPA) guidelines for health risk assessments to set the context for discussing hazard identification in ecological risk assessment. Generic issues related to paradigm flexibility, scope of ecological risk assessment, the role of uncertainty in research, and the role of nonscientific consideration were discussed. Specific issues were examined for each case study in terms of ecological hazard.
Generic Issues
There was general agreement that flexibility existed (even if not always applied) in the 1983 paradigm and in forthcoming EPA health guidelines. Flexibility is desirable for ecological risk assessment. Although the four components of the paradigm—hazard identification, dose-response assessment, exposure assessment, and risk characterization—are appropriate for any ecological risk paradigm, they may be combined in different ways. For example, hazard identification may be combined with other steps or treated separately case by case. The group also agreed that uncertainties that were not fully analyzed for hazard
OCR for page 310
Issues in Risk Assessment
identification in the case studies are as important in the presentation of hazard data as they are for health risk assessment.
Discussion of other questions suggested that the scope and definition of ecological risk assessment might be broader than the scope and definition of human health risk assessment in the Red Book. For example, risk management considerations (management and political pressures, social costs, economic considerations, and regulatory outcomes) were ingredients in all case studies and related discussions. Much attention was paid to the influence of management on the scope and design of assessment. Such considerations are absent from discussions of health risk assessment. Some participants also felt that generation of new data should be treated as an aspect of risk assessment, rather than restricting risk assessment to evaluation of data that are already in hand.
Discussion leaders questioned the role of valuation in hazard identification, but this issue was not discussed in detail. In view of repeated references to the question of end point selection as a valuation decision, additional examination on this point is needed.
The case studies illustrated the importance of a systematic presentation and evaluation of data used to identify hazard. Discussion leaders noted that presentation of hazard data was highly variable in the case studies and suggested that some of the hazard identification principles that guide health hazard evaluation might be useful, including emphasis on a complete and balanced picture of relevant hazard information. Specific criteria and questions that are critical to identifying ecological risk are needed to develop an operational definition of complete and balanced.
Analysis of Case Studies
Examination of the case studies revealed a variety of approaches to ecological hazard identification.
For the tributyltin study, hazard identification was based initially on field studies. Retrospective epidemiological studies included a monitoring program (both biological and chemical) and laboratory investigation of cause-effect relationships.
In pesticide risk assessments, as exemplified by the agricultural chemicals case study, neither laboratory nor field studies are required to establish a hazard. Instead, there is a regulatory presumption of hazard.
OCR for page 311
Issues in Risk Assessment
A similar presumption of hazard is used by the U.S. Department of Agriculture in evaluating proposed species introductions for biological control purposes.
The polychlorinated biphenyls and 2,3,7,8-tetrachlorodibenzo-p-dioxin study did not explicitly discuss hazard identification. Regulatory actions on both substances are strongly influenced by human health risks, so it is not clear that any explicit ecological hazard identification was needed or performed.
For the spotted owl study, hazard identification occurred through environmental impact studies undertaken by federal agencies to comply with National Environmental Policy Act that identified this species as being vulnerable to loss of habitat due to old-growth forest clearing.
In fisheries management, it might be assumed that fishing is by definition a hazard. Within limits, fishing confers no greater risk to a population than does predation or even the killing of small numbers of fish by toxic chemical spills. Detailed assessments, such as those described in the case study, appear to be triggered by observations of declining catch or by other evidence (e.g., from modeling studies) that suggests that sustainable yields are being exceeded.
The case studies demonstrate that ecological hazard identification can take many forms and can involve both scientific data and policy decisions. The group discussed two possible modifications of the Red Book paradigm to accommodate the clear influence of policy on the conduct of ecological risk assessment: addition of a ''scoping" component before hazard identification and expansion of the definition of hazard identification to include management inputs. No consensus was achieved on which alternative is preferable, but the group agreed that flexibility is important, the separation between risk assessment and risk management must be retained, a distinction is needed between socially relevant and biologically relevant end points for assessment, a social consensus as to which environmental values should be protected is needed, and scientists should communicate knowledge, not policy.
DOSE-RESPONSE ASSESSMENT
J. Bailar and J. Meyer
Discussion in this session focused first on the need to generalize the
OCR for page 312
Issues in Risk Assessment
concept of dose-response assessment for ecological applications and then on the complexities that need to be addressed in practice. The group agreed immediately that for ecological assessments it is better to talk about stress-response than about dose-response relationships. Scientifically, the stress-response concept, as it applies to ecological risk assessment, is complex and involves many considerations that are absent from the usual understanding of dose-response relationships in human health risk assessment. The bulk of the session was devoted to identifying those considerations and discussing how assessments should be structured to address them.
Aspects of An Adequate Stress-Response Analysis for Ecological Risk Assessment
Selection of End Points
The group argued that end point definition is critical for ecological stress-response assessment. Responses can be assessed at all three hierarchical levels of ecological organization: population, community, and ecosystem. Because of the inherent linkages between the levels, it is important to assess how an effect at one level can affect the other levels. No standard methods exist for making those linkages. Because empirical studies of different levels of organization usually also involve different spatial and temporal scales, the decision about which levels to study must be made before studies are initiated.
Final end points must be expressed as measurable characteristics, such as minimal sustainable population or maximal damage that permits the continued viability of a complex ecosystem. Both structural end points and functional end points should be considered. Structural end points include descriptive characteristics of an ecosystem, such as abundance, species composition, and trophic structure. Functional end points include energy/material flows and other transformation processes (i.e., what the organisms do, as distinct from what they are). The choice of end points must be responsive to both technical and policy concerns, including the following:
Values (what do we really care about?),
OCR for page 313
Issues in Risk Assessment
Measurability (can we get the data we need to do the assessment?);
Correlation (there might be little value in studying an end point that is highly correlated with one already selected);
Policy relevance (can the end point be linked to feasible policy options?);
Tracking and enforcement (can future efforts tell whether the management actions based on risk assessment have been effective?).
Many ecological risk assessments necessarily deal with complex systems that offer an abundance of possible end points for study, and selection of one or a few of them for the intense effort required in a full-scale risk assessment is likely to be time-consuming and expensive—perhaps as long and expensive as the risk assessment itself.
As a strategy for selecting end points, the group consensus favored starting with a broad focus and then narrowing to the appropriate level of detail to define the design of the assessment. Taking an initially broad approach prevents missing the broader implications of hazard and stress. Institutional forms of risk assessment, such as premanufacture reviews, are so routinized that the level of organization (e.g., population) is predetermined. For noninstitutional applications, the ability to quantify will probably dictate the level of organization.
Consideration of Nonlinearities And Discontinuities
Nonlinearities and discontinuities are likely in the response of ecological systems to stress. The group consensus was that the likelihood of observing a threshold or mean-threshold in the stress-response function increases with system complexity. Because thresholds are common in ecological systems, goals of stress-response analysis should include identification of degrees of stress at which thresholds occur and estimation of the upper ends of the threshold ranges.
The slope of the stress-response curves might be steeper as the scale of organization increases—and might approach a step function for communities and ecosystems. Therefore, the assessor needs to be sensitive to the probabilities of catastrophic changes that have few analogues at lower levels of organization and, consequently, use a greater margin of
OCR for page 314
Issues in Risk Assessment
safety. Work is needed to understand the mechanisms of the response that occur at the threshold in the stress-response function.
Expression of Uncertainty
The functional expression of the stress-response relationship is stochastic and distributional; the assessor must consider extremes and discontinuities, not just central tendencies. Assessments should recognize the natural variability in systems, and conclusions should be accompanied by a description of uncertainty and probability.
Understanding the Stressor
Qualitative and quantitative aspects of the stressors should be clearly articulated without bias with respect to desirability of outcome. The effect of other anthropogenic or natural stressors should be included in the analysis, because most ecological systems are affected by multiple stresses. For example, assessments of ecological risks of chemicals could increase reliance on field experiments in which test organisms are exposed to a suite of compounds and a range of natural conditions (this approach is already being widely used to set water quality criteria). One might also use a stressor classification to locate sensitive systems and sensitive components (e.g., species). Such a classification could include the components of the system potentially affected by the stressor, the pathway(s) of movement of the stressor, and the capacity of the affected component(s) to recover. Assessors should consider developing a matrix that considers the analytical method used to quantify stress versus class of stressor.
A good understanding of mechanisms of action can substantially improve understanding of stress-response relationships. Knowledge of mechanisms is not, however, a prerequisite for a useful risk assessment. Before a theory of mechanisms is used in a risk assessment, it must be validated in a realistic and comprehensive fashion.
OCR for page 315
Issues in Risk Assessment
Understanding the Response
Stressors have both direct and indirect effects, and both should be incorporated in the stress-response analysis. Timing of the stress (e.g., seasonality) can be critical in determining the response. A given stressor can have system-specific responses that vary geographically and with time.
Other Considerations
For the near term, some of the best data for assessing stress-response relationships in ecological systems are published empirical data on naturally or intentionally manipulated systems. For example, many whole-lake experiments have been conducted in which nutrient loading, acid deposition, and food-web structure have been manipulated (Schindler et al., 1985; Carpenter, 1989). Temporal and spatial scales of analysis should be appropriate to the stress and to the responses. For a point-source chemical release, the appropriate scale might be relatively small and short-term, depending on the dispersion pattern and degradation rate of the chemical and the life histories of the potentially affected organisms. For contamination or habitat change that affects large areas and threatens extinction of species, regional or global assessments that cover decades or centuries are appropriate.
Additions to the 1983 Paradigm Needed for Ecological Risk Assessment
The paradigm should include consideration of the resiliency of the ecological systems in question and the time to recovery. Both would vary with the type of system, geographic location, structure or process of interest, and type of stress.
The paradigm should discuss the choice of end point(s), including the relation of an end point to other major technical and societal concerns.
OCR for page 316
Issues in Risk Assessment
Application of Stress-Response Analysis In Case Studies
The group found that some form of stress-response analysis was used in nearly all the case studies, most obviously in the chemical-related assessments. Both laboratory-derived and field-derived stress-response information was used in the TBT studies. The agricultural chemical case study used various kinds of stress-response information, from biochemical studies to field experiments, although all focused on individual species populations. Consideration of multiple stressors would be useful. The TCDD modeling studies were driven by human health considerations and, therefore, used individual-level stress-response information.
In the spotted owl case, the stress-response relationship applies to habitat loss as the stressor and population viability as the response. In the case of harvesting, fishing is clearly the stressor, and yield and future recruitment are the end points. The latter study in particular was a good example of potential discontinuities in stress-response relationships, in that the Georges Bank haddock population, once depressed by overharvesting, did not recover after fishing pressure was relaxed. It is not clear whether the concept of stress-response relationships applies to species introductions.
Modeling Needs for Stress-Response Relationships
The group agreed that models are needed to deal with changes of state (e.g., shifts from bicarbonate to aluminum buffering), to incorporate multiple nonlinearities and discontinuities in multispecies systems, to extrapolate across ecological levels of organization (e.g., to assess the ecosystem consequences of a loss of a population or the population consequences of a loss in ecosystem function), and to make use of knowledge about synergies.
OCR for page 317
Issues in Risk Assessment
EXPOSURE ASSESSMENT
B. Leaderer and D. Porcella
The exposure assessment group agreed that, for applicability to ecological problems, the definition of exposure assessment should be generalized to accommodate both nonchemical and chemical stressors. The following definition was proposed: assessment of the extent and nature of the stressor and its co-occurrence with the target. Stressors can be physical, chemical, or biological. Examples of physical stresses are habitat loss, thermal loadings, and UV radiation. Chemical stressors include toxicants and nutrients. Biological stressors include species introductions and pest organisms.
Targets for exposure assessment can be at any level of biological organization, from individual organisms to ecosystems and the biosphere. Exposure assessment can involve direct measurement, indicators of exposure, and modeling. Extent refers to the magnitude and spatiotemporal distribution of the stressor. Nature refers to the characteristics peculiar to the stressor (e.g., physical and chemical properties of a chemical contaminant).
Methods of Measuring Stressors for Ecological Exposure Assessment
The group identified a wide range of methods applicable to measuring ecological exposures. The most obvious methods are the same kinds of direct and indirect methods used in human exposure assessment, including measurements of environmental contaminant concentrations in media to which organisms are exposed, measurements of uptake or body burden, and measurements of biochemical markers correlated with contaminant exposure.
Larger-scale tools for exposure assessment include remote sensing (habitat, productivity, and albedo) and aerial and ground-based mapping. Those methods are especially appropriate for such assessments as in the spotted owl study, in which habitat change, rather than a contaminant, is the stressor. Some participants suggested that ecosystem characteristics (measures of structure and function) can be used to quantify exposure.
OCR for page 318
Issues in Risk Assessment
There was a consensus that modeling for retrospective and prospective analysis will likely play a more important role in ecological exposure assessment than it normally does in human health risk assessment. Direct measurements with personal monitors or tissue-fluid analysis, the preferred methods of human exposure assessment, usually are not feasible or are prohibitively expensive in ecological assessments. Modeling was at least an underlying concept in all the case studies.
Test of the Definition
The group tested its proposed definition by attempting to fit it to the case studies and the 13 issues addressed by the EPA Science Advisory Board (SAB) Relative Risk Reduction Project (EPA, 1990). The group concluded that the new definition fit all six case studies, but that the definition provided in the Red Book fit only the three chemical case studies. Similarly, the new definition fit all 13 of the issues addressed by SAB, but the Red Book definition fit only about half. Two of the six case studies (on species introductions and harvesting) were related to issues not addressed by SAB.
RISK CHARACTERIZATION
G. W. Suter II and W. A. Farland
This group first developed a definition of risk characterization for ecological assessment and then applied the definition to the six case studies. As in health risk assessment, the principal objectives of risk characterization are to integrate information on exposure and effects and organize the results for presentation to risk managers, stake-holders, and the public.
Definition of Risk Characterization
The group determined that integration of exposure and exposure-response assessments is a complex process that requires a great deal of expert judgment. For the relatively straightforward case of predictive
OCR for page 319
Issues in Risk Assessment
assessment of risks associated with chemicals, ecological effects result from exposure to contaminants in ambient media, such as air, water, and sediment. These effects are functions of both the magnitude and duration of the exposure. Risk characterization requires that the assessor identify the dimensions of exposure and effects that are relevant to the risk estimate. The assessor must determine the relative positions of the expected exposure and effects in the (perhaps multidimensional) exposure-response space and then estimate the probability that the exposure exceeds some criterion of effects.
The above description applies well to predictive risk assessments of chemicals; however, other integration approaches might be more appropriate to other types of assessments. For example, when epidemiological methods are used to assess risks associated with apparent environmental damage (decline of a population, decline of forest stands, etc.) risk characterization would include evaluation of the strength of association, the plausibility of causation (given information on mechanisms), and the extent and magnitude of the observed effects.
Components of Risk Characterization
The group determined that characterization of risks for each end point and action should include the following components (not all must be included in presentations to all audiences):
An estimate of effects, including severity, frequency, spatial scope, temporal scope, and probability;
A description of the sources and magnitudes of uncertainty in the risk estimate;
An explanation of the assumptions used and a discussion of the plausible alternatives;
A discussion of the nature and quality of the models used (types used and existence and credibility of validation studies);
A discussion of the nature and quality of the data (quality assurance, quality control, relevance to the site, etc.);
Supporting lines of evidence (alternative models, different types of laboratory test data, and field monitoring that support the risk estimate);
OCR for page 326
Issues in Risk Assessment
the evaluation of species-recovery options. An alternative approach would be to develop a landscape-level model that would be used to evaluate habitat management options.
Georges Bank Fishery
As described in the case study, models are used extensively to assess the status of exploited fish stocks to quantify the relationship between fishing intensity and future abundance. Risk characterizations clearly delineate the effects of alternative harvesting strategies. However, the management decision-making process was described by the case study author as being disconnected from the scientific risk assessment exercise. The consensus of the group was that an adaptive management process, in which management itself is viewed as an experimental tool, is needed. The implementation of such an approach would require a closer connection between stock-assessment scientists and fishery managers.
General Discussion: Models and Risk Assessment
There was general agreement that modeling should have a prominent role in risk assessment. The participants agreed that models provide the only means to perform ecological risk assessments on large physical and organizational scales. Modeling should prove especially valuable for the more complex risk assessments required in the future (e.g., for release of genetically engineered organisms). It was clear from the case studies that models are being used in some settings. However, no consistent integration of modeling into risk assessment was evident. In particular, models were not routinely used in risk characterization or in evaluation of management alternatives. The Georges Bank assessment made the most extensive use of models, but even here the results of modeling did not appear to influence decision-making.
The group advanced a number of explanations for the lack of influence of models on risk management decisions:
OCR for page 327
Issues in Risk Assessment
Models are perceived as being too difficult to use and requiring too many data;
Risk managers do not understand the models and have little faith in their results;
The models are too difficult for risk assessors to use routinely;
Models sometimes lack credibility with decision-makers, because of lack of validation or conflicting results from alternative models.
The group agreed on four possible steps to increase the use of models in ecological risk assessment:
Development of a collaborative approach to risk assessment that includes both managers and modelers (risk assessment should be regarded as a process, not a discrete event);
Development of models with easier-to-use front ends or expert systems to ease risk assessors into the routine use of models;
Development of databases in tandem with models and risk assessments to provide means of validation and evaluation;
Encouragement of quantification of uncertainty through the use of Monte Carlo methods and multiple models that incorporate alternative process formulations.
UNCERTAINTY
R. Kimerle and E. P. Smith
Evaluation of uncertainty is a critical component of all risk assessments. Sources of uncertainty include limitations in knowledge, limitations in the use of models to approximate the physical world, and limitations in the parameters that are estimated and used in models to predict risk.
Uncertainties Identified In the Case Studies
The discussion group identified three general categories of uncertainty common to all six case studies:
OCR for page 328
Issues in Risk Assessment
Measurement uncertainties, e.g., low statistical power due to insufficient observations, difficulties in making physical measurements, inappropriateness of measurements, and natural variability in organic responses to stress;
Conditions of observation, e.g., spatiotemporal variability in climate and ecosystem structure, differences between natural and laboratory conditions, and differences between tested or observed species and species of interest for risk assessment;
Inadequacies of models, e.g., lack of or knowledge concerning underlying mechanisms, failure to consider multiple stresses and responses, extrapolation beyond the range of observations, and instability of parameter estimates.
Implications of Uncertainty for Ecological Risk Assessment
Most of the above uncertainties affect human health risk assessments, as well as ecological risk assessments. The consensus of the group was that knowledge-based uncertainties are often more important than uncertainties in parameter estimates. The usual statistical measures of uncertainty, p values and variance, measure only uncertainty due to random variation within the model; they do not account for uncertainties due to use of an incorrect model.
It was generally felt that the degree of uncertainty in ecological risk assessments increases with the level of biological organization. Models of ecosystem stress have higher uncertainties than models of populations and models of individual organism response. That is due in part to the increase in the number of end points available for modeling. Organism-level studies, such as single-species toxicity tests, usually have simple end points, such as survival and reproductive success. Ecosystem studies have the same end points plus additional ones that account for species interactions and measure community effects. Because of those uncertainties, ecological risk assessments still require substantial reliance on expert judgment and cannot be strictly model-based. Judgment-based approaches, such as the quotient approach to pesticide hazard assessment (described by Dr. Slimak in his plenary presentation) are often preferable to models for regulatory risk assessment.
OCR for page 329
Issues in Risk Assessment
The group noted that the degree of uncertainty that is acceptable in a study depends on the costs associated with the outcome of the risk assessment, the magnitude of expected effects, and the availability of alternatives to the hazardous agent being addressed. In the TBT study, although there were many uncertainties, once the risk to oysters was established, uncertainties about effects on other organisms were unimportant. The availability of alternatives to TBT as an antifouling agent further reduced the importance of the uncertainties.
Recommendations for Dealing With Uncertainty
A discussion of uncertainty should be included in any ecological risk assessment. Uncertainties could be discussed in the methods section of a report, and the consequences of uncertainties described in the discussion section. End point selection is an important component of ecological risk assessment. Uncertainties about the selection of end points need to be addressed.
Where possible, sensitivity analysis, Monte Carlo parameter uncertainty analysis, or another approach to quantifying uncertainty should be used. Reducible uncertainties (related to ignorance and sample size) and irreducible (stochastic) uncertainties should be clearly distinguished. Quantitative risk estimates, if presented, should be expressed in terms of distributions rather than as point estimates (especially worst-case scenarios). Power analysis or a discussion of sample size should be included in all studies involving collection of data and testing of hypotheses.
A continuing program of monitoring and experimental testing is needed to improve the accuracy and credibility of the process of ecological risk assessment. There are few standards for judging the accuracy of assessments, and continuing checks need to be made to increase confidence in the process.
VALUATION
W. Desvousges and R. Johnson
The discussion leaders began by summarizing their view of the role
OCR for page 330
Issues in Risk Assessment
of valuation in ecological risk assessment. Managing ecological risks requires a consistent means of comparing alternatives. Monetary values are an appropriate basis for such comparisons. Economic concerns influence several components of the risk assessment process, including hazard identification (which end points are worthy of societal concern?) and risk characterization (what are the economic implications of uncertainty?). Cost-benefit analyses are frequently a key aspect of risk management decisions.
The discussion leaders presented some methods for valuing ecological resources based on two assumptions of classical welfare economics—that societal values are sums of individual values and that people know and can express their willingness to pay (or accept compensation) for various risk policies. They then discussed some aspects of risk that influence individual decisions about willingness to pay or accept compensation:
Amount, content, frame, and source of information;
Decision heuristics;
Cause of damage;
Responsibility;
Degree of suffering;
Immediacy or delay of effects;
Morbidity or mortality.
They then discussed specific issues related to determining willingness to pay for preserving ecological resources. For recreational-use values (such as fishing, hunting, and birdwatching), techniques for valuation are reasonably well established. Current research in valuation focuses on nonrecreational values. There are two principal types of such values: ecological services (sometimes called services of nature) and existence value. Ecological services are services provided by ecosystems that otherwise would have to be provided by technology. The role of wetlands in pollution abatement and flood control is a good example of an ecological service. Existence values, more vaguely defined and more controversial, are defined by people's willingness to pay for the existence of particular populations or ecosystems, even if they never expect to use or see them.
The discussion leaders presented a tutorial on methods used to elicit existence values with questionnaires. There was much heated discussion.
OCR for page 331
Issues in Risk Assessment
In his summary, Dr. Johnson suggested that much better communication is needed between ecologists and the public and between ecologists and economists. Ecologists need to educate the public about the importance of preservation and must learn which aspects of nature the public values most highly. Economists need help from ecologists in educating people about the interactions between natural and human systems and in understanding motives for nonuse values. Ecologists need economists to help them understand both what people care about and how intensely they care. Ecologists also need economists to communicate effectively with risk managers who face competing demands for budgetary and regulatory resources.
RISK ASSESSMENT AND THE REGULATORY PROCESS
W. Cooper and D. W. North
Risk Assessment Has Many Uses
Because there are many uses for risk assessment, many forms of risk assessment are needed. The methodological approach and the level of detail in each form might differ a great deal, depending on the purpose for which risk assessment is carried out.
For strategic planning and setting priorities, it might be appropriate to conduct risk assessments that rely on expert judgment for direct assessments of relative risk. An example of the use of such an approach is the ecological risk portion of the recent EPA Science Advisory Report on Reducing Risk (EPA, 1990). With the direct approach, risk is assessed on the basis of overall integrated judgment to summarize each of the risks being compared. Modeling and other analytical tools are not used directly, but they can play an important role in providing the basis for expert judgment. The result of the risk assessment is a set of risk rankings that reflect the judgment of the assessors. The assessment also includes a discussion of the reasoning underlying the assessments, with explanation for differences among the experts. Because the direct approach relies on expert judgment, rather than mathematical formalism such as model calculations or statistical analysis to reach conclusions, the direct approach can be perceived as lacking in scientific rigor. However, the direct approach can be carried out quickly and might
OCR for page 332
Issues in Risk Assessment
provide extremely important guidance to nontechnical decision-makers, especially in the absence of any other form of integrated comparison among risks that are competing for scarce resources. In particular, such methods permit regulatory agencies to set priorities and research budgets in a proactive fashion. Such activities can counter the tendency to set priorities and research expenditures based on recent crises and public pressures—reaction to the pollutant of the month—rather than a comprehensive overview of competing risks.
Risk assessment is most often viewed as a quantitative process that is used to support specific risk management and resource management program decisions and policies. Among the biggest policy issues that involve ecological risk are acid deposition and global climate alteration. Neither of those was formally presented in the workshop, but participants in this work group frequently brought them up as examples of the most complex problems for ecological risk assessment. Application to problems of this scale is a massive undertaking. The six case studies were selected to be representative of major ecological issues of concern to government agencies. The case studies illustrate the complexities and uncertainties that the agencies must deal with on such issues. Participants observed that a complete risk assessment was not presented for any of the case studies. Yet, for each case study, a massive amount of information and analysis was described. At the local level, analytical resources are rarely available to deal with such a large amount of detail. But local communities and agency offices must deal with problems, such as remediation of hazardous waste sites, management of wildlife resources, and many other small-scale matters.
Risk assessment can provide scientific support to state or local agencies that are responsible for managing risk issues but lack the scientific and analytical resources of large federal agencies. Citizens groups might also have a strong interest in risk issues, but lack scientific capabilities and resources to carry out research and analysis. Risk assessment databases and monitoring efforts carried out by federal agencies to obtain baseline data can be useful to state, local, and citizens groups. Examples include the EPA-maintained IRIS database on toxic substances and the EMAP program the EPA is developing to obtain and make available data on ecological systems.
Risk assessment can provide guidance for identifying needed data and research. Such needs often become obvious when a risk assessment has
OCR for page 333
Issues in Risk Assessment
been carried out, and the conclusions on risk are found to depend on critical assumptions or elements of data, for which existing uncertainties can be resolved through research.
Risk assessment can be used for ''early warning"—a determination that an issue is of sufficient concern to place on the agenda so that existing policies, regulation, or legislation can be reconsidered. Advances in scientific understanding or changes in the stresses affecting an ecological system might indicate a potential for adverse changes that were not previously recognized. Recognition of the potential for adverse changes might allow these changes to be avoided through appropriate actions. Risk assessment can facilitate evaluation that permits earlier recognition and enables timely action.
In addition to situations in which timely warning of adverse changes is important, risk-based measures of ecological systems might facilitate continuing management activities to maintain, enhance, or restore the systems. Human-induced and other stresses interact in complex ways to affect ecological systems. Understanding how management policies affect the ability of an ecosystem to withstand or recover from stress will permit more effective management policies to be selected.
Different Risk Assessment Methods Are Suited to Different Risk Assessment Needs
The discussion above on the varied uses of risk assessment implies that there is not only one correct way to do risk assessment. Rather, risk assessment methods should be considered as a collection of tools from which analysts must select for the task at hand. In some cases, the tools must be developed, because the tools needed for a particular risk assessment task do not yet exist.
Risk assessment applications in similar situations might benefit from the same or very similar methodological approaches. Therefore, it will be important for public agencies and private organizations with similar needs for risk assessment to learn from each other's experiences. Both positive and negative experiences with models, databases, statistical procedures, and methods for assessing expert judgment can provide useful lessons for improving risk assessment practice. One lesson
OCR for page 334
Issues in Risk Assessment
learned from ecological risk assessments is that the power of an expensive test to reject or confirm a hypothesis should be evaluated before data collection. If the data are unlikely to provide a basis for rejecting or confirming a hypothesis important to the risk assessment, then it might not be worth the expense to obtain the data.
As the volume of risk assessments grows, it will be particularly important to ensure quality and consistency. Development and use of formal guidelines, training of risk assessors, and communication of examples of good risk assessment practice will help agencies and organizations to ensure quality and consistency in their applications.
Consistency and flexibility must be balanced appropriately in the risk assessment process. Consistency motivates doing risk assessments for similar situations in a similar and predictable way. Flexibility motivates departures from a standard risk assessment approach when scientific information indicates that differences are important for the proper assessment of risk. In practice, it might be appropriate to have standard or default procedures that are used when scientific information is not sufficient to motivate a different approach, and provisions for innovative exceptions that are supported by applicable scientific information.
Risk assessment should not become too rigid. Its purpose is to summarize and communicate applicable science to meet the needs of policy makers. That task by its very nature requires flexibility and creativity, not reliance on formulas or cookbook recipes evolved from past practice.
Risk Assessors and Risk Managers Need to Communicate
Managers responsible for ecological systems must be responsive to the public, and risk assessors should recognize that their task supports risk management. Risk assessment can help risk managers to explain the basis for their decisions to interested and potentially affected groups. Risk assessment, therefore, has an important function as communication.
As was stressed in a recent National Research Council report on risk communication (NRC, 1989), such communication should be two-way. To ensure that communication is effective and that public concerns are addressed, it is generally useful to involve the public while the risk
OCR for page 335
Issues in Risk Assessment
assessment is being prepared. Interested and affected groups should be informed in advance about the risk assessment. They should have the opportunity to express their concerns and contribute information to the risk assessment process while the process is being carried out.
Issues that seem obvious to the expert scientists participating in the risk assessment might not be obvious to laypersons, but it is important for both to understand each other if there is to be effective bridging between the scientific knowledge available to the experts and the concerns of the public. As one working-group participant expressed it, "if risk assessment opens up a dialogue, then it serves an appropriate objective."
Communication between modelers, risk assessors, and managers should be mutual, iterative, timely, and flexible. Risk assessments will be valuable as support to the risk management process only if the assessments address the right problem and if the managers who are the users of the products of risk assessment understand them. One suggestion offered at the workshop is that an agency assign someone the task of being the translator, or liaison, between the group that has carried out the risk assessment and the users of the risk assessment.
Credibility is Crucial
Risk assessments will be useful to the extent that they are perceived to be effective in accomplishing a difficult task: summarizing what science can tell us about the possible consequences to an ecological system. If a risk assessment is perceived to be incomplete or biased toward a particular point of view, it will not be trusted for risk management decision-making. It is therefore essential that a risk assessment be a comprehensive and balanced summary of the applicable science.
How can comprehensiveness and balance be achieved? The recommendations on health risk issues from the 1983 report appear equally applicable to ecological risk issues:
Regulatory agencies should take steps to establish and maintain a clear conceptual distinction between assessment of risks and the consideration of risk management alternatives; that is, the scientific findings and policy judgments embodied in risk assessments should be explicitly distinguished from the politi-
OCR for page 336
Issues in Risk Assessment
cal, economic, and technical considerations that influence the design and choice of regulatory strategies.
Before an agency decides whether a substance [ecosystem stressor] should or should not be regulated, … a detailed and comprehensive written risk assessment should be prepared and made publicly accessible …
An agency's risk assessment should be reviewed by an independent science advisory panel before any major regulatory action or decision not to regulate.
In those recommendations, it might be appropriate to replace "regulatory" language with more general terms relevant to the broad range of decision alternatives available for the management of ecological risks. However, the principles embodied in the recommendations can be applied essentially unchanged: to promote credibility, establish and maintain the conceptual distinction between risk assessment and risk management; place risk assessments in a written, publicly accessible form; and subject them to peer review by outside scientists.
Representative terms from entire chapter:
risk characterization
