|
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 293
Issues in Risk Assessment
Appendix E
Case Studies and Commentaries
CASE STUDY 1:Tributyltin Risk Management In the United States
R. J. Huggett and M. A. Unger, Virginia Institute of Marine Sciences
Tributyltin (TBT) is a chemical with a variety of biocidal applications, including use as an antifouling agent in boat paints (Blunden and Chapman, 1982). Biological effects of TBT on marine and estuarine organisms and the concentrations of TBT that induce them vary widely among species (Huggett et al., 1992). A water concentration of 1,000 ng/L (1 part per billion) is lethal to larvae of some species, and nonlethal effects have been observed at concentrations as low as 2 ng/L (2 parts per trillion, ppt). Both laboratory and field studies of toxicity were initially hampered by difficulties in measuring the low concentrations that were toxic to some organisms.
Adverse effects on nontarget organisms, including commercially valuable species of shellfish, were observed in Europe in the early 1980s (Alzieu, 1986; Abel et al., 1986). Abnormal shell growth was documented in Crassostrea gigas (European oyster) and linked through laboratory experiments to TBT leached from antifouling paints. That connection led to restrictive regulations in France (in 1982) and Great Britain (in 1985 and 1987). In the United States, concentrations exceeding those determined experimentally to be effective have been found in many
OCR for page 294
Issues in Risk Assessment
areas, particularly in harbors with large marinas. Snails in the vicinity of a marina on the York River, Virginia, were shown to have an abnormally high incidence of imposex (expression of male characteristics by female organisms), an effect previously observed under laboratory conditions in female European oysters, Ostrea edulis (Huggett et al., 1992). EPA began to assess effects of TBT in 1986, but has not yet issued any regulations. Meanwhile, restrictive actions have been taken by states and by the Congress.
A proposal by the U.S. Navy to use TBT paints on its entire fleet was prohibited by Congress in 1986, despite a Navy study that predicted no adverse environmental impact. Virginia enacted legislation and an emergency regulation in 1987, and Maryland, Michigan, and other states have since taken similar actions. Congress enacted national legislation restricting use of TBT paints in 1988. Those actions generally banned or restricted the use of TBT paints on small boats (less than 25 m long) and placed limits on leaching rates from paints used on larger vessels. Studies in Virginia had shown that most TBT releases were from small boats. Small-scale monitoring studies (e.g., in France and Virginia) have shown that the restrictions have been effective in reducing environmental concentrations and adverse impacts of TBT.
Risk management of TBT has been unusual in several ways. The initial basis for concern was field observation of adverse effects, not extrapolation from laboratory bioassays and field chemistry data. Risk assessment and risk management were conducted by state agencies and legislatures, rather than by EPA. Although the risk assessments were made without formalized methods, the results of the independent assessments were the same. Finally, TBT is the first compound banned by the Congress and the first regulated for environmental reasons alone.
Discussion
(Led by L. Barnthouse, Oak Ridge National Laboratory, and P. F. Seligman, Naval Ocean Systems Center)
The case study addressed, with differing completeness, each of the five recommended steps in risk assessment and management. Hazard identification included the observation of abnormalities in the field and the same effects in experimentally exposed animals. Dose-response identification included data both from the field (correlative) and from the laboratory (experimental). Exposure assessment was based on estimated
OCR for page 295
Issues in Risk Assessment
use and release rates rather than on monitoring or modeling studies. Risk characterization was only qualitative; it did not address such issues as the number and distribution of species that were vulnerable, or the degree of damage to the shellfish industry. Risk management actions were based on the demonstrable existence of hazard, on societal concern for the vulnerable species, and on the ready availability of alternative antifouling agents.
Some workshop participants were critical of the risk assessment approach adopted by Congress and state regulatory agencies. No attempt was made to plan and execute a formal risk assessment. Risk identification was based primarily on data on nonnative species. The Eastern oyster and blue crab, the species putatively at greatest risk, have been found to be less sensitive. Regulatory responses were based on findings of high environmental concentrations of TBT in yacht harbors and marinas, rather than in ecologically important regions such as breeding grounds. The central issue is whether a safe loading capacity (environmental concentration) of TBT for nontarget organisms can be defined, given substantially reduced rates of input. Recent information on fate and persistence, chronic toxicity, and dose-response relationships could support a more quantitative risk assessment with the possibility of more or less stringent restrictions.
CASE STUDY 2:Ecological Risk Assessment for Terrestrial Wildlife Exposed to Agricultural Chemicals
R. J. Kendall, Clemson University
The science of ecological risk assessment for exposure of terrestrial wildlife to agricultural chemicals has advanced rapidly during the 1980s. EPA requires detailed assessments of the toxicity and environmental fate of chemicals proposed for agricultural use (EPA, 1982; Fite et al., 1988). Performance of an ecological risk assessment requires data from several disciplines: analytical toxicology, environmental chemistry, biochemical toxicology, ecotoxicology, and wildlife ecology.
Addressing the ecological risks associated with the use of an agricultural chemical involves a complex array of laboratory and field studies—in essence, a research program. This paper provides examples of
OCR for page 296
Issues in Risk Assessment
integrated field and laboratory research programs, such as The Institute for Wildlife and Environmental Toxicology (TIWET) at Clemson University. Preliminary toxicological and biochemical evaluations include measurements of acute toxicity (LC50 and LD50), toxicokinetics, and observations of wildlife in areas of field trials. Assessment of reproductive toxicity includes studies with various birds and other wildlife, particularly European starlings that nest at high densities in established nest boxes; these studies include measurements of embryo and nestling survival, postfledgling survival, behavior, diet, and residue chemistry (Kendall et al., 1989). Nonlethal assessment methods include measurement of plasma cholinesterase activity associated with organophosphate pesticide exposures (Hooper et al., 1989). A wide variety of birds, mammals, and invertebrates have been used in these studies.
End points evaluated in wildlife toxicological studies include mortality, reproductive success, physiological and biochemical changes, enzyme impacts, immunological impairment, hormonal changes, mutagenesis and carcinogenesis, behavioral changes, and residues of parent compounds and metabolites (Kendall, 1992).
The paper includes a case history of a comparative evaluation of Carbofuran and Terbufos as granular insecticides for control of corn rootworms. Carbofuran has been responsible for many incidents of wildlife poisoning and is recognized as being very hazardous to wildlife. In contrast, although Terbufos is highly toxic to wildlife in laboratory studies, exposure of wildlife under field conditions appears generally to be relatively low, and widespread mortality is not evident. Field studies of Terbufos conducted by TIWET might be the only ones conducted to date that satisfy EPA's requirements for a Level 2 field study, a more quantitative assessment of the magnitude of the effects of a pesticide than the qualitative Level 1 studies. (Level 2 studies are performed when toxicity tests and use patterns suggest a detailed study is warranted.) Data generated in those studies support an ecological risk assessment for Terbufos that is reported in the paper. However, the research program on Terbufos represents many years of effort with integration of laboratory and field research to achieve a full-scale level 2 study in just one geographic area on one crop. Ecological modeling techniques will be needed to generalize the results to other chemicals or to other situations.
OCR for page 297
Issues in Risk Assessment
Discussion
(Led by B. Williams, Ecological Planning and Toxicology, Inc., and J. Gagne, American Cyanamid Company)
Dr. Williams noted that each step in ecological risk assessment is more complex and less understood than the corresponding step in human health risk assessment. Although hazard can be assumed when a toxic chemical is released, the species and populations at risk must first be defined. The appropriate selection of surrogate species for testing in the laboratory is usually unclear. Measurement of environmental concentrations is only the first step in exposure characterization. Exposure assessment also requires consideration of foraging behavior, avoidance, and food-web considerations, as well as spatial and temporal variability. Risk characterization involves comparison of exposure estimates with measures of hazard; this process might result in compounding of errors. Ecological risk assessments do not track individuals over time and so do not accurately reflect population changes.
The activities presented in the case study have a large research component, which is focused on dose-response assessment and exposure assessment. One discussant characterized risk assessment, as presented in the case study, as a retrospective exercise based on focused characterization of hazard and exposure in wildlife. Given the difficulties in conducting environmental risk assessments, the four-part paradigm might not be applicable at levels of organization above that of the population.
CASE STUDY 3A:Models of Toxic Chemicals in the Great Lakes: Structure, Applications, and Uncertainty Analysis
D. M.DiToro, Hydroqual, Inc.
This paper reviewed and summarized efforts to model the distribution and dynamics of toxic chemicals in the Great Lakes, with applications to PCBs, TCDD, and other persistent, bioaccumulated compounds. The models were based on the principle of conservation of mass (Thomann
OCR for page 298
Issues in Risk Assessment
and Di Toro, 1983). Analysis proceeded through five steps: water transport, dynamics of solids, dynamics of a tracer, dynamics of the toxicant, and bioaccumulation in aquatic organisms. Mechanisms considered include settling, resuspension, sedimentation, partitioning, photolysis, volatilization, biodegradation, growth, respiration, predation, assimilation, excretion, and metabolism. The model of toxicant dynamics considered three phases (sorbed, bound, and dissolved) in each of two media (water column and sediments) and 21 pathways into, out of, or between these phases. The model of bioaccumulation included 25 compartments (four trophic levels with one to 13 age classes at each level) with five pathways into or out of each compartment. Because of the large number of coefficients (rate constants), sparseness of knowledge of inputs, and little opportunity for field calibration, uncertainty analysis was important in all the modeling exercises.
The first example modeled the dynamics of total PCBs in Lake Michigan (Thomann and Connolly, 1983). Plutonium-239 was used as a tracer to analyze sediment dynamics, and the model suggested that resuspension is an important mechanism. Calculation of PCB concentrations was limited by an order-of-magnitude uncertainty in the mass loading. Predictions of PCB concentrations and their rate of decline were sensitive to the value assumed for the mass-transfer coefficient for volatilization.
The second example modeled TCDD in Lake Ontario and attempted to predict the relationship between one source of input and the resulting incremental concentrations of TCDD (Endicott et al., 1989). In the absence of knowledge of other inputs, field data could not be used to calibrate the model. Hence, a formal uncertainty analysis was performed with Monte Carlo methods and assumed probability distributions of the rate coefficients. The 95% confidence limits of predicted TCDD concentrations in water and sediment differed by a factor of 10-100. Uncertainties in rate constants for photolysis and volatilization were the most important sources of uncertainty in predicted TCDD concentrations.
The third example extended the Lake Ontario TCDD model to eight other hydrophobic chemicals and incorporated a food-chain model to predict concentrations in lake trout (Endicott et al., 1990). The model predicted wide differences in toxicant concentrations, depending primarily on the degree of hydrophobicity as indexed by the octanol-water
OCR for page 299
Issues in Risk Assessment
partition coefficient, Kow. The range of uncertainty in the predicted concentrations also varied among the chemicals. In-lake removal processes (sedimentation, volatilization, and degradation) were important for all chemicals.
CASE STUDY 3B: Ecological Risk Assessment of TCDD and TCDF
M. Zeeman, U.S. Environmental Protection Agency
This paper is based on a full-scale ecological risk assessment of chlorinated dioxin and furan emissions from paper and pulp mills that use the chlorine bleaching processes (Schweer and Jennings, 1990). Although the risk assessment addressed potential risks to terrestrial and aquatic wildlife exposed to TCDD and 2,3,7,8-tetrachlorodibenzofuran (TCDF) via a number of environmental pathways, the case study was limited to exposure of terrestrial wildlife to TCDD resulting from land disposal of paper and pulp sludges. This route of exposure was identified as one of the most hazardous in the multiroute risk assessment.
The specific exposure pathway considered was uptake of TCDD by soil organisms (earthworms and insects) from soil to which pulp sludge has been applied, and the consumption of soil organisms by birds and other small animals. Transfer factors were estimated both by modeling and from data collected in a field study in Wisconsin, in which an average soil TCDD concentration of 11 ppt led to concentrations of up to 140 ppt in a composite of six robin eggs. The models used three alternative sets of assumptions: low estimate, best estimate, and high estimate. The best estimates of tissue concentrations derived from the model were often similar to those observed in the field study: the low and high estimates were lower and higher, respectively, by a factor of roughly 10.
Risk estimates for terrestrial wildlife were derived by comparing exposure estimates (usually converted to daily intake rates) with benchmark toxicity values. The values used as benchmarks were either lowest-observed-adverse-effect levels (LOAELs) or no-observed-adverse-effect levels (NOAELs) for reproductive toxicity in birds and mammals
OCR for page 300
Issues in Risk Assessment
—specifically, the lowest reported LOAELs and NOAELs. The risk quotient (RQ) for each species considered was defined as the ratio of the estimate of exposure to the corresponding benchmark value. On the basis of transfer estimates for land disposal of paper sludges, RQs could exceed 60:1 for the most exposed species (robins, woodcocks, and shrews). To estimate soil concentrations of TCDD ''safe" for these species, two uncertainty factors of 10 could be applied: one to allow for interspecies variability in sensitivity and one for an extrapolation from laboratory to field and/or the use of a LOAEL as the benchmark value. The corresponding estimates of safe concentrations were estimates that would lead to RQs less than 0.01:1 for the most heavily exposed species considered. Under those assumptions, soil concentrations of TCDD safe for highly exposed species would be about 0.03 ppt.
Discussion
Led by L. A. Burns, U.S. Environmental Protection Agency, and D. J. Paustenbach, McLaren/Hart)
These case studies present only estimates of environmental concentrations—i.e., exposure assessment—and do not address other elements of risk assessment. Compared with traditional human health assessments, they show a greater concern for accuracy (as opposed "policy-driven conservatism"), a greater use of formal uncertainty analysis, and better opportunities for verifying accuracy of exposure and uptake models.
Criticism of the models focused on the omission of processes and on the assumed linear relationship between loading and environmental concentrations. Omitted processes include in-lake generation of solids (phytoplankton), transport in the benthic boundary layer, effects of water clarity on photolysis rates, and daily cycles in pH. A nonlinear relationship between loading and toxicant concentrations might occur if the toxicant reaches high enough concentrations to change the processes that control its own fate. For example, reduction in fish populations might allow for higher populations of zooplankton, which clarify the water column by decreasing populations of phytoplankton, thereby increasing photolysis rates and stabilizing pH.
OCR for page 301
Issues in Risk Assessment
CASE STUDY 4: Risk Assessment Methods in Animal Populations: The Northern Spotted Owl as an Example
D. R. Anderson, U.S. Fish and Wildlife Service
This paper described an analysis of northern spotted owl population dynamics performed to support ongoing studies of the impacts of clear-cutting of old-growth forest on the prospects for future survival of this endangered species (Salwasser, 1986). The paper summarized a method for estimating rates of population increase or decrease based on capture-recapture techniques and illustrates the methods with data on the northern spotted owl. The method proceeds in three steps: use of capture-recapture data to estimate age-specific survival or fecundity rates, estimation of the finite rate of population change (Leslie's parameter λ), and experiments on samples of marked animals in natural environments. Mathematical models for estimating population parameters, including λ, have been developed extensively, and computer programs are available (Burnham et al., 1987). Experimental studies are desirable to test hypotheses about relationships between population parameters and risk factors.
The case study was of a population of northern spotted owls in California studied for 6 years (Franklin et al., 1990). Capture-recapture data yielded estimates of age-specific survival and fecundity for females, as well as estimates of mean population size (37 females) and annual recruitment (0 to 19 females; mean, 8). On the average, the eight females entering the population each year would have included six immigrants from outside the study area and only two locally raised recruits. The calculated value of λ was 0.952 ± 0.028, which indicated a decreasing population.
In this case, the risk factor was clearance of the old-growth forest on which the species is believed to depend. Although the study area contained much suitable habitat, the population appeared not to be self-sustaining, but to be maintained by immigration from remaining areas of old-growth. It was suggested that the study population is temporarily above the long-term carrying capacity because of the drastic loss of
OCR for page 302
Issues in Risk Assessment
habitat in surrounding areas; these circumstances lead to a large "floating" component of the population.
The paper concluded that risk assessment in higher vertebrate populations must often rely on analysis of samples of marked individuals. A robust theory exists for study design and the analysis of such data. Selection of appropriate models is critical for rigorous assessment of impacts. Analysis of capture-recapture data allows inferences about the separate processes of birth, death, emigration, and immigration. Risk to a population does not affect population size directly; rather, it acts on the fundamental processes of birth and death.
Discussion
(Led by M. E. Kentula, U.S. Environmental Protection Agency, and O. L. Loucks, Miami University)
Dr. Kentula commented that the case study (like others in the workshop) focused on individuals and populations and thus took a bottom-up approach. An alternative, top-down approach is to conduct an ecosystem risk assessment from a landscape perspective. For example, Kentula stated that EPA's Wetlands Research Program is developing methods to assess impacts on landscape function due to cumulative wetlands loss (Abbruzzese et al., 1990). The method proceeds in two-stages: a landscape characterization map is used to classify and rank units of the landscape according to relative risk, and can also be used to set priorities for effort and allocation of resources; a response curve expresses the hypothesized relationship between stressors (such as loss or modification of wetlands) and reduction in landscape functions (e.g., maintenance of water quality, or life support). The system can be used both to identify areas at risk and to guide management decisions for landscapes that are already affected.
Dr. Loucks commented that the case study presents the consequences of the stress to one local owl population at one time. For assessment of risk to the regional or total population, one would need to construct a "dose-response" relationship, in which "dose" would be a measure of the degree of stress (e.g., the percentage of the old-growth forest that has been destroyed) and "response" would be the probability of extinction of the population within an appropriate period (e.g., 250 years). Calcula-
OCR for page 303
Issues in Risk Assessment
tion of the probability from the birth, death, and dispersal rates estimated in the case study would require stochastic population modeling that takes account of uncertainty and variability in the population parameters.
The Endangered Species Act is an example of preemptive risk management, in that a high probability of extinction of a single species is designated as unacceptable. A species-by-species approach, however, does not lead to quantitative assessment of the risk of impoverishment of an ecosystem. Where possible, ecological risk assessment should work across levels of organization and should assess risks of reduction in system utility.
CASE STUDY 5:Ecological Benefits and Risks Associated with the Introduction of Exotic Species for Biological Control of Agricultural Pests
R. I. Carruthers, USDA Agricultural Research Service
The accidental or deliberate introduction of exotic species into regions where they are not native can cause positive, negative, or no observable effects, depending on a wide variety of biological, sociological, economic, and other factors. About 40% of the major arthropod pests (Sailer, 1983) and 50-75% of weed species (Foy et al., 1983) in the United States are introduced species, and introduced pests also include vertebrates, mollusks, and disease organisms that affect animals and plants. Many countries have developed formal programs to limit the introduction and establishment of unwanted exotic organisms, and many have developed methods to assess benefits and risks associated with planned introductions. The United States has no federal statute or set of statutes that governs introductions; instead, it has cumbersome and sometimes conflicting regulations, protocols, and guidelines.
This paper addressed assessment of risks and benefits of "classical biological control" (CBC): the planned introduction of exotic enemies of an introduced pest collected from the pest's home range (DeBach, 1974). Classical biological control (either alone or integrated with other pest management methods) has frequently been successful in controlling
OCR for page 304
Issues in Risk Assessment
introduced pests and often provides large economic or environmental advantages over alternative methods. An example given in the paper is control of the alfalfa weevil: introduction and widespread releases of 11 species of parasitic hymenoptera have yielded substantial control of this major pest with no known negative side effects and with an estimated benefit-to-cost ratio of 87:1.
Risks of CBC programs have three different sources: the organism itself (e.g., parasitism or predation on nontarget species), associated organisms (e.g., pests of the introduced beneficial organism), and unrelated passenger organisms arriving with shipments of the introduced organism. Some adverse effects of all three types have been documented (Pimentel et al., 1984, Howarth, 1991), including local extinctions of nontarget species, especially in island situations. Although there is little documentation of notable adverse impacts of CBC programs in the United States, more precise prediction of benefits and risks would be desirable. Unfortunately, accurate prediction of both positive and negative impacts (target and nontarget effects) of CBC programs has not been achieved. The lack of predictive ability leaves CBC risk assessments in the realm of informed scientific judgment-based on limited published data.
In addition to requirements of various federal laws, guidelines have been developed to improve safety in CBC. Agricultural Research Service protocols (now under revision) require federal permits for importation and movement of organisms, quarantine, authoritative identifications, environmental and safety evaluations, documentation of movements and releases, and retention of voucher specimens. Current policy requires an environmental assessment (EA) to accompany applications for permits for field release of exotic organisms. Although the components of an EA depend on the specific situation, the documentation required is fairly extensive. At any step in the process, a proposed introduction can be deemed inappropriate and the project terminated.
Discussion
(Led by J. T. Carlton, Williams College, and D. Policansky, National Research Council)
Classical biological control is only one kind of introduction of nonnative species. Others include range expansions (either natural or mediat-
OCR for page 305
Issues in Risk Assessment
ed by human modification of habitats), deliberate introductions to "improve nature" or for aquaculture or horticulture, and a wide variety of accidental introductions. CBC seems to have a better safety record than other types of introduction. It is not clear whether this is because the activity is basically benign, because the safety precautions work well, or because CBC involves small organisms that pose smaller risks than larger organisms. The worst failures in all categories have occurred in insular environments such as islands and lakes.
The assessment of risks posed by introductions has been addressed separately by scientists in different disciplines (e.g., agriculture, freshwater and marine ecology, and nature conservation). Communication between the disciplines is poor, and several sets of criteria, procedures, and protocols have been developed independently. Whereas the U.S. Department of Agriculture has adopted flow charts as a way to systematize decision-making, other agencies (e.g., the International Council for the Exploration of the Sea) have concluded that too little is known about ecosystem functioning for flow charts to be useful.
Dr. Policansky commented that risk assessment for species introductions is difficult to fit into the four-step Red Book paradigm. Hazard is taken for granted (because it is the introduction of the species itself); dose-response and exposure are yes-no categories, not continuous variables, because the more important point is whether the species is present or not, not how much of the species is present. A more suitable paradigm might be that presented in the 1986 NRC report Ecological Knowledge and Environmental Problem-Solving: Concepts and Case Studies, which placed more emphasis on problem-scoping and problem-solving than on categorical activities.
CASE STUDY 1:Uncertainty and Risk in an Exploited Ecosystem: A Case Study of Georges Bank
M. J. Fogarty, A. A. Rosenberg, and M. P. Sissenwine, National Marine Fisheries Service
This paper addressed the risks of overexploitation of harvested marine
OCR for page 306
Issues in Risk Assessment
ecosystems, with specific application to Georges Bank, a highly productive area off the northeastern United States. In this context, risk assessment involves determining the probability that a population will be depleted to an arbitrarily predetermined "small" (e.g., 1% or 5%) size. The "quasi-extinction" level may be defined (Ginzburg et al., 1982) as (1) the population level below which the probability of poor recruitment increases appreciably or (2) the smallest population capable of supporting a viable fishery.
The primary determinant of the long-term dynamics of any population is the relationship between the adult population (stock) and recruitment. The null hypothesis is that the relationship is linear, i.e., that recruitment is independent of density (Sissenwine and Shepherd, 1987). Compensatory changes in survival or in reproductive output result in nonlinear stock-recruitment curves. Nonlinearity permits stable equilibrium under harvesting pressure (i.e., under increased mortality rates), up to a critical exploitation level, beyond which the population will decline to quasi-extinction. Stochastic variation in the stock-recruitment relationship or in multispecies interactions can increase risks of adverse effects at moderate exploitation levels. In practice, because of uncertainties resulting from stochastic variations and measurement errors, it is often impossible to reject the null hypothesis of no compensation. Assuming there is no compensation will, in general, result in a conservative assessment of production capacity and its ability to withstand exploitation.
Haddock populations on Georges Bank fluctuated about relatively stable levels between 1930 and 1960 when the fraction of the total haddock population killed per year by fisherman (annual fishing mortality rate) varied between 0.3-0.6, but collapsed after the fishing mortality rate increased to 0.8 during the 1960s (Grosslein et al., 1980). The empirical relationship between stock and recruitment was extremely variable with little indication of the form of the underlying curve. Analysis of the population dynamics showed that a density-independent null model could not be rejected and gave a neutral equivalent harvest rate of 0.5, which agrees well with the stable period of the fishery. In contrast, the compensatory model is over optimistic with respect to the long-term harvest rate.
The decrease in populations of haddock and other groundfish was accompanied by increases in other species, notably elasmobranchs (rays and sharks). The biomass of predatory species increased dramatically
OCR for page 307
Issues in Risk Assessment
with attendant consequences for the overall system structure (Fogarty et al., 1989). Population modeling suggests that the stock-recruitment relationship for haddock might have been changed and that the population cannot now withstand as heavy fishing mortality as it could before the increase in predation pressure.
Risk assessment for exploited systems must take into account uncertainties in population abundance, harvest rates, and system structure. Adoption of risk-averse management strategies would minimize the possibility of stock depletion or undesirable alterations in the structure of the system.
Discussion
(Led by R. M. Peterman, Simon Fraser University, and J. L. Ludke, National Fisheries Research Center-Leetown)
Discussion focused on the idea of statistical power—the probability that an experiment (or set of observations) will correctly reject a null hypothesis that is false, i.e., the probability that an experiment will detect effects that actually exist. In fisheries cases, the high degree of variability in population parameters means that most studies have very low power to detect changes, unless the studies are continued for many years or involve frequent measurements (Peterman and Bradford, 1987). Published papers in fisheries biology (and in other disciplines related to risk assessment) rarely report statistical power and hence can misleadingly report negative findings. The case study recommended adopting a conservative null hypothesis to allow for the low power of the observational studies. Other approaches are to improve the design of studies (e.g., by more frequent sampling), to incorporate uncertainties into formal decision analysis, and to reverse the burden of proof (to put the burden of documenting whether detrimental effects are occurring on exploiters of the resource, rather than in the management agency). If "proof" of safety is required, a formal statement of the power of studies should be provided for a size of effect deemed relevant.
The Georges Bank fishery is only one of a long series of cases in which overexploitation has occurred despite a nominal system of scien-
OCR for page 308
Issues in Risk Assessment
tific stock assessment and fishery management. Discussants generally felt that overexploitation was due to failures of management, rather than to deficiencies in assessment or failure to communicate results to managers.
The assessment of the risk to fish populations associated with exploitation in the Georges Bank case study is implicitly consistent with the 1983 health risk assessment framework, although the explicit steps differ. The case study illustrates the 1983 risk assessment paradigm within the larger context of problem-solving. However, the dose-response and exposure steps might be only loosely analogous. Differing circumstances of function, scale, and certitude could require variation in the method of risk assessment.
The numerous sources of uncertainty in assessing risk associated with exploitation of fish populations vary and increase in magnitude with increase in scale. Regulation of harvest of geographically confined populations can be achieved with greater confidence than can regulation of wide-ranging populations such as Chesapeake Bay striped bass and Lake Michigan lake trout. Sources of uncertainty include variation in recruitment, measurement (which requires many assumptions), and management and institutional characteristics. Management techniques for reducing risks associated with overexploitation of populations are fairly blunt instruments, and strong actions are usually taken only after the fact. Rarely, if ever, are risk reduction measures considered until an actual impact is noticed or a potential threat emerges.
Subtle and cumulative factors that are unknown or are measured imprecisely—e.g., chronic or episodic changes in predation, migration, and disease—are some of the issues with information gaps that contribute to uncertainties in ecological risk assessment. The Georges Bank case study describes multispecies interactions and consequences of selective harvesting practices within the fish community, but falls short of a systematic understanding of cause and effect with regard to changes in multispecies abundance.
Representative terms from entire chapter:
georges bank
