Ecological Risk Analysis: Prospects and Problems

摘要

Two methods for quantifying ecological risks in probabilistic terms are discussed. In Analysis of Extrapolation Error, effects thresholds for fish species in the field are estimated, with error bounds, from observed effects on laboratory test species. In Ecosystem Uncertainty Analysis, Monte Carlo simulation and ecosystem models are used to extrapolate from laboratory toxicological data to estimated risks to populations and trophic levels. At the present state of development and validation of ecological risk models, uncertainties associated with all predictions are enormous. Although accurate predictions of absolute magnitudes of risks to populations and ecosystems are not now possible, risk analysis can be used to compare or rank sources of stress or rank species in order of relative sensitivity. Probabilistic risk models can also be used to compare the relative importance of disparate sources of uncertainty in risk estimates. Limited uses of risk analysis in environmental management and regulation appear feasible, provided they are based on either water quality or on organism-level effects rather than on ultimate effects on populations/ecosystems. More ambitious applications of risk analysis in ecology are currently limited by insufficient toxicological data and by insufficient understanding of how real populations and ecosystems respond to stress. (ERA citation 08:042256)