Ecological insights into risk analysis: the side-effects of pesticides as a case study

Abstract

Ecotoxicology addresses a vast array of pollution problems that span all classes of habitats, chemical pollutants and organisms. In order to manage this complexity and develop a more predictive and less descriptive approach, general methods of ecotoxicological risk analysis are needed which incorporate appropriate ecological theory with toxicology and environmental chemistry. The current emphasis on toxicological criteria, at the expense of ecological factors, has led to methodologies that describe local toxic effects in great detail, but which fail to consider the importance of these effects for populations of the organisms that are suffering them. Application of ecological ideas to risk analysis will lead to a more probabilistic approach, which predicts ranges of possible outcomes for a given pattern of exposure to a pollutant: this should indicate that there is no single concentration of a pollutant that guarantees protection to a particular community of organisms. It may also tell ecotoxicologists that they should be measuring different things in order to estimate risk. This paper describes qualitative and quantitative approaches to ecological risk analysis for the short and long-term effects of pesticides on non-target arthropods. Criteria for identifying sensitive indicators for short-term effects and indicators of longer-term effects are given. Research into the spatial dynamics of predatory invertebrates in farmland is then used to argue that the probability of local extinction may be an appropriate endpoint for long-term risk analysis. This analysis indicates that the pattern and frequency of exposure to the toxic chemical and the landscape structure are important determinants of risk for pollutants with short persistence, pollutants in temporary habitats or pollutants affecting dispersive invertebrates.