ABSTRACT
In order to obtain registration of plant protection products, extensive risk assessments are carried out to demonstrate that the products’ use will cause no unacceptable effects on nontarget organisms. Currently, ecological risk assessments of plant protection products make extensive use of toxicity exposure ratios (TERs) gained from single-species tests, which are normally performed under laboratory conditions. Such TERs are good measures of the risk posed to individuals (European Commission 1997, 2002a, 2002b), and combined with safety factors (assessment factors), they are highly conservative and therefore suitable for lower-tier risk assessment. However, for most species, the protection goal is not the individual but the population or community (European Commission 2002a, 2002b; Pastorok et al. 2002, 2003; Sibly et al. 2005). Model ecosystem tests aim at measuring effects on populations and communities. Thus, in higher-tier risk assessments, the TERs are often calculated using results from semield (e.g., aquatic mesocosm studies) or eld studies (Campbell et al. 1990). These types of studies do indirectly take some of the factors important for population-level effects into account. However, such studies are time consuming and expensive, so it is not possible to carry out large-scale experiments for all possible scenarios. Furthermore, species composition and population characteristics might differ from those in natural ecosystems. For instance, recovery cannot always be observed in these systems because of the isolated nature of the semield experiments and the short time span of most studies. Because recovery is an item of increasing importance, there is a need for a tool that can extrapolate recovery patterns from semield experiments to the ecosystem level. Such a tool would also enable the prediction of recovery patterns of species not always present in test systems, in particular those with a limited number of life cycles per year. Such a tool can also test the ecological signicance of laboratory-based ndings, for example, population-level impact of sublethal effects such as endocrine disruption.
