Current issues in statistics and models for ecotoxicological risk assessment.

A review is given on statistical and modelling issues in ecotoxicology. The issues discussed are: 1. How to estimate an (almost) no effect concentration chemicals in the laboratory. 2. Combining single-species acceptable effect levels to an acceptable effect level for a multi-species ecosystem. 3. The combined effect of exposure to several chemicals. 4. Bioavailability in the natural environment and food-web models. Most current procedures in setting standards allow the environmental concentration to be above the acceptable effect concentration for a small fraction of the species. It is shown that a considerable part of the fraction of the affected species will suffer a severe effect.