Statistical Treatment of Ecotoxicological Data with Continuous Response using biologically based models

When characterising complex chemical pollution i.e. pesticide polluted groundwater or testing the effects of chemical compounds on the environment, ecotoxicological tests are of great value. This is due to their sensitivity and their expression of the overall toxicity. Since environmental pollution is increasing it is judged that ecotoxicological assays will have increasing impact on the toxicity- and risk assessment of environmental pollution.
Evaluating results from ecotoxicological tests inevitably include estimation of dose-response relations. Conclusions about toxicity of the pollution or tested chemicals are summarised in characteristics extracted from the dose-response relationship.
A revision of the statistical methods for extracting these measures has been internationally recommended by the scientific community. Generalized Nonlinear regression methods using classical dose-response models, are used to estimate the variance heterogeneous dosis-response relations and estimate endpoints and their confidence limits.