Perpest model, a case-based reasoning approach to predict ecological risks of pesticides



The PERPEST model is a model that predicts the ecological risks of pesticides in freshwater ecosystems. This model simultaneously predicts the effects of a particular concentration of a pesticide on various (community) endpoints. In contrast to most effect models, PERPEST is based on empirical data extracted from the literature. This model is based on case-based reasoning, a technique that solves new problems (e.g., what is the effect of pesticide A?) by using past experience (e.g., published microcosm experiments). The database containing the past experience has been constructed by performing a review of freshwater model ecosystem studies. This review assessed the effects on various endpoints (e.g., community metabolism, phytoplankton, and macroinvertebrates) and classified them according to their magnitude and duration. The PERPEST model searches for analogous situations in the database, based on relevant (toxicity) characteristics of the compound. This allows the model to predict effects of pesticides for which no effects on a semifield scale have been published. The PERPEST model results in a prediction showing the probability of classes of effects (no, slight, or clear effects, plus an optional indication of recovery) on the various grouped endpoints. This paper discusses the scientific background of the model as well as its strengths, limitations, and possible applications.