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

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