A BAYESIAN-APPROACH TO THE ANALYSIS OF ENVIRONMENTAL FATE AND BEHAVIOR DATA FOR PESTICIDE REGISTRATION

With the harmonisation of data requirements for pesticide registration under EC Directive 91/414 there is need for progress on the technique s used to analyse such data and so help make consistent the judgements applied by national regulatory authorities. This paper proposes a Bay esian technique for combining data from environmental fate and behavio ur studies of pesticides in soil. The method uses expert knowledge, ba sed on degradation and adsorption data, and logistic regression method s to form a prior probability distribution for the probability that a given compound leaches. Results from lysimeter experiments are used up date the prior knowledge. Data for the compounds bentazone and triclop yr are used to illustrate the techniques. The advantages of the method ology and its implications for the pesticide registration procedure ar e discussed in the light of possible advances using modern Bayesian st atistical techniques and mathematical models. (C) 1998 Society of Chem ical Industry