ABSTRACT
While many ecological models addressing pesticide risk assessment can be found in the academic literature, they are rarely applied in pesticide registrations. In order to determine issues with existing models as published in literature, we assess the state of the art of ecological models addressing pesticide risk assessments. We focus on how toxic effects were modeled, how risk was quantied, and on the verication and validation of the models. From an extensive literature scan, 41 models were selected for review. Three main classes of models are used for pesticide risk assessment: difference and differential equations (14 out of 41 models), matrix models (17), and individual-based models (IBMs; 8). Additionally, 2 grid-based simulations were included. In the majority of models, toxicity was implemented as increased mortality, and risk was quantied by the effects on population dynamics. In some publications, both measures remained unclear. Apart from a subset of the IBMs, no validation was performed on the models. Another problem could be identied in the often ambiguous description of the models, and a lack of justication of the choice of model type. From this review of models as published in scientic literature, we conclude that most models do not address key issues of pesticide risk assessment sufciently. We provide recommendations for the future use of ecological models in pesticide risk assessment.
