PESTonFARM – stochastic model of on‐farm insect behaviour and their response to IPM interventions

Abstract

AbstractPESTonFARM is an agent‐based pattern‐oriented model with elements of cellular automata, developed to simulate insect behaviour within seasonally changing mosaic of farming landscapes, under the challenge of Integrated Pest Management (IPM) actions. The model is based on a software construct (‘virtual insect’), acting according to a pre‐defined set of behavioural rules. Each ‘virtual insect’ acts independently, developing its individually unique behavioural sequence and life history. The local on‐farm conditions and the planned IPM interventions can be represented through user‐defined spatial and temporal parameters. Incompatibility among IPM treatments, for example, pesticide vs. biological control, also could be taken into account. Unit costs of IPM interventions and crop values can be specified to obtain cost/benefit assessments for each IPM scenario. The model emulates the behaviour of a pest population during a virtual IPM experiment, and each run generates stochastically equivalent but unique sets of results, which functionally correspond to a replication of an on‐farm experiment. By taking into account the local farm structures and their spatiotemporal variations the model alludes to the concept of ‘precision agriculture’. Once customized to the local conditions and validated, the model can support site‐specific forecasting, decision‐making and training, serving as a component of the ‘precision IPM’ toolbox. To demonstrate the concept, an example of the model application in management of the cherry fruit fly, Rhagoletis cherasi, is briefly presented and discussed.