Forecasting impacts of biological control under future climates: mechanistic modelling of an aphid pest and a parasitic wasp.

• We develop a population dynamics model for a global aphid pest and parasitic wasp. • The model is calibrated and validated with field data from south-eastern Australia. • Model performance was improved by the inclusion of rainfall-induced mortality. • Biological control from the wasp is most effective under abiotic stress. • Results can directly inform pest management strategies under future climates. Climate change impacts agricultural pests in complicated ways, and accounting for these responses should form an integral part of pest management programmes. Process-based models are often used in agriculture to forecast population dynamics of pests within a growing season, however these are often constrained to assessing the impacts of temperature in isolation of other factors. These models are therefore unable to fully explore species' responses to climate change, which may be driven by multiple abiotic and biological stressors. Here, we build a mechanistic model of a globally distributed agricultural aphid pest (Myzus persicae (Sulzer)), and an associated parasitic wasp (Diaeretiella rapae (M'Intosh)) that is known to biologically control the aphid. We simulate temporal dynamics of the crop with a well-established canola growing degree-day model (APSIM) and incorporate the impacts of temperature and rainfall on insect survival. The model was parameterised with laboratory-measured datasets from around the globe, and we have calibrated and validated the model to Australian broadacre cropping systems using regional observation records. We then ran the validated models with future temperature and rainfall scenarios to reveal that suppression of aphid populations by the wasp is enhanced under stressful abiotic conditions, which are predicted to occur more frequently in the future. The process-based modelling approach affords valuable and novel insights into physiological traits that influence population dynamics of both species and highlights gaps in our current understanding of the system. In the future, farmers could have greater confidence in the bio-control potential of D. rapae under different conditions, and be in a position to adjust their M. persicae management programmes accordingly. This is the first model to explore the interaction of these two cosmopolitan species in the field, which is applicable across broad geographic regions, while also providing insights as to how both species could be better managed on a local scale. [ABSTRACT FROM AUTHOR]