PesTwin: A modular agent-based framework for pest and vector population control

Abstract

Species-specific pest and vector control strategies, including the sterile insect technique, Wolbachia-based interventions, and genetic control technologies, offer powerful alternatives to broad-spectrum chemical control, with applications ranging from targeted crop protection to large-scale disease control. Among these, genetic control technologies are advancing rapidly, but the pace of technological development is outstripping the modelling tools needed to predict outcomes, guide technology design and its implementation, compare alternative strategies across different use settings, and support regulatory and operational decision-making. Here we present PesTwin, an agent-based modelling framework for simulating genetic control technologies across species, ecological settings, and deployment strategies within a common computational environment. PesTwin captures stochastic demographic effects, species-specific life-history traits, heterogeneous dispersal, and temporal variation in resource availability and infestation pressure. We validate PesTwin against published laboratory cage data from four genetic control systems, drawn from three studies, in two insect species, showing close agreement between predicted and observed population trajectories, including their replicate-to-replicate variability. We then illustrate how the same validated models extend beyond the cage to spatially explicit, field-scale scenarios, using PesTwin to explore how the timing, density and spatial placement of releases shape suppression and spread across heterogeneous landscapes. By making genetic control systems testable in silico before they are built or released, PesTwin can shorten the path from laboratory construct to field intervention: informing which constructs to prioritise, how to design the experiments that test them, where and when to release, and what evidence is needed to evaluate them.