Local adaptation to biocontrol agents: A multi-objective data-driven optimization model for the evolution of resistance

Spatial and temporal variability in the application of biological control agents such as parasites or pathogenic bacteria can cause the evolution of resistance in pest organisms. Because biocontrol will be more effective if organisms are not resistant, it is desirable to examine the evolution of resistance under different application strategies.We present a computational method that integrates a genetic algorithm with experimental data for predicting when local populations are likely to evolve resistance to biocontrol pathogens. The model incorporates parameters that can be varied as part of pest control measures such as the distribution and severity of the biocontrol agent (e.g., pathogenic fungi). The model predicts the evolution of pathogen defense as well as indirect selection on several aspects of the organism's genetic system. Our results show that both variability of selection within populations as well as mean differences among populations are important in the evolution of defenses against biocontrol pathogens. The mean defense is changed through the pest organism's genotype and the variance is affected by components of the genetic system, namely, the resiliency, recombination rate and number of genes.The data-driven model incorporates experimental data on pathogen susceptibility and the cost of defense. The results suggest that spatial variability rather than uniform application of biological control will limit the evolution of resistance in pest organisms.