Delaying quantitative resistance to pesticides and antibiotics.

Abstract

How can we best vary the application of pesticides and antibiotics to delay resistance evolution? Previous theoretical comparisons of deployment strategies have focused on qualitative resistance traits and have mostly assumed that resistance alleles are already present in a population. But many real resistance traits are quantitative, and the evolution of resistant genotypes in the field may depend on de novo mutation and recombination. Here, I use an individual-based, forward-time, quantitative-genetic simulation model to investigate the evolution of quantitative resistance. I evaluate the performance of four application strategies for delaying resistance evolution, to wit, the (1) sequential, (2) mosaic, (3) periodic, and (4) combined strategies. I find that which strategy is best depends on initial efficacy. When at the onset, xenobiotics completely prevent reproduction in treated demes, a combined strategy is best. On the other hand, when populations are partially resistant, the combined strategy is inferior to mosaic and periodic strategies, especially when resistance alleles are antagonistically pleiotropic. Thus, the optimal application strategy for managing against the rise of quantitative resistance depends on pleiotropy and whether or not partial resistance is already present in a population. This result appears robust to variation in pest reproductive mode and migration rate, direct fitness costs for resistant phenotypes, and the extent of refugial habitats.