Biochemical preadaptations, founder events, and the evolution of resistance in arthropods

We conducted literature-based comparative analyses of pesticide resistance evolution to evaluate 2 aspects of microevolutionary response to strong novel selection pressures. First, it has been proposed that herbivorous arthropods are preadapted to evolve resistance by a system of detoxifying enzymes the evolution of which was elaborated in response to plant defensive chemicals. To test this hypothesis, we note that arthropods that feed on plant vascular tissues (phloem and xylem, which are less strongly chemically defended than other tissues) have lower levels of detoxifying enzyme activities than species that chew or feed on cell contents, and might therefore be expected to be less strongly preadapted to pesticides. The record of pesticide resistance evolution in 615 arthropod pests of North American agriculture is consistent with he biochemical preadaptation hypothesis; phloem-and xylem-feeding species have a significantly diminished ability to evolve resistance when compared with either chewing or cell content-feeding species. The same result is obtained in an analysis restricted to species within the order Homoptera. We caution, however, that feeding behavior is confounded with taxonomy in these analyses; thus, strong inferences of a causal relationship between feeding mode and resistance evolution would be premature. Second, conventional theory suggests that population bottlenecks that occur during the initiation of geographically isolated populations can have profound effects on the genetic structure of populations, including particularly the loss of allelic diversity. Arthropods that have been introduced to North America, and thus have undergone events, might therefore be expected to be depanperate for key resistance-conferring genetic variants. Our analysis find no support for this founder event hypothesis; native and introduced species do not differ significantly in resistance evolution. We evaluate the importance of possibly confounding variables in the interpretation of these results and emphasize the need to integrate experimental results with these comparative broad-scale results.