Gene flow as a factor in the evolution of insecticide resistance

Abstract

The importance of gene flow in maintaining genetic variation among local populations in heterogeneous habitats has been debated for natural populations as well as populations subject to selection in manipulated ecosystems such as agricultural fields. The diamondback moth, Plutella xylostella (L.), is a pest of cruciferous crops and frequently treated with insecticides. Despite the high dispersal capabilities of this moth, considerable local variation exists in its response to insecticides, suggesting that gene flow among subpopulations may be restricted. The central question addressed by this study is whether this local variation is due to restricted gene flow, or if variation in selection due to different pesticide strategies is sufficient to maintain the variation. Electrophoretic measurements of allozyme variation among 11 subpopulations on three islands in Hawaii indicate small amounts of genetic differentiation between subpopulations (overall FST estimates ranged from 0.038 to 0.028). This suggests that roughly 8 migrants are exchanged per subpopulation per generation. Smaller amounts of differentiation were observed between islands and between Hawaii and two continental populations, perhaps because insufficient time has passed for these larger population subunits to differentiate via drift. Simulation of resistance evolution among treated and untreated fields in the diamondback moth suggests that migration rates greater than 10% are required before emigration from untreated into treated fields will delay resistance evolution. Low rates of gene flow (<0.1%) delayed the overall rate of resistance evolution when resistance alleles were rare initially. The fastest rates of resistance evolution occurred at intermediate gene flow rates. The effect of gene flow on the establishment of a pesticide resistant parasitoid, Trioxys pallidus, is examined in Chapter 3. Increased migration, either by dividing the number of resistant parasitoids released into smaller groups, or by increasing the migration rate, decreased the time required to implement the new strain. The simulations indicate that at least three years will be required for establishment of the resistant strain. The results demonstrate that implementation of genetically altered insects may require a significant proportion of the time allocated to a genetic manipulation project. Studies to identify potential problems in the implementation phase should be part of the initial consideration of any such project.