Landscape genetics and the biological control of walnut aphids in California
The introduction of the parasitoid wasp Trioxys pallidus (Hymenoptera: Braconidae) from Iran led to the dramatically successful biological control of walnut aphids in California. Recently, increases in numbers of aphids in mid-late summer has required the use of insecticide treatments, and the factors responsible for the breakdown in control remain unknown. Classical biological control combines ecological and evolutionary theory to provide permanent and sustainable control of pest species. Landscape genetics also represents the synergistic combination of ecological and evolutionary studies by examining the effects of landscape on population structure and connectivity. While landscape patterns and processes have been examined in regards to patterns of diversity and abundance of pests and natural enemies in agricultural systems, little is known about how landscape genetics affects biological control. Therefore, we used 15 microsatellite markers to genotype 123 individuals of T. pallidus from walnut orchards in California. We then examined the population structure of these populations using two clustering approaches. The first was agnostic to the geographic location of individuals, and the second incorporated the geographic location of individuals into the analysis. Finally, we used a Bayesian framework to measure the connectivity of populations.
We found evidence that individuals of T. pallidus were fragmented into as many as 7 genetically distinct and geographically isolated populations in California with significant differentiation between some population pairs based on Fst values, and that genetic distance between populations was not correlated with geographic distance. We also found that levels of geneflow differed significantly between populations. Our results suggest that patterns at the landscape level are shaping the genetic structure of populations of T. pallidus in this important crop system, and that management for the sustainability of classical biological control systems may also require accounting for landscape level patterns and processes.