Gene flow networks among American Aedes aegypti populations

Abstract

The mosquito Aedes aegypti, the dengue virus vector, has spread throughout the tropics in historical times. While this suggests man-mediated dispersal, estimating contemporary connectivity among populations has remained elusive. Here, we use a large mtDNA dataset and a Bayesian coalescent framework to test a set of hypotheses about gene flow among American Ae. aegypti populations. We assessed gene flow patterns at the continental and subregional (Amazon basin) scales. For the Americas, our data favor a stepping-stone model in which gene flow is higher among adjacent populations but in which, at the same time, North American and southeastern Brazilian populations are directly connected, likely via sea trade. Within Amazonia, the model with highest support suggests extensive gene flow among major cities; Manaus, located at the center of the subregional transport network, emerges as a potentially important connecting hub. Our results suggest substantial connectivity across Ae. aegypti populations in the Americas. As long-distance active dispersal has not been observed in this species, our data support man-mediated dispersal as a major determinant of the genetic structure of American Ae. aegypti populations. The inferred topology of interpopulation connectivity can inform network models of Ae. aegypti and dengue spread.