No evidence for biased co-transmission of speciation islands in Anopheles gambiae

Genome-scale scans have revealed highly heterogeneous levels of divergence between closely related taxa in many systems. Generally, a small number of regions show high differentiation, with the rest of the genome showing no or only low levels of divergence. These patterns have been interpreted as evidence for ongoing speciation-with-gene-flow, with introgression homogenizing the whole genome except loci involved in reproductive isolation. However, as the number of selected loci increases, the probability of introgression at unselected loci decreases unless there is a transmission ratio distortion causing an over-representation of specific combinations of alleles. Here we examine the transmission of three 'speciation islands' that contain fixed differences between the M and S forms of the mosquito, Anopheles gambiae. We made reciprocal crosses between M and S parents and genotyped over 2000 F-2 individuals, developing a hierarchical likelihood model to identify specific genotypes that are under- or over-represented among the recombinant offspring. Though our overall results did not match the expected number of F-2 genotypes, we found no biased co-transmission among M or S alleles in the three islands. Our likelihood model did identify transmission ratio distortion at two of the three islands, but this distortion was small (approx. 3%) and in opposite directions for the two islands. We discuss how our results impinge on hypotheses of current gene flow between M and S and ongoing speciation-with-gene-flow in this system.