Whole-genome sequencing reveals absence of recent gene flow and separate demographic histories for Anopheles punctulatus mosquitoes in Papua New Guinea

Abstract

Anopheles mosquitoes are the vectors of several human diseases including malaria. In many malaria endemic areas, several species of Anopheles coexist, sometimes in the form of related sibling species that are morphologically indistinguishable. Determining the size and organization of Anopheles populations, and possible ongoing gene flow among them is important for malaria control and, in particular, for monitoring the spread of insecticide resistance alleles. However, these parameters have been difficult to evaluate in most Anopheles species due to the paucity of genetic data available. Here, we assess the extent of contemporary gene flow and historical variations in population size by sequencing and de novo assembling the genomes of wild-caught mosquitoes from four species of the Anopheles punctulatus group of Papua New Guinea. Our analysis of more than 50 Mb of orthologous DNA sequences revealed no evidence of contemporary gene flow among these mosquitoes. In addition, investigation of the demography of two of the An. punctulatus species revealed distinct population histories. Overall, our analyses suggest that, despite their similarities in morphology, behaviour and ecology, contemporary sympatric populations of An. punctulatus are evolving independently.

Keywords: Anopheles; genomics; inbreeding; malaria; population genetics.

Fig. 1

Multiple-alignment statistics. The Venn diagrams summarize (A) the number of alignment blocks >500 bp generated for all species combinations, and (B) the total number of aligned base pairs represented in these blocks (in millions).

Fig. 2

Amount and age of gene flow that can be excluded given the number of shared polymorphisms observed. The figure shows the probability that introgression would lead to significantly more shared polymorphisms than observed based on the amount of gene flow (y-axis, in 4N_em) and time since the last gene flow (x-axis, in 4N_e generations). The white surface in the graph represents combinations of parameters that are incompatible with the data observed. Note that this analysis assumes that all shared polymorphisms were genuine (i.e. not sequencing errors) and therefore likely overestimates possible gene flow.

Fig. 3

Demographic histories of AF4 and AP. The figure shows the estimated historical variations in effective population size for AF4 (green) and AP (red) based on PSMC analyses. The composite green (AF4) and red (AP) lines represent 100 bootstrap replicates. The y-axis represents the population mutation rate (in log10 of θ), and the x-axis represents the log10 pairwise sequence divergence.