Using spatial genetics to quantify mosquito dispersal for control programs

BackgroundHundreds of millions of people get a mosquito-borne disease every year, of which nearly one million die. Mosquito-borne diseases are primarily controlled and mitigated through the control of mosquito vectors. Accurately quantified mosquito dispersal in a given landscape is critical for the design and optimization of the control programs, yet the field experiments that measure dispersal of mosquitoes recaptured at certain distances from the release point (mark-release-recapture MRR studies) are challenging for such small insects and often unrepresentative of the insects true field behavior. Using Singapore as a study site, we show how mosquito dispersal patterns can be characterized from the spatial analyses of genetic relatedness among individuals sampled over a short time span without interruption of their natural behaviors. Methods and FindingsWe captured ovipositing females of Aedes aegypti, a major arboviral disease vector, across floors of high-rise apartment blocks and genotyped them using thousands of genome-wide SNP markers. We developed a methodology that produces a dispersal kernel for distance that results from one generation of successful breeding (effective dispersal), using the distances separating full siblings, 2nd and 3rd degree relatives (close kin). In Singapore, the estimated dispersal distance kernel was exponential (Laplacian), giving the mean effective dispersal distance (and dispersal kernel spread {sigma}) of 45.2 m (95%CI: 39.7-51.3 m), and 10% probability of dispersal >100 m (95%CI: 92-117 m). Our genetic-based estimates matched the parametrized dispersal kernels from the previously reported MRR experiments. If few close-kin are captured, a conventional genetic isolation-by-distance analysis can be used, and we show that it can produce {sigma} estimates congruent with the close-kin method, conditioned on the accurate estimation of effective population density. We also show that genetic patch size, estimated with the spatial autocorrelation analysis, reflects the spatial extent of the dispersal kernel tail that influences e.g. predictions of critical radii of release zones and Wolbachia wave speed in mosquito replacement programs. ConclusionsWe demonstrate that spatial genetics (the newly developed close-kin analysis, and conventional IBD and spatial autocorrelation analyses) can provide a detailed and robust characterization of mosquito dispersal that can guide operational vector control decisions. With the decreasing cost of next generation sequencing, acquisition of spatial genetic data will become increasingly accessible, and given the complexities and criticisms of conventional MRR methods, but the central role of dispersal measures in vector control programs, we recommend genetic-based dispersal characterization as the more desirable means of parameterization.