Identifying individuals to be genotyped

The mapping approach used here genotypes only those individuals with extreme phenotypes (cf. Bastide et al. 2013). As our goal is to map nucleotides associated with desiccation resistance inside In(2La) and In(2Rb), we also need to control for other general factors that may contribute to survival. It is well known from lab colonies that both sex and weight are important factors in desiccation resistance (Fouet et al. 2012), and here we were able to demonstrate these same effects among our wild-caught mosquitoes (Supporting information Figures S1, S2). Because of the large difference in survival between sexes, we chose to conduct mapping using only females; the larger body size of females also ensures enough DNA for genotyping. Survivorship curves of teneral An. gambiae females stratified by In(2La) karyotype are shown in Figure 1. In order to identify individuals with the most extreme resistance phenotypes due only to the karyotype of the 2La inversion they carry, we used the Cox proportional hazard model (Cox 1972) as implemented using the package “coxme” in R (www.r-project.org). We modeled survival using the covariates karyotype, weight (log-transformed), maximum temperature on the day of the experiment, day of experiment, breeding site, and vapor pressure deficit on the day of the experiment; model selection was carried out with the Akaike information criterion (AIC) using the function “aictab” from the “AICcmodavg” package in R. With respect to individuals homozygous for the standard arrangement of In(2La) (“STD”), we identified 50 with the highest survival and 50 with the lowest survival using the residual values from the full model. We similarly identified extreme individuals homozygous for the inverted arrangement (“INV”), 50 with the highest survival and 50 with the lowest survival using residuals.

Survival of teneral adult female An. gambiae under acute desiccation stress. Data are stratified by In(2La) karyotype. Dashed lines represent 50% survivorship.