Supplementary Figure 1: Diploid genotype quality control using a haploid drone sequenced to 20× depth of coverage.

Haploid drone data were intentionally misspecified to be diploid in the FreeBayes SNP calling process, and variable positions shared by the drone and the population data set were compared. This figure shows the distributions of heterozygous genotypes in the population data set conditional on their genotype in the resequenced drone (blue, all population data set SNPs; pink, SNPs where the drone had a homozygous difference compared to the reference genome; gray, SNPs where the drone had a heterozygous genotype). Chromosomal positions with heterozygous drone genotypes demonstrated extremely elevated levels of heterozygous genotypes in the population data set (gray) compared to the L-shaped distribution of all population SNPs (blue) and the positions where the drone had homozygous differences relative to the reference sequence (red). SNPs overlapping the heterozygous drone positions were therefore removed from the population data set, as they are unlikely to represent true SNPs.