Dataset characteristics
We sequenced 16.3 to 50.3 Gbp (mean = 33.7 Gbp) per individual, with 14.8 to 44.5 Gbp (mean = 30.1 Gbp) passing quality filtering (Table S1). Because we used a relatively divergent reference genome—not in the same genus as any of the study species—parts of the genome had no alignment coverage (range = 13.3 to 23.4% unaligned reference genome per individual) (Table S1). Overall, the data quality and quantity resulted in 6.4× to 23.9× genomic coverage per individual (mean = 14.5×) (Fig. S1), and after filtering low coverage and low-quality genomic regions, all individuals were genotyped in ~400 to 600 million sites each (Table S1).