GBS Data for Desmognathus cheaha and D. monticola

We sampled 51 additional specimens of Desmognathus monticola and D. cheaha from 17 sites across the putative hybrid zone in north-central Georgia. We sequenced 48 of these from 16 sites using Genotype-By-Sequencing with the ApeKI enzyme (GBS; Elshire et al., 2011) at the University of Wisconsin Bioinformatics Center. We combined this with our previous dataset of 71 individuals from across the range of both species (Pyron et al., 2023). We assembled the GBS data in ipyrad 0.9.104 (Eaton and Overcast, 2020) using ‘pairgbs’ reference-based mapping with strict filtering and default quality options. The reference is D. fuscus (aDesFus1-2.1; Myers and Pyron, 2024), Genbank accession JBBULT000000000, comprising a complete 16GB assembly. We assembled the full dataset of 119 samples (65 D. monticola and 54 D. cheaha) using these settings, retaining loci present in 71 individuals (60%), with singletons and polymorphic sites removed (Linck and Battey, 2019). This yielded an initial assembly of 62,103 SNPs from 3,251 loci totaling 674,637 bp. For applications requiring filtered bi-allelic SNPs, we screened for 60% completeness by locus for a final matrix of 12,239 bi-allelic SNPs in 2,758 loci. This compares to 7,809 SNPs from 1,138 loci totaling 106,377bp in our previous de novo assembly (Pyron et al., 2023). For analyses assuming unlinked SNPs, we randomly sampled one SNP per locus.