Inconsistent estimates of hybridization frequency in newts revealed by SNPs and microsatellites

Abstract Hybridization between the European smooth and palmate newts has recurrently been mentioned in the literature. The only two studies that attempted to quantify the frequency of hybridization and gene admixture between these two species came to strikingly opposite conclusions. According to Arntzen et al. (1998, 42 allozymes), hybrids are rare in nature and introgression negligible, while according to Johanet et al. (2011, 6 microsatellites), introgressive hybridization is significant and widespread across the shared distribution range. To clarify this question, we implemented high-throughput SNP genotyping with diagnostic biallelic SNPs on 965 specimens sampled across Europe. Our results are in line with Arntzen et al., since only two F1 hybrids were identified in two distinct French localities, and no further hybrid generations or backcrosses. Moreover, reanalysis of 78 of the samples previously studied by Johanet et al. (2011) using our SNPs panel could not reproduce their results, suggesting that microsatellite-based inference overestimated the hybridization frequency between these two species. Since we did not detect methodological issues with the analyses of Johanet et al., our results suggest that SNP approaches outperform microsatellite-based assessments of hybridization frequency, and that conclusions previously published on this topic with a small number of microsatellite loci should be taken with caution, and ideally be repeated with an increased genomic coverage.