A Comparison of Two Species‐Discriminatory SNP Sets for Taxonomic Assignment in Hybridizing European White Oaks ( Quercus sect. Q

ABSTRACT

Species identification is essential for conservation, forest management, and understanding ecological and evolutionary dynamics, such as adaptive introgression and maintenance of species boundaries in hybridizing taxa. Species‐discriminatory marker sets have been shown to be more accurate than morphological identification, especially in hybridizing species complexes, and provide a more affordable alternative to genome‐wide analyses. However, to our knowledge, no species complex exists for which several marker sets are available, let alone have been compared to each other. Moreover, it remains unclear how well marker sets can be applied to additional, closely related species. Here, we compared two independently developed SNP sets, one derived from amplicon sequencing and one from whole‐genome sequencing, for three morphologically similar and hybridizing European white oak species sensu lato ( Quercus robur , Q. petraea , and Q. pubescens ). We applied the SNP sets to 560 samples, including samples of known pure ancestry, controlled crosses, natural populations, and four additional white oak species. Congruency between SNP sets was very high, but species assignment was slightly more accurate for the reference samples used in the development of the respective SNP set. Ancestry coefficients for first‐generation hybrids were more precise with the SNP set based on whole‐genome data. In a natural population, there was limited congruency between SNP sets in quantifying admixture. Notably, the SNP sets were able to identify up to two additional white oak species. Our study highlights the great potential of species‐discriminatory SNP sets in science and practice, while demonstrating that ascertainment bias should not be neglected, and accurate introgression assessments may require genome‐wide analyses.