Comparing inferences derived from microsatellite and RADseq datasets: a case study involving threatened bull trout

Abstract

Technological advancements have allowed geneticists to exploit an increasing array of molecular markers, many of which have different properties and may provide contrasting insights into the evolutionary history and structure of populations. This has important consequences for conservation managers attempting to identify units at which to conserve intraspecific diversity. In this study we compared the inferences derived from nuclear microsatellites and restriction-site associated DNA (RADseq) data for a threatened freshwater fish, the bull trout Salvelinus confluentus. For both marker types we generated data for the same suite of individuals collected from 24 populations distributed across the species range. The RADseq data were low coverage (mean site coverage < 3X), so we implemented a probabilistic genotyping approach. We performed a comparable suite of analyses for both datasets. Both datasets revealed similar broad patterns of subdivision that reflected primary evolutionary lineages (Coastal and Interior clades). However, the RADseq more clearly and consistently identified the hierarchical phylogenetic structure. Some populations had varying assignments to these lineages depending on the dataset. RADseq data also suggested admixture has shaped the genomic character of several populations. Such a signal was not apparent with the microsatellites, suggesting that the datasets are revealing different aspects of population history. Our study provides a valuable case study in how advances in molecular technology can enhance our understanding of a relatively well-studied species. It also underscores the importance of framing findings generated with high-throughput sequencing technology within the context of past research to enhance conservation decision making.