Contrasting whole-genome and reduced representation sequencing pipelines for population demographic inference - high coverage genomes

Genome sequence captures the adaptative history and demographic story of a species. Choice of sequencing strategy and sample size, however, can impact such inferences. We evaluated a whole genome and a reduced representation sequencing approaches to a study of genetic diversity and population demography of an alpine ungulate, the North American mountain goat (Oreamnos americanus). We applied the ddRAD sequencing approach to 254 individuals and whole genome resequencing (WGS) approach to 35 individuals across the species range with a known demographic history. We used ANGSD to estimate the genotype likelihoods (GLs) and estimated the contemporary effective population size (Ne), population structure, and explicitly modelled the demographic history with DADI and MSMC2. Whole genome and reduced representation approaches were overall similar in supporting a glacial induced vicariance and low Ne in mountain goats; DADI parameters were very similar between sequencing strategies. We also evaluated a set of climatic variables and geographic location as predictors of genetic diversity using redundancy analysis. A proportion of total variance (~18% for WGS and ~5% for RADseq data set) was explained by geography indicating a paramount role of genetic drift, though climate explained a portion in both data sets suggesting some degree of local adaptation. The empirical similarities of WGS and RADseq presented herein reassuringly suggest that both approaches will recover large demographic and adaptive signals in a population. Considering the predicted large climate-induced changes in alpine environments and the genetically depauperate mountain goat, the long-term adaptive capabilities of this enigmatic species are questionable.