No abstract

Genomic approaches to the study of population demography rely on accurate SNP calling and by-proxy the site frequency spectrum (SFS). Two main questions for the design of such studies remain poorly investigated: do reduced genomic sequencing summary statistics reflect that of whole genome, and how do sequencing strategies and derived summary statistics impact demographic inferences? To address those questions, we applied the ddRAD sequencing approach to 254 individuals and whole genome resequencing approach to 35 mountain goat (Oreamnos americanus) individuals across the species range with a known demographic history. We identified SNPs with 5 different variant callers and used ANGSD to estimate the genotype likelihoods (GLs). We tested combinations of SNP filtering by linkage disequilibrium (LD), minor allele frequency (MAF) and the genomic region. We compared the resulting suite of summary statistics reflective of the SFS and quantified the relationship to demographic inferences by estimating the contemporary effective population size (Ne), isolation-by-distance and population structure, FST, and explicit modelling of the demographic history with δaδi. Filtering had a larger effect than sequencing strategy, with the former strongly influencing summary statistics. Estimates of contemporary Ne and isolation-by-distance patterns were largely robust to the choice of sequencing, pipeline, and filtering. Despite the high variance in summary statistics, whole genome and reduced representation approaches were overall similar in supporting a glacial induced vicariance and low Ne in mountain goats. We discuss why whole genome resequencing data is preferable, and reiterate support the use of GLs, in part because it limits user-determined filters.

The North American mountain goat (Oreamnos americanus) is an iconic alpine species that faces stressors from climate change, industrial development, and recreational activities. This species’ phylogenetic position within the Caprinae lineage has not been resolved and their phylogeographic history is dynamic and controversial. Genomic data could be used to address these questions and provide valuable insights to conservation and management initiatives. We sequenced short-read genomic libraries constructed from a DNA sample of a 2.5-year-old female mountain goat at 80X coverage. We improved the short-read assembly by generating Chicago library data and scaffolding using the HiRise approach. The final assembly was 2,506 Mbp in length with an N50 of 66.6 Mbp, which is within the length range and in the upper quartile for N50 published ungulate genome assemblies. Comparative analysis identified 84 gene families unique to the mountain goat. The species demographic history in terms of effective population size generally mirrored climatic trends over the past one hundred thousand years and showed a sharp decline during the last glacial maximum. This genome assembly will provide a reference basis for future population and comparative genomic analyses.

Acquisition of field data and analytical methods needed for conservation and management of wildlife populations represent significant challenges, particularly for species that inhabit landscapes that are difficult to access or species that persist in small, isolated populations. In such instances, integrating diverse and complementary data streams, such as genetic and non‐genetic data, can advance our understanding of population dynamics and associated management implications. We examined how genetic and morphologic data can be used to articulate population structure of a low‐density, peninsular population of mountain goats (Oreamnos americanus) on the Cleveland Peninsula, Alaska, USA, and surrounding areas, 2005–2018. We then use a population demographic modeling approach to examine how the use of population structure information influences sustainable harvest quotas, as compared to a panmictic, null model. Specifically, we conducted extensive field sampling of genetic (n = 446) and morphologic (i.e., horn length, n = 371) data to characterize population structure. We conducted demographic analyses and examined harvest modeling scenarios using a sex‐ and age‐specific matrix population modeling approach. Genetic and morphologic data analyses suggested peninsular subpopulations were demographically isolated, relative to surrounding mainland populations. Specifically, genetic structuring was evident and followed an isolation‐by‐distance, stepping‐stone pattern indicating limited interchange, low effective population sizes, and reduced genetic diversity along a peninsular extremity to mainland gradient. Harvest modeling indicated that overharvest would likely occur if the panmictic, null model was used to guide harvest because the smallest genetically defined population at the peninsular extremity was too small to permit any level of sustainable harvest. Our analyses illustrate the importance of using genetic and morphologic data, in combination with demographic modeling, to quantitatively delineate population boundaries and dynamics for ensuring viability of small, isolated populations. © 2020 The Wildlife Society.

No abstract