Spatial patterns of genetic diversity in eight bear (Ursidae) species

Abstract: Many of the 8 extant bear species have large ranges, yet range-wide studies of genetic diversity are often impractical because of logistic challenges or focus on local questions. However, understanding the levels of diversity among populations of a species can be useful for conservation and management. Bear researchers were at the forefront of using microsatellites to study the demographics and diversity of populations, such that 3 species have complete sampling and 3 others are represented across their range breadth. Yet there has not been a synthesis of these data within or among species because of difficulties comparing microsatellites. We extracted microsatellite summary statistics from 104 papers that sampled 284 populations of any species within Ursidae, then yardstick-transformed the data for direct comparison. Studies had a median of 2 geographic sites, 30 individuals sampled per site, and 12 loci genotyped. We identified 193 loci genotyped in bears and argue this is a limitation within and among species comparisons. Tremarctos ornatus had the lowest average range-wide genetic diversity (Ar = 2.5; He = 0.43), although ascertainment bias may affect the results, whereas Ursus arctos had the highest diversity (Ar = 6.4; He = 0.69). We argue that at the spatial scale of a species' range, variation due to phylogeography and anthropogenically influenced diversity will overwhelm accuracy issues between studies and reveal broad spatial patterns. Further, by comparing allelic richness to heterozygosity across the range of a species, managers may identify populations in need of genetic management. We end by summarizing what is known about within-species lineages and genetic diversity and identify priority areas for future studies.