Effects of sampling scale on American black bear spatial genetic structure

Sampling bias can lead to erroneous interpretations of spatial genetic structure that can subsequently impact conservation efforts and management decisions. Genetic sampling of rare and elusive species can be challenging and, by necessity, samples are often collected opportunistically from multiple sources that could differ in spatial dispersion and timing of collection. Here we quantified the effects of timing of sample collection and sampling methods on spatial and temporal variability in sample dispersion and measures of American black bear (Ursus americantts) local spatial genetic structure. Hair (N = 890) and tissue (N = 1,017) samples were collected from Michigan's Northern Lower Peninsula (USA) during the summer using non-invasive (hair snares) and autumn harvest methods during 2003, 2005, and 2009. Point pattern analyses of sample dispersion revealed that sample density did not differ significantly between seasons or among years. Measures of spatial genetic structure (i.e., spatial autocorrelation) revealed significant positive genetic spatial structuring at distances of 0-10 km when samples were analyzed separately by year, season, and when temporally distinct samples were grouped. Local genetic spatial autocorrelation analyses revealed spatial genetic patterns over the entire study area were consistent across seasons and years. Spatial genetic structure is indicative of the extent of dispersal and gene flow, which is crucial for developing management plans for harvested species. Collectively, our data show that non-invasive and harvest collection methods similarly capture spatial genetic heterogeneity at the scale and spatial extent appropriate for management.