Year of Award


Document Type

Thesis - Campus Access Only

Degree Type

Master of Science (MS)

Degree Name

Organismal Biology and Ecology

Department or School/College

Division of Biological Sciences

Committee Co-chair

Gordon Luikart, Fred W. Allendorf

Commitee Members

Michael Mitchell, Richard D. Mace


connectivity, gene flow, habitat fragmentation, landscape genetics, landscape resistance modeling, Ursus americanus


University of Montana


Habitat loss and climate change will likely reduce connectivity among many black bear populations across their distribution. I used individual-based landscape genetics analyses in 12 study areas in western Montana and northern Idaho to assess the influence of landscape features on gene flow among American black bears (Ursus americanus). I compared the influence of landscape features among the study areas, tested the effect of reducing the number loci on support for landscape genetic models, and examined possible explanations of why different landscape features were supported in the different study areas. To determine the influence of landscape features on gene flow, I examined the relationship between pair-wise genetic distances and ecological (cost) distances between individuals in each study area by applying a similar landscape genetic modeling approach as Cushman et al. (2006). Tests of isolation by distance (IBD) after removing landscape effects were non-significant in nearly all study areas. Five study areas had significant landscape genetic models (p <0.05, partial Mantel tests). These results suggest landscape features were a better predictor of gene flow than IBD, which re-affirms previous findings. When testing the effect of the reduction of loci, I found that study areas with significant effects of landscape features (p < 0.05) revealed in the same significant landscape features using subsets of fewer loci than in the original full data set. However, study areas with less significant effects of landscape features (p > 0.04) resulted in different significant landscape features using subsets of loci. The results suggest that researchers should subsample loci and use relatively low p-values (p < 0.04) to identify landscape features contributing to genetic structure. I investigated causes of the variation of significant landscape features among study areas by conducting a ‘limiting factor’ analysis. Landscape features were identified as influencing gene flow when the feature was variable (elevation) or non-continuous (forest). These results suggest critical landscape features will present strong relationships with genetic differentiation only when the feature is variable and thus potentially limiting to gene flow. Researchers need to consider studying multiple landscape areas to avoid erroneous conclusions about which landscape features generally limit gene flow before extrapolating a landscape resistance model from one area to another or to broader regions. These results also begin to identify thresholds of variability of elevation and forest cover that have detectable influence on gene flow.

This record is only available
to users affiliated with
the University of Montana.

Request Access



© Copyright 2010 Ruth Short Bull