My dissertation combines population genetics, landscape ecology and spatial statistics to examine connectivity and degree of gene flow in black bears (Ursus americanus). More specifically, I used the interdisciplinary approach of landscape genetics to address an array of questions, including identifying specific landscape features that impede or promote dispersal, identity source-sink dynamics, and inferring the effect of landscape change on connectivity of a large, highly harvested, isolated black bear population in the Northern Lower Peninsula (NLP) of Michigan. My dissertation consists of four chapters following the introductory chapter (Chapter 1). Chapter 2 evaluates the effects of timing and heterogeneity in collection methodology on our ability to draw inferences about the extent of gene flow in NLP black bears. This chapter shows black bears exhibit significant positive genetic spatial autocorrelation and results were concordant across seasons and years indicating the overall pattern of spatial genetic structure would be detected regardless of season, year, or collection methodology. Chapter 3 results indicate that two genetically distinct groups exist in the NLP, defined as a western and an eastern genetic cluster. In addition, I found land cover was the landscape feature most strongly correlated with genetic distance. Chapter 4 investigates the presence of source-sink dynamics in NLP black bear population. I found by asymmetric emigration and immigration, revealing black bears in the NLP is composed of source and sink areas. Further, I showed source strength was associated with black bear local harvest density (a proxy for bear density). Chapter five examines the influence of landscape change on spatial genetic structure over time. My times series analyses shows land cover was significantly correlated with gene flow; however, comparisons among sampling years revealed temporal variability in the predictive power and performance of landscape resistance models. I found the best model to explain temporal inconsistencies was land cover change over time.These data represents one of most comprehensive landscape genetic studies performed on a single black bear population. Our data enables managers to target regions or habitat types that are important for maintaining connectivity across anthropogenically-altered habitats. In addition, here I present two novel extensions of current landscape genetics analytical approaches that provide flexible frameworks for understanding connectivity and assessing impacts of future change landscape alternation which could be widely integrated into landscape genetics research and conservation planning at multiple spatial scales.
Read
