Print PageWednesday, August 10, 2011: 8:40 AM
12B, Austin Convention Center
Background/Question/Methods
The emerging interdisciplinary field of landscape genetics investigates the effect of biotic and abiotic landscape features on the movement and dispersal of organisms by quantifying the influence of these features on gene flow. We asked if changes in landscape pattern and connectivity resulting from recent habitat loss and fragmentation have influenced genetic structure in a threatened woodland caribou population (Rangifer tarandus caribou) in central Canada. To examine if contemporary spatial genetic relationships are better explained by current or historical landscape pattern, we used vegetation layers representing the landscape in 2006 and in 1960, immediately prior to the start of rapid landscape change caused by forest harvesting, road construction and wildfire. Because pattern occurring at more than one spatial scale may influence genetic structure, we developed a method based on landscape graph theory to capture landscape pattern at multiple spatial scales. In total, we built 7500 hypotheses describing how landscape may resist gene flow, and tested the fit of each to genetic distances among 95 caribou, estimated using markers at 10 microsatellite loci. Tissue for genetic analysis was obtained from the outer mucosal layer of fecal pellets, collected by systematic survey during the winters of 2007 and 2008.
Results/Conclusions
We found evidence of landscape resistance to gene flow in 37 hypotheses describing the contemporary landscape, and no evidence in hypotheses describing the historical landscape, when we controlled for a null model representing isolation by geographic distance (paired partial Mantel tests, alpha level = 0.05, 105 permutations). Among the 37 significant hypotheses, we identified 4 clusters of similar hypotheses in a hierarchical clustering. Hypotheses from three of these clusters were simultaneously influencing contemporary genetic structure. When interpreted in terms of landscape features, the hypotheses represented three spatial scales and together emphasized the costs to caribou movement and dispersal associated with major forestry roads, a provincial highway, and forest harvest cut blocks in combination with the natural constraints imposed by large lakes on the landscape. Our results provide evidence of the early impacts of habitat loss and fragmentation on the genetic structure of a population, and demonstrate the importance of multiple scales of analysis in landscape genetics.
The emerging interdisciplinary field of landscape genetics investigates the effect of biotic and abiotic landscape features on the movement and dispersal of organisms by quantifying the influence of these features on gene flow. We asked if changes in landscape pattern and connectivity resulting from recent habitat loss and fragmentation have influenced genetic structure in a threatened woodland caribou population (Rangifer tarandus caribou) in central Canada. To examine if contemporary spatial genetic relationships are better explained by current or historical landscape pattern, we used vegetation layers representing the landscape in 2006 and in 1960, immediately prior to the start of rapid landscape change caused by forest harvesting, road construction and wildfire. Because pattern occurring at more than one spatial scale may influence genetic structure, we developed a method based on landscape graph theory to capture landscape pattern at multiple spatial scales. In total, we built 7500 hypotheses describing how landscape may resist gene flow, and tested the fit of each to genetic distances among 95 caribou, estimated using markers at 10 microsatellite loci. Tissue for genetic analysis was obtained from the outer mucosal layer of fecal pellets, collected by systematic survey during the winters of 2007 and 2008.
Results/Conclusions
We found evidence of landscape resistance to gene flow in 37 hypotheses describing the contemporary landscape, and no evidence in hypotheses describing the historical landscape, when we controlled for a null model representing isolation by geographic distance (paired partial Mantel tests, alpha level = 0.05, 105 permutations). Among the 37 significant hypotheses, we identified 4 clusters of similar hypotheses in a hierarchical clustering. Hypotheses from three of these clusters were simultaneously influencing contemporary genetic structure. When interpreted in terms of landscape features, the hypotheses represented three spatial scales and together emphasized the costs to caribou movement and dispersal associated with major forestry roads, a provincial highway, and forest harvest cut blocks in combination with the natural constraints imposed by large lakes on the landscape. Our results provide evidence of the early impacts of habitat loss and fragmentation on the genetic structure of a population, and demonstrate the importance of multiple scales of analysis in landscape genetics.
