Large carnivores are considered at risk of local extirpation in areas of increased urbanization and human development largely due to habitat reduction and fragmentation. Pumas (Puma concolor; also called mountain lions and cougars) in southern California live in a highly urbanized landscape fragmented by interstate highways, and they face mortality from road strikes and other conflicts with humans. Although these populations have persisted through increasing urbanization thus far, populations isolated in coastal mountain ranges are predicted to be at serious risk of extinction. We used double-digest Restriction Site Associated DNA sequencing (ddRAD) to determine the landscape features associated with gene flow, compare putatively neutral and potentially adaptive genomic variation, and compare genomic approaches to previous work with microsatellites for southern California mountain lions. We conducted landscape genomic analyses on a regional scale (across four populations) as well as within-populations analyses to evaluate how landscape features that influence gene flow vary across spatial scales and among neighbouring populations.
Results/Conclusions
We found genomic structure that was largely consistent with previous microsatellite analyses, reflecting distinct populations within separate mountain ranges separated by urban valleys. Landscape genomic analyses reveal that on a regional scale gene flow is negatively associated with urbanization. Within populations, however, landscape features associated with gene flow varied, although gene flow was primarily associated with natural landscapes. Within more urbanized populations (e.g., the Santa Ana mountains), riparian corridors were found to be particularly important for mountain lion gene flow, suggesting these landscape features as likely important habitat for the resiliency of large carnivores in increasingly urbanized ecosystems. Importantly, our results reveal that landscape genomic models explain a greater proportion of gene flow within more isolated populations than in populations with greater connectivity. These findings reiterate the importance of spatial scale in landscape genomics and reveal that landscape features associated with gene flow may differ between populations even when in close proximity to each other.