Urban development is an increasing cause of habitat fragmentation and modification and can have a variety of impacts on wildlife. These include losses in functional connectivity that can isolate populations, leading to reductions in genetic diversity and increased potential for inbreeding depression. In southern California, extensive urbanisation over a region that contains many endemic species has produced a hotspot for extinction risk. Bobcats (Lynx rufus) in southern California persist in relatively small habitat patches and can utilise modified habitat. However, bobcats remain sensitive to habitat fragmentation, and require habitat of sufficient quality to support prey populations. We implemented a landscape genomic approach using 13,520 SNP loci to characterise broad- and fine-scale genomic variation among 281 bobcats sampled from around Los Angeles and San Diego and to assess drivers of functional connectivity. Population genomic structuring indicated five distinct populations, which we treated as replicates in our landscape genomic design to assess the generality of our findings. We used population-specific landscape resistance analyses, as well as a region-wide resistance analysis, to test the influence of anthropogenic and natural landscape features on bobcat gene flow at fine and broad spatial scales.
Results/Conclusions
Population genomic analyses indicated that our five genetic populations were separated by major highways that were fringed by urban development. This was reflected in our region-wide landscape genomic analysis, which found a negative association between urban development and gene flow. Population genomic structuring was largely consistent with previous microsatellite studies, but we detected a higher proportion of migrants overall, and more admixture in some areas. Interestingly, no population structuring was detected around San Diego, with bobcats from this region forming a single genetic cluster with individuals from Chino Hills State Park in Los Angeles. This population is divided by several major highways, but these roads are not accompanied by urban development as extensively as in other populations. Fine-scale landscape genomic analyses indicated a positive association of riparian corridors with gene flow within the three least urbanised populations. Only within the smallest and most urbanised population was urban landcover negatively associated with gene flow, while topographic roughness and vegetation were positively associated with gene flow within both highly urbanised and less urbanised populations. This study provides insights into functional landscape connectivity in a region of high conservation concern, and highlights how landscape variables driving this connectivity can vary among different populations according to patch size and urban landscape configurations.