Landscape genetic analysis of population structure and barriers to gene flow in boreal woodland caribou (Rangifer tarandus caribou)

Abstract

This study examines patterns of population genetic structure and gene flow of boreal woodland caribou (Rangifer tarandus caribou), which are experiencing declining population sizes across North America. Compared to previous studies, I used fine-scale landscape genetic analyses with intensive sampling to identify genetic subdivisions within a single range and anthropogenic and natural drivers of genetic discontinuity. The Brightsand Range of Ontario is among the southernmost boreal woodland caribou populations and contains actively managed and unmanaged forests. This range provided a unique opportunity to examine the drivers of population subdivision using fecal DNA samples (n = 788) previously obtained from non-invasive surveys. I used 12 microsatellite markers to investigate genetic diversity, identify patterns of genetic structure, and delineate barriers to gene flow. I found high connectivity among most sites, with low but significant population genetic substructure (Fst=0.009, p<0.001). The Mantel test identified a weak pattern of isolation by distance, and genetic clustering algorithms failed to identify a biologically meaningful pattern of population substructure. MEMGENE analysis and multiple regression analysis based on univariate resistances in CIRCUITSCAPE indicated that wildfires acted as a barrier to gene flow, with sites separated by burned areas having higher genetic differentiation than expected due to isolation by distance alone. The POPGRAPH analysis identified genetically isolated sites among the managed portion of the range, and CIRCUITSCAPE analysis showed that the range is highly fragmented within the managed portion and contains limited connectivity corridors, whereas the unmanaged portion had high connectivity throughout. Overall, this study suggests that boreal woodland caribou are weakly genetically differentiated across the Brightsand Range, with isolation by distance and isolation by resistance contributing to variation in allele frequencies. However, while genetic differentiation was weak, conservation efforts will be required within the managed forest area to reduce the loss of genetic diversity by improving landscape connectivity.