Linking landscape structure to woodland caribou density using non-invasive spatial capture-recapture

Abstract

Assessing how landscape composition and anthropogenic disturbance influence spatial variation in abundance is essential for effective conservation planning in threatened species. For woodland caribou (Rangifer tarandus caribou), there remains limited knowledge of spatially explicit population density, potential differences in density and space use between males and females, and the specific habitat and disturbance features that structure their distribution across managed forest landscapes. We applied spatial capture-recapture (SCR) modeling to non-invasive fecal DNA samples collected in the Churchill Range of northwestern Ontario to estimate caribou density, space-use parameters, and evaluate how landscape composition and disturbance influence spatial variation in abundance. A total of 589 genotyped samples identified 245 unique individuals (132 females, 113 males). The best-supported SCR model estimated a population density of 43.6 individuals per 1,000 km² (95% CI: 31.8 – 59.6) and a mean space-use parameter (σ) of 5.4 km, corresponding to an estimated home range radius of 25.6 km. Sex-specific models revealed no strong differences in density or space use between males and females. Landscape models indicated that caribou density was positively associated with conifer forest and greater distance to roads, and negatively associated with deciduous cover, highlighting the influence of forest management on caribou distribution in this region. This study provides the first SCR-based population density estimate for woodland caribou in Ontario. Our findings offer critical baseline data for conservation planning and recovery efforts and demonstrate the value of non-invasive SCR for linking population abundance to habitat features in wide-ranging, low-density species.