An increasing number of species face large-scale range contractions and predicted distributional shifts in response to climate change, shifting forest characteristics, and anthropogenic disturbances. Canada lynx (Lynx canadensis) are listed as threatened under the U.S. Endangered Species Act and recently recommended for delisting; however, delisting and predicted climate-driven losses in habitat quality and quantity may accelerate the decline of the northeastern Minnesota lynx population, one of six remaining resident populations in the contiguous United States. Our object was to develop a regional monitoring protocol and dynamic occupancy modeling framework to track spatiotemporal lynx occurrence, link occupancy dynamics to measurable habitat covariates, and identify multi-year core use areas at the southern periphery of the species range.
To do so, we implanted a large-scale survey across Northeastern Minnesota in Superior National Forest and surrounding designated critical habitat. Spatially and temporally replicated snow track surveys were used to collect lynx detection/non-detection data across five winters (2014-15 to 2018-19) covering >17,000 km within the 22,100 km2 study area. We used a dynamic occupancy model to evaluate lynx occupancy, persistence, colonization, and habitat covariates affecting these processes.
Results/Conclusions
Lynx occupancy probabilities displayed high spatial and temporal variability, with grid cell-specific probabilities near 0.0 in periphery regions to consistently near 1.0 in multi-year core use areas. Lynx were more likely to colonize and persist in areas with more evergreen forest and greater average snowfall, while forest characteristics (3-5 m and 10-30 m vegetation density) had mixed relationships with occupancy dynamics. We identified 55 grid cells classified as multi-year core use areas across a relatively contiguous region of high average snowfall and high conifer forest densities.
This study demonstrates a landscape-scale multi-year monitoring program to assess the effects of habitat characteristics, land use, and anthropogenic factors on species distributional changes and landscape-level occupancy dynamics. Our framework incorporating landscape-scale resource selection, core use area concepts, and dynamic occupancy models provides a flexible approach to identify population-level mechanisms driving species persistence and key areas for conservation protection.