PATTERNS IN LANDSCAPE-WIDE SPATIAL HETEROGENEITY OF AMERICAN BLACK BEAR (URSUS AMERICANUS) POPULATIONS IDENTIFIED THROUGH GENETIC AND NONINVASIVE APPROACHES

Population-level patterns reflect the aggregation of individual-level movement, survival, and recruitment processes over a landscape. Estimating population density, distribution, and genetic structure is important for understanding species ecology, monitoring viability, and for developing effective management plans. Long-term monitoring is particularly necessary for detecting changes that have management implications. However, limited resources often impede the collection of sufficient high-resolution demographic data for robust population-level inferences, especially for species with extensive distributions and large ranges of individual movement. The American back bear (Ursus americanus) is a game species in New York (NY) that has been growing in abundance and expanding in distribution. However, robust knowledge of spatial variation in population density or genetic structure informative about current and future population trajectories is lacking. This research estimated patterns of landscape-wide spatial heterogeneity in NY bear populations using noninvasive, cost-efficient methods. First, I investigated the genetic structure of bears in NY and across the northeastern United States using neutral markers to reveal differentiation and patterns of restricted gene flow that may pre-date historical human disturbances. Genetic connectivity across political borders supports previous hypotheses of bear movement that motivate continued monitoring and coordination between management units. Second, I developed a citizen science (CS) program and conducted simulations with a novel integrated model to assess the utility of opportunistic CS data in augmenting systematic data to estimate population parameters. Then, I estimated bear density and patterns in bear density, distribution, and occupancy related to landcover types in southern NY with systematic spatial capture-recapture, occupancy, and CS approaches from 2015-2018. Across years, mean predicted density was 7.3 bears /100 km2 (95% CI: 4.7 – 11.5) with population growth, survival, fecundity, and landcover patterns suggesting that bears may continue to expand into areas with more human-impacted landscapes. Accounting for dependence between collocated sampling methods increased overall detection probability and highlighted the importance of appropriate spatial scales of different sampling methods for inference on population density. These findings provide the first spatially explicit, non-harvest based estimates of black bear population patterns across southern NY, and offer insights into the design of large scale, multi-method, long term population monitoring