Sells, S.N., and Costello, C.M., 2023, Grizzly Bear Space Use in the US Northern Rocky Mountains (ver. 3.0, July 2024): U.S. Geological Survey data release, https://doi.org/10.5066/P91EWUO8.
Grizzly bear populations in the continental US are fragmented and connectivity among federal recovery areas is a conservation goal. Our objective was to identify potential connectivity pathways for grizzly bears, i.e., areas predicted to facilitate movements of individuals between populations. Our study centered on the three occupied (NCDE, CYE, and GYE) and one unoccupied (BE) ecosystems in and adjacent to Montana, as this area contains most of the potential landscape for connectivity within the conterminous US. To achieve our study objective, we built on a larger study initiated in Sells et al. (2022). In this first phase, the authors developed integrated step selection functions within currently occupied range in the NCDE to better understand how grizzly bears use habitat. Subsequent application of these models to the NCDE demonstrated high predictive power. In a second phase, Sells et al. (2023) demonstrated that models developed for the NCDE accurately predicted habitat use in nearby populations and were therefore expected to be transferable and reliable for predicting space use beyond the NCDE. In this present third phase, we demonstrate that multiple simulation methods help predict connectivity pathways and where to focus conservation efforts. This general study framework can be easily applied to other species to enhance understanding of animal space use, potential for functional connectivity, and conservation needs. Our present work expands on Peck et al. (2017), who employed step selection functions and randomized shortest path simulations to predict pathways for male grizzly bear movements between the NCDE and GYE. Randomized shortest paths enable simulating varying degrees of optimal versus exploratory movements between a given start and end node. We employed iSSFs to model directed movements (i.e., randomized shortest paths with start and end nodes) and undirected movements (i.e., from start nodes only with no predetermined end nodes; Sells et al. 2022). iSSFs extend traditional step selection functions to mechanistically model movement. We used Sells et al. (2022)’s iSSFs, built using movement data from GPS-collared grizzly bears monitored during 2003 – 2020 in the NCDE. Sells et al. (2022)’s iSSFs represented hypotheses that landscape features influencing grizzly bear habitat selection include food availability, terrain ruggedness, forested areas, forest edges, riparian areas, building densities, and distance to secure (unroaded) habitat. Because Sells et al. (2022)'s iSSFs demonstrated high individual variation in spatial behavior, our connectivity simulations were likewise individual-based to account for variations in movement behaviors. By taking a functional approach that used highly transferable movement models based on actual GPS-collared grizzly bears (Sells et al., 2023, 2022), we gained further insight than studies that focus on structural connectivity, or the degree to which patches are connected by similar habitat types. Our approach was not intended to predict areas where grizzly bears might settle, although our predictions likely suggest areas with good potential for occupancy. Instead, our focus was to identify potential dispersal pathways among ecosystems. Actual dispersal movements by grizzly bears are highly individualized and have rarely been documented, making them difficult to simulate. Conceptually, bears likely disperse into unoccupied range in two general ways, either by making long-distance, directional movements away from occupied range or by making shorter-distance meandering movements that encompass occupied range or stretch just beyond it. By predicting movement pathways using both directed and undirected simulations, our study likely accounted for both potential behaviors. Overall, we expect it will be helpful to consider both modeling approaches when interpreting the potential for habitat to provide connectivity between ecosystems.
