In the Greater Yellowstone Ecosystem (GYE), elk movement between subpopulations is likely to explain the spatial expansion of brucellosis in elk, and could increase the risk of transmission from elk to livestock. The highest probability for Brucella abortustransmission among elk occurs during February and March when elk are largely aggregated on winter range, but transmission can also occur during the spring months when migration to summer range begins. Little is known, however, about how elk movement and regional connectivity varies during the transmission period, which limits our ability to predict the future rate and direction of disease spread. Therefore, we have compiled GPS collar data from 850 GYE elk collected over 15 years to evaluate time-varying movement and connectivity models in an effort to better understand elk connectivity during the brucellosis transmission period. Connectivity models rely on landscape resistance surfaces that are typically estimated using habitat suitability derived by resource selection functions (RSF). However, habitat suitability may not explain fast directed movements such as migration or dispersal important to elk connectivity. Thus, we also compared landscape resistance derived from continuous time Markov chain (CTMC) models of movement rate to those derived from RSF.
Results/Conclusions
The highest degree of empirical connections between elk subpopulations occurred in May, during the start of spring migration. During this time, coefficient estimates from CTMC and RSF were negatively related, suggesting that elk move fast in undesirable habitat. In this case, CTMC-derived resistance would provide a better understanding of landscape attributes important to migration, and help identify key corridors for those directed movements. Both methods, however, revealed significant individual variation in response to landscape variables, thereby affecting connectivity model accuracy and our understanding of how brucellosis has spread across the region. Our study examined the use of connectivity models for understanding disease transmission across broad scales and highlights the importance of considering speed and individual variation in connectivity models.