How humans shape wolf behavior in Banff and Kootenay National Parks, Canada

摘要: This paper describes the conceptualization and implementation of an agent-based model to investigate how varying levels human presence could affect elements wolf behavior, including highway crossings; use areas in proximity roads trails; size home ranges; activities, such as hunting, patrolling, resting, feeding pups; survival individuals Banff Kootenay National Parks, Canada. The consists a module primary component with five packs represented cognitive agents, grizzly bear, elk, modules that represent dynamic components environment. A set environmental data layers was used develop friction serves base map representing landscape over which wolves moved. decision built simulate sequence activities. implemented Java Programming Language using RePast, modeling library. Six months activities were simulated from April 16 October 15 (i.e., season coherent regard known behaviors), calibrated GPS radiocollars (n = 15) deployed 2002 2004. Results showed trajectories movements correlated observed (Spearman's rho 0.566, P < 0.001); other critical behaviors, time spent at den not traveling also correlated. simulations revealed behaviors noticeably affected by intensity presence. packs’ ranges shrank crossed highways less frequently increased In extreme example, pack whose range is traversed high-traffic-volume extirpated due inability hunt successfully under scenario wherein 10-fold. prototype developed this study may serve tool test hypotheses about effects on mammals, guide decision-makers designing management strategies minimize impacts species functionally related ecosystem.