B0578
Title: Accounting for unobserved spatial variation in step selection analyses of animal movement via spatial random effects
Authors: Rafael Arce Guillen - University of Potdam (Germany) [presenting]
Finn Lindgren - University of Edinburgh (United Kingdom)
Stefanie Muff - Norwegian University of Science and Technology (Norway)
Thomas Glass - University of Alaska Fairbank (United States)
Greg Breed - University of Alaska Fairbanks (United States)
Ulrike Schlaegel - University of Potsdam (Germany)
Abstract: Step selection analysis (SSA) is a common framework for understanding animal movement and resource selection using telemetry data. Such data are, however, inherently autocorrelated in space, a complication that could impact SSA-based inference if left unaddressed. Accounting for spatial correlation is standard statistical practice when analyzing spatial data, and its importance is increasingly recognized in ecological models (e.g., species distribution models). Nonetheless, no framework yet exists to account for such correlation when analyzing animal movement using SSA. The popular method, integrated step selection analysis (iSSA), is extended by including a Gaussian field (GF) in the linear predictor to account for spatial correlation. For this, the Bayesian framework R-INLA and the stochastic partial differential equations (SPDE) technique are used. Through a simulation study, the method is shown to provide accurate fixed effects estimates, quantify their uncertainty well and improve the predictions. In addition, the practical utility of the method is demonstrated by applying it to three wolverine (Gulo gulo) tracks. The method solves the problem of assuming spatially independent residuals in the SSA framework. In addition, it offers new possibilities for making long-term predictions of habitat usage.
