Quantifying permeability of linear barriers to animal movement: The permeability R package

Le résumé fourni par la source

Abstract Animals have always navigated environments characterized by linear features that influence movement, whether rivers, ridges or ravines. Large‐scale changes in land use have led to increasing interactions with anthropogenic features, especially roads and fences. These features can act as barriers, altering or impacting pathways, and understanding their impact is important for spatial planning and maintaining conservation priorities We present a straightforward method to estimate permeability of linear barriers to animal movement, available in the R package, permeability . We define an intuitive and flexible measure of permeability () that captures total impermeability (), semi‐permeability () or ‘hyper‐permeability’ (). Using maximum likelihood techniques, this measure can be estimated from a set of movement tracks and linear feature data, and modelled against any number of covariates that characterize a potential crossing (e.g. traffic, sex, season, time of day, kilometre marker, etc.), with tools available for obtaining confidence intervals, prediction outputs, model selection and mapping. We apply this tool to a large GPS tracking dataset of boreal woodland caribou ( Rangifer tarandus caribou ) which use habitat intersected by several highways in the Northwest Territories, Canada. Permeability of these highways is generally very low (), is influenced by traffic, and varies spatially and seasonally in ways that can readily be estimated and visualized. Our method provides concrete, quantitative estimates of barrier permeability and has broad applicability to any number of mobile and migratory species that interact with barriers. Code is provided in user‐friendly functions within the R package permeability , available for open‐source use.

Ce résumé expose les affirmations des auteurs. BNTIC ne l’interprète pas comme une validation indépendante des résultats.