Periodic continuous-time movement models uncover behavioral changes of wild canids along anthropization gradients

Most species exhibit periodic behaviors in response to cycles in resources and risks in the environment (circadian, lunar, seasonal, and so on). The ability to respond to anthropogenic perturbations by modifying periodic behaviors remains little studied, as does the question of whether and how periodic behaviors translate into periodic patterns in animal space use, on which we focus. Extending existing continuous-time stochastic movement models, we propose two new parametric approaches to detect and quantify periodic patterns of space use in animal tracking data, via periodicity in the expected position or circulation in the stochastic component of the path. We use them to study the movements of maned wolves (Chrysocyon brachyurus) and coyotes (Canis latrans) along anthropization gradients. These case studies illustrate how periodic patterns can be of natural origin (cycles in the environment) or anthropogenic origin (periodicity in human activity or restrictions on available habitat), suggesting a role for periodic patterns of space use in species persistence in anthropized areas. The method builds upon and extends existing functionalities in the R-package ctmm, in which the necessary tools are made available.