Time-dependent memory and individual variation in Arctic brown bears (Ursus arctos)

Abstract

Animal movement modelling provides unique insight about how animals perceive their landscape and how this perception may influence space use. When coupled with data describing an animal's environment, ecologists can fit statistical models to location data to describe how spatial memory informs movement. We performed such an analysis on a population of brown bears (Ursus arctos) in the Canadian Arctic using a model incorporating time-dependent spatial memory patterns. Brown bear populations in the Arctic lie on the periphery of the species' range, and as a result endure harsh environmental conditions. In this kind of environment, effective use of memory to inform movement strategies could spell the difference between survival and mortality. The model we fit tests four alternate hypotheses (some incorporating memory; some not) against each other, and we found a high degree of individual variation in how brown bears used memory. We found that 52% (11 of 21) of the bears used complex, time-dependent spatial memory to inform their movement decisions. These results, coupled with existing knowledge on individual variation in the population, highlight the diversity of foraging strategies for Arctic brown bears while also displaying the inference that can be drawn from this innovative movement model.

Reinforced diffusions as models of memory-mediated animal movement

“…Explorations of such ideas could proceed mathematically by modifying the memory functional so that it was discontinuous with non-zero values only in some vicinity of each landmark. Another option would be to integrate reinforced diffusion models with parameter estimation approaches recently developed to gauge the importance of memory in a 'time since last visit' framework (see [52] for a case study involving seasonal movement of grizzly bears). Movement data are typically collected in discrete intervals (e.g.…”