Detecting effects of spatial memory and dynamic information on animalmovement decisions

1. Animals of many species demonstrate movement behaviour in which decisions are based on a variety of infor-mation. Effects of resources have been studied widely, often under the assumption that the environment is con-stant over the course of the study. Much less understood is the role of dynamic information about continuously changing resource availability and past experiences. Such information can be acquired during movement bouts and used for future decisions viamemory. 2. We present a new class of animal movement models, which incorporates a dynamic interplay of movement and information gain processes. Information is contained in a dynamic cognitive map. As an example, we con-sider time since last visit to locations and how this interacts with environmental information to shape movement patterns. Our models can be fitted to empirical movement trajectories and are therefore amenable to statistical inference (parameter estimation andmodel selection). 3. We tested the functionality of our method using simulated data. Parameter estimates were in accordance with true values used in the simulations, and model selection via Bayesian information criterion (BIC) was able to identify true underlying mechanisms of simulated trajectories. Thus, if time since last visit to locations influences movement decisions, ourmethod is able to detect thismechanism. 4. The use of dynamic information such as the one demonstrated in our example models likely requires cognitive abilities such as spatial memory. Therefore, our method can be used to reveal evidence of spatial memory in empirical movement data. Understanding the components of individual movement decisions and their interac-tions ultimately helps us to predict how population distribution patterns respond to environmental changes, such as landscape fragmentation and changing climate