A synoptic model of animal space use: Simultaneous estimation of home range, habitat selection, and inter/intra-specific relationships

Abstract

We propose a simple multivariate model for describing and understanding animal space use that estimates an animal's probability of occurrence as an explicit function of the animal's association with a fixed spatial area (i.e., home range), the spatial distribution of resources within that area, and the occurrence of other animals. We begin with a null model of space use to describe an animal's utilization distribution in the absence of effects from environmental covariates. We then use this null model as the foundation for a set of candidate models of space use that incorporate different combinations of environmental covariates where each model is chosen to reflect various hypotheses about important drivers of space use. Models are parameterized via maximum likelihood using location data collected from individuals at discrete times (e.g., telemetry) and spatially explicit environmental covariates. Information theoretic criteria are used to select the model(s) with most support from the data. The best model(s) is then used for both estimating the animal's home range and for inferring the relative importance of various environmental factors on space use. As an example, we applied our approach using male white rhino ( Ceratotherium simum) location data collected in Matobo National Park, Zimbabwe. The best synoptic model was able to capture the complexities of the utilization distribution while the model structure and parameter estimates provided a basis to infer the importance of various ecological factors affecting male rhino space use.