Wednesday, August 8, 2007 - 10:10 AM

COS 78-7: Finding biological relevance in the home range concept: The application of a Bayesian kriging approach to analyzing and predicting animal space use

Amanda V. Bakian, Ryan R. Wilson, and Kimberly A. Sullivan. Utah State University

The application of home range estimators for the quantification of animal space use is ubiquitous in animal ecology.  Animal home ranges are constructed for the purpose of deducing individual and population level resource use and selection.  The habitat configuration of home ranges is often heterogeneous impacting the distribution of resources. Home range core areas (areas of intense use) develop as a result of a clustered or non-uniform pattern in resource selection. The accurate and non-arbitrary delineation of core areas is challenging. We present a novel approach to home range and core area analysis using a geostatistical method that directly links space use and resource selection through observations of animal behavior.  Specifically, we develop a generalized linear spatial model and apply a Bayesian kriging method to the analysis of a behaviorally explicit Willow Flycatcher (n = 21) radio telemetry data set collected over the 2002 and 2003 breeding seasons.  This method proves to be a powerful tool for predicting resource use throughout a spatially heterogeneous home range.  Strengths of Bayesian kriging include its inclusion of covariates and accommodation of spatial autocorrelation. When used in collaboration with more traditional core area estimators, Bayesian kriging enables the fine-tuning of core area boundaries.  In our Willow Flycatcher case study, the method is applied for the purpose of predicting foraging and vocalization behavior in the home ranges of male and female flycatchers.