A scaled line-based kernel density estimator for the retrieval of utilization distributions and home ranges from GPS movement tracks

Utilization distributions (UDs) can be used to describe with what intensity an animal or human may use a certain geographical location within the environment it is living in. Such a density distribution model represents one way to describe and obtain an animals’ home range in wildlife ecology. Several methods to derive UDs and subsequently home ranges have been developed, for instance Kernel Density Estimation (KDE) and Brownian Bridges (BB), two probabilistic approaches, and Local Convex Hull (LoCoH) methods. KDE and LoCoH have been developed with point-based datasets in mind that describe the observation of an animal, and hence don’t utilize additional information that comes with GPSbased tracking data from collars. We have extended the point-based KDE approach to work with sequential GPS-point tracks, calling it a line-based KDE. We (i) introduce the basicapproach, (ii) refine it by introducing a scaling function to achieve a better model for the utilization function of space, and (iii) compare results of our approach with point-KDE and BB. Advantages of the line-based KDE by design are a better representation of utilization density near GPS points in comparison with the BB approach, and the ability to model and retain travel corridors in comparison with the point-KDE.