Application of geovisual analytics to modelling the movements of ruminants in the rural landscape using satellite tracking data

Abstract

Geovisual analytics provides a framework for the development of decision support tools for landscape design, analysis and optimisation. An important application is modelling the spatial-temporal movements of ruminants and their grazing behaviour using global positioning system (GPS) collar units. This study describes the mapping and analysis of spatial distributions of animal waste products (which correlate with farm nitrogen [N] emissions) and also determination of animal feeding preferences (which correlate with animal welfare and production). Segmentation of local regions of animal N emissions provides support in meeting targets for local and international N leaching and greenhouse gas emissions. An agent-based model was used for pre-screening in order to gain insights into the clustering behaviour of sheep during feeding activities. Subsequent spatial analysis demonstrated that livestock excreta are not always randomly located, but concentrated around highly localised animal gathering points, separated by the nature of the excretion. In a separate study, the statistical significance of feeding choices was determined by testing a null hypothesis on animal boundary transitions between adjacent pastures using the binomial approximation. The analysis also included compensation for the precision of the GPS sensor, which produced a fuzzy decision boundary.