Modelling the movements of the Eurasian sparrowhawk Accipiter nisus from radio-tracking data
This thesis explores ways of modelling of animal movement that are applicable to empirical data, using the Eurasian sparrowhawk Accipiter nisus as the model species. Radio-tracking data on sparrowhawks were gathered during the breeding seasons of 2002 and 2003 in rural Aberdeenshire. In all, 15 sparrowhawks were tracked including provisioning males, females, non-breeding adults and dispersing juveniles. An analysis of quantified movements’ characteristics (move length, stopping times and distance from the nest site) in relation to behavioural, climatic and temporal variables, aggregated at various spatio-temporal scales, indicated that ambient temperature and wind speed affected sparrowhawk foraging range and the duration of foraging trips. There was however, little temporal pattern in recorded movements. A method of quantifying home range movements is proposed whereby habitually used sites were identified using cluster analysis of radio-tracking locations. The discrete movements around these sites by individual sparrowhawks can be considered as a Markov process and quantified in terms of a transition matrix. The equilibrium properties of such a matrix are then used to quantify the expected long term distribution of the individual within the home range. Modelling of transition probabilities suggests that much of the variation in observed movement patterns of three provisioning male sparrowhawks was explicable in terms of the spatial arrangement of the sites within the home range and that there was little variation between the behaviour of individual sparrowhawks. Finally, a mechanistic discrete space movement model was developed based on a few simple behavioural rules.