Using Presence-Absence Data on Areal Units to Model the Ranges and Range Shifts of Select South African Bird Species

Abstract

The study of where species occur is an important concern in ecology. Over the last decade, the occupancy model has been the primary tool used in attempts to answer where, when and why species occur where they do. In this dissertation, the occupancy model is improved upon by adding a spatial component to account for similarities between adjacent sites. The model was applied to the Southern ground hornbill (<italic>Bucorvus leadbeateri<italic>), an elusive species that occurs in low densities through much of sub-Saharan Africa; South Africa is the southernmost end of its range and the distribution of the species there is unknown. The model uncovered new areas of potentially high occupancies, quantified the strong associations between hornbill occurrences and the availability of protected areas, and provided the appropriate structure to ask additional biological questions. The spatial occupancy model was then further adapted to include a temporal component in order to quantify range expansions or contractions when data are collected over time. This model was used on the common myna (<italic>Acridotheres tristis<italic>), an invasive species in South Africa whose range has been expanding in recent decades. The results suggest that the range of the myna continues to expand at a rate of 3\% a year and that its occurrences are associated with high human population densities.