The confinement of many elephant (Loxodonta africana) populations in southern Africa to fenced reserves has made the management of such reserves difficult as elephants are a keystone species. Elephants are also water-dependent; therefore the availability of water affects the location, extent and intensity of elephant impacts on vegetation. Majete Wildlife
Reserve (Malawi) has undergone reform, during which it was fenced, artificial waterholes (AWPs) were created and wildlife reintroduced, including 220 elephants. Concerns have arisen as to the impact elephants may be having on the vegetation. In this thesis, two studies were conducted, along with a review of literature on elephant interactions with surface water. Woody vegetation changes in Majete were assessed by comparing woody vegetation cover datasets (based on remote-sensed vegetation classifications) of the reserve for 1985, 1990, 2000 and 2010. Woody cover loss was high between 2000 and 2010, therefore points of woody cover loss were further analysed in a spatial analysis. Using spatial and non-spatial environmental data, the effects of rainfall, fires, terrain (altitude, aspect, slope, hill and valley characteristics) and proximity to perennial water on woody vegetation cover were tested. Data analyses indicated that woody cover loss may have been caused by differing combinations of drought and herbivory or fire in different areas of the reserve. Where woody cover loss was attributed to herbivory, points of loss were not associated with proximity to perennial water. It was suggested that this is due to high perennial water availability in Majete, which would not limit herbivore foraging ranges in the dry season. Woody cover loss could not, however, be attributed to elephants in this study and further information on their use of and impacts around perennial water points was required. In a further study, the hypothesis tested was that different water point types (rivers, AWPs and springs) would be used at different intensities by elephants, and that perennial rivers would experience the most use. Elephant usage (including visits to water points, browsing levels and path use around water points) of selected perennial water points in Majete was monitored in the wet and dry season. The effects of season, water point characteristics (type, size and water quality) and habitat context (surrounding vegetation type, elevation and proximity to other water points) on elephant water point use were then tested. Elephant water point use was affected by season, as well as water point altitude and surrounding vegetation type. In areas of high perennial water availability, elephant browsing around water points did not decrease with increasing distance. It was suggested that this too could be because elephant browsing activity is not limited by water availability in Majete. Based on the findings of both studies, recommendations for water, elephant and fire management in Majete were proposed and discussed.