Does fencing reduce the carrying capacity for populations of large herbivores?

The population size of migratory species like wildebeest (Connochaetes taurinus Burchell 1823) and zebra (Equus burchelli Gray 1824) could alter following their confinement in smaller areas (Whyte & Joubert 1988), and such confinement may result in the local denudation of herbaceous grassland vegetation by intensive grazing and subsequent general habitat degradation (Sinclair & Fryxell 1985). Data collected over 25 years provided an opportunity to evaluate some of the factors which govern the fluctuations of ungulate populations in a semi-arid nature reserve. Variables measured included managerial intervention through bush clearance for the purpose of reclaiming grassland habitats (Patchett 1978), annual rainfall and lion predation. The Sabi-Sand Wildtuin (SSW) is a fenced game reserve of 540 km2 which borders the western part of the central region of the Kruger National Park (KNP) (240 50' S 310 30' E; 22,000 km2). Rainfall patterns show a summer maximum and the region that includes both reserves lies between a southern 700 mm and a northern 550 mm isohyets. Average daily maximum/minimum temperatures are 300C/ 180C in January and 230C/80C in July (Schulze 1965). The terrain is gently undulating and composed largely of catenary sequences on granitoid rocks, supporting typical vegetation communities (Venter 1986). Numbers of wildebeest and zebra were recorded annually in the SSW from 1962 by ground transects that followed the strip method, where an area is sampled by counting all the individuals that occur within a strip whose boundaries are determined by a visibility factor (Collinson 1985). Data on kills by lion (Panthera leo Linnaeus 1758), hyaena (Crocuta crocuta Erxleben 1777) or leopard (Panthera pardus Linnaeus 1758) and mortality due to other factors (e.g. hunting old age, drought or unknown reasons) were collected on a monthly basis by the reserve Warden, based on direct sightings by game scouts, private