Tracking in sand revealed data on hunting and kill rates, range, movements, activity, cover and water use, reproduction and interactions with other carnivores, by the leopard Panthera pardus in the Kalahari Gemsbok National Park. For leopards in the interior, 812,5 km of tracks were followed for 54 days, and 205,1 km for 15 days for females with cubs. In the Nossob riverbed 30,2 km of tracks were folowed in eight days. Medium-sized mammals featured prominently in the diet of all leopards, with prey used influenced by habitat type. Leopards in the interior moved greater distances than those in the Nossob riverbed. Leopards rested frequently at the onset and end of activity and used dense vegetation and aardvark Orycteropus afer and porcupine Hystrix africaeaustralis burrows as daytime cover. Leopards are independent of water, and females apparently have no definite breeding season. Lions Panthera leo dominate leopards, but the outcome of leopard/spotted hyaena Crocuta crocuta encounters depend on the size of the leopard and the number of hyaenas in the pack. Leopards in the Kalahari Desert are opportunists which occupy this harsh envi- ronment successfully.

No abstract

The concept of carrying capacity originally was developed for domestic grazers and does not make provision for the wide variety of diets found in wild African herbivores. In the African savannas, herbivores are classified into 4 basic dietary classes: low‐selective grazers, high‐selective grazers, mixed feeders, and browsers. Given these conditions, a more appropriate approach was needed than the agricultural method based on livestock units to determine stocking rates for wildlife. Consequently, a new approach was developed that recognizes plant resource variation at the plant community level and differentiates between the grazing and browsing component in the diet of herbivores in the African savannas. The model used to calculate the grazing and browsing capacity on wildlife ranches provides for rainfall variability, quality and quantity of available grazing and browse, dietary requirements of each type of wildlife, and availability of suitable habitat. In this model the conventional conversion of wildlife to a Large Stock Unit has been replaced by a Grazer Unit, which is the equivalent of a 180‐kg blue wildebeest (Connochaetes taurinus), and a Browser Unit, which is the equivalent of a 140‐kg greater kudu (Tragelaphus strepsiceros). By separating the grazing and browsing components in the diet of wildlife for stocking density calculation, the diversity in the vegetation resources is optimally utilized. This concept is exemplified with an actual case study of a wildlife ranch in a South African savanna. In doing so, quantitative and clearly defined parameters are provided on which to base wildlife management decisions.

Vegetation changes around artificial watering points were monitored in the Kalahari Gemsbok National Park from 1978 to 1989. Both rainfall and grazing influenced the vegetation but rainfall appeared to be the more significant factor. The above-average rainfall during the years preceding 1978 contributed to a relatively high basal cover in 1978. Since 1978, however, below-average rainfall was recorded resulting in a decline in basal cover, presence, frequency and density of most of the plant species. It seems likely that a wet cycle commenced in 1988, which is reflected in higher rainfall and an increase in basal cover, presence, and density. Some species maintained a relatively constant presence, some increased, and some declined during the study period. The monitoring is being continued

No abstract