Status and habitat preferences of Uganda cheetahs: an attempt to predict carnivore occurrence based on vegetation structure

In this paper we examine whether the occurrence of cheetahs (Acinonyx jubatus) in Uganda can be predicted from habitat characteristics extracted from a vegetation map. We first established the status of the cheetah in Uganda through field-interviews that Gros conducted in 1990. Cheetahs occurred almost exclusively in the Karamoja region where we estimated 53–310 individuals. Based on 216 sightings, the average number of adults in all-adults sightings was 1.65 + SD 0.95 and the average number of cubs in family groups 2.5 + SD 1.65. Compared to Graham and Parker's 1965 East African survey, average adult group size was slightly smaller in 1990 and large family groups were rarer. Comparison with Gros 1990 survey showed considerably lower cub-to-adult ratio and percent of observations with cubs in Uganda than in Kenya. A Geographic Information System (GIS) analysis of vegetation structure in areas where cheetahs were observed and in those where none were reported suggested that cheetahs might favor habitats with 25–50% woody cover and grasses of medium height (50–100 cm). A discriminant analysis correctly classified 72.1% of used habitats and 70.4% of no-report habitats. A logistic regression analysis improved the correct allocation of used habitats by 2.2%. Either the discriminant function or the logistic regression, which require only four easily obtainable vegetation characteristics, may help to pinpoint suitable cheetah habitats for conservation purposes. Our approach could be adapted for analyzing habitat suitability for other species of carnivores.