Factors influencing cheetah (Acinonyx jubatus) distribution in Kafue National Park, Zambia

Globally, effective management of large carnivores, particularly threatened and endangered species like cheetah (Acinonyx jubatus), hinges upon a comprehensive understanding of their spatial distribution and conservation status. Despite being among the most threatened species within the Felid family, cheetah are data deficient throughout much of their known range, including within protected areas of Zambia, a potential stronghold for the species. In this study, existing camera trap data collected between 2018 and 2023, were used to assess cheetah occupancy estimates within Kafue National Park. These camera trap surveys were designed and run by the NGO Panthera Wildcat Conservation, in collaboration with Department of National Parks and Wildlife (DNPW) to estimate leopard density across the north, central, and southern regions of the park. To adapt this survey for use in cheetah occupancy analyses, I overlaid a 6 x 6 km grid cell onto the surveyed area (KNP), where each grid subsumed multiple camera traps stations, and was treated as an independent site. All subsequent analyses were conducted at this site level. Due to low detection of cheetah, I used a stacked occupancy model to analyse cheetah spatial distribution in relation to landscape/site covariates within KNP, interpreting the results as probability of site use rather than the true occupancy. Data on cheetah detection histories (presence and absence) were analysed against nine covariates, including closed versus open habitat types (namely open forest, closed forest and grassland), the relative abundance of both prey and other large predators (lion [Panthera leo] and spotted hyena [Crocuta Crocuta]), distance to water, anti–poaching patrol effort and distance to legal camps (lodges, NGO and DNPW bases). The results confirmed the presence of cheetah in KNP with 170 cheetah detections, 110 of which were independent. The detection probability for cheetah increased significantly with the sampling effort (total number of days cameras were active in a site) but the addition of other detection variables decreased the predictive performance of the models. The naïve occupancy varied from 28% in the south to 50% in the northern region. Similarly, cheetah detections were highest in the north and lowest in the south. Among the nine site covariates analysed, only prey RAI (β = 2.08), and proportion of open forest (β = –3.55) had a significant influence on cheetah probability of site use. Using cheetah by–catch data from a leopard survey provided a first estimate of cheetah occupancy in a Zambian National Park. These estimates could be improved through the addition of more sites and a longer survey duration (range: 80 –100 days) to improve estimates for this important carnivore species.

Grantor dc:publisher.institution

Department of Biological Sciences

Year dc:date.issued

2025

Author dc:creator

• Mulenga, Choolwe

Advisors dc:contributor.advisor

• O'riain, Mannus
• Woodgate, Zoe
• Overton, Kim-Young
• Durant, Sarah

• Acinonyx jubatus
• Kafue National Park
• Zambia

Language dc:language.iso

en

Handle dc:identifier.uri

http://hdl.handle.net/11427/42554

OAI identifier oai:identifier

oai:open.uct.ac.za:11427/42554