Intraspecific and interspecific olfactory interactions and density of leopards Panthera pardus (Linnaeus, 1758) in Sabi Sands Game Reserve, South Africa

The olfactory behavioural ecology of large solitary carnivores is still poorly understood. Meanwhile, these species represent challenging cases in conservation and management. They are frequently involved in conflicts with farmers, depredating livestock and being killed in retaliation. Low densities and large territories also make it hard to assess their population status accurately. This thesis aimed to improve our understanding of the African leopard (Panthera pardus) olfactory ecology and investigate whether this knowledge could help mitigate farmer-leopard conflicts or facilitate the management of the species. Chapter 1 described the critical role of olfactory communication in animals, large carnivores, and leopards. I reviewed the evidence for the use of intraspecific and interspecific carnivore scents as deterrent strategies and movement-manipulating tools in wildlife management. Peer-reviewed evidence is limited, and conclusions vary between studies. Manipulative experiments observed target carnivores being either repelled or attracted to the scent-marked area. I concluded that, if any, the deterrence potential of scent marks against leopards likely relied on its two strongest sympatric competitors’ scents, lions (Panthera leo) and spotted hyenas (Crocuta crocuta). Chapter 2 experimentally characterized leopard intra and interspecific olfactory interactions at marking sites and assessed the effects of lions’ and spotted hyenas’ scents on leopards. Camera traps monitored leopard scent-marking sites in the western sector of the Sabi Sands Game Reserve. Lion and spotted hyena activity at marking sites was simulated with the addition of their scats. If successful in deterring leopards from an area, these scents could further be synthesized and used around farms to prevent conflicts. Results supported territory advertising and reproductive functions of leopard scent marking and revealed a higher marking frequency in females than previously reported. Frequent olfactory behaviours included spraying, rubbing, and investigating scents. Generalized linear mixed effect models revealed no deterrent effects of lion and spotted hyena scents on leopards. Additionally, leopards spent significantly more time marking and investigating sites where lion scats were present. Thus, olfactory cues of guild members did not spatially displace or affect leopards over time. Instead, leopards used scent-marking sites to gather information about conspecifics and heterospecifics and to advertise their presence reciprocally. This behavioural strategy does not enable the use of intraguild scent interactions to manipulate the movements of leopards. Chapter 3 estimated leopard density and assessed the impact of intraguild interactions on the estimate precision. I used data from the previous scent survey and implemented a control survey with 15 paired camera traps at road junctions along a 5-km² grid. Spatially-Explicit-Capture-Recapture analyses under both Bayesian and maximum likelihood frameworks estimated leopard densities. As expected, the accuracy and precision of estimates were relatively close between surveys. Neither the translocated scats or habitat types covariates impacted density results. However, the scent survey better described the sexually mature leopard population, whereas the control survey sampled individuals representative of the entire population structure. Cost-effectively and parameter-wise results provided more support for the control survey and Bayesian approach. Thereby, the estimated leopard density in the study area was 12.81 individuals (+/-0.07)/100 km². This type of small-scale in situ experimental study helps field biologists to make optimal decisions and better defines the range of management actions available to conservation.