Human-wildlife conflict issues on commercial farms bordering the Sperrgebiet and Namib-Naukluft National Parks borders, southern Namibia

Human-wildlife conflict is a global and growing problem, threatening carnivore conservation as well as the viability of farming practices in many areas. This study represents the first human-wildlife conflict study on commercial farmlands in southern Namibia, an arid environment with difficult farming conditions and low density carnivore populations which are highly susceptible to lethal control. Through novel combinations of ecological and questionnaire techniques to link the human perspective of the problem to the ecological data collected, the study aimed to identify areas in which to focus mitigation measures. It addressed two key research questions; which are the key farmland features carnivores are attracted to, and are these features identified by farmers as associated with higher levels of livestock losses? Additionally, as arid environments present a challenging environment for researchers, the study aimed to examine which methods are effective for surveying carnivores in such conditions. Using camera trapping, non-invasive hair collection devices and stable isotope analysis of diet, the study identified water sources, habitats with cover, and anthropogenic food sources as attractive to carnivores. Both water sources and mountain habitat were positively associated with carnivore risk scores and livestock losses respectively, and represent areas to avoid when kraaling smallstock. However, perceived risk of carnivore species was not associated with levels of livestock losses. The study revealed that whilst camera trapping is an effective method of surveying carnivores in an arid environment, the most efficient positioning of traps is dependent on the aim of the survey, with non-invasive hair collection devices being unsuitable for this environment. Being one of the first human-wildlife conflict studies to use both ecological and human perspective data simultaneously, it has demonstrated the need for this approach, suggesting the use of just one method is unlikely to provide a thorough understanding of this complex problem.