Fence management and time since pack formation influence African wild dog escapes from protected areas in South Africa

In human-dominated and highly fragmented landscapes, keeping wildlife within reserve boundaries is vital for conservation success. In South Africa, fences are a widely employed conservation management tool for protected areas and are successful in mitigating human-wildlife conflict. However, fences are permeable, and predators are able to cross through reserve fences. African wild dogs (Lycaon pictus) often leave fenced boundaries, resulting in high capture and translocation costs. Moreover, when wild dog packs (up to 30 individuals) leave fenced reserves they enter human-dominated landscapes where they face strong persecution and livestock predation incurs high costs. The factors driving packs to leave managed reserves are poorly understood, thus, to effectively manage wild dogs in fenced systems, it is important to understand why they leave reserve boundaries. There are several hypotheses as to why wild dogs cross through reserve fences, including inter- and intra-specific competition, social behaviour, management, prey density and environmental variability. Using a long-term dataset comprising 32 resident packs across five reserves, we investigated the relative strength of these hypotheses on the probability of wild dogs exiting a fenced reserve. During the 14-year study period, we recorded 154 exit events. We found that the interaction of fence integrity and time since pack formation were the primary factors affecting the probability of a pack leaving a reserve. When fence integrity was poor, escape probability decreased with pack age likely due to the exploratory behaviour of new packs. When fence integrity was average, escape probability increased with pack age likely due to the fitness benefits of holding larger and more exclusive territories as packs age. When fence integrity was good, the probability of a pack escaping was very low (only 1% occurrence). The implications of this research suggest that the primary management consideration for reducing wild dog escapes from fenced reserves should be maintaining adequate reserve-wide fence integrity, rather than focusing on social structure or drivers of inter- and intra-specific competition.