Assessing occupancy of mammals in the Waterberg Mountain Catchment area across spatial scales

Abstract

The on-going global decline of large mammals mainly due to anthropogenic activities, necessitates enhanced management and conservation efforts. Hence, there is a considerable need for monitoring ecological estimates such as mammal occupancy and distribution in space and time. Monitoring the impact of factors responsible for mammal occupancy and distribution is equally important. Additionally, the scale at which such investigations are carried out is also key to ensure reliable outcomes for informed decision making in the management and conservation of mammals. The Waterberg Mountain catchment (WMC) is situated in the Limpopo province of South Africa. This a very biodiverse area, however, few comprehensive studies on the biological diversity and associated aspects from this area have been carried out. This area consists of several small, fenced wildlife properties. As such, the main activity in the area is eco-tourism (wildlife-based) for economic purposes as well as contributing to conservation. The small, fenced, and the isolated nature of the wildlife properties in the WMC necessitates a considerable level of human intervention in the management of mammals and their environments. I used camera trap data collected between December 2021 and April 2024 from the WMC from several wildlife properties. I analysed the data using multi-species Bayesian occupancy models, modelling mammal occurrence as a function of several site-level covariates, including distance to nearest permanent water sources, distance to roads, normalized difference vegetation index (NDVI), elevation, among others to investigate the key drivers of mammal occupancy in the WMC. Furthermore, I measured site-level covariates at multiple spatial scales to examine if variation exists in mammal occurrence patterns across scales. I also sought to compare occupancy patterns observed between managed (fenced and managed) and non-managed mammal species (free ranging with the ability to cross through fences between properties). The results generally showed that anthropogenic variables were more impactful than environmental variables in driving mammal occupancy. Moreover, deriving covariates at multiple spatial scales had an impact on the mammal occurrence patterns observed. Also, differences between mammal occupancy responses were observed between managed and non-managed mammal species. These general observations were made even though most of the results were not statistically significant (using 95% confidence intervals). These findings highlight the importance of anthropogenic infrastructures including roads, buildings, and fences on mammal occurrence and space use patterns within small, fenced wildlife properties. In addition, the results suggest that scale should always be considered when designing species-environment relationship studies. Finally, the results highlight the importance of the environmental context (small, fenced wildlife properties) on mammals’ occurrence patterns and distribution, which in turn influences mammal ecology.