Species richness and spatial use patterns of medium and large mammals in Majete Wildlife Researve, Malawi

Abstract

Human population numbers are increasing exponentially across Africa, with the associated effects of urban and agricultural expansion fragmenting landscapes and largely confining medium and large mammals to protected areas (PAs). The PAs themselves do not escape the anthropogenic pressure on their boundaries. Smaller reserves are particularly vulnerable, needing intensive management and interventions to maintain core habitat characteristics. To ensure that management actions have the desired outcomes it is important to understand how interventions influence the mammal community and associated biodiversity, within the context of each PA’s unique environmental characteristics. Monitoring variables such as species richness and mammal space use can provide valuable insight on how anthropogenic and environmental variables influence mammal species and communities. Majete Wildlife Reserve (MWR) is a small (691 km2) and isolated PA in southern Malawi within the Miombo woodland ecoregion. The reserve lies within a transformed matrix of agricultural and rural land use. MWR has had a turbulent past with most large mammal species lost to poaching by the end of the 20th century. However, since 2003 numerous large mammal reintroductions have taken place and the reserve is recovering. Historical species lists suggest that at least 41 medium and large mammal species (> 0.5 kg) used to occur in MWR, but no rigorous inventory has been made and the space use of these mammals is largely unknown. The aim of this study was to first determine the species richness of the terrestrial medium and large mammals at MWR, and secondly, assess the space use patterns and anthropogenic and environmental drivers thereof for the ungulate species. A six month systematic camera trap study was undertaken at MWR during the 2018 dry season (July – December) whereby 140 locations, spaced evenly across the landscape, were sampled for 40 days each, using 47 camera traps moved across three blocks. Species richness estimators (Frequentists and Bayesian) were used to estimate species richness. Ungulate space use was assessed against a suite of nine potential environmental and anthropogenic drivers (i.e. landscape curvature, fire, vegetation type, visibility, grass biomass, relative predator abundance, distance to water, distance to road and distance to fence) using an occupancy modelling framework. Over a period of 5 456 camera days, a total of 120 239 photographs were recorded, of which 12 202 were independent detections of 35 medium and large mammal species (20 herbivores, seven carnivores, six omnivores and two insectivores). This figure represents 85% of the medium and large mammal species historically present. The species richness estimators estimated that between 1-5 species were missed by the current survey, and both direct and indirect observations confirm the presence of another four species in MWR. Most species missed have specific habitat requirements such as aquatic habitat and rocky outcrops which were not specifically accounted for in the camera trap survey design. The mammal community structure was found to be atypical for Miombo woodland with megaherbivores (specifically, elephants Loxodonta africana) underrepresented. Ungulate space use patterns were mainly driven by environmental drivers with distance to water, vegetation type and visibility strongly influencing species. Distance to water affects hippo (Hippopotamus amphibius), elephant (Loxodonta africana), plains zebra (Equus quagga), eland (Tragelaphus oryx livingstonii), kudu (Tragelaphus strepsiceros) and impala (Aepyceros melampus), while bushbuck (Tragelaphus scriptus), common duiker (Sylvicapra grimmia), nyala (Tragelaphus angasii), impala (Aepyceros melampus), black rhino (Diceros bicornis), elephant, waterbuck (Kobus ellpsiprymnus) and hartebeest (Alcelaphus lichtensteinii) are affected strongly by the distribution of the different vegetation types. Waterbuck use dense vegetation more, and buffalo (Syncerus caffer) eland, impala, zebra and sable (Hippotragus niger) prefer more open areas. The anthropogenic activity around the reserve impacts black rhino and plains zebra space use negatively with their probability of space use increasing away from the reserve boundary. The space use of elephant and impala decreases with distance from roads. This study demonstrates the efficacy of a systematic camera trap survey and occupancy modelling framework in producing medium and large mammal species distribution data in Africa. Furthermore, a scientifically verifiable baseline measure of species richness and ungulate space use has been produced providing the basis from where the impact of future changes (natural and anthropogenic) can be assessed, and the success of conservation objectives evaluated.