Impacts of anthropogenic disturbance on persistence patterns of forest mammals in an urban-forest mosaic of eThekwini Municipality, Durban, South Africa.

Abstract

The burgeoning human population size and the consequent land development pressures to meet its various needs has led to an unparalleled increase in the rates at which natural environments are converted for anthropogenic purposes. Among the major drivers of landscape modification by humans, urbanisation is arguably the most damaging, persistent and rapidly expanding across the globe. With the expansion of urbanisation, an increasing proportion of global biodiversity will be affected. The Ethekwini Municipality Area (EMA) is an urbanised landscape with high levels of biodiversity in South Africa. However, this area has a relatively large human population (~3.5 million), most of which is poor. Furthermore, it is located within a region (KwaZulu-Natal) undergoing rapid landscape changes. High development pressures, unauthorised development practices, conflicting governance systems and the preponderance of development priorities over environmental concerns has subjected most of the EMA to human activity. Considering these pressures, it is important to understand how wildlife adapt and persist in this human dominated landscape in order to guide conservation action. The aim of this study was to assess the impacts of anthropogenic disturbance on persistence patterns of forest mammals (excluding bats) in an urban-forest mosaic in the EMA, Durban, South Africa. The objectives were to: (1) determine factors affecting the occupancy of forest mammals; (2) determine the effects of landscape context on mammalian richness; (3) determine the effects of patch attributes and species’ ecological and life-history traits on nestedness patterns and (4) determine the effects of anthropogenic disturbance and abiotic factors on activity patterns and temporal niche overlap of mammals. Between May–September 2016 and December 2016–April 2017, mammalian surveys were conducted in forest patches within the study area using remote-triggered camera traps. Furthermore, data on vegetation structure at each camera trap location was recorded in order to better understand the habitat requirements of species. The results showed varying responses of mammals to landscape and habitat structural variables. The blue duiker (Philantomba monticola), bushbuck (Tragelaphus scriptus), bushpig (Potamochoerus larvatus), and Cape porcupine (Hystrix africaeustralis) were negatively affected by the loss and degradation of forest habitat and the increase in matrix development intensity whereas the large-spotted genet (Genetta tigrina) and vervet monkey (Chlorocebus pygerythrus) were relatively unaffected by such changes. Among habitat variables, an intact undergrowth and a high density of large trees were found to be important for the occurrence of many species found in this area as they provide sufficient breeding, roosting and browsing resources for specialist species that respond negatively to urban development. The results also demonstrated the importance of habitat area and its spatial configuration to the occurrence and persistence of mammals in this area. Patches that supported a high diversity of mammals were significantly larger, closer together, more contiguous and less separated by roads and urban development than patches that supported fewer species. The mammalian assemblage in the EMA was found to be significantly nested, with nestedness patterns related to patch size and isolation. This suggested that both the ability of species to persist on patches of various sizes and the ability to move to patches with different degrees of isolation affects the distribution and abundance of mammals in this area. Ecological specialists were found to be under more severe threat from further loss of forest habitat, which will likely threaten the long-term ecosystem functioning of forest habitat. When the activity patterns of species in the EMA were compared with activity patterns of the same species occurring in a less-disturbed landscape (Isimangaliso Wetland Park), no significant shifts in activity patterns were observed for most species except for common duiker (Sylvicapra grimmia), vervet monkey and Cape porcupine. Furthermore, there was only partial support for higher temporal overlap in activity patterns of ecologically similar species in the highly-disturbed EMA. Overall, the results presented in this thesis have provided insights into the persistence abilities of mammal species found in the EMA. The results have also provided basic ecological information on poorly known taxa, which will advance our understanding of their ecology locally and regionally. To ensure the continued persistence of mammals in the EMA, conservation efforts should prioritise the prevention of further loss of habitat, particularly large tracts of contiguous habitat. Furthermore, measures aimed at improving matrix permeability (e.g. stepping stones or corridors) should be promoted. This will help in reducing the negative effects of roads, which will ultimately increase landscape connectivity. Nevertheless, this is a metropolitan area with high development pressures, which are expected to increase even further in the future given the rapid population growth rate and the need to provide basic services to the people. Therefore, town planners, land owners, ecologists, and other decision makers need to consider the whole landscape, including the matrix, in the planning phase of future development projects in this area in order to minimise potentially negative effects on biodiversity.