Blair, Amy Marshall2021-12-082021-12-082021https://hdl.handle.net/10539/32241A thesis submitted in fulfilment of the requirements for the degree of Doctor of Philosophy to the Faculty of Science, School of Geography, Archaeology & Environmental Science, University of the Witwatersrand, Johannesburg, 2021The iconic species Sclerocarya birrea subsp. caffra, locally known as marula, is an ecologically important savanna tree, valued as much culturally as it is economically. Marula is a keystone species and vital natural resource with limited demographic data on the current population status outside of conservation areas. Communities harvest marula within non-conservation area communal lands in the lowveld region of far north-east South Africa. Across these non conservative savannas, marula is an important source of non-timber forest products (NTFP), with populations affected by a range of socio-economic factors. Specifically, marula products provide livelihood sustenance for local communities in these savanna woodlands of southern Africa. However, the fruit products are increasingly being commercialized, which in addition to heavy reliance on fuelwood, is increasing resource pressure and starting to degrade this resource. As an ecologically important savanna species, it is imperative to understand if this resource use is sustainable. Understanding environmental degradation requires examining landscape changes over multiple spatial and temporal scales. Landscape ecology emphasises patterns, processes and scales that result from combined ecological and social drivers of change. Addressing environmental degradation within complex socio-ecological systems (SES) necessitates the coupling of ecological and social sciences in order to interrogate the mutual interactions between the social and environmental elements of the system. Applying a socio-economic lens to environmental degradation further links social meaning and symbolism to physical findings, allowing social ecologists to comprehend SES contexts where ecological data alone tells an incomplete story. With increasing environmental degradation, the imperative to combine quantitative data with qualitative research, in order to address the underlying causes of SES change, has never been more pertinent. Considering the context of the complex SES of the lowveld marula bioeconomy, marula trees are utilised for a variety of socio-economic reasons across different land-use types in these non-conservation areas. Understanding if resource use is sustainable requires therefore interrogating population demography across different land-use types. The aim of this study was to provide an updated inventory of marula population demographics linked to resource utilisation practices. As a heavily utilised natural resource, marula demographic data is used as a key indicator of population changes across not only land-use types, but also geographical zones. The key ecological indicators of marula density, gender ratios, mortality, tree damage, maturity ratios and size-class distributions were used to gauge marula population demographics within the lowveld areas of Phalaborwa and Bushbuckridge. Additionally, temporal comparisons are provided for the Bushbuckridge area, where a previous data set exists, allowing for a 15-year comparison. Ecological data from 5832 surveyed marula trees coupled with 240 individual household interviews and numerous informal focus groups were combined to understand the marula bioeconomy catchment. Social data for different behaviour patterns were used to provide context and explain anomalies in the population demographic data. Marula tree population data were gathered in rangeland transects and randomly selected fields and yards from eight human settlements, divided equally between the Bushbuckridge and Ba Phalaborwa municipalities. Population structures can be gauged by comparing tree density and size-class distribution (SCD) profiles across the three land-use types (homestead yards, fields and rangelands) prevalent in these savanna socio-ecological systems. When combining density 4 data with SCD profiling, it is apparent that land-use and harvesting practices are sustainable at present and that marula population structures are stable. However, certain trends in the ecological data suggest that there is cause for concern. In both Bushbuckridge (mean = 25.7±4.1 trees/ha) and Phalaborwa (mean = 13.9±1.9 trees/ha), overall density was highest in yards. While this would appear to reject the hypothesis of decreasing density with increasing anthropogenic interference, the ecological data alone only tells part of the story. Juvenile populations comprising large numbers of seedlings were common in yards, which the social data revealed was the direct result of discarded kernels from marula beer making. These seedling populations caused elevated overall mean densities in yards. Respondents indicated that these individuals are usually removed after a season, meaning that the majority do not recruit through to adulthood. While comparisons with previous studies showed that overall marula densities have increased in homestead yards, they have been rapidly decreasing in rangelands over a 15-year period, suggesting intense resource use, particularly for fuelwood. Female densities were higher than male densities in yards in both Bushbuckridge and Phalabarwa but not in either fields or rangelands, with the social data once again revealing that this selection is due to the prioritisation of female trees as the fruit producers. Male trees will therefore be selectively cut when land is cleared for residential plots. The preference for the female tree with it’s valuable fruit has resulted in a departure from the 1:1 ratio expected in dioecious species, specifically in yards where the female tree is maintained for it’s combined cultural and economic status. These findings are important as they indicate that value attribution can lead to species conservation where the cultural capital of a specific natural resource is high. Due to the numerous seedlings in homesteads, yards had negative size-class distributions (SCD) in both locations, indicating strong recruitment, whereas rangelands displayed positive slopes in both locations, indicating the opposite. This suggests possible overharvesting in rangelands, with the collection of all available fruit under the parent tree as a likely key-contributing factor to diminished recruitment. In both locations, adult and large adult populations were highest in rangelands with the social surveys indicating that the larger trees are chopped down in yards due to the concern that marula’s extensive branch and root systems might destroy roofs and building foundations respectively. Furthermore, according to the social accounts provided by respondents, the declining large trees across land-use types are attributed to extended drought conditions. Mortality increased across the land-use types from yards to rangelands in Phalaborwa, while the opposite trend occurred in Bushbuckridge with mortality increasing from rangelands to yards. In both locations, primary damage increased along the gradient from yards to rangelands. While instances of mortality and tree damage were minimal, complete removal of chopped trees in yards and rangelands means that there is likely insufficient evidence to create an accurate overall picture of negative resource practices. Understanding the causes of resource conservation and degradation in the context of important NTFPs is a key knowledge challenge. In light of demographic changes from increased resource use in recent years, this study serves as an updated inventory benchmark for marula populations in the area. Important lessons learned here can be applied to socio-ecological contexts where key natural resources are responsible for sustaining livelihoods. By linking marula population demographic data with the specific social behaviour patterns that shape these data, it is possible to provide new insights regarding the use of a key non-timber forest product (NTFP) within the complex socio-ecological system of the lowveld marula bioeconomy catchmentenThesis