MENVSC / Department of Ecology and Resource Management / Wetlands are amongst the world’s most important ecosystems providing many direct and indirect benefits to local communities. The majority of South Africans residing in rural areas depends mostly on natural resources for their livelihood. However, wetlands in South Africa continue to be the most threatened ecosystems primarily due to unsustainable use and poor resource management. Additionally, the history of South Africa has been characterised by exclusion of local communities in the process of decision-making and general management of natural resources. The aim of the study was to investigate possible challenges affecting the management of Thulamela wetlands by assessing the level of interaction and conflicting interest amongst participating stakeholders, including role of wetlands on local communities for possibly improved management scenarios. The study used questionnaires, interviews and observations to capture data on the local communities and management stakeholders. Seven wetlands within Thulamela were selected as study areas and the study population was selected based on their specialised expertise, involvement, and closeness to wetlands. Data was analysed using SPSS, Microsoft Excel and also using thematic analysis in NVIVO. The results show that selected wetlands are highly beneficial in supporting the local communities. Based on the socio-economic and demographic characteristics (family size, age, gender, and employment status) measured, the study deduced that unemployment rate or low income of the respondents is the main contributor to an increase dependency on wetland utilization. Additionally, the results revealed that unregulated use and excessive agricultural practices such as cultivation and livestock grazing are common in all study areas, hence further degradation of these wetlands. One of the findings from the study was the destruction of wetlands through expansion of human settlement. The population increase in the areas was found to be major drivers of socio-economic challenges causing people to spread through and exploit wetlands. Consequently, human settlement along the wetland area has resulted in the extensive clearance of natural wetland vegetation. Furthermore, the results show that there is poor wetland information transfer to the local communities most likely due to none/or insufficient outreach programmes. The current management arrangements and structures for selected wetlands are not being practiced through the unequal representation amongst management stakeholders and poor inclusion of local communities in management processes. Additionally, there are currently no openly known active platforms provided upon which stakeholders are able to air their views on wetlands management issues. The findings further show differences in perceptions amongst wetlands users, non-users and management stakeholders. The management stakeholders have a relatively strong focus on livelihood and environmental problems, they regard rules and regulations on wetland use’ as a relatively central variable. On the other hand, the local communities are currently more concerned about the benefits they receive from the wetlands than the conservation of those systems. The study also revealed unequal representation amongst participating management stakeholders. The findings show poor interaction between the management stakeholders and the local communities; differences in perceptions amongst resource users, non-users and managers; exploitation of wetlands resources; poor wetlands information transfer. The results suggest that centralised top-down rules and regulations on wetland use are not sufficient for maintaining the wetland ecosystem and this poses a challenge to sustainable wetland management. Therefore, there is a need to develop shared understanding through bottom-up approaches to wetland management that are nested within national regulatory frameworks, ideally combined with awareness building and knowledge sharing on the ecological benefits and management of wetland. / NRF

Overfishing in the Benguela Current Large Marine Ecosystem (BCLME) resulted in degradation of the ecosystem. This study used the Driver-Pressure-State-Impact-Response (DPSIR) indicator framework to determine whether the ecosystem is now recovering. Indicator trends were analysed using various data sources that included government institutions and intergovernmental institutions. The results showed that the overall effect of Driver indicators was negative. This was mainly because of socio-economic pressure such as the need to create more jobs in light of rising national unemployment and the declining contribution of the fisheries sector to Gross Domestic Product (GDP). In addition to scientific advice, socio-economic factors also influenced the determination of Total Allowable Catches (TACs). The overall trend of Pressure indicators was positively influenced by the effect of TACs. The TACs reduced the quotas allocated for commercial fishing. Environmental factors did not seem to play a significant role in this study. State indicators had mixed results with the indicators assessed almost split in the middle between those showing a positive trend and those showing a negative trend. On the other hand, Impact and Response indicators showed overall positive results. Therefore, the conclusion of the study was that the degradation of the BCLME has slowed down and there are some signs of recovery. / Environmental Sciences / M.Sc. (Environmental Science)

Savannahs regulate an agro-ecosystem crucial for the production of domestic livestock, one of the main sources of income worldwide as well as in South African rural communities. Nevertheless, globally these ecosystem functions are threatened by intense human exploitation, inappropriate land use and environmental changes. Leaf area index (LAI) defined as one half the total green leaf area per unit ground surface area, is an inventory of the plant green leaves that defines the actual size of the interface between the vegetation and the atmosphere. Thus, LAI spatial data could serve as an indicator of rangeland productivity. Consequently, the accurate and rapid estimation of LAI is a key requirement for farmers and policy makers to devise sustainable management strategies for rangeland resources. In this study, the main focus was to assess the utility and the accuracy of the PROSAILH radiative transfer model (RTM) to estimate LAI in the South African rangeland on the recently launched Landsat 8 sensor data. The Landsat 8 sensor has been a promising sensor for estimating grassland LAI as compared to its predecessors Landsat 5 to 7 sensors because of its increased radiometric resolution. For this purpose, two PROSAIL inversion methods and semi- empirical methods such as Normalized difference vegetation index (NDVI) were utilized to estimate LAI. The results showed that physically based approaches surpassed empirical approach with highest accuracy yielded by artificial neural network (ANN) inversion approach (RMSE=0.138), in contrast to the Look-Up Table (LUT) approach (RMSE=0.265). In conclusion, the results of this study proved that PROSAIL RTM approach on Landsat 8 data could be utilized to accurately estimate LAI at regional scale which could aid in rapid assessment and monitoring of the rangeland resources. / Environmental Sciences / M. Sc. (Environmental Science)

The KwaZulu-Natal Drakensberg, South Africa, is predominantly a grassland ecosystem maintained by fire. The effect of the current burning regime on invertebrate and flowering forb diversity in this ecosystem is poorly understood. The overall aim ofthis study was to contribute towards the development of an effective burning regime for the KwaZulu-Natal Drakensberg that will conserve invertebrates and indigenous forbs, two major components of biodiversity. The objectives were to examine the effect of fire and fire frequency on flowering forb and invertebrate species diversity, to determine whether fire frequency, time since last burn or locality were influencing species composition, and to identify potential biodiversity indicators that reflect overall species richness for use in monitoring of invertebrates and forbs. Sampling took place in March, September and November of 2002 at Giants Castle Game Reserve. Invertebrates were sampled using sweep netting and targeted netting along transects, yellow pan traps and soil quadrats. Invertebrate taxa sampled were ants (Formicidae), butterflies (Lepidoptera), grasshoppers (Orthoptera), leafboppers (Cicadellidae), bees (Apoidea), bee flies (Bombyliidae), hover flies (Syrphidae), robber flies (Asilidae), spiders (Araneae), earthworms (Oligochaeta) and millipedes (Diploda). These were identified to species level with the assistance of taxon experts. Flowering forbs were sampled using five replicates of five by five metre quadrats randomly placed in each site. Overall flowering forb and invertebrate species diversity was higher in grasslands that were burnt for two consecutive years in 2001 and 2002 than in grasslands that were not burnt during those two years. Frequently (annual) and intermediately (biennial) burnt grasslands had significantly higher invertebrate and flowering forb diversity than infrequently (five years without burning) burnt grasslands. This, together with the fact that grasslands burnt during the year of sampling had higher species richness than grasslands burnt two and five years previously suggests that invertebrates and forbs are generally resilient to fire and many forb species appear to be stimulated by fire. However, each burn frequency had its own suite of unique flowering forb and invertebrate species. Invertebrate communities were influenced mostly by locality and the length of time past since the last fire and flowering forb communities were influenced mostly by the length oftime past since the last fire. Fire frequency had the least influence on both invertebrate and forb communities. Ecological succession occurred after each fire in the invertebrate communities but forb communities appear to need more than five years without fire for ecological succession to occur. The findings of this study therefore suggest that using a combination of three fire frequencies would result in patches of grassland in various stages of ecological succession, and would conserve species unique to each burning frequency, and would therefore conserve maximum diversity. Flowering forb species richness and certain invertebrate taxa (ants, leafboppers, spiders and bees) have the potential to act as indicators of overall invertebrate species richness for use in monitoring programmes. / Thesis (M.Sc.)-University of KwaZulu-Natal, Pietermaritzburg, 2006.

Overfishing in the Benguela Current Large Marine Ecosystem (BCLME) resulted in degradation of the ecosystem. This study used the Driver-Pressure-State-Impact-Response (DPSIR) indicator framework to determine whether the ecosystem is now recovering. Indicator trends were analysed using various data sources that included government institutions and intergovernmental institutions. The results showed that the overall effect of Driver indicators was negative. This was mainly because of socio-economic pressure such as the need to create more jobs in light of rising national unemployment and the declining contribution of the fisheries sector to Gross Domestic Product (GDP). In addition to scientific advice, socio-economic factors also influenced the determination of Total Allowable Catches (TACs). The overall trend of Pressure indicators was positively influenced by the effect of TACs. The TACs reduced the quotas allocated for commercial fishing. Environmental factors did not seem to play a significant role in this study. State indicators had mixed results with the indicators assessed almost split in the middle between those showing a positive trend and those showing a negative trend. On the other hand, Impact and Response indicators showed overall positive results. Therefore, the conclusion of the study was that the degradation of the BCLME has slowed down and there are some signs of recovery. / Environmental Sciences / M.Sc. (Environmental Science)

This study examined the current plant diversity status and the impact of drivers of change on the peat basins of the Kromme River peatland. It was conducted at six sites over sixty one years in the Eastern Cape Province of South Africa. I reviewed the rapid habitat and biodiversity loss of wetlands globally and discussed the distribution of wetlands and specifically peatlands in South Africa. Plant species diversity was assessed using Modified- Whittaker plots. The influence of environmental variables on floristic composition and distribution was investigated using ordination techniques (DCA and CCA). Land use dynamics were assessed by applying GIS techniques on orthorectified aerial images. Six different peat basins were subjectively classified into good, medium and poor condition peat basins. The good condition peat basin (Krugersland) was the most diverse in plant species (4.1 Shannon-Weiner’s index) (p> 0.20; F = 11.04; df = 2), with the highest mean number of plant species (32.5 ± 3.4). This was followed by the medium condition class (Kammiesbos) (26.5 ± 9.0) and poor condition class (Companjesdrift) (22.5 ± 8.9). On average, species composition was not evenly distributed across the peat basins (p> 0.21; F = 0.94; df = 2), since 77.8% of the Shannon-Weiner evenness index obtained were less than one. However, there were variations in plant species richness across six peat basins as confirmed by Oneway ANOVA test (p= 0.0008, F = 1241.6, df = 4). Key environmental variables that influenced plant species distribution and structure were erosion and grazing intensity, potassium, phosphorus, soil pH and calcium. Total species variance accounted for in the first two axes for ground cover and plant height were 40.7% and 56.4% respectively. Alien species (e.g. Acacia mearnsii and Conyza scabrida) were common in degraded peat basins, whereas good condition peat basins supported indigenous species (e.g., Cyperus denudatus, Chrysanthemoides monolifera and Digitaria eriantha). Analysis of aerial images revealed a general progressive decrease in the peatland area between 1942 and 1969 in the good (Krugersland) and poor (Companjesdrift) condition class, with a marginal increase from 1969 to 2003. Peatland area in the good and poor condition class decreased by 5.3% and 8.3% respectively between 1942 and 1969, with a marginal increase of 1.5% and 4.1% respectively from 1969 to 2003. Annual net rate of change in peatland area over the 61 year period was -0.32% (good condition class) and - 0.79% (poor condition class). Transformed lands were impacted by drivers of change such as alien invasives, agricultural activities, erosion and sediment transport. The area under alien invasives increased by 50% between 1942 and 2003, with an annual net rate of change of +0.82 (good condition class) and +1.63% (poor condition class).

Savannahs regulate an agro-ecosystem crucial for the production of domestic livestock, one of the main sources of income worldwide as well as in South African rural communities. Nevertheless, globally these ecosystem functions are threatened by intense human exploitation, inappropriate land use and environmental changes. Leaf area index (LAI) defined as one half the total green leaf area per unit ground surface area, is an inventory of the plant green leaves that defines the actual size of the interface between the vegetation and the atmosphere. Thus, LAI spatial data could serve as an indicator of rangeland productivity. Consequently, the accurate and rapid estimation of LAI is a key requirement for farmers and policy makers to devise sustainable management strategies for rangeland resources. In this study, the main focus was to assess the utility and the accuracy of the PROSAILH radiative transfer model (RTM) to estimate LAI in the South African rangeland on the recently launched Landsat 8 sensor data. The Landsat 8 sensor has been a promising sensor for estimating grassland LAI as compared to its predecessors Landsat 5 to 7 sensors because of its increased radiometric resolution. For this purpose, two PROSAIL inversion methods and semi- empirical methods such as Normalized difference vegetation index (NDVI) were utilized to estimate LAI. The results showed that physically based approaches surpassed empirical approach with highest accuracy yielded by artificial neural network (ANN) inversion approach (RMSE=0.138), in contrast to the Look-Up Table (LUT) approach (RMSE=0.265). In conclusion, the results of this study proved that PROSAIL RTM approach on Landsat 8 data could be utilized to accurately estimate LAI at regional scale which could aid in rapid assessment and monitoring of the rangeland resources. / Environmental Sciences / M. Sc. (Environmental Science)

Through a combination of theoretical discussion and case study analysis from two villages in the Eastern Cape Province, South Africa, this thesis explores the relationship between institutional change and ecosystem dynamics through a multi-scale approach that combines local and scientific knowledge. Several conceptual approaches were combined in this study. These included; the Resilience perspective, the Sustainable Rural Livelihoods approach, the Millennium Assessment framework and principles, and the Environmental Entitlements approach. Various participatory research techniques were employed which combined with GIS technologies, existing data sets, and historical records. Common pool resource areas are social spaces, where local values attached to resources are institutionally mediated, politically nuanced, economically interpreted and historically situated. Political driving forces at various scales have played a disproportionate role in local level institutional functioning in the case study area. In particular, inappropriate state-lead interventions into land use planning have weakened local level institutions, and have reduced the ability of the linked social-ecological system to cope with change and uncertainty. People and ecosystems become more vulnerable when driving forces such as political upheaval, economic depression and drought over-lap. However, rural people are not mere spectators in the face of these driving forces; they respond both reactively and proactively to ensure resilience to change and uncertainty. Ecosystems at Mt Coke play a key role during times of crisis in rural livelihoods. These ecosystems have undergone various short-term cyclical changes largely in response to rainfall fluctuations, and some longer-term changes linked to political events and trends that have affected management practices and local institutions over time. Orthodox ecological interpretations of ecosystem change appear to ignore four key factors identified in this study: 1) the role of institutions in shaping access to resources, 2) the demand for resources in rural livelihoods, 3) the dynamic interaction between social and natural systems, and 4) the interaction between social and natural systems across scales of analysis. The future of common pool resource management lies in the combination of local and scientific knowledge through an adaptive management approach that encourages learning and adaptation in local level institutional structures.

PhD (Zoology) / Department of Zoology / Ecosystem services and disservices are important features in agro-ecosystems and both can have relevant economic impacts on farmers. While there has been much research on the value of ecosystem services, especially that of pest control by bats and birds, ecosystem disservices are often overlooked or estimated separately. Both, services and disservices, however, change with landscape and may be supported by natural vegetation. It is hence inevitable to assess them together to uncover their relative value and importance. Habitat loss and fragmentation are major threats to biodiversity and ecosystem services. Agricultural intensification can cause habitat loss and can negatively affect natural ecosystems and their services. However, the removal of natural vegetation can also reduce ecosystem disservices that origin from natural areas. This study focused on the economic trade-off between biocontrol by bats and birds and crop raiding by vervet monkeys in South African macadamia orchards, using vertebrate exclusion experiments. Crop quality and yield were assessed and extrapolated to income values. The study furthermore focused in more detail on the impact of orchard edge habitat and farm management on bats as ecosystem service providers as well as the arthropod community, which can provide both services (pollination, biocontrol) as well as disservices (crop damage). The macadamia industry is highly profitable and plantings are expanding worldwide. South Africa is currently the largest macadamia producer in the world and the study area, Levubu, holds some of the oldest macadamia plantings in the country. The subtropical climate and high annual rainfall make this area an intensively farmed landscape. Levubu lies at the foothills of the Soutpansberg, a centre of biodiversity and patches of natural vegetation wind through orchards of macadamia, avocado, banana and timber. Insect pest pressure is high in the macadamia industry, with several stink bug species (Hemiptera) and two moth species (Lepidoptera) accounting for most of the damage to the crop. Chapter One provides a detailed overlook over the South African macadamia industry, describes the major pest insect species and their impact on the crop and highlights the role of ecosystem services and disservices and the potential importance of remaining natural areas. Over three consecutive seasons, between September 2015 and May 2018, bats, birds and monkeys were excluded, using cages with nets, from a total of 96 macadamia trees. Four treatments were applied (Full, Day, Night, Control) to distinguish between effects of the different functional groups. Exclusions took place at orchard edges with natural or human-modified edge habitat to assess the impact of vicinity to natural vegetation on pest control and crop raiding. Results presented in Chapter Two showed that biocontrol by bats and birds was stronger near natural vegetation and significantly impacted crop quality as well as yield. Effects of bats and birds were still significant at human-modified edges, whereas crop raiding by monkeys is limited to the natural orchard edges. The economic impact analysis shows that the benefits of biocontrol through bats and birds outweighed the income losses due to crop-raiding monkeys. The value of bats and birds was as high as USD 5000 per hectare per year through prevented insect damage and a 60% crop increase. Crop-raiding affected the yield, which dropped by about 26% and resulted in losses of about USD 1500 per hectare. These results highlight the importance of integrating natural areas into agricultural landscapes, even if they incur the risk of being disadvantageous. Farmers need to be made aware of the enormous economic benefits. Effective crop raiding mitigation strategies still need to be researched to avoid negative association with natural areas. These can, however, possibly also limit access for biocntrol agents, like bats and birds and thereby reduce their economic impact. Especially the removal of natural areas or establishment of buffer zones of unpalatable crops between agricultural land and forests is not advisable. Integrative research in agro-ecosystems on trade-offs between a variety of ecosystem services and disservices is necessary in future, rather than assessing them separately. The bat community on macadamia farms was monitored monthly and compared between orchard edges. Stationary bat detectors were placed at each set of exclosures for two consecutive nights a month, automatically recording from sunset to sunrise. Recorded bat calls were identified manually to species level and activity converted to Miller’s activity index, which counts the number of active minutes per species per night. We analysed the activity per feeding guild over season and landscape setting, estimated species richness and diversity (Hill’s numbers) and species turnover between farms, season and landscape setting, using diversity partitioning. The bat community was mostly influenced by seasonality with season turn-over accounting for 21% of total bat diversity (21.25 species). Edge habitat on the other hand only accounted for 5%. We found higher species richness and activity in the high season compared to the low. While there was no difference in diversity in the high season between the two orchard edge types, species diversity at the human-modified edge was lower than in the natural during the low season. Natural habitat might therefore be more important in the low season, while its effect is overwritten by high food availability during the high season. Clutter-edge species furthermore were more active at the natural orchard edges than open-air feeders, which were in turn more active at human-modified edges. Both activities dropped significantly during the low season. Chapter Three concludes that the macadamia landscape is able to support a high bat species diversity, which is affected by seasonal differences, probably due to food availability. Natural areas are important to be maintained to support sensitive species, relying on clutter habitat and natural roosts. Habitat features as well as farm management can also influence arthropod communities. Visual observations took place each month for 20 minutes per tree. Observations were counted and identified to at least order level. Chapter Four concentrates on the impact of orchard edge habitat and insecticide treatment on honey bees (83% of Hymenoptera observations) abundance. Honey bees were the only taxon significantly responding to both variables, with higher abundance close to natural areas and increasing abundance with increasing time since pesticide application. Honey bees furthermore seemed to recover slightly quicker from population crashes after insecticide treatments at natural edges than they did at human-modified edges. Hymenoptera can be highly beneficial to macadamia farmers, as farmers are heavily relying on pollination by honey bees and parasitoids are known to feed on major macadamia pest insect species. These ecosystem service providers were mostly affected by habitat and management practices, which may compromise their ecosystem service provision. Although commercial bee hives are exposed throughout macadamia orchards, they do not appear to fully replace the pollination and biocontrol services provided by feral species, which can be enhanced through resource supplementation by patches of nearby natural vegetation. Chapter Five concludes with emphasizing the general importance of natural vegetation in landscape planning of agricultural areas. This study proved confidently that benefits through ecosystem services largely outweigh negative impacts of ecosystem disservices, both stemming from these natural areas. Through the inclusion, maintenance and restoration of such patches of natural vegetation, farmers can potentially increase the value and effectiveness of biocontrol by bats and birds or other services. Seasonality largely determined bat species diversity, and a general high species diversity was observed in the macadamia landscape, possibly due to a certain degree of landscape heterogeneity and high food availability. Activity patterns of clutter-edge and open-air feeding bats varied with edge habitat, while clutter-dependent bat species / NRF

This study aimed to investigate the spatial extent of Land-Use Land-Cover (LULC) change and the implications for ecosystem services in order to develop a sustainable land-use management framework for traditional authorities in South Africa. Effectively, this study undertook an insightful examination of the impacts that arise from policy decisions and practices, which unfortunately were found to be ineffective. The methodologies and approaches used in this study included both quantitative and qualitative techniques. The critical quantitative method employed in this research was the use of survey questionnaires to collect primary data. Qualitative approaches, such as one-on-one and key informant interviews, were used to triangulate the findings. Remote sensing and geographic information system (GIS) methods were used to investigate changes in LULC from 1990 to 2018 through the use of data obtained from the South African National Land-Cover project. Stochastic models were used to predict future LULC changes from 2018 to 2050. The Co$ting Nature Policy Support System was used to identify and undertake economic valuation of services provided by ecosystems. Statistical analysis using the Statistical Package for the Social Sciences was used to identify correlations and the reliability of the data, while graphs and tables were generated to identify patterns and lessons from the research. Between 1990 and 2018, significant changes in land cover were noticed for thickets and dense bush, woodlands, waterbodies, subsistence agriculture, and built-up areas. Woodlands changed by over 1 000 hectares (ha) per year, while thickets decreased by over 900 ha per year. Drivers of these changes include deforestation, among others. Future predictions for LULC revealed that between 2018 and 2050, almost 500 ha of woodlands would be lost to built-up areas. The aggregate value of the services flowing from ecosystems was found to be R9 509 044 608.00. A significant issue was that 90% of the traditional leaders interviewed could not positively respond to whether they knew the extent of the land they presided over, which raised questions regarding the effectiveness of their management systems. Recommendations were made in this study to address the limitations identified in the land-use management practices by adapting elements of the main theoretical frameworks, namely the Intergovernmental Science-Policy Platform on Biodiversity and Ecosystem Services framework; the Drivers, Pressures, States, Impacts and Responses framework; the sustainability theory; and the hierarchy of plans, into a new framework designed specifically for traditional leaders, titled the “Traditional leaders land-use decision support framework”. / Environmental Sciences / D. Phil. (Environmental Management)