Southern Ground-Hornbills Bucorvus leadbeateri are listed as Endangered in South Africa and there is a concerted effort to reverse their population decline. Understanding the factors affecting their movement patterns and how they interact with their habitat is useful to inform conservation and habitat management options for the species, to select sites for artificial nests and to identify suitable areas for reintroduction initiatives. In this study, I report daily and seasonal patterns of habitat use as well as patterns of roost site use of four Southern Ground-Hornbill groups in the Associated Private Nature Reserves, northeast South Africa, based on data from GPS-satellite tags collected over one year from October 2010 to September 2011. Home ranges varied from 5.9-10.3 km² and were larger in winter than during the summer breeding season. Daily travel distances were greater during the breeding season, when birds were constrained to forage close to their nest, and were lower in winter, when birds ranged more widely. Hourly travel distances were affected by time of day, season, air temperature and group. Birds travelled farthest per hour in the morning, decreasing in the afternoon in winter. However, in summer hourly travel distances were bimodally distributed, with a minimum during the middle of the day when ambient temperatures exceed 25⁰C. Acacia-dominated vegetation and riparian habitats were favoured disproportionately during the heat of the day in summer, presumably because they offer more shade than other habitats. The number of roost sites used per month decreased progressively throughout the Early Wet season (October-December) and was lowest during the Late Wet season (January-March) for three groups that bred successfully. Mean monthly nights per roost were highest for these groups in the Early Wet and Late Wet seasons, specifically over December and January, coinciding with the peak breeding period. Throughout the Early Wet season, all four groups frequently roosted in close proximity to the nest, with 54-83% of roosts being within 1 000 m of the nest. During the Wet season, riparian habitats were favoured for roosting by the three groups that bred successfully, while during the dry season, disturbed areas, combretum-dominated habitats and mopane-dominated habitats were used. I conclude that the optimal habitat configurations for ground-hornbills include a mosaic of habitat types, including open areas for foraging and dense trees for shade as well as adequate large trees for nesting and roosting, particularly in riparian habitats.

This thesis examines the ecology of rare Proteaceae of the Fynbos Biome. South Africa. The aim was to determine whether there are any unifying ecological parameters which might be significant for their conservation. The current status of all Cape Proteaceae was initially assessed. A total of 124 taxa were ascribed the I U C N status of recently extinct (3 taxa), endangered (33 taxa), vulnerable (29 taxa) and naturally rare (59 taxa). The distributions of rare taxa are characteristically small in size and range, 59 taxa occurring in only one or two populations and 63 taxa being restricted to a range of less than 5 km². Small nature reserves are proposed for the protection of such restricted taxa. An examination of the distribution of all the rare taxa shows exceptionally high concentrations in the Cape Town urban area and the lowlands north of Cape Town. This indicates the need for conservation considerations by urban and regional planners in the long term development of this area. Naturally rare taxa exhibit nodes (areas of high concentrations) over the centres of well-defined centres of endemism. Such areas merit conservation as representative and diverse habitats of fynbos plant species. Threats occurring at rare Proteaceae sites were recorded. An inappropriate fire interval and invasion by the Argentine ant. Iridomyrmex humilis, were shown to be the probable cause of rarity in Proteaceae exhibiting short lifespans and myrmecochorous seed stores. The monitoring of fire intervals at rare Proteaceae sites and the removal of the Argentine Ant are considered priorities for conservation management of rare Proteaceae.

As trends in urbanisation continue globally, there is a growing need to understand the impacts of urban development on wildlife. Whilst urban impacts on patterns of diversity and abundance of species have been well-studied, there remains a distinct lack of understanding around the impacts on ecological interactions. Different species have different levels of tolerance to urban disturbance, some species even appear to thrive in urban areas and make use of human-subsidised resources; but the physiological costs and trade-offs faced by urban-dwelling species are poorly understood. Given that their range in South Africa has only recently expanded into the human-dominated landscapes of the Western Cape, the Black Sparrowhawk (Accipiter melanoleucus) provides an excellent opportunity to explore some of these questions. In this study we explored how urbanisation may affect the health of this raptor on the Cape Peninsula, South Africa. If the health of this species is negatively influenced by levels of urbanisation, this might be driven by differences in diet and prey availability across the urban spectrum. Thus, we explored this potential mechanism by examining diet composition and assessing prey abundance within different territories and habitat types. The health of nestling Black Sparrowhawks was evaluated through their immune response (Heterophil/Lymphocyte ratio in white blood cell counts), body condition (based on morphometric measurements) and blood parasite infection (presence and abundance of Haemoproteus and Leucocytozoon). Trends in diet composition along a gradient of urban cover were determined through the analysis of prey remains collected in the immediate nest surroundings and differences in prey abundance were determined through point counts in different habitat types.

Grasshoppers are sensitive indicators of the state of disturbance of grassland ecosystems. This study examined the grasshopper communities inside a game reserve, comparing those found on frequently grazed areas with communities inside plots that exclude megaherbivores. The vegetation inside the protected plots was found to differ from the openly grazed areas in terms of grass height and aerial cover, but not in % greenness or richness of forb species. Grass species varied with locality rather than grazing impact. Total numbers of grasshoppers did not differ significantly between the two contrasting areas (100.2 in vs 93.5 out), however grasshopper species richness did, with the outside, short-grass plots having on average 17 different species, and the inside tall-grass plots a higher mean of 24.5. Grasshopper communities responded primarily to grass height and vegetation cover, but not to grass species or greenness of vegetation.

We hypothesized that rainwater had a different and distinct isotopic signal to groundwater, and thus plants growing on groundwater-fed areas would have a unique and distinct isotopic composition compared to plants growing on freely-drained soils. We studied two groundwater-fed areas and adjacent freely-drained areas in the Cape Point Nature Reserve, Cape Peninsula. We sampled groundwater and analyzed the isotopic composition and compared this to the isotope composition of rainfall. We found that δ¹⁸O and oD values of rainwater (δ¹⁸O = O%o to -5.6%o and oD = +11 %o to -22%o) overlapped the isotopic composition of groundwater (δ¹⁸O = -4.2%o to -4.5%o and oD = -11.9%o to -13.2%o). Thus isotopic analysis could not determine which areas were groundwater-fed and which were rain-fed. We also sampled xylem water from five species, namely Leucadendron laureolum and Metalasia muricata (both of which grew on both dry and wetland areas), Mimetes hirtus (wetland only), Erica labialis (dry area only), and Erica multumbellifera (wet area orily). Our results suggest that both of L. laureolum and Metalasia muricata had access to groundwater on both groundwater-fed sites and the free-drained site. E. multumbellifera at a groundwater-fed marsh was likely also using groundwater, however on the adjacent freely drained site E. labialis was likely using rainwater. M. hirtus had a much more positive isotopic composition (δ¹⁸O = -1.5 ±1.2%o and δD = 5.1 ± 18.0%o) than either groundwater or rainwater, indicating that it was using evaporated water, which suggests the species were very shallow rooted. Thus M. hirtus would be very sensitive to drying out of the soil. This study has demonstrated that stable water isotopes cannot clearly distinguish rainwater and groundwater and we may have use other methods such as thermal imagery, chemical analysis, and xylem pressure potentials to determine the degree of dependency of vegetation on groundwater.

The biophysical drivers of fire; ignition, fire weather conditions, fuel biomass, and flammability, differ in varying ecosystems. The rates of occurrence of these factors influence fire regimes. This study investigated the causes and consequences of a severe wildfire that swept through the Hluhluwe Game Reserve, KwaZulu-Natal South Africa in September 2008. This fire was an extreme event that seems only to have been possible due to the combination of circumstances that occurred in the days prior to and during the 14th/15th September 2008. The event was considered extreme because it burnt from savanna into thicket and forest patches, which is atypical of fires. The fire caused large structural change in tree demography, however, not much change in densities. The results of this study indicate that coupled weather conditions conducive to fire; namely low relative humidity, high temperatures and high wind speeds, occurred at the time of the fire. The synergistic influences of fire weather conditions and the state of available fuel caused this severe fire. The fire continued to bum for 48 hours before weather conditions relaxed and became less dangerous. These data suggest extreme fires such as the September 2008 event may be exploited by managers to reclaim former grasslands and savannas that have suffered the effects of bush encroachment and/or create open areas allowing grasslands to develop. However, severe fires run the risk of leading to the loss of forests and the biodiversity that they support. This study has identified the conditions likely to promote such severe fires.

Understanding the processes and mechanisms governing animal movement is a fundamental goal in ecology. Processes driving movement can occur across multiple spatiotemporal scales and have important consequences for the structure and dynamics of populations, communities and ecosystems. The study of movement provides insights into the ecological resources and habitats necessary for persistence of species and communities. It also provides a theoretical and applied basis from which to formulate informed conservation plans. Waterbirds in semiarid southern Africa are an ideal study group for understanding interactions between movement and environmental factors because they exhibit a wide range of movement strategies and are located within a landscape in which resources are characterised by high levels of spatiotemporal variability. Emphasis has been placed on understanding movement phenomena from individually-tracked animals, but cases which consider this approach in conjunction with traditional community ecology perspectives are rare. In this thesis I explored questions of movement in both individuals and communities, and argue that an integrated multi-scale approach is necessary to advance our broader understanding of movement in waterbirds. In the first part of the study I addressed an individual-level movement perspective. I used fine-scale telemetry data from 35 individually tracked Egyptian Geese Alopochenaegyptiaca and Red-billed Teal Anas erythrorhyncha with novel analytical techniques to explore questions of trade-offs in habitat selection, functional responses and whether movement responses to landscape resources are reactive or prescient. My findings suggested that, at the home-range scale, both forage optimisation and predation risk were limiting factors of movement and habitat selection of Egyptian Geese. I also showed for the first time that waterbirds exhibit functional responses in relation to changes in the availability of habitat types. I subsequently showed that the proximate drivers of waterfowl movement are the dynamics of rainfall and primary productivity. Egyptian Geese and Red-billed Teal were able to perceive and respond to temporal shifts in resource conditions prior to habitat patch occupation. This in turn suggested that their movements in semi-arid landscapes are underpinned by an intimate knowledge of the local environment and that waterfowl exhibit a complex behavioural movement strategy. In the second part of the study I used waterbird count data collected from wetlands in KwaZulu-Natal, South Africa, to address the community-level movement perspective.

Forests are valuable ecosystems to society but are greatly threatened by changing factors from habitat conversion to climate change. South Africa's only extent of indigenous forest is predicted to disappear within the next 30 years. Many challenges are currently faced when trying to detect and interpret directional changes in forests which results in an urgent need to understand any effects that these change factors have on forest ecosystems. We investigate evidence for change in the old growth Lilyvlei Nature Reserve by monitoring growth and dynamics at various levels by examining a 20 year record of tree growth and stand dynamics. Through the inclusion of biodiversity measures and ecologically important plant traits, changes in forest dynamics and growth are investigated. Results show no total change in biomass across the 20 year period, although an intensification of extreme climatic events and dynamics indices were recorded for the second period. Significant correlations were found between community diversity measures and forest growth. Trait variables showed insignificant correlations with forest growth and dynamics. These results suggest that the Knysna forest is controlled by climatic variables and that increased diversity within communities result in increased growth. It is believed that changes in the forest may be masked by compositional shifts of just a few dominant species. These results become important, particularly in the light of changing climatic, atmospheric and environmental changes that threaten global ecosystems in the time to come. However, considering the brief 20 year period observed in a forest where the average individual has a life span of over a century, the importance of long term monitoring becomes an important component in the understanding of forest ecosystems.

Coastal cities have impacted negatively on freshwater and marine ecosystems - primarily through habitat loss, fragmentation and pollution. Globally, it has been found that otter's dependence on these ecosystems exposes them to a myriad of threats, including loss of habitat, human-wildlife conflict and the bioaccumulation of toxic pollutants. The Cape clawless otter, Aonyx capensis, is the most widely distributed otter species in southern Africa and persists in human-modified habitats, including large cities. The Cape Peninsula provides a unique opportunity to study the impacts of urbanisation on otters as it presents a gradient from densely populated urban areas in the north (City of Cape Town) to sparsely populated areas interspersed with large expanses of natural habitat (Table Mountain National Park) in the south. In this thesis, I investigate the distribution, diet and threats to otters living on the Cape Peninsula. I use sign-based occupancy surveys to determine both broad and fine scale drivers of otter presence within the Peninsula's river systems and predicted that otters would avoid densely populated urban areas and rivers or sections thereof that are heavily transformed and polluted. I collected spraint from living otters and vibrissae from dead otters to investigate their diet. I predicted that otters would show an increased reliance on marine foods in areas where freshwater habitats were degraded in addition to seasonal variation in diet associated with the marked seasonal variation in rainfall and primary productivity typical of temperate Mediterranean ecosystems. I explored both immediate and long-term threats to otters by collating all records of conflict, injury and mortality reported over 5 years in addition to determining PCB levels from road-killed otters. Contrary to my predictions, otters did not avoid urban areas, and were more frequently detected in transformed lowland freshwater river systems close to Marine Protected Areas (MPA). Within rivers otters avoided the relatively pristine, yet unproductive, upper reaches of rivers as well as canalised sections and those with consistently high E.coli counts. I found that otters were feeding on both marine and freshwater prey in both polluted and nonpolluted systems. Where large, transformed lowland wetlands were in close proximity to MPAs, otter diet consisted largely of a combination of freshwater crabs and exotic fish from the polluted systems as well as marine fish and rock lobster from the relatively unpolluted MPA. Isotope results complemented the faecal analyses and confirmed that otters show significant variation in diet between seasons, sites and individuals, suggesting an opportunistic and generalist foraging strategy. Importantly, the dietary results reinforce the distribution model that otters rely heavily on the interface between coastal and lowland wetland and river ecosystems close to the MPA, for both foraging and breeding habitat. However, these are the areas that are transformed and heavily impacted by urban development, and therefore the area where otters would most likely be at risk. I developed a hotspot map of otter conflict across the Peninsula and found that the Peninsula otter population experiences low to moderate levels of conflict throughout most of their current range. High conflict areas are associated with optimal habitat that has been fragmented by canalisation and urban development. Road-killed otters showed signs of accumulation of PCBs in liver tissue suggesting that despite otters being adaptable generalists, their dependence on polluted freshwater systems may have long-term health impacts. Mitigating these threats is possible with improved urban planning, waste water treatment and education of the public. However the success of these approaches requires long-term monitoring which is unlikely to be prioritised by resource constrained conservation authorities. I thus explored whether the large citizen science community in Cape Town can be used to monitor the population. I used Maxent to model otter distribution using citizen reported sightings over 5 years and compared the results with the occupancy model outputs. The predicted Maxent distribution mirrored that provided by occupancy models, and highlighted further areas of suitable otter habitat and routes for dispersal. Together my findings suggest that Cape clawless otters, like many other meso-carnivores in South Africa and globally, display a remarkable ability to adapt to human-modified environments using the interface between degraded freshwater systems and the inshore region to feed on a diverse range of prey. Of concern are the moderate to high levels of conflict with people and dogs, vehicle accidents and the accumulation of toxins. Long-term monitoring of the population and the effect of proposed interventions can be achieved by creating a platform for citizen sightings to be recorded in perpetuity at low cost. This platform can also serve as tool for educating the public on the global challenges of conserving biodiversity within and adjacent to large cities.

Abalone farms are becoming more common along the coast of South Africa. The effects these farms are having on the coastal habitats they occupy are becoming of increasing concern due to the possible negative effects of the effluent discharged from the farm. This study was conducted along the coast of Jacobsbaai (32°58'22.09" Sand 17°53'10.56" E) and Mauritzbaai (32°58'50.75"S and 17°52'59.44"E) near Saldanha on the South African west coast, situated approximately 120 Km from Cape Town. The aim of this study is to determine if the outflow of effluent from the Jacobsbaai Sea Products (pty) Ltd, abalone farm in Jacobsbaai can be traced using stable carbon and nitrogen isotopes. Kelp (Ecklonia maxima) and blue mussels (Mytilus galloprovincialis) along the coast near the farm (in Jacobsbaai) were collected from 8 sites and used as tracers of the farms effluent. A ninth site was sample in another bay (Mauritzbaai) which functioned as our control. We hypothesised that if the kelp and mussels are taking up the nitrogen and carbon from the farm, the amount of nitrogen and carbon taken up would decrease with decreasing distance from the effluent outfall. Using this we could map the geographic distribution of the abalone effluent alone the coast to measure the extent of eutrophication due to abalone effluent. Our results suggest that the effluent from the farm is having a localized effect on both Ecklonia maxima and Mytilus galloprovincialis.