Major influences on biodiversity in general and on species distributions in particular have occurred on South African landscapes over time as a result of human activities and climate change. Landscapes in southern Africa are not only a product of changes in climate, but they are also a product of the changes brought about by people and their land management practices. In recent decades, bush encroachment has increased dramatically throughout the Savanna Biome of South Africa. Bush encroachment is a vegetation dynamic of global interest. It refers to the thickening or increase of woody vegetation (trees and shrubs) at the expense of grasses and/or herbaceous vegetation. In many areas across South Africa, systems that were once open grassland have become dense thickets dominated by trees and shrubs. Bush encroachment is driven by global factors like increasing carbon dioxide (CO2) levels in the atmosphere as well as local factors like grazing pressures and fire management practices. In Africa, it has long been recognized that there is an influence of vegetation patterns and structure on the distribution and abundance of birds. Changes in habitats due to an increase or decrease in woody cover has been linked to changes in bird diversity and community structure. This thesis looks at what impacts bush encroachment has had on bird distributions in eastern South Africa, with a special focus on the Kruger National Park as well as Palearctic migrants. Chapter 2 of the thesis develops a new quantitative method to relate bird distributions to woody cover. The algorithm is experimental, and an important aspect of the chapter is to investigate whether the results obtained coincide with expectations. The approach first estimates the proportion of the bird species in an area which occur in each woody cover category in a region. It also calculates the proportion of the area of the region which is in each woody cover category. It then uses Jacobs Index to provide an estimate of the extent of preference/avoidance of each species for each woody cover category, on a scale that runs from +1 (the entire population of the species is in that woody cover category) to –1 (none of the population of the species is in that woody cover category). The method is tested on a region consisting of three one-degree grid cells spanning Mpumalanga, Limpopo and Gauteng which include a diversity of woody cover categories, from grassland (no woody cover) to savanna to dense woodland. Chapter 3 applies this approach to the birds of the Kruger National Park, examining which species have distributions which can, and which cannot, be explained to a greater or lesser extent by a preference for a particular range of woody covers. Bird species are selected for close monitoring for ecological management purposes of the Kruger National Park. Chapter 4 considers a set of 10 bird species, all of which are migrants to South Africa from Eurasia, and occur to a large extent during their non-breeding period within the savanna biome of South Africa. For each species, the extent of increase in relative abundance between the first and second bird atlas projects in the region (SABAP1 and SABAP2) is estimated. In most cases, the change of relative abundance can be explained in terms of bush encroachment. The likely provenance of these species in Eurasia is considered; for most species, this area lies in southwestern Asia, i.e. the Ukraine and adjacent Russia. Changes in habitat in this region are also considered and help to improve the understanding of how the populations in South Africa have changed.

Global human population growth has been predicted to grow exponentially, to a point where it exceeds the capacity of available resources to sustain it. The consequences that such exponential increase will have on the environment has also been the focus of several research. The spatial pattern of human population has reveal uneven pattern of human population with the urban areas being subject of increased influx of human population from the rural areas in search of better economic factors. The United Nations in 2007 revealed that at least half of the world’s 6.6 billion human population was living in urban areas. This number is expected to increase to over 60% of the world's population by the end of 2050. Most of this population growth is occurring in developing countries. While the health, security and town planning amongst other consequences of this global explosion in urbanization have been well-studied and documented, the impact which urbanization is having and will have on the ecosystem and on biodiversity, especially at regional and local scales has remained an a field of knowledge that has continued to evolve especially given the variable patterns and drivers of urbanization in different regions of the world as well as the different environmental factors and biodiversity in these regions. Biodiversity monitoring has been shown to be crucial to conservation goals aimed at accessing the state and condition of biodiversity. The Second South African Bird Atlas Project (SABAP2) is a citizen science atlas project which commenced in 2007. Over a decade, SABAP2 has produced a rich source of data, capturing bird distributions in South Africa. This makes SABAP2 a powerful tool for monitoring observed changes in bird communities and by extension biodiversity through time. I examined the effect that urbanization is having on the avian biodiversity in South Africa, one of the most urbanized countries in Africa. My research was focused on the 576 pentads in the four one-degree grid cells (25S 27E, 25S 28E, 26S 27E and 26E 28E) centered on the Gauteng province, referred to as Greater Gauteng region. In addition to being very urbanized, Greater Gauteng is also the most populated area in the country, and is home to 30% of the country’s 51 million people. The region is the most atlased SABAP2 region in the country, with each pentad having a minimum of 11 full-protocol SABAP2 checklists. It thus provide opportunities for the development of tools to monitor the temporal dynamics of bird communities. The first chapter is the general introduction where I did an extensive literature review of the research subject and gave an overview of the data chapters that make up the thesis. In the second chapter, I examined spatial patterns of urbanization and avian biodiversity. I assess avian species composition in the urban and rural areas of Greater Gauteng. I categorized bird data from SABAP2 for Greater Gauteng Urban and Rural subgroups. The dataset for this chapter had 700 bird species. 644 showed no range preference for either urban or rural areas. Five species showed a preference for rural areas while 51 species showed a preference for urban areas. The higher species richness recorded in urban pentads highlights the often overlooked benefits of biodiversity conservation efforts in urban areas such as green spaces and parks, gardens and water bodies. This chapter highlights the need for conservation efforts to be targeted at birds and other biodiversity in urban spaces. It is also raises the need to further promote policies aimed at having conservation efforts incorporated into town planning. In the third and fourth chapters, I used data from SABAP2 to investigate how different protected areas such as Important Bird Areas (IBAs) are to their surrounding areas by demonstrating how different the Devon Grasslands (Chapter 3) and Suikerbosrand Nature Reserve (Chapter 4) IBAs are to their immediate surrounding areas in terms of avian species richness and assemblage. Atlas data from the pentads covering these two IBAs were compared with data from the surrounding pentads. Both IBAs stand out as having more bird species than their immediate surroundings. The simple yet effective method used in this chapters can be applied in identifying potential sites for biodiversity conservation. In the fifth chapter, using a variation of the Shannon-Weiner species diversity index which is known to reach an asymptote rapidly even while species richness keeps increasing, to investigate patterns of spatial distribution of species richness and proportional diversity in Greater Gauteng. The chapter provides insights into pentads with the richest bird communities and also provides a method which can be applied to citizen science data such as SABAP2 to discover areas where particular groups of species, such as waterbirds and threatened species, are concentrated in the region. The sixth chapter examines the relationship between reporting rates of birds and human population in Greater Gauteng. With Greater Gauteng being the most populated region in South Africa, it presented an ideal situation to investigate patterns of correlation between human population and the reporting rates of bird species in the region. Based on the results obtained, the species were grouped into 18 groups categorized by the relationship pattern revealed by species reporting rates and human population. The Seventh chapter follows a similar pattern with chapter six. However, chapter seven, examines patterns between a socio-economic index, mean income per person, and the reporting rates of birds in Greater Gauteng. The eighth chapter is the conclusion. It gives a synthesis of the thesis and presents the implications for conservation of avian biodiversity in Greater Gauteng. Overall, this thesis highlights the contribution of citizen science can make to research. It also makes for a strong case showing fundamental importance of large volumes of data such as SABAP2 data, and the useful information that can be harnessed from this data. The conservation-relevant studies in the chapters of this thesis are a result of the spatial distribution patterns of the avifauna revealed by SABAP2 data from Greater Gauteng. It showed how we can detect changes in species abundance, richness and composition in a pentad or in any area, a method we can extend further to detect when bird species are starting to decline or drop out of the species list for a pentad. The results reported in this thesis provides a rich field of study for future research, especially in the field of urban ecology.

Includes bibliographical references. / Understanding breeding systems and gene flow is important for the conservation of rare species. Despite this, they remain relatively unknown for Fynbos species. Erica occulta is a rare, dull, closed-flower Erica that is restricted to the few limestone cliffs near Groot Hagelkraal on the Southern Agulhas Plain, South Africa. Its restricted range and small population make it a vulnerable species and determining its breeding system and genetic diversity may inform the best method of conservation. The aims of this study were to determine the genetic diversity of E. occulta and how it varies with distance and to make inferences about the breeding system and gene flow based on genetic diversity measures. Due to the population’s small, isolated nature, it was predicted that the species would have a low number of alleles and low heterozygosity. Given its floral morphology it was predicted to be a self-fertilising plant which may manifest as a further loss of heterozygosity and strong genetic isolation by distance.

The age and origins of large earth mounds (locally termed "heuweltjies") is under debate, with ages being proposed for between 4500 and 30 000 years old, and hypotheses including zoogenic and geological origins of the mounds. The widely accepted hypothesis for the origins of the heuweltjies found in the South Western Cape of South Africa is that mounds are ancient termitaria, belonging to the southern harvester termite Microhodotermes viator. Our idea was to use rocks and stone layer profiles to determine approximate ages of the heuweltjies as well as to gather evidence to either support or refute the termite hypothesis. We excavated ten mounds in Clanwilliam, South Africa, to gather stone and rock profiles throughout the mounds, as well as digging a trench through a heuweltjie with bedrock as its matrix in order to determine the extent of the bedrock into the mound. No stone layers as such were found. What we did find was that the mounds are a whole order of magnitude greater in volume than large termite mounds made by the northern harvester termite Macrotermes in tropical regions, and many of the mounds contain large rocks both throughout the mound as well as on the surface. The trench cut through the mound on bedrock revealed the bedrock extending all the way to the centre of the mound, on the same level as the surface of the surrounding matrix. This evidence can be used to refute the termite hypothesis for the heuweltjies of Clanwilliam as it is virtually impossible for termites to move rocks up to 25 kg to the top of a mound, and mining soil from beneath a mound with bedrock at its base is an impossibility.

Kelp forests are some of the most productive coastal ecosystems in the world and provide numerous ecosystem goods and services. Where they occur, they play a key role in coastal ecology and local fisheries. In South Africa, Ecklonia maxima and Laminaria pallida form kelp forests which support diverse ecological communities. The species composition of these communities has been shown to be spatially variable along the South African coast, controlled by abiotic processes and species interactions. Despite their importance in the southern Benguela Current Large Marine Ecosystem (BCLME), large-scale research directed towards these habitats has largely waned over the past 30–40 years, prompting a renewed focus on these systems. Stable Isotope Analysis (SIA) is an indispensable tool for investigating food web characteristics, with particular focus on trophic structure and functioning. SIA can be used to understand the basal isotope variability in producers, determine the primary carbon sources of food webs, and investigate the spatial and temporal patterns in consumer isotope values and trophic niches. Unlike in other global kelp ecosystems, this methodology has not yet been applied to the kelp forests in South Africa. A study among eight geographically separate sites and two seasons highlights the natural variability of stable isotope values (δ13C and δ15N), at different spatial and temporal scales, within the two dominant kelp species. Additionally, stable isotope variability was investigated within and among different tissues within both kelp species. Within a kelp plant, δ13C values had a range of 1.65‰ for E. maxima and 1.52‰ for L. pallida. The δ15N values had a range of 3.75‰ for E. maxima and 4.21‰ for L. pallida. There were also consistent variability patterns along the length of a single frond in both species, for both isotopes. Among the localities, E. maxima and L. pallida were highly variable with ranges in δ13C (9.37‰ and 11.22‰) and δ15N (3.44‰ and 4.51‰) for the two species respectively. The δ13C values of L. pallida and E. maxima displayed a clear pattern coinciding with depth, particularly for L. pallida. Within-site variability was a major contributor to the overall spatial variability for both species. This provides further evidence for the importance of understanding basal variability of stable isotope values when determining the carbon sources of bottom-up controlled ecosystems. Variability in particulate organic matter (POM) composition is hypothesized to be related to upwelling processes in, and around, a typical west coast kelp forest. Various variables were used to characterise the composition, and determine the dynamics, of the POM along two transects (alongshore and off-shore) originating within a kelp forest. SIA was employed to estimate the contribution of kelp-derived detritus (KDD) to the POM present in the water column, using a unique approach for isolating phytoplankton and kelp end-member values. Under upwelling conditions, stable isotope analyses confirmed the dominance (>70%) of kelp detritus in POM samples, even at distances of 7.5 km off-shore. Under downwelling conditions, however, phytoplankton was dominant (>60%) along both transects. This study therefore highlights the importance of coastal processes such as upwelling for controlling the composition of POM in kelp forests, as well as illustrating how the natural variability in POM composition created by upwelling processes can be used to gather POM endmember isotope values. Three ecologically distinct kelp forest communities were investigated using a combination of SIA and community-wide niche metrics ('Layman metrics’ and Bayesian inferences). Three kelp forests, with differing community composition were selected, including a west coast kelp forest, a False Bay kelp forest and a lobster-invaded kelp forest located east of Cape Hangklip at Betty’s Bay. Temporal and spatial variability in stable isotope values was identified in producer and consumer stable isotope values, but was variable among species. Community-wide metrics showed clear seasonal patterns, but despite large differences in community structure, metrics were not vastly different among sites. Seasonal variability was the largest observable trend in metrics at all sites. Community niche areas showed a high degree of overlap (80–95%) further illustrating the similarity among sites and seasons. These findings are largely contrary to expected patterns from community composition data. Trophic position and isotope niche of the West Coast Rock Lobster (Jasus lalandii) was determined at three ecologically different kelp forest habitats, located in Marine Protected Areas in south-western South Africa. Temporal and spatial variability in trophic position and niche size were detected. The stable isotope niche and trophic position of lobsters at Betty’s Bay were markedly different from those at Oudekraal and Bordjiesrif, with the summer niche being distinct (0% overlap). Trophic position was lowest at the lobster-invaded Betty’s Bay (2.52) and highest at Bordjiesrif (3.16). Similarly, the isotope niche of Jasus lalandii was significantly constricted at Betty’s Bay compared to the other two study sites. Primarily, these results show that in the region where these lobsters have invaded, the trophic niche is considerably constricted. This likely to be a consequence of the higher densities of lobsters in this area. These findings highlight the influence of the lobster invasion on Jasus lalandii itself, adding to the already published ecological effects on the kelp forest ecosystem as a whole. These findings also highlight the differences in trophic niche of this species, despite the overall community niche showing little difference among sites. The findings of this thesis provide estimates of the variability in stable isotope values of kelp forest seaweeds and consumers, as well as addressing the importance of ocean processes such as upwelling in controlling POM composition in kelp forests. Additionally, the trophic niche of a key kelp forest predator was characterised in ecologically different kelp forests, providing evidence of the effect of community structure on the niche of this species. This is the first concerted research effort into the stable isotope ecology of South African kelp forests, providing an updated look at the ecology of these important ecosystems, and serves as a foundation for future studies of this type.

Includes abstract. / Includes bibliographical references. / Ecosystem nutrient availability depends on the balance between rates of nutrient inputs and losses. Nutrients may be lost through fire and displacement of ash, herbivory, leaching and volatilization. The main pathways through which nutrients may be acquired are weathering of rock and atmospheric deposition. Symbiotic and free-living diazotrophic bacteria and blue green algae also contribute N. In ecosystems with limited occurrence of N₂-fixation and occurring on low-nutrient bedrock, atmospheric deposition is the most significant source of nutrients. Nutrients from atmospheric deposition may be of natural or anthropogenic origin, and can be “wet-deposited” dissolved in precipitation and “dry-deposited” when aerosols settle out of the atmosphere onto plant and soil surfaces. Studies on nutrient cycling around the world suggest that nutrient deposition can provide substantial amounts of nutrients to coastal ecosystems, although mineral weathering of rocks can also a significant source. Limited prior work on deposition in coastal areas of South Africa suggests that nutrient deposition could be an important component of nutrient budgets in the Cape Floristic Region. The west coast of South Africa borders a section of the Atlantic Ocean that is highly productive and characterized by strong seasonal winds, rough waters and strong wave action. This area is home to the Strandveld vegetation, which grows on marine-derived soils. Based on this, I hypothesized that marine aerosol deposition is a significant source of nutrients for the vegetation in west coast South Africa. To test this hypothesis, I examined the spatial and temporal characteristics of atmospheric deposition as well as the climatic and ecological characteristics of the area.

The boundary between forest and fynbos vegetation in the Southern Cape of South Africa present a dynamic ecotone in which forest has previously been noted to have expanded into fynbos territory. Forest colonization of fynbos has been shown to be primarily a function of nutrients and light environments in the understory which may or may not be conducive for the growth of forest species. Additionally, it is understood that fire regime is the primary agent in determining forest/fynbos boundaries. Virgilia oroboides is a fast-growing leguminous tree confined to the margin between forest and fynbos. It has long been thought to facilitate the growth of forest species through the heavy shading and nutrient enrichment of soils; both of which facilitate the growth of forest seedlings. Furthermore, V. oroboides is predicted to reduce fuel loads and as a result prevent fire penetration of forests. These possible functions were tested by sampling soils, nutrient content of leaves and the shade cast by plants 18 months after a burn on a forest margin near Swellendam, Southern Cape. In addition, the effects of varying density of Virgilia on surrogates for fuel biomass, and forest seedling growth, were observed at Silvermine on the Cape Peninsula. Shaded environments produced by V. oroboides are shown to have detrimental effects on the growth of fynbos, while facilitating the growth of forests. V. oroboides had the highest soil nutrient enrichment potential of all species investigated and is therefore predicted to produce nutrient cycling processes conducive to the growth and regeneration of forests. Estimates of fuel in the understory of V. oroboides are predicted to reduce fynbos fire severity upon reaching the margin. These results suggest that Virgilia does indeed facilitate forest tree species more than light-demanding fynbos species so that its presence should result in greater stability or slow expansion of forests into fynbos. If supported by further work, this facilitative role of Virgilia has management implications for the conservation of fynbos and forests in the Southern Cape of South Africa.

Endozoochorous plant species that have obligate dispersal mutualisms with megafauna are at risk of extinction because of the unique conservation threats associated with large animals. This study examines Balanites maughamii which appears to have similar dispersal attributes to Sclerocarya birrea and Balanites wilsoniana, both of which are considered to be obligate elephant fruits. Here, I emphasize the traits that make B. maughamii suitable for elephant dispersal and examine the cues required for seed germination. Large, fleshy fruits with a single, exceptionally hard seed which requires a crushing force of > 2.0 kN to promote germination are characteristic of this species. Using scent analysis of ripe fruits, the range of volatile that are responsible for the pungent odour were identified as being those which should function in mammal-attraction. Finally I examine the potential for long-distance dispersal by elephants using a gut passage time experiment, which revealed that elephants retain seed for approximately 48 hours, equating to a 4-8 km dispersal distance. Using fruit and seed bait stations to determine which other animals interact with B. maughamii, I found that small mammals may play a significant role in seed predation and that there is little evidence to suggest extensive secondary dispersal by rodents. These finding allude to the vulnerability of the long-term maintenance of B. maughamii in the absence of elephants, who appear to be their only legitimate disperser.

Interest in commercial sea urchin aquaculture is growing worldwide. This is because sea urchins have good quality roe which is a delicacy in many countries. Since the quality of sea urchin roe is dependent on what the sea urchin eats, increased research is being carried out to understand sea urchin feeding preferences. Feeding preference is related to the attractiveness/palatability of the feed, which is thought to be based on its chemical composition. Since an unpalatable feed will result in the poor quality of the organism, evaluating the palatability of sea urchin artificial feeds is important. In South Africa, the sea urchin Tripneustes gratilla has been selected for aquaculture due to its fast growth and high quality roe. Although it is a generalist herbivore, various studies have shown that T. gratilla has a significant preference for the macroalgae Ulva. The purpose of this study is to assess what chemicals contained in Ulva cause this preference. Chemosensory trials involving Ulva and its chemical constituents, as well as wounded Ulva and an artificial feed, were carried out using a Y-shaped maze. Results indicate that T. gratilla are not deterred by DMS or acrylic acid as literature suggests. Since previous studies showed that these two chemicals deter urchins, this study demonstrates that there are species-specific relationships to chemicals found in algal material. Results also indicate that T. gratilla are deterred by wounded Ulva, ulvan and ethanol, but cannot differentiate between these constituents when the constituents are compared against each other or ethanolic extract, fresh Ulva and feed. Since the sea urchins are not consistently deterred by wounded Ulva, ulvan and ethanol when compared with other constituents, this indicates that T. gratilla are not very selective in what they are attracted to or deterred from.

This work continues the marine tradition of the Department of Zoology but attempts to open-up a new field by tackling for the first time the ecology of the shallow sea bed. False Bay is accessible and provides a large area that is virtually unaffected by rivers or the works of man. Its situation happens to be particularly interesting. It is the nearest part of the flat, submarine shelf called the Agulhas Bank, where there are valuable fishing grounds; and it lies at the boundary of two very different masses of water. Study of False Bay should, in addition, be of interest in throwing light on the contrast between the shallows of the sea and those of lagoons and estuaries. Field work started with dredging from hired fishing boats. It was soon obvious that one of the most important regions of the sea was escaping notice, that above the shallowest depth in which dredges could be worked and below low tide level. This region could only be studied by diving and so a programme of diving with frogman kit was started. Dredging and diving meant considerable training in non-academic techniques not only in the field but in designing new gear (eg. dredges, underwater camera and underwater electronic flash).