Verbal accounts, supported by limited ground-based and satellite images, reveal decreasing riparian woodland and a loss of large trees along the rivers of the Kruger National Park (KNP) over the last century. These habitats occupy a tiny fraction of the park’s surface area but fulfil critical ecological functions and provide several ecosystem services. Little research has, however, focused on riparian woodland dynamics across the park. Here a multi-decadal time-series analysis of riparian woodland extent was conducted to identify trends in extent and possible drivers of riparian woody vegetation change. Aerial and satellite imagery (1936 to 2018) was used to measure changes in the extent of riparian woodland tree cover for 18 approx. 10 km long sites along five perennial and nine non-perennial rivers in KNP. This change was compared in a multivariate time-series with river flow and rainfall data from nearby gauging and weather stations, respectively. Particular attention was paid to cumulative flow effects, as well as the frequency and magnitude of large infrequent disturbances (LIDs) such as droughts and floods, which regulate the depth of the water table and may manifest as a physical disturbance. Tree cover fluctuated over the time period and the trajectory of change varied between sites. Most sites (n=11) experienced a decline in overall tree cover over the period while 14 showed a downward trend, six significantly. Overall tree cover increased at six sites, three of which showed an increasing trend (one significantly). There tended to be proportionately higher tree cover loss per year at sites with higher median tree cover. It appears that tree cover decreased substantially at a number of sites following the mega-flood event of 2000 and subsequent large floods over the last decade. It was not possible to generalise responses for the different sites, no doubt because of varying geology, flow regimes and vegetation characteristics at each site, resulting in differing responses to aspects of river flow and rainfall. Peak flow and maximum rainfall events, however, were the strongest significant association with decreases in riparian tree cover, indicating that floods are potentially the biggest drivers of tree loss. Flow variability and cumulative rainfall appear to significantly influence woodland expansion. The initial findings from this study should prompt increased attention to riparian habitats through fine-scale, detailed work aimed at further understanding the dynamics of these systems and determining thresholds for conservation concern in an attempt to ensure persistence of these important ecosystems.

Includes bibliographical references. / Urbanization, a fast growing and destructive human land use, causes local extinctions, biotic homogenization and fragmentation of natural habitats. Understanding how the nature of the urban matrix affects the species residing within a city’s fragmented habitats is an important founding component of urban conservation. This study investigated the influence that the urban matrix, as well as patch isolation, size and quality, had on weaver bird movement into and out of wetland sites in Cape Town, a growing city within a global biodiversity hotspot. Weaver bird movement data from 42 wetland sites were obtained through a long term mark - release - recapture project. Distance - based linear models revealed that site proximity was important as a predictor of weaver movement into and out of sites, while the site variables ( wetland size, bird abundance and weaver colony size) had limited and inconclusive influence. Once the variation explained by the proximity and site variables had been accounted for, the composition of the urban matrix and the presence of rivers as potential movement corridors (measured at three spatial scales) had little influence on weaver movement. The finding that proximity (or site isolation) influences weaver movement has important implications for maintaining current landscape connectivity. Habitat isolation, resulting from further habitat removal or destruction, could be expected to reduce movements of weavers, and potentially other species, among patches of favourable habitat. Weavers are robust, vagile birds that do well in the presence of humans and may not be highly sensitive to the nature of the urban matrix. Research into how other, less resilient and vagile species respond to the degree of urbanization in the matrix between wetlands would contribute further to our knowledge of urban biodiversity in this global biodiversity hotspot.

Adaptive genetic variability is vital to long-term species survival, as it presents the potential for evolutionary adaptive responses to environmental change. Genes of the Major Histocompatibility Complex (MHC) trigger the vertebrate adaptive immune response to pathogens through the recognition and presentation of foreign peptides, and thus provide an informative genetic marker for studying the adaptive potential of species with respect to disease. MHC class I loci mediate the immune system's recognition of intracellular pathogens, including protozoans, viruses and bacteria, and the high levels of genetic diversity reported at these loci is thought to be primarily the result of pathogen-mediated selection. Although variation within a number of MHC genes has been described in many felid species, the relationship between MHC diversity and pathogen prevalence within wild felid populations has rarely been examined previously. Because many wild felids are highly sensitive to the effects of anthropogenic-induced environmental changes, evolutionarily relevant adaptive genetic variation is particularly important for their conservation. This study explored the relationship between MHC class I exon 2 and exon 3 diversity and a number of tick-borne blood pathogens in a population of caracal (Caracal caracal) in the urban landscape of Cape Town, South Africa. The characterization of MHC class I diversity identified 38 and 45 unique, putatively functional alleles in the population, from exon 2 and exon 3 respectively. The influence of MHC allelic diversity and specific alleles, together with sex, age class, home range size and urban cover within home range, on individual pathogen prevalence was assessed using Generalized Linear Models. A positive relationship between Babesia infection and multiple measures of MHC diversity, including nucleotide diversity, average number of nucleotide differences and number of exon 3 alleles, was detected. Additionally, the presence of two specific alleles was significantly correlated with Babesia infection. High levels of infection by a number of tick-borne pathogens were observed in the population, suggesting environmental factors also have an important influence on pathogen prevalence. However, home range and urban cover within home range, as well as sex and age, were not significantly associated with Babesia infection. This study provides the first assessment of the relationship between MHC class I adaptive genetic variation and individual pathogen prevalence in caracal.

No description available.

Includes bibliographical references. / Disease outbreaks in both domestic and wild systems in recent years indicate the increasing potential for disease spill-over of generalist pathogens between domestic and wild species. Events of this nature are of considerable threat to rare or endangered species, while also being of significant economic concern for the farming industry. Understanding how disease moves within and between these contrasting systems is vital to prevent large-scale, multi-system epidemics in the future. This study utilized network analysis to understand how the movement of ostrich stock between farm locations in the Western Cape, South Africa may have contributed an epidemic outbreak of Highly Pathogenic Avian Influenza (HPAI) within the ostrich industry in 2011.

Includes abstract. / Includes bibliographical references. / Long-term monitoring of changes in species abundance and community composition within marine protected areas (MPAs) is essential to assess whether conservation goals are being reached. The costs, logistics and sampling biases inherent to traditional monitoring methods limit sustainable monitoring in all MPAs along the South African coastline. Baited remote underwater video (BRUV) technology offers non-extractive monitoring with lower labour and cost requirements, whilst eliminating inter-observer variability and increasing statistical power. Species richness and relative abundance were assessed employing BRUV technology in the Stilbaai MPA in the Western Cape, South Africa. ... [The] result corroborated findings from traditional underwater visual census (UVC) and controlled angling surveys, confirming BRUV technology as a sound monitoring tool.

Includes bibliographical references. / African elephant (Loxodonta africana) range and numbers have declined over the past century as a result of increasing human populations, agricultural development, and illegal hunting for ivory. Understanding the dynamics of wide-ranging animal migrations is important for the conservation of these species and their habitats in a rapidly changing world. The study of movement has greatly advanced in the past few decades and novel approaches for characterizing and interpreting complex movement data, predominantly collected through telemetry, have emerged. The Zakouma National Park elephant population has declined dramatically over the past ten years due to rampant ivory poaching. Several elephants in this population were fitted with satellite collars between 2011 and 2012. The telemetry data collected from these animals, in combination with data collected from the same population ten years earlier, provides spatiotemporal movement data from before and after a period of severe poaching. Broad scale shifts in seasonal movement patterns between these two time frames were explored. Movement behaviour was analysed at a finer spatial and temporal scale by comparing the rate of movement within different areas, during the day and night. A spatially-explicit approach for characterizing movement behaviour within discrete grid squares was used to identify seasonal patterns in the distribution of movement behaviour indicative of stress. After ten years of highly stressful poaching conditions, extensive seasonal migrations to the north and west of Zakouma National Park persist. At a finer spatial scale movement behaviour indicative of human-induced stress has emerged, especially in the areas utilized to the north of the park. Elephant behaviour to the north of Zakouma, particularly in response to the main road intersecting this area, suggests that barriers to dispersal may soon arise. Furthermore, the absence of migration in the wet season of 2013, after the data cut-off point for this study, suggests that the migration behaviour of this elephant population may be changing. In light of these findings, land-use planning in this area should be carried out in order to realign the boundaries of protected areas to include important elephant dispersal areas.

South Africa has 3650 km of coastline that spans the boundary between the Atlantic and Indo-Pacific biomes. The coastal waters boast a remarkable array of biological diversity and high levels of species abundance and endemism. Currently around 23% of the coastline is formally protected via marine protected areas (MPA) with 9% enforced as no-take zones. Even with this relatively high level of protection (as compared to other nations globally) the MPA network is still relatively sparse with protected areas that are on average ~110 km apart and unevenly distributed with the majority of MPAs situated along the species-rich east coast. This has led to concerns that the current MPA network is not protecting a representative sample of the genetic diversity among marine species nor is it sufficiently genetically connected via dispersal and gene flow to ensure their long-term persistence. To test a number of questions regarding the distribution of genetic diversity and degree of population genetic structuring along the South African coast we analyzed mitochondrial DNA sequence data for 10 sessile rocky-shore species and one reef-fish that represent three distinct life history strategies. We find that the distribution of genetic diversity across the South African coastline closely mirrors the distribution of species richness, increasing from west to east. We also find similar levels of population genetic structure among brooders, broadcast spawners and live-bearers, demonstrating that life histories are a poor predictor of genetic connectivity for South African marine species. Finally, we find that estimates of effective dispersal distance for taxa from each of the life history categories are low (~0.5-1.5 km per generation) suggesting that populations within MPAs are reliant on populations in unprotected areas via a steppingstone model of genetic connectivity. In light of these findings, we discuss a number of recommendations to enhance the role of the existing South African MPA network and echo previous calls for the establishment of protected areas along the west coast.

The Western Cape population of Blue Cranes (Anthropoides paradiseus) is of great importance as the largest and most stable population throughout its range. This species is strongly associated with agricultural lands in the Western Cape, and therefore may come into conflict with farmers who perceive them as damaging to crops. Blue Cranes are suspected to be locally nomadic, but little information has been collated to support this and they are also relatively understudied in terms of demographic parameters. This project investigates the viability of the Blue Crane population in three ways: exploring farmer attitudes towards cranes in two regions of the Western Cape (Swartland and Overberg) using 40 semi-structured interviews, generating estimates of survival using Mark-Recapture methods and exploring movement patterns using a long-term data set of resightings of marked individuals. These three components all add important aspects to the overarch ing goal of achieving a better understanding of threats to Blue Cranes in the Western Cape, and thus the population's long-term viability. Perceptions of cranes differed widely between regions: farmers in the Swartland perceived cranes to be particularly damaging to the feed crop sweet lupin (65% of farmers reported some level of damage by cranes), and 40% of these farmers perceived cranes as more problematic than other common bird pests. Farmers in the Overberg did not perceive cranes as highly damaging, although there was concern about cranes eating feed at sheep troughs. Survival was age-structured: individuals in their first year had a survival of 0.6, those in their second and third years that of 0.87 and adult individuals (4+) that of 0.72. The adult survival estimate is suspected to be underestimated due to ring loss. Resightings of colour-ringed cranes suggest that movements in the Western Cape were localized, with an average displacement of 24.6 km from their natal point. Only 3.8% of marked individuals were resighted in both the Overberg and the Swartland regions, indicating that movement between these regions was low and regional fidelity was high. There was significant movement within the Overberg however, and 90% of movements of >10 km were made within this region. This species therefore appears to be resident to locally nomadic in nature. Evidence for natal philopatry was also found: 57% of adults returned at least once to the area where they were ringed as juveniles. These results highlight the need for location-specific management solutions to crop-damage by cranes, and contribute to the understanding of basic demographics for this vulnerable species.

Includes bibliographical references. / Alien species are the second most important cause for the loss in biodiversity globally, after habitat destruction. Marine alien species are transferred across the globe through various vectors, including ballast water, hull fouling, aquaculture facilities and the aquarium and pet trade. Ballast water has previously been considered as the primary vector of alien species transfer. However, fouling is becoming widely recognised as an important vector for the transfer of marine alien species both internationally, as well as in South Africa, where it has been reported to contribute 48% of marine species introductions. The objectives of this study were to document alien species from fouling assemblages in six South African harbours (St Helena Bay, Saldanha Bay, Table Bay, Hout Bay, Gansbaai and Mossel Bay) and to use the data collected to identify factors (such as vectors and other harbour characteristics and activities), that could be used by management authorities to target harbours upon which to focus monitoring efforts. This was done by taking subtidal scrape samples and visual samples from harbour walls and pillars. The prioritisation of harbours was obtained through the use of regression tree models utilising CART (Classification and Regression Trees).