Browsing by Subject "Biological Sciences"

Now showing 1 - 20 of 223
  • Thumbnail Image
    Item
    Open Access
    A baited remote underwater video survey of the Goukamma Marine Protected Area's ichthyofauna and a subsequent community structure comparison with the Betty's Bay, Stilbaai, and Tsitsikamma Marine Protected Areas
    (2020) Dando, Jackson Willy; Attwood, Colin
    The Goukamma Marine Protected Area (GMPA) along the South African south coast has been in existence since 1990. The MPA encompasses 40.2 km2 of subtidal ocean, 76% of which is made up of sandy substrata and the remainder of which is made of rocky reefs. The imbalance in protected habitat type ratios prompted a proposal for an extension of the MPA's seaward boundary, referred to as the new no-take zone (NNTZ), and a restructuring of its eastern boundary, referred to as the new exploited zone (NEZ). The proposed boundary changes would increase the amount of protected reef by 53% and the overall size of the MPA by 38%. Goukamma has been surveyed using controlled angling surveys (CAS) and underwater visual census (UVC) but has yet to be surveyed using baited remote underwater video (BRUV). I collected and analysed mono-BRUV data over five years to determine patterns in fish community structure in Goukamma and compare it to the pre-existing CAS and UVC data. BRUVs are less invasive and more robust than the other two survey methods and have the potential to become the predominant method of surveying ichthyofaunal communities in South Africa. This work is therefore also intended as a baseline BRUV survey. BRUVs were deployed in Goukamma from 2013 to 2017. The survey produced 328 successful deployment records between 5-41.5 m across reef and sand sites. Date, site coordinates, depth, habitat type, protection zone were used as variables to explain patterns in the fish community data. Fish abundances were recorded using the MaxN metric. MaxN counts were recorded at the instance when the highest number of individuals of each species were present in a single video frame. The deployment records were converted into a single data frame and analysed using the RStudio integrated design environment. Ariids, scyliorhinids, serranids, sparids, and triakids were the most well represented ichthyofaunal families in Goukamma. Boopsoidea inornata, Cheimerius nufar, Chrysoblephus laticeps, Galeichthys feliceps, Mustelus mustelus, Pachymetopon aeneum, Poroderma africanum, Poroderma pantherinum, and Spondyliosoma emarginatum were the most frequently observed species throughout the MPA. Habitat type was identified as the primary determinant of diversity and abundance in the GMPA using multifactor analysis of variance (ANOVA) tests (species richness: F = 191.155, P < 0.001; relative abundance: F = 96.111, P < 0.001) and Wil coxon signed rank tests (Shannon-Wiener: W = 21 102, P < 0.001; Simpson: W = 18 553, P = 4.85x10-10). The reef sites supported a higher species richness and abundance than sandy sites throughout the MPA (Tukey: q = -4.41, P < 0.001 and q = -2.12, P < 0.001, respectively). Diversity and abundance were correlated with each protection zone's predominant habitat type. Exploited zones had significantly higher diversity and abundance than protected zones as a result of the imbalance in Goukamma's protected habitat type ratio (species richness: F = 27.740, P = 7.65x10-16; abundance: F = 10.438, P = 1.51x10-6 ; Shannon-Wiener: W = 17 314, P = 4.58x10-6 ; Simpson: W = 15 896, P = 3.42x10-3 ). The NNTZ had significantly higher species richness and abundance than the NEZ (Tukey: q = 3.07, P < 0.001 and Tukey: q = 1.48, P < 0.001, respectively). The proposed changes will therefore substantially boost diversity and abundance of protected fishes in Goukamma. BRUV samples in Goukamma recorded an overall higher species richness and abundance of sparids, chondrichthyans, and other reef-associated species than CAS and UVC samples. Over 90% more chondrichthyans were recorded in the BRUV samples than by the other two methods. BRUVs are therefore considered to be a suitable replacement for CAS and UVC surveys for the monitoring of South Africa's shallow subtidal ichthyofauna. BRUV data from Betty's Bay, Stilbaai, and Tsitsikamma were available for comparison with the Goukamma data, allowing for an extensive analysis of the south coast's ichthyofaunal communities. A combined data frame of 466 successful BRUV deployments from the four study areas was created. Multi-factor ANOVA tests indicated that location (F = 27.1, P = 1.00x103 ), depth zone (F = 17.4, P = 1.00x103 ), protection status (F = 23.1, P = 1.00x103 ), and habitat type (F = 91.8, P = 1.00x103 ) were all significant in determining community structure among the study areas. Reef sites had higher species richness and abundance than sand sites and species richness and abundance decreased from east to west along the south coast according to subtropical subtraction. However, the presence of an additional habitat type in Betty's Bay, namely kelp forests, resulted in it having a higher species richness and abundance than Stilbaai to the east. Betty's Bay's community structure was the least similar to the other three study areas as a result of the localised kelp forests in and around the MPA. These kelp forests shifted Betty's Bay's community structure away from the sparid-dominance observed in Stilbaai, Goukamma, and Tsitsikamma and towards a carangid- and scyliorhinid-dominance. However, cold-water associated sparids such as Pterogymnus laniarius were more abundant in Betty's Bay than the other study areas. Almost 80% of the species recorded among the study areas were represented in two or more of the four MPAs, indicating a good degree of redundancy of protection along the south coast within the depth ranges sampled. These data suggest that the Cape south coast is adequately protected from the perspective of fish representation. Review of the De Hoop, Sardinia Bay, and Bird Island MPAs should be conducted to further examine complementarity and redundancy of protection along South Africa's south coast.
  • No Thumbnail Available
    Item
    Open Access
    A contribution to understanding the primary moult of birds
    (2023) Scott, Tanya; Underhill, Leslie; Erni Birgit
    Moult is an essential component of the annual cycle of birds, along with breeding, and, if migratory, migration. The leading statistical model for the analysis of primary moult is the Underhill-Zucchini moult model. It estimates moult parameters (duration, the mean start date and its standard deviation). An objective of this thesis was to assemble all published results which had used the UnderhillZucchini moult and to examine the resulting database for patterns in the timing, duration and synchronisation of moult. A total of 242 analyses of 136 species were found. The initial analysis of the database generated two observations: firstly, that latitude played an important role in determining patterns of primary moult, and, secondly, that the available data were latitudinally, spatially across continents, and taxonomically not representative. These gaps in knowledge were too large to rectify within the confines of a single PhD thesis. The preliminary chapters of the thesis added 18 analyses, for 11 species, and have made a contribution towards representativity. Two of the preliminary chapters make substantive contributions to the methods of undertaking statistical studies of moult. A new moult index, the Relative Duration Index, is introduced. Moult studies are frequently undertaken by combining datasets collected in different years (or different places). This thesis contains the first practical application of a proposed method to identify which of the individual datasets have a large influence of the estimates of the moult parameters. Nine analyses extended the latitudinal range of studies southwards and were based on datasets assembled in New Zealand. These are the first analyses of primary moult using the UnderhillZucchini moult model from this country. For seven species, this thesis undertakes moult studies of a species in both its original country, the United Kingdom, and the country into which it had been introduced, New Zealand. These are the first analyses of primary moult of this kind. Preliminary patterns of moult strategies are discussed. Critical gaps in our understanding of primary moult are identified. Priorities for fieldwork and analysis are described in order that key gaps are filled and a global review of primary moult can be undertaken.
  • Thumbnail Image
    Item
    Open Access
    A mechanistic approach to understanding the colour polymorphism in black sparrowhawks (Accipiter melanoleucus)
    (2021) Nebel, Carina; Amar, Arjun; Sumasgutner, Petra
    Species that exhibit large variation in phenotypic traits, are commonly considered to have a stronger evolutionary potential. However, how they are capable to maintain polymorphisms remains a fundamental problem in evolutionary biology. Colour polymorphic species provide ideal study systems to explore the processes that lead to variation maintenance. About 3.5% of all bird species are colour polymorphic, but it is an especially common phenomenon in Accipitridae (22%), which indicates that it has an adaptive function in this bird group and makes them ideal model systems to study evolutionary processes. The black sparrowhawk (Accipiter melanoleucus) occurs in two discrete colour variants: dark and light. The two morphs differ in the expression of white and black feathers on the breast, belly and underwing coverts. The morph has been associated with ambient light-dependent foraging success and activity behaviour: Dark morphs forage more and have higher foraging success under low light conditions whereas light morphs forage independently of light levels but are better foragers under bright light conditions. This is hypothesized to be due to a crypsis advantage for the morphs under these conditions. During the winter breeding period, the predominating low light conditions on the Cape Peninsula (Western Cape, South Africa) could create an advantage for dark morphs, which should result in their higher survival and higher breeding success. However, this is not the case and only when the two morphs come together to breed, there is a fitness difference: mixed-morph pairs (that consist of a dark and a light morph) have higher breeding success than like-morph pairs (that consist of the same morph) and offspring of mixed-morphs have higher survival rates. This higher success of mixed-morph pairs is hypothesized to be due to emergent pair-level properties with the two morphs being able to expand the hunting niche as a pair. This ‘complementarity hypothesis' is based on previous research conducted on the study system. The aim of this PhD was to explore the mechanistic background of colour polymorphism maintenance in the black sparrowhawk. (i) I performed an experiment in which I test whether there is a morph- and ambient light-dependent crypsis advantage in the black sparrowhawk. I measure the reaction time of feral pigeons towards a simulated hawk attack but did not find indication of such an effect as pigeons reacted the same towards the two morphs. In line with the complementarity hypothesis, (ii) I found that mixed-morph parents provide food more consistently to the nest than like-morph parents. This results in a more predictable food supply for nestlings and buffers against long periods of malnourishment. However, (iii) I was unable to determine the mechanistic link between food supply and higher survival: nestlings of mixedand like-morph pairs had the same levels of innate immune function. Thus, an improved innate immune function in nestlings of mixed-morph pairs is unlikely associated with their higher survival rates. (iv) I performed individual-based model simulations which incorporate multiple key fitness parameters and found that complementarity – in combination with morphdependent seasonality-associated fitness effects – explains the stable colour morph equilibrium in this population. I conclude that emergent pair-level properties which arise due to the complementary nature of the two morphs play an important role in maintaining polymorphism in this species. Complementarity might not only be restricted to colour polymorphic species but could be present in other polymorphic traits that allow parents to behaviourally complement each other when raising their young.
  • Thumbnail Image
    Item
    Open Access
  • No Thumbnail Available
    Item
    Open Access
    A taxonomic revision of the squat lobsters (Anomura: Chirostyloidea and Galatheoidea) of South Africa
    (2025) Botha, Thomas; Griffiths, Charles
    South African squat lobsters (Chirostyloidea and Galatheoidea) are a severely understudied group. They have never been the subject of a formal monographic study and the last additions to the fauna were made almost 20 years ago. Thus, a revision of these superfamilies is long overdue. This study relied on examination of historic material housed in the Natural History Collections of the Iziko South African Museum, as well as ad-hoc sampling to carry out morphological analyses and molecular barcoding where possible. These methods were used to add to, revise and update the regional fauna. The thesis is presented in the form of six chapters, four of which each revise one of the families of squat lobsters present in South Africa and provide an illustrated key to the known regional species in that family. These chapters are supported by an introduction and synthesis chapter and a reference list. The first revisory chapter covers the Chirostyloidea (Chapter 2), and includes redescriptions of four species, reports on one species (and genus) from South Africa for the first time and expands the distribution and depth range for Uroptychus simiae. The third chapter describes two new species of Galatheidae, adds three species (and a genus) to the regional fauna list and resolves several dubious records. Galathea labidolepta Stimpson, 1858 is also redescribed and a neotype assigned, as the holotype was lost in the 1871 Chicago fire. The Munididae are revised in the fourth chapter, wherein two new species are described, seven species (four of which are new generic records for South Africa) are added as new distribution records, while five species are redescribed, two of which were previously only known from brief descriptions. The fifth chapter revises the deep-dwelling Munidopsidae and redescribes two species. This study marks the largest ever contribution to the South African squat lobster fauna with 15 species being added, comprising four new species and 11 new distribution records, while two species are removed due to synonymy. This increases the number of South African squat lobster species from 32 to 45. In the process, six genera are also reported on for the first time. Illustrated redescriptions are given for 12 species that were known from outdated or inadequate descriptions. Where possible, DNA sequences were produced as part of an integrative approach to delineate species, and this is the first study to provide such genetic data on South African squat lobsters, with 28 barcodes provided for eight species. Though this study adds valuable knowledge to this important group, it also highlights the need for more rigorous sampling, as more than half of the species are known from just a single sampling event and almost certainly more species remain to be discovered from the region.
  • Thumbnail Image
    Item
    Open Access
    Acoustic investigation of the shoaling dynamics of sardine sardinops sagax populations : implications for acoustic surveys
    (1997) Coetzee, Janet Claire; Field, John G; Champion, Harold
    The total allowable catches (TA Cs) of both anchovy and sardine in South Africa are determined by assessment methods which rely on absolute estimates of recruitment and spawner biomass each year. The validity of the TAC set for each species each year is heavily dependent on the accuracy and precision of the biomass estimates. The method used for both anchovy and sardine has been acoustic surveys on a random stratified grid covering the known range of the species. These surveys have, however, been designed according to the life history patterns of anchovy, which result in a particular occupation of the pelagic environment. This study investigated the shoaling dynamics of sardine at various scales in both the horizontal and vertical dimension in an attempt to ascertain whether these surveys are adequate to characterise the occupation of space by sardine. Hydro-acoustic data collected during 5 meso-scale surveys in 1994 and 1995 was used to study the shoaling dynamics of sardine, Sardinops sagax on the western Agulhas Bank of South Africa. Results of this study indicate the presence of spatial structure in sardine aggregations at both the shoal and meso-scale level.
  • No Thumbnail Available
    Item
    Open Access
    Acoustic manipulation by brood-parasitic honeyguide
    (2025) Blair, Cameron; Spottiswoode, Claire
    Brood-parasitic birds lay their eggs in the nests of other species (‘hosts'), and their chicks encounter the challenge of acquiring sufficient food from host parents. Many parasitic chicks require more food than host chicks, and so need to elicit high rates of provisioning through their begging displays. Greater (Indicator indicator) and lesser (I. minor) honeyguides are closely related brood parasites that require more food than host chicks, whom they kill soon after hatching. Previous research demonstrated that both species receive the same amount of food as an entire brood of hosts (little bee-eaters Merops pusillus and black-collared barbets Lybius torquatus respectively), and that their begging calls influence provisioning by host parents. However, the acoustic mechanisms facilitating this phenomenon remained unknown. In this dissertation, I first test whether greater and lesser honeyguides mimic the begging calls of their respective hosts and quantify the accuracy of acoustic mimicry. I then test the hypothesis that honeyguides elicit increased provisioning from host parents by mimicking an entire brood of host chicks. I found that while both honeyguides mimic their host's nestlings, they do not specifically replicate the sound of a brood rather than a single chick. Finally, I test the hypothesis that non-linear vocal phenomena (NLP) in honeyguide begging calls could facilitate a supernormal stimulus particularly when the pitch and temporal structure of begging calls is constrained (such as by selection for mimicry). NLP are known to indicate high-arousal states in other ecological settings, inspiring the hypothesis that elevated NLP could function to increase provisioning to parasitic chicks beyond levels individual host nestlings would receive. I found that greater honeyguides tended towards higher perceived roughness (a form of NLP) than single bee-eater nestlings, but not significantly so. Lesser honeyguide begging calls tended towards higher amplitude modulation frequencies and perceived roughness than single barbet nestlings, but this too was not significant. Since both these features are associated with the amplitude envelope, I conducted a playback experiment at black-collared barbet nests in the field in Zambia to test whether the shape of the amplitude envelope of honeyguide begging calls increases host provisioning. This cautiously suggested it did not, at least at the levels manipulated here. Together, these results show that honeyguides mimic the pitch and temporal structure of their primary hosts' begging calls but do not mimic a brood to increase host provisioning, contrary to previous suggestions. The results are inconclusive as to whether NLP could be exploited by honeyguides to increase host provisioning. NLP should be further examined as a potential signalling mechanism that could be exploited by brood parasites to elevate the level of provisioning by host parents. More generally, the signalling function of NLP in bird begging calls, which had not been examined prior to this dissertation, deserves further investigation.
  • Thumbnail Image
    Item
    Open Access
    Age, growth and per-recruit assessment of the Saldanha and Langebaan stock of Chelon richardsonii
    (2018) Horton, Matt; Kerwath, Sven; Parker, Denham
    Chelon richardsonii are omnivorous, particle feeders found specifically within inshore and estuarine habitats on the west and south coast of South Africa. They are the primary target of the gillnet and beach-seine fishery in this region. Despite being managed through a multifaceted approach of gear restrictions and total allowable e↵ort, the fishery is thought to be oversubscribed and the stock is regarded as being overfished. The social and economic importance of this fishery necessitates an update of the life history parameters of C. richardsonii to enable an accurate assessment the current status of the stock. The fishery in Saldanha and Langebaan was described via investigating changes in sexratio, mean length (mm) and standardised catch-per-unit-e↵ort (CPUE). Firstly, exploration of sex ratio indicated a significant switch between the two periods (1998-2002 and 2017), resulting in a predominantly male biased population (1.7 males: 1 female). Secondly, through investigation of three length-frequency distributions of commercial catch of C. richardsonii (1998-2002, 2009-2011 and 2017) a reduction in mean total length (TL) of 36.5 mm was observed. Lastly, the standardisation of the Netfishery CPUE for the time series of 2008-2016 through the application of a Generalised Additive Mixed Model (GAMM) showed a reduction of approximately 30% in relative abundance of C. richardsonii. Chelon richardsonii exhibited a fast growth, a maximum age of six and matured relatively early at two years old. Growth was best described using a three parameter von-Bertalan↵y growth model; where L1 is the asymptotic length, K is the rate at which L1 is reached and t0 is the age when the average length is zero. The data collected in 2017 expressed two problems. Firstly, as a result of high gillnet selectivity, smaller individuals within younger age classes were missing. Secondly, due to growth overfishing and/or a small sample size (n = 353) older and larger adults were missing. Consequently, this increasesd K and decreased L1 to biologically implausible values (female original growth: L1 = 257.450 mm, K = 0.610 year-1 and t0 = -0.040 year). As a result L1 was fixed in accordance to a historic Lmax, in order to overcome these issues and produce biologically plausible growth parameters. Growth di↵ered significantly between males and females, hence female growth was subsequently used for the spawner biomass-per-recruit analysis in the proceeding chapter (L1 = 347.400 mm, K = 0.235 year-1 and t0 = -0.833 year). Total mortality (Z) and average natural mortality (M ) were estimated as 1.466 year-1 and 0.329 year-1, respectively. i 0 ii Growth and mortality was constant in Chelon richardsonii throughout Saldanha and Langebaan and despite potential emigration out of the bay, the sup-population of C. richardsonii was considered to be a discrete stock for the purpose of this study. A spawner biomass-pre-recruit model, based on the growth and mortality parameters calculated in Chapter 3, revealed that the stock is heavily depleted and recruitment is likely to be seriously impaired (spawner biomass-per-recruit = 5.5% of pristine levels). It must be acknowledged that the results of a per-recruit stock assessment heavily depend on the growth model parameters. In contrast, the model indicated an optimally exploited stock when the original growth parameters were applied (spawner biomass per-recruit = 76.2% of pristine levels). Considering results from Chapter 2 and the justifications for fixing L1 the plausibility of the second scenario being true is less likely. Acknowledging the temporal, spatial and sample size limitations of this study conclusions made will require definitive future examination. Regardless, a reduction in fishing e↵ort and further restrictions in mesh sizes are suggested to facilitate the replenishment and sustainable use of the stock.
  • Thumbnail Image
    Item
    Open Access
    Aliens in the nursery : assessing the awareness and attitudes of Cape Town nursery managers in regard to invasive species regulations
    (2015) Cronin, Kate; Hoffman, Timm; Kaplan, Haylee
    The horticultural industry is recognised as one of the major pathways for the introduction and spread of invasive alien plants (IAPs). In recognition of this, the South African government has recently enacted a new set of Alien and Invasive Species regulations, under the National Environmental Management Biodiversity Act (NEMBA), that are intended, in part, to improve controls on the horticultural industry's role in the spread of IAPs. In order to assess, and possibly enhance, the likely effectiveness of NEMBA, it is critical to build an understanding of stakeholders' awareness and attitudes towards the control of IAPs and associated regulatory policies. A two-pronged approach – involving nursery manager interviews (n=30) and plant stock audit assessments (n=41) - was used to gauge the awareness, compliance and attitudes of Cape Town nursery managers towards the NEMBA regulations. Results showed that less than ten percent of the city's audited nurseries were compliant with the NEMBA regulations, and that over fifty percent were stocking IAPs that have been regulated for at least thirteen years under a previous set of regulations (the Conservation of Agricultural Resources Act, CARA). This is despite high levels of awareness about the CARA regulations reported in the interviews, reported enthusiasm for compliance, apparent concern for the environment, evidence that managers understand the problems that IAPs cause, extensive reported support for the control of IAPs, and a reportedly strong sense of duty to protect the environment. The vast majority (73.5%) of IAP species found in nurseries were NEMBA category 1b invaders such as Nerium oleander, Lantana montevidensis and Canna indica. These are species that are widespread and well-established invaders that require control. This study suggests that a range of factors are likely to negatively influence compliance including a perceived lack of enforcement, weak communication from government, issues related to the clarity of the regulations, the lack of inclusion of the industry in the regulatory process, and a lack of awareness, with at least half of the managers reporting that they had not heard about the enactment of NEMBA. Any attempt to improve the impact of the new regulations will need to adequately address each of these factors. The results of the study suggest that enhancing the impact of NEMBA will require improving aspects of the legislation itself, and supplementing the current top-down approach to regulation with an inclusive partner-centred approach.
  • No Thumbnail Available
    Item
    Open Access
    An intensive study of movement and population dynamics in Triakis megalopterus in the De Hoop Marine Protected Area, South Africa
    (2024) Wagstaff, Samantha; Attwood, Colin
    Coastal dwelling sharks are particularly at risk of decline due to the predicted rapid change in their environment and high incidences of accidental bycatch. Effective protection typically includes managing the fishing mortality and providing marine protected areas (MPAs) as refugia. To this effect, estimating natural mortality rates and home range size is vital. Triakis megalopterus, a commonly caught, endemic coastal shark of South Africa provides an excellent opportunity for this assessment. Mark-recapture data of 924 T. megalopterus caught at two sites in the De Hoop MPA from 1996 to 2020, a no-take reserve, were used to estimate home range size and natural mortality rates. Displacement frequencies were modelled to estimate home range size and space use within the MPA. Natural mortality rates for both sexes were estimated using two methods, one based on length data, combined with pre-established growth models, and another using the probability of recapture. Triakis megalopterus at De Hoop MPA displays a high level of philopatry. Individuals show consistent small movements over periods up to thirteen years. Multiple recaptures of the same sharks indicate frequent and repeated use of home ranges in the order of 1.0 km of coastline. The best fit model suggests a high degree of central tendency in space use. Skewed sex ratios towards females could be due to sex-specific longevity or mortality. The population had low mean natural mortality rates of 0.099/yr [95% C.I. 0.088/yr to 0.112/yr] for males and 0.072/yr [95% C.I. 0.062/yr to 0.082/yr] for females as estimated from the length data. These rates were close to published findings in other studies, which used models fitted to environmental and life-history data. Survivorship estimates for T. megalopterus based on tag-recapture probabilities were lower than expected compared to length-derived natural mortality rates, but still suggest a low natural mortality rate. The difference is likely caused by deflated length-based mortality estimation, caused by the high number of large female capture lengths, tag loss and predation. In conclusion, the small movements, philopatric tendencies, sex-specific movements, and rates, and small home range of T. megalopterus suggests that populations are highly susceptible to fishing and individuals are unlikely to radiate far and replenish diminished locations.
  • No Thumbnail Available
    Item
    Open Access
    An interdisciplinary study on the human-baboon interface in Rooiels, South Africa
    (2024) Mormile, Joselyn Elise; O'riain, Mannus
    The ongoing transformation of natural habitat for urban land use has negatively affected most wildlife, but some species are capable of adapting to novel human-dominated landscapes and even exploit anthropogenic resources. Primates are particularly adept at exploiting resources in human-modified landscapes, and in Africa, chacma baboons (Papio ursinus) are arguably the most successful primate urban exploiters. However, these spaces present novel risks to their health and survival, and baboons can damage urban infrastructure, access food from gardens, vehicles, and houses, and pose health and safety risks to people and pets. Consequently, in the interests of both people and baboons, the latter are typically actively discouraged from entering urban areas through a variety of interventions. There is, however, a notable exception to this exclusionary approach with the small coastal village of Rooiels (Western Cape Province, South Africa) having opted to share their properties and village with the local baboon troop. This provided a unique opportunity to explore the local community's acceptance of baboons, and to investigate how, when given a choice, baboons use urban versus natural habitat. Additionally, the absence of formal deterrents in the urban area allows for a rare opportunity to explore how the use of an urban area may influence the ranging patterns, activity budget, diet, and life history of baboons. I explored the spatial and behavioural ecology of the Rooiels troop using GPS data and behavioural observations, respectively, and used questionnaires to understand the attitudes, emotions, and value orientations of the local community towards the troop. The troop had a small home range of 5.78 km2 with a clear preference for both low-lying land and the village. Within the village, the troop selected for developed properties and roads over undeveloped erven, and amongst the developed erven they preferred those with larger areas of grass, properties where residents did not actively deter them and where food was provisioned. Their diet in urban areas was comprised predominantly of natural fynbos vegetation, with grass from lawns a commonly consumed fallback food. Aside from some deliberate provisioning, which greatly influenced the troop's space use within the village, human-derived foods, including foods from houses, vehicles, and bins, comprised <4% of their overall diet but their access to these foods peaked during the summer holidays when visitors were prevalent. The troop's activity budget more closely mimicked those of provisioned primates, with the least amount of time spent feeding, and considerable time spent resting and socialising. However, fecundity was not greater than troops living in natural habitats and mortality was high with 20 deaths occurring during the research period, of which 75% were due to anthropogenic causes, namely vehicles and dogs. Most residents held a mutualist wildlife-value orientation, had a positive attitude towards baboons, and expressed low levels of fear and stress in response to baboon visits. Consequently, most residents were willing to share the urban space with baboons, and this acceptance increased over time, developing into a culture of community stewardship which encouraged responsible behavior aimed at enabling the existence of the troop in the village. Despite this, the high level of anthropogenic mortality poses a risk to the persistence of the troop and suggests that this, and other urban areas, act as an evolutionary trap for baboons. Thus, the application of this inclusion model to other urban areas is ill-advised, although promoting community education and awareness to promote wildlife responsible behaviour and so reduce the potential for negative interactions associated with occasional urban incursions, is strongly advised. This interdisciplinary study has shown that communities can learn to adjust their behaviour and attitudes to wildlife in a manner that promotes coexistence, but if space sharing has unintended negative impacts on wildlife, it should not be encouraged until those impacts are mitigated.
  • Thumbnail Image
    Item
    Open Access
    Analysis of long-term changes in populations of the Clanwiliam Cedar (Widdringtonia cedarbergensis) using repeat photography
    (2013) White, Joseph Douglas Mandla; Hoffman, Timm; Jack , Sam; February, Edmund C
    Repeat photography photosets were used to analyse 20th century changes in populations of Widdringtonia cedarbergensis from four sites in the Cederberg Mountains, South Africa. Systematic evaluation of 115 photosets was combined with field observation and environmental data to determine the age class and degree of foliage cover, as well as fire frequency, rockiness of habitat, altitude, aspect of slope, annual mean temperature and annual precipitation for each tree viewed in photosets. Mortality and recruitment events were documented in each of the photosets. A total of 1315 trees were recorded in historical photographs (1931- 1982), with 968 of these shown to have died in the repeat photographs (2007-2013), indicating 74% mortality. With only 45 (3.4%) recruits being recorded, the total current living population of W. cedarbergensis in the repeat photographs was 392. There was no significant difference found in mortality or recruitment across the sampled sites. From the historical to repeat photosets there was an 8% increase in the proportion of mature adults and a 13% increase in the proportion of W. cedarbergensis individuals with sparse foliage cover. A generalized linear model was used to determine the effects of environmental factors on W. cedarbergensis mortality in natural populations.
  • Thumbnail Image
    Item
    Open Access
    Anthropogenic modification of the natural fire landscape and its consequences for vegetation patterns on the Cape Peninsula
    (2017) Rogers, Annabelle J; February, Edmund C; Moncrieff, Glenn R; Slingsby, Jasper A
    Understanding the spatial probability of fire and how urban development may alter natural patterns is particularly important in areas where alternate ecosystem states occur at fine spatial scales. The Cape Peninsula, South Africa, is a one such region where fire-sensitive forest patches occur interspersed in a sea of fire-dependent fynbos. Fire is believed to be an important determinate of forest distribution, with absence or occurrence of fires potentially allowing patch contraction and expansion. In this thesis I use a series of computer models to determine the extent to which anthropogenic development and land transformation have altered the spatial variation in fire likelihood, or the 'burn probability', and its consequence for the distribution of forest on the Cape Peninsula. The two multi-model, fire behaviour simulation systems I use are FlamMap and FARSITE. FARSITE is a deterministic simulation package used globally for discrete event simulation. In an effort to assess the viability of using the FARSITE model for fire prediction in fynbos and the determinants of model accuracy, I predicted fire area for a historical fire on the Cape Peninsula using a variety of fuel models and wind conditions. Following this validation, FlamMap was used to simulate the burn probability of the Cape Peninsula under natural conditions – no urban development present – and transformed conditions – where urban areas mapped as non-burnable fuel models. I then determined changes in forest distribution documented over the last 50 years relative to changes in burn probability as a result of urbanisation. My results show that an increase in urbanisation on the Cape Flats has produced a significant urban shadow effect due to the interruption of natural fire catchments. This urban shadow effect has resulted in an overall increase in area of fire refuges on the Peninsula and expansion of forest, particularly on the more mesic eastern slopes at Kirstenbosch and Newlands. The results strongly support that urban-mediated changes to fire patterns are drivers of forest expansion in this region, and adds further evidence to support the significance of fire in determining biome boundaries in the fynbos.
  • No Thumbnail Available
    Item
    Open Access
    Application of a multi-method approach to the stock identification and discrimination of kingklip, Genypterus capensis (Smith, 1847), off the South African coast
    (2024) Mobara, Ayesha; Underhill, Leslie
    Kingklip, Genypterus capensis, is a marine demersal fish distributed in shelf and slope waters of southern Africa, from Walvis Bay in Namibia to Algoa Bay in South Africa. This commercially important species is caught by demersal trawling and longlines. Whereas previous analyses of population structure in kingklip gave contrasting results, the most recent genetic and genomic analyses have indicated population structure in this species with three genomic clusters observed: one off the central Namibian coast, one off the southern Namibian coast and the South African West Coast, and one off the South African South Coast. The identification and delineation of sub-populations is necessary for effective management of exploited species, and determination of the stock identity of individual fish is most effectively and cheaply done using phenotypic characteristics. To assess whether the two genomically-differentiated stocks off South Africa are phenotypically differentiated, a holistic and integrated multi-method approach that analysed morphological (body and otolith), meristic, otolith elemental composition, and parasitological data of kingklip from the West and South Coast was applied. Nine body morphological (standardised for fish size) and three meristic variables were analysed. For both univariate and multivariate analysis, no significant differences between mature Genypterus capensis from the West and South coasts were observed in any of the 12 characteristics. For immature specimens, significant differences between regions were observed in four morphological (upper jaw, lower jaw, orbital diameter, and interorbital diameter), but not in any meristic characteristics. Univariate and multivariate analyses indicated that four (rectangularity, roundness, ellipticity, and aspect ratio) of the six otolith shape indices calculated showed significant spatial differences in mature specimens. In immature specimens, circularity, form factor, ellipticity, and aspect ratio, showed significant differences when using multivariate analysis, and roundness was also significantly different between regions when using univariate analysis. Elliptical Fourier Analysis of otolith shape was applied utilising 157 harmonics, of which seven in mature specimens, and nine in immature specimens were significantly different between regions. Seventeen elements were measured in kingklip otoliths and of these barium, phosphorus, and nickel showed significant differences between regions in immature but no differences were observed in mature fish. Regional differences in four kingklip parasites (Anisakis sp., Diphyllobothrium sp., Lecithochirium sp., Tentacularia sp.) previously identified as potential biotags were assessed, with univariate analysis indicating a significant difference for Tentacularia sp. for both mature and immature specimens. A Stock Differentiation Index (SDI) indicated weak (SDI = 0.26) spatial structure for mature, and a moderate (SDI = 0.33) spatial structure for immature Genypterus capensis in South African waters. These results indicate that the genomic differentiation of G. capensis off the South African coast does not translate into substantial and spatially differentiated phenotypic variation in mature specimens, however moderate stock delineation is observed in immature samples. This suggests that mixing between the two genomically-differentiated kingklip stocks is lowest for juvenile fish and increases with age, which accords with the hypothesis of separated and localised spawning locations for these two stocks off the West and South coasts of South Africa. That otolith shape indices had the strongest discriminatory power of the characteristics examined in this study suggests that analyses of otolith shape at an early stage inferred via measurement of the shape of the first otolith annulus (laid down when fish are 1-year old) may be a useful phenotypic characteristic for stock identity determination, and further research should be conducted in this regard.
  • Thumbnail Image
    Item
    Open Access
    Are there structural differences in the flight feathers among Procellariiformes related to the use of wings for underwater propulsion?
    (2019) Moloto, Ditiro Judith; Ryan, Peter G.
    Wings of birds differ significantly in relation to their flight mode, life-history, and habitat. Most seabirds have high aspect ratio wings for efficient gliding, whereas those that dive tend to have shorter wings with lower aspect ratio. Some seabirds including petrels, auks and, to a lesser extent, gannets use their wings to ‘fly’ underwater as well as in the air. These different environments differ greatly in terms of density. The aim of this study is to investigate if there are differences in the flight feathers of seabirds that use their wings for both underwater and aerial propulsion and seabirds that only use their wings in air. The study was restricted to the order Procellariiformes because of the wide range of divers and non-divers. I sampled 33 species ranging from albatrosses to diving petrels, including species that can dive quite well and those that barely dive at all. Due to the 800-fold difference in density between water and air, I expected to find structural differences in the flight feathers of petrels that use their wings underwater and those that do not. The investigation was based on feather length, mass, micro-structure and stiffness relative to body mass. I expected the feathers of birds that dive well (those that attain depths more than 10 m deep) to be shorter (derived from their short wings), heavier, and stiffer compared to the feathers of birds that seldom dive. Further, I expected the differences to be more marked in the primaries compared to the secondaries, because primary feathers are subject to greater forces during flight. Allometric comparisons showed both expected and unexpected results. The primary feathers of diving birds were shorter relative to body mass while the secondaries showed no differences between the two groups. As a result, non-diving birds had heavier primaries compared to the diving birds while there was no difference in the secondaries. As expected, diving birds had heavier feathers relative to feather length. There were minimal to no differences in feather microstructure or stiffness between divers and nondivers. However, diving birds exhibited slightly less flexure relative to the length of their flight feathers than non-diving birds, suggesting that their feathers are slightly stiffer. Diving birds appear to have adapted to utilising the two contrasting media by evolving short, slightly heavy and stiff outer primaries, but these differences are not marked. The main adaptation to flight underwater probably is to partly close the wing, reducing its area and increasing the overlap between adjacent feathers.
  • Thumbnail Image
    Item
    Open Access
    Aspalathus and Podalyria legumes balance acquisition of phosphorus and nitrogen for growth in nutrient poor fynbos soils
    (2015) Maistry, Pravin Mark; Chimphango, Samson B M; Muasya, A Muthama; Valentine, Alex J
    Legume species nodulate and grow successfully in the Core Cape Subregion, a Mediterranean-climate ecosystem with fynbos vegetation found on infertile soils. The physiological mechanisms enabling tolerance of low availability of phosphorus (P) are yet to be reported in Cape legume species such as Aspalathus linearis and Podalyria calyptrata; species that demonstrated traits typical of plants from nutrient poor soils. In the three research chapters of the thesis, it was anticipated that low P supply would limit plant growth and increase expression of traits for P acquisition.
  • No Thumbnail Available
    Item
    Open Access
    Aspects of population biology and behaviour of mobulid rays
    (2023) Carpenter, Michelle; Griffiths, Charles
    Manta and devil rays (Mobula spp.) are charismatic yet threatened species that remain poorly understood. The Inhambane Province in Mozambique is known to be home to the largest populations of manta rays in Africa, including the reef manta ray (M. alfredi) and the oceanic manta ray (M. birostris). However, research on other mobulid species, and in the region south of Tofo, including Závora in Mozambique and South Africa, has been limited. This thesis identifies critical sites and describes their characteristics, including habitat use, population abundance, and behaviour, for three species of mobulids: M. alfredi, M. birostris, and the shortfin devil ray, M. kuhlii. Chapter 1 comprises a literature review on the status and knowledge on mobulids in a global context, and within the southern African region, with focus on M. alfredi and M. birostris, due to the increased amount of information available on these species. Chapter 2 provides detailed information on the methodologies used throughout the thesis, some of which overlap. Chapter 3 utilises an 11-year photographic mark-recapture dataset spanning from 2010-2021 to investigate the size and structure of the M. alfredi population that aggregate at 'Red Sands' (RS), a reef cleaning station located at Závora, Mozambique. A total of 583 individual M. alfredi were identified using photographic mark recapture (photo-ID) of the unique ventral spot patterning. Pollock's Robust Design models were used to estimate annual abundances, emigration, annual apparent survival, and capture probability at RS during peak season, July-November, between 2016-2021. The resulting abundance estimates varied each year, with a range of 35 individuals (95% CI 30-45) up to 233 individuals (95% CI 224-249). Lagged Identification Rates were additionally used to describe the residency patterns of M. alfredi at Red Sands, revealing that despite site affinity to RS, individuals were overall transient within a given season, with an average residency of four days. Chapter 4 applies the methods used in Chapter 3 to report on South Africa's first photo-ID dataset for manta rays, including both M. alfredi and M. birostris, spanning from 2020-2022. Through the use of citizen science photographic donations and opportunistic snorkelling and diving, were used to document manta ray aggregation sites in the KwaZulu-Natal (KZN) Province. These same data were used to further document connectivity of M. alfredi between Mozambique and South Africa, revealing that at least 20% of all photographed individuals in South Africa had made international migrations to or from Mozambique. The distances travelled ranged from 435 km and 1,000 km, between the Aliwal Shoal Marine Protected Area (MPA), KZN and Závora, the second longest known migration for this species. Additionally, previously undocumented aggregations were found, including M. alfredi at the iSimangaliso Wetland Park and possibly M. birostris at Ballito. Aggregations in KZN were primarily related to surface feeding, with cleaning and courtship behaviours also documented. Chapter 5 analyses a 41-year catch dataset (1981-2021) provided by the KZN Sharks Board bather protection programme to investigate broad spatial-temporal trends in catch for manta rays (M. alfredi, M. birostris). Using Generalised Additive Models and the probability of encounter, the effects of annual and seasonal factors, moon phase, and location on catch rates are assessed. The size composition and demographics of the caught manta rays and the catch per unit effort at different beaches are also examined. The results revealed a significant decline in overall catches around 2000 (p<0.0001). Catches increased in summer, indicating seasonal visitation to South African waters. Moon phase had an effect on manta ray presence, with more catches during new and full moon phases. Over half of the catches included juveniles, whereby a total of 841 individuals (52% of total catch) of confirmed juvenile size class (1400-2500 mm disc width) were caught, particularly in the southernmost sampled area, from Hibberdene in the north to Mzamba in the south. Chapter 6 investigates a previously undocumented aggregation of M. kuhlii in the Aliwal Shoal MPA, and their cleaning behaviour by the bluestreak cleaner wrasse, Labroides dimidiatus. The study documented M. kuhlii sightings, behaviours, and associated environmental conditions during recreational snorkelling/diving. Generalised additive models (GAMs) were used to assess predictors of M. kuhlii presence and cleaning behaviour duration. The results showed that M. kuhlii were present in 56% of the survey days, with up to 150 individuals observed. The study recorded a total of 4.04 hours of M. kuhlii cleaning behaviour which was processed using frame by frame analysis. Sea surface temperature was found to be a significant predictor for M. kuhlii presence, while the absence of current and estimated horizontal visibility were significant predictors for longer cleaning durations. During recreational diving in the Aliwal Shoal MPA, first-time observations of M. kuhlii courtship were unexpectedly encountered and videoed in November 2020 and January 2021, which are detailed in Chapter 7. The common behaviours associated with mobulid courtship were documented: multiple males following a single female, rapid speed bursts, avoidance of the female, and swerving, which confirmed that M. kuhlii mating occurs in KZN waters. The collective results are concluded in Chapter 8, which indicate that both Závora and KZN waters provide important seasonal habitat for mobulid species along the southern African coastline. The findings of this work support other studies that show manta rays to use an expansive area, with seasonal affinities to certain locations, and transience during periods when conditions are not favourable for visitation. This demonstrates the importance of understanding site use within the larger home range of a given mobulid population. Therefore, this thesis has filled critical knowledge gaps and improved the overall understanding of mobulids in southern Africa, providing valuable information for conservation management efforts.
  • Thumbnail Image
    Item
    Open Access
    Aspects of the biology, ecology and fishery of the beaked clam Eumarcia paupercula (Holten, 1802), in Maputo Bay
    (2016) Mugabe, Eulália D.; Griffiths, Charles L; Macia, Adriano
    Clam populations globally have declined, or been depleted, with one of the major causes being uncontrolled human exploitation. This thesis investigates the population structure, growth, reproduction and exploitation of the beaked clam Eumarcia paupercula in Maputo Bay. The substantial commercial harvesting of this clam may lead to overexploitation of the resource. In this regard, some fundamental knowledge is necessary for the management of future exploitation. The data collection was based on an 18-months (November 2012 - April 2014) sampling for population structure across a tidal flat. The growth analysis was performed on FiSAT II, using mark-recapture experiments and length-frequency data. Monthly reproduction analysis was based on the fluctuation of body weight and gonad smear analysis. Furthermore, interviews were used to ascertain the importance of the resources for collectors and estimate landings of E. paupercula. The von Bertalanffy growth function and length-frequency analysis revealed that E. paupercula has a fast growth rate and a short life span. Eumarcia paupercula is a year-round spawner with higher peaks in the summer; recruitment follows a similar pattern, occurring over the year and after the spawning peaks. The clam collectors, the majority of whom are women, have experienced an increase in the effort required to collect clams, resulting in a decline of catches over the season. This study highlights that temporal population dynamics are influenced by collection and reproductive patterns, and that single environmental parameters do not explain the patterns of growth, reproductive cycles and spatial distribution. Findings of this study have relevance and application for the livelihood of the collectors, as well as the sustainability of the Eumarcia paupercula stock, by providing a basis for fishery governance. While recommendations are presented for the management of E. paupercula collection, the Ministry of Fishery in Mozambique also has to consider collecting data on bivalves, as they are an important source of income for artisanal fishers. Future research should include monitoring of a less exploited population, so as to understand better how collection impacts on the population dynamics of E. paupercula. Furthermore, laboratory studies of the larval cycle are necessary to gain thorough understanding of the species life cycle.
  • Thumbnail Image
    Item
    Open Access
    Aspects of the spatial and behavioural ecology of Hippopotamus amphibious in the Saint Lucia Estuary, KwaZulu-Natal, South Africa
    (2016) Prinsloo, Alexa Simone; Pillay, Deena; O'Riain, M Justin
    The common hippopotamus (Hippopotamus amphibius) is endemic to Africa but threatened by range restriction, exploitation and competition with humans and domestic livestock for access to critical resources such as fresh water and grazing. Collectively these impacts have resulted in population declines throughout most of their range with the consequence that hippos are classified as Vulnerable on the IUCN Red list. Despite these threats to their persistence and their important ecological role as ecosystem engineers there have been surprisingly few studies concerning factors that influence their current distribution and habitat use. Such research is important for developing effective conservation and management plans to improve the protection of vulnerable and ecologically important species such as hippo. South Africa is home to one of a few growing hippo populations and one of the last persisting in an estuarine habitat. The St Lucia Estuary, one of three Ramsar Wetlands of International Importance located in the iSimangaliso Wetland Park (itself a UNESCO World Heritage site), in northern KwaZulu-Natal is not only the largest estuarine system in Africa, but is also home to one of South Africa's largest hippo populations. In this study I explore aspects of the spatial and behavioural ecology of hippo within the St Lucia Estuary. More specifically, I investigate select abiotic and biotic predictors of hippo presence and the frequency with which they return to their diurnal lie-up sites. I used binary logistic regression models to evaluate the effects of the selected habitat features on hippo occurrence at two spatial scales (broad and fine). I used a negative binomial with log-link model to evaluate the effects of the selected habitat features on the frequency of use of sites (hippo persistence) in the fine-scale study. In addition, I employed both instantaneous scan sampling and continuous sampling methods in an attempt to determine the diurnal activity budgets and social interactions of a single group near the mouth of the St Lucia Estuary. I used non-parametric tests to evaluate differences in behaviour with daytime and across sampling days with a view to an improved understanding of how the environment influences the behaviour of hippo in an environment subject to ongoing human modification.
  • Thumbnail Image
    Item
    Open Access
    Assessing safety margins and hydraulic strategies in Restionaceae: evaluating xylem hydraulic traits in two Elegia species
    (2021) Atkins, Kayla; West, Adam; Skelton, Robert
    The movement of water through a plant is responsible for acquiring nutrients, avoiding desiccation and enabling gas exchange. Therefore, access to soil water is vital in determining vegetation structure in a landscape, which suggests that plant hydraulic strategies are an important ecological driver of responses to seasonality and intensity of drought events, especially in the face of climate and habitat change. Wide-spread mortality has been observed over recent decades as a result of drought stress. Many plant species have exhibited hydraulic niche segregation as a result of a trade-offs between stomatal regulation and xylem integrity, and the associated physiological adaptations. This is concerning in the face of climate change predictions of more frequent and intense drought and flooding events because predicted habitat water availability changes may exceed the physiological tolerance ranges of many species. The hydraulic safety margin (difference between xylem vulnerability (P50; MPa) to cavitation and point at which stomata are considered closed (turgor loss point; MPa)) displays the tolerance margins based on plant physiological capacity. Restionaceae have shown extensive evidence of hydraulic niche segregation and are a key demographic of fynbos vegetation but not much is known about their hydraulic strategies. The optical vulnerability method was used to construct vulnerability curves and the turgor loss point (ΨTLP) was acquired from pressure-volume curves. E. fenestrata, a localized seep species, was more vulnerable (P12 = −0.52 MPa; P50 = −1.07 MPa; ΨTLP = −1.57 MPa; negative safety margin = −0.5 MPa) than E. tectorum, a widespread, dry habitat species (P12 = −0.99 MPa; P50 = −1.6 MPa; ΨTLP = −1.64 MPa; slightly negative safety margin = -0.04 MPa). The Cape Floristic region is predicted to receive less rainfall and become drier with ongoing climate change. We expect that this overall drying trend will have a profound impact on the Restionaceae, particularly E. fenestrata that does not have the physiological capacity to deal with severe drought stress.