Gastropods are one of the most diverse species groups in mangrove habitats, however, many of their specific roles in relation to ecological patterns and processes are currently largely unknown. The overall aim of this research project was to provide basic ecological information for key gastropod species from subtropical mangroves within a protected area. South African mangroves cover relatively small areas and are restricted to estuaries, these habitats therefore present unique opportunities and challenges to the species that occur in them. Three gastropod species, Terebralia palustris, Cerithidea decollata, and Melanoides tuberculata, all occur at their natural southernmost range limit within South Africa and were selected based on their prominence and occurrence in mangrove habitats of the iSimangaliso Wetland Park, a UNESCO World Heritage Site. Trophic linkages and resource partitioning, resource utilization rates, and ecological resilience were investigated respectively using: 1) a stable isotope (δ15N and δ13C) approach; 2) an experimental approach to quantify feeding dynamics (ingestion rate, consumption/digestion efficiency and grazing impact); and 3) a mixed-effects modelling approach to relate population responses to environmental variables. The diet of T. palustris was seasonally variable and a number of sources were incorporated by different sized snails, but their grazing impact on microphytobenthos was not significant. The results also indicated an ontogenetic shift in the dietary niche for T. palustris through robust partitioning of resources between different size classes. The diets of C. decollata and M. tuberculata were dominated by different primary resources as a function of where they occurred in the mangroves. Melanoides tuberculata consumed a wide variety of primary resources, a typical trait of an opportunistic generalist species. The ingestion rate of M. tuberculata was not dependent on the availability of microphytobenthos, and was highest when conditions were oligotrophic. The resilience of C. decollata was related to the tree-climbing behaviour of this species and its occurrence was best explained by sediment conductivity. These responses were considered in conjunction to what has previously been reported on the resilience of the mangrove trees. The results of this research project have provided new basic ecological information for all three gastropod species in this data-deficient subtropical region. This information can potentially be used in comparative studies for these species in other regions or in broader scale ecological studies. Terebralia palustris has recently experienced a range contraction along the South African coastline. This research project has shown that the diet of this species is highly variable and that food limitation and competition for resources should be considered as potential drivers of the local decline. Cerithidea decollata has in contrast expanded its distributional range in this region. This research project has shown that this species has a generalist diet and exhibits traits in relation to tolerance that are expected to have facilitated its expansion into temperate saltmarsh habitats that occur in dynamic estuaries. Melanoides tuberculata is a globally invasive species, and as South African populations are within its native range, ecological information from this region is valuable as it can be used to investigate the potential ecological effects following introduction into new habitats beyond the native range. Biological drivers have a significant impact on mangrove ecosystem functioning, particularly in relation to recycling and the retention of organic carbon generated through primary productivity. Understanding the ecological linkages that maintain ecological functioning and stability is therefore an important step towards conserving and sustainably managing threatened ecosystems such as mangrove forests.

The St. Lucia Estuary, the largest estuary in South Africa, has been subject to many natural (a decade long drought) and anthropogenic impacts. A particular mouth manipulation activity, the artificial separation of the Mfolozi River and the St Lucia Estuary in 1952, was done to stop the perceived “silting up” of the estuary, but resulted in a decrease in freshwater supply. The changes in inflows (both fresh and marine) are controlled by management decisions and affect other system parameters such as salinity, water level and turbidity, which influence the distribution of biota. Therefore knowledge on the physico-chemical environment and eco-physiological tolerances of macrophytes will lead to informed future management decisions. The first of the three objectives carried out for this study determined the present state and distribution of the macrophytes of the St. Lucia and Mfolozi estuaries. The macrophyte habitats mapped in 2008 and 2013 were the submerged macrophytes, reeds and sedges, mangroves, grass and shrubs, salt marsh (succulent) and swamp forest. Results indicated that low salinity in the lakes and high water level in 2013 caused die-back or expansion of particular habitats. Submerged macrophytes, in particular Stuckenia pectinata, grows well in water with salinity <15 ppt, therefore this habitat increased by 412 ha (96%) in cover since 2008. Salt marsh decreased by 553 ha (57%) due to inundation. Since 2008 the reeds and sedges increased by 390 ha (in North Lake and the Narrows) due to the salinity decrease. In the Narrows the mangroves decreased by 28 ha (9%) in area cover. This was due to the drought that persisted for so many years, which caused low water levels and non-tidal conditions. The second objective was to determine the present state / health of the mangroves at four sites along the Narrows by assessing sediment condition and population structure of the trees. These results were compared to those obtained in 2010. The total density of Avicennia marina increased since 2010, however this was due to the large increase in seedlings at Site 1, the back channel site. The highest sediment salinity (26 ppt) and porewater salinity (29 ppt) was recorded for this site and these results show that this back channel site was strongly influenced by the open Mfolozi Estuary (increase in marine waters) and tidal conditions at the time of sampling in 2013. The total density of Bruguiera gymnorrhiza decreased, but an increase in adults was recorded at Site 2, the freshwater site. The soil collected from the Bruguiera quadrats was fresher and drier than that of the Avicennia quadrats of Site 2. Lack of seedlings (of both species) was due to the dense stands of Acrostichum aureum (mangrove fern) and Phragmites australis (common reed) and a thick, impenetrable mat of Avicennia pneumatophores. At Sites 3 and 4, the drier sites (where sediment moisture contents were the lowest for all sites at 43 and 42% respectively), seedling and sapling density was low, but adults increased in density since 2010. Recruitment and survival were impacted by the harsh environmental conditions that prevailed prior to 2013 (low water level and non-tidal influence), but adults survived. The results of the test that determined the percentage of aerenchyma of the pneumatophores indicated that waterlogging stress did not affect the aerenchyma of the pneumatophores. However the period of inundation was probably not significant enough to have affected the production of aerenchyma. The third objective formed part of an ongoing study by the Global Environmental Facility (GEF) on the feasibility of linking the Mfolozi River back to the St Lucia Estuary and the responses of these systems to different management scenarios: 1) the “do nothing” scenario; 2) maintain separate Mfolozi and St Lucia mouths representing an open mouth condition; and 3) actively facilitate a single mouth (therefore linking the Mfolozi and St Lucia mouths). Data gathered on the eco-physiological tolerances of the dominant macrophyte species was used to predict the response of the different habitats to these various management scenarios and the results indicated that the best management scenario would be to actively facilitate a single mouth (Scenario 3) as the estuary habitats would increase significantly because of preferred tidal and saline conditions, as this would represent more natural conditions. The results of the study will provide input to recommendations for future “adaptive management” strategies for the Global Environmental Facility (GEF) Project.

Estuaries require sufficient quantity, quality and appropriate timing of freshwater inflow, referred to as environmental flow requirements, to ensure adequate health and functioning. In South Africa the environmental flow requirements of estuaries is determined using Resource Directed Measures (RDM). The present health and importance of an estuary must be considered when determining the ecological flow required to maintain an estuary in its desired state. An Estuarine Health Index that quantifies changes in abiotic and biotic components from natural conditions to present day is used. Health of biotic components is assessed according to changes in species richness, abundance and community composition. There has, however, been debate regarding the calculation of these attributes. In particular, for macrophytes, the inclusion of all habitat within the Estuarine Functional Zone (i.e. 5 m topographical contour) would affect the health and changes over time determined in past assessments. This is due to different areas being included as different areas would be included The aim of this study was to test the validity and suggest improvements to South Africa’s RDM macrophyte health score determination. The health of macrophytes were assessed at varying levels of intensity from desktop studies to thorough field studies. Rapid field studies provided a visual estimate of macrophyte health; while the intermediate and comprehensive assessments quantified change by mapping the distribution of macrophyte habitats from aerial photographs. These approaches were applied to the 64 estuaries within the Mvoti-Mzimkulu Water Management Area (WMA 11), situated in KwaZulu-Natal (KZN) on the subtropical east coast. Historically, these estuaries have supported restricted macrophyte habitats. Consequently, estuaries that presently or historically supported range limited habitats such as mangroves and swamp forest are considered important. Submerged macrophytes are scarce in KZN estuaries due to siltation and low turbidity and thus estuaries supporting this macrophyte habitat are also important. Based on available literature important estuaries were highlighted and selected for field studies. Transects spanning from the estuary water channel to the boundary of the EFZ provided a generalised distribution of KZN vegetation along an elevation gradient. The transition from estuarine to terrestrial vegetation can be used to improve the current EFZ boundaries, as estuarine area has been found to occur outside of the 5 m contour in some estuaries. Results from the desktop assessment indicated that most of the estuaries are moderately modified. There was a 50 % similarity in the macrophyte health scores determined by the desktop assessment to the 2011 National Biodiversity Assessment (NBA). Field studies mostly confirmed the desktop assessment aside from seven estuaries that had different scores. There has been a significant loss of macrophyte habitat with 100% loss of certain habitats from 9 of the 22 estuaries for which there were field assessments. Submerged macrophytes were not found at any of the estuaries that were assessed in the field. Black mangroves, Bruguiera gymnorrhiza (L.) Lam., were rediscovered at Ngane Estuary, however the few individuals could not be considered a ‘mappable’ community (i.e. <0.5 ha). Mtamvuna and Mkomazi estuaries both supported small stands of mangroves that, compared to data from 2006, appeared healthy. Macrophyte habitats and surrounding coastal forest matched the species composition described in the Vegetation Map of South Africa. The freshwater mangrove or Powder puff tree, Barringtonia racemosa (L.) Roxb, was not abundant in the estuaries. It was only found at four (Damba, Fafa, Little Amanzimtoti and Mvoti) of the estuaries that were assessed in the field. Important estuaries, from a botanical perspective, included some of the larger estuaries such as Mgeni, Durban Bay and Sipingo. Smaller, healthier estuaries that presently support mangrove and swamp forest habitat were also identified as important from a botanical perspective. Macrophyte habitats have mainly been lost due to non-flow related pressures. Sugarcane cultivation occurred within the floodplain of 25% of the estuaries. Nutrient enrichment was evident for 12.5% of the estuaries and reed encroachment was evident for 27% of estuaries for which there were field assessments. Development, aside from the N2 road bridges, was evident for 40% of the estuaries assessed in the field. Invasive plant species, including aquatic invasive plants, were present in all estuaries assessed in the field. The extent of invasive plant species was related to the degree of disturbance and surrounding land use pressures. The Mkomazi and Mvoti estuaries that were assessed at a comprehensive level, as they have earmarked for further water abstraction, had the poorest macrophyte health scores. The macrophyte health of these estuaries was much lower than previously determined by the NBA. The decline in health was attributed to the removal of macrophyte habitat for sugarcane cultivation and development as well as displacement by invasive plant species. The areas covered by the macrophyte habitats also differed from the NBA highlighting the importance of updated mapping and ground truthing. Largely differing macrophyte health scores were produced when attributes were calculated using different combinations of macrophyte habitats. Results indicate that only macrophyte habitats and no other floodplain vegetation should be included when calculating abundance. All macrophyte habitats, physical habitats and floodplain should be incorporated when calculating community composition. The minimum of these attributes is used as the overall macrophyte health score as a precautionary approach is followed. Scoring of health is subjective and the only benchmark for determining the most appropriate method is comparion with previous RDM studies. This study illustrates the need for a standardised RDM scoring method that is presented in a manner that ensures the same results irregardless of the speciliast conducting the study. The updated estuary health and importance scores for WMA 11 are necessary to inform management, particularly as few of the estuaries receive formal protection. In conclusion, this study contributed to the limited knowledge of the estuaries of WMA 11 and assisted in determining appropriate methods for assessing the health and importance of estuary macrophytes.

Temporarily open/closed estuaries (TOCEs) are unpredictable environments that change in response to mouth condition, which is influenced by freshwater flooding or sea storm surges. The aim of the study was to determine whether macrophyte phenology in a TOCE was event driven rather than cyclically predictable and if it differed from permanently open estuaries (POEs). Macrophyte growth and flowering phenology in response to environmental conditions was investigated in the East Kleinemonde Estuary (TOCE) and the Kowie Estuary (POE) along the eastern coastline of South Africa. The lack of freshwater flooding due to low rainfall coupled with several overwash events resulted in a prolonged period of mouth closure from September 2008 to the end of this study period in 2010. This in turn caused the inundation of the supratidal and intertidal habitats, high water level (> 1.57 m amsl) and high salinity (30 - 42 ppt) in the TOCE. Principle Components Analysis showed that high water level and reduced sediments were the most significant environmental factors affecting macrophyte phenology. Macrophyte phenology in the POE was primarily driven by temperature, sediment redox potential and salinity. The saline high water level and reduced sediment significantly reduced macrophyte cover in all habitats in the TOCE. Macrophytes in the POE maintained high cover abundance due to seasonal re-growth compared to the TOCE where cover declined over the sampling period due to the high water level. Subsequent to water level dropping by as little as 11 - 20 cm in the TOCE, the intertidal species Sarcocornia tegetaria and Salicornia meyeriana completed their life-cycles and produced viable seeds within four and three months of germinating respectively. In contrast, the Sarcocornia hybrid and S. meyeriana in the POE took longer to complete their life-cycles, namely seven and nine months respectively, while S. tegataria did not germinate in situ but reproduced vegetatively despite producing seed. In the TOCE, the submerged species Ruppia cirrhosa and Chara vulgaris completed their life-cycles within five and three months and produced a maximum of 26 242 and 196 998 seeds m-2 respectively. Due to high water level and prolonged inundation, the reproductive periods were shorter for the intertidal and reed and sedge species in the TOCE compared to the POE. Seed output during the two reproduction periods varied between the two estuaries. Sarcocornia decumbens and S. tegetaria produced a substantially higher number of seeds in the TOCE compared to the POE, namely 0 - 102 847 versus 20 661 - 48 576 seeds m-²; and 7 001 - 45 542 versus 1 587 – 16 958 seeds m-² respectively. Seed output in the TOCE was significantly higher in S. tegetaria during the second reproduction period despite the significantly lower plant cover, which may be a function of the stressful environment in the TOCE. Seed production of S. meyeriana was significantly higher in the POE compared to the TOCE, with 264 224 - 640 292 compared with 24 050 - 27 643 seeds m-², due to higher plant cover in the POE. The research suggests that macrophyte phenology in the TOCE was significantly influenced by mouth condition. Further, macrophytes were able to demonstrate considerable phenotypic plasticity in response to changing and unfavourable environmental conditions. These data can be used in mouth management plans and freshwater requirement studies in TOCEs to ensure that macrophytes can complete their life-cycles and produce viable seeds for the safeguarding of habitat persistence and ecological processes. In impacted estuaries where artificial mouth opening is practised and the macrophytes have been severely degraded or extirpated, management should ensure that the intertidal and supratidal habitats are not inundated during peak flowering and seed production periods i.e. late spring to early autumn (November to March).

River flow is important in controlling phytoplankton distribution in estuaries. Data on the effect of river inflow on phytoplankton distribution patterns in temporarily open/closed estuaries is lacking. This study investigated the influence of river inflow on size-fractionated phytoplankton biomass (Chl a), community composition and environmental parameters measured monthly over three years in two temporarily open/closed estuaries in the Eastern Cape, South Africa. A once-off primary production study over an annual cycle was completed in the Van Stadens and Maitland estuaries. The study monitored physical, chemical and biological characteristics in both estuaries to examine the effects of changes in environmental factors and river inflow. Daily sampling of physico-chemical and biological variables from river to sea was carried out in the Van Stadens to investigate short-time scale effects of changes in environmental factors and river inflow on the phytoplankton biomass. Five and three stations in the main channel of the Van Stadens and Maitland estuaries respectively were sampled at 0.5 m below the water surface and 0.5 m above the sediment surface for biological and chemical variables and at the surface, 0.25 m and every 0.5 m thereafter for physical parameters. Five stations adjacent to the main channel along the estuary were monitored for groundwater macronutrient concentrations and five additional sites located within the upper catchment of the Van Stadens River were sampled on a quarterly basis over two years. Both estuaries were characterised by distinct hydrological conditions, an overwash, an open, a closed and a semi-closed mouth phase. Flooding in the Maitland and Van Stadens estuaries in 2001 and 2002 caused sediment scour, altered channel morphology and brought about breaching of the mouth. Flood driven mouth-breaching events occurred three and four times in each of the estuaries during the study. The mouth stayed open 20 – 25 percent and was closed 60 – 65 percent of the time. In the Van Stadens the closed overwash mouth condition occurred approximately 10 – 20 percent of the time while in the Maitland it occurred less with the semi-closed mouth condition occurring 10 – 20 percent of the time. Incidents related to mouth opening not associated with strong river floods occurred approximately 10 – 15 percent of the time, although in the Maitland a semi-closed mouth state persisted more frequently than in the Van Stadens Estuary. During flooding events salinity dropped to low levels (< 5 psu) but soon recovered to brackish conditions when river flow was reduced and marine water penetrated deep upstream. Reduction in river flow combined with marine sediment deposition resulted in the closure of the mouth. During closed mouth conditions strong onshore storm surges and spring high tides introduced marine water through overwash that kept salinity high. In both estuaries salinity showed a negative correlation with rainfall (R2 = 0.12), indicative of the strong influence of marine overwash that kept salinity high thus masking the influence of freshwater. High rainfall in the Van Stadens Estuary caused high levels of turbidity that reduced light penetration at depth. Light attenuation was positively correlated with the high rainfall (R 2 = 0.26) suggesting that increased turbidity was linked to rainfall induced discharge. In contrast, in the Maitland Estuary light attenuation did not show any correlation with increased rainfall possibly because of the reduced water depth and increased euphotic zone following the floods in 2002. High river inflow introduced macronutrients in both estuaries such that dissolved inorganic phosphates (DIP) and dissolved inorganic nitrogen (DIN) concentrations in the Van Stadens Estuary were strongly correlated with rainfall (R2 = 0.78 and 0.57 respectively). In the Maitland Estuary DIP and DIN concentrations remained significantly higher (p < 0.05) compared to that in the Van Stadens suggesting that the Maitland catchment contributed greater nutrient input into the estuary and may be associated with farming activities. Phytoplankton chlorophyll a (Chl a) ranged from 0.8 – 13.9 μg L-1 in the Van Stadens and in the Maitland Estuary from 5.3 – 138 μg L-1 during the 3-year study. During the open mouth condition Chl a biomass and primary production ranged from 5.4 – 52.9 μg Chl a L-1 and 1.2 – 11.7 mg C m-2 d-1 in the Maitland and in the Van Stadens from 1.6 – 9.8 μg Chl a L-1 and 1.2 - 14 mg C m-2 d-1 respectively. Maximum annual primary production in the Maitland and Van Stadens estuaries was 8.8 and 5.1 g C m-2 y-1 respectively. When the mouth was open in the Van Stadens Estuary the microphytoplankton (> 20 μm) accounted for > 65 percent of the Chl a, whereas during closed mouth conditions they accounted for about 55 percent of the Chl a biomass. Chlorophytes became the dominant taxon in the dry summer months but were replaced by cryptophytes and dinoflagellates during the wet season. When nutrient concentrations were low during low flow conditions in the Van Stadens Estuary mixotrophic microphytoplankton became an important fraction of the water column together with phototrophic dinoflagellates and cryptophytes. In the Maitland large sized chlorophytes were the dominant taxa in late spring and summer seasons and made up more than 80 percent of the cell numbers. In the Maitland before the floods in 2002 cyanophytes were the dominant group in late spring contributing more than 75 percent in cell abundance. Data from the short-term study in the Van Stadens Estuary showed similarities and differences in the Chl a response to increased river inflow. High river inflow initially reduced Chl a biomass followed by a recovery period of a couple of days compared to a 8 – 10 week recovery period in studies monitored over seasonal and annual temporal scales. The responses may be dissimilar but help to illustrate that there are similar response patterns to environmental forcing necessary to support phytoplankton biomass at different temporal scales. This study has demonstrated that flooding events caused by strong river flow cause breaching of the mouth, a reduction in salinity and marked nutrient input. Although the causes of flooding can be similar in both estuaries the resultant effects are varied and can alter the ability of the estuary to retain water. This study was able to demonstrate that the supply of macronutrients from the catchment was strongly correlated with rainfall (R2 = 0.67) and that phytoplankton growth mainly depended on an allochthonous source of macronutrients although internal supplies could be critical at times in controlling microalgal biomass.

Conventional knowledge of fynbos fire ecology is based on the summer-autumn fire regimes of the western Cape Floral Kingdom (CFK) where the climate is Mediterranean. However, the climate in the eastern coastal-CFK is milder and rainfall occurs year-round, with presumed effects on fire regimes. The Garden Route National Park (GRNP) has recently been established in the region, in a landscape where indigenous forests, fire-prone fynbos shrublands and fire-sensitive plantations of invasive alien trees are interspersed. The park faces considerable challenges related to the management of fire, including significant pressure from the adjacent plantation industry to reduce wildfire hazard by burning fynbos at short intervals, and high levels of invasion by alien trees (largely Pinus species originating from plantations). This study sought to improve understanding of fire regimes in eastern coastal fynbos shrublands, and to provide guidelines for ecologically sound management of fire in the area. My approach entailed (i) an assessment of the context within which fire management was practiced during the past century; (ii) characterisation of the recent fire history and fire regime (1900–2010); (iii) characterisation of the seasonality of fire weather and lightning; (iv) estimation of minimum fire return intervals (FRIs) from juvenile periods and post-fire recruitment success of overstorey proteoids (non-sprouting, slow-maturing, serotinous Proteaceae); and (v) determination of the ecologically appropriate fire season from post-fire recruitment seasonality of proteoids. I established that historically, plantation protection enjoyed priority over fynbos conservation in the area that is now the GRNP. Fynbos close to plantations has most likely been compromised by frequent and low-intensity burning in the past, as well as by invasion by alien trees. In terms of area burnt (1900–2010), natural (lightning-ignited) fires dominated the fire regime, particularly in the east, whereas prescribed burning was relatively unimportant. Typical fire return intervals (FRIs; 8–26 years; 1980–2010) were comparable to those in other fynbos protected areas and appeared to be shorter in the eastern Tsitsikamma than in the western Outeniqua halves of the study area. Proteaceae juvenile periods (4–9 years) and post-fire recruitment success (following fires in ≥7 year-old vegetation) suggested that for biodiversity conservation purposes, FRIs should be no less than nine years in moist, productive fynbos. Increases in the total area burnt annually (since 1980) were correlated with long-term increases in average fire danger weather, suggesting that fire regime changes may be related to global change. Collectively, findings on the seasonality of actual fires and the seasonality of fire danger weather, lightning, and post-fire proteoid recruitment suggested that fires in eastern coastal fynbos are not limited to any particular season, and for this reason managers do not need to be concerned if fires occur in any season. The ecological requirements for higher fire intensity may nonetheless be constrained by a need for safety. I articulated these findings into ecological thresholds pertaining to the different elements of the fire regime in eastern coastal fynbos, to guide adaptive management of fire in the Garden Route National Park. I also recommended a fire management strategy for the park to address the aforementioned operational considerations within the constraints posed by ecological thresholds. Finally, I highlighted further research and monitoring needs.

The Indo-Pacific bottlenose dolphin (Tursiops aduncus) is a common top predator along the KwaZulu-Natal coast. There have been several dietary studies done on this species, most of which have classed this predator as an opportunistic feeder preying on the most abundant prey species in that area. In KwaZulu-Natal (KZN) a dietary study was carried out in 1990, but there have been no followup studies to examine potential changes in diet. A number of bottlenose dolphins get entangled in the shark-nets set off the KZN coast each year. These dolphins provide a valuable source of information about different aspects of the ecology of this population. The aim of this study was to examine the diet of bottlenose dolphins off the KwaZulu-Natal coast with particular emphasis on temporal change in diet composition. The contents of 102 stomachs of bottlenose dolphins (58 females, 44 males) accidentally caught in the shark nets between 1985 and 2010 were analysed, and compared to historical data from dolphins caught between 1975 and 1986. A total of 94 different prey species from 37 families (32 fish families, five cephalopod families) were identified in the stomach contents. Teleosts made up the most important prey group in numbers and weight constituting 84 percent and 67 percent, respectively, while cephalopods made up 15 percent of the number and 30 percent of the weight of the prey items. Multi Dimensional Scaling (MDS) indicated that there is no resource partitioning between males and females, but significant partitioning between juveniles and adults. Six prey species contributed more than 60 percent of the weight of all prey species in the diet. These species are the most important according to their calculated index of relative importance score and are, in order of importance: Sepia sp. (cuttlefish)., Pomadasys olivaceus (piggy), Trachurus delagoa (maasbanker), Scomber japonicus (mackeral), Loligo sp. (squid). and Pagellus bellottii natalensis (pandora or red tjor-tjor). When compared to historic data, no changes in the main prey species has occurred, but the piggy has decreased in importance (1976-1986 percent IRI=45; 2000-2010 percent IRI=22), while the maasbanker has increased (1975-1986, percent RI=4; 2000-2010 percent RI=20). Stable carbon and nitrogen isotope analysis performed on tissue from the teeth of T. aduncus caught in the shark nets between 1970 and 2010 showed that there has been no significant shift in trophic level (δ15N) or in primary prey source (δ13C) over this time. The isotopic signatures of different tissues from an individual were used to create a base line and describe the differences in tissue signatures. Blubber was found to be significantly different from skin and muscle, which were found to be isotopically similar. Skin from remote biopsies can therefore probably be used in place of muscle for isotopic studies of bottlenose dolphins, while caution should be taken when using blubber. Given that T. aduncus is still listed as “data deficient” by the International Union for the Conservation of Nature (IUCN), and the potential for climate change and expanding fisheries to alter the distribution and abundance of their prey species, long term dietary studies such as this are important in order for their effective conservation and management.

The aim of this study is to gain a predictive understanding of the patterns and determinants of plant biodiversity in temperate, mesic grasslands of South Africa with a primary focus on the geographical area of the Eastern Cape. From a review of the literature on hypotheses explaining diversity (Chapter 2) it was possible to formulate a number of hypotheses that could be tested to explain species richness patterns in Eastern Cape grasslands. This thesis is organised so that each main chapter deals with a specific body of theory concerning the explanation of diversity patterns. A detailed description of the study area is provided (Chapter 3), including environmental variation and a description of major vegetation patterns. A summary is provided of grassland plant community patterns, as determined by phytosociological studies in the study area. A multivariate analysis of environmental variables was undertaken to determine which variables contributed the most towards explaining environmental variation in the study area and to determine whether any variables co-vary, a possible problem for any multivariate analysis in later chapters. Altitude produced one of the strongest gradients in the study area. There were a number of variables that were correlated with altitude, most notably temperature. Rainfall co-varied partially with altitude, but there was also a strong rainfall gradient perpendicular to the altitude gradient. A description of species richness, diversity and evenness patterns at the plot scale within different grassland plant communities of the Eastern Cape is provided in Chapter 4. To determine whether the environment acts differently on different growth forms, the contribution to species richness by different major growth forms is analysed. Furthermore, since the majority of literature attempts to explain diversity in terms of environmental factors, it was necessary to analyse the relationship between species richness and various environmental variables. The results indicate that there is high variation in species richness both within and among grassland communities. Forbs make the most significant contribution to overall species richness per 100 m2, followed by grasses. Variance in richness of all species together is not significantly related to environmental variables in mesic grasslands, but is significantly related to environmental variables in semi-arid grasslands. The result of greatest interest from this chapter is the fact that richness amongst different life-forms in the same place is explained by different environmental factors, indicating that the environmental factors that affect coexistence of species have a different effect on different life-forms. A classification of all the species of the dataset into plant functional types using a multivariate approach based on functional traits was conducted (Chapter 5). The grass species were classified into 16 functional types and the forbs into 14 functional types. The functional type classification provided the opportunity for undertaking analyses to develop an understanding of 8 the contribution by niche differentiation towards promoting species richness (Chapter 6). The results provide evidence of niche differentiation in the grasslands of the study area and also that niche differentiation promotes species richness in the grasslands of the study area. It was found that higher rainfall grasslands are less structured by niche differentiation than semi-arid grasslands. A regional / historical analysis is undertaken (Chapter 7) to investigate the relationship between the regional species pool and local richness, and the relationship between local richness and phytochorological diversity. Regional richness appears to have little effect in promoting local richness in grassland plant communities of the study area except at sites where there is high local richness. This provides an indication that regional richness only promotes local richness in the absence of local limiting factors. Phytochorological diversity promotes local richness, but mostly through diversity amongst species with narrow distribution ranges. Some theories ascertain that seasonal uncertainty may provide opportunities to species that would otherwise be outcompeted and thereby promote local richness. The degree to which seasonal uncertainty and seasonality promote local richness in the Eastern Cape grasslands was therefore investigated (Chapter 8). A weak relationship exists between these variables and local species richness in grassland communities of the study area, indicating that they do not promote niche differentiation to a significant degree in the study area. It is clear that in the grasslands of the Eastern Cape, environmental limiting factors are more important in semi-arid grasslands and species interactions are more important in mesic grasslands for structuring plant communities (Chapter 9, Discussion). Regional processes do not appear to be important in structuring local communities, but the analysis undertaken in this study shows that they may be significant when factors operating at the other two levels are overcome (species interactions and environmental limiting factors.

The Great Brak temporarily open/closed estuary was subjected to a drought during the spring and summer of 2009/2010 resulting in the mouth remaining closed for a prolonged period. According to the Great Brak Estuary management programme, the mouth of the estuary had to be open for a total of 308 days during spring and summer of 2009/2010, respectively, but was closed for almost the entire two years (693 days). The aim of this study was to assess monthly changes in the abiotic characteristics (salinity, temperature, oxygen, pH and nutrients) and the biotic responses of phytoplankton and macroalgae; identify sources of nutrient input into the estuary and determine the response of the salt marsh to water level and salinity changes. The results indicated that physico-chemical parameters were similar to that previously recorded during the closed mouth condition. However mouth closure combined with elevated nutrient concentrations led to a shift from rooted submerged macrophytes to one where either microalgae or macroalgae were dominant. Soluble reactive phosphorus concentrations were significantly higher in bottom compared with surface waters. There was a significant negative correlation with SRP and dissolved oxygen for the sampling period indicating potential release of phosphorus from the sediment during closed mouth conditions. Microalgal biomass increased in response to remineralised nutrients and freshwater pulses. Flagellates were the dominant microalgal group (21718 ± 3336 cells m l-1, p < 0.05) because of their morphological ability to migrate vertically within the water column. The macroalgal cover was highest during the closed mouth state but only during winter (August 2010) when temperatures were below 20 oC. Five major point sources of nutrient input into the Great Brak Estuary were identified during rainfall periods. Point sources 4 and 5 in the upper reaches of the estuary had the highest DIN input whereas point source 3 in the middle reaches of the estuary had the highest DIP input. As a result of the drought and low water level, the salt marsh was never inundated for longer than 3 months. Die-back of Sarcocornia decumbens (r 2= -0.62, p < 0.05) was related to smothering by dead macroalgae whereas dieback of Sporobolus virginicus was related to decreasing nutrient (r2 = 0.59, p < 0.05) and salinity (r2 = 0.55, p < 0.05) levels. The physico-chemical characteristics alone did not convey the true health status of the Great Brak Estuary for the duration of the sampling (April 2010-April 2011). The study showed that microalgae and macroalgae are valuable indicators of the status of the estuary. Therefore it is suggested that bio-indicators are incorporated into the management/monitoring plan in order to assist in improving the health assessment of the Great Brak Estuary.

The increasing number of proposed wind farm developments in South Africa provides an immediate reason to assess bat activity and diversity, with the expectation of developing measures to mitigate for negative impacts. The overall aim of this thesis was contribute to an understanding of the drivers of bat activity, relevant to wind turbines, and to determine what factors (exogenous – environmental, or endogenous – intrinsic physiological processes) contributed to observed bat activity patterns, at the Metrowind Van Stadens Wind Farm in the Eastern Cape of South Africa. A combination of active (mist netting) and passive (acoustic) bat monitoring techniques were used to determine free-ranging bat activity patterns (Chapter 3). A total of 889 bat passes were recorded over 323 detector nights from the beginning of May 2012 to the end of December 2012. The Cape serotine bat (82 per cent) and the Egyptian free-tailed bat (97 per cent) made up the majority of all bat passes recorded on site. Large variations in bat activity per month and per hour were apparent, with bat activity peaking in May 2012 and during the first few hours after sunset (18:00-23:00). Patterns in nightly, free-ranging bat activity at the site were modelled against various environmental conditions. Month, temperature, wind speed and an interaction between month and rainfall were the most significant predictors of bat activity, explaining 80 per cent of the variation observed on free-ranging bat activity patterns. A total of eight Cape serotine bats (Table 4.1.) were caught in mist nets on site and changes in the resting metabolic rate (RMR) of torpid (n = 6), and normothermic (n = 2) bats, over a 24 hr period, were measured and used to predict free-ranging Cape serotine bat activity (Chapter 4). Cape serotine bats showed a high proclivity for torpor in the laboratory and peaks in RMR were observed at 18:00 (0.89 ± 0.95 VO2 mℓ.g-1.hr-1) and again from 20:00-21:00 (0.89 ± 0.91 VO2 mℓ.g-1.hr-1). Peaks in RMR of torpid individuals coincided with peaks in the average hourly free-ranging activity of the Cape serotine bat, and RMR explained 33 per cent of the variation and was a good predictor of free-ranging bat activity (R2 = 0.2914). This study showed that both exogenous (Chapter 3) and endogenous (Chapter 4) factors drive bat activity in the wild. Although this dissertation was not intended for wind turbine management, the information presented on the biology and activity of bats is important for managing interactions between bats and wind turbines. By determining what factors influence bat activity, we are able to predict when bats will be most active and thus can develop mitigation measures to reduce the potential impacts that wind turbines will have on the bat community. In order to conserve bats and reduce potential bat fatalities from occurring at the site, mitigation measures should be concentrated to those times when bats are most active (May and during the first few hours after sunset – 18:00 to 22:00).