Includes bibliographical references. / The aims of this study were to investigate elements of the demographics, fine-scale habitat use and site fidelity of sevengill sharks. The study was conducted at an aggregation hotspot within a Marine Protected Area, near Miller’s Point in False Bay, South Africa.

Bibliography: pages 107-121. / A Monte-Carlo type model has been developed to investigate the importance of passive transport by currents above the thermocline for anchovy recruitment off South Africa. Simulation studies indicate that mean year-class strength of Cape anchovy is relatively robust to altered advective processes off South Africa. This occurs despite the fact that changed flow alters the likelihood of offshore advection and hence losses of anchovy from the system. Two different approaches have been taken to address the effects of altered advection, and the applicability of each is discussed. One approach involves altering westward advection in proportion to the mean current field (derived from Acoustic Doppler Current Profiler measurements), and the other, altering westward and northward advection by the addition of fixed offshore current velocities. The proportional approach did not affect year-class strength significantly, whereas the other approach, which incorporated large changes in the flow field, yielded statistically significant differences in predicted year-class strengths between advection scenarios. Reduced flow in the latter approach led to a mean year-class strength 2.7 times stronger than a proposed base flow scenario (which incorporated westward and northward drift in addition to the ADCP currents), whereas enhanced flow resulted in a mean year-class strength of similar magnitude to that of the base flow scenario. Changed flow may alter the geographic distribution of eggs and larvae, which might in turn influence recruitment of young-of-the-year anchovy to the South African purse-seine fishery. The north-flowing shelf-edge jet current plays an important role in transporting anchovy eggs and larvae from spawning grounds in the south to nursery areas frn1her north along the west coast of South Africa. Enhanced model advection westward and norward from the spawning grounds in the south serves to transport anchovy into the region of the jet current. However, advection into unproductive waters offshore is also enhanced and prevents good recruitment under these flow conditions. On the other hand, reduced westward and northward advection in the model, shown through wind records to characterise El-Nino years in coastal areas of South Africa, serves to retain anchovy reproductive products and often transports young anchovy into coastal areas, preventing offshore loss. Therefore the advection model suggests that good year-class strengths (in terms of numbers) are likely to be supported in years when westward and northward advection are reduced. A further reduction in westward advection may be less favourable by causing advective losses offshore along the south coast of South Africa. This may be viewed in terms of an "optimal environmental window" hypothesis, where reduced westward advection is favourable for anchovy survival off South Africa, but further reduction of westward advection as well as enhanced westward advection appear unfavourable. It is concluded that although passive transport, of anchovy in South African waters is relatively robust, it may account for a substantial proportion of recruitment variability.

The parapagurids Sympagurus dimorphus (Studer, 1883) and Parapagurus bouvieri (Stebbing, 1910) dominate South African deeper-water benthic invertebrate communities. Samples from the 2016 DAFF bi-annual hake demersal research trawl surveys provided data on population structure and morphology. Since these species forego the use of "traditional" gastropod shells in favour of symbiotically-associated Epizoanthus pseudoshells, these data were also used to investigate parameters of pseudoshell association. S. dimorphus to have a female-biased sex ratio of 1:1.5, and P. bouvieri of 1:1. The proportion of males in both species increased with cephalothoracic shield (CL) length. S. dimorphus males had positive allometric major right chelipeds, while P. bouvieri cheliped lengths were positively allometric in males, and negatively allometric in females. All collected P. bouvieri occurred in pseudoshells, but S. dimorphus showed a real shell occupancy rate of 3.39% (n = 38). A positive relationship was found between CL size and real shell occupation frequency (r2 = 0.96). S. dimorphus pseudoshell zoanthid polyps increased in number as a function of both total shell volume (R2 = 0.73) and colony height (R2 = 0.61), with polyps present on every available part of the shell. No such relationship was found for P. bouvieri, presumably because the polyps are distributed in a linear pattern around the back and outside of the pseudoshell. Juvenile parapagurids appear to select for Naticidae original shells, with Euspira napus being the dominant original shell species for both S. dimorphus (45.45 %) and P. bouvieri (39.00 %). Questions are raised about whether these hermits exist along an evolutionary continuum of traditional shell reliance, whereby the species utilise their pseudoshells differently - P. bouvieri appears to show a stronger mutualistic symbiosis with its exclusive pseudoshell species than S. dimorphus, and may have lost the ability to change shelters entirely, utilising the pseudoshell as a brood pouch covering only the abdomen, rather than as a shell in which to retract to escape predation. Data from 1987 - 2014 surveys were used to map the distribution patterns and densities of these species in the southern Benguela, and to investigate changes in the populations over time. As it stands, no correction factor need be applied to South African parapagurid catches as a result of the 2003-2004 change in survey gear. However, further work is needed. S. dimorphus occurred at a significantly greater mean sample mass (287.88 kg.km⁻²) than P. bouvieri (31.37 kg.km⁻²). The bulk of sample mass occurred within the West Coast Benguela upwelling region, with very few parapagurids caught over the Agulhas Bank, which may be too shallow for either species, given that S. dimorphus showed a preferential depth range of 200 - 299 m (range 30 - 814 m) and P. bouvieri of 400 - 499 m (range 62 - 700 m). The highest mean sample masses occurred on muddy sand (61.00 kg.km⁻²) and sandy sediments (45.68 kg.km⁻²). Overall catchability (mean = 43.35%) has remained constant over the 23 years (R2 = 0.13) with a slight trend towards increasing catchability over time, which may correspond to a change in survey design.

Includes abstract. / Includes bibliographical references (p. 53-59). / In 2007 the local government of Rodrigues gazetted four marine reserves in the north of the island based on knowledge and insights from stakeholders, mainly from the fishing community. In order to verify the stakeholder-based design, a marine reserve network was designed using Marxan, a systematic conservation planning programme.

The presence of white sharks Carcharodon carcharias in False Bay, South Africa follows a clear seasonal cycle. In the austral winter months, white sharks aggregate around Seal Island with female sharks in particular moving to the inshore areas in summer. What triggers these fine scale migrations at one of the largest white shark aggregation sites in the world remains a subject of debate. Previous research identified the environmental variables, water temperature and lunar phase, as significant influences on white shark presence inshore. In this study, I attempt to investigate the influence of prey fish availability on the presence of white sharks at two inshore areas of False Bay, namely Muizenberg and Fish Hoek. In addition, I explored the influence of sea surface temperature (SST), El Niño/Southern Oscillation (ENSO), dolphin presence, diatom patch presence and the occurrence of purse-seine fishing (treknetting) on shark presence. A total of 1209 shark sightings were recorded from 1 January 2006 - 31 December 2015 with 742 sightings at Muizenberg and 467 at Fish Hoek. I used randomization tests to determine whether the occurrence of shark sightings was randomly distributed across the environmental and biological variables and determined the dependencies between the predictor variables. I confirmed that shark sightings were clearly seasonal and sightings peaked between 17.49 - 18.57°C. Fish presence revealed a similar peak at 17.94°C. Randomization tests indicate that a spotter is 66% more likely to detect a shark when fish are present, which supports my prediction of a prey mediated cue. The occurrence of treknetting was also found to be non-randomly linked to white shark presence, increasing the probability to spot a shark to 38% instead of 20% at random. The presence of diatom patches was found to be negatively correlated with SST; with numbers increasing in colder waters. The presence of dolphins, as other predators in False Bay, didn't show a significant pattern with any of the analysed variables. The influence of ENSO on the presence of white sharks was also hypothesized and the data revealed that there was a significant influence of weak negative ENSO values on their inshore occurrence. In addition to providing a better understanding of the ecology of white sharks in False Bay, my results can also be used in shark safety and education programs to reduce the risk of shark human conflict.

Includes bibliographical references. / The high latitude coral communities of southern Africa suffered minimal impacts during mass bleaching events in the recent past. However, during the 2005 warm-water anomaly in the southern Indian Ocean, coral bleaching reached unprecedented levels. There is surprisingly little known about the fate of bleached corals, which may either regain their zooxanthellae and recover, or may die, in which case they generally become overgrown by macroalgae. The nature and dynamics of this algal overgrowth are not well understood. This study was done on Two-Mile Reef, Sodwana Bay, located in the iSimangaliso Wetland Park, a World Heritage Site. The first aim was to investigate the abundance and diversity of benthic algal communities colonising different hard substrata (comprising bleached digitate, brain and plate coral assemblages, and beach rock). The second was to compare the algal communities colonising various artificial hard substrata. The third was to document the species of non-geniculate coralline red algae found on the natural hard substrata during sampling. Fieldwork was carried out during the marine autumn (March) and spring (September) of 2010 using SCUBA. A total of 90 quadrats (10 cm x 10 cm) were sampled and the underlying substratum was recorded and classified. A Braun-Blanquet scale was used to assign cover-abundance values to each species within each quadrat. Additionally, the relative cover of different types of substrata was estimated using line-point intercept methods. Multivariate analysis (detrended correspondence analysis) and cluster analysis (complete linkage Bray-Curtis) were used to show how substrata and season relate with respect to their seaweed flora. Additionally, Kruskal-Wallis nonparametric tests with pairwise Mann-WhitneyU-tests were used to examine differences in macroalgal assemblages among substratum types.

No description available.

Ultrastructural studies of meiosis, as well as aspects of vegetative and reproductive development, in three species of red algae were conducted. An account of meiosis in Dasya baillouviana is presented and represents the first comprehensive account in red algae. In early prophase I, the nuclei of tetrasporangia have, first, axial cores and, later, synaptonemal complexes as well as an electron-lucent space of undetermined significance. In later stages, chromatin relaxes, presumed RNA synthesis occurs and cell size and organelle number greatly increase. Electron-dense bodies were found in the cytoplasm and compared to nuage, a substance found to be RNA in other organisms. Meiotic events from prometaphase I to telophase II were similar to accounts published for mitosis in higher red algae. A distinct interkinesis was present, a finding which disproved an earlier interpretation of uninuclear meiosis in red algae. Also of interest were large accumulations of smooth ER at the division poles and the presence in post-division of intranuclear structures thought to be polycomplexes. In addition, post-division nuclei were found to move simultaneously to the center of the cell, associate, possibly fuse and then move back prior to cytokinesis. This report recognizes, for the first time, the fibrous vacuole associated organelle (FVAO), a new organelle in red algae believed to give rise to fibrous vacuoles and possibly to process the contents of the vacuoles. of taxonomic significance was the correct identification of elements of the FVAO previously misinterpreted as intracellular flagella. A three-dimensional reconstruction of the chondriome of red alga (Rhodella reticulata) was made for the first time. The chondriome consisted of one large, ramifying mitochondrion with two very small fragments. Two other serially-sectioned cells were traced but not reconstructed. All the chondriomes were much like those found in other unicellular organisms from many phyla. This research indicated that although red algae have some very unique characteristics, cell morphology and development as well as characteristics of nuclear division do not indicate exceptionally primitive organisms.

Cultural eutrophication (CE) is the allochthonous input introduction of a quantity of matter, such as sediments, organic material, or nutrients, into a water body over the pre-anthropogenic (natural) levels. In most coastal estuaries CE has come to refer primarily to an increase in the concentration of phyto-nutrients. CE has been identified as the cause of very graphic phenomena such as hypoxia and fish kills. In this work I examine the potential for CE to alter the composition of the primary producer community and potentially alter or disrupt the benthic food web, using Macoma balthica as an indicator species. A series of surveys and experiments identified that clams in areas with greater than average nutrient concentrations had lower health, slower growth, and greater non-predatory mortality than clams in less eutrophic areas. Primary production, as estimated from chlorophyll a concentration, was greater at higher nutrient locations while the health and growth of clams was lower. The phytoplankton community in the more eutrophic areas had a lower proportion of diatoms relative to dinoflagellates. A biochemical analysis of clam tissue indicated that the clams from the less nutrient rich sites had a greater proportion of Eicosapentaenoic acid (EPA) relative to other fatty acids. Diatoms are rich in EPA compared to dinoflagellates. Thus, we hypothesize that CE induced shifts from diatom based production toward dinoflagellates may be limiting trophic transfer due to a lack of EPA. Using a series of models we were able to predict that trophic disruption could significantly reduce the scope for growth of the blue crab, Callenecties sapidus . Thus it is possible that the CE induced changes to primary producer community could disrupt the food web creating a trophic bottleneck.

Carbonate sands are an integral part of coral reef environments but their role in the cycling of matter in the reef is understudied. Methods such as micro-profiling, core incubations, in-situ chamber incubations, and aquatic eddy covariance measurements can be used to study solute fluxes and biogeochemistry of these sediments. The interfacial fluxes can reveal production and degradation processes in the sands and thereby provide key information on the role of these sediments in the cycles of matter in the coral reef. To date, the eddy covariance technique is the most advanced method for measuring the solute fluxes noninvasively. The traditional eddy covariance method employs a solute sensor (e.g. oxygen sensors) and an Acoustic Doppler Velocimeter (ADV). We developed a triple O2-sensor-eddy covariance instrument that provides a hardware solution minimizing time-lag errors arising from the physical separation of the O2 sensor and the velocity sensor measuring volume, which affects the measurements in the traditional instrument. These errors are most pronounced in environments with currents and waves, and are difficult to correct through data processing. Oxygen flux measurements with the aquatic eddy covariance technique in benthic environments with wave action and/or highly varying current direction can lead to erroneous flux estimates due to the transient time-lag between the velocity and oxygen concentration measurements. By positioning the oxygen sensors in the same horizontal plane and with 120 degrees radial spacing around the center point where current flow is measured, synoptic oxygen measurements are recorded with opposing time-lags. Averaging of the three sensor signals at each time point produces an oxygen concentration closely corresponding to that at the flow measuring point, and thus, at a position that does not require time-lag correction. Field tests with the new instrument in a coral reef sand flat affected by waves demonstrated the advantages of the new instrument setup, and simultaneously allowed evaluation of the magnitude of errors that are associated with the traditional time-lag correction. Model simulations of the 3 sensors system in a known oscillating oxygen distribution field suggests that the new 3 oxygen sensor eddy covariance (3OEC) system can reduce time-lag error by at least five-fold. We conclude that the new system can improve oxygen flux estimates significantly, while simplifying the processing of the aquatic eddy covariance data. Another factor that may impose errors in the calculated fluxes by the aquatic eddy covariance method is the slow response of the solute sensors which may dampen the recorded dataset. To overcome this issue, we developed an instrument by which reliable and reproducible measurements of the response time of the sensors is possible. We used this method to select the sensors that we installed on our improved eddy covariance instrument. We used the improved aquatic eddy covariance instrument in a Florida coral reef sand flat to quantify benthic oxygen fluxes as proxy for benthic metabolism. The non-invasive measurements characterize the carbonate sands as sites of intensive organic matter production and consumption, and underline their dependency on key environmental drivers such as light, water current velocity, and significant wave height. The positive response to light and increasing light intensity were characterized by large temporal dynamics even at ~9 m water depth. Daytime fluxes reached 2.3 ± 2.0 (Mean ± SE) mmol m-2 h-1 and nighttime fluxes -2.0 ± 0.7 (Mean ± SE) mmol m-2 h-1. Spring deployments indicated net autotrophy of the sedimentary environment, while summer and winter measurements implied a metabolic balance. During summer, an increase in bottom currents correlated with an increase in sediment oxygen uptake during daytime and nighttime, reflecting enhanced benthic organic matter mineralization activity during the warm season. The oxygen fluxes reveal their role in the reef sands as hotspots of benthic carbon cycling. / A Dissertation submitted to the Department of Earth, Ocean, and Atmospheric Science in partial fulfillment of the requirements for the degree of Doctor of Philosophy. / 2019 / November 8, 2019. / Aquatic eddy covariance, Benthic oxygen flux, Carbonate sediments, Ocean engineering, Permeable sediments, Sediment biogeochemistry / Includes bibliographical references. / Markus Huettel, Professor Directing Dissertation; Janie Wulff, University Representative; Amy Baco-Taylor, Committee Member; Peter Berg, Committee Member; Sven Kranz, Committee Member; Kevin Speer, Committee Member.