It is well documented that biomass (and catches) for many squid species varies considerably, and moreover, that recruitment strength is strongly related to the environment. The ramifications of catch fluctuations are significant and create uncertainty for resource managers and the fishing industry with the net result of increased risk levels of stock collapse, economic instability, long term investment; and for the semi-artisanal fisheries, socioeconomic hardship for the many fishers. Recruitmentâenvironment mechanisms which underpin biomass are region specific, and in the case of the chokka squid Loligo reynaudii, are unknown. This thesis addresses this knowledge gap and is fundamentally based on Bakunâs (1996)1 generalised triad of requirements for successful recruitment â enrichment, concentration and retention in the ecosystem. For chokka squid this implies that recruitment depends on the survival of paralarvae in terms of food availabilityâfeeding success (i.e. copepods biomass, density distribution and patchiness) and retention in the ecosystem. The nature of this investigation demanded a multidisciplinary approach comprising physical oceanography and biology, as well as a variety of scientific techniques. First a synthesis of basic ecosystem components for the domain in which chokka squid live (i.e. South Africaâs west coast and Agulhas Bank) was prepared using published and new data. It included bottom temperature, bottom dissolved oxygen, chlorophyll, and copepod abundance. Alongshore gradients of these indicated that the main spawning grounds on the eastern Agulhas Bank are positioned where bottom temperature and bottom dissolved oxygen are optimal for embryonic development. This location, however, appears suboptimal for hatchlings because the copepod maximum (food for paralarvae) is typically on the central Agulhas Bank some 200 km to the west. Data on currents suggest that this constraint may be overcome by the existence of a net west-flowing shelf current on the eastern Agulhas Bank, improving survivorship of paralarvae by transporting them passively towards the copepod maximum â a concept referred to as the western transport hypothesis or WTH (hypothesis 1). CTD (Conductivity, Temperature, Depth) data and a temporal analysis of AVHRR (Advanced Very High Resolution Radiometry) satellite imagery reveal the copepod maximum to be supported by a ââcold ridgeââ, a mesoscale upwelling filament present during summer when squid spawning peaks. In situ sea surface temperature (SST) data used as a proxy for cold ridge activity demonstrate considerable inter-annual variability of the feature, especially during El NiÑo-Southern Oscillation events. Negative linear correlations between maximum summer SST (monthly average) and squid biomass the following autumn (r2=0.94), and annual catch (r2=0.69), support the link between the ââcold ridgeâcopepod maximumââ and the early life cycle of chokka squid, and holds promise for prediction. Transport of squid paralarvae hatched on the inshore spawning grounds (<60 m) was also investigated using a bottom-mounted Acoustic Doppler Current Profiler deployed at 36 m on the Tsitsikamma coast (in the Tsitsikamma National Park). Analysis of 12 months of data showed that surface flow was mainly eastward (alongshore), with a maximum velocity of 115 cm sâ1 and an average of 24 cm sâ1. Generally, velocity decreased with depth, with a maximum bottom velocity of 65 cm sâ1 and an average of 10 cm sâ1. Data from a nearby thermistor array show that the water column was usually isothermal during winter (Julyâ September), with bottom flow in the same direction as the surface layer. In summer (DecemberâMarch), vertical stratification was most intense and surface and bottom flows differed in velocity and direction. Potential net monthly displacements calculated for three depths (5 m, 23 m and 31 m) indicate that passive, neutrally buoyant biological material (e.g. squid paralarvae) would likely be transported eastwards in the surface layer for eight of the 12 months, and would generally exceed distances of 220 km monthâ1. Displacement in the bottom layer was more evenly distributed between east and west, with net monthly (potential) transport typically 70â100 km, but reaching a maximum of 200 km. Wind-driven coastal upwelling prevalent during the summer, was observed to cause offshore flow for several days the surface layer of the coastal current resulting in potential displacement distances of 40 km from the coast. This mitigates eastward transport and potentially moves squid paralarvae in the Tsitsikamma current offshore and into the westward mid shelf current which flows towards the cold ridge. Realization that currents may not always be westward led to re-examination of the dependency of squid paralarvae on the cold ridge as the only rich feeding area on the Agulhas Bank. A synthesis of existing data and materials found that chlorophyll and copepods also exist at elevated levels on the thermocline in areas other than the cold ridge, and that currents therefore may not necessarily be so critical to connect hatching position with food. Moreover, varied current data suggested that currents may remove squid paralarvae from the Agulhas Bank ecosystem (hypothesis 2) through the leakage of shelf water into the Indian and Atlantic Oceans. Magnitude and timing of such a phenomenon will impair recruitment, cause biomass fluctuations, and ultimately affect catches. Leakage has been cited as the root cause of the sudden drop in annual squid catches experienced in 2001. To investigate the leakage hypothesis, a Lagrangian IBM (Individually-Based Model) coupled to a ROMS (Regional Ocean Model System) model was setup covering the west coast and Agulhas Bank to 24° E (St Francis Bay). Three simulations were performed for 12 model months using neutrally buoyant particles released from the seabed every second day on the mid shelf of the eastern, central and western parts of the Agulhas Bank. Boundary effects and resolution precluded the release of virtual particles on the inshore spawning grounds. Particles were given life spans of 40 days. Results demonstrated large particle losses from the eastern Agulhas Bank (76%) and the western Agulhas Bank (64%). In contrast few particles were lost from the central Agulhas Bank (2%) making this, in terms of the model, the most suitable place on the Agulhas Bank for spawning. Visualization of the ROMS outputs revealed that leakage on the eastern Agulhas Bank was caused by a cyclonic eddy resident in the Agulhas Bight. Similarly leakage from the western Agulhas Bank was caused by deep water cyclonic eddies in the adjacent Atlantic Ocean. A parallel study was also undertaken using four satellite tracked drifters released on the eastern Agulhas Bank to validate the ROMS-IBM experiments. Two drifters with drogues to 8 m were released on the inshore spawning grounds off the Tsitsikamma coast at the ADCP site. The other two drifters were released on the mid shelf where squid eggs had previously been found and âvirtual paralarvaeâ released in the IBM. One of the mid shelf drifters was tethered to a drogue at 70 m to measure advection in the bottom layer. Both inshore drifters were transported 70 km eastwards in the Tsitsikamma coastal current to Tsitsikamma Point. Here one beached while the other moved offshore onto the mid shelf and then southwards to leave the shelf 20 days after release. The surface drifter on the mid shelf was transported westward across the central and western Agulhas Bank (550 km) to leave the shelf after 58 days south of the Cape Peninsula. The deeper drifter also travelled westward, but remained on the shelf in the vicinity of the cold ridge and was recovered after 40 days and 100 km of the release position. Satellite SST and ocean colour images indicated frequent offshore flows of shelf water near the southern tip of the Agulhas Bank, as well as an intrusion of oceanic water onto the western Bank during this experiment. The latter caused an anti-cyclonic circulation which led to further leakage of shelf water from the inner central Agulhas Bank. The combination of drifters and satellite imagery in this experiment demonstrated that retention of chokka squid paralarvae in the Agulhas Bank ecosystem is not certain, even for the inshore spawning grounds, and that the risk would be less if paralarvae were found near the bottom. So far in situ sampling indicates they occupy the surface layer. Overall, this work has identified (1) environmental niches on the Agulhas Bank which the chokka squid life cycle has evolved to use (e.g. spawning grounds), (2) that the cold ridgeâ copepod maxima is a rich feeding ground on the Agulhas Bank and plays a role in chokka squid recruitment strength (biomass), and (3) that there is potential for chokka squid paralarvae to be advected off the shelf and removed from the Agulhas Bank ecosystem on the eastern and central Agulhas Bank possibly resulting in biomass crashes for the following year. Importantly, the cold ridgeâcopepod biomassâsquid biomass relationship has been quantified, and holds promise for prediction. Prediction will be further strengthened if, in the future, advective paralarval loss can be linked to early retroflection of the Agulhas Current.

Phytoplankton are a key component in the functioning of marine ecosystems and play a central role in the cycling of nitrogen and other elements. Metrics that can adequately represent the biogeochemical processes associated with phytoplankton diversity are needed in order to make use of remote sensing and modeling platforms. A single-value size proxy, effective diameter (Deff ), represents the mean volume to surface area ratio across the nano and micro plankton size fraction (2-200µm) in the southern Benguela, but has yet to be tested regarding its biogeochemical relevance. Cell size imposes overarching constraints on phytoplankton metabolism; there are therefore strong grounds for evaluating the usefulness of the metric (Deff ) in studies of nitrogen dynamics in diverse, natural assemblages. Three case studies were used to explore the nitrogen dynamics in naturally occurring assemblages and to evaluate the relationships between Deff and the uptake of the different sources of nitrogen. Two of the case studies comprised high biomass, harmful algal blooms observed off Lamberts Bay during an upwelling/downwelling cycle. The third case study used bi-monthly sampling over a full year in Saldanha Bay. The Lamberts Bay case studies involved blooms occasionally dominated by HAB-forming species: a mixotrophic ciliate, Myrionecta rubra, and a dinoflagellate, Prorocentrum triestinum. The nitrogen uptake rates followed the well observed pattern of high nitrate uptake by large cells and regenerated nitrogen uptake by small cells. Myrionecta rubra had a wide range of nitrate (O₃⁻ ) uptake rates (0.02-0.3 µmol N L⁻¹ h⁻¹). Prorocentrum triestinum showed slower rates of O₃⁻ uptake (0.01-0.2 µmol N L⁻¹ h⁻¹) and dominated in low O₃⁻ , stratified conditions. Diatoms were the most efficient utilisers of O₃⁻ and total nitrogen in these cases. The effective diameter was significantly related to the uptake rates of ammonium (NH₄⁺ ) (r=-0.54, p<0.005) and urea (r=-0.59, p<0.005), but not O₃⁻ (r=0.27, p=0.11). This was attributed to some instances of bi-modality in observed size distributions as well as potentially specialist nutrient uptake strategies employed by diatoms. The year-round data from Saldanha Bay indicated the system was diatom-dominated and was used to assess 1 how well Deff could represent the nitrogen uptake strategies employed by the diverse diatom assemblages. The Saldanha Bay system has O₃⁻ limited surface waters during summer, and light-limited bottom waters during winter. No significant relationship was found between Deff and the mass-specific uptake rates of the different nitrogen species in this data set. This was attributed to the complex shapes of the size distributions and the comparatively low biomass observed. Uptake kinetic experiments revealed high variability for maximum uptake rates (Vmax) and half saturation values (Ks) for both O₃⁻ and NH₄⁺ . For O₃⁻ : Vmax ranged 0.007-0.17 µmol N L⁻¹ h⁻¹, and Ks ranged between 0.2-42.5 µmol N L⁻¹. For NH₄⁺ Vmax was observed between 0.02-2.7 µmol N L⁻¹ h⁻¹; and Ks values ranged 0.1- 14.02 µmol N L⁻¹. Variability was observed in association with the availability of the ambient sources of nitrogen, but some variation was accounted for by the presence of different diatom species. From these three case studies it was concluded that the single-value size proxy was an adequate metric to quantify the uptake of regenerated nitrogen in scenarios of high biomass algal blooms. Such blooms are a pervasive feature in the southern Benguela Ecosystem. For lower biomass blooms, however, Deff did not adequately represent the nutrient dynamics of diverse diatomdominated assemblages. The variable shape of the size spectrum is an important factor in determining the rates of nutrient uptake and, in cases of bi- or multi-modality, this information could be lost when represented by a single descriptor such as Deff . It was subsequently hypothesised that size spectra could be used to accurately represent the nitrogen dynamics in diverse phytoplankton assemblages. This was tested by comparing the observed uptake rates of the three case studies to estimated uptake rates based on size spectra. Observed particle size distributions were used to estimate the uptake of O₃⁻ and NH₄⁺ , based on theoretical relationships to calculate size-dependent values of Vmax and Ks. Michaelis-Menten models were applied to measured ambient nutrient concentrations and particle size distributions, generating size-integrated estimates of O₃⁻ , NH₄⁺ and total N uptake rates. The variability in the estimated uptake rates was similar to that of the measured values. It was thus concluded that the representation of phytoplankton diversity by size spectra allowed modification of model parameters, such that improved estimates of uptake rates of O₃⁻ and NH₄⁺ could be obtained for a dynamic eutrophic environment.

Includes abstract. / Includes bibliographical references (p. 229-253). / This study investigates the climatology, intensification and ocean atmosphere interaction in relation to the passage of tropical cyclones (TCs) in the South-West Indian Ocean (SWIO). A Climatology of TCs in the SWIO including landfall in the area of Mozambique and Madagascar was developed for the 1952-2007 and 1980-2007 periods.

Includes bibliographical references (p. [147]-158). / Includes abstract. / The oceanic structure and circulation of the Agulhas Bank, the very wide continental shelf area off South Africa, has been explored in this thesis. The Agulhas Bank is a complex ocean region influenced by shelf processes as well as a nearby western boundary current, the Agulhas Current on its eastern margin and the Benguela Upwelling system on its western margin. In addition, a cold water feature, known as the cool ridge, has been observed on the Eastern Agulhas Bank. A consistent dynamical description is not available but it is commonly observed as a south-westerly flow of cold water in the upper water column, roughly following the 100m isobath and extending seawards off the coast. The formation of the cool ridge has also been investigated in this thesis by means of a numerical ocean model. Previous studies on the Agulhas Bank have been limited temporally and / or spatially. Thus, the Agulhas Bank as a whole has been inadequately sampled to provide a comprehensive representation. In order to remedy these deficiencies, the Regional Ocean Modeling System (ROMS) was used to study the Agulhas Bank dynamics. A large-scale model of the surrounding oceans, the SAfE (South African Experiment) configuration of the ROMS model, was used to force a one-way embedded finer-resolution model over the Agulhas Bank. This produced, 8 years of model data at an approximate horizontal resolution of 8km and 32 vertical terrain-following levels. Two main experiments were performed to understand the nature of the Agulhas Bank. Firstly, the "Reference Experiment" derived a seasonal ocean climatology of the Agulhas Bank. Secondly, the "No Agulhas Experiment" was carried out in an approach in which the Agulhas Current was removed from the shelf edge. A comparison of these two experiments yielded the influence of the Agulhas Current on the Agulhas Bank. The ROMS model was able to reproduce the main observed seasonal structure and circulation of the Agulhas Bank as well as the cool ridge. The Agulhas Bank showed marked seasonality, with its two-layer structure being significantly influenced by the Agulhas Current. The direct influence of the Agulhas Current on the Bank occurs on the Outer Agulhas Bank by the Agulhas Current itself or an Agulhas Current filament. Ekman veering by the interaction of the Agulhas Current with the bottom topography on the slope of the eastern Agulhas Bank advect cool water vertically onto the Bank. This strengthens the thermocline from below, in contrast to surface warming by solar insolation in summer and the Agulhas Current in winter. Cold waters, upwelled over the shelf edge, indirectly affect the greater Agulhas Bank by their advection by the predominantly westward mean currents. The most significant influence of the cold shelf-upwelled waters are in bringing cold waters to shallower depths over most of the Agulhas Bank. This may influence the waters that upwell at the coast, which on the eastern Agulhas Bank (without the Agulhas Current) are trapped under a thick warm surface layer.

No description available.

There is increasing interdisciplinary interest in phytoplankton community dynamics as the growing environmental problems of water quality (particularly eutrophication) and climate change demand attention. This has led to a pressing need for improved biophysical and causal understanding of Phytoplankton Functional Type (PFT) optical signals, in order that satellite radiometry may be used to detect ecologically relevant phytoplankton assemblage changes. This understanding can best be achieved with biophysically and biogeochemically consistent phytoplankton Inherent Optical Property (IOP) models, as it is only via modelling that phytoplankton assemblage characteristics can be examined systematically in relation to the bulk optical water-leaving signal. Harmful Algal Bloom (HAB) conditions in the Southern Benguela and various inland waters of Southern Africa require continuous observation by satellite due to the potential for significant negative environmental impacts. Current oceanic bio-optical models do not perform well in elevated Chlorophyll a conditions, but the high biomass conditions of Southern African inland and coastal waters lend themselves extremely well to the development of phytoplankton IOP models as the water-leaving signal is overwhelmingly phytoplankton-dominated. An initial validation of a new model of Equivalent Algal Populations (EAP) is presented here, and comparison is made with two prominent phytoplankton IOP models. The EAP model places emphasis on explicit biophysical modelling of the phytoplankton population as a holistic determinant of IOPs. By necessity due to its origins in highly scattering waters, a distinctive attribute of the EAP model is its comprehensive handling of the spectral and angular character of phytoplankton scattering. This emphasis is shown to have an impact on the ability to retrieve the detailed phytoplankton spectral scattering information necessary for PFT applications and to successfully simulate waterleaving reflectance across wide ranges of physical environments, biomass, and assemblage characteristics. The accurate description of a water body's Volume Scattering Function (VSF), and hence its phase functions, is critical to the determination of the constituent IOPs, the associated spectral water-leaving reflectance, and consequently the retrieval of PFT information. The EAP model offers the ability to provide phytoplankton population-specific phase functions, unveiling an opportunity to gain further insight into the causality of the PFT signal. This is a new modelling capability, and its application in case studies and sensitivity analyses has resulted in improved understanding of the PFT/assemblage-related signal, in particular the discovery that phytoplankton spectral scattering is the primary driver of the PFT-related signal. The required thresholds of PFT detection with respect to biomass, IOP budget and assemblage effective diameter are quantified. Key findings are that the backscattering-driven signal in the 520 to 600 nm region is the critical PFT identifier at marginal biomass, and that while PFT information does appear at blue and red wavelengths, it is compromised by biomass/gelbstoff ambiguity in the blue and low signal in the red, due primarily to absorption by water. The key findings and recommendations are hoped to provide considerable insight into PFT approaches with regard to in situ observation, sensor development and algorithm optimisation for the next generation of PFT investigations.

The ROMS is used to investigate the oceanic circulation in the Delagoa Bight region, near the southwestern end of the Mozambique Channel. The model is initially configured at a horizontal resolution of 1/10° (9.9km) over the domain 30.1-43°E,22.13-30.8° S. Subsequent configurations nested an inner grid of resolution 1/30° (3.3km) over the region 32.43-36.43°E,24.09-27.71°S Several sensitivity experiments were performed with and without the inner grid or with or without tidal forcing. Using only the outer coarse resolution grid, the first experiment (DELAGI) does not include tidal forcing whereas tides are included in the second (DELAG-II) experiment. DELAG-III and DELAG-IV both use the inner higher resolution grid but exclude and include tidal forcing respectively. The model was evaluated against observations, namely the WOA, Pathfinder SST and AVISO SSH. The results showed that ROMS adequately resolves the oceanic features in the region, namely the pathways of the anticyclonic eddies from the northern Mozambique Channel and from Madagascar, and the instances when the DBLE is present or absent. The model is also able to reproduce the main water masses and their sources in the region. Water masses found in the centre of the Bight enter through the northeastern sector, either by intrusion of pulses or instabilities of the southwards flowing current. When the DBLE is well established, upwelling is likely to contribute to the water masses in the lower layers. The transport of water towards the Bight from the east was found to be less than that from the north. The model also succeeds in representing the thermocline structure of the DBLE but it fails to capture the local salinity maximum. When tidal forcing is included, the speed of the flow close to the coast increases. The model also revealed the influence of the Inhambane Cyclone on the Delagoa Bight as well as on the region to its south. This cyclone, which is generated in the flow near Inhambane, is similar to Natal Pulses which occur in the Agulhas Current. An eddy detecting and tracking system was used with both the model outputs and VISO SSH to determine the statistics of the DBLE, namely its dimensions, amplitudes and life-times. A maximum radius of 59.52 km, life span of 126 days and an amplitude of 27.27 cm were found. It was also demonstrated that this feature is generated northeast of the Bight. When the inner grid was included in the simulations, anticyclonic features were generated within the Bight with a maximum diameter of 85.4 km and life span of 12 day. These anticyclonic features dominate the circulation when the DBLE is absent (less than 30 percent of the total period of the simulation). The possibility of several cyclonic cores existing simultaneously in the Bight was also demonstrated. Two cores were found with life spans of more than 12 days. When the number of cores is greater than two, they tend to be short lived. Analysis of energy conversion rates showed that the generation of both the DBLE and the Inhambane cyclones is mainly by barotropic instabilities, although in both regions of their generation, weak baroclinic instabilities were also found.

Includes bibliographical references (leaves 228-270). / Subtropical ocean gyres are typically characterised by low carbon export into the deep ocean. However, due to their large area, even relatively small average carbon export may be globally significant. Strong correlations observed between deep-sea organic carbon, calcite and opal suggest that mineral phases may enhance the export and survival of organic matter as it sinks (the 'ballast effect'). However, the processes underlying these correlations are not well understood and remain key uncertainties in models that predict global carbon cycling. To better constrain carbon and mineral export from the surface ocean of subtropical gyres, radioactive disequilibria between 234Th and 238U were used to estimate fluxes of particulate organic carbon (POe), calcite and opal in the North and South Atlantic subtropical gyres.

Includes abstract. / Includes bibliographical references (leaves 109-124). / The impacts of the Benguela Niño on southern African rainfall and circulation are investigated using an atmospheric general circulation model. The model used is the United Kingdom Met Office Hadley Centre Atmospheric General Circulation Model version 3 and experiments using idealizations of observed regional and remote SST anomalies during various Benguela Niño events were performed. It is found that SST forcing in tropical South East Atlantic induces a regional baroclinic response and that a Benguela Niño is capable of forcing anomalous wet conditions over western Angola on its own, via changes to uplift and evaporation over the SST forcing. It is also capable of forcing anomalous rainfall much further inland when the intensity is increased. An experiment with the tropical South East Atlantic SST anomaly shifted slightly further north produced a larger circulation and rainfall response in the model. Additional experiments with various SST anomalies in the South West Indian Ocean/central equatorial Pacific combined with those in the South East Atlantic were performed. These experiments are motivated by the fact that equatorial Pacific/South Indian Ocean SST anomalies of varying signs often occur at the same time as the Benguela Niño Events. The results suggest that depending on its sign, magnitude and location, SST forcing from the South West Indian Ocean may augment or oppose the southern African rainfall anomalies occurring during a Benguela Niño event to varying degree.

Includes bibliographical references (p. 257-271). / This thesis is an investigation of cut-off low pressure systems over South Africa. These weather systems have been responsible for many of the flooding disasters that have affected South Africa, particularly the coastal regions, over recent decades. The thesis has two main objectives, namely, to construct a 30-year climatology of cut-off lows over South Africa, and to further understanding of the evolution of the low-level flow that leads to these systems producing extreme quantities of rainfall.