Includes abstract. / Includes bibliographical references (p. 72-84). / The reproductive biology of the commercially exploited fish St. Joseph or Cape Elephant fish Callorhinchus capensis, is described based on 173 fishes caught by gillnets in October 2010 from Velddrif along the west coast of South Africa. Segregation by sex was reflected by a difference in sex ratio from 1:1 for females to males fishes collected from the fishery (?2 = 16.23, d.f. = 1, p = 5.6 x 10-5).

Includes bibliographical references. / It has been shown that sardine and anchovy populations in the southern Benguela have alternated dominance during the past 50 years, and a frame based model that explores the usefulness of frames within a marine ecosystem by modelling sardine/anchovy dominance shifts in the southern Benguela has been developed in a previous study. The existing model contains a sardine population sub-model without age structure and the inclusion of age-structure allows age-related dynamics to be incorporated in the model design. In this mini-dissertation, an age-structured sardine population is implemented in the frame based model by using the sardine assessment model from the joint sardine/anchovy Operational Management Procedure, OMP-02. Methods for incorporating biological processes - such as the hypothesised school trap phenomena - that were present in the existing frame based model into the OMP-02 equations are developed, critically analysed and tested.

Includes abstract. / Includes bibliographical references. / The life history strategy of the Cape horse mackerel, Trachurus capensis in the southern Benguela, relating to distribution and reproductive ecology, was re-evaluated. It is possible that certain aspects of the species' life history may have changed since the previous assessment by Barange et al. (1998), as a result in this study, the conceptual hypotheses about the life history of horse mackerel proposed by Barange et al. (1998) were re-evaluated and possible revisions proposed. Distribution patterns were analysed using acoustic and midwater trawl data (1997-2010), and demersal trawl data (1984-2011) collected during biannual surveys of pelagic and demersal fish.

The Sofala Bank in the western Mozambique Channel is an essential habitat for shallow-water penaeid shrimps. It supports an important multi-sector and -species fishery, with Fenneropenaeus indicus and Metapenaeus monoceros (banana shrimp) being the two main target species. Over the past decade this valuable resource has been declining, which has been attributed to environmental changes, but no conclusive evidence has been found. This PhD thesis aims to understand the interactive roles of biophysical processes on recruitment of banana shrimps, par ticularly their larvae on the Sofala Bank. It is hypothesized that shrimp larvae can be advected offshore by passing mesoscale eddies to regions where they are unable to survive and are thus lost. In the absence of both physical and biological observations, a modelling approach is used. A high-resolution, nested, coastal, Regional Ocean Modeling System (ROMS) of the Sofala Bank is developed. In general, the model agrees well with available observations and literature. The ROMS outputs and self-organizing map analysis indicate that the shelf circulation, structure and river plumes are strongly influenced by the highly energetic o_shore eddy activity. A biophysical, individual-based model (IBM) coupled to the ROMS was developed for early life stages of banana shrimps on the Sofala Bank. The IBM uses spawning patterns identified from analyses of both commercial and research survey data. Simulations indicate that shrimp larvae are lost offshore by entrainment in mesoscale eddies at inter- and intra-annual scales and eddies therefore are unlikely to produce a continuous declining in the catch. In contrast, these eddies induce onshore transport of larvae, promoting coastal settlement, compared with periods without eddies. Locations for simulated larval coastal settlement are identified: northern, central and southern. The roles of tides and larval diel vertical migration in influencing simulated larval settlement success was not conclusive; further research considering a selective tidal stream is needed. Simulated larvae were sensitive to low lethal temperature and river plumes. A conceptual model for the Sofala Bank circulation and another for the banana shrimps, early life history dynamics are proposed based on the results of this thesis.

Phytoplankton is an important component of the oceanic carbon cycle, and deriving a good estimate of its carbon biomass (Cphyto) at ocean scale is difficult due to the lack of automatic sampling procedures. This is particularly difficult in the Southern Ocean, where winter conditions limit the sampling. This study explored the opportunity of using a high resolution data from the glider tracks in the Sub-Antarctic Zone of the Southern Ocean. The data consisted of particulate backscattering and chlorophyll and four different methods of estimating phytoplankton carbon were used, three of them based on backscattering (named 30%POC, B05 and M13) and one on chlorophyll (S09). The methods are different in their empirical formulations and source of original data. Three methods showed similar results despite the fact that one of them makes use of chlorophyll to derive Cphyto. Method M13 doubles that of the 3 other methods (~80mg C m-³ vs 40-50 mg C m-³). It was observed that discrepancy between M13 and the other 3 methods decreases with depth and when biomass was low (~0.25 mg Chl-a m-³) e.g., at depth 80 m. Investigating the drivers of variability in chl-a:C phyto ratios with depth and MLD shows little response and highlighted the need for more research in this region. Although M13 has a very low chl-a:Cphyto ratios, the range of variability was similar to that of the 30%POC and B05 methods and likely driven by variability in light and Fe limitation and changes in community structure. Despite a similar magnitude, the S09 method show a tight constrain in chl-a:Cphyto ratios that were methodologically driven and thus less sensitive to physiological adjustments in cellular chl-a:Cphyto ratios. The analysis also confirms that each oceanic region has factors that drive their variability and care needs to be taken when applying a method that was derived from one oceanic region to another.

The Cape Elephant fish (Callorhinchus capensis) and two common sand shark species (Rhinobatos annulatus and Rhinobatos blochii) were caught off False Bay and Saldanha Bay and surveyed for their parasite community in 2013 and 2014. The surveys were used to build species accumulation curves (SAC) and calculate biodiversity indices, particularly, rarefied species richness, Shannon Weiner's diversity index, Simpson's index and Pielou's J index. The biodiversity indices were correlated with the host's biological data and parasite infection data, to determine the parasite community structure and provide insight into the host's community structure. The parasites identified in C. capensis (n=19) include a cestode (Gyrocotyle plana), two monogeneans (Callorhynchicotyle callorhynchi and Callorhinchicola multitesticulatus) and an isopod (Anilocra sp.). The cestode was the most prevalent at 68.4 % and the monogenean, C. callorhynchi was the most abundant (1.68 ±0.78) and had the highest infection intensity (4.00 ±1.45). The SAC and biodiversity measures indicate a uniform parasite community across the host population, suggesting a highly interactive shark community. Conversely, Rhinobatos annulatus (n=19) and R. blochii (n=17) had very limited parasite infection with two species of nematode found infecting the stomach (Proleptus obtusus) and encysted in the kidneys (Ascaris sp.) and a copepod species (Clavelottis sp.) found infecting the gills. Proleptus obtusus was the most prevalent (31.6 % and 29.4%), the most abundant (1 ±0.37 and 3.68 ±2.76) and had the highest mean infection intensity (3.17 ±0.4 and 14 ±1.5). A cestode (Trilocularia sp.) was found infecting three specimens of R. annulatus from False Bay. The SAC and biodiversity indices combined with the limited parasite infection indicate a non-uniform parasite community across the host population, suggesting an isolationist population. Within the parasite community discovered, a potential biological indicator for heavy metal accumulation was identified to determine the levels of heavy metal pollution within these two anthropogenically impacted bays. Gyrocotyle plana and Proleptus obtusus were chosen as potential indicators due to their high prevalence and the close relationship they have with their hosts. The results support the use of higher trophic level animals as biological indicators. The results also indicate that G. plana is an incredibly good accumulator of certain metals, particularly As (4073.52 ± 5561.54 μg/g), Mn (522.16 ± 578.21 μg/g), Pb (64.87 ± 101.7 μg/g), Ti (1821.42 ± 1348.16 μg/g), and Zn (12439.57 ± 9743.60 μg/g). Unfortunately water and sediment samples were not tested, however, concentrations were compared to baseline values, and the accumulation of these metals are orders of magnitude above the surrounding environment. Proleptus obtusus did not significantly accumulate metals from its surrounding environment. These results show that parasites can be used to infer their own and their host's community structure and confirm their usefulness as indicators of pollution in marine ecosystems.

Parasites are a ubiquitous but often unseen, unacknowledged and understudied component of biological communities. However, their roles in structuring ecosystems, their influence on the evolutionary history of host species and their potential applications are slowly being uncovered. The snoek, Thyrsites atun, is a nomadic predator native to the cold coastal waters of the southern Hemisphere. Being a major target of the South African inshore line-fishery, the snoek is a socioeconomically important species whose ecological significance in the southern Benguela should not be underestimated. This study aimed to survey the metazoan parasite community of snoek off South Africa, assess host-parasite relationships and evaluate the potential of parasites as biological tags for stock structure studies. Examination of 210 snoek (FL 411 - 1040 mm) revealed them to be host to 16 parasite taxa. These included 9 new host records (Tentacularia coryphaenae, Caligus coryphaenae, Caligus dakari, Corynosoma australe, Nothobomolochus fradei, Hatschekia conifera, Bolbosoma vasculosum, Rhadinorhynchus cadenati, Digenea sp.) and 4 new locality records (Molicola uncinatus, Pseudoterranova sp., C. dakari, B. vasculosum). A further three cosmopolitan taxa (Anisakis sp., Kudoa thyrsites, Hepatoxylon trichiuri) as well as Caligus zei were also recorded.

Bibliography: pages 108-123. / This study was initiated in response to allegations by the conservation lobby that the commercial beach-seine fishery in False Bay was jeopardizing fish stocks and detrimentally affecting the ecology of the Bay. Its main aims were to quantify the current catch and place it in an historical perspective, to describe seasonal patterns in catches and effort and to assess the impacts of netting on juvenile fish and benthic organisms. The overall objective was to provide a "scientific" basis for the resolution of the controversy surrounding this fishery.

No description available.

Includes bibliographical references / An Ecosystem Approach to Fisheries management (EAF) offers a holistic approach for sustainable fisheries management by extending the traditional target resources-orientated management (TROM) to include wider social-ecological dimensions of fisheries. An EAF requires balancing of multiple, often conflicting objectives, effectively dealing with complexity and uncertainty, and engaging with diverse groups of stakeholders. Various tools within the field of Multi-criteria Decision Analysis provide a formal approach which takes explicit account of multiple criteria, while effectively dealing with risk and uncertainty. A knowledge-based tool was developed in this thesis to assess the efficacy of EAF implementation for the ecological well-being dimension in the South Africa sardine fishery. An iterative, participatory approach was adopted for its implementation. The modelling philosophy applied a rapid prototyping approach, and an applied research perspective was employed to direct the research. A broad group of stakeholders participated in indicator selection, tool design, and interpretation. The knowledge-based tool provided a hierarchical framework for seven specific management objectives to which eleven ecological indicators were linked. Time series (1987-2009) were collated for each indicator, and a utility approach was used to transform indicators to a common scale. Weights for indicators and objectives were agreed to by stakeholders and combined through the objectives' hierarchy using weighted means. The resulting outputs were discussed in detail during focus group meetings to ensure that the tool was clearly presented and as intended helped improve the stakeholder's understanding of the process. It was confirmed that the 20 knowledge-based tool presents a transparent, repeatable and scientifically defensible approach, suitable to meet management requirements. The tool development process was useful in bringing diverse stakeholder groups together, and through applying the tool as a boundary object, has helped to bridge the boundary between the TROM and EAF research communities. Encouraging stakeholder interaction offers opportunities for social learning, which if carefully facilitated through the tool development process is likely to enhance the outcomes of this process and support more generally in bridging boundaries to EAF implementation. The combined focus on tool development and social processes supports effective implementation of an EAF in the South African small pelagic fishery and provide a model for other fisheries.