The South African hake resource faces divergent fishing pressures and management issues. Although the resource consists of two species, management was simplified because the resource was only subject to trawl effort and because of the similarity in population parameters, single species assessment models could be adopted. The impact of trawling on the stock is considered to be well understood and the resource has shown recovery since exclusion of foreign vessels. The 1990s were punctuated by major political change and the need for transformation has resulted in an expansion of the number of fishing rights holders. Longlining has been reintroduced and there is no clear understanding of how the combined fishing pressures will affect the population structure of either Merluccius capensis or M. paradoxus. Information is vital to successful management. Fisheries are complex and intricate, and at times appear impossible to control or monitor. An integrated information system provides easily understood graphical explanations of complex issues. This thesis assessed the dynamics of the trawl and longline fisheries between 1994 and 1999 using a geographical information system (GIS). The accessability of a GIS incorporates the needs of scientists, managers and fishing communities. The simple GIS developed in this study revealed shifts in effort, facilitated the calculation of spatially precise catches and biomasses and highlighted the inadequacy of current sampling coverage. Trawlers were shown to fish the same areas consistently during the years investigated, with highest fishing intensity and average CPUE achieved at depths between 300 and 500m. Analysis of the longline sector revealed several similarities to the trawl sector, fishing intensity was highest between 301 and 500m, suggesting that both sectors face a "friction of distance" dilemma. The distribution and abundance of hake, in particular the exploitable proportion of the population, was determined by a combination of depth and substrate type. The selectivity patterns of trawlers and longliners were briefly investigated with the results illustrating that gear selectivity of the M. capensis stock was depth dependent. The deeper fishing occurred, the larger the length-at-selection. As a result, the M. capensis parental stock faces unprecedented fishing pressure. In the absence of reliable species-specific catch data, logistic and linear regression models were developed to split the hake catch into its respective species components. Large discrepancies between the predicted M. capensis catch for the two models were noted. Regression estimates constructed at a finer spatial scale may provide a better fit than the current depth logistic employed by Marine and Coastal Management. A first attempt at assessing the M. capensis resource on the West Coast using an age-structured production model was presented. It was found that a lengthy and accurate M. capensis catch series is required before it is possible to successfully model the dynamics of the stock. It is necessary to incorporate finer spatial detail in the collation of catch data and collection of sampling data. It would be unadvisable to assume that the stock is stable or recovering. The implications of a size/sex relationship must be investigated and properly appraised.

The motivation for this study was to examine the exploitation patterns of the M. capensis and M. paradoxus hake fisheries on the Southeast Coast, and determine the size and species of hake caught in each of the hand-line, long-line, and trawl methods. The handline hake fishery has increased substantially over the last several years and concerns are beginning to emerge about the impact this will have on the inshore resource collected on the South Coast between August 1998 and July 1999 was used to describe the hand-line method and estimate annual landings. Data on the size and species in longline catches of hake caught during 1997 were already available for this study. Size distributions in trawl catches were determined from commercial category landing data reported by catch weight and depth. The species composition in these catches determined by comparison using RV Afrikana survey data collected in the same depth regions. Location plays a significant role in determining the sizes and species of hake caught by each gear. Hand-lines catch smaller sizes on average than do long-lines, inshore trawls target mainly M. capensis while offshore trawls catch both hake species. A substantial amount of the hand-line hake caught on the South Coast is not reported. Examination of the exploitation patterns reveal that intense trawling pressure is directed at the smaller sized M. paradoxus inhabiting the depth region between 160-400-meters. Inshore trawls discard a large amount of small sized M. capensis within the 100-meter isobath. A preliminary stock assessment on the status of each hake species found that M. paradoxus appears to be over-exploited while M. capensis was in better shape. However, length-based pseudo-cohort analysis, used in this assessment, is critically reliant on having length frequency data from a steady state population in equilibrium. This limits the application of this model for management purposes and this finding is purely theoretical at this stage. Results suggest that each hake species is under a different pattern and level of exploitation and the multi-species nature of hake stocks on the South Coast should be considered in developing optimum management policies. Future work should focus on developing appropriate age/length keys so that an age-based VPA, which is more powerful than the length-based approach, can be applied towards stock assessments on the South Coast. Alternatively, length-data covering a longer period should be compiled and the equilibrium assumption further investigated so that the results from length-based models can be used with more confidence.