The development and implementation of a joint management procedure for the South African pilchard and anchovy resources

Bibliography: p. 309-319.Pilchard and anchovy are the main targets of the pelagic fishery, South Africa's second most valuable fishery in monetary terms, but producing the highest annual yield in terms of landed mass. It is the most dynamic of South Africa's major commercial fisheries, because the species it targets are relatively short-lived, often occur in mixed shoals and experience large fluctuations in abundance. Mixed shoaling causes operational problems for the pelagic fishery because of the inevitability of juvenile pilchard by catch (more valued as adults for canning) in the anchovy directed fishery. The operational interactions imply a trade-off between these species, and necessitate that they are managed together. The thesis describes the development of a joint management procedure, which provides a framework for quantifying this trade-off such that acceptable levels of risk of resource ""collapse"" are not exceeded for either resource. A management procedure in this context is essentially a computer-simulation tested set of formulae used to recommend catch levels based upon the results of two research surveys of resource abundance, one based on adult fish, and the other on recruitment (juveniles). Parameter estimates for the underlying models used in these simulation tests are derived from assessments of each resource, which fit models to the observations from research surveys and commercial catches using maximum likelihood estimation. Assessment results indicate adult and juvenile natural mortality values of 0.4 and 1.0.year-1 for pilchard, and 0.9.year-1 each for anchovy. They also suggest that recruitment surveys for pilchard are much less precise than sampling CVs alone indicate, but this may be an artefact of problems in the age data, which require further investigation. A range of alternative assessments for pilchard and anchovy is selected for the purpose of performing management procedure robustness trials as part of the simulation tests. The thesis presents two methods for incorporating a trade-off decision between pilchard and anchovy into management procedures. The first develops the trade-off decision as ""external"" (the external trade-off decision, or ETD method), where the responsible Minister makes a trade-off decision for the industry as a whole. This method is problematic because it requires a compromise between those rights holders that prefer pilchard (for canning), and those that prefer anchovy (for reduction). The trade-off decision is then developed as ""internal"" (the internal trade-off decision or lTD method), where rights holders make their own trade-off decision based on their particular preferences. The ETD method specifies a pilchard anchovy ratio for the whole industry, calculates TACs (total allowable catches) on the basis of this selection, and then allocates quotas to each rights holder on the basis of a percentage right for each fishery (pilchard and anchovy). In contrast, the lTD method collapses separate rights for pilchard and anchovy into a single right for the pelagic fishery, and allocates quotas on the basis of trade-off decisions made by individual rights holders, thus avoiding an overall compromise choice. The introduction of non-linear adjustment factors ensures that (over time) allocations to individual rights holders would be near identical, whichever of the two methods is used. The lTD method is currently (2002) being used by Marine and Coastal Management as the basis to set pilchard and anchovy TACs and to allocate individual quotas. The thesis analyses two further aspects of management procedures. The first looks at whether the performance of management procedures can be improved if they incorporate information about regime shifts, modified as sinusoidal cycles. Results indicate that improvement is possible, but that, surprisingly, one would do better using a running mean of research survey estimates of abundance to track a regime cycle, than direct measurements of the actual position in the cycle. The second investigates the benefits of using environmental indices in management procedures, where these indices are used as short-term predictors of recruitment and replace the assumption (made because of lack of information) that forthcoming recruitment will be average. Results indicate that, given the uncertainties related to using such indices, they would need to explain at least 50% of the variation in recruitment before management procedures start showing appreciably improved performance. Priorities for future work include: investigating the implications of a possible changed perception of the bias associated with pilchard adult surveys than that adopted to test the present management procedure; developing the daily egg production method to provide unbiased estimates of pilchard abundance; and performing a thorough investigation of pilchard age data to isolate any problems