Using Management Strategy Evaluation to address problems arising as a result of competing users of the South African horse mackerel resource

Abstract The Cape horse mackerel (Trachurus trachurus capensis ) has traditionally made an important contribution to the South African fishing industry and is a key component of the Benguela ecosystem. This thesis concerns the assessment and management of the South African horse mackerel resource. It starts with a brief review of the biology of the Cape horse mackerel and the history of the fishery, as well as of the Management Strategy Evaluation approach, which was applied in this work. Assessments of the horse mackerel resource are currently undertaken through the combined efforts of the Demersal and Pelagic Scientific Working Groups (SWGs) of the South African Department of Agriculture, Forestry and Fisheries (DAFF). A joint effort is required because the resource is available to multiple fisheries: as directed catch to the midwater trawl fishery and as bycatch to the demersal trawl and pelagic purse-seine fisheries. Management of the resource is complicated by differences in the age-structures of the horse mackerel caught in each of these three fisheries. The data available for the assessments are described, including the details of their collection and processing. Four age-structured production models (each reflecting different assumptions about the horse mackerel resource) are fitted to those data using the maximum-likelihood estimation method, and are used to provide assessments. Estimates of the current status of the stock indicate that it is healthy, putting it well above its Maximum Sustainable Yield (MSY) level. For the directed midwater fishery, MSY is estimated to be in the region of 50 000–100 000 tonnes per annum. However, the results of constant catch projections suggest that there is a pronounced yield-per-recruit effect, with even small bycatches of juvenile horse mackerel in the pelagic fishery having a pronounced negative effect on the level of a catches in the midwater fishery that can be sustained. Since 2000, the annual bycatch of horse mackerel in the pelagic fishery had been regulated by a iiiPrecautionary Upper Catch Limit (PUCL) of 5 000 tonnes. However, it became apparent that this fixed allocation was insufficient during years with high juvenile horse mackerel availability when, in 2011, the fishery was threatened with closure long before the quotas for its target species could be filled. A Management Procedure (MP) therefore needed to be developed that would provide flexibility in the annual PUCLs and thereby minimise operational constaints on the pelagic fishery. The performances of a number of Candidate Management Procedures (CMPs) are evaluated in this thesis through simulation testing and expressed in terms of expected catch and risk of resource depletion. Following discussions in the DAFF SWGs, a MP that limits the total bycatch over a three year period was selected for implementation and has been used to provide PUCL recommendations for the 2013 and following fishing seasons. There are indications that estimates of horse mackerel abundance based on demersal swept-area surveys may be negatively biased, and hence that the resource is underutilised. Therefore, a MP is developed that experimentally increases the Total Allowable Catch (TAC) for the directed midwater fishery without increasing the risk of resource depletion. The CMPs considered use trends in abundance indices to monitor the response of the stock to management actions, and then adjust the annual allocation accordingly. Again, their performances are evaluated through simulation testing. The results show that the CMPs offer various trade-offs between improved catches and increased interannual TAC variability, and also in better short-term catch performance at the expense of lower long-term TACs. A MP that restricts the TAC to a 10% increase and 15% decrease per annum was chosen from the candidates considered and has subsequently been used by DAFF to recommend the midwater TACs for the 2013 and following fishing seasons. Future suggestions for improving the assessment and management of the South African horse mackerel resource are discussed. These include adopting a Bayesian approach to refine the estimates of assessment precision and allow the incorporation of prior knowledge, investigating alternate approaches to modelling pelagic bycatches during projections and improving the model fits to catch-at-length data. Future research needs to provide estimates of the bias in the horse mackerel abundance estimates that are currently based on demersal swept-area surveys, possibly by performing concurrent trawl and hydro-acoustic surveys. Additionally, a wider range of CMPs than those considered in this thesis should be investigated for both the pelagic and midwater fisheries—the recommended CMP for the former did not seem to appreciably decrease the probability of disruptions to that fishery, while that for the latter did not consistently award larger increases in TAC when the resource was assumed to be more productive. Furthermore, robustness trials need to be conducted for all CMPs to ensure that they exhibit reasonable ivperformance given a wider variety of plausible scenarios for the resource’s dynamics.