Towards sustainable fisheries: Improving the robustness and effectiveness of management procedures for data-limited fish stocks

Abstract

Marine fisheries are very important from a socio­economic point of view,providing food and employment for hundreds of millions of people around the world. However, evidence suggests that roughly half of all fish landed come from unsustainable fisheries. To ensure sustainability of fisheries resources, management procedures (MPs) capable of adequately assessing the current exploitation status of fish stocks and predicting future stock development under various exploitation scenarios are required. The reliable assessment of stock status and the resulting management advice, however, is complicated by a lack of available and informative data, as well as a high degree of scientific uncertainty. Scientific uncertainty typically arises from natural variability, sparse and inaccurate data collection, incorrect model structure, and the parameter estimation process. The aim of this thesis is to improve the robustness and effectiveness of data­limited MPs by (i) evaluating the performance of state­of­the­art length­based assessment methods, (ii) advancing a stochastic surplus production model (SPM) with time­variant productivity parameters, and (iii) accounting for the uncertainty quantified by SPMs in harvest control rules (HCRs) that are based on biological target and threshold reference points, and (iv) developing effective HCRs that are based on the predicted biomass trend. This is done by (closed­loop) simulation frameworks that allow for the validation and evaluation of existing and novel models and MPs by analysing their impact on simulated stock and fisheries dynamics. The stock dynamics are simulated by fully stochastic individual­based and age­structured population models with sub­annual time steps. The results of the thesis show that scientific uncertainty can lead to inaccurate and imprecise estimates of stock status by length­based assessment methods, and that the magnitude of bias and variance does not only depend on available data, but also on the life history parameters of the species/stock as well as the exploitation history of the stock. In particular, fish stocks associated with large population fluctuations (such as fast­growing, short­lived species) can complicate length­based assessments. In contrast, the stock status estimated by SPMs is relatively robust to scientific uncertainty. Absolute reference points and absolute biomass and fishing mortality, on the other hand, are biased when productivity varies over short seasonal and/or long­term scales and is not accounted for. Furthermore, the findings of the thesis demonstrate that in the face of high scientific uncertainty, HCRs based on target reference points alone can lead to a high risk of overfishing. However, by including biomass thresholds or accounting for scientific uncertainty quantitatively, the risk associated with the target reference points can be reduced, though at the cost of a short­term loss in yield. The amount of yield lost, depends on the definition of the HCR, the species­specific life history parameters, the degree of overfishing, and the time elapsed since the first application of the rules. Finally, the results of this work highlight a promising suit of HCRs that are based on the short­term forecast of the biomass trend and can outcompete current trend­based MPs in terms of their effectiveness and applicability. Overall, it can be concluded that biased and imprecise stock assessment methods, in combination with HCRs that do not account for uncertainty, compromise the robustness and effectiveness of MPs. The sustainable management of data­limited fisheries will not only rely on comprehensive guidelines on the application and implementation of assessment methods, but also on the further development of stochastic models, and effective HCRs that propagate the estimated uncertainty into management advice. This thesis offers a promising step in that direction and identifies avenues for further research within the field of fisheries science.