Management of Atlantic cod (Gadus morhua L.) under future climate change

Abstract

Living marine resources are increasingly threatened by combined pressures of fishing and climate change, which put the sustainability of harvesting for yields that can support a fishery at risk. Ocean warming frequently reduces the productivity of commercial fish stocks, aggravating efforts of fisheries management to reverse effects of overfishing incurred through massive buildup of fishing capacity in the 20th century. A substantial challenge for management approaches that aim for adaptation to climate change is the lack of comprehensive knowledge on the mechanistic nature of climate effects on the dynamics of harvested populations, especially on their little-observed early life stages. Unexpected management response, e.g. delayed or missing rebuilding dynamics, indicate that stocks react in highly non-linear or discontinuous and long-lasting ways to fishing and climate change; ways that can hardly be anticipated or projected within models. Yet, accounting for such dynamics is important for guiding management decisions, as ignoring them can lead to poor stock condition and fixation thereof and economic hardship for fishers. The present dissertation investigates the impact of deep uncertainty in the relationship between temperature, stock size and recruitment (i.e., early-life-stage- / juvenile survival until entry to the fishery), which partly results from the non-linear stock response to stressors, on the potential for long-term future sustainable harvesting under intensifying future climate change. To this end, the dissertation adopts the perspective of “Decision-Making under Deep Uncertainty” (DMDU) by performing a large number of model projections under a very large range of uncertain scenarios, defined i.a. by possible stock-environment-recruitment (SR-) relationships and climate scenarios, in combination with a range of harvesting policies. The aim is not to improve model predictive skill or to infer the most “likely” future development but instead to unveil the potential consequences of harvesting policies and ocean warming for stock health. The various SR relationships here are intended to represent the general uncertainty about the SR relationship but especially also different (hypothetical) regimes of stock productivity that might arise from shift-like non-linear stock response to warming and harvesting. Atlantic cod (Gadus morhua L.), a fish of high commercial and cultural relevance that was frequently over-fished in the last half-century and displays climate vulnerability and unexpected management response, is chosen as model species. The dissertation first investigates the impact of deep uncertainty about the SR relationship on the response of a typical moderately over-fished cod stock, the North-Sea cod, to harvesting under future climate change. Procedures developed from the concept of “Robust Decision-Making” (RDM), a particular analytical approach used in DMDU, are used to explore the existence and properties of harvesting strategies that achieve universal sustainability by maintaining healthy stock size under all deeply uncertain scenarios. It is found that this uncertainty acts equally strong as harvesting on the degree of sustainability, and that no temporally fixed harvesting strategy can attain the objective of continuous future sustainability under climate change under all SR scenarios. A risk analysis is conducted to assess the relative effect of harvesting intensities on future sustainability and economic viability as a surrogate for identifying a fully “safe” policy, finds that these two objectives are related to each-other in a negative trade-off. This relationship shows that higher chances for future sustainable harvesting are likely associated with lower-than-recent profits, at least when aiming for temporally stable harvesting. The dissertation then adopts this risk-analysis framework for developing “Safe-Operating- Spaces”- (SOS-) concept for sustainable harvesting under climate change of 19 stocks of Atlantic cod inhabiting the eastern- and western-Atlantic shelves and the northerly waters (> 60 °N). Model simulations are conducted for combinations of various levels of catch, temperature increase and initial stock biomass, resulting in a multivariate map of sustainability risk calculated over SR scenarios. While a fully safe harvesting level is not detected for any stock, the existence and extent of the scenario space yielding < 50 % risk, i.e. the SOS in the present study, shows distinct patterns between major geographical areas and levels of historical overfishing: Historically long-term-collapsed western-Atlantic stocks display hardly any SOS, while relatively well-rebuilt northerly eastern-Atlantic stocks appear strongly responsive to harvesting and more southerly eastern-Atlantic stocks additionally display a clear negative climate effect. Initial stock biomass is found to have a profound impact on maximum warming tolerable (i.e. yielding ≤ 50 % risk) for the stock and for harvesting the stock, indicating that building stock reserves could help mitigate future climate effects on many stocks. Finally, the dissertation develops a heuristic optimization procedure for uncovering sustainable (long-term) harvesting pathways under dynamic ocean warming for the currently depleted Western Baltic cod. The optimization algorithm is designed to respect both conservation- and, with secondary priority, yield objectives. The resulting harvesting trajectories force a quick rebuilding of stock size to sustainable levels and then, by gradually reduce fishing pressure as warming intensifies, maintain sustainable stock size while continuously decreasing catch. Introducing deep SR uncertainty leads to harvesting of similar dynamics but on a much reduced level. Risk of stock decrease to unsustainable levels is kept on a very low level for several decades but eventually increases under continuous warming, indicating a cessation of manageability under unfavorable SR scenarios. Risk of missing even a historically modest target yield is, in contrast, continuously comparatively high. Embedding harvesting optimization along with stock monitoring as a recurring procedure within a simulated management pathway, under a true but unpredictable trajectory of changing SR relationships, leads to high stock size and yields. The clear fulfillment of both objectives indicates the combination of uncertainty-conscious long-term harvesting planning, stock monitoring and short-term policy flexibility to be a valuable management approach under future climate change. The dissertation concludes that Atlantic cod will likely remain manageable under future climate change to a certain degree. Achieving future sustainability will depend strongly on the realized SR relationship(s), the degree of future warming, the degree of stock (re-) building achieved and likely on the severity of historical overfishing. Management strategies most likely to be successful will aim for quickly (re-) building stock biomass and adapt harvesting intensity to realized stock size and with the aim of minimizing long-term conservation risk under a wide-range of future SR scenarios.