Fish harvesting advice under climate change: A risk-equivalent empirical approach

and non-random process that potentially alters both resource state and productivity and could lead to unreliable advice for resource management that does not take it into account [1]. CC could affect the likelihood of achieving most objectives (including making some objectives unachievable) through inter-alia changing how a resource like a fish stock responds to pres- sures like fishing. During the current period of rapid CC [2,3], the consequences of manage- ment actions that occur on various temporal scales could be impacted by changes in climate- driven environmental variables that affect fish stock production and the ability to meet stock objectives. In such cases, taking CC impacts into account would change advice and decisions for stock management. It is therefore wise to consider approaches to operationally integrate climate considerations in the formulation of science advice, even in the absence of detailed process-based understanding of climate-fish production dynamics. Stock by stock, science advice that incorporates CC may or may not differ substantially from advice that does not incorporate it. However, as evidence of CC impacts on marine ecosystems continues to accu- mulate, science advice consistent with the precautionary approach (PA) should consider envi- ronmental variation and CC as factors that can change the risks associated with alternative fishing strategies and/or management options. CC-informed fisheries advice can be provided by adapting mechanistic population models

for stock dynamics to include a key climate forcing environmental variable on one or more specific productivity processes in the model. A simulated CC scenario can then be used to drive future stock states with this modelled mechanistic dependence. There are limitations to this approach however, as it is well-established that statistical relationships between individual environmental variables and specific biological processes often break down as new data becomes available [4]. Stock production processes are likely affected by a suite of environmen- tal factors, the effects of which may be non-linear, synergistic or antagonistic, and change quickly with the influence of different environmental factors on the production processes changing as different factors take on values that are typical (modal or median) or atypical (infrequent or extreme) [3]. Interest in these varying but important non-linear and non-stationary relationships is growing, and fostering research in areas like managing for resilience of systems and processes in the face of CC. CC-informed fisheries advice is already being proposed in efforts to improve setting objectives and developing higher level multi-year policy guidance. Nevertheless, how- ever robust the objectives and strategies may be to a non-stationary environment, the expert community must face the task of provision of reliable short-term operational advice relative to those objectives and strategies. If the computations underlying the operational advice assume

stationarity of the model parameterisation in relation to stock productivity condition (and they usually do), they are also by default saying that the environment is stable or it is unlikely to impact model parameterisation and associated reference points. Alternatively, they may consider a much larger uncertainty that can encompass non-stationarity within their predic- tion time frame but this has the impact of making all advice highly uncertain and inevitably leads to foregone yield that may be socio-economically difficult to implement and therefore ignored by decision makers. By assuming non-stationarity or by casting a very wide uncertainty window to derive sci- ence advice are both more vulnerable to error if some of the environmental variation is direc- tional and systematic–as is happening with climate change. If researchers can regularly update their knowledge of these environment-productivity relationships, and the ways that the envi- ronment is changing, there may be opportunities to update which environmental factors may be approaching or have passed key thresholds, and are affecting stock productivity. This knowledge can be used to condition risk-based advice based on empirical models parameter- ized to conditions as local in space and time as information allows. In this sense, we seek an PLOS ONE Climate-change conditioned risk-equivalent fisheries advice PLOS ONE | https://doi.org/10.1371/journal.pone.0239503 February 19, 2021 2 / 22 (ACCASP) know as ‘B-base’ (Sept 2019), was used to support travel after to the main work was

conducted to facilitate writing of this MS. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. No portion of any of the authors’ salaries were supported through application to specific programs from funders and salaries were paid fully through A-base funds of the DFO Funding (~5 k CAD) from a specific departmental program (ACCASP) know as ‘B-base’ (Sept 2019), was used to support travel of DD and M-JR after to the main work was conducted to facilitate writing of this MS in Nanaimo, BC Sept 2019 with KH. Competing interests: The authors have declared that no competing interests exist.