Accounting for shifting distributions and changing productivity in the development of scientific advice for fishery management

Abstract

Introduction

Changing climate and ocean conditions affect fish stocks, fishermen, resource managers, and coastal communities that rely on marine resources for food and economic security. Changing conditions, such as increasing water temperature, ocean acidification, changes in atmospheric and ocean circulation, and more frequent and stronger weather events, have had a discernible impact on some marine ecosystems and fisheries (The Royal Society, 2005; Diaz and Rosenberg, 2008; Hoegh-Guldberg et al., 2014; IPCC, 2014; FAO, 2016; Barange et al., 2018). With ocean temperatures expected to increase an additional 2°C–4°C by the end of the century, increasing effects to fish stocks and fisheries are of great concern (Hoegh-Guldberg et al., 2014; IPCC, 2014; Poloczanska et al., 2016).

For the United States, strong legislative mandates and investments in scientific programmes have helped to establish the United States as a global leader in sustainable fisheries management. However, changing climate and ocean conditions present growing challenges for fisheries science and management. To address these challenges, the United States National Oceanic and Atmospheric Administration National Marine Fisheries Service (NOAA Fisheries) developed a series of planning documents to better coordinate and communicate preparations for and responses to climate-related impacts and other stressors on marine ecosystems, fisheries, and coastal communities [e.g., NOAA Fisheries Climate Science Strategy (Link et al., 2015; Busch et al., 2016); Ecosystem-Based Fisheries Management (EBFM) Road Map (NMFS, 2016); and Implementing a Next Generation Stock Assessment Enterprise: An Update to the NOAA Fisheries Stock Assessment Improvement Plan (SAIP; Lynch et al., 2018)]. Building from those efforts, this article provides specific recommendations for addressing two effects of changing ocean conditions: shifting species distributions, and changing stock and ecosystem productivity (Nye et al., 2009; Lenoir et al., 2011; Pinsky et al., 2013; Lynch et al., 2015).

Shifting distributions

A distribution shift refers to a permanent, multi-decadal, or centennial, shift in the spatial distribution of a species or stock from its historical region or habitat to a new region or habitat. These shifts are often considered a result of changing climate or ocean conditions; although other factors (e.g. fishing, habitat degradation, trophic dynamics, etc.) may also cause shifts (Bell et al., 2014). Stocks are already on the move in response to changing oceanographic (e.g. temperature) and ecosystem (e.g. habitat loss, trophic dynamics) conditions in many regions (Nye et al., 2009; Lenoir et al., 2011; Pinsky et al., 2013; Bell et al., 2014; Lynch et al., 2015; Poloczanska et al., 2016).

Distribution shifts violate traditional assumptions of “stationarity” (variation around a constant mean value) common to fish stock assessments, and have implications for estimates of stock size, biological reference points, stock status, and ultimately harvest recommendations and allocation (Szuwalski and Hollowed, 2016). For example, when a stock shifts across a defined stock boundary, questions regarding stock status may arise as there would be an apparent decline in abundance in the area that it is leaving, and an apparent increase in abundance in the area to which it is moving, while the true stock abundance may actually remain constant (Link et al., 2011). Shifting species distributions also have consequences for access to the fish, particularly when stocks shift across management jurisdictions (Spijkers and Boonstra, 2017; Pinsky et al., 2018), which can have a direct economic impact on fishing communities. Additionally, management tools, such as catch limits, bycatch reduction measures, or time and spatial closures, may lose their effectiveness as species shift into or out of regions and interact with species and gear differently.

Changing productivity

Productivity in a fisheries context refers to the total biomass or number of fish that a stock can produce and relates to how much it can theoretically support for removal (i.e. by fishing). Climate, ocean, and ecosystem conditions can greatly affect fish stock productivity by altering habitat suitability, ecosystem-level productivity and dynamics (e.g. predator-prey interactions), and a stock’s life history parameters such as growth, maturation rate, natural mortality, and stock–recruitment relationships (Hare et al., 2010; Farley et al., 2016). For example, the recruitment of many cold water and temperate species, such as Atlantic cod (Hare et al., 2010), yellowtail flounder (Fogarty et al., 2008; Miller et al., 2016), and walleye pollock (Mueter et al., 2011; Stabeno et al., 2012; Duffy-Anderson et al., 2017) has been linked to water temperature and/or sea ice extent. These changes may occur concurrently with a stock’s distribution shift, but also may affect stocks that remain within their historical ranges.

When stock productivity changes, it directly affects the stock assessment and management process. Errors in estimated life history parameters that stem from productivity changes can lead to significant differences in inferred stock productivity and biological reference points, including estimation of the maximum sustainable yield (Whitten et al., 2013; Audzijonyte et al., 2016). The achievability of rebuilding plans may also be affected if biological targets are no longer feasible under current or future conditions (Bell et al., 2018). Changes in productivity may be unidirectional (as is expected with climate change), where the affected stock is not expected to return to its original state in the foreseeable future, or cyclical where productivity oscillates in cycles that last for years to decades before shifting back to a previous productivity regime (as has been the case with decadal-scale climate and oceanographic oscillations).

There is growing recognition among scientists and managers of the importance of accounting for the effects of changing climate conditions on species’ distributions and productivity; however, it remains challenging to effectively incorporate this information into the science-to-management process.

Six-step process: challenges and recommendations

We evaluated the U.S. fisheries science-to-management process, and developed recommendations for how the effects of climate change may be addressed at each step in the process, from the capacity to detect and understand changes, to the communication of results to managers and subsequent management actions (Figure 1). Recommendations (summarized in Table 1) were developed in the context of U.S. fisheries management, but many are generally applicable to fisheries managed by other jurisdictions.

Figure 1.

Climate-ready fisheries management process. Changing climate conditions are represented at the centre of the diagram as ocean acidification, temperature change, sea level rise, and extreme events. These cause changes in the biotic community, such as shifting distributions and changing productivity, as indicated in the next ring out from the centre. To enable managers to account for these changes and move toward climate-ready fisheries management (outermost ring), scientists and managers need to be able to detect changes, understand mechanisms of those changes, evaluate risks and priorities, conduct assessments and develop forecasts, and communicate results and advice to managers and stakeholders.

Open in new tabDownload slide

We discuss six steps of a climate-informed science-to-management system: (1) detect and anticipate changes, (2) understand mechanisms of change, (3) evaluate risks and priorities, (4) conduct assessments and develop forecasts, (5) communicate advice to managers and stakeholders, and (6) manage fisheries under changing conditions. Although we present these steps in a linear fashion, it is more often an iterative process where information gained during one step may be used to inform the previous or other steps. In the following sections, the major challenges to accounting for shifting distributions and changing productivity are identified, followed by a description of recommended actions for addressing those challenges.

Detect and anticipate changes

Addressing species distribution shifts and changing productivity may require scientists and managers to: determine that a change has occurred in the past; detect that a change is currently occurring; and predict whether a change is likely to occur in the near future. A crucial component of detecting and predicting changes is the ability to track and monitor shifts in oceanographic conditions and stock characteristics in near real-time. However, tracking changes and collecting data at the temporal and spatial scales necessary to provide near real-time data and early warnings that a change may occur is challenging under current monitoring capacity (Holsman et al., in press).

NOAA Fisheries relies on fishery-dependent and independent sources to collect data and monitor changes in oceanographic and biological variables. For stock abundance monitoring, survey and catch rate data are often standardized over space and time to the historical range of the species or to a jurisdictional domain. Historically, standardized sampling and analysis results in reduced sampling bias and facilitates spatial and temporal comparisons, yet standardization may not overcome error and/or bias if the approach does not account for major changes in oceanographic features that affect stock distributions or life history parameters. Additionally, many of the geostatistical techniques underlying spatial models rely on extensive and accurate spatial data (Berger et al., 2017), increasing the need for data collection programmes that accurately capture the spatial structure and movement of stocks and coinciding ecosystem information.

Facilitating survey flexibility and adaptability could greatly improve the detection and tracking of distribution shifts and changes in productivity, and ensure that the complete range of a stock is adequately sampled. Adapting surveys could involve both adjusting surveys (e.g. location or start/end date) in response to known changes in species behaviour (e.g. change in timing or location of migration or spawning aggregations) Pacific Sardine Surveys (Zwolinski et al., 2011) and expanding the spatial and temporal scope of surveys. It is imperative, however, to evaluate proposed changes in advance to understand their potential impact on the integrity of survey design and determine whether additional calibration studies are needed to maintain survey consistency. Additionally, we recommend more research regarding adaptive survey designs that maximize precision and minimize biases given shifting distributions. In addition to potential issues of survey bias when adapting survey design, adaptive surveys can be cost and time prohibitive. Therefore, the use of fishery catch data, citizen science, local ecological knowledge, and partnerships with fishermen should be encouraged and evaluated to see if their observations could be used to supplement monitoring capacity and inform survey adjustment decisions. An example of such a partnership is the “Environmental Monitors on Lobster Traps” or eMOLT project (https://www.nefsc.noaa.gov/epd/ocean/MainPage/lob/lob.html), a non-profit collaboration between NOAA scientists, industry, and academics established to expand the capacity to monitor the physical environment of the Gulf of Maine and Southern New England shelf. Participating fishermen collect data on bottom temperature, salinity, and current velocity using instruments attached to their lobster traps. The main motivation behind the partnership is to improve lobster science, but the data are made publically available and are also used by ocean circulation modellers. In the southeast region of the United States, fishermen regularly provide insights on relevant environmental and climate drivers via the Southeast Data, Assessment, and Review (SEDAR) process under which all stock assessments are conducted. Input from fishermen is routinely considered and incorporated into stock assessments; for example, stakeholder insights eventually led to the incorporation of red tide events as a major source of mortality on grouper stocks of the West Florida Shelf (SEDAR, 2006, 2015).

Movement of stocks across survey or jurisdictional borders poses an additional challenge for detecting changes in a stock’s distribution or productivity. Surveys are not always consistent in design, timing, or gear type across jurisdictions, which can complicate studies evaluating whether observed changes are due to distribution shifts, changing productivity, or other causes (Brown et al., 2016). Improving the coordination and communication of research, survey efforts, and results across regions and jurisdictions will allow better tracking of change.

Monitoring could also be improved through more effective use of ongoing data collection efforts. Early warnings and indicators of change can be developed using ecosystem or stock attributes that are relatively easy to track through current oceanographic and fishery surveys. For example, oceanographic cruises and fishery independent surveys routinely collect information on temperature, dissolved oxygen, pH, and salinity, which are likely to shift with changing climate and are potential drivers of fish stock dynamics (Hollowed et al., 2013). Additionally, catch data from fishery-dependent and independent surveys can be analysed to track biological indicators of change such as, centres of biomass, trophic and stock structure, connectivity, growth and size at age, recruitment patterns, and changes in habitat suitability. Monitoring these metrics consistently across relevant spatial and temporal scales and making the data available at near real-time is key to having useful physical and biological indicators of changing conditions.

Ocean observing systems can provide information needed to track large-scale oceanographic conditions and indicators of change at temporal and spatial scales not adequately captured by traditional fishery independent surveys (see Manderson et al., 2011; Kohut et al., 2012). Further, increased use of advanced sampling technologies, such as sailing drones and other autonomous vehicles, should be considered to help expand the spatial and temporal scope of fish stock and oceanographic data collection.

Anticipating changes in both the oceanographic and ecological conditions can be aided by the use of ocean and ecological forecasting (Hobday et al., 2016,, 2018). Skilful forecasts of ocean and climate conditions (e.g. sea surface temperature, air temperature, rainfall) or habitat suitability on time scales relevant to fisheries managers [seasonal (3 months) to decadal (1–10 years)] are now possible (Hobday et al., 2016; Kaplan et al., 2016; Payne et al., 2017). Scientists and managers can use such forecasts to anticipate changing ocean conditions and biological responses, and to optimize the allocation of resources for monitoring activities (Hobday et al., 2016; Payne et al., 2017). However, the development of such forecasts is dependent on data availability and having a sound understanding of the underlying mechanisms. See “Conduct assessments and develop forecasts” section for further discussion of important considerations when developing forecast models.

Understand mechanisms of change

Understanding the underlying mechanisms of distributional and productivity changes is key to improving scientific advice and developing forecasts to inform fishery management. The way in which managers respond to observed changes may vary depending on mechanistic understanding and system interactions.

When stocks shift their distributions as a result of changing growth rates and/or oceanographic conditions, there can be resultant effects on the “availability” (catchability, selectivity, sampling efficiency, etc.) of fish stocks to surveys or fisheries. This availability is important to quantify in stock assessments because it affects estimates of total abundance, productivity, and resultant stock status and sustainable catch levels, yet remains difficult to quantify due to limited mechanistic understanding and measurement abilities. Data collection and experimental approaches should be designed to estimate catchability, availability, and selectivity in relation to ambient environmental and habitat conditions. For example, the use of paired acoustic-trawl surveys can provide information on changes in vertical distribution of fish, and video surveys of fish behaviour around survey gear can improve understanding of changing interactions of fish with survey gear in response to environmental conditions (e.g. Kotwicki et al., 2018 as used in the 2017 Eastern Bering Sea Pollock assessment).

Collecting oceanographic, ecosystem, and habitat data at the same time as a stock’s biological information improves the ability to determine potential drivers of change by matching up climate, ocean, and habitat conditions with resultant changes in stock dynamics. Additionally, studies looking at the effects of changing environmental conditions on a species’ vital rates (e.g. growth, mortality, larval survival, maturity, fecundity, and recruitment) throughout its life cycle are needed (see Hare, 2014; Boldt et al., 2015; Stawitz et al., 2015; Thorson et al., 2015b; Barbeaux and Hollowed, 2017).

Even when data are available to provide information on key drivers of change, the time and personnel required to process and analyse these data can be a limiting factor that results in a mismatch or lag between the data collection and processing, and the subsequent use of the information in assessments to inform management decisions. Identifying and addressing personnel and training needs are essential to increasing staff capacity to confront the growing challenges facing fisheries management in a changing environment.

Evaluate priorities and risks

Not all stocks will respond the same to changing conditions, and not all observed changes may warrant response in the science or management processes. Ideally, the decision to expand a stock assessment to include climate and ecosystem information, and/or to take management action, will consider the degree of change observed, the stock’s vulnerability to changing conditions, and the potential benefits to conservation and management. There are a number of methods and tools available to help evaluate changes, identify species and regions at greatest risk, and inform prioritization.

Spatial-temporal models and geostatistical techniques can facilitate evaluation of changes in a stock’s spatial distribution and incorporation of spatial processes in stock assessments (e.g. Thorson et al., 2015a; Ianelli et al., 2016; Thorson, 2019). Advancements in geospatial statistics can help identify and evaluate the significance of distribution shifts and spatial changes in catch rates to support including these factors in further analyses.

Sensitivity analyses, which are a routine part of stock assessments, determine how perturbations affect the assumptions and uncertainties in the model. These analyses can help identify: factors that have an effect on stock productivity and distribution; when observed changes in productivity or distribution have a meaningful effect on the scientific advice; and when explicitly accounting for the change may be warranted (e.g. U.S. West Coast Sardine, see PFMC, 2014; Kvamsdal et al., 2016; Spencer et al., 2016). Results of these analyses can also inform the development of thresholds and triggers that indicate when changes are large enough to warrant a management response.

Ecosystems, species, and even different populations of the same species, can vary in their vulnerability and responses to changing environmental conditions. A population’s vulnerability to change depends on several factors such as, life history and location within a species’ range (Hare et al., 2016). For example, within species, populations at the range edges generally react more strongly to environmental change compared to populations at the centre of the species range (Rijnsdorp et al., 2009; Rehm et al., 2015; Robinson et al., 2015). These differences in population dynamics within species emphasize the need for stock assessment and fisheries management approaches to account for differences in vulnerabilities and the range of responses to climate-induced pressures that may be occurring within one species.

Ecological risk assessments can be used to identify the major threats facing groups of species and their relative vulnerabilities to those threats, and help prioritize species for expanded assessments and/or management action (Hobday et al., 2011; Gaichas et al., 2014; Holsman et al., 2017). Several ecological risk assessment methods are currently available. Hobday et al. (2011) proposed a hierarchical (tiered) ecological risk assessment approach for looking at the effects of fishing on species and ecosystems, a method later expanded by Holsman et al. (2017) to include assessment of risk due to any natural or anthropogenic pressure. In a tiered approach, analysts evaluate the data available for a species to decide whether to move to the next higher tier of assessment for a more rigorous analysis. The approach allows analysts to prioritize time and resources on developing quantitative assessments for only those species identified to be most at risk. Additionally, the lower tier qualitative assessment can provide a rapid and computationally inexpensive screening tool to identify key pressures that may affect a wide range of species, habitats, activities, or social components. These pressures can then be prioritized for more in-depth monitoring and analysis (Holsman et al., 2017).

Climate vulnerability analysis (CVA) is another form of ecological risk assessment and is specifically designed to identify and prioritize fish stocks of greatest risk from climate change (Link et al., 2015; Hare et al., 2016). Results from these various risk assessment approaches can help determine where to focus efforts for inclusion of climate and ecological information into the fishery science-to-management process.

Conduct assessments and develop forecasts

Accounting for changing conditions within a stock assessment model can be quite challenging. Presently, most stock assessment models assume that model parameters either remain constant through time or space, are measured empirically (typical for growth), or vary according to random processes (typical for recruitment). However, changing environmental conditions may alter life-history processes (e.g. growth, maturity, and fecundity), or catchability, thereby affecting estimates of productivity and abundance (Mackenzie et al., 2007; Pankhurst and Munday, 2011; Hollowed et al., 2013). When there is a clear or anticipated effect of changing climate, ocean, or ecosystem conditions on fish or fishery dynamics, explicit consideration of these factors in assessments may provide more accurate estimates of current and projected stock status advice (Keyl and Wolff, 2008).

To formalize the consideration of ecosystem drivers in a stock assessment, both the EBFM Road Map and the SAIP recommend that ecosystem considerations be included in stock assessment and assessment reviewer Terms of Reference (NMFS, 2016; Lynch et al., 2018), a recommendation that we re-iterate in this article. These considerations are best addressed when improvements to stock assessment methodology are being considered (e.g. during “research” or “benchmark” assessments). However, there is opportunity to evaluate hypotheses about climate-related effects on fish stocks during routine “operational” assessments, which may help establish research priorities.

It has been historically challenging to account for and forecast shifts in stock distribution or productivity when multiple drivers are plausible (e.g. temperature vs. age-structured dynamics vs. fishing mortality: Thorson et al., 2017). However, advances in statistical modelling allow analysts to interpret a structural model (e.g. stock assessment, ecosystem, or spatio-temporal model) as a tool for attributing changes in a response to multiple direct or indirect causal drivers (Bell et al., 2014; Kaplan et al., 2016; Tommasi et al., 2017). These modelling approaches allow for improved attribution of productivity and distribution shifts to climate vs. fishing impacts and should be further explored and developed.

When processes are not well understood, multiple hypotheses may need to be considered, wherein comparative assessments test alternative mechanistic formulations (see Hare 2014 for additional discussion on importance of considering multiple hypotheses). Comparing diagnostics and predictive skill across a suite of models can help determine which model, or suite of models, should form the basis of scientific advice (Thorson, 2019,). Multi-model techniques are a useful tool that may enable scientists to more fully evaluate alternative states of nature, clarify mechanistic relationships, and characterize the uncertainty in the system and resultant forecasts (e.g. Ianelli et al., 2016).

Uncertainty in model forecasts can contribute to a lack of confidence by managers and stakeholders in the results, limiting the use of potentially more accurate, but less certain, scientific advice. Therefore, there is a need for continued evaluation of the degree to which models with environmental and climate linkages are improvements over non-climate linked models. Practitioners should use various model diagnostics, including retrospective and sensitivity analyses, as well as other predictive skill metrics, to evaluate the benefits and risks of including environmental covariates or time-varying parameters in models (Hobday et al., 2019). Seasonal to decadal-scale climate projections can be used retrospectively to estimate forecast skill relative to persistence forecasts, or to inform forecasts for models with mechanistic environmental linkages at scales relevant to fishery management (Tommasi et al., 2017). Regardless of whether environmental variables are included, the evaluation and quantification of model predictive skill should be a routine part of the stock assessment process and effectively communicated to managers and stakeholders as uncertainty around the results (Hobday et al., 2019).

Communicate scientific information to management

To facilitate informed management decisions that respond to a dynamic and changing environment, scientific information must be communicated effectively and efficiently. Currently, information on shifting distributions and changing productivity is rarely used in fishery management decisions. Skern-Mauritzen et al. (2016), in a global review of fish stock assessments, found that only 2% of the 1250 assessments analysed incorporated ecosystem information into tactical management advice. A more recent review focused on U.S. stocks found that close to 24% of the 206 assessments analysed (about half of U.S. federally managed stocks) included ecosystem information in the assessment report (Marshall et al., 2019). Certainly, progress is being made, but more could be done. This limited use of ecosystem information is partially a result of uncertainty in the underlying mechanisms, and partially a result of how the information is communicated to managers, which often leaves managers unsure of when and how the information could be used to adjust management actions. This highlights the need to re-evaluate how climate change and ecosystem information is communicated to managers, and to identify ways to better communicate and operationalize this information.

In the United States, the presentation of information on ecosystem condition to managers varies across the regions. The lack of a protocol for reporting ecosystem conditions hinders the ability of managers to account for changes in ecosystem conditions that affect stock assessment outcomes. Routine reporting and standardized templates would improve the communication of information on the effects of ecosystem dynamics on species distributions and productivity to managers. As part of this reporting, scientists should provide contextual information that changes are occurring with either positive or negative implications for management objectives. This process would benefit from a more coordinated effort between stock assessment and ecosystem science teams to provide this contextual information, drawing connections between ecosystem changes and stock assessments where appropriate and highlighting potential drivers of change. It is important that these interdisciplinary teams work with management partners to determine the best format and timing for presenting the new information to managers. A national initiative is underway in the United States to create a generic template that communicates stock-specific ecosystem considerations, referred to as Assessment Profile, Ecosystem Considerations, and Socioeconomics (APECS; Lynch et al., 2018; Shotwell et al., 2018). These APECS will provide two-page summaries of the important ecosystem and stock assessment information, serving as a useful tool to improve communication and comparison of results across stocks and regions.

An important, but challenging element to communicate to managers is the uncertainty, trade-offs, and risks associated with various states of nature and management decisions. Decision support tools (e.g. decision tables, decision trees, maps, etc.) that provide action-oriented information (e.g. trade-offs regarding different catch limits associated with a probabilistic risk of overfishing) should be considered for inclusion in assessment reports. Additionally, maps of changes in current and future projected species distributions, as related to management jurisdictions, could help managers understand and respond to species shifts.

Lastly, communication among scientists, managers, and stakeholders must be multidirectional, whereby all perspectives are heard and participants learn from each other. A critical component of this communication is regular engagement and coordination with relevant management bodies and fisheries stakeholder groups. This can be achieved through regular and open dialogue at management meetings, workshops, and debriefs with stock and ecosystem assessment scientists. The goal of these engagements is for all parties to arrive at a clear understanding of what management actions should be considered in response to changing systems. Additionally, establishing work teams that include both scientists and managers could help ensure that scientists are aware of spatial management programmes affecting species, and that managers are aware of changing scientific information on stock distributions and habitat use. Increased dialogue among all parties and a more collaborative and transparent decision-making process are important to building trust and scientific understanding, especially given the uncertainty and complexity of the challenges posed by changing climate and ocean conditions.

Manage fisheries under changing conditions

Fishery managers face the daunting challenge of determining appropriate responses to current and predicted changes in stock distribution and productivity. Changing species distributions and productivity can result in managers having to make controversial decisions, such as changes to allocation, spatio-temporal closures, stock status determinations, and catch limits. To help prepare them to make such decisions, scientists should: provide managers with the ability to evaluate management action under potential future scenarios; plan for emerging fisheries; evaluate time and area management measures; develop adaptive and responsive harvest control rules (HCRs); and consider population resilience, age structure, and genetic diversity when making management decisions.

To increase resiliency across shifting ecological and social conditions, Holsman et al. (in press) advocate for climate-resilient management portfolios that integrate near-term dynamic, medium-term adaptive, and long-term fixed management approaches across spatio-temporal scales. Tools that can aid managers in this endeavour include structured scenario planning, holistic ecosystem models, climate vulnerability and risk assessments, and management strategy evaluations (MSEs). Structured scenario planning is a strategic planning method used to visualize how alternative plausible futures might emerge, explore potential ways to prepare for those alternative futures, and evaluate how different actions or strategies would play out under the different plausible futures (Moore et al., 2013). Risk assessments and CVAs (Morrison et al., 2015; Holsman et al., 2017) inform management actions by identifying species and ecosystems at most risk, and therefore in need of additional data, analysis, and management action (e.g. Holsman et al., 2017). MSEs can be used to test the efficacy of various alternative management options in achieving predetermined management objectives. These analyses can be used to evaluate the robustness of alternative strategies under different climate change scenarios by including climate change signals in developing alternative operating models (A’mar et al., 2009; Tommasi et al., 2017). Managers should work with their scientific counterparts to conduct these analyses and use the results to help prioritize future actions.

Managers should plan for emerging fisheries within the context of a changing climate. A prudent and rudimentary management action to address potential movement of new species into fishery management areas would be for jurisdictions to review and update their list of authorized fisheries and gear. Additionally, jurisdictions should develop and document a process for making allocation decisions when stocks change their distributions. Several approaches have been suggested for making these decisions (see Bailey et al., 2013; Morrison and Termini, 2016). For example, allocations could be based not just on historical distribution, but on forecasts of distribution over a time horizon meaningful to management, based on models that have shown greater skill than a persistence forecast. Additionally, management bodies can use early warnings and spatial distribution projections to identify species that may shift across jurisdictional boundaries to help prepare for potential emerging fisheries and allocation decisions.

Changing environmental and oceanographic conditions can influence effectiveness of spatial management approaches by altering the distribution, spatial extent, and use of spawning, migration, and nursery habitats. Regulatory regimes that manage through time and area closures to protect important habitats or prohibit fishing during vulnerable parts of a stock’s life history (e.g. spawning and juvenile stages) can experience mismatches or lags between when change occurs and when management actions are implemented. Changing ocean and climate conditions may therefore require more proactive and dynamic fishery management. For proactive and dynamic management, near real-time biological, oceanographic, social and/or economic data can be used to forecast changes and evaluate and adjust temporal and spatial management actions to better align with changes in the resources being managed (Maxwell et al., 2015; Hobday et al., 2016; Hazen et al., 2018). A combination of traditional and more responsive management procedures could increase the speed and effectiveness of management decisions and their outcomes (Maxwell et al., 2015).

Catch limits in the United States are usually set using HCRs that adjust fishing mortality or catch as a function of a stock’s biomass, in accordance with stock status relative to benchmarks. These rules generally include precautionary buffers to account for associated scientific and management uncertainty (Brunel et al., 2010). Stock status determinations and HCRs are often developed under the assumption of fixed, or stationary biological reference points over time. Many HCRs perform poorly when environmental conditions change stock productivity and reference points such that the HCR gets incorrect information on the true status of the stock or is unresponsive to changes (A’mar et al., 2009; Brunel et al., 2010). Developing HCRs that account for uncertainty and known changes in environmental conditions affecting productivity may increase the successful management of fish stocks influenced by environmental forcing (Kvamsdal et al., 2016; Tommasi et al., 2017). For example, the HCR for Pacific sardine (Sardinops sagax) includes an environmentally dependent exploitation parameter, which serves to reduce the recommended harvest when ocean temperature is unfavourable for recruitment (e.g. cold years) and increase the harvest when ocean temperatures are more favourable (e.g. warm years; PFMC, 2014; Kvamsdal et al., 2016).

Managers can strengthen the adaptive capacity and resilience of stocks by protecting and enhancing the age structure and genetic diversity of a population. Genetic adaptation to climate change may be necessary, and management should aim to increase or preserve current genetic diversity, which provides the building blocks needed to adapt to a changing environment. Similarly, stocks that have maintained their age structure are more resilient to environmental change (Planque et al., 2010; Rouyer et al., 2011). Large females tend to produce more eggs, which are often larger leading to better provisioned offspring, and spawn over a greater time period and depth gradient than smaller females (Hixon et al., 2014). Thus, the removal of large females from a population can decrease the range of conditions experienced by eggs and larvae, reducing the likelihood that at least some eggs and larvae encounter environmental conditions beneficial to growth and adaptation.

Conclusions

Changing climate and ocean conditions are affecting fisheries around the globe in ways that are not routinely addressed in the science-to-management process. Species are shifting distributions and experiencing changes in productivity in response to changing ocean conditions, altering both ecosystem and fishery dynamics. Traditional methods and assumptions used in the fishery management process need to be adapted to ensure effective stewardship of living marine resources under changing conditions. There is growing recognition of the need for a more holistic and ecosystem-linked approach to fisheries management which can take into account these climate and environmentally driven changes (e.g. NCSS, EBFM policy and roadmap, SAIP). This article describes six key steps where improving capabilities would enable science agencies and their management partners to better understand and account for climate and ecosystem changes in fisheries management and therefore continually advance the ecosystem approach to fisheries management. This will involve addressing the challenges and gaps identified in this article—from detecting and understanding changes to analysing and communicating impacts of those changes to the relevant decision makers. The recommendations and related actions are intended to be applied broadly to support the full range of fishery management decisions in different regions. However, resource and other limitations may require assessment of how best to implement the recommendations. Although the recommendations presented here were largely developed in the current single-species context, it can be envisioned how these recommendations can be applied in a multispecies context, and thereby also improve multispecies assessments. As one example, forecasts developed for single-species distribution changes can be compared to highlight potential changes in species composition or species interactions in a region. Further, single-species assessment results and forecasts within an ecosystem could be incorporated into an ecosystem-level simulation model to inform managers of the likely responses and interacting effects of proposed management action across stocks.

Faced with shifting species distributions and changing productivity, fisheries scientists and management bodies should explore future scenarios, plan for emerging fisheries, re-evaluate their spatial and temporal management procedures, and develop responsive HCRs. Ultimately, management approaches that account for changes in the environment depend on data collected from survey and monitoring efforts and process-oriented research. It is important that the basic science activities adapt to changing data needs to differentiate between changes in abundance and shifts in distribution or productivity. Opportunities exist to make better use of current data collection efforts through standardizing the type and frequency of data collected on existing surveys and across jurisdictions. Other key opportunities include leveraging the capacity of fishermen and citizen scientists, and operationalizing advanced sampling technologies that can collect information at appropriate temporal and spatial scales.

Each step in the science-to-management process benefits from close collaboration and communication among scientists, managers, and stakeholders. Regularly scheduled workshops, debriefs, meetings, or other mechanisms to allow open dialogue and engagement between scientists and managers, are essential for effective communication and collaboration. Regular meetings would provide scientists with feedback from managers to help prioritize the types of analyses and evaluations that would be most beneficial to sustainable fisheries management.

Overall the recommendations identified in this article, if enacted at appropriate regional and national levels, will better equip fisheries science and management agencies to prepare for and respond to changing climate and ocean conditions, and thereby improve stewardship of the living marine resources upon which many people, businesses and communities depend.

Acknowledgements

The views expressed are those of the authors, and do not necessarily represent findings or policy of any government agency. We thank Kenric Osgood, Stephen Brown, Karyl Brewster-Geisz, Jennifer Cudney, and two anonymous reviewers for their constructive feedback on earlier versions of the document.

Funding

This research was performed while LAKB held an NRC Research Associateship award at the Northwest Fisheries Science Center and his contribution to this publication is partially funded by the Joint Institute for the Study of the Atmosphere and Ocean (JISAO) under NOAA Cooperative Agreement No. NA15OAR 4320063 and the NOAA National Protected Species Toolbox Initiative.

References