ICES Journal of Marine Science

ICES Journal of Marine Science (2014), 71(5), 1170–1173. doi:10.1093/icesjms/fsu026

Contribution to the Themed Section: ‘Integrated assessments’ Introduction Integrating what? Levels of marine ecosystem-based assessment

and management Downloaded from

Jason S. Link1* and Howard I. Browman2 1National Oceanic and Atmospheric Administration, National Marine Fisheries Service, Office of the Assistant Administrator, 166 Water St, Woods Hole, MA 02543, USA 2Marine Ecosystem Acoustics Group, Institute of Marine Research, Austevoll Research Station, 5392 Storebø, Norway http://icesjms.oxfordjournals.org/ *Corresponding author: tel: 1 508 495 2340; fax: 1 508 495 2258; e-mail: [email protected] Link, J. S., and Browman, H. I. 2014. Integrating what? Levels of marine ecosystem-based assessment and management – ICES Journal of Marine Science, 71: 1170–1173 Received 19 September 2013; accepted 31 January 2014; advance access publication 19 March 2014.

Integrated assessment requires examination of factors across biological major policy objective of many marine-oriented organizations—as by Howard Browman on July 2, 2014 hierarchies, taxonomic groups, ocean-use sectors, management objec- is clear from a perusal of the strategic plans of organizations such as tives, and scientific disciplines. The articles in this theme set represent ICES, PICES, FAO, UNEP, and NOAA. The need for integrated attempts to clarify and elaborate upon what integrated assessments are, assessments frequently arises in the context of discussions over with a particular emphasis on how they are being implemented. The implementing EBM. aim of thisthemed article set isto clarify the use of integratedassessment The term “integrated assessments” is perceived as mysterious and terminology and demonstrate, by presenting case studies, examples in ultimately unhelpful because it suffers from a plurality of definitions which integrated ecosystem assessments serve as useful tools to imple- and it is used in a multitude of contexts—i.e. it has high linguistic un- ment ecosystem-based management (EBM) while also identifying certainty. Thatiswhywe pose the questioninthe title ofthisintroduc- challenges that must be overcome for this to succeed. tion: what are we integrating and, hence, what are we assessing? In theory, EBM seeks to address the various natural and an- Returning to the EBM context for sustainably managing marine thropogenic pressures faced by the key components of marine resources, we note that there are, in fact, multiple levels at which an systems simultaneously. EBM also attempts to account for “cumu- “ecosystem approach” can be adopted in practice. To illustrate, we lative impacts” that might otherwise be overlooked. Nascent focus on the fisheries sector. There are levels of application for EBM attempts to implement EBM highlight the need—in practice—to that focus solely on fish stocks, levels that focus on fish stocks but address trade-offs across multiple objectives for a given system, in with ecosystem considerations incorporated, ecosystem levels that a coordinated and comprehensive manner. During the past focus solely on the fisheries sector but for the full system of fisheries decade, the discussion over EBM has shifted from “what is it and and stocks, and the full set of ocean-use sectors impacted by and why should we do it” (Link, 2002; Browman and Stergiou, 2004, impacting the fisheries sector (Table 1). For example, consider 2005) to “how can we do it and when can we operationalize it” forage stocks such as small pelagic fish. For an ecosystem approach (Arkema et al., 2006; Link, 2010; Berkes, 2012). Marine EBM (e.g. to fisheries (EAF) that takes a stock focus, one would need to consider Levin and Lubchenco, 2008; McLeod and Leslie, 2009)andsimilar the effects of environmental factors (e.g. temperature changes or ecosystem-based efforts for more specific ocean-use sectors, such as NAO events) and ecological factors (e.g. predator removals or ecosystem-based fisheries management (EBFM; e.g. Pikitch et al., models of multispecies interactions) in addition to targeted fisheries 2004; Link, 2010) or integrated coastal-zone management (e.g. removals to truly grasp what is driving the population dynamics of Cicin-Sain and Knecht, 1998; Moksness et al., 2013), have become such stocks. Using the same type of focal species as an example, for the mandated approach to managing ocean resources. EBM is a EBFM that takes a system focus in the fisheries sector, one would

Published by Oxford University Press 2014. This work is written by (a) US Government employee(s) and is in the public domain in the US. Integrated assessments: theme set introduction 1171

Table 1. Levels of application for EBM for the fisheries sector, with notes on particular features of each level Level of EBM in fisheries sector

Feature EBM EBFM EAF Classical FM Sectoral focus All Fisheries Fisheries Fisheries Focus of the All Community/whole system Stock/population Stock/population biological hierarchy Aggregate Maybe multispecies Evaluation processes Integrated ecosystem Integrated ecosystem Integrated stock assessments Stock assessments used to do it assessments assessments, fisheries sector focus Primary objective of Address cross-sector Address fishery sector, living Delineate status of stock/s Delineate status of the analysis trade-offs marine resource trade-offs stock Ascertain ecosystem goods Ascertain ecosystem Ascertain stock productivity Ascertain stock

and services productivity productivity Downloaded from Identify best mix of services Identify best mix of goods Identify levels of optimal stock Identify levels of across goods and services across fisheries and stocks production cognizant of optimal stock ecosystem factors production Evaluate cross-sector Evaluate within-sector Evaluate within-stock effects of Evaluate within-stock cumulative effects cumulative effects multiple drivers effects of fishing

Primary outputs for Provide systemic reference Provide systemic reference Provide fishery stock reference Provide fishery stock http://icesjms.oxfordjournals.org/ decision criteria points points points reference points Analytical tools EMs EMs ESAMs SAMs available to do it RA MSMs MSMs MSE RA MSE MSE EMs, ecosystem models; RA, risk analyses; MSE, management strategy evaluations; ESAMs, extended stock assessment models; MSMs, multispecies models; SAMs, stock assessment models; EBM, ecosystem-based management; EBFM, ecosystem-based fisheries management; EAF, ecosystem approaches to fisheries; FM, fisheries management. For full EBM, this would be consideration of all ocean-use sectors. For EBFM, this would be only the fisheries sector with an ecosystem emphasis. For EAF, this would be fisheries management with inclusion of ecosystem considerations. For FM, there would be no explicit consideration of ecosystem features. by Howard Browman on July 2, 2014 have to consider not only the impacts of other factors on these forage stock assessments integrate a wide array of standard data streams (e.g. stocks, but also the dynamics of these forage stocks on other parts of Fournier and Archibald, 1982;Methot,1989;PuntandHilborn,1997; the ecosystem. For instance, there are seabirds or marine mammals ICES, 2012). Here, integrated stock assessments, although often that have some form of protected or conservation status and that termed integrated assessments or integrated analyses, are referring are highly dependent on small pelagic forage fish. There are commer- to other factors beyond the typical catch, abundance, and biology cially targeted groundfish that are also major predators of these small data,particularlybringingmorestatisticaltechniquesandapproaches pelagic forage fish. There are also multiple fisheries operating on both into consideration. These integrated stock assessments have been dis- the groundfish and the small pelagic species. In such a case, clearly a cussed in many other contexts (e.g. Fournier and Archibald, 1982; more integrated, “bigger picture” evaluation of the whole system and Maunder and Punt, 2013; Methot and Wetzel, 2013). In this EAF howitfitstogetherisneededtoaddressthepotentialtrade-offsamong context, such integrated stock assessments use extended stock assess- the different uses of and impacts to these forage stocks. Further, if ment models that incorporate other facets of the ecology, oceanog- these forage stocks represent a key pathway of energy from lower raphy, or environment to explain and predict stock dynamics (e.g. trophic levels to upper trophic levels (which they typically do), then Maunder and Watters, 2003; Methot and Wetzel, 2013). A growing the resilience, structure, and functioning of the system would need bodyofmultispeciesmodelscanalsoexploresomefacetsofthesecon- to be evaluated. For an EBM that covers all ocean-use sectors, consid- siderations (e.g. Townsend et al., 2008; ICES, 2011) and, although eration of these small pelagics and their role in the ecosystem is war- they could be used in an integrated stock assessment process and ranted in a broader context for anthropogenic drivers such as power their outputs are often used as inputs into extended stock assess- plant discharges (thermal impacts), eutrophication, toxin deposition, ment models, they are not typically termed integrated stock assess- hydroelectric energy generation, dredging for navigation safety, and ment models themselves. Mo¨llmann et al. (2014) provide an similar uses that might impact the habitats of these species. One can en- excellent example covering the range of assessment models used to vision similar examples for other ocean-usesectors,butfacetsofthisfull inform the management of living marine resources in the Baltic. range of considerations across different levels of EBM are demonstrated Their ensemble approach of using multiple models to inform the in Mo¨llmann et al. (2014).Thesalientpointbeingthatonecandointe- decision-making process has strong potential for both integrated grated assessments at all these levels of application, but they are called stock assessments and advancing EBFM more broadly by providing different things. multiple perspectives on the same problem without being biased For EBM that focuses only on single species of fish (Table 1,EAF), fromusingonlyalimitedsetofmodelassumptions, modelstructures, integrated stock assessments can/have been used. Certainly, classical or modelling frameworks. 1172 J. S. Link and H. I. Browman

For EBM in the fisheries sector that includes an ecosystem em- † Are integrated assessments (integrated stock assessments or inte- phasis (Table 1, EBFM), parts of integrated ecosystem assessments grated ecosystem assessments) an improvement over how we do can/have been used. Integrated ecosystem assessments are, by their assessments now and, if so, how? nature, designed to be multisectoral (Levin et al., 2008, 2009)but, † Are integrated approaches important to consider as we move clearly, the overall process can be adopted for sector-specific applica- towards implementing EBM at all of its various levels? tions. A large part of doing EBFM at this level entails exploring trade- offs among different fishery fleets, targeted species, and non-targeted † Are the strengths and weaknesses of such integrated approaches species. Mo¨llmannet al. (2014) describe somefacetsof integrated eco- readily apparent? system assessments that have been used in the fishery sector for the † What needs to happen next to better use such integrated Baltic, capitalizing on the other ecosystem factors that so readily and approaches? What is impeding the fuller implementation of so obviously impact fish in that Sea. Samhouri et al. (2014) note these approaches? how features of integrated ecosystem assessments have, to date, largely been used to advance EBFM in a fisheries context in the We trust that this themed set of articles will be as thought-provoking USA.Theyparticularlyemphasizetheimportantroleof contextualin- as it is informative. We hope that our start at delineating the levels of formation and the use of indicators for ecosystem considerations in EBM and the types of integrated assessments serves as a useful means the fisheries management process. The contributions to this article to minimize some of the linguistic uncertainty that often surrounds EBM. As we move forward with the implementation of EBM at all themesetpresentvariousrenditionsofanintegratedecosystemassess- Downloaded from ment process and, following from them, it is not difficult to envision of its various levels, we hope that the clarifications provided here, how this approach could be adapted to other ocean-use sectors. the elucidations of the different integrated assessments, identification For all ocean-use sectors, integrated ecosystem assessments are of key challenges, and the examples demonstrated in the various con- themostappropriatetool(Table1,EBM).Nascentintegratedecosys- tributions, serve to move forward the implementation of EBM. tem assessment efforts have been underway for approximately half a

Acknowledgements http://icesjms.oxfordjournals.org/ decade, though mostly piecemeal and usually not in a fully imple- mented manner (Dickey-Collas, 2014). Progress from the few HB’s editorial contribution to this article theme set was supported instances where integrated ecosystem assessments are beginning to by the Institute of Marine Research, Norway, Projects # 81529 be more fully utilized is reported in this theme set, with notes on (“Fine scale interactions in the plankton”) and 83741 (“Scientific lessons learned (Samhouri et al., 2014) and guidelines for imple- publishing and editing”). We thank the authors who contributed menting integrated ecosystem assessments (Levin et al., 2014) high- to this theme set and our many colleagues who have discussed inte- lighted. This is done with the intention of better informing multiple grated assessments with us over the years. ocean-use sector management and avoiding the clashes that often arise when competing ocean-use sectors do not communicate their References objectives. One of the key findings from both the Samhouri et al. Arkema, K. K., Abramson, S. C., and Dewsbury, B. M. 2006. by Howard Browman on July 2, 2014 andLevin et al. worksis theimportance of earlyandfrequent engage- Marine ecosystem-based management: from characterization to ment of stakeholders in the process. This observation is reinforced by implementation. Frontiers in Ecology and the Environment, 4: the frank insights about the development of the ICES approach to 525–532. integrated ecosystem assessment offered by Dickey-Collas (2014). Berkes, F. 2012. Implementing ecosystem-based management: evolu- tion or revolution? Fish and Fisheries, 13: 465–476. The North American examples on integrated ecosystem assess- Browman, H. I., and Stergiou, K. I. 2004. Perspectives on ecosystem- ments are complemented by broader European perspectives (Dickey- based approaches to the management of marine resources. Marine Collas, 2014; Walther and Mo¨llmann, 2014). These authors note pro- Ecology Progress Series, 274: 269–303. gressto-date on integrated ecosystem assessmentsinthe ICES context, Browman, H. I., and Stergiou, K. I. 2005. Politics and socio-economics particularly identifying some important challenges facing their imple- of ecosystem-based management of marine resources. Marine mentation. They note that these challenges could in fact impede the Ecology Progress Series, 300: 241–296. ability to fully embrace EBM. The European perspectives bemoan Cicin-Sain, B., and Knecht, R. 1998. Integrated Coastal and Ocean the generally slow implementation of integrated ecosystem assess- Management: Conceptsand Practices. Island Press, Washington DC. ments. Yet an important point that Samhouri et al. (2014) make is Dickey-Collas, M. 2014. Why the complex nature of integrated ecosys- that it takes time to develop new processes and protocols, so delays tem assessments requires a flexible and adaptive approach. ICES Journal of Marine Science, 71: 1174–1182. in the use of integrated ecosystem assessments and their products Fournier, D., and Archibald, C. P.1982. A general theoryfor analyzing catch should not necessarily be viewed as a hurdle to overcome but rather at age data. Canadian Journal of Fisheries and Aquatic Sciences, 39: a natural part of the research-to-operations developmental process. 1195–1207. One item that Walther and Mo¨llmann note is the need to (re)- ICES. 2011. Report of the Working Group on Multispecies Assessment consider institutional structures and frameworks for using infor- Methods (WGSAM), 10–14 October 2011, Woods Hole, USA. ICES mation generated from integrated ecosystem assessments. This Document CM 2011/SSGSUE:10. 229 pp. observation is repeated in other contributions to this theme set, es- ICES. 2012. Report on the Classification of Stock Assessment Methods pecially Samhouri et al. (2014). There may, in fact, be no clear, easy, Developed by SISAM. ICES Document CM 2012/ACOM/SCICOM: or imminent answers, but that discussion should continue. 01. 15 pp. Dickey-Collas (2014) notes that although the way forward may Levin, P. S., Fogarty, M., Matlock, G., and Ernst, M. 2008. Integrated ecosystem assessments. US Department of Commerce, NOAA indeed be messy, the need to consider broader factors when man- Technical Memorandum, NMFS-NWFSC 92. 20 pp. aging marine resources is highly warranted and an adaptable, in- Levin, P. S., Fogarty, M. J., Murawski, S. A., and Fluharty, D. 2009. novative approach across multiple considerations is very much Integrated ecosystem assessments: developing the scientific basis required—just as has been envisioned for integrated assessments. for ecosystem-based management of the ocean. PLoS Biology, 7: We conclude by posing a few questions for readers to ponder. e1000014. doi:1000010.1001371/journal.pbio.1000014. Integrated assessments: theme set introduction 1173

Levin, P. S., Kelble, C., Shuford, R., Ainsworth, C., deReynier, Y., Moksness,E.,Dahl,E.,andSto´ ttrup, J. 2013. Global Challenges Dunsmore, R., Fogarty, M., et al. 2014. Guidance for implementation in Integrated Coastal Zone Management. Wiley-Blackwell, of integrated ecosystem assessments: a USperspective. ICESJournalof Oxford, UK. Marine Science, this volume. Mo¨llmann, C., Lindegren, M., Blenckner, T., Bergstro¨m, L., Casini, M., Levin, S. A., and Lubchenco, J. 2008. Resilience, robustness, and marine Diekmann, R., Flinkman, J., et al. 2014. Implementing ecosystem- ecosystem-based management. Bioscience, 58: 27–32. based fisheries management: from single-species to integrated eco- Link, J. S. 2002. What does ecosystem-based fisheries management system assessment and advice for Baltic Sea fish stocks. ICES mean? Fisheries, 27: 18–21. Journal of Marine Science, 71: 1187–1197. Link, J. S. 2010. Ecosystem-Based Fisheries Management: Confronting Pikitch, E.K., Santora,C., Babcock, E. A., Bakun, A., Bonfil, R., Conover, Tradeoffs. Cambridge University Press, Cambridge, UK. D. O., Dayton, P., et al. 2004. Ecosystem-based fishery management. Science, 305: 346–347. Maunder, M., and Punt, A. 2013. A review of integrated analysis in fish- Punt, A. E., and Hilborn, R. 1997. Fisheries stock assessment and deci- eries stock assessment. Fisheries Research, 142: 61–74. sion analysis: the Bayesian approach. Reviews in Fish Biology and Maunder, M. N., and Watters, G. M. 2003. A general framework for inte- Fisheries, 7: 35–63. grating environmentaltime series into stock assessment models: model Samhouri, J.,Haupt, A., Levin, P.,Link, J.,and Shuford, R. 2014. Lessons description, simulation testing, and example. Fisheries Bulletin, 101: learned fromdevelopingintegrated ecosystemassessmentstoinform 89–99. marine ecosystem-based management in the USA. ICES Journal of McLeod, K. L., and Leslie, H. M. (Eds) 2009. Ecosystem-Based Marine Science, 71: 1205–1215.

Management for the Oceans. Island Press, Washington, DC. Townsend, H. M., Link, J. S., Osgood, K. E., Gedamke, T., Watters, G. Downloaded from Methot, R. D. 1989. Synthetic estimates of historical and current biomass M., Polovina, J. J., Levin, P. S., et al. (Eds). 2008. Report of the of northern anchovy, Engraulis mordax. American Fisheries Society NEMoW (National Ecosystem Modeling Workshop). NOAA Symposium, 6: 66–82. Technical Memorandum, NMFS-F/SPO-87. 93 pp. Methot, R. M., and Wetzel, C. R. 2013. Stock synthesis: a biological and Walther, Y., and Mo¨llmann, C. 2014. Bringing integrated ecosystem statistical framework for fish stock assessment and fishery manage- assessments to real life—a scientific framework for ICES. ICES

ment. Fisheries Research, 142: 86–99. Journal of Marine Science, 71: 1183–1186. http://icesjms.oxfordjournals.org/ by Howard Browman on July 2, 2014 ICES Journal of Marine Science

ICES Journal of Marine Science (2014), 71(5), 1174–1182. doi:10.1093/icesjms/fsu027

Contribution to the Themed Section: ‘Integrated assessments’ Original Article Why the complex nature of integrated ecosystem assessments

requires a flexible and adaptive approach Downloaded from

Mark Dickey-Collas1,2* 1ICES, H. C. Andersens Boulevard 44-46, DK-1553 Copenhagen, Denmark 2Wageningen IMARES, 1970 AB, IJmuiden, The Netherlands http://icesjms.oxfordjournals.org/ *Corresponding author: tel: +45 3338 6759; fax: +45 3338 6700; e-mail: [email protected] Dickey-Collas, M. 2014. Why the complex nature of integrated ecosystem assessments requires a flexible and adaptive approach. – ICES Journal of Marine Science, 71: 1174–1182. Received 21 December 2013; accepted 31 January 2014; advance access publication 6 March 2014.

This article considers the approach taken by the ICES to integrated ecosystem assessments (IEAs) in the context of the wider evolution of IEAs and the science/policy landscape within the ICES region. It looks forward and considers the challenges facing the development of IEAs, specifically those of scoping for objectives, participatory engagement, developing indicators and targets, risk analysis, and creating tools to evaluate management measures for marine anthropogenic activities. It concludes that expectations that the implementation of IEAs will by Howard Browman on July 2, 2014 take an ordered, stepwise approach will lead to disappointment and frustration. This is a consequence of the need to operate in an adaptive manner in a complex system. The ecosystem, the science support infrastructure, and the governance systems are all complex. Plus when engaged in a debate about societal objectives, we expect to encounter a complex and changing landscape. As a community, the challenge is to find leverage mechanisms to encourage IEA efforts to provide insights and tools within resources. We will need to innovate and be responsive to the complexity of the ecosystem and governance structures encountered when performing IEA. Keywords: ecosystem approach, fisheries, HELCOM, MSFD, OSPAR, regionalization.

Introduction Jennings and Rice (2011) suggest that progress towards an ecosys- The concept of integrated ecosystem assessments (IEAs), in the tem approach to fisheries has been slow because of the lack of specific marine sphere, germinated in response to the development of the environmental, social, and economic objectives, with no agreed ecosystem approach to fisheries management (EAFM, Rice 2011). guidance on trade-offs from policy-makers. They add that it is The political incentive behind EAFM is clear (Murawski, 2007; impacted by a framework for decision-making in Europe with a fish- Jennings and Rice, 2011), but operationalizing it has been, and eries policy that is skewed towards short-term national interests. It is still is, a challenge (Rice, 2011). The ecosystem approach is certainly likely that the lack of guidance for trade-offs and objectives and a re- not a new approach (Jennings, 2004; Garcia and Cochrane, 2005) luctance by those engaged in the debate to bring non-ecologiststo the but it requires management to take a different path, although table will hamper any push towards developing tools for the ecosys- perhaps not a paradigm shift, to be successful (Sissenwine and tem approach (Degnbol and McCay, 2006; Francis et al., 2011). Murawski, 2004; Murawski, 2007). Core to the ecosystem approach “The challenge for EAM/EBM will be to link the incremental is managing the impact, or the pressure of humans on the marine changes in selected indicators to a target end state to the societal ecosystem; the human dimension (De Young et al., 2008; Rice, costs and benefits of achieving the end state ” (Murawski, 2007). 2011). Jennings (2004) sums this up as “The ecosystem approach Part of making an ecosystem approach operational is ensuring is variously defined, but principally puts emphasis on a manage- that the interactions between fisheries and other sectors are equit- ment regime that maintains the health of the ecosystem alongside able, with inclusiveness in decision-making (Levin et al., 2009; appropriate human use of the marine environment, for the Rice, 2011). This has led to the development of IEAs as a concept benefit of current and future generations.” (Link, 2012), although like the ecosystem approach, it is difficult

# International Council for the Exploration of the Sea 2014. All rights reserved. For Permissions, please email: [email protected] ICES approach to IEA 1175 to provide a working definition for IEAs as they encompass a frame- understand all the processes, and even if we did, it is unlikely that full work and executable practice. The ICES “benchmark” group on IEA understanding would make our political decision-making easier. (WKBEMIA) defined IEA as an interdisciplinary process of com- Rice (2011) stresses that the process must maximize the use of avail- bining, interpreting, and communicating knowledge from diverse able information rather than emphasize the areas of uncertainty. scientific disciplines, in such a way that the interactions of a Likewise, the idea that we can “data-collect” our way to the ecosys- problem can be evaluated and thus provide useful information to tem approach, or an IEA, must be recognized as a fallacy. Levin et al. decision-makers (ICES, 2012a). Levin et al. (2009) define IEA as (2009) point out that massing data simply cannot tell us how to im- “a synthesis and quantitative analysis of information on relevant plement EBM or determine priorities for management. They say physical, chemical, ecological and human processes in relation to that reductionist approaches can result in researchers and policy- specified ecosystem management objectives”. Defining IEA is a makers “drowning in data but gasping for knowledge”. Tallying challenge, especially because the suggested approach to IEA is adap- the status or trends of various components cannot inform EBM tive and iterative; learning by doing (De Young et al., 2008). I see IEA (Levin et al., 2009). As Degnbol et al. (2006) claim, “improvements as a process that leads to the provision of joined-up and consistent in fisheries management will be realized not through the promotion advice that addresses society’s needs to manage anthropogenic of technical fixes but instead by embracing and responding to the pressures on the marine ecosystem. Importantly, it does not lead complexity of the management problem”. to one answer, but provides the information and knowledge to In this article, I reflect on my experience of working with IEA

facilitate exploring the space for decision-making and policy devel- groups over the last 2 years. I consider the approach taken by the Downloaded from opment. It is regional in scope. In the ICES context, thereis a distinct ICES community through it IEA groups in the context of the difference between IEA and developing models of ecosystem func- wider IEA debate. I also look forward and consider the challenges tioning. A framework to make the IEA approach operational has that face the development of IEAs, specifically those of scoping for been evolving and can be traced through the documents from objectives, developing indicators and targets, risk analysis, and cre- FAO (2003), ICES (2005), and Levin et al. (2009). The framework ating tools to evaluate measures to manage marine anthropogenic can be summed up as scoping, developing indicators and targets, activities. Almost all researchers involved in the development of http://icesjms.oxfordjournals.org/ risk analysis, assessing ecosystem status, then evaluating the strategy IEAs recognize that an adaptive and iterative process is crucial to and return to any previous stage that requires further development success (Samhouri et al., 2014). However, many members of the (Levin et al., 2009). Samhouri et al. (2014) explore this framework ICES community have expressed frustration over the “institutional further with specific case studies from the USA. and governance structures” and the limited resourcing of develop- In an inherently chaotic system, we should aim to manage an- ment that challenge implementation of IEAs (ICES, 2012a; thropogenic impacts in an integrated manner and respond to un- Walther and Mo¨llmann, 2014). In this article, I would like to foreseen changes as they occur without the requirement to be able explore whether a clearly prescribed framework or high-level stra- to quantify every functional relationship (Schellnhuber and tegic steering can be expected (Levin et al., 2014; Mo¨llmann et al., Sahagian, 2002; Clark et al., 2005). This requires an iterative 2014), and whether its absence is a block on progress? by Howard Browman on July 2, 2014 process linked to adaptive management. Linking a scientific inves- tigation to a societal debate on management objectives, trade-offs, ICES approach and tools for analysis may well challenge those that see science as a The organizational approach used to develop IEAs by ICES is search for pure truth and not part of a societal debate. Many described in Walther and Mo¨llmann (2014). This approach has argue that the science is not yet good enough, nor based on strong evolved mostly through the efforts of individual researchers to enough evidence to enter that debate. To paraphrase myth build multidisciplinary teams and is characterized by regional number 4 in Murawski (2007), people think that insufficient infor- groups, each acting in slightly different ways, to address slightly dif- mation limits the application of the ecosystem approach. This is also ferent challenges. There is one group based in the northwest allied to the idea that a quantitative model of species interactions Atlantic, but the majority of groups work on European regional among all components is necessary to guide the ecosystem approach seas. The new ICES strategic plan (ICES, 2014) places developing (myth 8 in Murawski 2007). It is necessary to understand the broad integrated ecosystem understanding at its core for the next dynamics of the ecosystem state and function (a “macroscope” ap- 5 years. This puts the regional groups at the centre of the ICES proach, Schellnhuber and Sahagian, 2002). However, we will never network. From the start, it is important to note that none of the

Table 1. ICES IEA Working Groups that were considered in this article, including plans for their future. Region Group Years active Future plans* Baltic Sea WKIAB 2006 WGIAB Annually 2007 to present 2014, 2015 Northwest Atlantic WGNARS Annually 2010 to present 2014, 2015, 2016 Northeast Atlantic WGEAWESS 2011 and 2013 2014, 2015, 2016 North Sea WGHAME 2010 WGINOSE Annually 2011 to present 2014, 2015, 2016 Norwegian Sea WGINOR 2013 2014, 2015 Barents Sea WGIBAR Yet to meet 2014, 2015, 2016 *Agreed terms of reference for these years. WKIAB- ICES/BSRP/HELCOM, Workshop on Developing a Framework for Integrated Assessment for the Baltic Sea; WGIAB-ICES/HELCOM, Working Group on Integrated Assessments of the Baltic Sea; WGNARS, Working Group on the Northwest Atlantic Regional Sea; WGEAWESS, Working Group on Ecosystem Assessment of Western European Shelf Seas; WGHAME, Working Group on Holistic Assessments of Regional Marine Ecosystems; WGINOSE, Working Group on Integrated Assessments of the North Sea; WGINOR, Working Group on the Integrated Assessments of the Norwegian Sea; WGIBAR, Working Group on the Integrated Assessments of the Barents Sea. More complete descriptions of the groups are given in Walther and Mo¨llmann (2014). 1176 M. Dickey-Collas groups have been asked, or proposed themselves, to carry out an (ii) Monitoring—collecting datasets and monitoring of the eco- actual ecosystem assessment as yet. The groups have been tasked system. This should be associated with ecosystem modelling with developing methods and tools for IEAs. When reading the with the overall aim to review the state of the system and reports of the ICES IEA groups (Table 1, and see Walther and monitor progress of indicators in relation to thresholds or Mo¨llmann, 2014), certain frustrations become apparent. The targets. groups ask for high-level stewardship, there is a crying out for a (iii) Developing indicators, targets, and reference points— simpler governance structure and clearer defined management developing indicators forecosystem functioning, performance objectives. A huge amount of work has been done by the groups of management measures, decision rules, or communication and the outputs have been substantial, with analysis of trends in of management objectives and action. These indicators must the state of the ecosystems, much focus on fisheries issues and be associated with targets and reference points. initial exploration of the Levin et al. (2009) approach. When reading the reports, it is also evident that some groups have fallen (iv) Assessing cumulative effects—develop methods (quantitative into traps that could hamper development of IEAs. or qualitative) for assessing a range of anthropogenic pressures In my mind, there are some key issues that researchers need to both unisectoral and multisectoral. Methods for combining consider when pushing forward IEA. These key issues are based indicators need to be developed to allow an IEA to reach an on the work by FAO (2003), ICES (2005), and Levin et al. (2009) overall conclusion about the state of the ecosystem.

and can be grouped into the concepts and techniques (the practical (v) Risk analysis—exploring risk analysis tools to provide infor- Downloaded from method development and implementation). The key concepts are: mation on potential scenarios and trade-offs associated with specific decisions. Examining trade-offs by looking at risk ana- lysis in relation to potential management objectives from (i) The human dimension—acknowledge that humans are across sectors, or choices within a single sector. central to the challenge. IEAs are management tools for main-

taining the health of the ecosystem alongside the appropriate http://icesjms.oxfordjournals.org/ What follows below is my personal interpretation of how we have human use of the marine environment. IEAs are as much explored the key ideas of IEAs in ICES. It is based on my reading about a societal debate as about the results of a scientific study. of reports and discussions with the active IEA groups. The aim of (ii) Ecosystems vary—thereis no such thing as a marine ecosystem my comments is to explore what we mean in practice in ICES in equilibrium. This means that predictions beyond are known when we work on IEAs and are aimed at setting the stage for the knowledge should be challenged and the state of the ecosystem further development of IEAs throughout the ICES community. should always be monitored. Links between components of ICES in its new strategic plan has set the target of carrying out the system and potential responses to management action example IEAs in the next few years, so the efforts of the ICES com- will not be linear. Approaches that suggest that the future munity now needs to include the execution of IEAs along with the will behave like the past should be challenged. further development of methods and tools. The current IEA by Howard Browman on July 2, 2014 (iii) Seek operational objectives—they are the cornerstone of IEAs. groups were tasked with method development. All groups have They are not usually predetermined, but best derived through made progress on building datasets and considering monitoring a debate where stakeholders explore the consequences of their needs. They acknowledge the need to set boundaries (regional, dis- objectives through models and other tools. The process of how cipline, and sector). Many suggest dividing their regions based on the objectives are derived might be complex and multilayered. ecosystem functioning, and a few appear to have considered the impact of these divisions on management. Almost none of the (iv) Integrate across sectors—IEAs should aim to integrate across groups have carried out an effective scoping exercise. This is issues, sectors, and stakeholders. Although individual projects despite links with regional seas conventions and probably reflects orteamsshould always definethe scope and limitsoftheir work. their interpretation of method development being their central (v) Understanding is what matters—data acquisition is different purpose. Most groups suggest that governance issues prevent from knowledge acquisition. You will never fully understand either participatory approaches or integration of management the system in which you operate, and it will always change. across marine sectors, which suggests that carrying out a scoping ex- Collecting data in isolation without synthesis and testing of ercise is crucial to method development and IEA implementation. ideas will not build knowledge but be aware that even a full With regard to the overall concept of IEAs, there still appears to understanding will not replace the societal debate required be a reluctance to think conceptuallyabout the human dimension. It to make decisions. is not clear whether all the groups see IEA as tools to balance diverse (vi) Do it again and again—it is an iterative process. The methods societal objectives. Some appear challenged with the concept that being developed should be part of an adaptive management indicators can exist that address societies priorities rather than eco- framework that will constantly evolve and change based on system functioning. All groups acknowledge that ecosystems are knowledge acquisition and the priorities of society. influenced by a range of drivers that impact at differing time-scales. Studies examining the trends in the state of the ecosystem dominate some of the groups. Every group reports that they are struggling to On the practical side, the techniques used should consider: work across sectors. Some groups appear unclear about the concept of adaptive management, whereas others are developing techniques (i) Scoping—of societal, managerial, and operational objectives. to make use of their improving understanding. Many groups In addition to this, it is important to set spatial boundaries and prioritize datacollection. As yet, there appears to be limited progress the subdivision of regions in relation to ecosystem dynamics on developing methods to synthesize knowledge to inform or and management objectives. It is also practical to set the sec- explore potential management action. In my mind, some IEA toral boundaries too. researchers see themselves as teams working on ecosystem ICES approach to IEA 1177 description rather than IEAs as defined by WKBEMIA (ICES, to “green growth” in the EU Horizon 2020 research programme. It is 2012a)orLevin et al. (2009). often said that scoping is an initial activity, but in reality scoping is The groups seem most confident and happiest when working in required continuously throughout the process (ICES, 2012a; the core ICES disciplines of fisheries and the ecology of marine verte- Samhouri et al., 2014). In ICES, a wealth of experience has been gen- brates and copepods. Biodiversity and habitat quality/connectivity erated in describing objectives through participatory modelling of have been considered by a minority of groups. On the European single stock fisheries management plans (where scientists and stake- side, up until 2013, no groups had really considered non-fishing- holders work together to build and explore the models used to test relatedpressures,butthathasrecentlychanged,especiallywiththede- management plans) and exploring policy objectives for marine velopment of the ecosystem overviews (ICES, 2013a; Walther and spatial planning (e.g. Schwach et al., 2007; Degnbol and Wilson, Mo¨llmann, 2014). Some participants appear to be aware of the mul- 2008; Mackinson et al., 2011; Dankel et al., 2012; Ro¨ckmann et al., tiple roles played by indicators, and the contrasting ideas ofindicators 2012). This participatory approach is really what is required in and targets for management and monitoring of the ecosystem state. IEA. Scoping must be a dialogue between stakeholders. Often the The groups appear challenged by the prospect of proposing indica- policy developers, industrial concerns or NGOs are not aware of tors; this is despite the large amount of work on indicators that has the potential options and trade-offs until they see the ability of the been published by the ICES group WGECO (ICES, 2013b)overthe tools or the dynamics of the models. Scientists can operate as “trans- last decade. Most groups have had terms of references that asked parency brokers” exploring policy options with the range of stake-

for indicator development and delivery, but few indicators have holders (P. Degnbol, ICES, pers. com.), thus providing useable Downloaded from been forthcoming. Few IEA groups, as yet, have suggested thresholds, knowledge (Haas, 2004). Our experience in ICES of developing targets, or reference points for integrated advice, although the Baltic multispecies advice for the Baltic and North Sea shows that a key IEA group has provided advice on future scenarios for multispecies aspect of providing advice is finding the appropriate communica- options for Baltic fisheries (Mo¨llmann et al., 2014). Proposing tion approach to describe the trade-offs (ICES, 2012b). This com- targets or reference points would facilitate a dialogue to start munication approach comes before any consideration of the with policy developers. In recent years, some groups have begun management/policy objectives. Assuming that policy developers http://icesjms.oxfordjournals.org/ developing risk-based approaches and exploring methods to know what they want before an IEA exercise is unrealistic. explore trade-offs. Assuming that there is one set of unified managers, with one object- The conclusion of the ICES Benchmarking workshop on IEA ive, is even more unlikely (van Leeuwen et al., 2012). (ICES, 2012a) was that almost none of these key concepts or activ- One of the most important parts of the scoping exercise is to es- ities can be carried out in isolation or in a specific order. They devel- tablish boundaries around the conceptual space being covered by oped the idea of an IEA cloud, where the iterative process is not the IEA, the factors that will be integrated and the objectives that circular, or linear, but is structured as a multidimensional matrix. will be included. The range of relevant time-scales, from short-term It is well known that a multidimensional approach is difficult to goals to longer-term objectives, must also be scoped so that develo- steer, and no one individual can be fully aware of all activities pers of IEAs are aware of the multiple spatial and temporal scales at by Howard Browman on July 2, 2014 and all links, so the approach stresses the importance of teams which IEAs operate (De Young et al., 2008). working towards a shared vision, with a strong need for communi- In some cases, the policy agenda jumps ahead of science cation between the main players. know-how (Rice, 2011). A good example of this is the Marine Strategy Framework Directive (MSFD; EC, 2008) in the EU. Looking forward The MSFD can be viewed as an imposition and clarification of the Considering all these issues, we must explore the opportunities higher management objectives (Good Environmental Status, GES) offered by the current system and view the perceived limitations for the EU marine environment including fisheries (Ratza et al., as challenges. Governance mechanisms are always challenging; 2010). However, in practice, many suggest that the MSFD is ambigu- whether in Europe or North America. Resources are becoming ous with unclear boundaries and conflicting objectives (van Leeuwen ever more limited, despite large increases in governance challenges. et al., 2012; Ounanian et al., 2012). Trying to define the operational Society expects informed decisions from policy developers and yet objectives and the indicators from the MSFD is difficult with the seems unprepared for the resource implications of that expectation. GES descriptors being a mixture of ecosystem components, attri- The sections below with further explore how we can use the butes, and pressures (S. Jennings, CEFAS, pers. com.). However, it strengths in the ICES community to further develop IEAs specific- can be viewed as the result of a politically imposed scoping exercise. ally on the key techniques of scoping, use of indicators, risk analysis, Thus, in the EU, despite the MSFD’s ambiguities, researchers have a and tools for management strategy evaluations. set of objectives by which to operate. The remaining challenge is to resolve the issue of trade-offs and priorities for the multitude of Scoping GES descriptors. Few natural scientists have experience operating in A concern of the ICES IEA groups, and a key factor in an IEA, is a a participatory process as “transparency brokers” and developing scoping exercise to establish higher and lower (operational) objec- indicators that relate to societal interests/priorities. Researchers tives for managing the impact of humans on the ecosystem (Levin may feel awkward moving from their training in biology/ecology et al., 2009). You should “balance diverse societal objectives” into the participatory decision-making realm, but this is what is (FAO, 2003). This is no mean feat as the perceived reality of each required. stakeholder will be based on a different understanding of the func- tioning of the system (Verweij and van Densen, 2010) and, thus, Indicators and targets probably lead to a different notion of the impact of a management Indicators can be viewed as an interface between science and policy action on their activity (Delaney and Hastie, 2007). Policy objectives (Heink and Kowarik, 2010) with some arguing that they have more and research initiatives are in a state of continuous flux; observe the value to policy-makers than to scientists (Levin et al., 2010). shift in terms of research priorities from environmental protection Indicators are needed to provide information on the state of the 1178 M. Dickey-Collas ecosystem, and the progress of management in relation to objectives research community and have been further developed by ICES (Jennings, 2005). Indicators are quantifiable metrics that reflect (2013b). The 8-stage framework can lead to subjective choices changes in ecosystem attributes (Samhouri et al., 2009) and the per- (Rochet and Rice, 2005; Piet et al., 2008). Fulton et al. (2005), formance of an indicator is quantified by the ability to detect and/or Samhouri et al. (2010, 2012), and Kaplan et al. (2012) use ecosystem predict trends in ecosystem attributes (Fulton et al., 2005). We also models to assist in the selection of appropriate indices. There is a require metrics for monitoring overall state, and I would class these danger that when working with indicators you assume that causal- monitoring indicators as conceptually different from IEA indica- ities are linear (Samhouri et al., 2010, 2012). This approach appears tors, which are target-oriented and specific for management action. to ignore the likelihoodof alternative stable states, hysteresis, and in- There is a wealth of studies on defining indicators and targets herent non-linearity in the underlying responses of the ecosystem for both the marine ecosystem-based approach and IEAs. Most (Heath, 2012; Denderen and Kooten, in press). are based on case studies and often use different ecosystem or As part of their method development, it would be constructive trophic models to test the how the indicators are chosen and the for the IEA groups in ICES to further consider indicators and suitability of the indicators themselves (Trenkel and Rochet, 2003; engage with policy developers to show the ramifications of decisions Fulton et al., 2005; Methratta and Link, 2006; Samhouri et al., about targets. Many nations have independently proposed indica- 2009; Kershner et al., 2011). It is paramount to remember the tors through the MSFD, and both OSPAR and HELCOM (the re- purpose of indicators. As Levin et al. (2010) says: “It is tempting gional sea conventions for the NE Atlantic and the Baltic,

for natural scientists to advocate indicators that provide maximal respectively) now have long lists of either proposed or potential Downloaded from information about ecosystem structure and function without con- indicators. So some management objectives and associated indica- sideration of societal values”. Providing maximal information on tors have been provided. The ICES IEA groups are in a strong pos- ecosystem structure and function may negate the indicators’ role ition to explore whether these indicators are of value and show their in melding social and natural science (Levin et al., 2010). relative utility. Informing society that an ecosystem is changing in a dramatic manner without indicating what it means or how to resolve the Risk http://icesjms.oxfordjournals.org/ problem would negate an indicators role. In the ICES context, Risk analysis provides an accountable and transparent framework the policy objectives cannot be forgotten. The optimal portfolio of for prioritizing actions in management, particularly in the indicators is one that ensures appropriate scientific information is broader context of the ecosystem approach to fisheries (ICES, captured while also maximizing the value of the indicators for 2009). Samhouri and Levin (2012) propose that ecosystem-based policy-makers (Levin et al., 2010). As indicators sit at the interface risk should be scored along two axes of information: the exposure between science and policy they should be expected to change of a population to an activity and the sensitivity of the population over time as society’s objectives change (Samhouri et al., 2012). to that activity. Following the same line of argument about There are often very pressing needs which can be quickly addressed scoping and indicators, the subjective value-based nature of objec- (Jennings and Le Quense, 2012) and marine scientists that work in tivesrequires recognition that riskanalysisisa decisionsupport tool, by Howard Browman on July 2, 2014 the applied field should be careful not to deflect focus away from the not a decision-making tool. Risk analysis can also bring in factors main issues. such as cost of monitoring or research into the decision-making Indicators (with associated targets/reference points) serve process. It is often argued that first a qualitative assessment of several purposes; to define the expected objectives, track progress risks be undertaken (a screening out of low-risk activities) followed towards those objectives and communicate trends in complex by a detailed assessment (often quantitative). Some IEA groups impacts to a non-specialist audience (Fulton et al., 2005; Jennings, suggest that risk analysis should be operationalized into a predictive 2005). The portfolio of indicators should contain complementary measure of the effectiveness of management actions, rather than an indicators that exclude redundant metrics, i.e. be resource efficient exploration of trade-offs. (Samhouri et al., 2009; ICES, 2013b; Large et al., 2013). Rice and Accounting for the unpredictability of human behaviour is also a Rochet (2005) stress that the number of indicators should be challenge when investigating fisheries systems (Fulton et al., 2011). limited, as indicators will be incorporated into a range of roles The following uncertainties may need to be assessed when consider- and must be communicable to a wider audience. Some potential ing any fisheries management cycle; resource dynamics, reporting, indicators will reflect the same trends in ecosystem attributes over monitoring, assessment, management decisions, implementation, time; Levin et al. (2010) comment that this might be the case for and fishing activity (Fulton et al., 2011). This approach appears to some societally important indicators (e.g. abundance of different be just as valid when considering any anthropogenic pressure on species of large sea mammals) which also do not provide that the system. When thinking about IEAs specifically, the error and much information on ecosystem functioning. The building of the scale of the chosen indicators should be considered, thus accounting portfolio will involve compromise. for the degree of signal and noise (Jennings, 2005). If managers Much has been said about the selection process for indicators. respond and act based on noise rather than signal, they risk squan- Levin et al. (2010) describe this process as indicator mapping. dering the credibility of everyone involved (Jennings, 2005). Thus, Nonetheless, there is broad agreement about the criteria for asses- an assessment of risk is crucial. sing indicators (ICES, 2013b). The indicators must be sensitive and specific to changes in ecosystem attributes and responsive to Developing tools and evaluations of strategy changes in human activities (Rice, 2003; Jennings, 2005; Rice and ICES has many in-house experts in management strategy evalu- Rochet, 2005). Rice and Rochet (2005) provide an 8-step framework ation, but they tend towards evaluating the effect of single-species for selecting indicators, including scoping. They list the important fisheries on single-species populations (Kraak et al., 2010). There criteria as concreteness, theoretical basis, public awareness, cost, are notable exceptions where scientists have started to consider measurement, availability of historical data, sensitivity, responsive- the other effects of fishing activities on multispecies targets, the ness, and specificity. The criteria appear to be accepted by the effect of mixed fisheries, closed areas, and impact on benthic ICES approach to IEA 1179 communities and on elasmobranchs (Mackinson et al., 2009; Ulrich Jennings and Rice (2011) suggest a two-tier approach to evaluat- et al., 2011; Gascuel et al., 2012; Kraak et al., 2012; Rijnsdorp et al., ing impact and potential management measures, by assessing both 2012, etc.) but studies that integrate across pressures or that investi- single-sector and multisectoral activities. There appears to be pro- gate cumulativeeffects arestill rare (Stelzenmu¨lleret al., 2010; Foden gress in understanding single-sector impacts, with researchers con- et al., 2011). There are various examples of good practice of manage- sidering the effects of one type of activity on many components of ment evaluation coming from IEA work (Sainsbury et al., 2000; the ecosystem. Cumulative effects should consider how social and Smith et al., 2007; Ojaveer and Eero, 2011; Heath, 2012; Kaplan economic aspects of ocean uses covary or compete and how et al., 2012, 2013). impacts of multiple ocean uses contribute to status and trends of Downloaded from http://icesjms.oxfordjournals.org/ by Howard Browman on July 2, 2014

Figure 1. The description of the science and policy landscape in which ICES operates, taken from the ICES strategic plan (2014–2018; ICES, 2014). # ICES. The examples are not exhaustive. Acronyms: Baltic Marine Environment Protection Commission (HELCOM), Convention for the Protection of the Marine Environment of the North-East Atlantic (OSPAR), Convention on Biological Diversity (CBD), Convention on International Trade in Endangered Species of Wild Fauna and Flora (CITES), European Environment Agency (EEA), European Union (EU), Food and Agriculture Organization of the United Nations (FAO), General Fisheries Commission in the Mediterranean (GFCM), International Arctic Science Committee (IASC), International Commission for the Conservation of Atlantic Tunas (ICCAT), International Council for the Exploration of the Sea (ICES), International Council for Science (ICSU), Intergovernmental Panel on Climate Change (IPCC), Intergovernmental Panel on Biodiversity and Ecosystem Services (IPBES), Intergovernmental Oceanographic Commission (IOC), Mediterranean Science Commission (CIESM), Northwest Atlantic Fisheries Organization (NAFO), North Atlantic Salmon Conservation Organization (NASCO), North Atlantic Marine Mammal Commission (NAMMCO), North East Atlantic Fisheries Commission (NEAFC), North Pacific Marine Science Organization (PICES), Scientific Committee on Oceanic Research (SCOR), Statistical Office of the European Communities (EUROSTAT), United Nations Development Program (UNDP), United Nations Environment Program (UNEP). 1180 M. Dickey-Collas ecosystem components or attributes. Cumulative effects will not be accepted a triumvirate EU governance mechanism (Commission, additive but be complex and responsive (ICES, 2012c). A multisec- Council, and Parliament) that led to this reform and the mechanism torial approach may lead to the conclusion that one sector domi- is complex and the outcome unpredictable. Now that the CFP nates the anthropogenic impact (e.g. the impact of fishing on reform has been agreed upon, ICES scientists have to assess its con- benthic communities; Foden et al., 2011), which is a useful sequences for the sustainability of fisheries. We will be adaptive and finding when looking at the next stages of implementing manage- iterative in our approach to the reformed CFP. Should we expect any ment actions, but you need a multisectoral approach to arrive at less of our approach to IEAs? this conclusion. It has also been suggested that cumulative assess- ments across multiple sectors are extremely sensitive to the Conclusion assumed weightings between sectors (Stelzenmu¨ller et al., 2010). There have been significant advances in the development of This area of research needs much more attention and Rice suggests integrated ecosystem approaches in ICES, especially on investigat- that the initial barrier to development is a lackof cross-sectoral data- ing variability in environmental drivers and the incorporation of sets, and institutional blocks to exchange between researchers from fisheries management trade-offs in a dynamic ecosystem individual sectoral fields (J. Rice, pers. com., DFO, Canada). The (e.g. multispecies catch options in Baltic Sea fisheries). IEAs are move to integrated data management policies and in Europe, now a central element of the ICES strategy. The community is closer collaborations with (and between) regional sea commissions dynamic and looking forward to making further progress on the

will hopefully aid this process. challenges of IEAs. There are opportunities and, as shown by Downloaded from The ICES IEA groups were instigated to develop methods for the Walther and Mo¨llmann (2014), incremental progress has been advisory process. Their rolewasto providetools and methods. Thus, made. Ecosystem-based management (EBM) is a US national prior- the groups were not expected to carry out complete IEAs as ity within the National Oceans Policy, and to affect EBM, the IEAs described by FAO (2003), ICES (2005), and Levin et al. (2009), are seen as best practice. In Europe, the MSFD and HELCOM and namely to assess the ecosystem status and evaluate potential man- OSPAR are demanding a different approach to science and new agement strategies. They should, however, be encouraged to pick methods to providing advice on the management of anthropogenic http://icesjms.oxfordjournals.org/ up any developmental requirements after the evaluation of the strat- pressures. In my view, IEA research should be encouraged that con- egy and work on following iterations of the process. It is probably siders the portfolios of indicators, should develop tools for evaluat- now time for the advisory process of ICES to use the tools that ing management measures (that incorporate risk), and should have been developed by carrying out example IEAs. further develop dialogue with stakeholders. When engaging in a debate about societal objectives, we should The science and policy landscape for IEAs in ICES expect to encounter a complex and changing landscape. Thus, our re- Governance and research structures are layered, complex, and often sponse should be adaptive and iterative. The expectation of an contradictory with poor communication between those layers. ordered process is not only likely to lead to disappointment and frus- There is no central funding mechanism, no simple decision-making tration, but is also actually not an optimum approach. We should by Howard Browman on July 2, 2014 body, and society does not speak with one voice. Within ICES, we plan, and build strategies that can be measured, but those need to find a leverage mechanism that can encourage the IEA plans should be responsive as we expect change and unforeseen groups to continue progress. Society also has the right to make deci- consequences of our actions. The field of managing anthropogenic sions based on its evolving political processes. We must note myth pressures on the marine ecosystem is complex with many inter- number 5 of Murawski (2007); that the ecosystem approach is too dependencies, but the challenge is to provide insights from our difficult to apply in multinational regional management organiza- science with the resources available. We will need to innovate and tions. Complexity in governance should not be an excuse to avoid find new ways to respond to the complexity of the management developing approaches. The balance between conservation and util- problem. A straight path might not necessarily be the route that ization strategies will be a matter of societal choice and negotiation, gets you home. consistent with the operating principles of any agreement (Murawski, 2007).Therefore, to succeed,wemustembrace the com- Acknowledgements plexity of the decision-making and implementation framework The article is based on an internal ICES report that I wrote in 2012. It (Degnbol et al., 2006). was widely circulated and I received much input from across the The science and policy landscape in which ICES scientists ICES community. This feedback was greatly appreciated. I would operate is complex (Figure 1, taken from ICES 2014). ICES scientists like to thank Simon Jennings, Jake Rice, Jason Link, Howard areexperienced in describing and exploring complexmarine ecosys- Browman, and Steve Murawski for their help. Poul Degnbol is tems. When considering IEAs, they must prepare to enter the also thanked for his comments on the manuscript. complex system of governance and participatory engagement in the translation of societal objectives into their assessments, or at References least work in teams that include researchers that are prepared to Clark, W. C., Crutzen, P. J., and Schellnhuber, H. J. 2005. Science for engage in dialogue with the governance system. We can learn Global Sustainability: Toward a New Paradigm. John F. Kennedy from “complex systems science” that our frameworks are useful to School of Government Faculty Research Working Paper Series, conceptualise our challenges, but new structures and approaches RWP 05-032, pp. 1–28. emerge from engagement in a complex system (Mitleton-Kelly, Dankel, D. J., Aps, R., Padda, G., Ro¨ckmann, C., van der Sluijs, J. P., Wilson, D. C., and Degnbol, P. 2012. Advice under uncertainty in 2003). Adaptability of the approach taken is key. It could be the marine system. ICES Journal of Marine Science, 69: 3–7. argued that ICES scientists had little direct input into the two Degnbol, P., Gislason, H., Hanna, S., Jentoft, S., Nielsen, J. R., main outcomes of the 2014 EU Common Fisheries Policy (CFP) Sverdrup-Jensen, S., and Wilson, D. C. 2006. Painting the floor reform, namely the landing obligation (discard ban) and the with a hammer: technical fixes in fisheries management. Marine firming up of the political meaning of the MSY. EU society has Policy, 30: 534–543. ICES approach to IEA 1181

Degnbol, P., and McCay, B. J. 2006. Unintended and perverse conse- ICES. 2013b. Report of the Working Group on the Ecosystem Effects of quences of ignoring linkages in fisheries systems. ICES Journal of Fishing Activities (WGECO). ICES Document CM 2013/ACOM: Marine Science, 64: 793–797. 25. 117 pp. Degnbol, P.,and Wilson, D.C. 2008. Spatialplanning on the NorthSea:a ICES. 2014. ICES strategic plan 2014–2018. ICES Copenhagen. 19 pp. case of cross-scale linkages. Marine Policy, 32: 189–200. Jennings, S. 2004. The ecosystem approach to fishery management: a Delaney, A. E., and Hastie, J. E. 2007. Lost in translation: differences in significant step towards sustainable use of the marine environment? role identities between fisheries scientists and managers. Ocean and Marine Ecology Progress Series, 274: 279–282. Coastal Management, 50: 661–682. Jennings, S. 2005. Indicators to support an ecosystem approach to fish- van Denderen, P. D., and van Kooten, T. in press. Size based species eries. Fish and Fisheries, 6: 212–232. interactions shape herring and cod population dynamics in the Jennings, S., and Le Quesne, W. J. F. 2012. Integration of environmental face of exploitation. Ecosphere, doi.org/10.1890/ES13–00164.1. and fishery management in Europe. ICES Journal of Marine Science, De Young, C., Charles, A., and Hjort, A. 2008. Human Dimensions of 69: 1329–1332. the Ecosystem Approach to Fisheries: an Overview of Context, Jennings, S., and Rice, J. 2011. Towards an ecosystem approach to fish- Concepts, Tools and Methods. Fisheries Tech. Paper, 489. FAO, eries in Europe: a perspective on existing progress and future direc- Rome. 165 pp. tions. Fish and Fisheries, 12: 125–137. EC. 2008. Directive 2008/56/EC of the European Parliament and of the Kaplan, I. C., Gray, I. A., and Levin, P. S. 2013. Cumulative impacts Council. Establishing a framework for community action in the field of fisheries in the California Current. Fish and Fisheries. of marine environmental policy (Marine Strategy Framework doi:10.1111/j.1467–2979.2012.00484.x. Directive). EC, Brussels. Kaplan, I. C., Horne, P. J., and Levin, P. S. 2012. Screening California Downloaded from FAO. 2003. Fisheries management. 2. The ecosystem approach to Current fishery management scenarios using the Atlantis fisheries. FAO, Rome. 121pp. end-to-end ecosystem model. Progress in Oceanography, 102: 5–18. Foden, J., Rogers, S. I., and Jones, A. P. 2011. Human pressures on UK Kershner, J., Samhouri, J. F., James, C. A., and Levin, P. S. 2011. Selecting seabed habitats: a cumulative impact assessment. Marine Ecology indicator portfolios for marine species and food webs: a Puget sound Progress Series, 428: 33–47. case study. PLoS one, 6: e25248. Francis, T. B., Levin, P. S., and Harvey, C. J. 2011. The perils and promise Kraak, S. B. M., Kelly, C. J., Codling, E. A., and Rogan, E. 2010. On scien- http://icesjms.oxfordjournals.org/ of futures analysis in marine ecosystem-based management. Marine tists’ discomfort in fisheries advisory science: the example of Policy, 35: 675–681. simulation-based fisheries management-strategy evaluations. Fish Fulton, E. A., Smith, A. D. M., and Punt, A. E. 2005. Which ecological and Fisheries, 11: 119–132. indicators can robustly detect effects of fishing? ICES Journal of Kraak, S. B. M., Reid, D. G., Gerritsen, H. D., Kelly, C. J., Fitzpatrick, M., Marine Science, 62: 540–551. Codling, E. A., and Rogan, E. 2012. 21st century fisheries manage- Fulton, E. A., Smith, A. D. M., Smith, D. C., and van Putten, I. E. 2011. ment: a spatio-temporally explicit tariff-based approach combining Human behaviour: the key source of uncertainty in fisheries man- multiple drivers and incentivising responsible fishing. ICES Journal agement. Fish and Fisheries, 12: 2–17. of Marine Science, 69: 590–601. Garcia, S. M., and Cochrane, K. L. 2005. Ecosystem approach to fisher- Large, S. I., Fay, G., Friedland, K. D., and Link, J. S. 2013. Defining trends ies: a review of implementation guidelines. ICES Journal of Marine and thresholds in responses of ecological indicators to fishing and Science, 62: 311–318. environmental pressures. ICES Journal of Marine Science, 70: by Howard Browman on July 2, 2014 Gascuel, D., Merino, G., Do¨ring, R., Druon, J. N., Goti, L., Gue´nette, S., 755–767. Macher, C., et al. 2012. Towards the implementation of an integrated van Leeuwen, J., van Hoof, L., and van Tatenhove, J. 2012. Institutional ecosystem fleet-based management of European fisheries. Marine ambiguity in implementing the European Union Marine Strategy Policy, 36: 1022–1032. Framework Directive. Marine Policy, 36: 636–643. Haas, P. M. 2004. When does power listen to truth? A constructivist ap- Levin, P. S., Damon, M., and Samhouri, J. F. 2010. Developing meaning- proach to the policy process. Journal of European Public Policy, 11: ful marine ecosystem indicators in the face of a changing climate. 569–592. Stanford Journal of Law, Science and Policy, 2: 36–48. Heath, M. R. 2012. Ecosystem limits to food web fluxes and fisheries Levin, P. S., Fogerty, M. J., Murawski, S. A., and Fluharty, D. 2009. yields in the North Sea simulated with an end-to-end food web Integrated ecosystem assessments: developing the scientific basis model. Progress in Oceanography, 102: 42–66. for ecosystem-based management of the ocean. PLoS Biology, 7: Heink, U., and Kowarik, I. 2010. What are indicators? On the definition e1000014. of indicators in ecology and environmental planning. Ecological Levin, P. S., Kelble, C. R., Shuford, R., Ainsworth, C., de Reynier, Y., Indicators, 10: 584–593. Dunsmore, R., Fogarty, M. J., et al. 2014. Guidance for implementa- ICES. 2005. Guidance on the application of the ecosystem approach to tion of integrated ecosystem assessments: a US perspective. ICES management of human activities in the European marine environ- Journal of Marine Science, 71: 1198–1204. ment. ICES Co-operative Research Report 273, 22 pp. Link, J. S. 2012. Overview of Ecosystem-based Management and ICES. 2009. Report of the Study Group on Risk Assessment and Integrated Ecosystem Assessments. CalCOFI Reports, 53: 91–92. Management Advice (SGRAMA). ICES Document CM 2009/ Mackinson, S., Deas, B., Beveridge, D., and Casey, J. 2009. Mixed-fishery RMC: 14. 36 pp. or ecosystem conundrum? Multispecies considerations inform ICES. 2012a. Report of the Workshop on Benchmarking Integrated thinking on long-term management of North Sea demersal stocks. Ecosystem Assessments (WKBEMIA). ICES Document CM 2012/ Canadian Journal of Fisheries and Aquatic Sciences, 66: 1107–1129. SSGRSP: 08. 30 pp. Mackinson, S., Wilson, D. C., Galiay, P.,and Deas, B. 2011. Engaging sta- ICES. 2012b. Report of the Workshop on North Sea and Baltic Sea keholders in fisheries and marine research. Marine Policy, 35: 18–24. Multispecies Trade-offs (WKM-Trade). ICES Document CM Methratta, E. T., and Link, J. S. 2006. Evaluation of quantitative indica- 2012/ACOM: 71. 15 pp. tors for marine fish communities. Ecological Indicators, 6: 575–588. ICES. 2012c. Report of the Working Group on the Ecosystem Effects of Mitleton-Kelly, E. 2003. Complex Systems and Evolutionary Perspectives Fishing Activities (WGECO). ICES Document CM 2012/ACOM: on Organisations: the Application of Complexity Theory to 26. 192pp. Organisations. Advanced Series in Management. Elsevier Science ICES. 2013a. Report of the Working Group on the ICES ACOM/ Ltd, Oxford, UK. 257 pp. SCICOM Workshop on Ecosystem Overviews (WKECOVER). Mo¨llmann, C., Lindegren, M., Blenckner, T., Bergstro¨m, L., Casini, M., ICES Document CM ACOM/SCICOM: 01. 135 pp. Diekmann, R., Flinkman, J., et al. 2014. Implementing ecosystem- 1182 M. Dickey-Collas

based fisheries management: from single-species to integrated eco- Samhouri, J. F., Lester, S. E., Selig, E. R., Halpern, B. S., Fogarty, M. J., system assessment and advice for Baltic Sea fish stocks. ICES Longo, C., and McLeod, K. L. 2012. Sea sick? Setting targets to Journal of Marine Science, 71: 1187–1197. assess ocean health and ecosystem services. Ecosphere, 3: 41. Murawski, S. A. 2007. Ten myths concerning ecosystem approaches to Samhouri, J. F., and Levin, P. S. 2012. Linking land- and sea-based activ- marine resource management. Marine Policy, 31: 681–690. ities to risk in coastal ecosystems. Biological Conservation, 145: Ojaveer, H., and Eero, M. 2011. Methodological challenges in assessing 118–129. the environmental status of a marine ecosystem: case study of the Samhouri, J. F., Levin, P. S., and Ainsworth, C. H. 2010. Identifying Baltic Sea. PLoS One, 6: e19231. thresholds for ecosystem-based management. PLoS One, 5: e8907. Ounanian, K., Delaney, A., Raakjaer, J., and Ramirez-Monsalve, P. 2012. Samhouri, J. F., Levin, P. S., and Harvey, C. J. 2009. Quantitative evalu- On unequal footing: stakeholder perspectives on the marine strategy ation of marine ecosystem indicator performance using food web framework directive as a mechanism of the ecosystem-based ap- models. Ecosystems, 12: 1283–1298. proach to marine management. Marine Policy, 36: 658–666. Piet, G. J., Jansen, H. M., and Rochet, M-J. 2008. Evaluating potential Schellnhuber, H. J., and Sahagian, D. 2002. The twenty-three GAIM indicators for an ecosystem approach to fishery management in questions. Global Change Newsletter, 49: 20–21. European waters. ICES Journal of Marine Science, 65: 1449–1455. Schwach, V., Bailly, D., Christensen, A-S., Delaney, A. E., Degnbol, P., Ratza, H. J., Dorner, H., Scott, R., and Barbas, T. 2010. Complementary van Densen, W. L. T., Holm, P., et al. 2007. Policy and knowledge roles of European and national institutions under the Common in fisheries management: a policy brief. ICES Journal of Marine Fisheries Policy and the Marine Strategy Framework Directive. Science, 64: 798–803.

Marine Policy, 34: 1028–1035. Sissenwine, M., and Murawski, S. A. 2004. Moving beyond ‘intelligent Downloaded from Rice, J. 2003. Environmental health indicators. Ocean and Coastal tinkering’: advancing an Ecosystem Approach to Fisheries. Marine Management, 46: 235–259. Ecology Progress Series, 274: 291–295. Rice, J. 2011. Managing fisheries well: delivering the promises of an eco- Smith, A. D. M., Fulton, E. J., Hobday, A. J., Smith, D. C., and Shoulder, system approach. Fish and Fisheries, 12: 209–231. P. 2007. Scientific tools to support the practical implementation of Rice, J. C., and Rochet, M.-J. 2005. A framework for selecting a suite of ecosystem-based fisheries management. ICES Journal of Marine

indicators for fisheries management. ICES Journal of Marine Science, 64: 633–639. http://icesjms.oxfordjournals.org/ Science, 62: 516–527. Stelzenmu¨ller, V., Lee, J., South, A., and Rogers, S. I. 2010. Quantifying Rijnsdorp, A. D., van Overzee, H. M. J., and Poos, J. J. 2012. Ecological cumulative impacts of human pressures on the marine environment: and economic trade-offs in the management of mixed fisheries: a a geospatial modelling framework. Marine Ecology Progress Series, case study of spawning closures in flatfish fisheries. Marine 398: 19–32. Ecology Progress Series, 447: 179–194. Trenkel, V. M., and Rochet, M-J. 2003. Performance of indicators Rochet, M-J., and Rice, J. C. 2005. Do explicit criteria help in selecting derived from abundance estimates for detecting the impact of indicators for ecosystem-based fisheries management? ICES fishing on a fish community. Canadian Journal of Fisheries and Journal of Marine Science, 62: 528–539. Aquatic Sciences, 60: 67–85. Ro¨ckmann, C., Ulrich, C., Dreyer, M., Bell, E., Borodzicz, E., Ulrich, C., Reeves, S. A., Vermard, Y., Holmes, S. J., and Vanhee, W. Haapasaari, P., Hauge, K. H., et al. 2012. The added value of partici-

2011. Reconciling single-species TACs in the North Sea demersal by Howard Browman on July 2, 2014 patory modelling in fisheries management—what has been learnt? fisheries using the Fcube mixed-fisheries advice framework. ICES Marine Policy, 36: 1072–1085. Journal of Marine Science, 68: 1535–1547. Sainsbury, K. J., Punt, A. E., and Smith, A. D. M. 2000. Design of oper- ational management strategies for achieving fishery ecosystem Verweij, M. C., and van Densen, W. L. T. 2010. Differences in causal rea- objectives. ICES Journal of Marine Science, 57: 731–741. soning about resource dynamics and consequences for the participa- tory debate on North Sea fisheries. Marine Policy, 34: 1144–1155. Samhouri, J.F., Haupt, A. J.,Levin, P.S., Link, J.S., and Shuford, R. 2014. Lessons learned from developing integrated ecosystem assessments Walther, Y., and Mo¨llmann, C. 2014. Bringing integrated ecosystem to inform marine ecosystem-based management in the USA. ICES assessments to real life: a scientific framework for ICES. ICES Journal of Marine Science, 71: 1205–1215. Journal of Marine Science, 71: 1183–1186.

Handling editor: Jason Link ICES Journal of Marine Science

ICES Journal of Marine Science (2014), 71(5), 1183–1186. doi:10.1093/icesjms/fst161

Contribution to the Themed Section: ‘Integrated assessments’ Original Article Bringing integrated ecosystem assessments to real life:

a scientific framework for ICES Downloaded from

Yvonne M. Walther1* and Christian Mo¨llmann2* 1Department of Aquatic Resources, Swedish University of Agricultural Sciences, Uto¨va¨gen 5, SE-37137 Karlskrona, Sweden 2Institute for Hydrobiology and Fisheries Science, Center for Earth System Research and Sustainability (CEN), University of Hamburg, KlimaCampus, Grosse Elbstrasse 133, D-22767 Hamburg, Germany http://icesjms.oxfordjournals.org/ *Corresponding author: tel: +46 709 35 92 82; fax: +46 10 478 40 75; e-mail: [email protected] Walther, Y., and Mo¨llmann, C. 2014. Bringing integrated ecosystem assessments to real life: a scientific framework for ICES. – ICES Journal of Marine Science, 71: 1183–1186. Received 10 March 2013; accepted 4 September 2013; advance access publication 13 November 2013.

An ecosystem approach to management (EAM) aims to secure a healthy ecosystem along with sustainable use of its goods and services. Although the main principles of EAM are agreed upon and desirable, wider implementation of EAM is still a challenge. The difficulties stem from unclear definition and communication of the EAM, lack of routines or protocols to develop ecosystem-based advice, inappropriate institutional struc- by Howard Browman on July 2, 2014 tures, and communication issues between scientists, advisers, and managers. Integrated ecosystem assessment (IEA) is an instrument that has proven to help in the implementation of EAM. For successful implementation of EAM and IEA in the European regional seas context, an inter- national forum is required that develops tailor-made EAM tools specific to the different regional ecosystems to overcome fragmented national strategies. We describe a multinational peer network of working groups developed within the International Council for the Exploration of the Sea (ICES) under the auspice of Science Steering Group on Regional Sea Programmes. Available is a wealth of data, expertise, scientific methods, and models for each regional sea. This network can be instrumental in advancing IEA for the implementation of EAM in the North Atlantic seas. Keywords: ecosystem approach to management (EAM), ecosystem-based management (EBM), integrated ecosystem assessment (IEA), International Council for Exploration of the Sea, The Science Steering Group on Regional Sea Programmes, Marine Strategy Framework Directive.

Introduction do EAM, the wider implementation of an EAM remains a challenge The idea of an ecosystem approach to management (EAM) is to (Murawski, 2007; Rice, 2011). manage natural resources in a holistic way, by considering the inter- Reasons for the difficulty in implementing EAM are manifold, acting influences of multiple use sectors on the environment among them unclear definition and communication of EAM to (McLeod and Leslie, 2009). The overall goal is to assure the health important user groups (e.g. fishers), resistance of advice and man- of the ecosystem alongside appropriate use of the environment for agement bodies to develop or revise their routines and protocols the benefit of future generations (Jennings, 2004). EAM has been towards ecosystem-based approaches, inappropriate institutional a core debate of natural science and resource management for structures (i.e. sector-specific independent agencies), and commu- several decades (Grumbine, 1994) and the main principles of nication issues between scientists, advisers, and managers. We EAM are generally agreed upon and desired by most policy-makers, believe that overcoming the problems of implementing EAM managers, and scientists. Moreover, the intention to move towards needs a forum that channels and translates relevant scientific the adoption of a comprehensive and integrated EAM is no longer results into practically applicable EAM tools; a forum that develops just an obscure future vision (Levin et al., 2009; Lubchenco et al., tailor-made EAM tools specific to different regional seas’ challenges 2010; Tallis et al., 2010). But despite the political and societal will, and needs, thereby overcoming national defragmentation as seen in as well as the availability of scientific concepts and information to the European Union; and last but not least, a forum that furthers the

# 2013 International Council for the Exploration of the Sea. Published by Oxford University Press. All rights reserved. For Permissions, please email: [email protected] 1184 Y. M. Walther and C. Mo¨llmann communication between science and advisory bodies, and eventu- the Baltic Sea (WGIAB; ICES, 2007), an initiative taken together ally to managers. with the Helsinki Commission (HELCOM). Initially, the ecosystem Here we describe the bottom-up development of, the present assessments mainly assessed the impact of climate, fisheries, work on, and the potential future for implementing EAM via a and eutrophication to combine the major drivers that previously multinational peer network of working groups within the were evaluated in isolation (Mo¨llmann et al., 2013). These International Council for the Exploration of the Sea (ICES), analyses demonstrated pronounced climate, fisheries, and under the auspice of the SCICOM Steering Group on the eutrophication-related structural changes for several Baltic subsys- Regional Sea Programmes (SSGRSP). We suggest that this tems (Mo¨llmann et al., 2009; Diekmann and Mo¨llmann, 2010; network will be instrumental in developing and advancing inte- Lindegren et al., 2010). The WGIAB serves as a counterpart for grated ecosystem assessments (IEA) towards further implementa- the ICES Baltic Fisheries Assessment Working Group (WGBFAS), tion of EAM in the North Atlantic seas. and has over several years incrementally provided ecosystem knowl- edge for more ecosystem-based fish stock assessments and advice IEA: a concept for EAM implementation (Ga˚rdmark et al., 2011;Mo¨llmann et al., 2013). The WGIAB now Often, EAM is seen as a complex, scientifically driven process that is develops tools to be applicable in future IEAs for the Baltic Sea hampered by lack of knowledge about the ecosystem. But the percep- (ICES, 2013a). The WGNARS was the first group to introduce tion that knowledge is insufficient to apply EAM has been effectively and explore the full IEA concept developed by NOAA (US

contradicted (Murawski, 2007). Rather, it is more important to maxi- National Oceanic and Atmospheric Administration) (Levin et al., Downloaded from mize the use of available information than to emphasize areas of un- 2009).Earlyonthe group focused onthe importance of humanpres- certainty (Rice, 2011). Despite the scientific fac¸ade, EAM is sures and the value of indicator-based approaches. The WGNARS foremost a decision process (Walters and Martell, 2004)andthe has now continued towards closing the IEA cycle by showing exam- desired status of an ecosystem is not decided in a scientific process. ples of how to include human pressures and socio-economic aspects Yet science needs to assist that process with knowledge that helps the (ICES, 2010, 2011, 2012a) and has developed into a focal point for decision-making. Looking more closely, successful cases of EAM are exploring and testing facets of IEAs in the USA and Canada. http://icesjms.oxfordjournals.org/ successful not because of the designation of a scientific programme, The IEA concept has developed within ICES from a few solitary but rather because of amultidisciplinary, multi-actorprocess integrat- expert groups to a wide network under the auspice of SSGRSP. This ing research and management, i.e. a framework connecting scientists, has come true by a “sticky” process where existing groups help to managers, and stakeholders (deReynier et al., 2010). IEA is a proposed attract scientists from different areas of expertise, geographical but framework that has the ability to synthesize and analyse scientific in- also topical, to create new regional IEA groups. Through this formation in relation to specified ecosystem management objectives sticky process, knowledge and methodology developed and and is especially suitable for engaging different stakeholder and deci- applied has been transferred between regions. Today, SSGRSP is sion maker groups (Levin et al.,2009; Tallis et al.,2010; Levin et al., composed of a network of IEA groups including the Baltic Sea 2013). The main purpose of IEAs is to integrate physical, biological, (WGIAB), the North Sea (WGINOSE), the Western European by Howard Browman on July 2, 2014 and socio-economic data to evaluate trade-offs among objectives Shelf Seas (WGEAWESS), the Norwegian Sea (WGINOR), and across different ecosystem goods and services. the North West Atlantic Regional Sea (WGNARS) in the USA– Canada. The current and planned groups with their acronyms are ICES and the development of a framework for IEA shown in the schematic structure which is affectionately termed The ICES, as an intergovernmental organization with 20 member the “Fishbone” of the SSGRSP (Figure 1). The structure also states, has the task to organize science and provide advice for the man- includes groups that work on topical issues that interact with and agement of the marine environment and associated living resources in can add value to IEAs, e.g. Ecosystems economics (SGIMM), Best the North Atlantic (Rozwadowski, 2002).Theadviceprovidediswell Practices for Large Marine Ecosystems (WGLMEBP), comparing known and relates to the, for example, the EU Common Fisheries ecosystems (WGCOMEDA), and linking contaminants to IEA Policy (CFP) and EU Marine Strategy Framework Directive (WKLINCON). Eventually, it is envisaged that there will be an (MSFD). The ICES anticipates and is prepared for future demands ecosystem group for all regional seas that are covered by ICES. of ecosystem considerations into this advice. The new “ICES The ICES IEA process is also complemented by ecosystem over- Strategy” under construction (including the Strategic, Advisory, and views that are being conducted or planned (WKECOVER; ICES, Science Plans) prepares for the anticipated request for specific consid- 2013b). These overviews effectively are reports on ecosystem erations of ecosystem drivers and ecosystem impacts on different status that can be largely considered a product of the IEA process. ocean-use sectors, as well as interactions among them. IEA will be a Similar and related work is espoused in the ICES context (e.g. mod- strong component in the new “ICES Strategy” and development of elling, risk assessments, etc.). The salient point being that the IEA IEAs for the European regional seas covered by ICES has been con- process can serve as a useful mechanism to collate and communicate ducted under the auspices of Science Steering Group on Regional a wide range of disparate information into packages that are useful Sea Programmes (SSGRSP; Walther, 2011). for particular management issues being asked. Ecosystem assessments within ICES started in 2006, conducted by the Regional Ecosystem Study Group for the North Sea The future of IEA within ICES (REGNS; ICES, 2006) based on methodology developed in the Despite the early successes of the regional groups, a common frame- USA and Canada (Link, 2002; Choi et al., 2005; Mo¨llmann and work for IEA, especially in the European regional seas, is still lacking. Diekmann, 2012). The REGNS conducted the first assessment of To ensure further progress, SSGRSP initiated a series of Workshops the state and development of abiotic and biotic variables as well as on Benchmarking Integrated Ecosystem Assessments (WKBEMIA). natural and human pressures in the North Sea. The template of In the first workshop (ICES, 2012b), progress of each regional group REGNS was then taken up by Baltic Sea scientists leading to the was evaluated to identify common approaches, results, and chal- formation of the Working Group on Integrated Assessments of lenges. The workshop adopted the NOAA IEA approach (Levin Bringing integrated ecosystem assessments to real life 1185 Downloaded from http://icesjms.oxfordjournals.org/

Figure 1. Expert Groups under the auspice of ICES Science Steering Group on Regional Sea Programmes. Expert groups on side bones are IEA groups; expert groups on the main bone have a more overarching and/or topical function. WGIAB, Working Group on Integrated assessments in the Baltic Sea; WGINOSE, Working Group on Integrated Assessments of the North Sea; WGINOR, Working Group on the Integrated Assessments of the Norwegian Sea; WGEAWESS, Working Group on Ecosystem Assessment of Western European Shelf Seas; WGNARS, Working Group on the Northwest Atlantic Regional Sea; WGLMEBP, Working Group on Large Marine Ecosystems Best Practices; WKLINCON, Workshop on Linking Contaminant Issues to Integrated Ecosystem Assessments; WKBEMIA, Workshop on Benchmarking Integrated Ecosystem Assessments; SGSPATIAL, Study Group on Spatial Analyses for the Baltic Sea; SGIMM, Study Group on Introducing Coupled Ecological–Economic Modelling and Risk Assessment into Management Tools; WKCOMEDA, Working Group on Comparative Analyses between European Atlantic and

Mediterranean marine ecosystems to move towards an Ecosystem-based Approach to Fisheries. by Howard Browman on July 2, 2014 et al., 2009) as a framework for the SSGRSP regional groups, but also and meet the considerations in directives, e.g. the EU’s MSFD considered approaches developed and discussed by the UN Food need for Good Environmental Status indicators (EU-COM, 2008) and Agriculture Organisation (FAO) and the ICES Working and HELCOM Baltic Sea Action Plan (HELCOM, 2007). The Group on Ecosystem Effects of Fishing (ICES, 2013c). The region- approaches can be implemented in relation to applied directives specific application of the different methods and model require- on ecosystem level or even on single sectors, pressures, and compo- ments for each component of the Levin et al. (2009) cycle were nents when necessary, though using the common IEA framework. identified, and now need to be further developed by the regional A future challenge will be a stronger consideration of cumulative groups. An important step will be to conduct a rigorous scoping effects of anthropogenic pressures. Further workshops on IEA will exercise that a priori develops the extent of the EAM to be developed, therefore focus on engaging a wider science community, i.e. includ- i.e. which sectors, pressures, and ecosystem components need to be ing other sectors (e.g. traffic, eutrophication) and socio-economics. included in an IEA for a specific region. Then other components of The first WKBEMIA workshop emphasized that IEAs will be the IEA cycle, i.e. indicator development, risk assessments, and important for improving the single sector, issue-based (e.g. fisheries) management strategy evaluation, can be developed (Levin et al., advice for ICES by providing ecosystem context. Results from an IEA 2009). Plenty of data, expertise, scientific methods, and especially process should be linked to traditional stock assessments and incre- models are available for every regional sea. mentally ecosystem and potentially socio-economic knowledge Future benchmarking workshops will continue to develop IEA would be included in existing advice and management schemes methods specific to the needs of the different regional seas. leading to an Ecosystem Approach to Fisheries Management (EAF; Methods will include the OSPAR Quality Status Report (QSR) Rice, 2011). First steps have already been conducted for Baltic fish approach (Knights et al., 2011), the REGNS approach (ICES, stock advice (Mo¨llmann, 2013). 2006), Bayesian Belief Networks (BBN) (Stelzenmu¨ller et al., Implementing the EAM and IEAs still requires some rethinking 2011), and the Ocean health index (OHI; Halpern et al., 2012). and restructuring of institutional and governance structures. For These methods have often been seen as competing with IEAs, but ICES, this means a closer cooperation between science and advice. WKBEMIA noted that instead they are in fact quite complementary. Operationally, this would be performed through a closer linkage Each has strengthsthatcan be used in aspects of IEAquantitative nu- of the regional IEA groups with stock assessment and advice drafting merical processes, i.e. REGNS and BBN can be used when data are groups. In a longer perspective, stock assessment groups could or available, and expert judgement-based methods (OSPAR QSR and should be complemented by or even developed into ecosystem OHI) when data are sparse. The methods used should investigate assessment groups (ICES, 2012b). 1186 Y. M. Walther and C. Mo¨llmann

Conclusions ICES. 2013b. Report of the Working Group on the ICES ACOM/ SCICOM Workshop on Ecosystem Overviews. ACOM/SCICOM We think that for successfully implementing the EAM and IEAs in 01. 130 pp. the European regional seas, an international forum is required ICES. 2013c. Report of the Working Group on the Ecosystem Effects of thatdevelops tailor-made EAM tools specific to the different region- Fishing Activities (WGECO). ICES Document CM 2013/ACOM:25. al ecosystems to overcome fragmented national strategies. We 115 pp. believe that SSGRSP and its regional IEA groups provides such a Jennings, S. 2004. The ecosystem approach to fishery management: a forum for that, particularly with the potential to translate science significant step towards sustainable use of the marine environment? into management application and further the communication Marine Ecology Progress Series, 274: 279–282. between science advisory bodies, and managers. In a few decades, Knights, A., Koss, R., Papadopoulou, N., Cooper, L., and Robinson, L. from now we may consider EAM and IEAs as common as we 2011. Sustainable use of European regional seas and the role of the think the Internet is in our society today. Yet, as noted in the Marine Strategy Framework Directive. Deliverable 1, EC FP7 Project (244273) “Options for Delivering Ecosystem-based Marine context of developing the Internet, “The point is, that when we Management.” succeed, we succeed because of our individual initiative, but also Levin, P. S., Kelble, C. R., Shuford, R. L., Ainsworth, C., deReynier, Y., because we do things together” (The International Herald Dunsmore, R., Fogarty, M. J., et al. 2014. Guidance for implementa- Tribune, 22–23 September 2012, p. 14 continued on p. 16). tion of integrated ecosystem assessments: A US perspective. ICES Journal of Marine Science, 71: 1198–1204.

Levin, P. S., Fogarty, M. J., Murawski, S. A., and Fluharty, D. 2009. Downloaded from Integrated ecosystem assessments: developing the scientific basis References for ecosystem-based management of the ocean. PLoS Biology, Choi, J. S., Frank, K. T., Petrie, B. D., and Leggett, W. C. 2005. Integrated 7: e1000014. assessment of large marine ecosystem: a case study of the devolution Lindegren, M., Mo¨llmann, C., Nielsen, A., Brander, K., MacKenzie, of the Eastern Scotian Shelf, Canada. Oceanogr. Mar. Biol. Ann. B. R., and Stenseth, N. C. 2010. Ecological forecasting under Rev. 43: 47–67. climate change: the case of Baltic cod. Proceedings of the Royal http://icesjms.oxfordjournals.org/ deReynier, Y. L., Levin, P. S., and Shoji, N. L. 2010. Bringing stake- Society B: Biological Sciences, 277: 2121–2130. holders, scientists, and managers together through an integrated Link, J. S. 2002. What does ecosystem-based fisheries management ecosystem assessment process. Marine Policy, 34: 534–540. mean. Fisheries (Bethesda), 27: 18–21. Diekmann, R., and Mo¨llmann, C. 2010. Integrated ecosystem assess- Lubchenco, J., Sutley, N., and Approach, A. E. 2010. Proposed U.S. ments of seven Baltic Sea areas covering the last three decades. Policy for Ocean, Coast, and Great Lakes Stewardship, 328: ICES Cooperative Research Report 302. 1485–1486. EU-COM. 2008. Directive 2008/56/ECof the European Parliament and McLeod, K. L., and Leslie, H. M. 2009. Ecosystem-based management of the Council of 17 June 2008 establishing a framework for commu- for the oceans. Island Press, Washington, DC. nity action in the field of marine environmental policy (Marine Mo¨llmann, C., Lindegren, M., Blenckner, T., Bergstro¨m, L., Casini, M., Strategy Framework Directive: 19–40. Diekmann, R., Flinkman, J., et al. 2014. Implementing ecosystem- Ga˚rdmark, A., Nielsen, A., Floeter, J., and Mo¨llmann, C. 2011. Depleted based fisheries management: from single-species to integrated eco- by Howard Browman on July 2, 2014 marine fish stocks and ecosystem-based management: on the road to system assessment and advice for Baltic Sea fish stocks. ICES recovery, we need to be precautionary. ICES Journal of Marine Journal of Marine Science, 71: 1187–1197. Science, 68: 212–220. Mo¨llmann, C., and Diekmann, R. 2012. Marine ecosystem regime shifts Grumbine, R. E. 1994. What is ecosystem management? Conservation induced by climate and overfishing: a review for the Northern hemi- Biology, 8: 27–38. sphere. In Advances in Ecological Research, vol. 47, pp. 303–347. Halpern, B. S., Longo, C., Hardy, D., McLeod, K. L., Samhouri, J. F., Ed. by U. Jacob, and E. O’Gorman. Academic Press, London. 510 pp. Katona, S. K., Kleisner, K., et al. 2012. An index to assess the health Mo¨llmann, C., Diekmann, R., Mu¨ller-Karulis, B., Kornilovs, G., Plikshs, and benefits of the global ocean. Nature, 488: 615–620. M., and Axe, P.2009. Reorganization of a large marine ecosystem due HELCOM. 2007. Baltic Sea Action Plan, HELCOM min. 101 pp. to atmospheric and anthropogenic pressure: a discontinuous regime ICES. 2006. Report of the Regional Ecosystem Study Group of the North shift in the Central Baltic Sea. Global Change Biology, 15: Sea (REGNS). ICES CM, RMC:06. 111 pp. 1377–1393. ICES. 2007. Report of the ICES/HELCOM Working Group on Murawski, S. 2007. Ten myths concerning ecosystem approaches to Integrated Assessments of the Baltic Sea (WGIAB). ICES CM, marine resource management. Marine Policy, 31: 681–690. BCC:04. 53 pp. Rice, J. 2011. Managing fisheries well: delivering the promises of an ICES. 2010. ICES WGNARS REPORT 2010 Report of the Working Group ecosystem approach. Fish and Fisheries, 12: 209–231. on the Northwest Atlantic Regional Sea (WGNARS) International Rozwadowski, H. M. 2002. The Sea KnowsNoBoundaries. Universityof Council for the Exploration of the Sea. ICES CM, SSGRSP:03. 67 pp. Washington Press, Seattle and London. ICES. 2011. ICES WGNARS REPORT 2011 Report of the Working Stelzenmu¨ller, V., Schulze, T., Fock, H. O., and Berkenhagen, J. 2011. Group on the Northwest Atlantic Regional Sea (WGNARS) Integrated modelling tools to support risk-based decision-making Halifax, Canada International Council for the Exploration of the in marine spatial management. Marine Ecology Progress Series, Sea. ICES CM, SSGRSP:01. 69 pp. 441: 197–212. ICES. 2012a. ICES WGNARS REPORT 2012 Report of the Working Tallis, H., Levin, P. S., Ruckelshaus, M., Lester, S. E., McLeod, K. L., Group on the Northwest Atlantic Regional Sea (WGNARS) Fluharty, D. L., and Halpern, B. S. 2010. The many faces of International Council for the Exploration of the Sea. ICES CM, ecosystem-based management: Making the process work today in SSGRSP:01. 83 pp. real places. Marine Policy, 34: 340–348. ICES. 2012b. Report of the Workshop on Benchmarking Integrated Walters, C. J., and Martell, S. J. D. 2004. Fisheries Ecology and Ecosystem Assessments (WKBEMIA). ICES CM, in press. Management. Princeton University Press, Princeton, NJ. ICES. 2013a. Report of The Working Group on Integrated Assessments Walther, Y. 2011. SSGRSP—the square peg in a round-peg society. ICES of the Baltic Sea (WGIAB). ICES CM, SSGRSP:05. 53 pp. Insight, 48: 28–32.

Handling editor: Jason Link ICES Journal of Marine Science

ICES Journal of Marine Science (2014), 71(5), 1187–1197. doi:10.1093/icesjms/fst123

Contribution to the Themed Section: ‘Integrated assessments’ Original Article Implementing ecosystem-based fisheries management: from

single-species to integrated ecosystem assessment and advice Downloaded from for Baltic Sea fish stocks

Christian Mo¨llmann1*, Martin Lindegren2, Thorsten Blenckner3, Lena Bergstro¨m4, Michele Casini5, 1 6 7 8 9 Rabea Diekmann , Juha Flinkman ,Ba¨rbel Mu¨ller-Karulis , Stefan Neuenfeldt ,Jo¨rn O. Schmidt , http://icesjms.oxfordjournals.org/ Maciej Tomczak7,Ru¨diger Voss9, and Anna Ga˚rdmark4 1Institute for Hydrobiology and Fisheries Science, Center for Earth System Research and Sustainability (CEN), Klima Campus, University of Hamburg, Grosse Elbstrasse 133, D-22767 Hamburg, Germany 2Scripps Institution of Oceanography, University of California San Diego, La Jolla, CA, USA 3Stockholm Resilience Centre, Stockholm University, Kra¨ftriket 10, SE-106 91 Stockholm, Sweden 4Department of Aquatic Resources, Institute of Coastal Research, Swedish University of Agricultural Sciences, Skolgatan 6, SE-742 22 O¨regrund, Sweden 5Department of Aquatic Resources, Institute of Marine Research, Swedish University of Agricultural Sciences, PO Box 4, SE-54330 Lysekil, Sweden 6Marine Research Center, Finnish Environment Institute (SYKE), Mechelininkatu 34a, PO Box 140, FI-00251 Helsinki, Finland 7Stockholm University Baltic Sea Centre, Stockholm University, SE-106 91 Stockholm, Sweden by Howard Browman on July 2, 2014 8Technical University of Denmark, National Institute of Aquatic Resources, Charlottenlund Castle, DK-2920 Charlottenlund, Denmark 9Department of Economics, University of Kiel, Wilhelm-Seelig-Platz 1, 24118 Kiel, Germany *Corresponding author: tel: +49 40 42838 6621; fax: +49 40 42838 6618; e-mail: [email protected] Mo¨llmann, C., Lindegren, M., Blenckner, T., Bergstro¨m, L., Casini, M., Diekmann, R., Flinkman, J., Mu¨ller-Karulis, B., Neuenfeldt, S., Schmidt, J. O., Tomczak, M., Voss, R., and Ga˚rdmark, A. 2014. Implementing ecosystem-based fisheries management: from single-species to integrated ecosystem assessment and advice for Baltic Sea fish stocks. – ICES Journal of Marine Science, 71: 1187–1197. Received 10 March 2013; accepted 24 June 2013; advance access publication 24 August 2013.

Theory behind ecosystem-based management (EBM) and ecosystem-based fisheries management (EBFM) is now well developed. However, the implementation of EBFM exemplified by fisheries management in Europe is still largely based on single-species assessments and ignores the wider ecosystem context and impact. The reason for the lack or slow implementation of EBM and specifically EBFM is a lack of a co- herent strategy. Such a strategy is offered by recently developed integrated ecosystem assessments (IEAs), a formal synthesis tool to quan- titatively analyse information on relevant natural and socio-economic factors, in relation to specified management objectives. Here, we focus on implementing the IEA approach for Baltic Sea fish stocks. We combine both tactical and strategic management aspects into a single strategy that supports the present Baltic Sea fish stock advice, conducted by the International Council for the Exploration of the Sea (ICES). We first review the state of the art in the development of IEA within the current management framework. We then outline and discuss an approach that integrates fish stock advice and IEAs for the Baltic Sea. We intentionally focus on the central Baltic Sea and its three major fish stocks cod (Gadus morhua), herring (Clupea harengus), and sprat (Sprattus sprattus), but emphasize that our ap- proach may be applied to other parts and stocks of the Baltic, as well as other ocean areas. Keywords: Baltic Sea, indicator approaches, integrated advice, integrated ecosystem assessment, strategic modelling.

# 2013 International Council for the Exploration of the Sea. Published by Oxford University Press. All rights reserved. For Permissions, please email: [email protected] 1188 C. Mo¨llmann et al.

Introduction generally strategic in that they explore possible future trajectories Ecosystem-based management (EBM) is now a central paradigm of human and natural systems under a multitude of natural and an- underlying living marine resource policy worldwide. Contrary to thropogenic pressures (Weyant et al., 1996; Levin et al., 2009). conventional resource management approaches, EBM addresses However, present EU fisheries management requires annual stock the cumulative impacts of multiple ocean uses and climate change assessment and advice. Hence, we attempt to combine both require- on multiple ecosystem components (Pikitch et al., 2004; Leslie ments into a strategy that supports the present Baltic Sea fish stock and McLeod, 2007; Marasco et al., 2007). Hence, the goal of EBM advice, conducted by the International Council for the Exploration is to find trade-offs between a diverse set of ecosystem services of the Sea (ICES). We first review the state of the art in the develop- and often conflicting, management goals (McLeod and Leslie, ment of EBFM and IA within the current management framework. 2009). Although the theory behind EBM is well developed, its imple- We then outline an IEA-based approach that integrates fish stock mentation lacks behind (Berkes, 2012). A clear example is fisheries advice and IEAs for the Baltic Sea. We intentionally focus on the management in Europe that is still largely based on single-species central Baltic Sea and the three major fish stocks cod (Gadus assessments and ignores the wider ecosystem context and impacts. morhua), herring (Clupea harengus), and sprat (Spratus sprattus), The reason for the lack or slow implementation of EBM is a lack but emphasize that our approach may be applied to other parts of a coherent strategy. Such a strategy is offered by Integrated and stocks of the Baltic, as well as other ocean areas. Assessment (IA). IA is commonly defined as an interdisciplinary process of combining, interpreting, and communicating knowledge Elements of EBM in present operational fish Downloaded from from diverse scientific disciplines, in such a way that the whole set of cause–effect interactions of a problem can be evaluated from a syn- stock assessment optic perspective with two characteristics: (i) it should have added ICES advice on total allowable catches (TACs) to the EU Commission value compared with single disciplinary assessments and (ii) it is currently based on targets of fishing mortality at maximum sus- should provide useful information to decision makers (Rotmans tainable yield (FMSY) and, for Eastern Baltic cod, on an EU manage- http://icesjms.oxfordjournals.org/ and Dowlatabadi, 1997; van der Sluijs, 2002). ment plan (EC, 2007; ICES, 2012a). The analytical assessments of IAs were initially developed during the 1970s, fuelled by scientific Baltic fish stocks delivering the basis for the advice are conducted and public policy efforts to understand and control acid deposition in on a stock-by-stock basis. Extended survivor analysis or a state- Europe and North America (van der Sluijs, 2002). Since the 1980s, space fish stock assessment model (www.stockassessment.org), both they have played a role in the development of the international based on commercial catch-at-age data supplemented by fishery- climate policy leadingto,andbeingpartof,theworkof theIntergovern- independent survey indices, are applied to obtain estimates of key menta lPanel on Climate Change (IPCC). In this context, IAs were stock parameters, such as spawning-stock biomass (SSB), recruit- largely based on IA models (IAMs), computer simulation models com- ment, and fishing mortality (F; Shepherd, 1999; ICES, 2012b). biningknowledgefrommultipledisciplines,andhencesuitedtoanalyse Multispecies interactions are currently only considered for the main forage fish species, namely sprat and central Baltic herring, environmental problems in an integrated fashion (Rotmans and by Howard Browman on July 2, 2014 Dowlatabadi, 1997). Nowadays, it is widely recognized that a complete and for cannibalism on juvenile cod. For these stocks, predation IA methodology combines IAMs with otheranalytical tools for integra- mortality by the top predator cod is derived from multispecies as- tion, such as analyses of large datasets and participatory approaches. sessment models (previously multispecies virtual population ana- Early developments of IAs within the field of marine living re- lysis; currently stochastic multispecies model; ICES, 2012a) and source management were started in North America and based on used to update the natural mortality input to the single species analyses of large datasets showing substantial changes in ecosystem forage fish stock assessments. states due to climate and human impacts (Hare and Mantua, 2000; As part of the assessment process, short-term forecasts (3 years) Link et al., 2002; Choi et al., 2005). Only recently, integrated ecosys- are conducted starting with the latest assessment year and using esti- tem assessments (IEAs) have been developed as a formal synthesis mates of recruitment based on empirical spawning stock–recruit- tool to quantitatively analyse information on relevant natural and ment relationships. Environmental data to inform these forecasts socio-economic factors, in relation to specified EBM objectives are only used for the herring stock in the Gulf of Riga (ICES, (Levin et al., 2009). IEAs provide a strategy to overcome the still pre- 2012b). Data used are the copepod Eurytemora affinis, the main vailing single-species and single-sector approaches; they organize prey for larval herring, and water temperature determining the science in order to inform decisions in marine EBM at multiple timing and distribution of herring spawning (Cardinale et al., 2009). scales and across sectors (Levin et al., 2009; Tallis et al., 2010). Previously, the winter index of the North Atlantic Oscillation The goals and objectives of EBM and IEA need to be specified be- (NAO) and water temperature have been used in forecasts of sprat forehand (Jennings, 2005; Levin et al., 2009). The specified goals year-class strength (MacKenzie and Ko¨ster, 2004; ICES, 2006), but then define the focal components of the EBM and IEA approach these are no longer used due to problems with matching environ- (Kershner et al., 2011), which may range from a specific ecosystem- mental time-series updates with the assessment meetings. based fisheries management (EBFM) to a full blown cross-sector Multispecies modelling has a long tradition for the Baltic Sea eco- EBM approach (Leslie and McLeod, 2007; McLeod and Leslie, system (Sparholt, 1994; Gislason, 1999; Ko¨ster et al., 2001) but has 2009). Although the ambitious objectives of the European Union never been used fully in operational fish stock assessment (aside (EU) Marine Strategy Framework Directive (MSFD) call for a full from providing input to single species stock assessments, see EBM approach, fisheries management within the reformed EU above).Howeverrecently, inresponse toa request by the EUcommis- Common Fisheries Policy (CFP) still relies on the development sion, a multispecies assessment has been conducted as a pilot study and implementation of a more sector-specific EBFM approach. towards implementing the ecosystem approach (ICES, 2012d; Here, we focus on the process of implementing EBFM for the STECF, 2012). The assessment was conducted using SMS that Baltic Sea using elements of the IEA approach. IAs and IEAs are accounts for the mortality inflicted by cod on sprat, herring, and ju- venile cod. Simulations revealed all species-specific multispeciesFMSY Implementing EBFM 1189 to be higher than their single-species counterparts. However, particu- reorganizations in ecosystem structure and function bear important larly for cod and sprat, simulations show that higher fishing mortality social and economic costs and may be difficult to reverse (Scheffer will give very similar long-term yields, but result in lower SSBs and a et al., 2001; Suding et al., 2004). Hence, management strategies higher riskofstockcollapse.Model resultsfurtherindicatethathigher that help prevent unwanted change are necessary, and these rely F on cod will result only in a minor increase in cod yield, but in higher on indicators showing warning signs well before a transition sprat and herring yields. However, modelling of FMSY currently might occur (Scheffer et al., 2009). Although the theoretical founda- ignores structural uncertainty such as variable predator–prey tion of early warning indicators has greatly improved (Scheffer et al., overlap and density-dependent growth. Hence, ICES (2012d) and 2012), the application of potential early warning indicators in real STECF (2012) concluded that more work is needed to fully under- ecosystems is stillvery limited. Lindegren et al. (2012b) applied mul- stand the results of the multispecies runs and their implications for tiple early warning indicators to monitoring data of key ecosystem the Baltic fish stock advice and management. components of the Baltic Sea, namely the zooplankton species Pseudocalanus acuspes and Acartia spp. They demonstrated that Indicator approaches to IAs the ability of the indicators to forewarn the major ecosystem Integrated trend and status assessments regime shift in the central Baltic Sea is variable depending on the indicator time-series used. Hence, a multiple method approach Despite efforts towards multispecies considerations, Baltic fish for early detection of ecosystem regime shifts is proposed that can stock assessments are mostly single species, ignoring the larger eco-

be useful in informing timely management actions in the face of Downloaded from system context (Casini et al., 2011a). As a first step towards develop- ecosystem change (Lindegren et al., 2012b). ing more integrative assessments, analyses on the state and development of the various Baltic ecosystems have been conducted using large multitrophic datasets (ICES, 2008; Mo¨llmann et al., Indicator approach in support of single-species fish 2009; Diekmann and Mo¨llmann, 2010; Lindegren et al., 2010a, stock advice

2012a). These so-called integrated trend analyses (ITAs) used multi- Results and data from the above reported integrated trendand status http://icesjms.oxfordjournals.org/ variate statistics based on an approach developed for the North assessments have also been used for developing a set of ecosystem Pacific and Atlantic, as well as the North Sea ecosystems (Hare and indicators that can support the assessment of the stock status of in- Mantua, 2000; Link et al., 2002; Choi et al., 2005; Weijerman et al., dividual species (Ga˚rdmark et al., 2011). Using eastern Baltic cod as 2005; Kenny et al., 2009; Mo¨llmann and Diekmann, 2012). a case study, they developed a set of indicators including size/age Analyses of these large datasets were conducted using dimension re- structure of the target stock, predator, and prey levels for early life- duction techniques, mainly principal component analysis (PCA; e.g. history stages and physical oceanographic conditions. Individual Legendre and Legendre, 1998) and methodologies to identify step indicators were evaluated for showing support for or against the as- changes in biotic and abiotic variables, such as STARS (Rodionov, sessment model-derived stock status and trends. Additionally, indi- 2004) or chronological clustering (Legendre et al., 1985). vidual indicators were combined into an ecosystem-based indicator

Here, we present a reanalysis of data from the Central Baltic Sea of cod recruitment potential using fuzzy-logic networks (Jarre et al., by Howard Browman on July 2, 2014 using 57 annual time-series (1979–2010), in two separate PCAs on 2008). Ga˚rdmark et al. (2011) thus show how to use ecosystem indi- 28 biotic variables from phytoplankton to fish and 29 abiotic cators for reducing the uncertainty in model-based estimates of variables describing the physical conditions and anthropogenic stock status and for determining more precautionary harvest driving forces such as nutrient concentrations and fishing pressure control rules (HCRs). (Figure 1). Connecting biotic year scores chronologically on the first factorial plane (PC1 vs. PC2) results in a time trajectory that shows Integrated modelling approaches the regime shift during the late 1980s/early 1990s (Figure 1a). The Ecological models change in the ecosystem is displayed by the opposition of cod and Like in many other areas, fish stock assessments for the Baltic Sea use herring as well as sprat and the copepod Acartia spp., showing the single-species fisheries modelsasa basis for short-term tacticalsetting strongest loadings on PC1 (Figure 1b). According to Mo¨llmann of TAC. However, to evaluate the impact of alternative management et al. (2009), the period between 1987 and 1992 can be interpreted strategies on the state and dynamics of exploited fish populations, as a transition period, in which a major reorganization of the ecosys- strategic modelling approaches are essential tools. Obviously, an tem occurred, due to the interaction of abrupt climatic changes, un- exploited population does not exist in isolation, and its response to sustainable fishing pressure and eutrophication. Our updated PCA a fishing pressure emerges from the feedbacks caused by its interac- confirms that after 1992, abiotic conditions largely returned to tions with other species in the foodweb, as well as from direct and in- values similar to the initial state (Figure 1c), indicating hysteresis direct effects of abiotic pressures. The essence of EBFM is to account and stabilizing feedbacks in the foodweb (Casini et al., 2009, 2010; for such foodweb mediated feedbacks and indirect effects (McLeod Mo¨llmann et al., 2009). The analysis of the major abiotic loadings and Leslie, 2009). Hence, strategic modelling for EBFM involves confirmed that a temperature increase and salinity decrease were evaluating exploitation effects in the presence of species interactions major drivers of the central Baltic regime shift (Figure 1d). and under alternative future environments. Similar major reorganizations during the late 1980s/ early 1990s For the main exploited species in the central Baltic Sea, strategic were found synchronously for almost all studied Baltic ecosystems modelling for EBFM has been performed using a few approaches. ¨ (Diekmann and Mollmann, 2010; ICES, 2012c). Lindegren et al. (2009) developed a stochastic multivariate autore- gressive model (BALMAR) of cod, sprat, herring, and their zoo- Early warning indicators plankton prey including the influence of fishing and climate Ecosystem regime shifts, like those observed in the Baltic Sea, are variables. Such a model represents statistically derived relationships reported for many other marine ecosystems in the world between the species, and any management impact analyses using (Mo¨llmann and Diekmann, 2012). These large-scale this model (e.g. Lindegren et al., 2010b) thus relies on the 1190 C. Mo¨llmann et al. Downloaded from http://icesjms.oxfordjournals.org/ by Howard Browman on July 2, 2014

Figure1. ResultsofanITAfor thecentralBalticSeaecosystemofbiotic(aand b) andabiotic (c andd)variables.(a andc)TimetrajectoriesofPCscores on the first factorialplanebased on a normalized PCA.(band d) Biplotsshowing the ten variables that were best represented on the first factorial plane (time trajectory in the background; variables and years are scaled symmetrically by square root of eigenvalues). Cod/Her/Spr, cod/herring/sprat; R, recruitment;F, fishing mortality; Ac1,Acartia spp.biomass;Chla2,4,chlorophylla summerconcentrationinBornholmand Gotland Basin,respectively; Dino 3, dinoflagellate spring biomass in Bornholm Basin; X11psu, depth of the 11psu isohaline; T 5,6,7,8, midwater temperature in Bornholm and Gotland Basin in spring and summer, respectively; Anoxic, area with anoxic bottom water; S4, deepwater salinity in the Gotland Basin. assumption that the species interactions (i.e. parameters) do not isto explicitly model individual levelprocesses, such as feeding (pre- change outside the estimated range obtained during the period dation), metabolism, energy allocation, and resulting growth, mor- used for model fitting. An alternative is to represent trophic interac- tality, and reproduction, as done in physiologically structured tions between species in more detail as, for example, is possible with population models (Metz and Diekmann, 1986; de Roos and the Ecopath with Ecosim (EwE) modelling approach (Christensen Persson, 2001). Thistype of models has been developed for the inter- and Walters, 2004). The EwE approach has been applied to the actions between Baltic cod, sprat, and their resources (van Leeuwen central Baltic Sea foodweb (Harvey et al., 2003; O¨ sterblom et al., et al., 2008, 2013), and between herring, sprat, and their resources 2007) and in its present form includes 22 functional groups (Huss et al., 2012), to assess qualitative responses of cod, sprat, (Niiranen et al., 2012; Tomczak et al., 2012). However, since inter- and herring to environmental changes under alternative types of actions occur between individuals rather than populations, a interactions. As exemplified by these three modelling approaches, more mechanistic approach towards modelling trophic interactions available foodweb (or multispecies) models of the central Baltic Implementing EBFM 1191

Sea differ greatly in fundamental assumptions on how to represent maximizing economic rent, employment, and biomass extraction ecological processes. Simulated responses of exploited species and (Quaas et al., 2012). foodwebs to fishing may therefore depend on model choice. A future IEA strategy for Baltic fish stock advice Biological ensemble modelling approach and management To overcome the issue of model choice for impact analyses, In the following section, we outline how the approaches and tools Ga˚rdmark et al. (2013) developed a biological ensemble modelling described above can be combined into an IEA strategy that facilitates approach (BEMA), which can be used to assess the relative influence the implementation of EBFM for the Baltic Sea (Figure 3). The strat- of model structure on simulated species responses, as well as to for- egy includes three components: (i) the transition from existing mulate fisheries management advice robust to such uncertainty. In single-species to a multispecies stock assessment, (ii) an ecosystem the BEMA, a set of seven ecological models (ranging from single- assessment that integrates environmental information into the species to foodweb models of varying complexity) was subjected to single-/multispecies assessment, and (iii) a strategic component the same initial conditions and external forcing based on a combin- that conducts long-term management strategy evaluation using ation of exploitation and climate change scenarios to simulate the coupled ecological and economic models. Hence, our strategy responses of eastern Baltic cod to historically observed high fishing accounts for both the short-term needs of annual fish stock assess- levels vs. fishing at management target levels, including a number ments, conducted for most of the European fish stocks, but also the

of future climate change scenarios (Figure 2). The BEMA ensemble long-term needs of future strategic EBM advice (Figure 3). Downloaded from was then used to: (i) identify model assumptions causing key diver- gence in simulated species responses (by contrasting ensemble Towards multispecies stock assessments subsets), (ii) evaluate the relative importance of model structure un- Single-species assessments are to a large degree still the basis for certainty for overall uncertainty in simulated responses (by contrast- present day fish stock advice. However, we envisage an increasing ing variation among models within a climate trajectory with use of multispecies assessment models in the future for a more real- variation within each model among all climate trajectories), and istic assessment of multispecies MSY reference points, population http://icesjms.oxfordjournals.org/ (iii) formulate robust advice for fisheries management (by identify- sizes relative to target levels, and TACs. Multispecies models can ing conclusions on management effects common for the whole en- be applied indirectly by providing predation mortality rates to be semble). Ga˚rdmark et al. (2013) demonstrated that assumptions of used in single-species models (Figure 3). This procedure is already species interactions greatly impacted simulated cod dynamics, with common practice for some Baltic fish stocks but should be con- models lacking stabilizing predator–prey feedbacks showing large ducted on a more regular, preferably annual, basis. Multispecies interannual fluctuations and a greater sensitivity to the underlying assessments may in the future replace the single-species procedures, uncertainty of climate forcing. Nevertheless, robust conclusions when the implications for multispecies interactions especially for regarding the effects of alternative fishing levels could be found, MSY calculation are sufficiently evaluated, as is recently initiated e.g. in all models, intense fishing prevented recovery and climate (ICES, 2013). Furthermore, additional effort is required for improv- by Howard Browman on July 2, 2014 change further decreased the cod population. Although the BEMA ing multispecies assessment models, e.g. with respect to implement- (Ga˚rdmark et al., 2013) has so far only been applied to single-species ing predator–prey feedbacks and density-dependent growth responses, the approach is equally suitable to study indirect effects of (Ga˚rdmark et al., 2013) as well as the dependence of population exploitation on non-target species (ICES, 2009), groups of species processes on environmental conditions. Equally important are providingparticularecosystemservices,aswellasindicators ofenvir- improved monitoring systems needed to account for the increased onmental status (ICES, 2012c) as developed for the MSFD. data requirements of multispecies models, especially with respect to data on predator diets and key functional groups affecting fish Coupled ecological–economic modelling performance, such as zooplankton. Overall, a replacement of When formulating new fishing rules, basic economic conditions for single- by multispecies assessments which account for key species the relevant fishery are often not understood and therefore ignored. interactions and environmental influences would be a first step Coupled ecological–economic optimization models have been towards integrated fish stock advice for the Baltic Sea. developed and analysed for Baltic cod, herring, and sprat stocks. These models are available as single-species (Quaas et al., 2013)or The importance of ecosystem assessments as multispecies type, i.e. accounting for cod preying upon herring Accounting for ecosystem effects of fishing, as well as effects of en- and sprat (Nieminen et al., 2012). These models are also able to vironmental variability on fish stock productivity is a crucial com- address climate change impacts by using environmentally sensitive ponent of EBFM (Pikitch et al., 2004). Fish stock assessments stock–recruitment functions. Changes in optimal fishing strategy traditionally rely on assumptions regarding stationary (equilib- under climate change scenarios can therefore be computed (ICES, rium) dynamics of populations, communities, and the carryingcap- 2010; Voss et al., 2011). Easily understandable, well-defined indica- acity of ecosystems. However, a large body of research has shown tors have proven to be especially helpful to foster transfer and appli- that abrupt changes in productivity, e.g. due to climate, may cation of economic analyses into “real-world” fisheries management. occur (e.g. Brander, 2007), with far reaching consequences for eco- In this context, a new ecological–economic indicator was devel- system structure and functioning, as well as the provision of import- oped: the shadow interest rate, SIR (Quaas et al., 2012). The SIR ant goods and services (Scheffer et al., 2001; Millennium Ecosystem extends economic concepts and considers fish stocks as natural Assessment, 2005). The Baltic Sea, having experienced an ecosystem capital stocks. Furthermore, the SIR, as a generic measure, allows regime shift (Mo¨llmann et al., 2008), including trophic cascading quantifying and comparing the economic success or failure of fish- (Casini et al., 2008; Mo¨llmann et al., 2009), as well as a collapse of eries management across stocks. It captures biological information the cod stock, clearly illustrates that changes in the environment on stock productivity as well as economic information and allows as well as foodweb interactions need to be considered for sustainable the trade-off of management objectives to be assessed, e.g. management (Lindegren et al., 2009). Hence, ecosystem 1192 C. Mo¨llmann et al. Downloaded from http://icesjms.oxfordjournals.org/ by Howard Browman on July 2, 2014

Figure 2. The BEMA (Ga˚rdmark et al., 2013) can be used to derive robust, strategic advice for EBFM. Combinations of management and environmental scenarios are used to force a set of ecological models of varying complexity (e.g. single species, multispecies, and foodweb models) to simulate fish stock responses to exploitation in future environments. The performance of management strategies can then be compared across (i) ecological models, (ii) environmental scenarios, and (iii) environmental variation to identify management strategies and HCRs robust to the uncertainties of ecological processes and future conditions. Implementing EBFM 1193

relation to targets) according to the state of the EHI relative to GES. An example for a precautionary approach for Baltic cod is sug- gested by Ga˚rdmark et al. (2011). Into the future: long-term management strategy evaluation The fish stock and ecosystem assessments described above primarily focus on short-term, tactical management aspects. However, envir- onmental conditions and impacts on ecosystems and their fish stocks may vary over longer time-scales, especially in the light of expected future climate change (MacKenzie et al., 2012; Meier et al., 2012). Hence, strategic long-term simulations are needed to define management goals (such as FMSY) and to evaluate how robust these goals are to future climate change (Lindegren et al., 2010a, b), to fishing effects on the ecosystem (Lassen et al., 2013), and to our ignorance of foodweb interactions in general. In add-

ition, Ga˚rdmark et al. (2013) showed that a multimodel approach Downloaded from (i.e. BEMA) would be beneficial for developing management Figure 3. Schematic outline for a future IEA strategy for Baltic fish goals robust to uncertainties inherent in such simulations due to stock advice and management including (1) multispecies fish stock model structure and parameterization (Niiranen et al., 2012). The assessments, (2) ecosystem assessments, and (3) long-term management strategy evaluation: F, fishing mortality; MSY, maximum proposed long-term management strategy evaluations will be in- sustainable yield; TAC, total allowable catch; i, fish species, i.e. cod, strumental in developing multispecies long-term management http://icesjms.oxfordjournals.org/ herring, and sprat; stock, indicators of stock status and structure; Env, plans (STECF, 2012). Eventually, coupled ecological–economical indicators on the state of the abiotic and biotic environment; EW, early models can be used to evaluate the economic implications of man- warning indicators; EHI, ecosystem health index; GES, good agement strategies relative to environmental conditions (Voss et al., environmental status; BEMA, biological ensemble modelling approach; 2011; Lassen et al., 2013). ECON, coupled ecological/economical modelling. Discussion We have proposed a strategy that integrates the present Baltic Sea assessments are an important component of an effective EBFM ap- single-species fish stock assessment and advice with elements of proach. For the Baltic Sea, we have identified three groups of indica- IEAs. The approach focuses on (i) integrating environmental infor- tors that are needed to inform fish stock assessments (Figure 3). mation into the short-term tactical fish stock assessment and (ii) by Howard Browman on July 2, 2014 First, indicators of stock status and structure (“Stock” in using existing ecological and coupled ecological–economic Figure 3), other than biomass and abundance, need to be considered models in simulations to evaluate management strategies and to an- (Ga˚rdmark et al., 2011; Eero et al., 2012). These are also required by ticipate environmental productivity changes. The strategy is largely the MSFD in its “Descriptor 3” (EC, 2008) and should describe based on existing studies and models for the Baltic Sea and should be growth and recruitment of the stocks. Second, the state of the readily implemented and operational. abiotic and biotic environment needs to be evaluated (“Env” in However, although many indicators for the ecosystem assess- Figure 3). Ga˚rdmark et al. (2011) propose indicators for the Baltic ment have been proposed (e.g. Ga˚rdmark et al., 2011; Eero et al., cod stock related to physical oceanographic parameters important 2012), a standard set of indicators needs to be developed. Hence, in- for recruitment, as well as prey and predator effects on early life dicator development is a crucial part of the IEA framework (Levin stages. Similar indicators have been identified also for the other et al., 2009) and should comprise (i) aformal and objective indicator fish stocks in the Baltic Sea (Lindegren et al., 2011; Eero et al., selection routine, involving both scientists and stakeholders (Levin 2012). In addition, the PC-based holistic indicator from ITAs (see et al., 2010; Kershneret al., 2011), and (ii) an approach for determin- above) can provide important information about the ecosystem dy- ing target and reference levels based on time-series and ecosystem namics and state in a holistic way (Link et al., 2002; Mo¨llmann et al., modelling (Samhouri et al., 2009, 2010, 2011, 2012). 2009). Finally, early warning indicators (“EW” in Figure 3) are im- After indicator selection, a further important step in our strategy portant in providing timely detection of ecosystem change would be an analysisthatidentifiesthe risk to the indicators posed by (Lindegren et al., 2012b). human activities and natural processes. Risk analyses have the goal A standing difficultyof using ecosysteminformation in an assess- to determine the probability that an ecosystem indicator will reach ment framework is to translate ecosystem indicator information or remain in an undesirable state (Levin et al., 2009). Various tech- into decision criteria for management (Link, 2005). A first step niques exist that use qualitative expert opinion or quantitative tech- would be to combine individual ecosystem indicators into an eco- niques such as statistical analyses and ecosystem modelling (Smith system health index (EHI) (Figure 3). EHI indicates the state of et al., 2007; Samhouri and Levin, 2012). As demonstrated here, the environment relative to its good environmental status (GES), the necessary data and modelling tools are available to employ a goal of many environmental policy drivers such as the MSFD these techniques for the Baltic Sea ecosystem to implement this im- (EC, 2008). A number of procedures is available for combining in- portant step of IEA in the here proposed strategy. dividual indicators, such as fuzzy logic and various weighting The implementation of the outlined approach depends on the schemes (Ga˚rdmark et al., 2011; Ojaveer and Eero, 2011; Halpern development of an integrated monitoring programme that routine- et al., 2012). Eventually, rules have to be defined that modify the ly measures the selected indicators. The monitoring programme HCR (i.e. rules for how the TAC shall be set depending on state in should combine the present fish stock surveys with environmental 1194 C. Mo¨llmann et al. monitoring, such as conducted by HELCOM (2011). At present, a models able to address multisector use of ocean ecosystems are number of key indicators are insufficiently sampled, e.g. an indica- rare. Hence, multiple model sets are potentially needed to cover tor of GES of zooplankton in relation to fish stocks (Mo¨llmann et al., multiple goals, and rigouros management goal-specific selection 2008). Access to complex biological data, such as zooplankton and criteria for model inclusion need to be developed. In general, phytoplankton, is still difficult and reduced funding poses a serious models should be prioritized that include stabilizing predator– threat to the important continuation of monitoring and the main- prey feedbacks (Ga˚rdmark et al., 2013), climate as well as anthropo- tenance of long-term time-series. Furthermore, monitoring should genic drivers, and that allow studying indirect exploitation effects have a wide spatial distribution at an appropriate resolution, since among important target and non-target species. recent studies have shown important changes in the distributions Eventually implementing our approach requires a fundamental of key species such as sprat (Casini et al., 2011b) and cod (Casini change in how fish stock assessment and advice is conducted et al., 2012) which have important implications for management (Casini et al., 2011a). More effort needs to be shifted from the (Eero et al., 2012). Eventually, regular predator stomach sampling regular single-species procedure towards implementing multispecies is required for conducting reliable multispecies assessments. and ecosystem assessments. This potentially requires a reduction The utility and implementation of the proposed approach hinges in the temporal frequency of single-species assessments, allowing on the development of management plans that can account for the more effort towards the development of ecosystem assessments. type of information that integrated fish stock advice would provide. Moreover, fish stock and ecosystem assessments should be combined

Management plans need to include explicit rules on how TACs (and into an integrative, interdisciplinary framework. Within the ICES Downloaded from other management measures) are to be determined using informa- framework, the required integration of fish stock and ecosystem tion provided by, e.g. ecosystem assessments. Ga˚rdmarket al. (2011) assessments is facilitated by the ICES SCICOM Steering Group on givean example how HCRs can account for uncertainty in the advice Regional Sea Programmes (SSGRSP) and currently supported by for Baltic cod, using among others ecosystem information. They the planned revision of the ICES science plan, which gives the devel- suggest HCRs to be modified depending on environmental state, opment of IEA a prominent position in the organization. But IEA e.g. reflected by the EHI (Figure 3). Furthermore, harvest strategy implementation needs an enhanced cooperation with regional con- http://icesjms.oxfordjournals.org/ frameworks able to deal with a broad range of information are ventions such as HELCOM (Helsinki Commission) and OSPAR being used in fisheries management in Australia using indicators (Convention for the Protection of the marine Environment of of fishing and stock trends, combined in a hierarchical strategic ap- the Northeast Atlantic) and related EU bodies such as STECF proach (Smith et al., 2007). The approach to long-term manage- (Scientific, Technical, and Economic Committee for Fisheries). ment strategy evaluation we propose can be instrumental in However, in the Baltic Sea, as well as in other areas, the data, knowl- setting up these new EBM plans. edge, and tools for IEA and EBFM are readily available and should be Our strategy described here implicitly demands for a resilience applied and implemented without further delay. approach to ecosystem management (Folke et al., 2004; Hughes et al., 2005; Levin and Lubchenco, 2008; Fujita et al., 2012). by Howard Browman on July 2, 2014 Reducing resilience of populations, communities, and foodwebs Acknowledgements to environmental change, through anthropogenic impacts, can We thank two anonymous referees and Jason Link for comments cause critical transitions, i.e. regime shifts, as observed in the greatly improving the manuscript. A part of the work leading to Baltic Sea. Implementing early warning indicators and evaluating this paper was funded through the European Union Framework potential future scenarios accounts for a precautionary approach Seven project MYFISH (“Maximising yield of fisheries while balan- that ensures the health and resilience of Baltic Sea fish stocks. cing ecosystem, economic and social concerns”). The paper is to a Furthermore, regime-based approach to EBFM can easily be imple- large degree the outcome of work conducted by the “ICES/ mented in our framework (King and McFarlane, 2006; Fu et al., HELCOM Working Group on Integrated Assessments of the 2013; Szuwalski and Punt, 2013). Baltic Sea (WGIAB)”. We are grateful to all past and present The approach developed here benefits from the relative simple members of WGIAB for their support and input. We dedicate the three species fisheries system of the central Baltic Sea and the avail- article to the late Johan Modin, who was a supporter of WGIAB ability of long-term datasets and modelling approaches (Casini from its start. et al., 2011a). However, our strategy can readily be applied to more complex fisheries ecosystems that usually have (i) fairly devel- References oped multispecies fisheries models (e.g. Garrison et al., 2010; Berkes, F. 2012. Implementing ecosystem-based management: evolu- Howell and Bogstad, 2010; Kempf et al., 2010; Link et al., 2011a), tion or revolution? Fish and Fisheries, 13: 465–476. (ii) long-term time-series for ecosystem assessment (Link et al., Brander, K. M. 2007. Global fish production and climate change. 2002; Kenny et al., 2009), and (iii) foodweb and ecosystem models Proceedings of the National Academy of Sciences of the USA, 104: for strategic long-term management strategy evaluation (e.g. Link 19709–19714. et al., 2011a, b; Kaplan et al., 2012). Naturally the higher complexity Cardinale, M., Mo¨llmann, C., Bartolino, V., Casini, M., Kornilovs, G., in terms of foodweb structure and related mixed fisheries in ecosys- Raid, T., Margonski, P., et al. 2009. Effect of environmental variabil- tems such as the North Sea (Ulrich et al., 2012) makes ecosystem ity and spawner characteristics on the recruitment of Baltic herring assessments and long-term modelling approaches more challen- Clupea harengus populations. Marine Ecology Progress Series, 388: 221–234. ging. However, we are convinced that conducting rigorous and goal- Casini, M., Bartolino, V., Molinero, J. C., and Kornilovs, G. 2010. specific indicator selection and risk assessment help overcome these Linking fisheries, trophic interactions and climate: threshold dy- challenges. The multimodel approach for strategic long-term simu- namics drive herring Clupea harengus growth in the central Baltic lations advocated here (i.e. BEMA) is less straightforward when Sea. Marine Ecology Progress Series, 413: 241–252. applied to more than one target species (as in Ga˚rdmark et al., Casini, M., Blenckner, T., Mo¨llmann, C., Ga˚rdmark, A., Lindegren, M., 2013) or multiple management goals, especially since foodweb Llope, M., Kornilovs, G., et al. 2012. Predator transitory spillover Implementing EBFM 1195

induces trophic cascades in ecological sinks. Proceedings of the Biological ensemble modelling to evaluate potential futures of National Academy of Sciences of the USA, 109: 8185–8189. living marine resources. Ecological Applications, 23: 742–754. Casini, M., Hjelm, J., Molinero, J. C., Lo¨vgren, J., Cardinale, M., Garrison, L. P., Link, J. S., Cieri, M., Kilduff, P., Sharov, A., Vaughan, D., Bartolino, V., Belgrano, A., et al. 2009. Trophic cascades promote Muffley, B., et al. 2010. An expansion of the MSVPA approach for threshold-like shifts in pelagic marine ecosystems. Proceedings of quantifying predator–prey interactions in exploited fish communi- the National Academy of Sciences of the USA, 106: 197–202. ties. ICES Journal of Marine Science, 67: 856–870. Casini, M., Kornilovs, G., Cardinale, M., Mo¨llmann, C., Grygiel, W., Gislason, H. 1999. Single and multispecies reference points for Baltic Jonsson, P., Raid, T., et al. 2011b. Spatial and temporal density de- fish stocks. ICES Journal of Marine Science, 56: 571–583. pendence regulates the condition of central Baltic Sea clupeids: com- Halpern, B. S., Longo, C., Hardy, D., McLeod, K. L., Samhouri, J. F., pelling evidence using an extensive international acoustic survey. Katona, S. K., Kleisner, K., et al. 2012. An index to assess the health Population Ecology, 53: 511–523. and benefits of the global ocean. Nature, 488: 615–622. Casini, M., Lo¨vgren, J., Hjelm, J., Cardinale, M., Molinero, J. C., and Hare, S. R., and Mantua, N. J. 2000. Empirical evidence for North Pacific Kornilovs, G. 2008. Multi-level trophic cascades in a heavily regime shifts in 1977 and 1989. Progress in Oceanography, 47: exploited open marine ecosystem. Proceedings of the Royal 103–145. Society of London, Series B, 275: 1793–1801. Harvey, C. J., Cox, S. P., Essington, T. E., Hansson, S., and Kitchell, J. F. ¨ ¨ Casini, M., Mollmann, C., and Osterblom, H. 2011a. Ecosystem ap- 2003. An ecosystem model of food-web and fisheries interactions in proach to fisheries in the Baltic Sea: present and potential future the Baltic Sea. ICES Journal of Marine Science, 60: 939–950. applications in assessment and management. In Ecosystem Based

Management for Fisheries—an Evolving Perspective, pp. 9–31. HELCOM. 2011. HELCOM Activities 2011 Overview (Baltic Sea Downloaded from Ed. by A. Belgrano, and C. Fowler. Cambridge University Press, Environment Proceedings No. 132). Published by the Helsinki Cambridge, UK. Commission. Choi, J., Frank, K., Petrie, B., and Leggett, W. 2005. Integrated assess- Howell, D., and Bogstad, B. 2010. A combined Gadget/FLR model for ment of a large marine ecosystem: a case study of the devolution of management strategy evaluations of the Barents Sea Fisheries. the eastern Scotian Shelf, Canada. Oceanography and Marine ICES Journal of Marine Science, 67: 1998–2004.

Biology: an Annual Review, 43: 47–67. Hughes, T. P., Bellwood, D. R., Folke, C., Steneck, R. S., and Wilson, J. http://icesjms.oxfordjournals.org/ Christensen, V., and Walters, C. J. 2004. Ecopath with Ecosim: methods, 2005. New paradigms for supporting the resilience of marine ecosys- capabilities and limitations. Ecological Modelling, 172: 109–139. tems. Trends in Ecology and Evolution, 20: 380–386. de Roos, A. M., and Persson, L. 2001. Physiologically structured Huss,M., Ga˚rdmark, A., van Leeuwen, A., and de Roos,A.M. 2012. Size- models—from versatile technique to ecological theory. Oikos, 94: and food-dependent growth drives patterns of competitive domin- 51–71. ance along productivity gradients. Ecology, 93: 847–857. Diekmann, R., and Mo¨llmann, C. (Eds) 2010. Integrated ecosystem ICES. 2006. Report of the Baltic Fisheries Assessment Working Group assessments of seven Baltic Sea areas. ICES Cooperative Research (WGBFAS). ICES Document CM 2006/ACFM: 24. Report, 302. 90 pp. ICES. 2008. Report of the ICES/HELCOM Working Group on Diekmann, R., Otto, S., and Mo¨llmann, C. 2012. Towards Integrated Integrated Assessment of the Baltic Sea (WGIAB). ICES Document Ecosystem Assessments (IEAs) of the Baltic Sea—investigating eco- CM 2008/BCC: 4. by Howard Browman on July 2, 2014 system state and historical development. In Climate Impacts on the ICES. 2009. Report of the ICES/HELCOM Working Group on Baltic Sea: From Science to Policy, 1st edn. Ed. by M. Reckermann, K. Integrated Assessments of the Baltic Sea (WGIAB). ICES Document Brander, B. R. MacKenzie, and A. Omstedt. Springer. 260 pp. CM 2009/BCC: 2. Heidelberg, Germany. ICES. 2010. Workshop on including socio-economic considerations EC (European Commission). 2007. Council Regulation (EC) No. 1098/ into the climate-recruitment framework developed for clupeids in 2007 establishing a multi-annual plan for the cod stocks in the Baltic the Baltic Sea (WKSECRET). ICES Document CM 2010/SSGRSP: 9. Sea and the fisheries exploiting those stocks, amending Regulation ICES. 2012a. International Council for the Exploration of the Sea. / / (ECC) No 2847 93 and repealing Regulation (EC) No 779 97. Advice 2012, Book 8. EC (European Commission). 2008. Directive 2008/56/EC of the ICES. 2012b. Report of the Baltic Fisheries Assessment Working Group European Parliament and of the Council of 17 June 2008 establishing (WGBFAS). ICES Document CM 2012/ACOM: 10. a framework for community action in the field of marine environ- mental policy (Marine Strategy Framework Directive). ICES. 2012c. Report of the ICES/HELCOM Working Group on Integrated Assessment of the Baltic Sea (WGIAB). ICES Document Eero, M., Lindegren, M., and Ko¨ster, F. W. 2012. The state and relative CM 2012/SSGRSP: 2. importance of drivers of fish population dynamics: an indicator- based approach. Ecological Indicatos, 15: 248–252. ICES. 2012d. Report of the Workshop on Integrated/Multispecies Folke, C., Carpenter, S. R., Walker, B. H., Scheffer, M., Elmqvist, T., Advice for Baltic Fisheries (WKMULTBAL). ICES Document CM / Gunderson, L. H., and Holling, C. S. 2004. Regime shifts, resilience 2012 ACOM: 43. and biodiversity in ecosystem management. Annual Review in ICES. 2013. Report of the Benchmark Workshop on Baltic Multispecies Ecology, Evolution and Systematics, 35: 557–581. Assessments (WKBALT). ICES Document CM 2013/in press. Fu, C., Perry, R.I., Shin, Y.-J., Schweigert, J., and Liu, H. 2013. An eco- Jarre, A., Paterson, B., Moloney, C. L., Miller, D. C. M., Field, J. G., and system modelling framework for incorporating climate regime Starfield, A. M. 2008. Knowledge-based systems as decision support shifts into fisheries management. Progress in Oceanography, tools in an ecosystem approach to fisheries: comparing a fuzzy-logic doi:10.1016/j.pocean.2013.03.003. and a rule-based approach. Progress in Oceanography, 79: 390–400. Fujita, R., Moxley, J. H., DeBey, H., Van Leuvan, T., Leumer, A., Honey, Jennings, S. 2005. Indicators to support an ecosystem approach to fish- K., Aguilera, S., et al. 2012. Managing for a resilient ocean. Marine eries. Fish and Fisheries, 6: 212–232. Policy, doi:10.1016/ j.marpol.2012.05.025 Kaplan, I. C., Horne, P. J., and Levin, P. S. 2012. Screening California Ga˚rdmark, A., Nielsen, A., Floeter, J., and Mo¨llmann, C. 2011. Depleted Current fishery management scenarios using the Atlantis marine fish stocks and ecosystem-based management: on the road to end-to-end ecosystem model. Progress in Oceanography, 102: 5–18. recovery, we need to be precautionary. ICES Journal of Marine Kempf, A., Dingsør, G. E., Huse, G., Vinther, M., Floeter, J., and Science, 68: 212–220. Temming, A. 2010. The importance of predator–prey overlap: pre- Ga˚rdmark, M., Lindegren, M., Neuenfeldt, S., Blenckner, T., dicting North Sea cod recovery with a multispecies assessment Heikinheimo, O., Mu¨ller-Karulis, B., Niiranen, S., et al. 2013. model. ICES Journal of Marine Science, 67: 1989–1997. 1196 C. Mo¨llmann et al.

Kenny, A., Skjoldal, H., Engelhard, G., Kershaw, P., and Reid, J. 2009. An Shelf Large Marine Ecosystem: towards ecosystem-based fisheries integrated approach for assessing the relative significance of human management. US Department of Commerce, Northeast Fisheries pressures and environmental forcing on the status of large marine Science Center Reference Document 11-23. 89 pp. ecosystems. Progress in Oceanography, 81: 132–148. Link, J. S., Gamble, R. J., and Fulton, E. A. 2011b. NEUS—Atlantis: con- Kershner, J., Samhouri, J. F., James, C. A., and Levin, P. S. 2011. Selecting struction, calibration, and application of an ecosystem model with Indicator Portfolios for Marine Species and Food Webs: a Puget ecological interactions, physiographic conditions, and fleet behav- Sound Case Study. PLoS One, 6: e25248. doi:10.1371/journal.- ior. NOAA Technical Memorandum, NMFS NE-218. National pone.0025248. Marine Fisheries Service, Woods Hole, MA. 247 pp. http://www King, J. R., and McFarlane, G. A. 2006. A framework for incorporating .nefsc.noaa.gov/nefsc/publications/. climate regime shifts into the management of marine resources. MacKenzie, B., and Ko¨ster, F. 2004. Fish production and climate: spratin Fisheries Management and Ecology, 13: 93–102. the Baltic Sea. Ecology, 85: 784–794. Ko¨ster, F., Mo¨llmann, C., Neuenfeldt, S., St. John, M., Plikshs, M., and MacKenzie, B. R., Meier, H. E. M., Lindegren, M., Neuenfeldt, S., Eero, Voss,R.2001.DevelopingBalticcodrecruitmentmodels. I.Resolving M., Blenckner, T., Tomczak, T., et al. 2012. Impact of climate change spatial andtemporal dynamics ofspawningstock andrecruitment for on fish populationdynamics inthe Baltic Sea: adynamical downscal- cod, herring, and sprat. Canadian Journal of Fisheries and Aquatic ing investigation. Ambio, 41: 626–636. Sciences, 58: 1516–1533. Marasco, R. J., Goodman, D., Grimes, C. B., Lawson, P. W., Punt, A. E., Lassen, H., Pedersen, S. A., Frost, H., and Hoff, A. 2013. Fishery manage- and Quin, T. J. 2007. Ecosystem-based fisheries management: some ment advice with ecosystem considerations. ICES Journal of Marine practical suggestions. Canadian Journal of Fisheries and Aquatic

Science, 70: 471–479. Sciences, 64: 928–939. Downloaded from Legendre, P., Dallot, S., and Legendre, L. 1985. Succession of species McLeod, K. L., and Leslie, H. M. (Eds) 2009. Ecosystem-Based within a community: chronological clustering, with applications Management for the Oceans. Island Press, Washington, DC. to marine and freshwater zooplankton. American Naturalist, 125: Meier, H. E. M., Andersson, H. C., Arheimer, B., Blenckner, T., 257–1298. Chubarenko, B., Donnelly, C., Eilola, K., et al. 2012. Comparing Legendre, P., and Legendre, L. 1998. Numerical ecology, 2nd edn. reconstructed past variations and future projections of the Baltic

Elsevier, Amsterdam, The Netherlands. Sea ecosystem—first results from multi-model ensemble simula- http://icesjms.oxfordjournals.org/ Leslie, H. M., and McLeod, K. L. 2007. Confronting the challenges of tions. Environmental Research Letters, 7: 034005. implementing marine ecosystem-based management. Frontiers in Metz, J. A. J., and Diekmann, O. 1986. Formulating models for struc- Ecology and the Environment, 5: 540–548. tured populations. In The Dynamics of Physiologically Structured Levin, P. S., Damon, I. A., and Samhouri, J. F. 2010. Developing mean- Populations (Amsterdam, 1983), Volume 68 of Lecture Notes in ingful marine ecosystem indicators in the face of a changing climate. Biomath, pp. 78–135. Springer, Berlin. Standford Journal of Law, Science and Policy, 2: 36–48. Millennium Ecosystem Assessment. 2005. Ecosystems and Human Levin, P. S., Fogarty, M. J., Murawski, S. A., and Fluharty, D. 2009. Wellbeing: Synthesis. Island Press, Washington, DC. Integrated ecosystem assessments: developing the scientific basis Mo¨llmann, C., and Diekmann, R. 2012. Marine ecosystem regime shifts for ecosystem-based management of the ocean. PLoS Biology, 7: induced by climate and overfishing—a review for the Northern e14. doi:0.1371/journal.pbio.1000014. hemisphere. Advances in Ecological Research, 47: 303–347. Levin, S. A., and Lubchenco, J. 2008. Resilience, robustness, and marine Mo¨llmann, C., Diekmann, R., Mu¨ller-Karulis, B., Kornilovs, G., Plikshs, by Howard Browman on July 2, 2014 ecosystem-based management. Bioscience, 58: 27–32. M., and Axe, P. 2009. Reorganization of a large marine ecosystem Lindegren, M., Blenckner, T., and Stenseth, N. C. 2012a. Nutrient reduc- due to atmospheric and anthropogenic pressure: a discontinuous tion and climate change cause a potential shift from pelagic to regime shift in the Central Baltic Sea. Global Change Biology, 15: benthic pathways in a eutrophic marine ecosystem. Global Change 1377–1393. Biology, 18: 3491–3503. Mo¨llmann, C., Mu?ller-Karulis, B., Kornilovs, G., and St. John, M. 2008. Lindegren, M., Dakos, V., Gro¨ger, J. P., Ga˚rdmark, A., Kornilovs, G., Effects ofclimateand overfishingon zooplankton dynamics and eco- Otto, S. A., and Mo¨llmann, C. 2012b. Early detection of ecosystem system structure: regime shifts, trophic cascade, and feedback coops regime shifts: a multiple method evaluation for management appli- in a simple ecosystem. ICES Journal of Marine Science, 65: 302–310. cation. PLoS One, 7: e38410. Nieminen, E., Lindroos, M., and Heikinheimo, O. 2012. Optimal bioe- Lindegren, M., Diekmann, R., and Mo¨llmann, C. 2010a. Regime shifts, conomic multispecies fisheries management: a Baltic Sea case study. resilience and recovery of a cod stock. Marine Ecology Progress Marine Resource Economics, 27: 115–136. Series, 402: 239–253. Niiranen, S., Blenckner, T., Hjerne, O., and Tomczak, M. T. 2012. Lindegren, M., Mo¨llmann, C., Nielsen, A., Brander, K., MacKenzie, B., Uncertainties in a Baltic Sea food-web model reveal challenges for and Stenseth, N. C. 2010b. Ecological forecasting under climate future projections. Ambio, 421: 613–625. change: the case of Baltic cod. Proceedings of the Royal Society of Ojaveer, H., and Eero, M. 2011. Methodological challenges in assessing London, Series B: Biological Sciences, 277: 2121–2130. the environmental status of a marine ecosystem: case study of the Lindegren, M., Mo¨llmann, C., Nielsen, A., and Stenseth, N. C. 2009. Baltic Sea. PLoS One, 6: e19231. doi:10.1371 /journal.pone.0019231 Preventing the collapse of the Baltic cod stock through an ecosystem- O¨ sterblom, H., Hansson, S., Larsson, U., Hjerne, O., Wulff, F., Elmgren, based management approach. Proceedings of the National Academy R., and Folke, C. 2007. Human-induced trophic cascades and eco- of Sciences of the USA, 106: 14722–14727. logical regime shifts in the Baltic Sea. Ecosystems, 10: 877–889. Lindegren, M., O¨ stman, O., and Ga˚rdmark, A. 2011. Interacting trophic Pikitch, E.K., Santora,C., Babcock, E. A., Bakun, A., Bonfil, R., Conover, forcing and the population dynamics of herring. Ecology, 92: D. O., Dayton, P., et al. 2004. Ecosystem-based fishery management. 1407–1413. Science, 305: 346–347. Link, J. S. 2005. Translating ecosystem indicators into decision criteria. Quaas, M. F., Froese, R., Herwartz, H., Requate, T., Schmidt, J. O., and ICES Journal of Marine Science, 62: 569–576. Voss, R. 2012. Fishing industry borrows from natural capital at high Link, J. S., Brodziak, J., Edwards, S., Overholtz, W., Mountain, D., Jossi, shadow interest rates. Ecological Economics, 82: 45–52. J., Smith, T., et al. 2002. Marine ecosystem assessment in a fisheries Quaas, M. F., Requate, T., Ruckes, K., Skonhoft, A., Vestergaard, N., and management context. Canadian Journal of Fisheries and Aquatic Voss, R. 2013. Incentives for optimal management of age-structured Sciences, 59: 1429–1440. fish populations. Resource and Energy Economics, 35: 113–134. Link, J. S., Gamble, R. J., and Fogarty, M. J. 2011a. An overview of the Rodionov, S. N. 2004. A sequential algorithm for testing climate regime NEFSC’s ecosystem modeling enterprise for the Northeast US shifts. Geophysical Research Letters, 31: L09204. Implementing EBFM 1197

Rotmans, J., and Dowlatabadi, H. 1997. Integrated assessment of Szuwalski, C. S., and Punt, A. E. 2013. Fisheries management for regime- climate change: evaluation of methods and strategies. In Human based ecosystems: a management strategy evaluation for the snow Choice and Climate Change: an International Social Science crab fishery in the eastern Bering Sea. ICES Journal of Marine Assessment. Ed. by S. Rayner, and E. Malone. Battelle Press, Science, 70: 955–967. Washington, USA. Tallis, H., Levin, P. S., Ruckelshaus, M., Lester, S. E., McLeod, K. L., Samhouri, J. F., Lester, S. E., Selig, E. R., Halpern, B. S., Fogarty, M. J., Fluharty, D. L., and Halpern, B. S. 2010. The many faces of Longo, C., and McLeod, K. E. 2012. Sea sick? Setting targets to ecosystem-based management: making the process work today in assess ocean health and ecosystem services. Ecosphere, 3: 1–18. real places. Marine Policy, 34: 340–348. Samhouri, J. F., and Levin, P. S. 2012. Linking land- and sea-based activ- Tomczak, M. T., Niiranen, S., Hjerne, O., and Blenckner, T. 2012. ities to risk in coastal ecosystems. Biological Conservation, 145: Ecosystem flow dynamics in the Baltic Proper—using a multitrophic 118–129. dataset as a basis for foodweb modelling. Ecological Modelling, 230: Samhouri, J. F., Levin, P. S., and Ainsworth, C. H. 2010. Identifying 123–147. thresholds for ecosystem-based management. PLoS One, 5: e8907. Ulrich, C., Wilson, D. C. K., Nielsen, J. R., Bastardiea, F., Reeves, S. A., Samhouri, J. F., Levin, P. S., and Harvey, C. J. 2009. Quantitative evalu- Andersen, B. S., and Eigaard, O. R. 2012. Challenges and opportun- ation of marine ecosystem indicator performance using food web ities for fleet- and me´tier-based approaches for fisheries manage- models. Ecosystems, 12: 1283–1298. ment under the European Common Fishery Policy. Ocean and Samhouri, J. F., Levin, P. S., James, A., Kershner, J., and Williams, G. Coastal Management, 70: 38–47. 2011. Using existing scientific capacity to sett argets for ecosystem- van der Sluijs, J. P. 2002. Integrated assessment. In Encyclopedia of based management: a Puget Sound case study. Marine Policy, 35: Downloaded from 508–518. Global Environmental Change, 4, pp. 250–253. Ed. by R. E. Munn. Wiley, Chichester. Scheffer,M., Bascompte, J.,Brock,W., Brovkin,V.,Carpenter, S., Dakos, V., Held, H., et al. 2009. Early-warning signals for critical transitions. van Leeuwen, A., De Roos, A. M., and Persson, L. 2008. How cod shapes Nature, 461: 53–59. its world. Journal of Sea Research, 60: 89–104. Scheffer, M., Carpenter, S., Foley, J. A., Folke, C., and Walker, B. 2001. van Leeuwen, A., Huss, M., Ga˚rdmark, A., Casini, M., Vitale, F., Hjelm,

Catastrophic shifts in ecosystems. Nature, 413: 591–596. J., Persson, L., et al. 2013. Predators with multiple ontogenetic niche http://icesjms.oxfordjournals.org/ Scheffer, M., Carpenter, S. R., Lenton, T. M., Bascompte, J., Brock, W., shifts have limited potential for population growth and top-down Dakos, V., van de Koppel, J., et al. 2012. Anticipating critical transi- control of their prey. American Naturalist, 182: 53–66. tions. Nature, 338: 344–348. Voss, R., Hinrichsen, H.-H., Quaas, M. F., Schmidt, J., and Tahvonen, O. Shepherd, J. G. 1999. Extended survivors analysis: an improved method 2011. Temperature change and Baltic sprat: from observations to for the analysis of catch-at-age data and abundance indices. ICES ecological-economic modeling. ICES Journal of Marine Science, Journal of Marine Science, 56: 584–591. 68: 1244–1256. Smith, A. D. M., Fulton, E. J., Hobday, A. J., Smith, D. C., and Shoulder, Weijerman, M., Lindeboom, H., and Zuur, A. 2005. Regime shifts in P. 2007. Scientific tools to support the practical implementation of marine ecosystems of the North Sea and Wadden Sea. Marine ecosystem-based fisheries management. ICES Journal of Marine Ecology Progress Series, 298: 21–39. Science, 64: 633–639. Weyant, J., Davidson, O., Dowlatabadi, H., Edmonds, J., Grubb, M., by Howard Browman on July 2, 2014 Sparholt, H. 1994. Fish species interactions in the Baltic Sea. Dana, 10: Richels, R., Rotmans, J., et al. 1996. Integrated assessment of 131–162. climate change: an overview and comparison of approaches and STECF. 2012. Scientific, Technical and Economic Committee for results. In Climate Change 1995–Economic and Social Fisheries (STECF). Multispecies Management Plans for the Baltic Dimensions of Climate Change. Contribution of Working Group (STECF-12-06). III to the Second Assessment Report of the Intergovernmental Suding, K., Gross, K., and Houseman, G. 2004. Alternative states and Panel on Climate Change (IPCC). Ed. by J. P. Bruce, H. Lee, and E. positive feedbacks in restoration ecology. Trends in Ecology and F. Haites. Cambridge University Press, Cambridge. Evolution, 19: 46–53.

Handling editor: Jason Link ICES Journal of Marine Science

ICES Journal of Marine Science (2014), 71(5), 1198–1204. doi:10.1093/icesjms/fst112

Contribution to the Themed Section: ‘Integrated assessments’ Original Article Guidance for implementation of integrated ecosystem

assessments: a US perspective Downloaded from

Phillip S. Levin1*, Christopher R. Kelble2, Rebecca L. Shuford3, Cameron Ainsworth4, Yvonne deReynier5, Rikki Dunsmore6, Michael J. Fogarty7, Kirstin Holsman8, Evan A. Howell9, Mark E. Monaco10, Stephanie A. Oakes3, and Francisco Werner11 http://icesjms.oxfordjournals.org/ 1National Marine Fisheries Service, Northwest Fisheries Science Center, 2725 Montlake Blvd. E., Seattle, WA 98115, USA 2NOAA Office of Oceanic and Atmospheric Research, Atlantic Oceanographic and Meteorological Lab, Miami, FL, USA 3National Marine Fisheries Service Office of Science and Technology, Silver Spring, MD, USA 4University of South Florida, St. Petersburg, FL, USA 5National Marine Fisheries Service, Northwest Regional Office, Seattle, WA, USA 6NOAA National Ocean Service, National Marine Sanctuary Program, Monterey Bay National Marine Sanctuary, Monterey, CA, USA 7National Marine Fisheries Service, Northeast Fisheries Science Center, Woodshole, MA, USA 8Joint Institute for the Study of the Atmosphere and Ocean, University of Washington, Seattle, WA, USA 9National Marine Fisheries Service, Pacific Islands Fisheries Science Center, Honolulu, HI, USA 10NOAA National Ocean Service, National Centers for Coastal and Ocean Science, Silver Spring, MD, USA by Howard Browman on July 2, 2014 11National Marine Fisheries Service, Southwest Fisheries Science Center, La Jolla, CA, USA *Corresponding author: tel: +206-860-3473; e-mail: [email protected] Levin, P. S., Kelble, C. R., Shuford, R., Ainsworth, C., deReynier, Y., Dunsmore, R., Fogarty, M. J., Holsman, K., Howell, E., Monaco, M., Oakes, S., and Werner, F. 2014. Guidance for implementation of integrated ecosystem assessments: a US perspective. – ICES Journal of Marine Science, 71: 1198–1204. Received 11 April 2013; accepted 7 June 2013; advance access publication 4 September 2013.

Ecosystem-based management (EBM) has emerged as a basic approach for managing human activities in marine ecosystems, with the aim of recovering and conserving marine ecosystems and the services they deliver. Integrated ecosystem assessments (IEAs) further the tran- sition of EBM from principle to practice by providing an efficient, transparent means of summarizing the status of ecosystem components, screening and prioritizing potential risks, and evaluating alternative management strategies againsta backdrop of environmental variability. In this paper, we draw upon lessons learned from the US National Oceanic and Atmospheric Administration’s IEA programme to outline steps required for IEA implementation. We provide an overview of the conceptual framework for IEAs, the practical constraints that shape the structure of individual IEAs, and the uses and outcomes of IEAs in support of EBM. Keywords: ecosystem-based management, ecosystem indicator, ecosystem risk, IEA, integrated ecosystem assessment, management strategy evaluation.

Introduction that humans are integral components of ecosystems that interact The need fora more holistic and integrated approach to the manage- with all other components in diverse, complex ways. Despite the ment of ocean resources is now widely appreciated in the scientific appreciation for integrated EBM, there remain few examples of suc- and management communities (U. S. Commission on Ocean cessful implementation (Cowling et al., 2008). This is in part, Policy, 2004; MEA, 2005; Murawski and Matlock, 2006; Agardy because we need tools to make the scientific principles of EBM et al., 2011). Ecosystem-based management (EBM) recognizes useful to resource managers (Arkema et al., 2006). A preeminent

Published by Oxford University Press on behalf of the International Council for the Exploration of the Seas 2013. This work is written by US Government employees and is in the public domain in the US. Guidance for implementation of IEAs 1199 need for implementing EBM is a framework to assess the status of status, trends and risk, an overall assessment of ecosystem status, an ecosystems relative to specific management goals and objectives evaluation of alternative management strategies, and monitoring and to evaluate the probable outcomes and trade-offs of alternative and evaluation (Figure 1). This process is iterative, allowing for management strategies. Integrated ecosystem assessments (IEAs) improved understanding and management of the coupled human- are intended to provide just such a framework (Levin et al., 2009). natural system over time. We use this framework to organize our IEAs provide a structured approach to ecosystem assessment and discussion below. evaluation that serves as an integrative counterpart to single-species and single-sector assessments now applied in resource manage- Step 1: defining EBM goals ment. IEAs are driven by clearly defined management objectives; conse- A number of IEA frameworks are currently being considered quently, the IEA approach purposefully begins by identifying prior- across the globe. For example, the ICES Working Group on the eco- ity management objectives to be addressed. This requires that system effects of fisheries activities, reviewed IEA approaches used in scientists, managers, and stakeholders work together to define the the Northeastern Atlantic, North Sea, Canada, and the United States broad vision and objectives of EBM, the spatial scale or scales of (ICES, 2010). These approaches differ in the degree to which pres- interest, and the ecosystem components and ecosystem threats sures are linked to ecosystem states, the degree of integration that will be included in the effort. Below we detail each one of across human and natural dimensions, and the regional consistency these elements.

that the frameworks promote. However, they all share a motivation Downloaded from to describe the status of the ecosystem relative to some desired state. Articulating the objectives to be addressed by the IEA process Here, we focus on NOAA’s IEA framework, while acknowledging The range of potential issues that could be addressed by an IEA is that IEAs are an emerging tool, which is being approached different- enormous, and any IEA effort must begin with transparent articula- ly by different countries and agencies. tion of the vision and objectives of the assessment. Sainsbury and The basic structural elements of NOAA’s IEA framework have Sumaila (2003) provide a useful framework for thinking about eco- been described elsewhere (Levin et al., 2008, 2009). Since NOAA system objectives. They define an ecosystem vision as a statement of http://icesjms.oxfordjournals.org/ first published its IEA framework, the NOAA IEA programme has the way “things should be”. For example, in 2005, Washington grown, and the IEA approach is now being implemented through- Governor Christine Gregoire’s ecosystem vision for Puget Sound, out the United States (www.noaa.gov/iea). In this document, we USA, was that it will “forever be a thriving natural system, with build on the initial IEA formulation of Levin et al. (2008) and clean marine and freshwaters, healthy and abundant native draw upon lessons learned from NOAA’s IEA programme to species, natural shorelines and places for public enjoyment and a outline practical steps required for IEA implementation. We vibrant economy that prospers in productive harmony with a provide an overview of the conceptual framework for IEAs, the prac- healthy sound.”(Puget Sound Partnership, 2006). Although such tical constraints that shape the structure of individual IEAs, and the vision statements are necessary, they are too vague to be practically uses and outcomes of IEAs in support of EBM. useful. Thus, ecosystemvisions need to be decomposed into concep- by Howard Browman on July 2, 2014 tual and operational objectives (O’Boyle and Jamieson, 2006). A A stepwise process for developing an IEA conceptual objective is a high-level statement of what is to be In this paper, we follow Levin et al. (2009) and define an IEA as a attained. As examples: (i) manage resources sustainable for formal synthesis and quantitative analysis of existing information human nutritional, economical, and social goals or (ii) protect on relevant natural and socio-economic factors in relation to speci- rare or fragile ecosystems, habitats, and species. An operational ob- fied ecosystem management objectives. IEAs are a stepwise process jective is an objective that has a direct and practical interpretation. consisting of goal and target setting, defining indicators, analysis of Formulating effective operational objectives requires thinking

Figure 1. A stepwise process for completing an IEA. An IEA begins with a scoping process, identifies appropriate indicators and reference levels, assesses risk, and evaluatesthe potential ofdifferentmanagement strategiesto alterecosystemstatus. Monitoringand evaluationoccur throughout the process, and the IEA cycle is repeated in an adaptive manner. 1200 P. S. Levin et al. carefully about the specific outcomes of EBM and how success or where they are needed most. The analytical efforts described failure will be measured and detected. below can be useful in identifying key threats. Additionally, The US National Ocean Policy provides many conceptual objec- formal or informal discussions with partners and stakeholders can tives that will require the development of operational objectives at na- be useful for focusing efforts. tional and regional scales. Fortunately, NOAA and other federal and The identification of threats during the initial stage of the IEA state agencies have a long history of working with Congress and man- process allows regional experts an opportunity to highlight ecosys- agementbodiestotranslatetheconceptualobjectivesoflegislationinto tem componentsthatare highlyexposed to human and naturalpres- operational objectives. For example, in coastal-zone management, sures. In essence, the identification ofthreats isthe base fromwhich a NOAA and the EPA provide broad guidance to the states, which formal risk analysis is launched (risk analysis is discussed below). then implement the objectives of the Coastal Zone Management Act and Clean Water Act through state-specific programmes. Putting it all together: conceptualize the ecosystem and the IEA Defining the spatial scale of an IEA Developing a common understanding of the context of the IEA, in- cluding the biophysical, socio-economic, and management systems The spatial scale of an IEA is largely determined by the management that affect the ability to achieve the vision of the IEA is the ultimate questions being addressed, and thus IEAs can vary in scale and step in scoping an IEA. This step builds upon previous work in extent. As a result, a key step in IEA scoping is to determine the which the objectives, focal ecosystem components, and threats are

scale of a particular IEA as it relates to defined management objec- Downloaded from identified. tives. Ecosystem boundaries are human constructs based on bio- Conceptual ecosystem models have proven useful for under- physical and political distributions; consequently, defining the standing ecosystems and can be helpful for synthesizing diverse sci- scale of an IEA is an important exercise that ideally considers bio- entific information. Conceptual models have a successful history in physical, human dimension, and management considerations. As resource management (Bowen and Riley, 2003), especially when a consequence, there is no “correct” scale at which to conduct an

scientists, resource managers, and stakeholders jointly develop http://icesjms.oxfordjournals.org/ IEA, but each implementation of the IEA framework will have a spe- models in workshop settings (Svarstad et al., 2008). Collaborative cified scale and extent appropriate to the management objectives model development assists in building consensus and helps move driving the assessment. beyond disagreements to a focus on how the ecosystem is structured IEAs will almost certainly have to contend with ecosystem pro- and functions. cesses and human activities that extend beyond the boundaries of an IEA. Oceanographic processes, terrestrial-based inputs, and climate change are just a few examples of this problem. This issue Step 2: defining ecosystem indicators and reference levels can be addressed by identifying drivers and pressures that occur Description of different approaches to select and evaluate within the IEA vs. those that emanate from outside the IEA, but ecosystem indicators must be considered. This delineation has the benefit of defining A critical step in the IEA process is to select indicators that capture by Howard Browman on July 2, 2014 pressures over which local management agencies have jurisdiction the key aspects of the focal ecosystem components identified in step and the externalities that affect the system, but are outside of local 1. Indicators are quantitative measures that serve as proxies for char- control. acterizing key attributes of biogeochemical and human systems (Heinz Center for Science and the Environment, 2008). Effective Identifying focal ecosystem components of an IEA indicators serve as the measures of the many ecosystem services Components of an ecosystem include everything: the physics, that concern policy-makers and stakeholders (Link, 2005) and are geology, chemistry, biology, and human dimensions. Focal ecosys- one of the primary contact points between policy and science. tem components are the major elements of an ecosystem that can Hundreds if not thousands of indicators have been proposed for be used to organize relevant information in a limited number of dis- use in EBM (e.g. Kershner et al., 2011; James et al., 2012). Rice and crete, but not necessarily independent categories. In the California Rochet (2005), Methratta and Link (2006), and Shin and Shannon Current IEA, a workshop with NOAA managers was used to (2010) outline valuable frameworks for sorting through volumin- develop a set of focal components that included: ecological structure ous lists and building a portfolio of informative indicators for and function, fisheries, protected species, habitat, and human com- EBM. These authors argue that indicators should be directly observ- munities (Levin and Schwing, 2011). Similarly, the focal compo- able and based on well-defined theory while also being understand- nents of the Kona ecosystem were identified as: coral reefs, able to the general public, cost effective to measure, supported by ornamental, recreational, and commercial fisheries, open ocean historical time-series, sensitive, and responsive to changes in ecosys- and coastal aquaculture programmes, tourism, shared-use areas in- tem state (and management efforts), and responsive to properties volving industry and natural resources (e.g. manta ray habitat and they are intended to measure. Levin et al. (2011) adapted these fra- commercial diving), critical cetacean habitat, and natural energy fa- meworks by soliciting and organizing expert judgement from the cilities. The Northeast Continental Shelf IEA defined focal ecosys- scientific community. Then, building on Rice and Rochet (2005), tem components as ones that require management interventions a team of scientists representing different agencies and areas of ex- to ensure their continued viability, such as species directly or indir- pertise worked through proposed indicators and determined how ectly affected by fishing activities and other anthropogenic impacts, well they met criteria related to public awareness, cost effectiveness, protected species, and human communities dependent on the eco- theoretical foundation, measurability, and availability of historical system for food, recreation, and other uses. data (Levin et al., 2011). This type of screening process is a necessary first step, but it cannot rigorously evaluate key indicator traits such Identifying key threats to ecosystem components as sensitivity, responsiveness, or specificity. Similar screening pro- Identification and prioritization of threats to achieving EBM objec- cesses have been adopted in other regions (Link et al., 2002; tives are important so that the IEA team can concentrate efforts Ecosystem Assessment Program, 2009; Shin et al., 2012). Guidance for implementation of IEAs 1201

Efforts in the Gulf of Mexico, Northeastern US, and California Step 3: risk analysis—impacts of natural perturbations Current are using ecosystem models to evaluate the diagnostic qual- and human activities on ecosystem status ities of indicators (Fulton et al., 2005; Samhouri et al., 2009; Levin Description of different approaches for conducting risk analysis and Schwing, 2011). These models are used to simulate varying Once ecosystem indicators and reference levels are selected, the next levels and different types of perturbations to the ecosystem. IEA step evaluates the risk to the indicators posed by human activ- The performance of each indicator in tracking perturbations ities and natural processes. The goal of these risk analyses is to quali- or ecosystem structure and function can be assessed using a tatively or quantitatively determine the likelihood that an ecosystem variety of statistical techniques (e.g. time-series analysis, cross- indicator will reach or remain in an undesirable state (i.e. breach a correlations, etc.). reference limit). Ecosystem modelling and analysis are important As with all phases of the IEA, indicators need to be regularly in determining incremental improvements in ecosystem indicators updated and revisited as more information becomes available, as en- in response to changes in human-induced pressures. Risk analysis vironmental conditions change, and as new threats emerge. Finally, must explicitly consider the inevitable uncertainties involved in because indicators are a key point of connectivity between science understanding and quantifying ecosystem dynamics and their posi- and policy, it is important to generate portfolios of indicators that tive and negative impacts on social systems. are not only scientifically rigorous but are also understandable Risk analysis must include pressures that occur on land (e.g. and salient to stakeholders (Levin et al., 2010). coastal development, agriculture, changing river flows, etc.), in

the air (e.g. weather, climate), and in the ocean itself (e.g. shipping, Downloaded from naval exercises, fishing, energy extraction, and physical and chem- Science to inform the establishment of ecosystem management ical conditions; Halpern et al., 2009). Thus, an ecosystem risk ana- reference levels lysis ideally requires an understanding of the distribution and Establishing a set of indicator values that reflect progress towards intensity of land-, air-, and sea-based pressures, as well as their specific management objectives is critical for successful EBM. impacts on ecosystem components. Additionally, because the cu- http://icesjms.oxfordjournals.org/ Such reference levels provide context for evaluating performance mulative effect of multiple stressors may not simply equal the sum and progress towards EBM goals. Reference levels can be diverse of the individual stressors’ effects, risk analysis should consider cu- and include both ecosystem state variables of interest (e.g. habitat mulative impacts (Crain et al., 2008; Ban et al., 2010; Kaplan et al., area, measures of diversity, etc.) as well as metrics of ecosystem pres- 2012). sures (e.g. shoreline development, nutrient, or contaminant input). There are a number of approaches to risk assessment; however, These levels can be drawn from the underlying properties of the most forms of risk assessment can be used within the ecological natural and human systems or they can be designated as part of risk assessment framework described by Hobday et al. (2011). the process of setting management goals. Establishing a reference Briefly, this is a hierarchical approach that moves from qualitative level is informed by science, but ultimately reference levels are set but comprehensive analyses (level 1) to a less comprehensive, semi- to achieve a desired policy outcome. quantitative analysis (level 2) and to a focused, fully quantitative by Howard Browman on July 2, 2014 For an IEA, reference levels serve a variety of purposes. The most analysis (level 3). obvious isthe role they play in the establishment of targets for restor- A level 1 analysis for each pressure qualitatively scores each ing and protecting the ecosystem (Sainsbury et al., 2000; Bottrill human activity or natural perturbation for its impact on the focal et al., 2008). A reference target is the level of an ecosystem indicator ecosystem components of the IEA. Those pressures receiving a that represents the desired state of the ecosystem. A reference limit, high impact score move onto level 2 analyses. As part of the on the other hand, is the attribute level that marks an ecosystem state ongoing engagement process described above, scientists and man- to be avoided (Samhouri et al., 2011). In fisheries management, for agers need to define what levels of impact score constitute “high” instance, “optimum yield” is a reference target and “overfished” is a or “low” for this process. Thus, a level 1 analysis separates out reference limit. Because some ecosystem indicators respond slowly those activity-component pairs that warrant further investigation to management action or natural drivers, there is a need for bench- from those that are given the all clear. marks or intermediate indicator values that demonstrate progress A level 2 analysis considers the exposure of an ecosystem compo- towards those levels. nent to a pressure and the sensitivity of the component to that pres- Ecosystem-based reference levels are needed to refine the limits sure. This framework has proven useful for conducting of acceptable resource use, plan for future climate- and human- semi-quantitative risk analysis in ecosystem-based fisheries man- induced changes to ecosystem services, and inform conservation agement (Stobutzki et al., 2002; Hobday et al., 2006, 2011; Patrick and recovery plans for sensitive species and habitats. Scientific ana- et al., 2010) and has been useful for IEAs, as well (e.g. Samhouri lysis can contribute to the setting of reference levels in several ways. and Levin, 2012). In this approach, risk to an indicator can be For example, perturbations to ecosystem models have been used to defined as the Euclidean distance of the indicator from the origin explore non-linearities between pressures and ecosystem state (e.g. in a space defined by exposure and sensitivity to particular human Samhouri et al., 2010). To complement these modelling approaches, activities. Exposure can be estimated as a function of spatial and statistical models may prove useful for investigating the relationship temporal overlap of activities, and sensitivity can be estimated by between ecosystem state variables and natural and anthropogenic the degree to which life-history attributes or behaviour affects an pressures. No matter what approaches are used, the goal of this indicator’s ability to resist or recover from exposure to a human ac- step is to identify thresholds and inflection points that may tivities. If this level 2 analysis determines the impact of an activity on provide a basis for the identification of reference levels. a species or other ecological component is high and there are no Importantly, as we noted at the beginning of this section, the iden- planned management interventions to remove it, the reasons are tification of reference levels is a societal choice and the role of IEAs is documented and the assessment moves to level 3. to inform that choice. 1202 P. S. Levin et al.

The level 3 analysis takes a quantitative approach such as is used opportunity for stakeholder involvement (e.g. workshops) and is in stock assessments and population viability analyses. A number of greatlystrengthened by discussion regarding management strategies modelling approaches lend themselves to level 3 analyses, but all are to be evaluated, target species to include, incorporation of long- dependent on the amount and quality of the available data. What term monitoring data, comparison of model outcomes (including model is used in a level 3 analysis depends on the nature of the socio-economic impacts), and discussion of management trade- data and other available information. offs. (v) The MSE process often identifies data and knowledge gaps, which in turn can be used to inform future research. Step 4: evaluation of management strategies for protection or restoration of ecosystem status Step 5: monitoring and evaluation Approaches for evaluating management strategies Monitoring and evaluation of chosen indicators and management The next step in the IEA process uses simulation and analytical or strategies is an integral part of the IEA process. Monitoring and conceptual modelling to evaluate the potential of different manage- evaluation is necessary to determine whether management strat- ment strategies to influence the status of natural and human system egies improve ecosystem services and sustainability and quantifies indicators and to achieve our stated ecosystem objectives. the trade-offs that have occurred since implementation of the man- Scenario analysis generates the multiple alternative descriptions agement strategy. of potential outcomes, including processes of change, thresholds, and uncertainties (Alcamo, 2008).Scenarios explorealternativeper- Monitoring Downloaded from spectives about underlying system processes and can illuminate key At its core, monitoring is straightforward; it is the collection of issues, by using a consistent set of assumptions about the system biotic, abiotic, and human dimension data. In the context of IEAs, state to broaden perspectives (Raskin, 2005; Refsgaard et al., monitoring is the systematic collection of data to reliably answer 2007). They generate alternative, internally consistent, logical clearly articulated management questions (Katz, 2013). For IEA descriptions of the future. Scenarios can be qualitative, in which indicators, monitoring must directly address the operational objec- http://icesjms.oxfordjournals.org/ “storylines” are developed, or quantitative, in which the outcomes tives developed as part of the IEA process. While apparently simple, of numerical models are explored (Refsgaard et al., 2007). monitoring is costly and subject to the changing priorities of Formal management strategy evaluation (MSE) is a modelling funding agencies. Thus, successful monitoring depends on develop- approach that can be used to analyse posited scenarios. MSE is ing efficient sampling programmes that allow a cost-effective deter- widely used in the management of protected species (e.g. marine mination of the state of the ecosystem and the effectiveness of mammals) and fisheries and is beginning to see use in the EBM management actions. (Sainsbury et al., 2000; Fulton et al., 2007). By evaluating a range In general, there are two types of monitoring that are particularly of management scenarios using multiple performance indicators important to IEAs. Trend monitoring is a systematic series of obser- (and potentially multiple operating models), formal MSE can be vations over time to detecting change in the state of an ecosystem used to test the utility of modifying indicators, management component (MacDonald et al., 1991). Typically, the observations by Howard Browman on July 2, 2014 targets, assessments, monitoring plans, management strategies, are not taken with the aim of evaluating management actions, al- and decision rules.Importantly, theobjective of MSE is not optimal- though such data may prove useful in this context as well. Trend ity. Rather, MSE is used to screen out poorly performing manage- monitoring focuses on the indicators of ecosystem state developed ment strategies, identify trade-offs among objectives, and in step 2 of the IEA. Effectiveness monitoring is used to evaluate distinguish approaches robust to various types of uncertainty. whether specific management actions had the desired effect. Increasingly, MSEs are being used to evaluate interactions Effectiveness monitoring focuses on changes in threats identified between separate management tactics and interactions between in the scoping phase of the IEA and links threat reduction to management, ecosystem processes, and large-scale drivers like changes in the status of key ecosystem components. Thus, effective- climate change. For example, recent applications have attempted ness monitoring requires the observations of threats as well as the to the minimize risk of uncertainty on target and bycatch species ecosystem component(s) targeted by the management action. (e.g. Stram and Ianelli, 2009), ESA at-risk species, critical habitats, Katz (2013) notes that the key elements of monitoring are deter- and human communities under various short- and long-term mining “what, where,(and sometimes) how” to measure the system. climate scenarios (Hollowed et al., 2009) He also notes that in the best cases, a monitoring programme also MSE incorporates a number of important features (Sainsbury confronts the issue of how well one wants to know the answer. et al., 2000) that make it an ideal supporting process for IEAs. (i) Successfully addressing these elements defines the indicators Simulations are performed in the operating model on the (what and how) and the sampling design (where, when, and how managed system as a whole. For management towards ecological well). Importantly, monitoring includes not only the measurements objectives, we require the explicit representation of the ecological of the biophysical environment but also includes social and eco- system; similarly, for economic objectives, we require explicit nomic systems (McLeod and Leslie, 2009). ecological-economic linkages. (ii) Performance metrics are evalu- ated quantitatively in a simulation framework utilizing the indica- Evaluation tors developed earlier in the IEA process. (iii) A variety of models Evaluation of ecosystem status involves using data generated from or submodels may be used in the evaluation process; thus, scenarios trend monitoring to assess the condition or status of particular eco- may explore alternative hypotheses on ecosystem functioning or system components (Stem et al., 2005). In contrast to status evalu- may use a consistent set of assumptions about the system state to il- ation, evaluations for measuring management effectiveness are luminate key issues and broaden perspectives (Raskin, 2005; necessarily linked to discrete management actions and obviously Refsgaard et al., 2007). (iv) The MSE process allows ample are directly linked to effectiveness monitoring. Stem et al. (2005) Guidance for implementation of IEAs 1203 describe two types of effectiveness evaluations. Impact evaluations Crain, C. M., Kroeker, K., and Halpern, B. S. 2008. Interactive and cu- are generally one-time assessments frequently performed at the con- mulative effects of multiple human stressors in marine systems. clusion of a management project. The goal of impact evaluations is Ecology Letters, 11: 1304–1315. to determine how well a particular project performed. A second Ecosystem Assessment Program. 2009. Ecosystem Status Report for the form of effectiveness evaluation is adaptive management—an itera- Northeast U.S. Continental Shelf Large Marine Ecosystem. US Department of Commerce, Northeast Fisheries Science Center tive process that integrates the design of management strategies and Reference Document, 09-11. 34 p. monitoring to systematically evaluate management actions Fulton, E. A., Smith, A. D. M., and Punt, A. E. 2005. Which ecological (Walters, 1986). Successful IEAs will evaluate the effectiveness of indicators can robustly detect effects of fishing? ICES Journal of management actions and provide information to managers so Marine Science, 62: 540–551. they can adjust actions, as needed. Fulton, E. A., Smith, A. D. M., and Smith, D. C. 2007. Alternative man- agement strategies for Southeast Australian Commonwealth Fisheries: Stage 2: Quantitative management strategy evaluation. Conclusions Commonwealth Scientific and Industrial Research Organisation After decades of struggle over what EBM means, whether it is pos- (CSIRO), Hobart, Tasmania, Australia. sible, and if it is needed, we have arrived at a time when comprehen- Haller, J. 1976. The Blue Strawberry Cookbook: Cooking (Brilliantly) sive EBM is emerging and is supporting efforts to recover and Without Recipes. Harvard Commons Press, Cambridge, MA. conserve marine ecosystems (e.g. Lubchenco and Sutley, 2010). To Halpern, B. S., Kappel, C. V., Selkoe, K. A., Micheli, F., Ebert, C. M., further the transition of EBM from principles to practice, we must Kontgis, C., Crain, C. M., et al. 2009. Mapping cumulative human Downloaded from develop approaches that provide an efficient, transparent means impacts to California Current marine ecosystems. Conservation Letters, 2: 138–148. of summarizing the status of ecosystem components, screening Heinz Center for Science and the Environment. 2008. The State of the and prioritizing potential risks, and evaluating alternative manage- Nation’s Ecosystems 2008: Measuring the Lands, Waters, and ment strategies against a backdrop of environmental variability. Living Resources of the United States. Island Press, Washington DC.

IEAs provide a means to do just this. Importantly, while the Hobday, A. J., Smith, A., Webb, H., Daley, R., Wayte, S., Bulman, C., http://icesjms.oxfordjournals.org/ science of IEAs is progressing, Samhouri et al. (in review) note Dowdney, J., et al. 2006. Ecological Risk Assessment for the Effects that operationalizing IEAs in management will require additional of Fishing: Methodology. Report R04/1072 for the Australian work and provide insight into how this might be facilitated. Fisheries Management Authority, Canberra. In this paper, weopted out of producing an IEA “cookbook” with Hobday, A. J., Smith, A. D. M., Stobutzki, I. C., Bulman, C., Daley, R., simple recipes for accomplishing each IEA step. Ecosystems, includ- Dambacher, J. M., Deng, R. A., et al. 2011. Ecological risk assessment for the effects of fishing. Fisheries Research, 108: 372–384. ing their associated institutions, are too variable for such a prescrip- tive treatment. Instead, we provide a cookbook without recipes (cf. Hollowed, A. B., Bond, N. A., Wilderbuer, T. K., Stockhausen, W. T., A’mar, Z. T., Beamish, R. J., Overland, J. E., et al. 2009. A framework Haller, 1976), with the goal of providing basic guidance that can be for modelling fish and shellfish responses to future climate change. tailored to the specific management needs, scientific capacity, and ICES Journal of Marine Science, 66: 1584–1594. governance structures in any region. Like all science-management ICES. 2010. Report of the Working Group on Ecosystem Effects of by Howard Browman on July 2, 2014 processes, we expect that as IEAs propagate and the scientific and Fishing Activities (WGECO), 7–14 April 2010, Copenhagen, management communities gain experience with them, a collection Denmark. ICES Document CM 2010/ACOM: 23. of best IEA practices will emerge. Currently, however, we look James, C. A., Kershner, J., Samhouri, J., O’Neill, S., and Levin, P. S. 2012. forward to a period of creativity and innovation that will rapidly A methodology for evaluating and ranking water quantity indicators advance the rigor and utility of this try. in support of ecosystem-based management. Environmental Management, 2012: 1–17. Kaplan, I. C., Gray, I. A., and Levin, P. S. 2012. Cumulative impacts of References fisheries in the California Current. Fish and Fisheries, in press. Agardy, M., Davis, J., Sherwood, K., and Vestergaard, O. 2011. Taking Katz, S. L. 2013. Monitoring endangeredspecies. InExtinction. Ed. by N. steps toward marine and coastal ecosystem-based management-an MacLeod, D. Archibald, and P. S. Levin. Gale Publishing, introductory guide. UNEP Regional Seas Reports and Studies, 189. Farmington Hills, MI. 68 pp. Kershner, J., Samhouri, J. F., James, C. A., and Levin, P. S. 2011. Selecting Alcamo, J. 2008. Environmental Futures: the Practice of Environmental indicator portfolios for marine species and food webs: a Puget Sound Scenario Analysis. Elsevier, Amsterdam. case study. PLoS One, 6: e25248. Arkema, K. K., Abramson, S. C., and Dewsbury, B. M. 2006. Marine Levin, P. S., Damon, M., and Samhouri, J. F. 2010. Developing meaning- ecosystem-based management: from characterization to implemen- ful marine ecosystem indicators in the face of a changing climate. tation. Frontiers in Ecology and the Environment, 4: 525–532. Stanford Journal of Law Science and Policy, 2: 37–48. Ban, N. C., Alidina, H. M., and Ardron, J. A. 2010. Cumulative impact Levin, P. S., Fogarty, M., Matlock, G., and Ernst, M. 2008. Integrated mapping: advances, relevance and limitations to marine manage- ecosystem assessments. US Department of Commerce. NOAA ment and conservation, using Canada’s Pacific waters as a case Technical Memorandum, NMFS-NWFSC 92. 20 pp. study. Marine Policy, 34: 876–886. Levin, P. S., Fogarty, M. J., Murawski, S. A., and Fluharty, D. 2009. Bottrill, M. C., Joseph, L. N., Carwardine, J., Bode, M., Cook, C., Game, Integrated ecosystem assessments: developing the scientific basis E. T., Grantham, H., et al. 2008. Is conservation triage just smart de- for ecosystem-based management of the ocean. PLoS Biology, 7: cision making? Trends in Ecology and Evolution, 23: 649–654. e1000014 doi:1000010.1001371/journal.pbio.1000014. Bowen, R. E., and Riley, C. 2003. Socio-economic indicators and inte- Levin, P. S., James, C. A., Kershner, J., O’Neill, S., Francis, T. B., grated coastal management. Ocean and Coastal Management, 46: Samhouri, J. F., Harvey, C. J., et al. 2011. The Puget Sound 299–312. Ecosystem: what is our desired outcome and how do we measure Cowling, R. M., Egoh, B., Knight, A. T., O’Farrell, P. J., Reyers, B., progress along the way? Accessed from pugetsoundscienceupdate.- Rouget, M., Roux, D. J., et al. 2008. An operational model for main- com. Puget Sound Partnership, Tacoma, WA. streaming ecosystem services for implementation. Proceedings of Levin, P. S., and Schwing, F. 2011. Technical background for an IEA of the National Academy of Sciences of the USA, 105: 9483–9488. the California Current: ecosystem health, salmon, groundfish and 1204 P. S. Levin et al.

green sturgeon. NOAA Technical Memorandum, NMFS-NWFSC-109. Sainsbury, K. J., Punt, A. E., and Smith, A. D. M. 2000. Design of oper- 330 pp. ational management strategies forachievingfisheryecosystem objec- Link, J. S. 2005. Translating ecosystem indicators into decision criteria. tives. ICES Journal of Marine Science, 57: 731–741. ICES Journal of Marine Science, 62: 569. Samhouri, J. F., Haupt, A., Levin, P. S., Link, J., and Shuford, R. In Link, J. S., Brodziak, J. K. T., Edwards, S. F., Overholtz, W. J., Mountain, review. Lessons learned from developing integrated ecosystem D., Jossi, J. W., Smith, T. D., et al. 2002. Marine ecosystem assessment assessments to inform marine ecosystem-based management in in a fisheries management context. Canadian Journal of Fisheries the U.S. ICES Journal of Marine Science. and Aquatic Sciences, 59: 1429–1440. Samhouri, J. F., and Levin, P. S. 2012. Linking land-and sea-based activ- Lubchenco, J., and Sutley, N. 2010. Proposed U.S. Policy for Ocean, ities to risk in coastal ecosystems. Biological Conservation, 145: Coast, and Great Lakes Stewardship. Science, 328: 1485–1486. 118–129. MacDonald, L. H., Smart, A. W., and Wissmar, R. C. 1991. Monitoring Samhouri, J. F., Levin, P. S., and Ainsworth, C. H. 2010. Identifying guidelines to evaluate effects of forestry activities on streams in the thresholds for ecosystem-based management. PLoS One, 5: e8907. Pacific Northwest and Alaska. Region 10, US EPA. https:// doi:10.1371/journal.pone.0008907 digitallibwashingtonedu/dspace/handle/1773/17068. Samhouri, J. F., Levin, P. S., Andrew James, C., Kershner, J., and McLeod, K., and Leslie, H. 2009. Ecosystem-Based Management for the Williams, G. 2011. Using existing scientific capacity to set targets Oceans. Cambridge University Press. for ecosystem-based management: a Puget Sound case study. Methratta, E. T., and Link, J. S. 2006. Evaluation of quantitative indica- Marine Policy, 35: 508–518. tors for marine fish communities. Ecological Indicators, 6: 575–588.

Samhouri, J. F., Levin, P. S., and Harvey, C. J. 2009. Quantitative evalu- Downloaded from Millennium Ecosystem Assessment (MEA). 2005. Ecosystems and ation of marine ecosystem indicator performance using food web Human Well-being: Synthesis. Island Press. models. Ecosystems, 12: 1283–1298. Murawski, S., and Matlock, G. 2006. Ecosystem science capabilities required to support NOAA’s mission in the year 2020. NOAA Shin, Y. J., Bundy, A., Shannon, L. J., Blanchard, J. L., Chuenpagdee, R., Technical Memorandum, NMFS-F/SPO-74, Silver Spring, MD. Coll, M., Knight, B., et al. 2012. Global in scope and regionally rich: O’Boyle, R., and Jamieson, G. 2006. Observations on the implementa- an IndiSeas workshop helps shape the future of marine ecosystem tion of ecosystem-based management: experiences on Canada’s indicators. Reviews in Fish Biology and Fisheries, pp. 1–11. http://icesjms.oxfordjournals.org/ east and west coasts. Fisheries Research, 79: 1–12. Shin, Y. J., and Shannon, L. J. 2010. Using indicators for evaluating, Patrick, W. S., Spencer, P., Link, J., Cope, J., Field, J., Kobayashi, D., comparing, and communicating the ecological status of exploited Lawson, P., et al. 2010. Using productivity and susceptibility indices marine ecosystems. 1. The IndiSeas project. ICES Journal of to assess the vulnerability of United States fish stocks to overfishing. Marine Science, 67: 686–691. Fishery Bulletin, 108: 305–322. Stem, C., Margoluis, R., Salafsky, N., and Brown, M. 2005. Monitoring Puget Sound Partnership. 2006. Sound health, sound future: protecting and evaluation in conservation: a review of trends and approaches. and restoring Puget Sound. Olympia, Washington. Conservation Biology, 19: 295–309. Raskin, P. D. 2005. Global scenarios: background review for the Stobutzki, I., Miller, M., and Brewer, D. 2002. Sustainability of fishery Millennium Ecosystem Assessment. Ecosystems, 8: 133–142. bycatch: a process for assessing highly diverse and numerous Refsgaard, J. C., van der Sluijs, J. P., Hojberg, A. L., and Vanrolleghem, bycatch. Environmental Conservation, 28: 167–181. by Howard Browman on July 2, 2014 P. A. 2007. Uncertainty in the environmental modelling process - A Stram, D., and Ianelli, J. 2009. Eastern Bering Sea pollock trawl fisheries: framework and guidance. Environmental Modelling and Software, variation in salmon bycatch over time and space. American Fisheries 22: 1543–1556. Society Symposium, 70: 827–850. Rice, J. C., and Rochet, M. J. 2005. A framework for selecting a suite of Svarstad, H., Petersen, L. K., Rothman, D., Siepel, H., and Wa¨tzold, F. indicators for fisheries management. ICES Journal of Marine 2008. Discursive biases of the environmental research framework Science, 62: 516–527. DPSIR. Land Use Policy, 25: 116–125. Sainsbury, K., and Sumaila, R. 2003. Incorporating ecosystem objective U. S. Commission on Ocean Policy. 2004. An ocean blueprint for the is into management of sustainable marine fisheries, including best 21st century. Final Report to the Present and Congress. practice reference points and use of marine protected areas. In Responsible Fisheries in the Marine Ecosystems. Ed. by M. Walters, C. J. 1986. Adaptive Management of Natural Resources. Sinclair, and G. Valdimarsson. FAO, Rome. Mac-Millan, New York.

Handling editor: Sarah Kraak ICES Journal of Marine Science

ICES Journal of Marine Science (2014), 71(5), 1205–1215. doi:10.1093/icesjms/fst141

Contribution to the Themed Section: ‘Integrated assessments’ Original Article Lessons learned from developing integrated ecosystem

assessments to inform marine ecosystem-based management Downloaded from in the USA

Jameal F. Samhouri1*, Alison J. Haupt2, Phillip S. Levin1, Jason S. Link3, and Rebecca Shuford4

1Conservation Biology Division, Northwest Fisheries Science Center, National Marine Fisheries Service, National Oceanic and Atmospheric http://icesjms.oxfordjournals.org/ Administration, 2725 Montlake Boulevard East, Seattle, WA 98112, USA 2West Coast Governors Alliance, Northwest Fisheries Science Center, National Marine Fisheries Service, National Oceanic and Atmospheric Administration, 2725 Montlake Boulevard East, Seattle, WA 98112, USA 3National Marine Fisheries Service, National Oceanic and Atmospheric Administration, 166 Water St., Woods Hole, MA 02543, USA 4Marine Ecosystems Division, Office of Science and Technology, National Marine Fisheries Service, National Oceanic and Atmospheric Administration, Silver Spring, MD 20910, USA *Corresponding author: e-mail: [email protected] Samhouri, J. F., Haupt, A. J., Levin, P. S., Link, J. S., and Shuford, R. 2014. Lessons learned from developing integrated ecosystem assessments to inform marine ecosystem-based management in the USA. – ICES Journal of Marine Science, 71: 1205–1215. by Howard Browman on July 2, 2014 Received 2 May 2013; accepted 1 August 2013; advance access publication 7 September 2013.

Borne out of a collective movement towards ecosystem-based management (EBM), multispecies and multi-sector scientific assessments of the ocean are emerging around the world. In the USA, integrated ecosystem assessments (IEAs) were formally defined 5 years ago to serve as a scientific foundation for marine EBM. As outlined by the US National Oceanic Atmospheric Administration in 2008, an IEA is a cyclical process consisting of setting goals and targets, defining indicators, analysing status, trends, and risk, and evaluating alternative potential future management and environmental scenarios to enhance information needed for effective EBM. These steps should be hierarchical, iterative, non-prescriptive about technical implementation, and adaptable to existing information for any ecosystem. Despite these strengths and some initial successes, IEAs and EBM have yet to be fully realized in the USA. We propose eight tenets that can be adopted by scientists, policy-makers, and managers to enhance the use of IEAs in implementing EBM. These tenets include (i) engage with stakeholders, managers, and policy-makers early, often, and continually; (ii) conduct rigorous human dimensions research; (iii) rec- ognize the importance of transparentlyselecting indicators; (iv) set ecosystem targets to create a system of EBM accountability; (v) establish aformal mechanism(s) for the review of IEA science; (vi) serve current management needs, but not at the expense of more integrative ocean management; (vii) provide a venue for EBM decision-making that takes full advantage of IEA products; and (viii) embrace realistic expecta- tions about IEA science and its implementation. These tenets are framed in a way that builds on domestic and international experiences with ocean management. With patience, persistence, political will, funding, and augmented capacity, IEAs will provide a general approach for allowing progressive science to lead conventional ocean management to new waters. Keywords: ecosystem-based management, indicator, integrated ecosystem assessment, marine policy, risk, sustainable marine management and conservation.

Introduction are affected by people’s actions and influences (Halpern et al., People are unquestionably dependent upon the services that marine 2012). Marine resources face a well known and growing list of pres- ecosystems provide (Guerry et al., 2012). In turn, ocean conditions sures and demands—from long-standing activities like fisheries and

Published by Oxford University Press on behalf of the International Council for the Exploration of the Seas 2013. This work is written by (a) US Government employee(s) and is in the public domain in the US. 1206 J. F. Samhouri et al. extraction of fossil fuels, to newer uses such as wind and wave energy coastal resource management. Indeed, the premise of EBM is now (Halpern et al., 2008). Because of existing governance structures, codified in the US National Ocean Policy (Obama, 2010; USNOC, ocean managers have historically tackled the most pressing activities, 2013). In response to recommendations of these reports suggesting and addressed them one-by-one (Kareiva and Marvier, 2011). the need for technical information to support EBM decisions, However, in the face of emerging uses and increasing demands NOAA’s Science Advisory Board commissioned an external from existing activities, this reactive, sectoralized approach to ocean management has been called into question (Levin and Lubchenco, 2008). Protected areas, for instance, may achieve some conservation objectives within their boundaries, but without coordinated management outside of their boundaries, dis- placed fishing effort may result in the degradation of an ecosystem as a whole (Hilborn et al., 2006; Jennings, 2009). Similarly, reductions in chemicalpollution and improvements inwaterquality have facili- tated the recovery of endangered coastal species such as the bald eagle, but recovery of eagles has led to the decline of some other seabird species and overall diversity (Hipfner et al., 2012). As

these examples illustrate, well intentioned but segmented manage- Downloaded from ment has the potential to interfere with other ecosystem goals and perhaps even degrade the overall state of marine environments. A potential solution is a more systematic, integrated approach known as ecosystem-based management (EBM; McLeod and Leslie, 2009). The transition to EBM will not be easy, fast, or simple. It is likely to be gradual and iterative, building on existing single species http://icesjms.oxfordjournals.org/ and single-sector governance where possible and characterized by trial-and-error learning where such building is not possible. The evo- lution of comprehensive management is likely to benefit from inte- grated science—a rarity in many regions. In the USA, the National OceanicandAtmosphericAdministration(NOAA)hasbeendevelop- Figure 1. Conceptual schematic describing the cyclical, iterative ing the integrated ecosystem assessment (IEA) as a means to conduct nature of IEAs at the US NOAA. This figure is an update of the and deliver integrated, cross-sectoral science to support EBM (Levin characterization of the approach depicted in Levin et al. (2009). Define et al., 2009). IEAs transparently identify socio-economic and bio- EBM goals and targets: the IEA process involves managerengagement to physical attributes that maintain ecosystem structure and function, identify critical ecosystem management goals and targets to be by Howard Browman on July 2, 2014 assess human activities and their interdependence with the natural addressed through and informed by the IEA approach. The rest of the ecosystem, and evaluate management alternatives that will maintain process is driven by these defined objectives, which can be informed by or improve the coupled social-ecological system (Figure 1). available scientific syntheses during the development of the first IEA in any region. Engagement is continual throughout the entire IEA process. While IEAs are a major step forward for developing the science to Develop indicators: indicators represent key components in an support EBM (Foley et al., 2013), both in the USA and around the ecosystem and allow change to be measured. They provide the basis to world (Smith et al., 2007; European Commission, 2008; assess the status and trends in the condition of the ecosystem or of an Katsanevakis et al., 2011; ICES, 2011, 2013a, b, c, Mo¨llmann, in element within the system. Indicators are essential to all subsequent press), widespread confusion exists about what, exactly, defines an steps in the IEA approach. Assess ecosystem: ecosystem indicator data IEA, and how to use an IEA so that it informs on-the-water are assessed together to evaluate overall ecosystem status and trends actions. A companion paper by Levin et al. (this volume) and previ- relative to ecosystem management goals and targets. Individual ous descriptions (Levin et al., 2008, 2009) delineate one perspective indicators are assessed to determine the underlying cause for the of what an IEA is. In this article, we explore the question of how such observed ecosystem status and trends. Analyse uncertainty and risk: IEAs can be used in US ocean management now (through several ecosystem analyses and models evaluate risk to the indicators (and thus the ecosystem) posed by human activities and natural processes. These proposed tenets), discuss reasonable expectations of IEAs in the methods incorporate the degree of uncertainty in each indicator’s context of implementation of EBM, and offer some suggestions response to pressures. This determines incremental improvements or about what needs to happen next to make better use of them. Our declines in ecosystem indicators in response to changes in drivers and perspective on these topics is based on our collective experience pressures and to predict the potential that an indicator will reach or attempting to inform EBM via IEAs in the USA, largely learning remain in an undesirable state. Evaluate strategies: management by trial and error, while doing our best as scientists to create a strategy evaluation (MSE) is useful to help resource managers consider solid technical basis from which to conduct EBM. We hope this trade-offs and potential for success in reaching targets. It relies on mini-review of what has worked well, what has not worked well, ecosystem modelling and empirical analysis to evaluate the potential and what may work well in the USAwill be of interest to those devel- for different management strategies to influence the status of natural oping integrated scientific assessments, andtrying to avoid potential and human system indicators and to achieve stated ecosystem objectives. Taking, monitoring, and assessing action: Based on the MSE, impediments to their use, for ocean management around the world. an action is selected and implemented. Monitoring of indicators is important to determine if the action is successful; if yes, the status, Brief history of IEAs in the USA trends, and risk to the indicators continue to be analysed for High-profile reports by the US Oceans Commission (USCOP,2004) incremental change. Otherwise as part of adaptive management, the and the Pew Oceans Commission (POC, 2003) stressed the import- outcomes need to be assessed and evaluated to refine goals, targets, ance of incorporating ecosystem principles into US ocean and and/or indicators. Lessons learned from developing IEAs in the USA 1207 review of NOAA’s ecosystem research. The resulting report comprehensive fashion (e.g. regional planning bodies and regional (Fluharty et al., 2006) recommended that NOAA prioritize the pro- ocean partnerships); and (ii) marine resource managers who are duction of IEAs, although the report did not precisely define an IEA. required by law to incorporate ecosystem considerations into their As a result, NOAA developed the US IEA framework (Levin et al., decisions but need tools to do so in a useful way. 2009) as a cyclical process consisting of setting goals and targets, de- As an example, in 2008, the West Coast Governors Alliance fining indicators, analysing status, trends, and risk, and evaluating (WCGA) created the WCGA IEA Action Coordination Team to alternative potential future management and environmental scen- address the following issue (their Action 3.2): Assess physical, biologic- arios (Figure 1). The NOAA approach is a direct descendant of al, chemical, and socio-economic factors in ecosystem health across the approaches advocated by Caddy (1999), Sainsbury et al. (2000), West Coast to establish standards and indicators for ocean health.The and Smith et al. (2007), but it is related to decades of EBM thinking WCGA IEA team created a plan to divide the dynamic and heteroge- around the globe (reviewed by Rosenberg et al., 2009; also see neous CaliforniaCurrentLargeMarineEcosystemintosix subregions European Commission, 2008; Katsanevakis et al., 2011; ICES, to allow for local customization. For example, the Oregon Coast 2011, 2013a, b, c; Mo¨llmann, in press). sub-IEA may focus on interactions of fisheries and renewable ocean Underpinning the distinct IEA steps (Figure 1) are four integra- energy projects, whereas the Southern California sub-IEA may be tive ideas. First, an IEA is hierarchical: all indicators, status assess- more concerned with water quality issues that can lead to frequent ments, risk analyses, and scenario evaluations must map back to beach closures in popular recreation areas.

the established ecosystem management goals. Second, IEAs must Although IEAs area nascent enterprise in the USA, therehas been Downloaded from be considered an iterative process. This idea complicates an other- ample opportunity for growth and progress. While no region in the wise hierarchical structure during the development of the first IEA USA can yet boast a fully mature IEA, the IEA approach must none- for a region. The process of setting ecosystem management goals theless look forward to ensure that it is actually used to facilitate for the first time can definitely benefit from synthesis of scientific in- EBM. Below we identify eight key tenets (Table 2) that we have formation that allows a description of what is possible from a multi- extracted from the US experience, with the objective of improving sector perspective. For example, in what has emerged as a national the implementation and practice of IEAs. These tenets are by no http://icesjms.oxfordjournals.org/ example of EBM implementation, a synthetic scientific report was means unique to the USA. Indeed, they build on best practices pro- produced in Puget Sound, WA (Ruckelshaus and McClure, 2007), posed elsewhere (e.g. UNEP and IOC-UNESCO, 2009) but are before the delineation of ecosystem goals by the regional manage- focused on areas in which emphasis is needed to see IEAs implemen- ment agency. Indeed, the first IEA for any region, and any major ted in the USA: transparency, accountability, assurance of scientific changes to ecosystem management goals, requires an iterative dia- rigor, and forging effective links between IEA science and ocean logue between scientists and those charged with setting goals. policy. It is our hope that clear articulation of these tenets will Third, the IEA is not prescriptive about analytical approaches. help scientists, managers, and policy-makers avoid potential pitfalls There are a variety of qualitative and quantitative tools that can be as ocean management transitions from its historical focus on single used to conduct an IEA, and the appropriate tool depends on the species and single sectors to the one that is ecosystem-based. by Howard Browman on July 2, 2014 particulars of data availability, scientific capacity, and ecosystem goals. Thus, because it is not prescriptive, an IEA can be implemen- ted now, over a range of spatial and temporal scales, for a variety of Key tenets of IEA implementation and practice ecosystem objectives (Tallis et al., 2010). Fourth, an IEA is funda- 1. Engage with stakeholders, managers, and policy-makers mentally cross-sectoral. That is, IEAs address drivers, pressures, early, often, and continually and impacts on ecosystem services that cross ecosystems, habitats, Good communication underlies almost all successful applications species, and human activities and institutions. Thus, ecosystem of science to management and policy (Olson, 2009; Baron, 2010). assessments that address a single sector (e.g. fisheries) cannot be The success of IEAs depends fundamentally on the scientist– considered truly integrated (Levin et al., 2009). manager–stakeholder dialogue in which scoping sessions serve to IEAs are currently under development throughout the USA in a support the policy-making process that leads to the establishment phased implementation strategy (Table 1). NOAA selected the of a shared vision of the current and futuredesired states of theireco- California Current for the first full IEA development, and work system. This dialogue provides an opportunity for scientists to began in 2009. There is also formal IEA work being conducted in evaluate realistic alternative approaches for management to the Gulf of Mexico, the US Northeast Shelf, Alaska, and the achieve desired states. Interactions between scientists, managers, Pacific Islands. Although IEAs were introduced only 5 years ago as and stakeholders must not be limited to a one-time event, but an approach to enable marine EBM (Levin et al., 2008), there are should be maintained to adapt scientific products, management already examples that illustrate how they can and will be used. strategies, and ecosystem goals over time (Watson, 2005; Fletcher, A key success of the US IEA programme has been the adoption of 2007; UNEP and IOC-UNESCO, 2009; deReynier et al., 2010). a standard conceptual framework that facilitates customized imple- There areanumberof meansthatIEAteams can use to engage the mentation. While IEA products are best tailored to regional goals larger EBM community. In countries like the USA, these include and institutions, the existence of a common, cyclical approach presenting IEA science in person or via webinars, developing web (Figure 1) has provided for national consistency while enabling content, engaging in social media, producing web-based videos and encouraging regional flexibility. It is already clear that this pro- (http://www.noaa.gov/iea/multimedia/index.html), and conven- gramme design eases internal and external communication and ing workshops, among others. Critically, effective communication enhances the potential for regional comparisons. must consider the technical and scientific knowledge of each user. Perhaps as a result of national consistency and regional flexibil- Sophisticated data analyses and mathematical models that are a ity, the “clientele” for IEAs is substantial and growing. The primary large part of IEAs may not be accessible to all users (cf. Gleason IEA clients include (i) new decision-making entities that have et al., 2010). Such hesitancy can be assuaged by communicating in emerged to address increasingly diverse ocean uses in a a way that is tailored to specific interests and the knowledge base 1208 J. F. Samhouri et al.

Table 1. Snapshot examples of regional implementation of IEAs in the USA, as described on regional IEA websites accessible at http://www. noaa.gov/iea/. Key management Region Select examples of implementation progress topics Current clientele Alaska Complex † Annual ecosystem considerations chapter provided to † Fisheries † North Pacific Fisheries NPFMC † Climate Management Council † † Ecosystem indicator selection process through † Energy Arctic Research Commission multistakeholder workshops † North Pacific Marine Science † Use of multiple ecosystem models to develop and test Organization (PICES) indicators † Bureau of Ocean Energy † Development of metrics to represent condition of Management ecosystems; help establish reference points and comparisons across ecosystems (part of Risk Assessment) California Current † Two IEA reports, including all steps within IEA process † Climate † Pacific Fisheries Management (2011, 2013) † Fisheries Council † † Downloaded from Ecosystem considerations report delivered to PFMC (2011) † Energy West Coast National Marine † Integration of IEA science into management discussions Sanctuaries (e.g. Sanctuaries, Puget Sound, Sacramento River) † West Coast Governors Alliance † IEA “toolkit” for CC † North Pacific Marine Science † Next steps: habitats, highly migratory species Organization (PICES) http://icesjms.oxfordjournals.org/ Gulf of Mexico † Demonstration project with GMFMC SEDAR process to † Commercial and † Gulf of Mexico Fisheries provide ecosystem considerations Recreational Fishing Management Council † Development of ensemble set of ecosystem models † Energy † Gulf of Mexico Alliance † Ecosystem Assessment Management Report for 4 † Population † National Marine Sanctuaries estuarine ecosystems † Recreation and † Development of digital trophic database and Data Atlas Tourism † Next: Ecosystem Status report for 2013 † Shipping Northeast US Shelf † Development of multispecies and ecosystem models † Fisheries † New England and Mid-Atlantic by Howard Browman on July 2, 2014 † Indicator development using DPSIR † Wind energy Fisheries Management Councils † † Development of spatial tools for EBFM † Protected species Northeast Regional Ocean Council † Semi-annual production of Ecosystem Advisory report † Climate † Mid-Atlantic Regional Council on † Biannual Ecosystem Status report † Recreation and the Ocean Tourism † North Atlantic Fisheries † Shipping Organization † International Council for Exploration of the Seas Pacific Islands † Studies on effects of ocean circulation on larval retention † Shared use resources † Western Pacific Regional Fisheries † Development of “reef” and “coastal” ecosystem model to † Aquaculture Management Council † understand energy flows and interactions † Climate change and The Kohala Center † Studies on human dimensions in ecosystem functions in Kona Sentinel Site † Hawaii Division of Aquatic Kona Coast (incl. indicator development, identification of † Fisheries and Resources drivers and pressures, communication and networking aquarium fish † Pacific Islands and Hawaii Ocean with managers) collection Observing Systems † Hawaiian Islands Humpback Whale National Marine † National Marine Sanctuaries Sanctuary plan review † Cetacean habitat modelling † Kona IEA research cruises Over time, examples of progress, key management topics, and current clientele will expand and there are aspirations to add Caribbean, Southeast, and Great Lakes regions for IEA implementation. of each audience. For some groups, using marine spatial planning trophic relationships as well as environmental conditions. The as an entre to the utility of IEAs may be useful, while for others, it bottom line is that one size will not fit all, and that meeting audi- may be better to explain an IEA as an extension of standard stock ences where they are will be crucial in engaging the diverse consti- assessments that includes such considerations as multispecies and tuents of IEAs. Lessons learned from developing IEAs in the USA 1209

Table 2. Tenets for IEA implementation and practice. We offer three illustrative examples below that underscore the im- portance of including such work in IEAs. 1. Engage with stakeholders, managers, and policy-makers early, often, and continually Cultural keystone species 2. Conduct rigorous human dimensions research Clearly, in evaluating trade-offs among management options, it is 3. Recognize the importance of transparently selecting indicators 4. Set ecosystem targets to create a system of EBM accountability critical to know if a particular species holds importance beyond 5. Establish a formal mechanism(s) for the review of IEA science its monetary value or ecological role. For example, the red alga 6. Serve current management needs, but not at the expense of more Porphyra abbottiae has been dubbed a “cultural keystone species” integrative ocean management for the role it plays in indigenous cultures in the Northern 7. Provide a venue for EBM decision-making that takes full advantage of California Current (Turner, 2003). Porphyra is highly valued for IEA products its nutritional content, as a gift and trade item, and for its medicinal 8. Embrace realistic expectations about IEA science and its properties (Garibaldi and Turner, 2004). As the term cultural key- implementation stone species suggests, small changes in the availability of Porphyra can have disproportionate impacts on the cultures that depend on it. Elders believe declines in gathering this alga will The Puget Sound Partnership (PSP) provides one model of en- lead to a substantial loss of knowledge and tradition (Garibaldi gagement for IEAs. The PSP was established by the Governor of

and Turner, 2004). Thus, ignoring the cultural role species like Downloaded from Washington state, USA, in 2005 to protect and restore the Puget Porphyra play in human communities will miss fundamental attri- Sound ecosystem. The PSP worked with the general public, diverse butes of the social-ecological system. stakeholders, a science working group, and regional managers and policy-makers to craft an “Action Agenda” to achieve this goal. The Ecosystem service modelling Action Agenda was developed using a community-based approach Marine ecosystems produce goods and services that sustain human thatmadeuseofanumberofpublic forumsandthatpromoted exten- life, and increasingly, modern conservation science is focused on http://icesjms.oxfordjournals.org/ sive public and scientific input. These included workshops, expert protecting ecosystems for the benefit of humanity (Kareiva and topic meetings, and local meetings with the general public. In all, Marvier, 2011). Process-based ecosystem service models that quan- 1600 people attended workshops, 75 presentations were given to titatively link change in marine ecosystemsto change in humanwell- community and business groups, and some 12 300 public comments being will substantially enhance our ability to analyse the status and were received (PSP, 2009). This level of engagement has been central forecast the future of coupled social-ecological systems (Guerry to the success of the PSP in obtaining funding to support restoration et al., 2012). They offer a productive avenue for quantifying efforts$—$230 million has been spent annually on Puget Sound impacts of EBM strategies on livelihoods and communities. These Restoration since the creation of the Action Agenda (http://www. models cannot be developed, refined, and put to use for manage- kitsapsun.com/news/2012/oct/27/little-progress-reported-in-puget- ment planning without deep interdisciplinary collaboration by Howard Browman on July 2, 2014 sound-health/#axzz2RtAGFF1l). between economists, social scientists, and natural scientists. Many examples of fruitful ecosystem service modelling efforts 2. Conduct rigorous human dimensions research exist around the USA (Barbier et al., 2011; Tallis et al., 2013; The notion that humans are integral to marine ecosystems is now White et al., 2012b). Perhaps the area ripest for further development commonplace (McLeod and Leslie, 2009). However, there is related to services provided by coastal habitats (e.g. saltmarshes, remains a surprising asymmetry in the attention researchers coral reefs, mangroves, oyster beds, barrier islands, etc.). The devote to natural vs. human systems when it comes to marine value of coastal habitats can be enormous—in many areas, they EBM (cf. Kittinger et al., 2012). We contend that for EBM, an provide a large variety of benefits to people, including coastal absence of rigorous investigations on the human dimensions of eco- defence from storms, aesthetic value, climate change mitigation systems is akin to ignoring basic elements of ecosystems such as up- via carbon storage and sequestration, and more. For instance, a welling, plankton productivity, and fish population dynamics. recent analysis examined how natural habitats across the entire Beyond simple definitions of EBM (McLeod et al., 2005), there USA modify hazards due to sea-level rise, suggesting that the are a number of reasons why social science is fundamental. First, number of people, poor families, elderly, and total value of residen- EBM is predicated on managing the actions of people based on tial property most exposed to hazards can be reduced by half if exist- their needs, and understanding people and their motivations is fun- ing coastal habitats remain fully intact (Arkema et al., 2013). This damental to the achievementof ecosystem goals (Pollnac et al., 2010; study underscored just one of the important functions natural habi- Gutierrez et al., 2011). Second, understanding and addressing trade- tats play in the coupled social-ecological system. The development offs is acentral EBM pursuit; consequently, identifying howand why of models that adequately capture the layered complexity of ecosys- cultures value different ecosystem services derived from marine tem services provided by coastal habitats will require significant in- systems is a basic necessity of EBM. Additionally, including tellectual consideration and resources, but are central to the human dimensions of ecosystems can enhance the implementation development of socially resonant IEAs. of EBM by quantifying impacts on livelihoods and communities, in- creasing buy-in, reducing conflict (Evans and Klinger, 2008), and Behavioural science as a source of insight helping to develop alternatives that address concerns of those A deficiency of standard economic models is that they assume affected by ecosystem changes (Turner et al., 2008). people make decisions rationally (Amir et al., 2005). The evidence The ways in which human dimensions research could be incor- across a wide range of examples suggests, however, that humans porated into IEAs are as varied as the academic disciplines that act in predictably irrational ways (Ariely, 2009). For example, study humans in marine ecosystems (e.g. history, political science, Johnson and Goldstein (2003) showed that organ donation rates geography, anthropology, sociology, economics, psychology, etc.). across several European countries were highly predictable based 1210 J. F. Samhouri et al. on default options: participation rates were much higher in coun- properly, will generate a transparent and inclusive indicator set that tries where citizens “opted in” to donor programmes by default truly allows for monitoring and measuring of progress towards eco- than in countries where the default option was to “opt out”. system goals. Further, it can be used to generate a subset of standard Peoplemay bemotivated toactionbased onsimilarlysubtle differ- indicators that will facilitate comparisons among IEA regions. ences in choices about how to report and deliver IEA science. For in- An example of culling these indicators, as developed in an IEA stance, Levin et al.(2010)showed that in some cases, indicators with context, comes from the Northeast US shelf ecosystem where over equivalent ecological information content have widely varying social 300 indicators have been documented (Link and Brodziak, 2002). resonance. Similarly, Beaudreau et al. (2011) demonstrated that After an iterative selection process conducted to winnow this alternative views of biodiversity can lead to radically different percep- number down (Link, 2005; Methratta and Link, 2006), many con- tions of risk to ecosystem components. Moreover, recognition of sultations among scientists, and several stakeholder workshops, boundaries of acceptable ecosystem conditions for different ocean the number of indicators routinely reported (http://www.nefsc users and stakeholders can help define the “safe” operating space .noaa.gov/publications/crd/crd1207/crd1207.pdf)is30–50 for EBM decisions (Rockstro¨m et al., 2009). Successful implementa- (EcoAP, 2009, 2012). These ecological, economic, and social indica- tion of EBM will require people to support management actions, and tors are intended to represent aspects of the marine social-ecological in some cases, to take actions as individuals (e.g. voting, clean up environment that affect and are affected by human use of the ecosys- efforts, compliance with regulations, etc.). Social science can facilitate tem. This much more concise list allows for multiple uses of the in-

anunderstanding ofpeople’s motivations for supporting oropposing formation in many contexts, regular updates (every 2–3 years), and Downloaded from such actions. Insights from behavioural science can help guide the the examination of novel hypotheses and processes related to development of socially meaningful IEA products most likely to observed patterns. An even smaller subset of indicators is reported garnerpublic supportfor, and participationin, EBM implementation more frequently to track trends in ecosystem dynamics and (cf. Thaler and Sunstein, 2003). provide alerts in the event of major shifts (http://www.nefsc.noaa. gov/ecosys/advisory/current/advisory.html). 3. Recognize the importance of transparently selecting http://icesjms.oxfordjournals.org/ indicators 4. Set ecosystem targets to create a system of EBM The development of a performance evaluation system is basic to accountability good management of businesses, environmental issues, human Alone, a transparently selected, ecologically defensible, politically health, and more (Otley, 1999; Link, 2005; Rice and Rochet, 2005; acceptable set of ecosystem indicators is of limited value for EBM. Manski, 2013).Indicators are the basic building block fora perform- To guide effective management, indicators must be associated ance management system: they provide a means to track progress with targets, or values of the indicators equated with successful towards management goals. At first glance, indicator selection achievement of management goals (Samhouri et al., 2011, 2012). may seem like a straightforward process. IEAs require indicators We distinguish between targets and other reference points such as that capture key ecosystem and socio-economic states and processes ecosystem thresholds, in that targets need not be defined objectively by Howard Browman on July 2, 2014 (Levin et al., 2009), and a number of frameworks exist to rigorously through analysis of biophysical or socio-economic data. Rather, determine an indicator set (e.g. Methratta and Link, 2006; Shin et al., targets should reflect desired ecosystem states, as articulated 2010a, b; Kershner et al., 2011; James et al., 2012). through people’s stated or revealed preferences for alternative eco- In practice, deriving a set of indicators is more challenging than system conditions. Active monitoring of indicators in relation to simply determining proxies for key ecosystem components or pro- such targets can be used to instigate, cease, or adapt EBM actions. cesses. This difficulty emerges from the truism that IEAs require However, generating consensus on target values for indicators at some indicators, but not too many (cf. Lægreid et al., 2006). IEAs the ecosystem level is no small task, for at least two reasons. First, require that the performance characteristics of the indicators are perceptions of ecosystem status are just that: a personal, subjective understood and that the status and trendsand current values relative assessment that is very likely to differ from stakeholder to stakehold- to EBM targets can be interpreted correctly (Rice and Rochet, 2005). er depending on how s/he uses and values marine ecosystems (e.g. Using too fewindicators limitsthe scope foradequately assessing the Loring, 2013). Thus, it is realistic to expect target setting to be a pol- ecosystem, but too many indicators reveal a lack of prioritization of itical activity, informed by scientific information. Such is the case in goals and objectives (Aucoin and Jarvis, 2005; Fulton et al., 2005; many familiar single-sector management examples (e.g. targets for Methratta and Link, 2006). The fact that there is a limit on the protected areas, fisheries harvest guidelines, water quality, etc.). Yet number of indicators means that choices must be made, and not means of objectively determining ecosystem thresholds in relation all potential indicators can be used. to environmental and anthropogenic pressures show some The key to whittling down a potentially infinite list of candidate promise of at least identifying regions where target levels are apt indicators is a transparent selection process. Transparency can to be most effective (Samhouri et al., 2010; McClanahan et al., avoid, or at least reveal, institutional and personal biases, helping 2011; Large et al., 2013). The actual control rules used will still to create an indicator portfolio that meets the needs of EBM by need to be specified, but many other disciplines (e.g. ecotoxicology, being scientifically, politically, and socially meaningful. Formal water quality, species-focused fisheries management) have used evaluation of indicators following one of many indicator-screening similar non-linearities to support the establishment of these target frameworks (see references above) ensures the transparency of indi- levels. A related idea advocated by Rockstro¨m et al. (2009), among cator selection. Final selections can be made based on additional others, is to define nature’s safe operating space, or boundaries, input from complementary public engagement efforts and conven- via objective analysis. Policy-makers can use this information to de- tional political processes. Indicators that are reported regularly can lineate targets within or outside of the natural system boundaries. represent a subset of those that are tracked for the purposes of eco- The second challenge in defining target values at the ecosystem system status assessment, providing a hedge in the event of ecologic- level is that they force trade-offs to the fore. It is relatively easy to al surprises and evolving management goals. This recipe, if executed develop ambitious targets for individual indicators. However, Lessons learned from developing IEAs in the USA 1211 achievement of individual targets is complicated by interrelation- however, as it should not be applied in a way that impedes innov- ships among ecosystem components. Improvements towards one ation and adaptability. While we do not intend to be prescriptive target may impede or enhance progress towards another regarding what standardized products or the review mechanism(s) (Samhouri et al., 2011; Fay et al., 2013). Thus, it is critical for IEA shouldlook like, wedofeel that thedevelopment ofthese processes is science to outline how establishment of alternative targets for indi- essential to guaranteeing rigor in IEAs and for more effective use of vidual indicators influences the full suite of indicators related to IEA science in EBM implementation. EBM goals (for a related discussion on an ecosystem approach to fisheries, see Rice, 2005). 6. Serve current management needs, but not at the expense There are many, non-mutually exclusive ways for IEA teams to of more integrative ocean management illustrate the inter-relations of different target sets. For example, IEAs were borne out of the need to provide integrative science to simulation models can describe what is ecologically and socio- inform comprehensive, ecosystem-based ocean management. economically feasible for the full ecosystem, given a set of targets While genuine EBM must operate across different ocean use for a particular subset of indicators (e.g. Kaplan and Leonard, sectors, in the USA, it is clear that this vision is constrained by exist- 2012; Fay et al., 2013). For example, Kaplan and Leonard (2012) illu- ing governance structures and is more of an aspiration than the strated the direct, indirect, and induced effects of management strat- current reality (Rosenberg and Sandifer, 2009). The current reality egies with alternative groundfish targets for the broader ecological bears more of a resemblance to many individual actors operating

communities of which groundfish are a part and for fishery in a common space and using common pool resources (Crowder Downloaded from sectors, industry suppliers, and household spending patterns. et al., 2006), without much accommodation for the joint social Scenarios such as these can be developed into survey instruments, value of their decisions (White et al., 2012a). There are some exam- designed to solicit thresholds of acceptability for different ecosystem ples available, however, suggesting that accounting for the joint states. Similarapproaches arecommonlyused to evaluate normative social value of single-sector management decisions can lead to beliefs associated with wildlife management actions (Zinn et al., improved outcomes (Hannesson et al., 2009; Allnutt et al., 2012; 1998) and outdoor recreational activities (Vaske et al., 1993). Levi et al., 2012). The contrast between aspiration and reality http://icesjms.oxfordjournals.org/ With an understanding of what is acceptable across the full ecosys- creates a tension between meeting single-sector ocean management tem, targets for individual indicators can be deduced. Just as im- needs today and making directed progress towards true, cross- portant, these approaches can identify ecosystem states that are sectoral EBM. categorically non-desirable by all stakeholders or those that are tech- In the near-term, as EBM and IEAs gain trust, traction, and nically infeasible to achieve. momentum, it is imperative that IEA products support current By specifying indicator targets, IEAs have the potential to create a national ocean resource mandates as they are prescribed now. If system ofEBM accountability. Givensuchasystem,itis up to policy- IEA products do not provide information relevant to single makers to decide on a course of management action that most ap- sector and single species marine management decisions (e.g. fish- propriately reflects stakeholder desires for the ecosystem. eries, protected species, marine sanctuaries, etc.), then the IEA en- by Howard Browman on July 2, 2014 terprise will not be viewed (by some) as useful. However, this 5. Establish a formal mechanism(s) for the review existential risk need not be a major concern. Because they are syn- of IEA science thetic “scientific warehouses”, IEAs can be leveraged to augment A basic step in seeing IEA science used to inform EBM implementa- single-sector management. For instance, IEAs summarize infor- tion is likely to involve the development and institutionalization of a mation about climatic conditions, species abundance, and mechanism(s) for peer review. Ideally, managers, stakeholders, and human activities in the ocean, and these data can be packaged scientists will work with policy-makers to develop a template for to inform the emerging needs of ecosystem-based fisheries man- the review of IEA science. Key elements of a review process may agement. In fact, “ecosystem considerations” reports and fisheries include committees tasked with external, independent review of ecosystems reports that place individual stocks at the centre of the IEA scientific products, opportunity for public comment, and a ecosystem, and describe climate, prey availability, predator abun- clear commitment to ensure the best available science is used to dance, and non-fisheries stressors, have already been requested inform management decisions. In the USA, federal fisheries manage- and delivered to several Regional Fisheries Management ment practices provide a useful analogue, as they are conducted in a Councils across the USA (e.g. Levin and Wells, 2011; EcoAP, highly reviewed context. National Standard 2 of the Magnuson- 2012; Zador, 2012). Similarly, within National Marine Stevens Sustainable Fisheries Act requires documentation that a Sanctuaries, habitat risk assessments can be used to inform stock assessment has used best available science. A significant spatial planning decisions regarding species of concern such as process to review information relied upon and produced by stock corals and sponges (Samhouri et al., 2012). assessments has thus developed. The easier path forward is to repackage existing scientific pro- The success of the precedent set by US species-focused fisheries ducts and market them as IEAs. However, in the longer term, if management is encouraging, although from our perspective, the IEAs focus exclusively on the scientific needs of individual ocean precise format of an IEA review process is less important than ensur- use sectors, they will fail to fulfil the goals of EBM. While potentially ing that it occurs. The expectations from an IEA, whatever they may useful, a report describing ecosystem considerations for an individ- be, should be specified in a formal terms of reference to clarify sci- ual species or an individual sector is not an IEA. Such reports will entific deliverables and their intended uses. The review process help to justify the existence of an IEA programme in today’s ocean could lead to standardized outputs tailored for use by EBM decision management spheres, but will not help to realize EBM. Individual makers. Forexample, the terms of referencemaystatethatIEAs must sectors aim to maximize utility based on their independent goals include a minimum of five management strategy evaluations asso- for marine ecosystems (White et al., 2012a), yet at some point, a ciated with corresponding risk levels for stakeholder-designated comprehensive evaluation of resulting trade-offs is needed. To ecosystem goals. We urge caution with any type of standardization, truly blaze a new trail for a sustainable ocean future, IEA science is 1212 J. F. Samhouri et al. better served by striving for the anticipatory development of tools Endangered Species Act (ESA) of 1973. Population viability analysis that can assist in actualizing better, more efficient marine EBM. is considered one of the best quantitative methods for evaluating ex- Only by integrating biophysical and socio-economic information tinction risk, yet it remains infrequently used in endangered species across an ecosystem can an IEA—and by extension EBM—help to recovery planning (Morris et al., 2002). This reality is natural, shape a new reality characterized by coordinated and sustainable expected, and to a limited extent even desirable—unfamiliar scien- ocean uses. tific products require time for vetting and shaping to practical appli- cations. However, lack of familiarity should not be an excuse for a 7. Provide a venue for EBM decision-making that takes lack of reliance on best available science in guiding management full advantage of IEA products decisions. An IEA is in many ways widely marketable; it is designed to provide Given the lag time often associated with new management strat- scientific information that can inform more sustainable, efficient, egies, EBM and IEAs could take a generation or more to establish a and compatible ocean uses across multiple sectors. Perhaps most strong foothold across all sectors of US ocean management. The importantly, an IEA strives to clarify the varied outcomes of alterna- application of stock assessment science in US fisheries management tive management scenarios, by drawing connections between land provides a useful corollary. In his book, Smith (1994) elegantly and sea, nearshore and offshore, fisheries and water quality, tracked the development of the theory and associated scientific energy extraction and coastal recreation, and whale watching and underpinnings of fish stock recruitment and production dynamics,

shipping, just to note a few possible contrasts. However, although largely in the middle of last century. He noted that it was at least a Downloaded from IEAs create a broader perspective for decision makers, more integra- decade, or more, before such information was used as the basis for tive ocean management will not just happen. setting reference points and harvest control rules to manage living As is true on the international stage (Gjerde et al., 2008), chal- marine resources. It has taken even longer for the full suite of govern- lengesto coherent management and regulation in the USA represent ance structures and review venues to evolve into something even re- both a policy implementation gap and a governance gap (Hildreth, motely familiar to what we now observe in the USA with National 2008). Governance processes have been exploring ways for effective Standards,RegionalFisheriesManagementCouncils,andrelatedma- http://icesjms.oxfordjournals.org/ coordination of cross-sector decision-making ever since the Law of chinery to conduct fishery management. A similar evolution is to be the Sea came into force, and many hurdles remain (Rice, 2011). expected for IEAs as they begin to support EBM. Nonetheless, in the long term, we believe that EBM will require IEA science is young and necessary, but not sufficient for making the creation of a venue for making decisions about ecosystems, in EBM happen. The National Ocean Policy, the National Environ- contrast to decisions about individual species, individual issues, mental Policy Act, the National Coastal Zone Management Act, or individual sectors. and other statutes do provide inroads for implementing facets of Getting there will not be easy. In the USA, IEAs do not yet fold EBM (Keiter, 1998; Abbott and Marchant, 2010; Fox et al., 2013). neatly into any one existing governance structure, and the creation However, it could be argued that the readiness of IEA products pre- of a new one is an enormously complex undertaking. Such a cedes the preparedness of legal and management structures to use by Howard Browman on July 2, 2014 decision-making body would need to be charged with maintaining the information. Happily, there are several examples from across a cross-sectoral focus, possibly by way of an inter-sectoral set of con- theworld,suggestingthatevencomplex, internationalpolicy-making stituent members. The US National Ocean Policy recognizes the im- organizations are capable of adapting quickly to new scientific devel- portance of Regional Ocean Partnerships (ROPs) that have opments and including them in their processes (e.g. IUCN extinction identified priorities relevant to their specific regions, and could risk assessments have used “quantitative population models” since provide a preliminary venue for the delivery of IEA science. The 1994; in 1992, Rio Agenda 21 created a policy framework for precau- ROPs may serve as an initial template, and could work to leverage tionary fisheries management based on stock–recruitment product- existing legal mandates and apply them within existing single-sector ivity dynamics, etc.). Our observation is that these efforts tend to be management frameworks, while remaining cognizant of the cross- particularly effective when top-down actions to implement policy sectoral trade-offs that deserve consideration. and management are coupled to bottom-up efforts to develop the The formation of such a venue is a step that will facilitate dia- relevant science. logue, the adjudication of compatible vs. incompatible ocean uses, and the coordination of management decisions. Ideally, it could serve a facilitative role by tracking actions taken within individual sectors and providing a forum to negotiate trade-offs that influence The future of IEAs for ocean management the joint social value of an ecosystem with respect to stakeholder- in the USA and beyond determined goals (White et al., 2012a). Even with due diligence by To date, IEAs are most useful as syntheses of information and for in- scientists to see IEAs used in management, EBM is unlikely to fluencing the way scientists, managers, and policy-makers think become a reality without the creation of a group charged with main- about ocean issues. IEAs take the conceptually appealing idea of taining a broad perspective to coordinate trade-offs across individ- EBM, and provide some structure around which to consider the ual ocean-use sectors. otherwise overwhelming task of identifying, evaluating, and devel- oping management strategies for multiple ecosystem components 8. Embrace realistic expectations about IEA science and sectors. These characteristics make them a next-generation and its implementation tool for ocean and coastal management. Those involved with devel- Historically, there is a time-lag between the development of the oping IEAs in the USA are learning by doing, improvising as they go, science for a new environmental management approach and the and subsequently trying to improve the implementation of EBM. adoption of this information into policy and management. For in- The eight tenets introduced in this article (Table 2) represent stance, in the USA, extinction risk is one of the main criteria evalu- lessons we have learned along the way that may enhance the ated in determining whether a species should be listed under the uptake of integrated scientific assessments in the USA and beyond. Lessons learned from developing IEAs in the USA 1213

IEAs facethe tall orderofmeeting managers wheretheyaretoday, Beaudreau, A. H., Levin, P. S., and Norman, K. C. 2011. Using folk tax- while at the same time remaining sufficiently visionary that they can onomies to understand stakeholder perceptions for species conser- show managers where they could be tomorrow. IEA scientists will be vation. Conservation Letters, 4: 451–463. challenged to remain nimble and responsive to sectoral needs, and at Caddy, J. F. 1999. Fisheries management in the twenty-first century: will the same time steadfast and persistent in the pursuit of the larger new paradigms apply? Reviews in Fish Biology and Fisheries, 9: 1–43. goal of EBM. In the short term, we expect that the most efficient way forward is to tailor IEAs, so that they are clear where and Crowder, L. B., Osherenko, G., Young, O. R., Airame, S., Norse, E. A., Baron, N., Day, J. C., et al. 2006. SUSTAINABILITY: resolving mis- how cross-sectoral, coordinated management can increase the matches in US ocean governance. Science, 313: 617–618. probability of attaining ecosystem goals relative to single-sector, deReynier, Y. L., Levin, P. S., and Shoji, N. L. 2010. Bringing stake- siloed decision-making (White et al., 2012a). If successful, IEAs holders, scientists, and managers together through an integrated will lead managers and policy-makers to create the infrastructure ecosystem assessment process. Marine Policy, 34: 534–540. capable of ingesting and applying principles of ecosystem science Ecosystem Assessment Program. 2009. Ecosystem status report for the to foster sustainable ocean uses in the future. Northeast US Continental Shelf Large Marine Ecosystem. Northeast The full utility of IEAs in ocean management will not be apparent Fisheries Science Center Reference Document, 09-11. 34 pp. overnight. We have spent centuries studying the parts and only Ecosystem Assessment Program. 2012. Ecosystem status report for recent decades in more formally developing the science to inform the Northeast US Continental Shelf Large Marine Ecosystem— 2011. Northeast Fisheries Science Center Reference Document, management of the whole (Jackson, 2011). IEAs are science for Downloaded from tomorrow’s oceans, and we urge patience as it matures, and with 12-07. 32 pp. it, the readiness of ocean managers to use it. The eight tenets out- European Commission. 2008. Directive 2008/56/EC of the European lined above summarize our perspective on how to carry IEAs Parliament and of the Council of 17 June 2008 establishing a frame- work for Community actions in the field of marine environmental from the realm of interesting scientific products to the realm of policy (Marine Strategy Framework Directive). Official Journal of useful EBM tools. Common to all of them is a need for patience, per- the European Communities, L164/19 25.06.2008. http://icesjms.oxfordjournals.org/ sistence, political will, funding, and augmented scientific capacity Evans, K. E., and Klinger, T. 2008. Obstacles to bottom-up implementa- (especially human dimensions capacity). It will also take innov- tion of marine ecosystem management. Conservation Biology, 22: ation, engagement, and open dialogue to see ecosystem science 1135–1143. implemented into effective, ecosystem-based ocean management, Fay, G., Large, S. I., Link, J. S., and Gamble, R. J. 2013. Testing systemic but we assert that it can be done using IEAs. fishing responses with ecological indicators: an MSE approach. Ecological Modelling, 265: 45–55. Acknowledgements Fletcher, S. 2007. Converting science to policy through stakeholder in- volvement: an analysis of the European Marine Strategy Directive. This article was inspired by dialogue at a national IEA programme Marine Pollution Bulletin, 54: 1881–1886. meeting in September 2012. We thank all participants of that Fluharty, D., Abbott, M., Davis, R., Donahue, M., Madsen, S., Quinn, T., meeting, participants in the National NOAA IEA programme, and and Rice, J. 2006. Evolving an ecosystem approach to science and by Howard Browman on July 2, 2014 IEA partner NOAA ESRL in Boulder, CO, USA, for hosting the management throught NOAA and its partners. Available from meeting. Jake Rice and an anonymous reviewer provided comments www.sab.noaa.gov/reports/eETT_Final_1006.pdf (last accessed 14 that vastly improved the article. March 2013). Foley, M. M., Armsby, M. H., Prahler, E. E., Caldwell, M. R., Erickson, References A. L., Kittinger, J. N., Crowder, L. B., et al. in press. Improving ocean management through theuse of ecological principles and inte- Abbott, K. W., and Marchant, G. E. 2010. Institutionalizing sustainabil- grated ecosystem assessments. Bioscience, 63: 619–631. ity across the federal government. Sustainability, 2: 1924–1942. Allnutt, T. F., McClanahan, T. R., Andre´foue¨t, S., Baker, M., Lagabrielle, Fox, E., Miller-Henson, M., Ugoretz, J., Weber, M., Gleason, M., Kirlin, E., McClennen, C., Rakotomanjaka, A. J. M., et al. 2012. Comparison J., Caldwell, M., et al. 2013. Enabling conditions to support marine of marine spatial planning methods in Madagascar demonstrates protected area network planning: Califronia’s Marine Life value of alternative approaches. PLoS ONE, 7: 28–969. Protection Act Initiative as a case study. Ocean and Coastal Management, 74: 14–23. Amir, O., Ariely, D., Cooke, A., Dunning, D., Epley, N., Gneezy, U., Koszegi, B., et al. 2005. Psychology, behavioral economics, and Fulton, E. A., Smith, A. D. M., and Punt, A. E. 2005. Which ecological public policy. Marketing Letters, 16: 443–454. indicators can robustly detect effects of fishing? ICES Journal of Marine Science, 62: 540–551. Ariely, D. 2009. Predictably Irrational, Revised and Expanded Edition: the Hidden Forces That Shape Our Decisions. HarperCollins Garibaldi, A., and Turner, N. 2004. Cultural keystone species: implica- Publishers, New York, USA. tions for ecological conservation and restoration. Ecology and Arkema, K. K., Guannel, G., Verutes, G., Wood, S. A., Guerry, A., Society, 9: 1. Ruckelshaus, M., and Kareiva, P. et al. 2013. Coastal habitats shield Gjerde, K., Dotinga, H., Hart, S., Molenaar, E., Rayfuse, R., and Warner, people and property from sea-level rise and storms. Nature R. 2008. Regulatory and Governance Gaps in the International Climate Change, advance online publication. http://www.nature Regime for the Conservation and Sustainable Use of Marine .com/nclimate/journal/vaop/ncurrent/full/nclimate1944.html Biodiversity in Areas beyond National Jurisdiction. IUCN, Gland, (last accessed 29 July 2013). Switzerland. 70 pp. Aucoin, P., and Jarvis, M. D. 2005. Modernizing government account- Gleason, M., McCreary, S., Miller-Henson, M., Ugoretz, J., Fox, E., ability: a framework for reform. Canada School of Public Service, Merrifield, M., McClintock, W., et al. 2010. Science-based and Canada. stakeholder-driven marine protected area network planning: a suc- Barbier, E. B., Hacker, S. D., Kennedy, C., Koch, E. W., Stier, A. C., and cessful case study from north central California. Ocean & Coastal Silliman, B. R. 2011. The value of estuarine and coastal ecosystem Management, 53: 52–68. services. Ecological Monographs, 81: 169–193. Guerry, A. D., Ruckelshaus, M. H., Arkema, K. K., Bernhardt, J. R., Baron, N. 2010. Escape from the Ivory Tower: A Guide to Making Your Guannel, G., Kim, C-K., Marsik, M., et al. 2012. Modeling benefits Science Matter. Island Press, Washington, DC. 272 pp. from nature: using ecosystem services to inform coastal and 1214 J. F. Samhouri et al.

marine spatial planning. International Journal of Biodiversity Lægreid, P., Roness, P. G., and Rubecksen, K. 2006. Performance man- Science, Ecosystem Services & Management, 8: 107–121. agement in practice: the Norwegian Way. Financial Accountability & Gutierrez, N. L., Hilborn, R., and Defeo, O. 2011. Leadership, social Management, 22: 251–270. capital and incentives promote successful fisheries. Nature, Large, S. I., Fay, G., Friedland, K. D., and Link, J. S. 2013. Defining trends advance online publication. http://dx.doi.org/10.1038/nature and thresholds in responses of ecological indicators to fishing and 09689 http://www.nature.com/nature/journal/vaop/ncurrent/ environmental pressures. ICES Journal of Marine Science, 70: abs/nature09689.html#supplementary-information. 757–767. Halpern, B. S., Longo, C., Hardy, D., McLeod, K. L., Samhouri, J. F., Levi, T., Darimont, C. T., MacDuffee, M., Mangel, M., Paquet, P., and Katona, S. K., Kleisner, K., et al. 2012. An index to assess the health Wilmers, C. C. 2012. Using grizzly bears to assess harvest-ecosystem and benefits of the global ocean. Nature, 488: 615–620. tradeoffs in salmon fisheries. PLoS Biology, 10: e1001303. Halpern, B. S., Walbridge, S., Selkoe, K. A., Kappel, C. V., Micheli, F., Levin, P. S., Damon, M., and Samhouri, J. F. 2010. Developing meaning- D’Agrosa, C., Bruno, J. F., et al. 2008. A global map of human ful marine ecosystem indicators in the face of a changing climate. impact on marine ecosystems. Science, 319: 948–952. Stanford Journal of Law, Science, & Policy, 1: 36–48. Hannesson, R., Herrick, S., Jr, and Field, J. 2009. Ecological and eco- Levin, P.S., Fogarty, M. J.,Matlock, G. C., and Ernst, M. 2008. Integrated nomic considerations in the conservation and management of the ecosystem assessments. US Dept.of Commerce, NOAATech.Memo., Pacific sardine (Sardinops sagax). Canadian Journal of Fisheries NMFS-NWFSC-92. 20 pp. and Aquatic Sciences, 66: 859–868. Levin, P. S., Fogarty, M. J., Murawski, S. A., and Fluharty, D. 2009. Hilborn, R., Micheli, F., and De Leo, G. A. 2006. Integrating marine pro- Integrated ecosystem assessments: developing the scientific basis tected areas with catch regulation. Canadian Journal of Fisheries and for ecosystem-based management of the ocean. PLoS Biology, 7: Downloaded from Aquatic Sciences, 63: 642–649. e14. doi:10.1371/journal.pbio.1000014 Hildreth, R. G. 2008. Place-based ocean management: emerging US law Levin, P. S., and Wells, B. 2011. Discussion document: development of and practice. Ocean & Coastal Management, 51: 659–670. an annual report on conditions in the California Current ecosystem. Hipfner, J. M., Blight, L. K., Lowe, R. W., Wilhelm, S. I., Roberston, G. J., http://www.pcouncil.org. Barrett, R. T., Anker-Nilssen, T., et al. 2012. Unintended conse- Levin, S. A., and Lubchenco, J. 2008. Resilience, robustness, and marine quences: how the recovery of sea eagle Haliaeetus spp. populations ecosystem-based management. BioScience, 58: 27–32. http://icesjms.oxfordjournals.org/ in the northern hemisphere is affecting seabirds. Marine Loring, P. A. 2013. Alternative perspectives on the sustainability of Ornithology, 40: 39–52. Alaska’s commercial fisheries. Conservation Biology, 27: 55–63. ICES. 2011. Report of the Working Group on Ecosystem Assessment of Western European Shelf Seas (WGEAWESS). 3–6 May 2011, Link, J. S. 2005. Translation of ecosystem indicators into decision cri- Nantes, France. 175 pp. teria. ICES Journal of Marine Science, 62: 569–576. ICES. 2013a. Report of the Working Group on Integrated Assessments Link, J., and Brodziak, J. (Eds.). 2002. Report on the status of the NE US of the North Sea (WGINOSE). 11–15 February 2013, Lisbon, Continental Shelf Ecosystem. NEFSC Ecosystem Status Working Portugal. 76 pp. Group. Northeast Fisheries Science Center Ref. Doc. 02-11. 245 pp. ICES. 2013b. Report of the Working Group on the ICES ACOM/ Manski, C. F. 2013. Diagnostic testing and treatment under ambiguity: SCICOM Workshop on Ecosystem Overviews (WKECOVER). using decision analysis to inform clinical practice. Proceedings of the 7–11 January 2013, ICES HQ, Copenhagen, Denmark. National Academy of Sciences of the United States of America, 110: by Howard Browman on July 2, 2014 2064–2069. ICES. 2013c. Report of the Workshop on Benchmarking Integrated Ecosystem Assessments (WKBEMIA). 27–29 November 2012, McClanahan, T. R., Graham, N. A. J., MacNeil, M. A., Muthiga, N. A., ICES Headquarters, Copenhagen, Denmark. 27 pp. Cinner, J. E., Bruggemann, J. H., and Wilson, S. K. 2011. Critical Jackson, J. B. C. 2011. Ecological extinction and evolution in the brave thresholds and tangible targets for ecosystem-based management new ocean. Proceedings of the National Academy of Sciences of of coral reef fisheries. Proceedings of the National Academy of the United States of America, 105: 11458–11465. Sciences of the United States of America, 108: 17230–17233. James, C. A., Kershner, J., Samhouri, J. F., O’Neill, S., and Levin, P. S. McLeod, K. L., Lubchenco, J., Palumbi, S. R., and Rosenberg, A. A. 2005. 2012. A methodology for evaluating and ranking water quantity Scientific consensus statement on marine ecosystem-based manage- indicators in support of ecosystem-based management. Environ- ment. Signed by 221 academic scientists and policy experts with rele- mental Management, 49: 703–719. vant expertise and published by the Communication Partnership for Science and the Sea at http://compassonline.org/?q=EBM. Jennings, S. 2009. The role of marine protected areas in environmental management. ICES Journal of Marine Science, 66: 16–21. McLeod, K. L., and Leslie, H. 2009. Ecosystem-based Management for the Oceans. Island Press, Washington, DC. Johnson, E. J., and Goldstein, D. 2003. Do Defaults Save Lives? Science, 302: 1338–1339. Methratta, E. T., and Link, J. S. 2006. Evaluation of quantitative indica- Kaplan, I. C., and Leonard, J. 2012. From krill to convenience stores: tors for marine fish communities. Ecological Indicators, 6: 575–588. forecasting the economic and ecological effects of fisheries. Marine Mo¨llmann, C. In press. Implementing ecosystem-based fisheries man- Policy, 36: 947–954. agement—from single-species to integrated ecosystem assessment Kareiva, P., and Marvier, M. A. 2011. Conservation Science. Roberts and and advice for Baltic Sea fish stocks. ICES Journal of Marine Science. Company, Greenwood Village, CO. Morris, W. F., Bloch, P. L., Hudgens, B. R., Moyle, L. C., and Katsanevakis, S., Stelzenmu¨ller, V., South, A., Sørensen, T. K., Jones, Stinchcombe, J. R. 2002. Population viability analysis in endangered P. J. S., Kerr, S., Badalamenti, F., et al. 2011. Ecosystem-based species recovery plans: past use and future improvements. Ecological marine spatial management: review of concepts, policies, tools, Applications, 12: 708–712. and critical issues. Ocean & Coastal Management, 54: 807–820. Obama, B. 2010. Executive Order 13547. Stewardship of the ocean, our Keiter, R. B. 1998. Ecosystems and the law: toward an integrated ap- coasts, and the Great Lakes. Executive Order 13547. proach. Ecological Applications, 8: 332–341. Olson, R. 2009. Don’t Be Such a Scientist: Talking Substance in an Age of Kershner, J., Samhouri, J. F., James, C. A., and Levin, P. S. 2011. Selecting Style. Island Press, Washington, DC. indicator portfolios for marine species and food webs: a Puget Sound Otley, D. 1999. Performance management: a framework for manage- case study. PLoS ONE, 6: e25–248. ment control systems research. Management Accounting Research, Kittinger, J. N., Finkbeiner, E. M., Glazier, E. W., and Crowder, L. B. 10: 363–382. 2012. Human dimensions of coral reef social-ecological systems. POC. 2003. America’s living ocean: charting a course for sea change. A Ecology and Society, 17: 17. report to the nation. Pew Trusts, Washington, DC. Lessons learned from developing IEAs in the USA 1215

Pollnac, R., Christie, P., Cinner, J. E., Dalton, T., Daw, T. M., Forrester, Shin, Y-J., Shannon, L. J., Bundy, A., Coll, M., Aydin, K., Bez, N., G. E., Graham, N. A. J., et al. 2010. Marine reserves as linked Blanchard, J. L., et al. 2010b. Using indicators for evaluating, com- social-ecological systems. Proceedings of the National Academy of paring, and communicating the ecological status of exploited Sciences of the United States of America, 107: 18262–18265. marine ecosystems. 2. Setting the scene. ICES Journal of Marine PSP. 2009. Public Involvement and outreach summary. Puget Sound Science, 67: 692–716. Partnership. http://psp.wa.gov/aa_action_agenda.php. Smith, A. D. M., Fulton, E. J., Hobday, A. J., Smith, D. C., and Shoulder, Rice, J. 2005. Will an ecosystem approach mitigate the factors of unsus- P. 2007. Scientific tools to support the practical implementation of tainability? In International Workshop on the Implementation of ecosystem-based fisheries management. ICES Journal of Marine the International Fisheries Instruments and Factors of Unsustain- Science, 64: 633–639. ability and Overexploitation in Fisheries. Siem Reap, Cambodia, Smith, T. D. 1994. Scaling Fisheries: The Science of Measuring the 13–16 September 2004. Ed. by J. Swan, and D. Gre´boval. FAO, Effects of Fishing, 1855–1955. Cambridge University Press, Rome, Italy. http://www.fao.org/docrep/009/a0312e/A0312E12 New York. .htm#ch2.10. Tallis, H. M., Levin, P. S., Ruckelshaus, M., Lester, S. E., McLeod, K. L., Rice, J. C. 2011. Achieving coherent policies for conservation and sus- Fluharty, D. L., and Halpern, B. S. 2010. The many faces of tainable use of marine ecosystems. Conservation Biology, 25: ecosystem-based management: making the process work today in 1065–1068. real places. Marine Policy, 34: 340–348. Rice, J. C., and Rochet, M. J. 2005. A framework for selecting a suite of Tallis, H.T., Ricketts, T., Guerry, A.D., Wood, S. A., Sharp, R., Nelson, E., indicators for fisheries management. ICES Journal of Marine Ennaanay, D., et al. 2013. InVEST 2.5.5 User’s Guide. The Natural

Science, 62: 516–527. Downloaded from Capital Project, Stanford. Rockstro¨m, J., Steffen, W., Noone, K., Persson, A˚ ., Chapin, F. S. I., Lambin, E., Lenton, T. M., et al. 2009. Planetary boundaries: explor- Thaler, R. H., and Sunstein, C. R. 2003. Libertarian paternalism. ing the safe operating space for humanity. Ecology and Society, 14. American Economic Review, 93: 175–179. http://www.ecologyandsociety.org/vol14/iss2/art32/. Turner, N. J. 2003. The ethnobotany of edible seaweed (Porphyra abbot- Rosenberg, A. A., Mooney-Seus, M. L., Kiessling, I., Mogensen, C. B., tae and related species; Rhodophyta: Bangiales) and its use by First Nations on the Pacific Coast of Canada. Canadian Journal of

O’Boyle, R., and Peacey, J. 2009. Lessons from national-level imple- http://icesjms.oxfordjournals.org/ mentation across the world. In Ecosystem-Based Management for Botany, 81: 283–293. the Oceans, pp. 294–313. Ed. by K. McLeod, and H. Leslie. Island Turner, N. J.,Gregory, R., Brooks, C., Failing, L., and Satterfield, T. 2008. Press, Washington, DC. From invisibility to transparency: identifying the implications. Rosenberg, A. A., and Sandifer, P. A. 2009. What do managers need? In Ecology and Society, 13: 7. Ecosystem-based management for the oceans, pp. 13–30. Ed. by UNEP and IOC-UNESCO. 2009. An Assessment of Assessments, K. McLeod and H. Leslie. Island Press, Washington, DC. Findings of the Group of Experts. Start-up Phase of a Regular Ruckelshaus, M., and McClure, M. 2007. Sound science: synthesizing Process for Global Reporting and Assessment of the State of the ecological and socioeconomic information about the Puget Marine Environment Including Socio-economic Aspects. Sound ecosystem. US Dept. of Commerce, National Oceanic and USCOP. 2004. An ocean blueprint for the twenty-first century. United Atmospheric Administration (NMFS), Northwest Fisheries States Commission on Ocean Policy, Washington, DC.

Science Center, Seattle, WA, USA. 93 pp. USNOC. 2013. National ocean policy implementation plan. United by Howard Browman on July 2, 2014 Sainsbury, K. J., Punt, A. E., and Smith, A. D. M. 2000. Design of oper- States National Ocean Council, Washington, DC. ational management strategies forachievingfisheryecosystem objec- Vaske, J., Donnelly, M., and Shelby, B. 1993. Establishing management tives. ICES Journal of Marine Science, 57: 731–741. standards: selected examples of the normative approach. Samhouri, J. F. 2012. Risk assessment for ecological integrity. In Environmental Management, 17: 629–643. Integrated Ecosystem Assessment of the California Current. Ed. by P. S. Levin, and B. Wells. U.S. Dept. of Commerce, NMFS- Watson, R. T. 2005. Turning science into policy: challenges and experi- NWFSC, Seattle. ences from the science–policy interface. Philosophical Transactions of the Royal Society B: Biological Sciences, 360: 471–477. Samhouri, J. F., Lester, S. E., Selig, E. R., Halpern, B. S., Fogarty, M. J., Longo, C., and McLeod, K. L. 2012. Sea sick? Setting targets to White, C., Costello, C., Kendall, B. E., and Brown, C. J. 2012a. The value assess ocean health and ecosystem services. Ecosphere, 3: art41. of coordinated management of interacting ecosystem services. Ecology Letters, 15: 509–519. Samhouri, J. F., Levin, P. S., and Ainsworth, C. H. 2010. Identifying thresholds for ecosystem-based management. PLoS ONE, 5: e8907. White, C., Halpern, B. S., and Kappel, C. V. 2012b. Ecosystem service Samhouri, J. F., Levin, P. S., James, C. A., Kershner, J., and Williams, G. tradeoff analysis reveals the value of marine spatial planning for mul- 2011. Using existing scientific capacity to set targets for ecosystem- tiple ocean uses. Proceedings of the National Academy of Sciences of based management: a Puget Sound case study. Marine Policy, 35: the United States of America, 109: 4696–4701. 508–518. Zador, S. (Ed.). 2012. Ecosystem Considerations 2012. North Pacific Shin, Y-J., Bundy, A., Shannon, L. J., Simier, M., Coll, M., Fulton, E. A., Fisheries Management Council, Anchorage, AK. 230 pp. Link, J. S., et al. 2010a. Can simple be useful and reliable? Using eco- Zinn, H. C., Manfredo, M. J., Vaske, J. J., and Wittmann, K. 1998. Using logical indicators for representing and comparing the states of normative beliefs to determine the acceptability of wildlife manage- marine ecosystems. ICES Journal of Marine Science, 67: 717–731. ment actions. Society & Natural Resources, 11: 649–662.

Handling editor: Howard Browman