Tinkering with a tanker—slow evolution of a Swedish ecosystem approach

The ecosystem approach is a salient policy paradigm originating from a scientific understanding of the reality of complex ecosystem dynamics. In this article, we investigate how Swedish national marine policies and practice between 2002 and 2015 have changed towards an ecosystem approach. Government documents, the scientific literature, institutional changes, changes in legislation, pilot projects, and changes in science and public opinion were reviewed and combined with information from expert interviews. We found that changes in policy and practice have slowly stimulated the development of an ecosystem approach, but that limited political leadership, challenges of coordination, different agency cultures, and limited learning appears to be key barriers for further and more substantial change. We compare and contrast the Swedish national process of change with other documented experiences of implementing an ecosystem approach and find that several countries struggle with similar challenges. Substantial work still remains in Sweden and we provide suggestions for how to stimulate further and more substantial change at the national level.

Introduction

A scientific paradigm shift is defined as a revolutionary change due to new and surprising insights (Kuhn, 1962). Shifts in policy can also be described at three hierarchical levels, where a first-order change includes continuous policy adaptations, a second-order change includes alterations of policy instruments without changing the underlying policy goals, and a third-order change (analogous to a paradigm shift) include a complete re-organization of the policy goals (Hall, 1993). Large-scale shifts in natural resource governance have been termed “transformations”, which in many ways can be described as paradigm shifts (Olsson et al., 2006). In such shifts, new visions and approaches have changed the way in which people relate to ecosystems and how stakeholders interact.

A shift to an ecosystem approach to marine resource management has been described as a paradigm shift (Browman and Stergiou, 2005), from a traditional, top-down and sector-based approach, to a multilevel, cross-sectoral, and integrated approach (McLeod and Leslie, 2009; Tallis et al., 2010; Berkes, 2012; Engler, 2015). The demand for this shift is resulting from a growing recognition that sector based, “command and control” management (Holling and Meffe, 1996) is insufficient and that science, policy and practice need to co-develop approaches to address messy and “wicked” environmental problems (Jentoft and Chuenpagdee, 2009; Berkes, 2012). However, despite a broad recognition of the necessity for an ecosystem approach (Browman and Stergiou, 2005; Engler, 2015), establishing marine zoning and other forms of governance that require coordination across sectors is often constrained by inflexible institutions, unclear rules and responsibilities, power dynamics, lack of public support, insufficient legislation and limited learning (Leslie and McLeod, 2007; Kittinger et al., 2011; Richter et al., 2015).

In this review, we describe how the ecosystem approach has developed in Sweden between 2002 and 2015. We used government documents and scientific reports, reviewed changes in relevant institutions, legislation, and public opinion, and studied pilot projects. We also carried out eighteen interviews with politicians, parliamentarians and experts with present or previous employment in relevant agencies and ministries, including in the Swedish parliament and government office (six interview), the Environmental Protection Agency (EPA, six interviews), and the Swedish Agency for Marine and Water Management (SWaM, six interviews). Three of the latter had previously worked at the National Board of Fisheries (NBoF).

The starting point for this study is the “Commission on the Marine Environment”, (CME) established by the Swedish Government in 2002 (Figure 1) and tasked to work for nine months to propose measures for major policy change, based on an ecosystem approach. The Minister of Environment at the time wanted to make the marine environment a political priority:

Figure 1

Summary of key Swedish marine policy and practice developments.

“In the coming years the most important issue in our surroundings will be the marine environment. My ambition is that the commission lays the foundation necessary to lift up marine environmental issues” (SOU, 2003).

First, we describe how CME proposals have been associated with tangible outcomes in policy and practice. Then, we compare these developments to other experiences with operationalizing an ecosystem approach, specifically in relation to a number of “consensus elements” or “key principles” of the ecosystem approach (Engler, 2015; Long et al., 2015) and case studies describing the process of change towards more integrated and ecosystem-based management approaches (e.g. Schultz et al., 2015). Finally, we discuss whether observed changes can be described as a paradigm shift, and provide suggestions for how further change could be stimulated.

Development of a Swedish ecosystem approach (2002–2015)

Swedish Government Official Reports (“Statens Offentliga Utredningar”, or SOU), such as the one prepared by CME, provide a mechanism for expert input and external policy advice for the government. Policy makers are not bound by these recommendations, but act on them if they align with political priorities. All suggestions for policy change are negotiated within the government offices, including all the ministries concerned, and the parliament, which inherently leads to trade-offs between multiple political priorities. A Minister of Environment, for instance, will thus act on recommendations published in a Government Official Report to the extent he or she is willing and able to do so. When significant recommendations are operationalized, a Government bill (“Proposition”) is prepared and made into law or regulation through adoption in the parliament.

The CME proposals aimed at both integrating across sectors, and improving the environmental performance of individual sectors (e.g. fisheries or agriculture). Here, we focus on the integrated and area-based aspects of the ecosystem approach, rather than progress in individual sectors, see (Engler, 2015). The main proposals focused on (i) improving government agency leadership and coordination, (ii) developing a national marine strategy, with a particular focus on inclusive spatial planning, (iii) increasingly using adaptive management, (iv) stimulating public awareness, and (v) improving the multidisciplinary scientific basis and creating a stronger link between science and policy (Supplementary Table S1). We describe how additional investigations, policy- and action plans, policy and legal change, reorganization of government agencies, and the scientific community developed along the lines of these proposals.

Establishing leadership and improving coordination across sectors

CME identified the lack of a designated lead agency and lack of coordination across sectors as a key barrier for implementing the ecosystem approach. To overcome this barrier, the CME recommended that the EPA should be given an overarching responsibility for the marine environment (Supplementary Table S1). However, the subsequent Government Official Report Developed Management of the Marine Environment (SOU 2008:48, Supplementary Table S2) concluded that the EPA did not cover all necessary aspects to legitimately act as the central agency. The Government Official Report An Agency for Water and Marine Management (SOU 2010:8, Supplementary Table S2) and the Government bill Investigation Regarding Establishment of a New Agency for Marine and Water Management (Anonymous, 2010) described how a new agency would incorporate all relevant responsibilities carried out by the EPA and merge them with the NBoF.

A new national agency, SWaM, with a broad mandate to manage marine and freshwater issues (including spatial planning) was established in June 2011 (Figure 1 and Supplementary Table S2). The purpose was to improve and strengthen marine management thorough a holistic and comprehensive agency for all water and marine issues.

Despite the Government’s ambitions to integrate across sectors; however, freshwater management was not integrated with marine management when SWaM was established. Freshwater management in Sweden is guided by the implementation of the European Water Framework directive (2000/60/EG) (Anonymous, 2000). In 2004, the operationalizing of the directive resulted in the adoption of national regulation 2004:660 (Anonymous, 2004) and the establishment of five regional water agencies. When SWaM was established in 2011, they were tasked with coordination and development of the management of freshwater resources, carried out by the five regional water agencies. Our interviews suggest that these regional agencies are perceived as unable to deliver the active local and regional management that was intended. Instead, regional interest groups have influenced priorities and their effectiveness is being reviewed.

Marine policy is also guided by Sweden’s environmental quality objectives (Government bill 2000/01:130, 2004/05:150, and 2009/10:155) which cover all areas of environmental policy—from unpolluted air, lakes free from eutrophication and acidification, to functioning forest and farmland ecosystems. For each objective there are a number of specifications, clarifying the desired state of the environment which is to be attained (EPA, 2012). The desired state has to be achieved within one generation (the generational goal) and all environmental policy should be directed towards ensuring that (EPA, 2012). EPA coordinates these policies, but the responsibility for freshwater and marine resource management was transferred to the regional water agencies and SWaM respectively, and then reported to EPA.

In response to the operationalizing of the government bills on environmental quality objectives in 2004–2011, the EPA developed substantial competence in the field of marine resource management. With the establishment of SWaM, these responsibilities were transferred from Stockholm, on the East Coast of Sweden, to the new agency in Gothenburg, on the West Coast. The move was carefully prepared with written documentation of each policy area, as very few staff members were willing to make the move.

The new agency was established together with the NBoF. The NBoF had, for decades, worked closely with and supported the fishing industry (Ask et al., 2015), which resulted in frequent and heated conflicts with both EPA and environmental non-governmental organizations (NGOs). The last Director General of the NBoF (2005–2011) had made a concerted effort to “green the culture” of the agency. However, the interviews all describe it as a major challenge to bring cohesion to the new agency where the NBoF culture “was infused in the walls” and “resisted” the new and holistic ecosystem-based approach (e.g. regarding fish as part of a greater ecosystem rather than just as a commodity). Moreover, interviewees expressed frustration with SWaM for lack of experience with relevant topics and inability to deliver with adequate quality.

SWaM was thus forced to operate in a fragmented institutional landscape, and was influenced by substantial loss of institutional memory and conflicting agency cultures. According to our interviews, the agency was perceived as unable to provide the necessary leadership and its perceived inefficiency was a conversation piece among external stakeholders, both public and private. The agency underwent a major reorganization in 2013 aimed to address some of these challenges (although internal documentation of SWaM did not specifically highlight that cultural differences represented a challenge).

Given the strong emphasis on different agency cultures in our interviews, and the history of conflict and tension between different agencies and actors, we found it surprising to note that the preparatory work for proposing the establishment of SWaM [An Agency for Water and Marine Management (SOU 2010:8), see Supplementary Table S2], primarily focused on cost effectiveness and administrative advantages, and that it did not address cultural differences between NBoF and EPA.

Developing a system of inclusive spatial planning

Establishing a system of spatial plans was perceived by CME as a main tool to go from piecemeal and reactive management, to an integrated and proactive approach (Supplementary Table S1). Spatial planning has received substantial attention in subsequent government activities, including in the National Strategy from 2004 (Figure 1), but most significantly through the Government Official Report Developed Management of the Marine Environment (SOU 2008:48) (conducted between 2006 and 2008, Figure 1), the bill A Congruent Marine Policy for Sweden 2008/09:170 (presented in 2009, Figure 1), the Government Official Report Deep Planning—Physical Mapping of the Sea (SOU 2010:91) (conducted between 2009 and 2010) and the spatial planning bill Householding Marine Areas 2013/14:186 (presented to and passed by Parliament in 2014, Figure 1, see also Supplementary Table S2).

Spatial planning of the marine environment hardly existed in Sweden prior to 2002 but is becoming a national priority for the government and its agencies (SWaM, 2012, 2015). The adoption of a legal framework has made the development of spatial plans for three major areas of the Swedish economic zone possible and will be presented to the government in 2019. It has been a long process and the links to other relevant legislation have been thoroughly addressed. SWaM receives a clear mandate and responsibility in implementation and coordination. There is, however a lack of clear directives in case of conflict or dissenting views between the central agency and local or regional authorities, which may prove to be problematic. It therefore remains to be seen whether spatial planning will become an inclusive and effective tool for improving ecosystem integrity and sustainability, or whether it will primarily represent a planning tool for area-based marine activities.

Learning by doing: adaptive management

The CME emphasized the importance of adaptive management and here, we identify a number of relevant initiatives that has taken place during the time period (Table 1 and Supplementary Table S3), all to various extent, designed to learn from, and evaluate practical experience with ecosystem-based management approaches. For instance, the NBoF initiated and co-ordinated six pilot projects in 2004 of local fisheries co-management, designed to test and evaluate how this management approach could complement existing approaches (Figure 1). An evaluation of this initiative in 2006 (when it was formally discontinued) illustrated that users actively engaged in adaptive and learning based approaches, and by doing so, were able to developed shared visions and novel forms of collaboration. They were also able to better mitigate conflicts and agree on measures to address perceived problems (NBoF, 2007). One of these co-management initiatives (designed to balance impacts from small-scale Northern shrimp Pandalus borealis trawling with consideration to highly vulnerable cold water corals) has been institutionalized through the establishment of the first Swedish marine national park in 2009. The NBoF proposed to institutionalize fisheries co-management as a general method for managing Swedish coastal fisheries (NBoF, 2007), but this proposal has not led to any action by the government.

Table 1

Getting the ingredients right. How does 8 “consensus elements” and 15 “key principles” of the ecosystem approach compare to Swedish observations?

Eight consensus elements (Engler, 2015)	Fifteen key principles (Long et al., 2015)	The Swedish equivalent
Ecosystem approach is a holistic, or system, approach	Consider Ecosystem Connections	Holistic and system-based approaches in science, policy and practice are emerging, but are still in an early stage of development.
	Integrated management
Humans are part of nature	Recognize coupled social-ecological systems	Policies generally tended to treat humans as an external disturbance to the marine environment, rather than as an integrated part
Place-based management with ecologically defined boundaries	Distinct boundaries	Current focus on marine spatial planning will put a strong focus on place-based management. Local experiences of co-management can contribute with important knowledge
Management should be decentralized to the lowest appropriate level	Appropriate spatial and temporal scales	Swedish municipalities have a strong mandate to develop their own policies, guided by national policies
Management should be based on collaborative decision making	Stakeholder involvement	Most of the development 2002–2015 has been top-down and expert driven, coupled to national consultations of policy development and co-management
	Decisions reflect societal choice
Management should focus on the long term	Sustainability	Sweden operates under a system of 16 environmental quality objectives http://www.miljomal.se/sv/Environmental-Objectives-Portal/, with specific targets for the marine environment, eutrophication, biodiversity, toxic pollutants, etc. They should guide all policies in all sectors and be fulfilled by 2020.
	Ecological integrity and biodiversity
Ecosystem approach is knowledge based	Use of scientific knowledge	Substantial national and international funding has been devoted to increase the knowledge base for the marine environment. Most scientific efforts are however undertaken within specific disciplines.
	Interdisciplinary
Dealing with uncertainty: precautionary approach and adaptive management	Adaptive management	A number of attempts to use adaptive management have been carried out at smaller geographical scales. Adaptive management, which explicitly embraces uncertainty and coupled to appropriate monitoring, is still limited.
	Accounting for dynamic nature of ecosystems
	Appropriate monitoring
	Acknowledge uncertainty

In 2008, the Swedish EPA analogously initiated five pilot projects (Figure 1, Supplementary Table S3) tasked to develop collaborative and multiple stakeholder-based local and adaptive management plans (EPA, 2011). These pilot projects were discontinued in 2011 but a formal evaluation (Norrby et al., 2011) and scientific studies (Sandström et al., 2014,, 2015; Bodin et al., 2016) highlighted that important social capital and learning were developed in all areas. However, given the long time it takes to establish collaboration between diverse interests, it appears as if the short time period during which the pilot projects were in operation was insufficient (Sandström et al., 2014). For instance, measures implemented by the EPA previously in these regions, but without stakeholder involvement and perceived as having limited legitimacy, had generated mistrust that took time to overcome. Unclear instructions from the EPA and uncertainty about the purpose of the pilot projects also delayed action. However, large personal commitment from project leaders, stand out as a key factor for making progress, as were previous experiences in organization around similar issues (Ö. Bodin, pers. comm.).

Scientific experiments can also help developing practical experiences with adaptive management and the ecosystem approach (Österblom et al., 2016). A science-based adaptive management experiment, focusing on nutrient mitigation strategies coupled to a major nutrient reduction plant, illustrates the importance of collaborative learning between scientists and practitioners (Franzén et al., 2011). Philanthropy-funded biomanipulation initiatives to reduce eutrophication (Elmgren et al., 2012) or improve water quality through the introduction of predatory fish, i.e. pikeperch Sander lucioperca to trigger a trophic cascade (Hansson, 2013) represent notable examples where management measures have been designed as closely monitored experiments. Research and testing of measures to bind phosphorous in sediments by applying aluminium appear promising for improving water quality, but the introduction of predatory fish has not generated any measurable effects (Hansson, 2013).

The Swedish government also led an ambitious and long-term (7 years) ecological adaptive management experiment called “PLAN-FISH” (Planktivore management – linking food web dynamics to fisheries in the Baltic Sea) in order to investigate the strength of trophic cascading effects (Figure 1). The project investigated whether targeted fisheries for sprat Sprattus sprattus could be used as a method for improving the prospects for the recovery of cod Gadus morhua and other predatory fish in the ecosystem (scientific studies had indicated that a large sprat stock would e.g. compete with larval cod for food). The project developed important insights in relation to the seasonal strength of top-down, versus bottom-up trophic cascades, as well as possible threshold levels above which the sprat stock would have negative effects on cod. These findings improved the prospects for understanding how targeted fishing could improve the recovery of depleted species and is currently processed in order to be further developed in to scientific advice when setting fishing quotas (Appelberg et al., 2013).

These examples illustrate that experiments with adaptive management provide important insights and stimulate an understanding of how to implement an ecosystem approach, while also highlighting challenges, including the importance of developing a long-term commitment from policy makers and funders, and the complexity of ecosystems.

Public awareness

CME underlined the importance of public awareness campaigns (Supplementary Table S1). The frequency and intensity of the national public debate increased substantially following a suggested cod-fishing moratorium in 2002, which became headline news and contributed to widespread public awareness, resulting in changes in consumer behaviour (partially spurred by a consumer guide developed by the World Wide Fund For Nature, WWF) (Crona et al., 2015). Re-occurring events of cyanobacterial blooms during the peak of summer vacations have further contributed to environmental awareness as they are frequently the subject of summer time newspaper headlines, combined with TV documentaries and regularly reoccurring NGO campaigns.

Swedish public opinion has been measured on an annual basis by the Society Opinion Media institute http://som.gu.se/som_institute. In 2013, a deteriorating marine environment was ranked second, after environmental degradation, but ahead of e.g. organized crime, increased xenophobia and high unemployment, as major concerns for Swedish citizens (Bergström and Oscarsson, 2014; Ekengren Oscarson and Bergström, 2014). The public in all Baltic Sea countries are aware of (and concerned with) the challenges in the Baltic Sea and the willingness to pay for an improved marine environment significantly exceeds the estimated costs of action (BalticSTERN, 2013). Judged by this high level of public awareness and interest, there appears to be a large interest for substantial investments to improve the marine environment.

Bridging multidisciplinary science and policy

CME emphasized that cross-sectoral and interdisciplinary science represented an important support for decision makers. Swedish marine ecosystem scientists were pioneers in integrated analyses of ecosystem dynamics and linkages between the ecosystem, economics and social changes (Hammer et al., 1993; Holmlund and Hammer, 1999; Jansson and Dahlberg, 1999). This work triggered wide scientific discussion, but the governance model in operation at that time, was arguably designed in a way that made cross-sectoral planning and decision-making difficult. A second wave of research triggered a renewed discussion on resilience, ecosystem change, ecological cascading effects and regime shifts (Österblom et al., 2007; Casini et al., 2008,, 2009).

The Government Official Report A Swedish Marine Institute (SOU 2006: 112) (Figure 1, Supplementary Table S2) proposed improving the performance of four regional research institutes (three of which were established already in 1989), under the umbrella of one national marine institute. The merger was expected to address a number of deficiencies in coordination and collaboration, improve the scientific support for design and evaluation of environmental monitoring, as well as contribute to multidisciplinary scientific syntheses and evaluations, which would support relevant agencies.

Consequently, the Marine Institute (HMI) was established in 2008 (in Gothenburg) but an evaluation of its activities between 2008 and 2012 concluded that coordination had not worked well and that cooperation had been limited (Statskontoret, 2013). The merger resulted in duplication of efforts and inefficient use of resources (Statskontoret, 2013). Competition between participating universities, insufficient governance mechanisms, strong internal cultures and insufficient (economic) incentives were all identified as reasons for the limited cooperation (Statskontoret, 2013).

The marine science centre at Stockholm University was one of the four research institutes connected to HMI. In 2013, Stockholm university established a multidisciplinary Baltic Sea Centre, with a similar focus to that of HMI. The evaluation of HMI (Statskontoret, 2013) concluded that although this centre represents an important addition of resources, its establishment also risk contributing to additional competition between HMI and Stockholm.

Swedish national funding agencies are supporting multidisciplinary research programmes, such as the BONUS programme, designed to stimulate international and multidisciplinary marine science in the Baltic Sea (Kononen et al., 2014), and existing integrated marine ecosystem assessments, models and indicators are advanced (Diekmann and Möllmann, 2010; Möllmann et al., 2013; Meier et al., 2014; Wulff et al., 2014; Lade et al., 2015; Undeman et al., 2015). These multidisciplinary efforts improve the prospects for better coordination and interaction between science and policy, since they can stimulate collaboration between academics and provide the ecosystem-level type of advice that decision makers are looking for. However, the future development of such activities, e.g. through collaboration between HMI and the Baltic Sea Centre (located at two of the largest universities and on the west and east coasts respectively), depends critically on how turf wars are addressed, how funding is distributed and how collaborative projects are co-conceived.

Why has the Swedish process of change been so slow?

Our review and interviews suggests that the national process of change towards a more holistic, adaptive, and ecosystem-based management approach has been moving along at a steady but slow speed. It has been argued that sectorial implementation of the ecosystem approach should be evolutionary (Cowan et al., 2012), whereas operationalization of the type of cross-sectoral and area-based management considered here, which requires substantial institutional and legal change, should rather be revolutionary (Berkes, 2012; Engler, 2015). Our review and comparison with identified “consensus elements” or “key elements” of the ecosystem approach (Table 1) indicates that progress has been made, but that changes are evolutionary rather than revolutionary. They resemble the minor adjustment of the course of a tanker, steadily moving along the same trajectory, but with substantial path dependency (Boonstra and de Boer, 2014). The extent to which recent adjustments influence the present course is unclear, as is when or how it will influence the “final” destination.

A recent synthesis of three case studies of adaptive and ecosystem-based management (Schultz et al., 2015) illustrates a common pattern of development over time. Specifically, that: (i) policy entrepreneurs have been able to reframe the perception of the ecosystem in question and create new “umbrella concepts” that served as a common vision for action, that (ii) bridging organizations could channel resources, mobilize knowledge and connect sectors and scales and that c) a crisis (of some form) contributed substantially to triggering change (Schultz et al., 2015). Adaptive management initiatives can however not only be guided by dynamics and processes from the bottom-up, but need to be adequately supported and framed by enabling legislation and appropriate institutions (Olsson et al., 2002), developed from the top-down. Such development requires political leadership.

Political leadership

The political leadership has changed twice during the study period, from a Social Democratic Party-led government in 2002, to a coalition of liberal and conservative parties in 2006, and a coalition of Social Democrats and the Green Party in 2014 (Figure 1). Starting from 2007, the Swedish government allocated additional resources to stimulate measures that were anticipated to be particularly effective to achieve the national marine environmental quality objectives (Regeringen, 2011; OECD, 2015). A recent evaluation of the effect of these and other measures however, could not identify any progress towards reaching the environmental quality objectives. In fact, there were no positive trends for either objective 4: A non-Toxic Environment, objective 7: Zero Eutrophication, or objective 10: A balanced Marine Environment, Flouring Coastal Areas and Archipelagos—the three key environmental objectives for the marine environment. The evaluation concluded that none of these objectives will be met by 2020 (EPA, 2015).

Our interviews indicated a wavering level of ambition among central agencies and inconsistent political will in the government office to enable more comprehensive change. Although a number of political actions aimed to improve the marine environment have been initiated during the time period, key Ministers have been unable, or unwilling, to make the marine environment a main, long-term, priority. Strong political leadership has thus been lacking. The Swedish Society for Nature Conservation (a major national environmental NGO) studied the environmental policy promises made by parties in the parliament before the election 2010 and after 4 years in office 2014. The period, which was followed by another 4-year period with a conservative government, was labelled “Four lost years for the environment” and was characterized by loss of tempo, delays and changes in priorities (SSNC, 2014).

Potentially as a consequence of this, we also found discrepancies between expert-based advice and subsequent political decisions. The Government Report Developed Management of the Marine Environment (Figure 1), for instance, included recommendations for legally binding spatial plans and protection of biological diversity, whereas the bill Householding Marine Areas (Figure 1) included non-binding plans and a more utilitarian approach (Supplementary Table S2), consistent with priorities of the government in office at the time. A utilitarian approach is also prevalent in the “Blue Growth” concept, which is increasingly advocated for by national and international government agencies and among NGOs (COM, 2014; WWF, 2015).

Preceding the election in 2010, social-democratic Governments had been in power since 1994. During this period, the environment was a clear and strong policy focus under the umbrella of “building the green welfare state”, which gained popular approval. Global financial crisis and related national economic challenges reduced the interest for this focus as elections approached in 2010. The shifts in government and resulting fragmentation of decisions pertinent for coherent, integrated marine management have weakened the implementation and underlying intentions. An emphasis on jobs and growth, together with limited progress in reaching environmental goals, raise important concerns related to how sustainability of ecosystems are considered and integrated in practical policies. Similar limitations have also been observed for the implementation of the Australian Oceans Policy, where the complexity of issues, inadequate legislation and funding, and limited ownership of the policy process substantially limited the scope from original intents, e.g. when implementing spatial planning (Vince et al., 2015)

Non-state actors as policy entrepreneurs?

Non-state actors have mobilized capacity to compensate for the perceived limited political leadership and are demanding more action, but there is limited coordination between such actors. Individual annual campaigns, evaluations of policy measures, and collaborative scenarios (WWF, 2010; SSNC, 2014) are common products from such NGOs.

Two philanthropic foundations, BalticSea2020, established in 2006, and Zennström Philanthropies, established in 2007, where new organizations during this time period. BalticSea2020 focuses on improving the environmental condition of the Baltic Sea, using a private donation of 500 million SEK (50–60 million €) and intends to only be in operation during a limited time period (until 2020). The foundation has funded policy campaigns, television films and research projects (e.g. on biomanipulation). Zennström Philanthropies carried out campaigns, trying to fill the space not already claimed by existing, established NGOs. Today they work exclusively with a programme focusing on supporting local governments in their efforts to improve the Baltic Sea.

It seems that even with significant funding, neither the old or newer NGOs has been able to operate as a policy entrepreneur and stimulate transformative change (Schultz et al., 2015) on a large scale, partly perhaps because, umbrella concepts that integrate social innovation, social movements, socio-technical transitions, and ecosystem stewardship (Olsson et al., 2006,, 2014) have not been used or integrated in their activities. Limited coordination of efforts may also affect their ability to influence change.

Challenges with establishing bridging organizations

Legal and institutional change have been demonstrated as important for removing formal barriers for coordination across sectors in the United States (Gunderson et al., 1995), but e.g. different agency cultures or management styles, and informal power dynamics, can also represent important barriers (Kittinger et al. 2011). It takes time to become a credible coordinator and Swedish efforts to establish bridging organizations that support a transition towards more integrated management have not been successful. Neither SWaM, nor HMI, has yet been able to fill this role, nor any of the NGOs.

Recent geographical shifts in balance and power of scientific institutions and agencies have changed the conditions for different actors to engage in formal and informal dialogues between science and policy. Reorganization of scientific networks (Misund and Skjoldal, 2005; Stange et al., 2012) or government agencies elsewhere (Kittinger et al., 2011; Vince et al., 2015) illustrates similar challenges with turf wars, a lack of legitimacy and losses in institutional memory and capacity to those experienced in Sweden. Strategies and targeted efforts for mitigating barriers between powerful and potentially competing scientific institutions, or between old and new agencies, could improve institutional and collaborative learning (Kittinger et al., 2011). Co-development of priorities, and repeated exchange of knowledge and experiences (mentoring) between different scientific institutions and between the EPA and SWaM could be part of a strategy for improving collaboration and ensuring that institutional memory is valued and developed. Simple measures, such as more equal distribution of funding between HMI and participating institutions, would likely have improved scientific collaboration (Statskontoret, 2013).

The location of both SWaM and HMI on the Swedish west coast, away from the national centre of political power on the East Coast is analogous to the National Oceans Office Agency (NOO) in Australia, established to increase integration across sectors and facilitate implementation of the National Oceans Policy. NOO was located in Hobart, far from other key agencies in Canberra, which had adverse effects on its operations and eventually led to the loss of its executive agency status (Vince et al., 2015). Geographic co-location of relevant agencies appears to be a necessary, but not in itself, sufficient condition for the development of ecosystem-based management.

Using crises as opportunities

The status of the marine environment along the Swedish coasts is frequently highlighted as a major crisis in Swedish media and in reports from NGOs. Regular reports of anoxic sediments and depleted fish stocks have increased public awareness, but also likely contributed to policy fatigue. Limited ecological recovery potential (Österblom et al., 2007; Savchuk and Wulff, 2009; Blenckner et al., 2015) suggest that effects of (expensive) actions taken today may not be evident for decades, which may limit political will to invest substantially in the marine environment. However, crises that stimulate a transition to an ecosystem approach may not necessarily originate from the same geographical scale or even from the same policy area. Both ecological crisis (Olsson et al., 2008) and shifts in political power (Gelcich et al., 2010), can be important for “the opening of an opportunity” for change, provided that individual agents and policy entrepreneurs can skilfully “navigate the transition” (Folke et al., 2005). In the Swedish example, it is currently unclear whether or not such entrepreneurs exist, and/or whether crises associated with eutrophication or overfishing have been perceived as sufficiently serious to open up an opportunity for change.

Moving beyond tinkering

Marine spatial planning is emerging as an important tool around the world, but the original focus on ecosystems is often lost along the way as the tool is implemented (Merrie and Olsson, 2014). As Sweden applies spatial planning, ecosystem sustainability and environmental quality objectives should represent the foundation for decisions taken. It is also important to engage with the most recent thinking on adaptive and dynamic oceans management (Maxwell et al., 2015; Dunn et al., 2016), which could stimulate interesting technological developments for real-time management and novel collaboration between government agencies, scientists, policy entrepreneurs, technology firms, and financial actors.

The Swedish experiences with pilot projects to investigate the potential for adaptive management illustrate that important insights can be harvested from such social-ecological innovations (Olsson and Galaz, 2012). The relatively short time period during which these initiatives were operational, however, appear ill suited for the long-term processes associated with establishing trust and enabling change. Without formal (scientific) evaluation and synthesis of such pilot projects, important insights and social learning will be lost (Leslie and McLeod, 2007; Berkes, 2012). Adequate funding for associating such projects with science-based, monitoring and evaluation (of both social and ecological dynamics), in turn, could lay the foundation for novel and long-term investments to scale up and further develop ecosystem-based management initiatives.

The scientific literature is becoming increasingly clear with what the ecosystem approach can and should mean in practice (Curtin and Prellezo, 2010; Kittinger et al., 2014; Engler, 2015; Long et al., 2015; Richter et al., 2015). Diverse empirical insights of how interconnected social-ecological systems represent challenges and opportunities for implementing an ecosystem approach are rapidly accumulating (Murawski, 2007; Ruckelshaus et al., 2008; Österblom et al., 2010; Tallis et al., 2010; Leslie et al., 2015). Similar approaches for implementing the ecosystem approach in Australia, Canada, Norway (Sainsbury et al., 2014), and Sweden (this study) suggest that important insights could be derived from collaborative learning across countries.

The national policy changes described in this article indicate that there has been a distinct movement towards implementing the ecosystem approach in Sweden over the study period. We particularly observe changes in funding and policy instruments (first and second order outcomes). However, the changes have come about gradually and in an uncoordinated manner, in part due to political changes and we argue that the observed changes cannot be described as a third-level outcome or paradigm shift. The re-organization of SwAM illustrates the tension between gradual adaptions within the current policy paradigm (evolutionary, second level outcomes) vs. a paradigm shift (revolutionary, third level outcome). The agency was originally organized to break down the traditional barriers between fisheries management and management of the marine environment (i.e. with an intention to achieve an organizational structure that could better integrate different areas from a marine perspective), but by merely focusing on the marine environment, not comprehensively addressing the complexity of freshwater and land-based pollution sources affecting the ecosystem, combined with institutional inefficiency, the agency was re-organized towards a more sectorial structure.

A remaining question is whether a paradigm shift is needed to achieve holistic management of the marine environment (as has been argued by Berkes, 2012), or if the gradual changes observed in the article can eventually deliver the same end results? The answer to this question is yet unclear. We suggest that Sweden need to invest substantially in collaborative learning between central agencies and that scientific networks and NGOs think creatively about new ways to collaborate and coordinate their activities. Sweden should also engage in national and international innovation investments coupled to the marine environment, with a vision to move beyond evolutionary tinkering, see (Jacob, 1977). Such investments should be guided by clearly defined visions and objectives from political leaders and central agencies, while also harnessing the creative power and abilities of local and regional actors. Such engagement would provide international leadership, enable a national paradigm shift and stimulate innovative research. Importantly, such efforts would also substantially improve resource management and the chances of Sweden reaching its environmental quality objectives.

Conclusions

Efforts by Swedish scientists, policy-makers and practitioners have improved the prospects for making the ecosystem approach operational. New national legislation for marine spatial planning is in place and a new agency (SWaM) is developing capacity for cross-sectorial collaboration. Increasingly experienced practitioners and a high degree of public concern are creating conditions for making the ecosystem approach operational, but political leadership, learning and improved coordination is also required. We argue that the time period studied represents a phase of “preparing the system for change” (Folke et al., 2005), but further dynamics require leadership, coordination, bold experiments and possibly, a crisis of sort.

Supplementary data

Supplementary material is available at the ICESJMS online version of the article.

Acknowledgements

H.Ö. was a working member of the CME and KV led the secretariat of CME. This article benefited from the kind participation of multiple experts in interviews and constructive comments provided by Dr B. Bohman, Dr A. Nilsson, and Dr Ö. Bodin.

Funding

The study received support from the Baltic Ecosystem Adaptive Management (BEAM) Programme and Mistra, through a core grant to the Stockholm Resilience Centre.