Dilemma 1: imbalance of natural and social science

MPAs may be defined as ‘a clearly defined geographical space, recognized, dedicated and managed, through legal or other effective means, to achieve the long-term conservation of nature with associated ecosystem services and cultural values’ (Dudley Reference Dudley2008, p. 8).

This spatial management tool has many goals, ranging from environmental restoration, to tourism development, to preservation of historic sites (World Bank 2006). The rules associated with MPAs include complete prohibition of extractive activities to limitation of particular forms of fishing or other activities (World Bank 2006). MPA governance ranges from community-based, to co-management, to centralized approaches (Christie & White Reference Christie and White2007). The specific research methods used to design and evaluate vary in response to the MPA goals and form. MPAs developed to improve fisheries management may require stock assessments, in addition to habitat studies while MPAs developed for biodiversity conservation may emphasize habitat and life history studies. All MPA types benefit from social science studies of, for example, governance and social dynamics (Mascia Reference Mascia, Sobel and Dahlgren2001; Christie et al. Reference Christie, Buhat, Garces, White, Brechin, Wilshusen, Fortwangler and West2003c). MPA ecological goals are unattainable in the long term if social and governance dynamics are poorly understood and managed (Christie Reference Christie and Shipley2004). The need for balanced and integrated assessments is increasingly apparent and acknowledged in the practitioner and scientific literatures (Kaufman et al. Reference Kaufman, Bunce Karrer, Peterson, McLeod and Leslie2009; Shackeroff et al. Reference Shackeroff, Hazen, Crowder, MacLeod and Leslie2009). A relatively small, but growing, group of scholars and practitioners are researching critical MPA interdisciplinary knowledge gaps (for example Aswani & Hamilton Reference Aswani and Hamilton2004; Christie Reference Christie and Shipley2004; Cinner et al. Reference Cinner, Marnane, McClanahan and Almany2005; Pollnac et al. Reference Pollnac, Christie, Cinner, Dalton, Daw, Forrester, Graham and McClanahan2010), but interdisciplinary information is generally inadequate and underused in policy making.

A comprehensive MPA design and evaluation research programme would, at a minimum, consider a range of issues and social and natural science methods (Table 2). However, with the possible exception of the Great Barrier Reef National Marine Park, few MPAs have undergone such this type of interdisciplinary review.

Table 2

MPA-related research activities. NA = not applicable.

The relative balance of research remains skewed toward accounts of declining ocean conditions and debates over ecological impacts of marine and coastal policies and MPAs (Campbell et al. Reference Campbell, Gray, Hazen and Shackeroff2009). There are many plausible explanations for this lack of balance and integration between MPA-relevant social and natural science research. The next two dilemmas introduce two explanations.

Dilemma 2: Overreliance on a particular worldview and science-policy epistemic community

Levins and Lewontin (Reference Levins and Lewontin1985, p. 267) stated: ‘Scientists, like other intellectuals, come to their work with a worldview, a set of preconceptions that provides the framework for their analysis of the world. These preconceptions enter at both an explicit and an implicit level, but even when invoked explicitly, unexamined and unexpressed assumptions underlie them’. The goals and objectives of policies are profoundly influenced by shared social constructions of the ocean. Whether the ocean is conceived of as a place of commerce, military opportunity or locus for important biodiversity, is directly related to distinct social constructions which are historically embedded and discernable (Steinberg Reference Steinberg2001).

Berger and Luckmann (Reference Berger and Luckmann1967) wrote that ‘. . .reality is socially constructed and the sociology of knowledge must analyze the processes in which this occurs. . . It will be enough. . .to define ‘reality’ as a quality appertaining to phenomena that we recognize as having a being independent of our own volition (we cannot ‘wish them away’), and to define ‘knowledge’ as the certainty that phenomena are real and that they possess specific characteristics.’

Berger and Luckmann (Reference Berger and Luckmann1967) further clarified other empirically grounded and practical considerations, that:

• • knowledge and perceived reality are socially relative, and influenced by factors such as class, education, race and gender;

• • specific patterns of knowledge and perceived reality pertain to certain social contexts which are describable;

• • over time, subjective meanings tend to evolve into positions assumed to be factual;

• • and, most importantly, human thought is subject to the influences of its social context which leads to ideological positions of a social group.

The internalization of a particular worldview is unavoidable. Few scientists or policy makers are aware of the dimensions of their worldview regarding key variables such as knowledge and resource abundance. The lack of realization and understanding of the biasing tendencies of an individual's worldview is limiting, and raises the potential for conflicting and uncompromising ideological positions. The future of marine resource management, especially in a globalized and diverse context, will hinge on a degree of self-awareness, compromise and flexibility. Marine-policy relevant social research has usually focused on the resource users and constituencies associated with marine policy intervention rather than on the wider community conducting the science and setting policy priorities. Reviews of the culture of science and the science-policy complex, usually by critical theorists interested in science studies (for example Lowe Reference Lowe2006), generally remain outside the management and policy-making discourse.

The influence of worldviews on scientific research is complex, subtle and undeniable and occurs at individual and societal levels (Kuhn Reference Kuhn1970). Whether those perspectives become ideological, or a shared belief system, and influential depends on whether opportunities and resources emerge to create or influence potent epistemic communities through which to diffuse ideas (Haas Reference Haas1990; Morris & Mueller Reference Morris and Mueller1992; Rogers Reference Rogers2003). An epistemic community is an influential group of socially connected people who share common worldviews.

One such epistemic community influenced policy development for the Mediterranean Sea in a process that parallels the current discourse and process for global MPA networks. ‘The Med Plan's successful creation was promoted by a community of ecologists and marine scientists. . . According to the explanation suggested here, if a group with a common perspective is able to acquire and sustain control over a substantive policy domain, the associate regime will become stronger and countries will comply with it. Such groups are more likely to be consulted after a crisis, especially when policy makers are uninformed about the technical dimensions of the problem at hand and are uncertain about the costs and benefits of international cooperation. New national policies, often in compliance with the regime, would then reflect the interests of the group consulted and empowered, and the duration of the new policies would depend upon the group's ability to consolidate and retain its bureaucratic power. The substantive nature of the regime would reflect the group's cause-and-effect beliefs . . .’ (Haas Reference Haas1990, p. 380).

The ways in which environmental problems are constructed and, in turn, what types of information are considered critical to their resolution determine the range of policy options. Adler and Haas (Reference Adler and Haas1992, p. 374) concluded after reviewing various empirical studies that ‘The impact of epistemic communities is institutionalized in the short term through the political insinuation of their members into the policy making process and through their ability to acquire regulatory and policymaking responsibility and to persuade others of the correctness of their approach’. The EBM and MPA advocacy epistemic community has become remarkably influential and created a well-defined policy agenda which has diffused through influential conservation, resource management and donor institutions (and some government institutions). One example of the growing influence of the EBM epistemic community is the appointment of academic, marine biologist and enthusiastic MPA and EBM proponent Dr Jane Lubchenco (Lubchenco et al. Reference Lubchenco, Palumbi, Gaines and Andelman2003) as head of the US National Oceanographic and Atmospheric Administration (NOAA).

The discourse within and produced by this epistemic community to influence policy has distinct characteristics which make it particularly influential by some accounts (for example journal impact factor, raising awareness of ocean decline), but also potentially limited in its ability to develop feasible and IR-informed policies at a global scale. High impact scientific assessments present early accounts of how ocean conditions are declining on a large scale and what should be done to reverse these trends. The common form of a number of these articles (for example Pauly et al. Reference Pauly, Christensen, Dalsgaard, Froese and Torres1998; Roberts et al. Reference Roberts, Bohnsack, Gell, Hawkins and Goodridge2001; Lubchenco et al. Reference Lubchenco, Palumbi, Gaines and Andelman2003; Mora et al. 2003; Pandolfi et al. Reference Pandolfi, Bradbury, Sala, Hughes, Bjorndal, Cooke, McArdle, McClenachan, Newman, Paredes, Warner and Jackson2003) follows the outline of: (1) ocean conditions are declining globally and precipitously, (2) society is responsible for this decline, (3) MPAs are a preferred and under-realized policy option, and (4) ambitious systems of large MPAs are required. These MPAs are commonly offered as a means to return to historic baseline conditions in at least some places, a recommendation that suggests a particular relationship (as minimal as possible) and position of society with regard to nature (separate) (Campbell et al. Reference Campbell, Gray, Hazen and Shackeroff2009). It is notable that most of the high impact global assessments and proponents for ambitious global MPA systems are generated by scientists and conservationists from the global North. This may be owing partly to the fact that the Northern academic community has greater access to and interest in publishing in these scientific journals, or could be because the academic community in the South is less comfortable with such large-scale policy agendas and favours priorities that elevate the needs of impoverished people. This topic would benefit from additional investigation.

Advocating for global MPA networks commonly overlooks various important social considerations. How MPAs affect various and complex human communities and the level of acceptance of MPAs by a diverse society is little documented and examined. Alternatives, trade-offs and governance feasibility are rarely considered. A global network, especially one proposed by Northern scientists and conservation groups, is almost certain to create controversy in the global South (Agardy et al. Reference Agardy, Bridgewater, Crosby, Day, Dayton, Kenchington, LaVoley, McConney, Murray, Parks and Peau2003; Oracion et al. 2007; ICSF [International collective in Support of Fishworkers] Reference Kumar2008; Aswani et al. Reference Aswani, Christie, Muthiga, Mahon, Primavera, Cramer, Barbier, Granek, Kennedy, Wolanski and Hacker2011) and may create a sense of disempowerment and isolation in marginalized fishing communities in developing countries (Jones Reference Jones2009). The under-examination of power dynamics between social groups has been identified as a key challenge to functional governance (Jentoft Reference Jentoft2007). MPA-advocating studies tend to conflate utilitarian with biodiversity goals, giving the impression that biodiversity conservation and fisheries benefits, for example, may be equally maximized (Jones Reference Jones2006). In other words, all boats may (and likely will) not rise equally with MPA implementation and rent capturing by influential social groups is likely without mechanisms for equitable distribution of benefits (Christie Reference Christie and Shipley2004; Christie et al. Reference Christie, Lowry, White, Oracion, Sievanen, Pomeroy, Pollnac, Patlis and Eisma2005; Oracion et al. Reference Oracion, Miller and Christie2005).

Simplified arguments are often applied in the explanation of complex environmental problems and to advocate for particular solutions. Human overpopulation is frequently offered as the fundamental cause of Malthusian environmental collapse and suffering (Hardin Reference Hardin1968, Reference Hardin1974). Malthusian conceptualizations are cited as an explanation for overfishing (Pauly Reference Pauly1990). Disagreement with this line of reasoning may be viewed as disinterest in protecting the Earth's vulnerable ecosystems, as if arguing for the rights of marginalized people implies a lack of interest in protecting the world's environment (but see Vandermeer Reference Vandermeer1996 and Brechin et al. Reference Brechin, Wilshusen, Fortwangler and West2003 for alternative views). Constrained lines of reasoning, while readily understood, reduce and limit the role of IR within policy design or evaluation. The argument's construct allows for little meaningful engagement of social sciences beyond economics focused on determining ecological services associated with marine systems and MPAs (Campbell et al. Reference Campbell, Gray, Hazen and Shackeroff2009).

In conclusion, while consensus is emerging that ocean conditions are frequently in a state of decline and that well-designed and implemented MPAs have potential to redirect society's overexploitation of resources, the creation of partially, and mainly ecologically, informed MPA mandates is contested by those at the periphery of MPA science (Trist Reference Trist1999; Walley Reference Walley2004; Oracion et al. Reference Oracion, Miller and Christie2005; ICSF Reference Kumar2008). The current scientific discourse is heavily influenced by the perspectives of a particular epistemic community's views which tend to under-represent complicating social dynamics, alternative policy options and social science inputs. It is not merely a matter of overlooking social sciences, rather this group actively selects a particular narrative and suite of scientific information. This narrative agrees with their worldviews and social constructions of the environment (Berger & Luckman Reference Berger and Luckmann1967). The implications for people and ocean conditions are serious, since MPAs founded on an incomplete understanding are likely to fail in most places in the long term (Christie Reference Christie and Shipley2004; Christie et al. Reference Christie, Lowry, White, Oracion, Sievanen, Pomeroy, Pollnac, Patlis and Eisma2005; Charles & Wilson Reference Charles and Wilson2009; Pollnac et al. Reference Pollnac, Christie, Cinner, Dalton, Daw, Forrester, Graham and McClanahan2010). Policies based on dire assessments without consideration of critical process elements might ultimately undermine longer term and widespread MPA success (Agardy et al. Reference Agardy, Bridgewater, Crosby, Day, Dayton, Kenchington, LaVoley, McConney, Murray, Parks and Peau2003) and generate conservation islands surrounded by a sea of highly degraded areas (Agardy et al. Reference Agardy, Notarbartolo di Sciara and Christie2011; Vandermeer Reference Vandermeer1996).

Considerable IR effort will be required to uncover and systematically consider why the above narrative is generated, why it has become so influential, and how it shapes the range of management options available. This research should be conducted in the spirit of creating a full and balanced scientific understanding and feasible policies, rather than, on the one hand, dismissing the utility of scientific documentation of environmental decline or, on the other hand, justifying MPAs and EBM.

Dilemma 3: Emergent integration of natural and social sciences and inherent tensions

Many of the underlying epistemological foundations, methods, and goals of natural and social sciences are distinct (Pickett et al. Reference Pickett, Burch and Grove1999; Campbell Reference Campbell2005; Campbell et al. Reference Campbell, Gray, Hazen and Shackeroff2009). Excessively reductionist and positivist approaches to understanding biological or social processes have been challenged in the social and natural sciences, but for different reasons. There has been a movement toward synthetic thinking within the environmentally-focused natural sciences. The emergence of ecology and conservation biology is an attempt to view biological relations from a synthetic non-reductionist perspective (Vandermeer Reference Vandermeer1996). The emergence of resilience theories in ecology raises the importance of non-linear, dramatic and somewhat unpredictable change in ecosystems (Gunderson & Holling Reference Gunderson and Holling2001). Within the social sciences, the very role of scientific methods has been questioned. The emergence of postmodernism, a movement which questions Cartesian modes of research and objectivity, has been particularly influential in the social sciences. Political ecology, as an influential progeny of postmodernism, emphasizes issues of power, oppression and social construction of the environment (Peet & Watts Reference Peet and Watts2004; Vandermeer & Perfecto Reference Vandermeer and Perfecto2005). The critique of ‘mainstream’ environment management efforts, which are cast as hegemonic, is controversial, partly because it questions the privileged and influential role that natural science and scientists occupy in defining issues and prescribing solutions (Chapin Reference Chapin2004; Lowe Reference Lowe2006). In large part, postmodern social scientists are at the periphery of the environmental management and associated natural science discourse, while there are rare exceptions that generate discussions among donors and conservationists (Chapin Reference Chapin2004).

Besides epistemological debates, additional methodological and ethical constraints tend to separate natural and social sciences. For the ecological assessment of MPAs, the gold standard of study design is the before–after control-impact (BACI) design (Kaufman et al. Reference Kaufman, Bunce Karrer, Peterson, McLeod and Leslie2009), a standard rarely attained since pre-MPA data are rare and environmental conditions inside and outside MPAs vary. For social systems, pre and post data are almost never collected. In this sense, the challenges of natural and social MPA science overlap. But the social scientist faces unique challenges; isolating particular variables in complex social systems is exceedingly difficult and humans communicate about and adapt quickly to change (Pickett et al. Reference Pickett, Burch and Grove1999). People are engaged in many economic, social and cultural activities simultaneously in relation to any environmental condition or policy. In a complex world, the variables of interest cannot be adequately isolated, nor can social experiments be conducted which may have unintended negative consequences for society. For example, poverty may influence coastal community receptivity to MPAs, but poverty cannot be isolated from many other associated variables, and there is general unwillingness to apply the treatment (MPAs) across a spectrum of relatively wealthy to relatively poor communities as part of an ethically questionable social experiment.

The optimal means to generate social knowledge of power and oppression is debated within the social sciences. Traditional social science, which ‘extracts’ knowledge from informants, and participatory social science, which ‘co-creates’ knowledge with informants, coexist within the social sciences (Fals-Borda Reference Fals-Borda1987; Freire Reference Freire1993). While most academic institutions favour scientific over participatory methods, academic communities engaged in participatory research exist on most campuses with underpinnings from seminal works (see for example Freire Reference Freire1993). The emergence of participatory or rapid rural appraisals in the 1990s (Chambers Reference Chambers1994) and subsequent participatory coastal resource assessments (Deguit et al. Reference Deguit, Smith, Jatulan and White2004) represent the practical application of participatory theories and methods expressed by intellectuals largely from the global South (for example Fals-Borda Reference Fals-Borda1987; Freire Reference Freire1993).

Bridging the epistemological, methodological and ethical divides of the natural and social sciences or, greater yet, overcoming the opposing scientific and postmodern intellectual camps, is extraordinarily difficult. Perhaps the full integration of social and natural sciences is neither possible nor desirable, especially if important social MPA design criteria are obfuscated by ecological ones. Stoffle et al. (Reference Stoffle, Minnis, Murphy, Van Vlack, O'Meara, Smith and McDonald2010, an anthropologist, recommended separate but parallel social and natural science MPA research, to be eventually rectified in the design process. The natural tensions and distinctive perspectives within parallel processes may be useful and result in a plurality of understandings. Integrative analysis may explain how societies and individuals respond to environmental cues or environment shifts (Aswani & Hamilton Reference Aswani and Hamilton2004; Cinner et al. Reference Cinner, Marnane, McClanahan and Almany2005; Pollnac et al. Reference Pollnac, Christie, Cinner, Dalton, Daw, Forrester, Graham and McClanahan2010), but may not be the most effective approach to investigating other important processes such as the detailed understanding of formal institutional design, environmental histories, or class and race relations (Campbell et al. Reference Campbell, Gray, Hazen and Shackeroff2009).

Dilemma 4: Balancing mandate dependent and mandate independent research

The role of science within environmental policy making is changing. The marine scientist as an advocate of social policy is relatively new and simultaneously applauded and vilified (Pauly Reference Pauly2009). Discussions about desired environmental states, and more specifically MPAs, tend to polarize constituency groups, including researchers (Campbell et al. Reference Campbell, Gray, Hazen and Shackeroff2009). Some of the polarization may be attributed to the definitions used by ecologists and fisheries scientists to set thresholds for overexploitation (Hilborn Reference Hilborn2007) and differences between researchers’ interpretations of the goals and consequences of MPAs. The Souffrière MPA in St Lucia provides an illustrative example of polarization based on distinct interpretations of outcomes. Roberts (Reference Roberts, Bohnsack, Gell, Hawkins and Goodridge2001) characterized the Souffrière MPA in Saint Lucia as an unqualified success story with both positive ecological impacts and social acceptance. Sandersen and Koester (Reference Sandersen and Koester2000) provided a detailed description of the management process for this same MPA and identified management conflicts and difficulties. Trist (Reference Trist1999), a political ecology geographer, provided a detailed historic review of the evolution of the Caribbean from a place of subsistence to one of global tourism; the Souffrière MPA was characterized as part of this transformation, with marginalization of, and resistance by, local fishers. Notably, the more recent and influential paper by Roberts (Reference Roberts, Bohnsack, Gell, Hawkins and Goodridge2001) did not cite the work of the other more critical scholars. The distinct methods, perspectives and journals allow for narrow interpretations. The examination of reasons for such distinct interpretations is uncommon in the marine policy and scientific literature used by ocean policy makers. The emphasis has been on describing the proximate environmental condition and policy efficacy (what Christie et al. Reference Christie, Buhat, Garces, White, Brechin, Wilshusen, Fortwangler and West2003c refer to as ‘MPA mandate dependent’ topics), rather than examinations of the appropriateness of the policy (i.e. MPA) and why scientists reach distinct interpretations of common cases (examples of mandate independent topics). Scientists, including social scientists, aspire to studying both applied and basic questions. Stated more forcefully, social scientists working on environmental issues are not willing to limit their studies and research output to ‘solve problems’ of conservation and articulation of conservation planners and communities (Campbell Reference Campbell2005).

Dilemma 5: Balancing global and local studies

The widespread debates over whether MPAs are intended mainly for biodiversity conservation, fisheries management, or both (Hilborn et al. Reference Hilborn, Stokes, Maguire, Smith, Botsford, Mangel, Orensanz, Parma, Rice, Bell, Cochrane, Garcia, Hall, Kirkwood, Sainsbury, Stefansson and Walters2004; Jones Reference Jones2006) are rarely located within comparisons of contextual particulars. Increasingly, with the development of spatially explicit global datasets and global information system (GIS) enabled presentations, important large ecological changes have been discerned (Pauly et al. Reference Pauly, Watson and Alder2005; Worm et al. Reference Worm, Barbier, Beaumont, Duffy, Folke, Halpern, Jackson, Lotze, Micheli, Palumbi, Sala, Selkoe, Stachowicz and Watson2006; Halpern et al. Reference Halpern, Walbridge, Selkoe, Kappel, Micheli, D'Agrosa, Bruno, Casey, Ebert, Fox, Fujita, Heinemann, Lenihan, Madin, Perry, Selig, Spalding, Steneck and Watson2008). The sophisticated ecological trend analysis is not matched with similarly impressive analytics for ocean and coastal use, values and ecosystem services, governance and MPA policies, either as a result of limited data or by design (Campbell et al. Reference Campbell, Gray, Hazen and Shackeroff2009; Shackeroff et al. Reference Shackeroff, Hazen, Crowder, MacLeod and Leslie2009). The mismatch of scale for current global analyses with policy recommendation is methodologically multifaceted. First, the prioritization of MPA evaluative metrics is often arbitrary. The selection of data layers for GIS interpretations can be driven by the availability of data and influenced by subjective prioritization. For example, Mora et al. (Reference Mora, Andréfouët, Costello, Kranenburg, Rollo, Veron, Gaston and Myers2006) raised awareness of the condition of reefs and the inadequacy of coral reef MPA coverage. They advocated for large and closely-spaced MPAs across the world. This interpretation is based on a layering of data in descending order of importance: regulation on extraction, incidence of poaching, risk to external impacts, MPA size and isolation. Little beyond anecdotal justification for inclusion of variables was provided, including the supplemental methods section of the paper, and no MPA social science studies were cited, despite the policy subject matter (Mora et al. Reference Mora, Andréfouët, Costello, Kranenburg, Rollo, Veron, Gaston and Myers2006). The minimum size of the MPAs that met Mora et al.'s (Reference Mora, Andréfouët, Costello, Kranenburg, Rollo, Veron, Gaston and Myers2006) criteria for an effective coral reef MPA (an area ≥10 km²), was based solely on ecological criteria and thus they regarded most community-based MPAs (which are understandably below this size cut-off) as ineffectual. Such categorization contradicts the extensive positive accounts of moderately sized MPAs (Halpern Reference Halpern2003; Russ et al. Reference Russ, Alcala, Maypa, Calumpong and White2004) especially when networked (Gaines et al. Reference Gaines, White, Carr and Palumbi2010). The constrained selection of variables to measure the broad concept of MPA effectiveness is representative of a particular suite of disciplinary perspectives; mostly notably, that what really counts is the protection of the ecosystem.

Ideally both global and local studies inform understanding of MPA impact and strategy. The initial MPA literature consisted primarily of site-specific case studies, which limited potential for generalization (Russ & Alcala Reference Russ and Alcala1999). Subsequent meta (Halpern Reference Halpern2003) and comparative studies (Pollnac et al. Reference Pollnac, Crawford and Gorospe2001; Christie et al. Reference Christie, Pollnac, Fluharty, Hixon, Lowry, Mahon, Pietri, Tissot, White, Armada and Eisma-Osorio2009a) allowed for generalization. While meta and comparative studies elucidate general principles, robust inductive case studies provide the details necessary to understand local dynamics, which are clearly important to MPA impact and success (Pickette et al. 1999; Christie et al. Reference Christie, Pollnac, Fluharty, Hixon, Lowry, Mahon, Pietri, Tissot, White, Armada and Eisma-Osorio2009b; Shackeroff et al. Reference Shackeroff, Hazen, Crowder, MacLeod and Leslie2009). They allow for an understanding of context-dependent cultural constructs of the environment (Campbell et al. Reference Campbell, Gray, Hazen and Shackeroff2009). The explanations for, and exceptions to, general patterns emerge best from carefully constructed local case studies. Case studies can provide detail accounts of complex interactions within the specific policy realm and will not dismiss important outliers based only on statistical trends (Pickett et al. Reference Pickett, Burch and Grove1999; Russ & Alcala Reference Russ and Alcala1999; Christie et al. Reference Christie, White and Deguit2002; Yin Reference Yin2008). Outliers either in favour of or opposed to MPAs can have a marked impact of their progress (Christie et al. Reference Christie, Buhat, Garces, White, Brechin, Wilshusen, Fortwangler and West2003a). The scientific community may be compelled by the general and comparative studies, while the practitioner and policy maker remain interested in the particulars of their site of interest. The diffusion of case-specific information through social and learning networks clearly influences perceptions by resource users and other MPA constituencies (Rogers Reference Rogers2003; Pietri et al. Reference Pietri, Christie, Pollnac, Diaz and Sabonsolin2009).

Reassessing the starting point and disciplinary balance of knowledge and policy creation

Providing a more pluralistic form of research to guide coastal and marine policy will require reconceptualization of environmental problems and solutions. A fully social ecological conceptualization of problem and solution implies equal attention is paid to both social and ecological aspects. Environmental frameworks and policies, which are social constructs of which ecological conditions are only one of many considerations, are most effective when grounded in reliable and detailed understandings. For ethical, theoretical and practical reasons, the human dimension should not be reduced to mainly economic calculations of, albeit important, ecosystem services or quantified general principles (Campbell et al. Reference Campbell, Gray, Hazen and Shackeroff2009). Just as robust ecological research must span the breadth of natural history, population dynamics and genetics, social research should include, at a minimum, attempts to understand the social context over time, the environmental management process, institutional design principles, human adaptation and social impacts of policy with a consideration for justice (Campbell et al. Reference Campbell, Gray, Hazen and Shackeroff2009; Jones Reference Jones2009).

Recasting the position of society within social ecological research will create opportunities for balanced IR. In the predominant narrative of ocean decline and global policy response, society is generally reduced to the role of perpetrator of environmental degradation, with humans located outside of nature (Campbell et al. Reference Campbell, Gray, Hazen and Shackeroff2009). The conclusion that ocean resources are in a state of decline in many places is important and accurate, and has generated considerable impetus to alter ocean policy. Relatively little is known about the conditions and mechanisms through which society either prevents environmental degradation or actively restores environment. For example, until the seminal work of political scientist Ostrom (Reference Ostrom1990) and others, the seemingly inevitable ‘tragedy of the commons’ was a predominant explanation for why much of the non-private environment was in decline (Hardin Reference Hardin1968). Casting society as both perpetrator of environmental degradation and advocate of environmental sustainability will allow for more meaningful research, theory and policy.

Understanding the historic and contemporary influences on worldviews of social groups and how these shape responses to particular ocean policy is a practical and largely unexploited opportunity. Making known the worldviews of ocean natural and social scientists and how these worldviews influence scientific descriptions of the ocean and ocean policy will improve ability to interpret scientific conclusions. The lack of self-reflective research within the science policy making community is problematic on many levels. First, on a scientific basis, it results in an obvious gap in understanding how science and policy priorities are created. Second, and pragmatically, it reduces the ability to create successful, broadly acceptable and sustainable policies. Policies should not be predicated on constrained information sources and assumptions of what policies are preferable and feasible. The opportunities for IR on such topics are broad and could make an important contribution to the creation of successful ocean and coastal environmental policy.

Another pragmatic area for coastal and ocean IR is the creation and application of multi-method evaluative research. To date, despite mandates to establish global MPA networks and large marine ecosystem management systems (Sherman et al. Reference Sherman, Sissenwine, Christensen, Duda, Hempel, Ibe, Levin, Lluch-Belda, Matishov, McGlade, O'Toole, Seitzinger, Serra, Skjoldal, Tang, Thulin, Vandeweerd and Zwanenburg2005), and rapidly growing interest in EBM (McLeod & Leslie Reference McLeod, Lubchenco, Palumbi and Rosenberg2005; CEQ 2009), the evaluative efforts are commonly ad hoc and inappropriately scaled. Evaluation began with useful case studies of management effectiveness of MPAs (see for example White et al. Reference White, Salamanca and Courtney2002) and LMEs (for example Cochrane et al. Reference Cochrane, Augustyn, Fairweather, Japp, Kilongo, Iitembu, Moroff, Roux, Shannon, Van Zyl and Vaz Velho2009; Fanning et al. Reference Fanning, Mahon and McConney2009). Only few and limited comparative studies of the human and management dimensions of MPAs (Pollnac et al. Reference Pollnac, Crawford and Gorospe2001, Reference Pollnac, Christie, Cinner, Dalton, Daw, Forrester, Graham and McClanahan2010; Cinner et al. Reference Cinner2007; Christie et al. Reference Christie, Pollnac, Fluharty, Hixon, Lowry, Mahon, Pietri, Tissot, White, Armada and Eisma-Osorio2009a) and marine EBM (Christie et al. Reference Christie, Pollnac, Fluharty, Hixon, Lowry, Mahon, Pietri, Tissot, White, Armada and Eisma-Osorio2009b) exist. Global assessments attempting social ecological metrics are challenged by limited data which, in turn, limits metrics and ultimately the usefulness of such assessments. Mora et al. (Reference Mora, Andréfouët, Costello, Kranenburg, Rollo, Veron, Gaston and Myers2006) reduced the human dimensions of coral reef MPA success to an overly constrained suite of variables. Pollnac et al. (Reference Pollnac, Christie, Cinner, Dalton, Daw, Forrester, Graham and McClanahan2010) reduced ecological effectiveness to a ratio of fish density inside and outside MPAs. The reliance on primarily social or ecological, superficially global or context-bound case studies is problematic and should limit confidence that global MPA schemes are feasible (Christie et al. Reference Christie, Pollnac, Fluharty, Hixon, Lowry, Mahon, Pietri, Tissot, White, Armada and Eisma-Osorio2009b). In short, the evaluative research completed or underway is not scaled properly to the problems or the management interventions underway to help ensure feasibility or capture lessons.

Applied marine IR would need to consider methodological issues of specificity and generalization. Consistent methods will need to be used for monitoring particular outcomes of management interventions (for example MPA impact on biodiversity and biomass inside and outside MPAs or socioeconomic impacts of MPAs). The bias toward quantitative or modelled results, based on a tendency to view quantitative results as rigorous while qualitative results are dismissed as anecdotal and subjective, should be resisted on methodological, practical and educational grounds (Pickett et al. Reference Pickett, Burch and Grove1999; Campbell et al. Reference Campbell, Gray, Hazen and Shackeroff2009). Qualitative social data which maintain the narrative of informants and complement general patterns with specific and, perhaps non-statistically significant, results are essential to a complete understand of complex phenomena. Such site-specific and detailed research will improve ability to explain, at a scale meaningful to policy makers and impacted communities, why patterns emerge from comparative research. Creation of publication review standards for such IR is underway, but requires editorial guidance to reduce the tendency of reviewers to retreat to their disciplinary and methodological biases (Pickett et al. Reference Pickett, Burch and Grove1999; Campbell Reference Campbell2005; Campbell et al. Reference Campbell, Gray, Hazen and Shackeroff2009).

Social ecological research will need to be conducted over time to identify patterns of change. Social ecological outcomes are related to the duration of implementation (see Russ et al. Reference Russ, Alcala, Maypa, Calumpong and White2004) and management concerns change from attainment of primary MPA implement (for example enforcement) to distribution of benefits (Pollnac & Pomeroy Reference Pollnac and Pomeroy2005; Pajaro et al. Reference Pajaro, Mulrennan, Alder and Vincent2010). The costs and benefits of MPAs change over time.

IR will need to bridge the scientific/non-scientific gulf. Environmental policy, perceived as legitimate by diverse social groups, can be generated by participatory IR (Fals-Borda Reference Fals-Borda1987). Interdisciplinary participatory research offers an opportunity to create partnerships with communities impacted by environmental policies, many of which have important non-scientific knowledge and may, through the research process, become aware of social ecological dimensions of the ocean and policy processes in a manner that fosters their commitment to sustainable practices (Pomeroy et al. Reference Pomeroy, Parks and Watson2004). Their knowledge and perspectives may complement and challenge the perspectives of scientific researchers. Tropical coastal inhabitants are commonly generalist farmer-fishers, and therefore are aware of critical social ecological linkages. The poor and marginalized are obvious ‘experts’ on the impacts of poverty, a key social variable defining policy options (Fals-Borda Reference Fals-Borda1987; Freire Reference Freire1993). These social groups are most interested in research which directly addresses a tangible problem which they face.

The current separation of research and action is grounded in a commitment to a particular form of scientific inquiry. Praxis, the melding of research with action in an iterative process, provides a potentially complementary research frame to applied research which typically ends with policy recommendations (Lather Reference Lather1986). Pragmatically speaking, self-compliance with ocean policy is essential given limitations in enforcement capacity. Rules which are derived of participatory processes are frequently perceived of as more legitimate, which in turn increases compliance (Kuperan & Sutinen Reference Kuperan and Sutinen1998; Pollnac et al. Reference Pollnac, Crawford and Gorospe2001; Christie et al. Reference Christie, Pollnac, Fluharty, Hixon, Lowry, Mahon, Pietri, Tissot, White, Armada and Eisma-Osorio2009a). Therefore, fostering policy processes based on inclusive research programmes is practical and increases the likelihood of success.

Setting high standards for all relevant research will foster IR. Meaningful and useful IR will not occur if policy makers mainly rely on rigorous ecological impact assessments and anecdotal socioeconomic information or policy evaluations. Marine policy practitioners and independent researchers should collaborate in developing supportive applied research programmes to identify theory and practical lessons in a systematic manner. Researchers will need to retain some independence to ensure objectivity and freedom of expression. Their efforts, however, can be collaboratively designed with practitioners and programme constituents to help ensure that questions are practical and broadly interesting.

While it is not commonly discussed, the conduct of successful IR research is, as with environmental management, dependent on the quality of interactions between individuals involved in the research endeavour. The balance of disciplinary and multidisciplinary perspectives working in concert is complex and requires mutual respect, trust and transparency (not unlike any complex institution; Pickett et al. Reference Pickett, Burch and Grove1999; Tress et al. Reference Tress, Tress and Fry2005). Such teams should be guided by individuals with interdisciplinary training and leadership qualities (Tress et al. Reference Tress, Tress and Fry2005). The research team may wish to integrate policy makers and practitioners as researchers, research design consultants and reviewers. A degree of theoretical, methodological and epistemological ‘tension’ is appropriate to ensure that researchers are challenged to respond to constructive criticism while conducting research. The design of multidisciplinary research that uses complementary methods (for example quantitative and qualitative, natural and social science) will increase the likelihood of robust conclusions based on triangulation of methods and completeness of analysis.

One such programme, the ICM Sustainability Research Project (ICMSRP), relied on social and natural science, and both quantitative and qualitative research to identify why ICM programmes tended to collapse after donor and technical assistance withdrawal in the Philippines and Indonesia (Christie Reference Christie2005; Christie et al. Reference Christie, Lowry, White, Oracion, Sievanen, Pomeroy, Pollnac, Patlis and Eisma2005; Lowry et al. Reference Lowry, White and Courtney2005; Pollnac & Pomeroy Reference Pollnac and Pomeroy2005). This three-year effort involving an international team of 15 researchers from natural and social sciences, research and practice began with a commonly-defined question, planned for integration from the onset, emphasized basic and applied research, and resulted in a sense of collegiality; some of the hallmarks of successful IR (Pickett et al. Reference Pickett, Burch and Grove1999; Campbell Reference Campbell2005; Tress et al. Reference Tress, Tress and Fry2005). Within this realm of inquiry, the following constitutes an example of how complementary methods function. Quantitative survey research methods determined, amongst other variables, that participatory processes were strong predictors of MPA success (as measured in multivariate manner) in nine sites (Pollnac & Pomeroy Reference Pollnac and Pomeroy2005) and that MPA success was reduced when tourism business owners were directly involved in community-based MPA implementation processes (Thiele et al. Reference Thiele, Pollnac and Christie2005) or captured the majority of economic benefits (Oracion et al. Reference Oracion, Miller and Christie2005). Qualitative research provided detailed narratives of why conflict emerged between tourism and fishing social groups based on distinct definitions of MPA goals and distribution of benefits (Christie et al. Reference Christie, Buhat, Garces, White, Brechin, Wilshusen, Fortwangler and West2003b; Christie Reference Christie and Shipley2004). These findings were related directly to long-term biological conditions in MPA study sites (Christie Reference Christie2005). Findings were translated into empirically-grounded education materials used in training programmes (Milne et al. Reference Milne, Christie, Oram, Eisma and White2003) and served to validate findings and encourage realistic policy recommendations. The creation of coherent, multidisciplinary and multi-method explanations for complex phenomena is attainable and productive, but takes time and commitment to generating a common language and common meaning (Tress et al. Reference Tress, Tress and Fry2005; Pickett et al. 2009).

The reformation of educational programmes to train interdisciplinary and integrative researchers and policy makers requires ongoing support if environmental IR is to flourish (Zarin et al. Reference Zarin, Kainer, Putz, Schmink and Jacobson2003). Environmental studies programmes at North American and European universities are in a process of rapid evolution. Leading universities have created environment programmes drawing together diverse disciplines and interdisciplinary faculty members and students. Education programmes should include coursework and support for student research in epistemology, science studies, interdisciplinary theory and methods, and scientific and participatory methods. Students will need to be educated in the theories and methods across disciplinary boundaries and exposed through practical experiences to the perspectives and interests of diverse social communities. There is room for growth; educational materials to train the next generation of interdisciplinary scholars and policy makers are not as advanced as disciplinary programmes.

Finally, an investment in applied IR commensurate with the problems and impacts of global ocean policy trends is needed. As a provocative but reasonable solution to the challenges identified in this review, I suggest that equal amounts of funding be applied to the social and natural science components of marine IR, with a prioritization of research that is comprehensively and simultaneously social and natural science inquiry. Such a reworking of funding priorities based on parity would increase the possibility for balanced and meaningful IR. While some may argue that the current funding imbalance is appropriately due to the elevated expense of natural science versus social research, the paucity of ocean social science highlights the immediate need for rebalancing. High quality, comparative social science requires significant funding commensurate with the scale of the issues identified in order to be successful. While a plea for additional social science research funding may seem self-serving, this suggestion is made from a pragmatic perspective. Investment in appropriately scaled and complete IR research could improve success for policies which are expensive and necessary, given the decline of ocean resources and biodiversity. Marine and coastal policies are in a state of evolution, and such IR research could support a constructive recalibration of concepts at a critical juncture.

Having identified areas for improvement and opportunities for IR within the realm of marine and coastal policy, further attention to balance within and between disciplines and research and policy making will continue to improve understanding of the ocean and associated societies. The need and impetus for such research is greater now than ever.