Toward Operationalizing Resilience Concepts in Australian Marine Sectors Coping with Climate Change

We seek to contribute to the scholarship on operationalizing resilience concepts via a working resilience indicator framework. Although it requires further refinement, this practical framework provides a useful baseline for generating awareness and understanding of the complexity and diversity of variables that impinge on resilience. It has potential value for the evaluation, benchmarking, monitoring, and reporting of marine system resilience. The necessity for such a framework is a consequence of the levels of complexity and uncertainty associated with climate change and other global change stressors in marine socialecological systems, and the problems involved in assessing their resilience. There is a need for: (1) methodologies that bring together knowledge from diverse sources and disciplines to investigate the complexity and uncertainty of interactions between climate, ocean, and human systems and (2) frameworks to facilitate the evaluation and monitoring of the social-ecological resilience of marine-dependent sectors. Accordingly, our main objective is to demonstrate the virtues of combining a case study methodology with complex adaptive systems approaches as a means to improve understanding of the multifaceted dynamics of marine sectors experiencing climate change. The resilience indicator framework, the main product of the methodology, is developed using four case studies across key Australian marine biodiversity and resource sectors already experiencing impacts from climate and other global changes. It comprises a set of resilience dimensions with a candidate set of abstract and concrete resilience indicators. Its design ensures an integrated approach to resilience evaluation.


INTRODUCTION
Marine systems are recognized as complex adaptive systems that are under considerable stress from a range of anthropogenic impacts (Levin and Lubchenco 2008). In many locations, nonclimate anthropogenic impacts on marine ecosystems from overfishing, bycatch, habitat destruction including from coastal development, and chemical and nutrient pollution are being further exacerbated by climate change (Crowder et al. 2008). Although we consider the consequences of wider global change, our particular focus is on the effects of climate change on marine-dependent industries and associated human populations, because these are expected to be especially affected by climate change (Badjeck et al. 2010, Brander 2010. Indeed, the future of services supplied by marine ecosystems is becoming increasingly uncertain (Gunderson 2003).
Australia's marine systems and biota are exposed to a range of likely impacts from human-induced climate change, including warming ocean temperatures, ocean acidification, sea level rise, changes in nutrient availability, and changes in variability and extremes such as storminess, rainfall intensity and runoff, and associated variation in salinity levels (Poloczanska et al. 2007(Poloczanska et al. , 2012. Diverse marine environments are already exhibiting climate change impacts, including extensive coral bleaching along the Great Barrier Reef (GBR; , Hughes et al. 2010, poleward range shifting of species (Last et al. 2010), increasing frequency of harmful algal blooms (Hallegraeff 2010), habitat damage from changes in storm frequency and distribution, and ocean acidification (Howard et al. 2009, Poloczanska et al. 2012. All of these have knock-on effects for marine biodiversity and the resilience of marine social-ecological systems (SES; Poloczanska et al. 2012). Ocean warming of Australia's highly productive southeast and southwest marine waters (Holbrook and Bindoff 1997, Pearce and Feng 2007, Ridgway 2007, which are warming faster than 90% of oceans elsewhere, so-called "hotspots" (Tittensor et al. 2010), has serious implications for dependent marine sectors. As well as these significant climate stressors, marine sectors are subject to a range of nonclimate drivers that often interact with the former and can have a compounding or dampening effect.
As SES, the complex interactions between the social and ecological dimensions of marine sectors are influenced by nonlinearity of feedback effects between the two systems, by associated thresholds, surprises and perverse effects, legacy effects, resilience status, and by spatial, temporal, and organizational variation (Liu et al. 2007). To understand and temper the resulting levels of complexity and uncertainty, a Additionally, operationalizing resilience is associated with some more general conceptual and practical challenges. The first set of challenges includes developing the commonplace acceptance of SES as complex adaptive systems (Walker and Salt 2006), definitional problems resulting from the ambiguity of the 'resilience' term (Brand and Jax 2007), its dynamic context (Nelson et al. 2007, Bohensky 2008, and constraints on further conceptual development by high levels of system complexity and dynamism . The second set of challenges involves the practical difficulties of resilience measurement (Carpenter et al. 2005), such as determining which variables to measure , developing standard metrics (Cutter et al. 2008), making resilience observable (Nyström et al. 2008, Robinson andBerkes 2010), locating and finding measures for thresholds Meyers 2004, Eakin andLuers 2006), acquiring sufficient data (Malone and Brenkert 2008), and measuring resilience in a context of multiple fast-and slowmoving drivers of change (Nelson et al. 2007).
For our purposes, resilience operationalization is the practical application of resilience concepts in decision making and planning. Operationalization entails making resilience concepts useful and useable beyond their theoretical context to policy makers and managers in marine SES and using the lessons from such application to further inform resilience's conceptual and practical development.
As a contribution to the literature on operationalizing resilience, we develop and present a resilience indicator framework, based on investigations of the system dynamics of four Australian marine sectors experiencing impacts from climate and other sources of global change. We propose an approach to resilience diagnosis that reflects the multidimensionality and complexity of marine SES. We use four case studies to develop a set of critical resilience dimensions to underpin our framework.

Resilience and complex adaptive systems
The theoretical grounding for this study is in the approach to SES proposed by the Resilience Alliance (2007) and its associates. This work owes much to ecologists such as Gunderson and Holling (2002; see also Holling 2001) and the concepts they propose for understanding complex SES, such as the adaptive cycle and panarchy. A key tenet of this theory is that change rather than equilibrium is the normal state of complex adaptive systems. As a result of enhanced interconnectedness between social and ecological systems, it is becoming clearer that change is increasingly predictable, whether it be gradual or unexpected change (Nelson et al. 2007). In marine environments subject to climate change and variability, such changes are a function of the complex nonlinear feedbacks among human production, ocean, and climate systems. The operation of feedbacks can generate unexpected disturbances and outcomes, which, in turn, create an environment of uncertainty for marine managers. One of the virtues of a resilience approach is that it opens up the possibility of operating in this "zone of uncertainty" (Bourdieu 1999).
If we follow Bourdieu's line of reasoning, we understand this zone to be one that allows for transformative spaces to be created in which the ways of behaving and acting that are taken for granted can be unsettled and interrogated so that novel responses to complex problems can emerge and be tested. Although predictions cannot be made with confidence, causes may be unclear, and contradictory conditions are evident, operating in the zone of uncertainty can enable the sort of reflexivity and adaptive practice that support rapid reevaluation of dominant conceptualizations of conditions. The zone of uncertainty is paralleled in complex adaptive systems by the 'back loop' of the adaptive cycle in which levels of system resilience are low, and the system is open to external influences, novelty, innovation, experimentation, learning (Holling 2001), and 'windows of opportunity' (Olsson et al. 2004).
The potential of the back loop is complemented by the understanding that complex adaptive systems are capable of operating in multiple states, effectively allowing for the Ecology and Society 18(3): 4 http://www.ecologyandsociety.org/vol18/iss3/art4/ possibility of changing the system state. An associated concern for complex adaptive systems is avoiding transformation into a qualitatively different and undesirable state as a result of disturbance. This is a function of their resilience, which describes the amount of change they can undergo and retain the same controls on structure and function, their capacity for self-organization, and their ability to build capacity to learn and adapt (Walker et al. 2002, Folke 2006; see also http://www.resalliance.org/index. php/resilience). Resilience analysis therefore needs to account for those slowly changing variables on which resilience depends, the key feedbacks operating among the different systems, nearness to thresholds that might carry the system into an undesirable state, and the capacity for reorganization in the face of both gradual and transformative change.
A case study approach to operationalizing resilience Empirical robustness is achieved by using case examples of sectors currently dealing with climate change impacts. The particular case studies discussed here are instrumental case studies (Stake 2000), selected to provide insight into the interactions and interdependencies between linked social and ecological systems threatened by novel climate change impacts. The cases are also of the extreme and critical kind (Flyvberg 2006) because they comprise different types of marine-dependent sectors already dealing with climate change effects, although they utilize marine resources in different ways. As extreme cases, they exemplify instances in which climate change impacts are particularly problematic, e.g., the effects of species shifting their ranges in response to climateinduced changes in ocean circulation, impacts of ocean warming on fisheries and aquaculture in a global warming hotspot, and the impacts of extensive coral bleaching on marine tourism in the GBR. As critical and extreme cases, it was anticipated they would provide both the maximum amount of information and understanding about the dynamics of four different SES and therefore generate a valid set of data as the basis for an indicator framework.
The main purpose in combining case study and complex adaptive systems approaches to appraising marine sector resilience was to ensure that due consideration was given to the complexity, uncertainty, and multidimensionality that is inherent in such an enterprise in the context of global climate change, especially in marine-dependent sectors.
The four studies of marine sector resilience were undertaken in 2009/10 under the auspices of Australia's Climate Change Adaptation Research Network for Marine Biodiversity and Resources, one of the National Climate Change Adaptation Research Facility's eight national adaptation networks. The case studies, i.e., range shifting of marine species in response to ocean warming, the Tasmanian commercial rock lobster (Jasus edwardsii) fishery, oyster aquaculture in southeastern Australia, and tourism in the GBR were selected on the basis of climate change impacts of current concern, researcher expertise, and data availability. The sectoral cases were approached from a disciplinary perspective first, focusing initially on ecological (species range shifting), economic (rock lobster), institutional (oyster aquaculture), social (GBR tourism) resilience perspectives. Key aspects of the case sectors are outlined in Table 1.
Conclusions about the resilience of each case study system cannot easily be drawn at this stage. The evidence for rangeshifting species is still emerging; however, some shifts will have dramatic repercussions for receiving ecosystems whereas others will be viewed as benign or even beneficial from a human perspective (Madin et al. 2012). Those in the former category may possibly cause regime shifts as is happening with the invasion of the urchin species, Centrostephanus rogersii, which is damaging the resilience of rocky reef ecosystems and dependent fisheries along Tasmania's east coast (Johnson et al. 2011). The Tasmanian rock lobster industry is rated as having high economic resilience in relation to its governance and management institutions and for its fleet capacity; however, the sector is vulnerable in terms of fuel costs, supply chain components such as information flow and innovation, and financial security (Pecl et al. 2009, van Putten and. Oyster aquaculture's vulnerabilities include water quality impacts resulting from catchment activities, absence of integrated terrestrial-marine governance, and lack of understanding of the biophysical basis of the industry (Leith and Haward 2010). However, emergence of collaborative management approaches between government and growers and improvements in oyster species' resistance to disease, maintenance of productive environmental conditions, and management improvements will contribute to sector resilience. The New South Wales industry is more vulnerable to climate change impacts, through outbreaks of disease and flooding, than the South Australian and Tasmanian segments. Great Barrier Reef tourism is especially vulnerable to climate change impacts through increased risk of vector-borne disease, increased intensity of natural hazards, e.g., cyclones, and reduced biodiversity (Marshall et al. 2009). However, the existence of multilevel, collaborative governance arrangements and interactions among scientists, environmental managers, tourism operators, fishing industries, and the broader community provides a high level of institutional and social resilience.
As a prelude to the presentation of the resilience indicator framework, we discuss the challenges for resilience frameworks, establish a rationale for the use of frameworks in SES studies, discuss the role that indicators can play in operationalizing resilience concepts, and consider precursors for resilience indicators. Reef based tourism is vulnerable to climate change impacts: sea level rise, increased water and air temperatures, increased storm/ cyclone frequency and severity, ocean acidification, increased windspeed, changed rainfall and runoff, cloud cover affecting visibility, and changes in El Niño Southern Oscillation (Coghlan andPrideaux 2009, Wilson andTurton 2010).
Substantial falls in visitor numbers are expected if environmental conditions, e.g., water quality, are degraded (De'ath and Fabricius 2008) with significant implications for regional annual income (Huybers and Bennett 2000).
Sea and air temperatures are increasing; there is observed sea level rise, ocean acidification and more intense storms and more frequent rainfall (Poloczanska et al. 2012) Diversity of small-medium businesses (retail, accommodation, and tour based) Resilient to change  Spatially differentiated operations across inner and outer coral reefs and islands Global economic activity and other factors influencing businesses' profitability Coral bleaching events on nearshore reefs have increased in frequency and severity since 1990 (Thompson and Dolman 2010) (con'd) http://www.ecologyandsociety.org/vol18/iss3/art4/ Marine tourism activities and operations are diverse, encompassing live-aboard vessels, cruise ship operators, catamaran and kayaking tours, fishing, and diving.

Existence of alternative reef destinations and competition with and economic viability of other iconic tourist attractions, such as North Queensland's Wet Tropics
Reef bleaching influences the number of visitors (Oxford Economics 2009).
Long lived corals are calcifying 15% less than prior to 1990 , De'ath et al. 2009). Contributes to a growing hospitality industry of resort-style accommodations and restaurant services.
Recreational and tourism services are strongly related to coral reef biodiversity, coral cover, and water clarity (Wielgus et al. 2004).
Rapid coastal population growth (Great Barrier Reef Marine Park Authority 2009).

Challenges for resilience frameworks
One of the challenges of resilience approaches to SES is to integrate understanding from multiple disciplines, methods, and perspectives (Berkes 2007). Indeed, the study of complex systems necessitates interdisciplinarity because of their multifaceted dimensions, limited predictability, and dynamism (Newell 2001). A further challenge is to enable improved management of uncertainty and surprise. Currently, uncertainty is most often handled within a risk management framework. In highly dynamic contexts, uncertainty can be managed more effectively using an emergent/contingent framework that explicitly deals with surprise. This is necessary because the kinds of events or outcomes associated with the nonlinear feedbacks that characterize complex interactions between the climate, ocean, and human production systems are concerned with what van der Heijden (1996) refers to as "structural uncertainties," i.e., possible events for which there is little or no evidence to judge the likelihood of an outcome, and "unknowables" or unimaginable events.
Consequently, some of the key functions of a resilience indicator framework should be to support ways of operating in the zone of uncertainty, facilitate identification of windows of opportunity and potential transformative spaces, and inform capacity building to better prepare for and respond to surprise.

Operationalizing resilience through indicator frameworks
Although resilience indicator frameworks are in their infancy, scholarship around the use of sustainability and environmental indicator frameworks hints at how indicators could help in operationalizing resilience. From a policy perspective, indicators can enhance the overall understanding of resilience as a concept because the resulting reflection of ongoing assessments can lead to the gradual incorporation of resilience goals and standards into policies and organizations, the socalled enlightenment effect (Gudmundsson 2003). Indicators are valuable in providing information on complex issues in a way that is accessible to decision makers (Niemeijer and de Groot 2008). They can also stimulate change in stakeholders and systems through ongoing processes of negotiation and learning (Reed et al. 2006), with the eventual and desirable outcome of such processes being legitimization of a resilience orientation (Cabell and Oelofse 2012).
From the perspective of dealing effectively with change, indicators perform several functions. First, they can be used to establish baselines and to determine the direction of change in relation to a particular condition of resilience such as a threshold. In monitoring change, thresholds, targets, or baselines, beyond which problems become critical, serve to trigger remedial action (Reed et al. 2006). Information from monitoring indicators can be used as the basis for adaptive management strategies that help stakeholders adapt to and manage change. Lastly, indicators can enhance processes of social learning through stakeholder participation in indicator development processes (Pretty 1995).
From a resilience perspective, the sustainability indicators literature is deficient in that, with few exceptions (for example, Grosskurth and Rotmans 2007), it has so far not captured the broader dynamics of systems, and these are critical to both sustainability and resilience. Although monitoring indicators can adequately support incremental adaptive change, a different class of indicators is required in uncertain contexts to capture the complexities of system dynamics. These indicators could be developed through futures planning techniques such as scenario analysis (Haward et al. 2013), thus facilitating the consideration of potential windows of opportunity and suggesting areas to build capacity to better respond to surprise. With respect to the latter, resilience indicators should be significant sources of social-ecological learning, identified as a critical element of resilience building and for coping with uncertainty and surprise .
Although aware that the selection of indicators ultimately depends on the research question being asked or the objectives of a particular study, we sought to identify a set of indicators that would be useful in a general sense in the diagnosis and monitoring of marine sector resilience. The decision to http://www.ecologyandsociety.org/vol18/iss3/art4/ develop indicators as measures of resilience rather than resilience metrics per se is supported by the opinions of others that the value of resilience thinking is more likely to be realized in industry sectors and systems by identifying general rules of thumb that can guide sectors toward a resilience orientation (Bennett et al. 2005, Carpenter et al. 2005, Darnhofer et al. 2010, Cabell and Oelofse 2012. Consequently, we were not concerned at this stage to adhere closely to accepted ideals of data availability, measurability, and cost effectiveness in indicator selection. Presentation of indicators in a framework format is supported by Walmsley (2002) who suggests that the use of frameworks for sustainability indicators is crucial to identifying, summarizing, and reporting on key issues because it enables the logical grouping of information and thus the promotion of indicator interpretation and integration. Frameworks also help in the identification of data collection needs and gaps. Similarly, Ostrom (2011:8) advocates the use of frameworks in diagnostic work because frameworks establish the "elements and general relationships among these elements that one needs to consider for ... analysis and they organize diagnostic and prescriptive enquiry. They attempt to identify the universal elements that any theory relevant to the same kind of phenomena needs to include." A key purpose of the framework presented here is to ensure that important aspects of resilience concerns are identified and considered in studying and managing marine sector resilience.

The process of resilience indicator framework development
The framework has three main components: a set of critical resilience dimensions, a capitals or assets framework to organize the indicators, and indicator subsets, both abstract and concrete. Following initial assessment of the respective social, economic, institutional, and ecological resilience foci for each case study, the lead researchers overseeing each case study combined their expertise at a workshop held in January 2011, where they drew on their links with marine researchers, policy makers, and managers to provide a more comprehensive, interdisciplinary assessment of marine system resilience dimensions across the four sectoral case studies. The methodology was guided by the systems dynamics approach developed by the Resilience Alliance (2007) and informed by Bennett et al.'s (2005) work on resilience surrogates. The researchers identified a set of eight resilience dimensions that could be applied to sectors to describe current and potential resilience (Table 2). Thus the resilience dimensions set, grounded in a rigorous interdisciplinary process, provided the ideal basis for a comprehensive indicator framework.
The case information was synthesized into a matrix composed of the eight resilience dimensions and five asset classes. The latter, which encompass a range of livelihood resources, i.e., ecological, social/human, economic/financial, political/ institutional, and infrastructural/technological (physical) assets, adapted from Scoones' (1998) sustainable livelihoods framework, are used to elaborate the resilience dimensions (Table 3). Livelihood perspectives have been shown to be useful in complex, highly dynamic development contexts (Scoones 2009). Applying this framework helped to ensure the balanced treatment of all relevant system components, although the addition of political and institutional assets, in particular, secured the inclusion of governance and power factors.
To formulate the indicator component of the framework, exemplars were extracted for each resilience dimension and for as many asset classes as possible from the case study data (refers to columns 2 and 3 in Table 4). The exemplars represent abstract indicators of sector resilience, which, although useful at a conceptual level, are not usually sufficiently concrete as the basis for data collection (Niemeijer and de Groot 2008). In column 4, we suggest potential concrete indicators for each abstract indicator and provide a rationale for each indicator in column 5.

ANALYSIS AND DISCUSSION
Evaluation of the resilience indicator framework's utility as a first step in system resilience appraisal requires some consideration of its capacities for benchmarking, monitoring, evaluating, and reporting on SES resilience. In part, this capacity is dependent on the comprehensiveness of resiliencerelevant content, and, in part, on the ease with which the framework can be operationalized. Operationalization will be influenced by (1) the development of appropriate resilience metrics, currently limited by data availability, and (2) the framework's further refinement. In our integrated approach to SES resilience, we addressed a range of resilience preconditions, including opportunities for social-ecological learning, preparedness for surprise, ability to cope with uncertainty, ways of dealing with complexity, and the presence of transformative spaces and windows of opportunity.
Social-ecological or resilience learning is catered for by ensuring multiple stakeholder perspectives and knowledge systems are incorporated in problem solving. This capacity for systematic learning through dialogue, deliberation, and meaningful social interaction to enhance long-term sustainability and resilience under uncertain conditions is determined by indicators for trust building and purposeful strategies to enhance social capital (Béné et al. 2011). Social capital is also a crucial ingredient in the necessary collaboration of industry, managers, policy makers, and other stakeholders to effect transformative action.
Preparedness for surprise and capacity to cope with uncertainty needs a substantial monitoring program, although there will always be unforeseeable events. In turn, monitoring aids the social learning needed in responding to change. Other  Awareness of factors contributing to social-ecological system (SES) vulnerability is needed to manage their capacity to produce ecosystem services.
Poor water quality caused by runoff from adjacent catchments is a significant stressor of North Queensland coral reefs (Fabricius 2011). Management authorities established the Reef Rescue program to improve agricultural practices and monitor water quality (Eberhard Consulting 2011). Key slow variables affecting resilience Slow variables are controlling variables that are buffered by stabilizing feedbacks and determine the ability of a system to stay in a particular system state ).
The resilience of coral reefs to cyclones, warming sea surface temperatures, and anthropogenic stressors determines whether or not reefs shift into less productive algal-dominated systems .

Key fast variables affecting resilience
Fast variables are those operating at shorter temporal and smaller spatial scales that can cause changes in slow variables operating at longer time scales.
Fishing effort increases through technological intensification in conjunction with recent climate change-induced oceanographic changes contribute to localized rock lobster (Jasus edwardsii) depletion in southeastern Australia .

Key feedbacks
Feedbacks between biotic and abiotic components of marine systems, and climate and socioeconomic systems can act synergistically to drive SES into less desirable states (Harley et al. 2006).
A synergistic interaction between climate variation (warming waters), fishing pressure (through technological intensification), and long-spined sea urchin (Centrostephanus rodgersii) predation on kelp beds affects abalone and lobster stocks (Last et al. 2010) and can be a first step transformation to a new SES state. The recently arrived urchins can now be harvested, partially replacing declining abalone and rock lobster fisheries and also redirecting diving effort and labor. Coastal communities and their fishing fleets may consequently change in composition and size.

Likelihood of crossing thresholds
Identifying the likelihood of a system crossing a threshold into a less desirable state will indicate its resilience and what should be done to strengthen adaptive capacity and increase sectors' or systems' ability to move toward institutions/practices that allow sectors to learn and innovate (Berkes 2007).
Overgrazing of Tasmania's productive east coast kelp forests by the rangeextending long-spined sea urchin from warm temperate waters is contributing to a catastrophic regime shift (Ling et al. 2009). Restocking of rocky reefs with large lobsters (urchin predators) is intended to counter the effects of their earlier overfishing, which facilitated successful urchin invasion in the first place.

Response to uncertainty and surprise
The ability of a society to live with surprise and uncertainty is a key factor in building resilience (Folke et al. 2003, Berkes 2007).
Successive major cyclones and coral bleaching events along North Queensland coral reefs have put many coastal communities at risk of permanently losing the ability to attract tourists. In recognition of future change, tourism operators have implemented eco-efficiency measures such as risk management, energy reduction, and building climate change into business plans (Zeppel 2012).

Openness to resilience ideas
Openness to resilience ideas acts as a proxy for SES preparedness to adapt to change. This is especially relevant given increasing evidence that future changes may be sudden and disruptive. (2010) is encouraging Australian fisheries to adapt to climate change through providing research support (Hamon et al. 2013, Pecl andHobday 2011) and fostering initiatives, such as conducting vulnerability assessments. Potential to reorganize SES resilience is largely dependent on ability to reorganize in the event of disturbance. Reorganization can be directed to a degree if critical capacities are maintained (Folke et al. 2003).

The Fisheries Research and Development Corporation
'No take' areas on coral reefs help to maintain biodiversity (ecological memory) which is crucial to regeneration after disturbance by tropical cyclones (Mumby et al. 2006). indicators of enhanced response capacity under conditions of surprise and uncertainty comprise perceptions of risk , infrastructure planning and flexibility, availability of a diverse range of responses, ongoing learning, planning for extreme events, and futures planning.
Indicators that account for complexity include planning for extreme events, openness to innovation, response diversity, resilience building, collaborative management and governance that can help to create learning (Booher and Innes 2010), selforganizing processes, diversity of risk responses, trends in economic diversification, adaptive management, and innovative approaches to environmental management. It is expected that guidance from these indicators would help to manage complexity.
Prospects for transformational change are indicated by the presence of transformative spaces in which accepted practices can be unsettled and interrogated. These could be evident where there are processes of critical reflection in place, such as in shadow networks, i.e., networks that operate outside the mainstream testing new or innovative ideas, practices, and approaches (Olsson et al. 2006). However, any threat to the current stability regime is also a potential site for transformation, e.g., natural disasters, declines in keystone species, a drop in fishing effort, catch trophic changes, stock collapses or declines, pest invasions, and economic crises, http://www.ecologyandsociety.org/vol18/iss3/art4/ Although unexpected events may open windows of opportunity on an irregular basis, a more important question is: Can the framework facilitate the purposeful creation of windows of opportunity? Westley (2002) argues that, to allow a policy window to open, all the relevant actors and organizations at all levels have to create the right links, at the right time, and around the right issues. The fora of operators, managers, and scientists conducted within the aquaculture and fishing industries have the potential to create appropriate alignments of actors, organizations, and issues at crucial times but policy entrepreneurs need to provide leadership and generate the political will to push in new directions (Olsson et al. 2004). This reliance on the serendipitous alignment of the appropriate factors points to the importance of systematically prefiguring institutional reform to be prepared for brief windows of opportunity (Young 2010). Such fora must purposively incorporate processes of critical reflection on current practice and industry or sector direction. This aspect is reflected in the indicator requiring the installation of processes for critical reflection to reevaluate norms, values, rules, and practices.
Lastly, for transformation to be initiated, the system must be open to external influences, and have capacities for novelty, innovation, and learning. The relevant indicators that represent these values include maintenance of species diversity from an ecological perspective, flexibility of location or equipment from a marine sector perspective, innovative approaches to environmental management, potential for consumer preference revision, individuals' preparedness for change , Marshall 2010, educational attainment, existence of multilevel networks, and stakeholder inclusiveness.

CONCLUSIONS AND FUTURE POTENTIAL
We conceptualized resilience as a complex and dynamic multidimensional model of change in SES. We anchored the resilience indicators framework in a case study methodology and a systems dynamics approach for the purpose of capturing this complexity. Key resilience dimensions and a related set of resilience indicators applicable to marine sectors experiencing climate change were identified. The comprehensiveness of the indicator framework relied on taking an interdisciplinary approach to data collection and a suitable framework to ensure that all relevant elements were considered. Although the method of indicator development is readily replicable, it should be understood that a different group of research participants may identify a different indicator set. Although the framework requires further refinement, we have been able to demonstrate through the methodology that it is possible to capture the complexity and variety of variables impinging on marine sector resilience.
Although the framework is not yet ready for immediate implementation, it provides a baseline that can be used for discussion and to focus attention on the diversity and complexity of factors influencing resilience. The indicators act as prompts for the kinds of variables that should be considered. They suggest negative factors or constraints on resilience that should be taken into account and positive or negative trends that may influence resilience-building efforts. They include reminders to allow for the unexpected and for the importance of having adequate responses to deal with uncertainty. They provide examples of critical feedback signals and possible signs of impending thresholds. Lastly, they suggest potential demonstrations of resilience action and the kinds of capacities and precursors required to respond to system change.
The immediate practical value of the indicator framework resides in its potential use in: Ecology and Society 18(3): 4 http://www.ecologyandsociety.org/vol18/iss3/art4/ Warming sea temperatures are having a substantial impact on ecosystems and will likely affect the future shape of marinedependent sectors E Economic pressures Number of people entering or exiting the industry When people are constrained by the costs associated with entry into an industry, replacement of exiting operators is slow leading to industry stagnation Stagnation can also be due to retention of nonresilient individuals who have limited adaptation options I Institutional constraints Integrated governance/ management approaches Governance and management of social-ecological systems is complex and it is essential that governance bodies and instruments are connected and coordinated across multiple levels and that governance is perceived as legitimate Acceptability of rules and management approaches P Longevity of infrastructure Replacement of infrastructure The longevity of some infrastructure and associated sunk (irrecoverable) costs may slow adaptation to change 3. Key fast variables Ec Occurrence and frequency of natural disasters Frequency of cyclones/storms Natural disasters may have unexpected and unpredictable effects on larger cycles; these effects may be catastrophic or open up windows of opportunity for management and/or emergence of novel species Annual catchment runoff Torrential stream outflow events S Changes in consumer preferences

Changes in seafood and recreation preferences
Changes in preferences for particular seafoods or tourism experiences may result in further pressures on overloaded ecosystems thus pushing them toward an irreversible threshold but also causing economic and social instability (con'd) Operators also need to be flexible to take advantage of the flexibility of gear, area, etc., and make rapid changes when the need arises † Ec refers to ecological assets; S includes social and human assets; E refers to economic and financial assets; I includes institutional, policy, and political dimensions; and P refers to physical, including infrastructural, technical, and technological assets. q raising awareness of the breadth of internal and external preconditions for marine sector resilience, i.e., economic, financial, ecological, social, institutional, political, and physical; q raising public awareness of resilience problems and their interconnectedness; q making complex concepts meaningful and comprehensible by helping to develop a common language for discussion; q helping stakeholders to understand resilience and to read resilience trends; q informing decision making so that it is founded on logical, coherent, and transparent information; q setting targets to improve resilience of a sector or sphere of activity that scores low on specific resilience dimensions or variables; and q highlighting trends that can strengthen general and specific resilience of stakeholders and that of their sectors.
With refinement, this framework can ultimately be expected to support the operationalization of resilience concepts by: (1) guiding policy analysis and formulation toward more resilient marine sectors either directly, conceptually, or symbolically (Gudmundsson 2003); (2) developing operational approaches to benchmark, monitor, evaluate, and report on marine sector resilience; and (3) assisting marine sector decision makers and managers to embrace complexity and operate more effectively and easily in a context of uncertainty. The overall purpose of the framework is to guide marine sectors toward a more resilient orientation (Darnhofer et al. 2010, Cabell andOelofse 2012).
To advance the framework, the next steps include its further testing and refinement within applied resilience-based management contexts. This could be achieved through participatory action research on resilience metrics and techniques to facilitate selection of key indicators, i.e., those that relate to multiple resilience dimensions and are representative of overall resilience performance, and so reduce the number of indicators. As a starting point, in Table 5, we offer a list of candidate variables, those found to recur in the framework. These are categorized in terms of generic resilience perspectives that could help facilitate dialogue to identify a subset of predictive or leading indicators, which are used to signal potentially significant change toward or away from desirable resilience states. For example, it may be important for resilience planning to identify which social or ecological components are more or less vulnerable, resistant, or resilient to change. Leading indicators are therefore essential to long-range or strategic planning, monitoring progress on resilience, and anticipatory adaptation.
Further indicator development would ideally involve collaboration with stakeholders in the diagnosis process to ensure relevance and social-ecological learning. Although much of the data for the framework originated in work undertaken by researchers with stakeholder groups, the data were inevitably filtered through expert perspectives.
Ultimately, the expectation for a more mature framework would be one that is able to map resilience, measure progress, and assist in setting priorities, while lessons from its application would further inform the conceptual and especially practical development and implementation of resilience.