King’s Research Portal DOI: 10.1002/fee.1283 Document Version Publisher's PDF, also known as Version of record Link to publication record in King's Research Portal Citation for published version (APA): Fischer, A. P., Spies, T. A., Steelman, T. A., Moseley, C., Johnson, B. R., Bailey, J. D., ... Bowman, D. MJS. (2016). Wildfire risk as a socioecological pathology. Frontiers in Ecology and the Environment, 14(5), 276-284. https://doi.org/10.1002/fee.1283 Citing this paper Please note that where the full-text provided on King's Research Portal is the Author Accepted Manuscript or Post-Print version this may differ from the final Published version. If citing, it is advised that you check and use the publisher's definitive version for pagination, volume/issue, and date of publication details. And where the final published version is provided on the Research Portal, if citing you are again advised to check the publisher's website for any subsequent corrections. General rights Copyright and moral rights for the publications made accessible in the Research Portal are retained by the authors and/or other copyright owners and it is a condition of accessing publications that users recognize and abide by the legal requirements associated with these rights. •Users may download and print one copy of any publication from the Research Portal for the purpose of private study or research. •You may not further distribute the material or use it for any profit-making activity or commercial gain •You may freely distribute the URL identifying the publication in the Research Portal Take down policy If you believe that this document breaches copyright please contact librarypure@kcl.ac.uk providing details, and we will remove access to the work immediately and investigate your claim. Download date: 18. Jun. 2020 CONCEPTS AND QUESTIONS 276 Wildfire risk as a socioecological pathology A Paige Fischer1*, Thomas A Spies2, Toddi A Steelman3, Cassandra Moseley4, Bart R Johnson4, John D Bailey5, Alan A Ager6, Patrick Bourgeron7, Susan Charnley8, Brandon M Collins9, Jeffrey D Kline2, Jessica E Leahy10, Jeremy S Littell11, James DA Millington12, Max Nielsen-Pincus13, Christine S Olsen5, Travis B Paveglio14, Christopher I Roos15, Michelle M Steen-Adams16, Forrest R Stevens17, Jelena Vukomanovic7, Eric M White18, and David MJS Bowman19 Wildfire risk in temperate forests has become a nearly intractable problem that can be characterized as a socioecological “pathology”: that is, a set of complex and problematic interactions among social and ecological systems across multiple spatial and temporal scales. Assessments of wildfire risk could benefit from recognizing and accounting for these interactions in terms of socioecological systems, also known as coupled natural and human systems (CNHS). We characterize the primary social and ecological dimensions of the wildfire risk pathology, paying particular attention to the governance system around wildfire risk, and suggest strategies to mitigate the pathology through innovative planning approaches, analytical tools, and policies. We caution that even with a clear understanding of the problem and possible solutions, the system by which human actors govern fire-prone forests may evolve incrementally in imperfect ways and can be expected to resist change even as we learn better ways to manage CNHS. Front Ecol Environ 2016; 14(5): 276–284, doi:10.1002/fee.1283 F ire-prone temperate forests are becoming increasingly risky places for humans. Despite massive and increasing investments in firefighting, wildfire risk – the probability and potential losses associated with fire – is increasing. The problem is global in scale: Australia and countries in North America and the Mediterranean Basin have experienced substantial losses in life and property to wildfires in temperate forests in recent years (Chapin et al. 2008; Bowman et al. 2011; Dennison et al. 2014; Moritz et al. 2014; Stephens et al. 2014). Length of fire seasons and extent of land area burned have increased in these regions, as have economic losses from wildfire and expenditures on fire suppression (Jolly et al. 2015). In In a nutshell: • Wildfire risk in temperate forests can be considered a socioecological pathology: a set of interrelated social and ecological conditions and processes that deviate from what is considered healthy or desirable • Finding solutions to the problem of wildfire risk requires a more complete specification of fire-prone temperate forests as coupled natural–human systems, and more attention to the complex interplay between the social and ecological conditions and processes that influence human decision making (ie the wildfire governance system) • Building social networks of stakeholders and engaging stakeholders in scenario planning exercises can foster creative problem solving to reduce wildfire risk and restore fire to fire-prone temperate forests 1 University of Michigan, Ann Arbor, MI *(apfisch@umich.edu); US Department of Agriculture (USDA) Forest Service, Corvallis, OR; 3University of Saskatchewan, Saskatoon, Canada; 4University of Oregon, Eugene, OR; 5Oregon State University, Corvallis, OR; continued on last page 2 www.frontiersinecology.org the US, economic losses from wildfires doubled and suppression costs tripled in the decade after 2002 as compared with the previous decade (Headwaters Economics 2013; Reuters 2013). Nevertheless, fire is an essential ecological process in many temperate forest ecosystems, playing a critical role in maintaining native plant and wildlife diversity. The nearly intractable problem of wildfire risk in temperate forests can be characterized as a socioecological pathology: a set of interrelated social and ecological conditions and processes that deviate from what is considered healthy or desirable. Another example of a socioecological pathology is the desiccation of the Aral Sea in central Asia and the subsequent decimation of its fishing industry and coastal human communities, which resulted from a narrow societal focus on the rapid spread of irrigated agriculture for cotton monoculture that led to the overuse of water resources (Gunderson and Pritchard 2002). The wildfire risk pathology, which should not imply that all wildfire is undesirable, can be traced to a complex set of interacting factors. Conditions in forests have become more hazardous due to accumulation of abundant flammable vegetation, in many cases a result of disrupted traditions of indigenous fire management, practices of fire exclusion and suppression, establishment of weeds and other flammable plants, and a warming climate (Moreira et al. 2011; Williams 2013). Population change has also affected fire risk. In some regions, such as the western US, expansion of exurban areas has increased the probability of ignitions and placed more assets at risk in forested fire-prone areas. Accompanying demographic shifts have engendered new social values, policies, and decisions that favor reduction of short-term fire risk to homes and other structures at the expense of longterm risk to forest landscapes (Williams 2013). In other © The Ecological Society of America AP Fischer et al. Wildfire risk as a socioecological pathology ble to identify key human components of the system that control attitudes, behaviors, and policies; it is also possible to develop strategies and analytical tools that human actors in the system can leverage to create more adaptive feedback loops in which wildfire risk reduction accompanies reduction in human and ecological vulnerability. J The nature of the pathology Although global in scale, the socioecological pathology of wildfire risk is clearly demonstrated in the western US. During the 20th century, suppression and exclusion of fire (ie fire protection) allowed flammable vegetation to accumulate in this region’s temperate forests, including scenic areas along the wildland–urban interface (WUI) where amenity-seeking Figure 1. Wildfire risk in fire-prone temperate forests is a result of interacting positive migrants (people who relocate to areas feedback loops that link wildfire and human vulnerability through key drivers of land use based on non-consumptive values such as scenery and recreation) settled and natural resource management. beginning in the 1970s, and increasingly in the 1990s (Theobald 2001; areas, such as southern Europe, rural exodus has led to aban- Johnson and Beale 2002). The extent of area burned donment of land management activities and accumulation and the social and ecological impacts of wildfire in the of hazardous vegetation (Moreira et al. 2011). These drivers western US have increased as the climate has warmed have co-evolved over time, creating a maladaptive, positive over the past two decades (Dennison et al. 2014; NIFC feedback loop in which wildfire risk increases despite poli- 2015), although the proportion of high-severity fires cies and practices designed to reduce it. As wildfires become that is increasing is debatable (Baker 2015). The result larger and less controllable and forested areas become more has been a destabilizing feedback loop in which spiraling vulnerable, society demands more fire protection, pushing fire losses are a direct consequence of policies intended agencies toward suppressing rather than using fire as a tool to protect people and resources from wildfire (Figure 1). (North et al. 2015). The challenge of understanding the The wildfire risk pathology can be viewed as the result problem of wildfire risk and developing solutions is com- of a set of social and ecological regime shifts (Figure 2; pounded by variability and complexity in: (1) fire regimes, Folke et al. 2004). Forests that historically experienced not all of which exhibit the same positive feedbacks, (2) frequent, low- and mixed-severity fires have been effectiveness of fuels management strategies, and (3) insti- homogenized by widespread infilling with smallertutions involved in the governance of fire-prone forests diameter, shade-tolerant tree species, and selective logging (Price et al. 2015). of large, fire-resistant tree species. These changes created We use a coupled natural and human systems (CNHS) new successional pathways and primed forests for large, perspective (Liu et al. 2007) to understand the pathology uncontrollable fires under changing climatic conditions of wildfire risk in fire-prone temperate forests and suggest (Stephens et al. 2013; Stavros et al. 2014). New states and strategies to mitigate it. Applying CNHS concepts to wild- dynamics may be emerging in social systems as well. fire risk has been identified as a prerequisite for under- Expanded populations of WUI residents may be less tolerstanding the problem and framing appropriate policies ant of smoke from fire than their early 20th century natu(Chapin et al. 2008; Moritz et al. 2014; Spies et al. 2014). ral resource-dependent counterparts and earlier native Although some researchers have attempted to address ele- peoples, who relied on forests for consumptive and proments of the pathology, we submit that their effectiveness ductive uses and often actively used fire as a management has been limited by incomplete specification of the CNHS, tool. Fires burning in forested areas raise legitimate conespecially the interplay between the social and ecological cerns about effects on scenic beauty and human health. conditions and processes that influence human decision The potential for fires to escape containment, as well as making – what we call the wildfire governance system. By debates about the effectiveness of controlled burning, including governance in the CNHS framework, it is possi- impose particular constraints on the use of prescribed fire © The Ecological Society of America www.frontiersinecology.org 277 Wildfire risk as a socioecological pathology AP Fischer et al. 278 (a) (b) (c) (g) (d) (e) (f) Figure 2. Social and ecological regime shifts: transition of ecological system from fire-dependent ponderosa pine (Pinus ponderosa) woodland to fire-intolerant early-successional mixed-conifer forest (top); transition of social system from fire-dependent hunting culture to fire-intolerant amenity-oriented culture (bottom). Note the last two pictures in the social regime change series are from Mirror Pond, on the Deschutes River, in Bend, OR, where use has gone from wood processing to recreation and shopping. Courtesy of Amon Carter Museum, Fort Worth, Texas, Deschutes County Historical Society, Tumalo Creek Kayak & Canoe, and Elmer Fredrick Fischer/Corbis. to manage forest vegetation, although the public generally supports activities that mitigate forest fire risk (Shindler and Toman 2003; Maguire and Albright 2005; Wilson et al. 2011; McCaffrey and Olsen 2012). Furthermore, while managed wildfire (eg lightning-ignited fire allowed to run its course within well-defined and maintained perimeters) can contribute to reducing the fuels that support high-severity fires, economic and social factors and attitudes severly limit its use, despite policies that allow it (North et al. 2015). The current wildfire governance system in the western US evolved as part of the positive feedback loop and accompanying regime shifts that comprise the wildfire risk pathology (Figure 1). Governance systems are “messy” collections of diverse parties with different levels of authority at different scales, whose aim is to create stable expectations, norms, and institutions to address complex problems (Duit and Galanz 2008). The wildfire governance system in the western US consists of many state and non-state actors with competing goals, policies, and practices. Long-standing www.frontiersinecology.org federal actors such as the US Forest Service (USFS) and the Bureau of Land Management, as well as state-level departments of natural resources, administer divisions that simultaneously hold different and conflicting aims. For instance, one division within a natural resource agency may aim to restore ecological conditions and processes on historically fire-prone forestlands while another division will aim to suppress fire on those same lands. Departments of natural resources at the state level also provide fire protection to private industrial and nonindustrial landowners, and forest management assistance to nonindustrial owners. A variety of nonprofit organizations are also active in the wildfire governance system, advocating for ecological restoration and fire protection, and providing technical assistance to homeowners and nonindustrial private forest landowners. While based on well-intentioned strategies, the current wildfire governance system has made changing the pathology extremely difficult. Despite the recognized importance of restoring ecological conditions and processes on historically fire-prone forestlands, including reintroducing fire, © The Ecological Society of America AP Fischer et al. Wildfire risk as a socioecological pathology current forest management policies, as implemented, continue to prioritize fire protection (Steelman and Burke 2007). State and federal agencies continue to focus on fire suppression (North et al. 2015) and face numerous challenges that make it difficult to encourage use of thinning, prescribed burning, and managed wildfire to restore forests and reduce future fire risk (Maguire and Albright 2005; Wilson et al. 2011). Expanding state and federal fire suppression budgets creates a disincentive for agencies to shift toward thinning and use of fire as a management tool (North et al. 2015). Moreover, land-use policies and property insurance practices can subsidize the risk of settling in hazardous areas (Yoder and Blatner 2004; Donovan and Brown 2007), although there is no empirical evidence for the strength of this feedback. In addition, the combined influences of climate change and land-use change appear to be leading to longer fire seasons and increased wildfire activity in the western US (Westerling et al. 2006), strongly suggesting that ineffective greenhouse-gas emissions policies in tandem with regional land-use policies have amplified the problem. The result has been a set of complex interactions between fire protection behaviors, hazardous fuels, human settlement patterns, wildfire ignitions, and climate change, which have given rise to everincreasing wildfire risk (Figure 1). For better or worse, the wildfire governance system, in turn, reinforces the wildfire risk perceptions and management behaviors of individual property owners. Such owners often do not make short-term investments in reducing flammable vegetation to diminish their long-term exposure (McCaffrey 2004), in part because the probability of a wildfire damaging their property is relatively low in any given year, but also because they can benefit from the risk reduction activities of other landowners nearby (Busby and Albers 2010). Furthermore, the public generally expects government agencies to protect them when wildfires occur (Canton-Thompson et al. 2008). The resultant human decisions to reduce flammable vegetation (or not to do so) can influence risk at large spatial scales. Unlike other natural hazards, a fire can be ignited by a single individual and can cause widespread impact, and owners who fail to reduce hazardous vegetation around structures and along property lines can enable the spread of wildfire to larger areas (Calkin et al. 2014). J Policy innovation in a complex coupled system Ultimately, the remedy to the wildfire risk pathology is a governance system that transforms maladaptive feedbacks into adaptive feedbacks. Creating such a governance system requires policies that influence human– land–forest and fire-management behaviors and that account for socioecological interactions at multiple scales: spatial (ownership, landscape, ecoregion), temporal (short- and long-term), and organizational (individuals, groups, institutions). Recent US federal policy innovations such as Stewardship End Result Contracting and the Collaborative Forest Landscape Restoration Program, both permanently authorized in 2009, have, to some extent, moved toward this ideal. These initiatives encourage local variation in planning and management such that actions can be coordinated and adapted across larger spatial scales and longer time frames than are typically seen in forest management (Table 1). Similarly, Table 1. Examples of US policies that account for socioecological interactions at multiple scales Policy Intent Demonstrated ability to account for key types of cross-scale interactions Spatial Temporal Organizational Collaborative Forest Landscape Restoration Program (CFLRP) of 2009 Promotes landscape-scale restoration on national forests by making long term financial investments where stakeholders are already working together Engages managers and stakeholders in landscape in planning and management Fosters longer planning horizons than typical in forest management Integrates decision making at local, state, and regional scales Stewardship End Result Contracting (first passed in 1999, permanent authority in 2014) Creates mechanisms for forest management that allow for integration of timber removal and restoration activities to benefit local communities Integrates forest management projects across landscapes Fosters longer implementation horizons than typical in forest management Integrates considerations of local economic, social, and ecological benefits with forest management and wildfire protection goals The National Cohesive Wildland Fire Management Strategy (mandated as part of Federal Land Assistance, Management and Enhancement [FLAME] Act of 2009) Promotes fire-resilient landscapes, fire-adapted communities, and effective and efficient wildfire protection through multi-scalar strategy development and implementation Integrates responses by federal and state agencies, state and local government, and tribes across regional, state, and local scales Will be revised at least every five years to consider changes with respect to landscape, vegetation, climate, and weather Engages federal and state land management and fire protection agencies, state and local governments, tribes, and other stakeholders in analyzing alternatives © The Ecological Society of America www.frontiersinecology.org 279 Wildfire risk as a socioecological pathology 280 AP Fischer et al. the intent of the National Cohesive Wildland Fire Management Strategy of 2009 – mandated as part of the Federal Land Assistance, Management, and Enhancement (FLAME) Act – is to balance local, state, and federal fire protection goals with the need to restore fire-adapted landscapes and create human communities that can plan for, respond to, and recover from wildfires. Policy innovation has already occurred on multiple scales of social organization. A growing number of networks of non-state actors have emerged to address wildfire in the western US by supplementing the work of long-standing state and federal actors. Across the wildfire governance system, networks of diverse stakeholders are operating at various spatial and organizational scales. These include collaborative activities at the national level, such as in the area of interagency wildfire Figure 3. Components of a framework for addressing the pathology of wildfire risk in response, and at the local level, as fire-prone temperate forests through broad human engagement in complex thinking about with neighborhood organizations multi-scalar policies and adaptive planning and management. seeking to reduce wildfire risk. Federal agencies are heavily involved with many of these governance system that itself operates at multiple efforts, such as the Fire Learning Network, a USFS- organizational scales. Furthermore, formal policies do funded project of The Nature Conservancy (an environ- not change human behavior in straightforward ways. mental nonprofit organization). Other efforts have been Change is often resisted, as in the case of the Federal initiated with limited government intervention, as with Wildland Fire Management Policy of 1995, which forprescribed fire councils where local landowners, land mally moved federal policy away from absolute fire managers, and other stakeholders are organizing to suppression. In practice, however, suppression remains increase social and political support for using fire as a the default choice of wildfire management, even as fedmanagement tool and building capacity to implement it eral agencies experiment with more complex strategies across jurisdictional lines. (Steelman and Burke 2007). What is needed is a more While these new cross-scalar policy interventions have fire-adapted governance system that leads to reduced created opportunities to weaken maladaptive feedbacks fire risk through better-targeted fuel treatments, coorbetween wildfire and human vulnerability, their effects dinated efforts, and restoration across whole landare not yet visible. Property losses from wildfires continue scapes. to grow and the annual rate of restoration needed to reduce risk remains well beyond current treatment rates J CNHS planning approaches and analytical tools (Stephens et al. 2013). With projected climate change and further development in the WUI, the problem of In a fire-adapted governance system, actors from across wildfire risk is outpacing the human capacity to adapt. spatial, temporal, and organizational scales would be Perverse incentives continue to encourage not only resi- engaged in interactive, collaborative efforts to develop dential development in fire-prone forests in the WUI but solutions to the wildfire risk pathology (Figure 3). also fire suppression instead of management to reduce risk Social network analysis offers an efficient path to in forested areas (North et al. 2015). Moreover, jurisdic- understanding the complex social structure of a govtional heterogeneity has added new layers of complexity ernance system. The patterns of interaction within a to the governance system, making progress uneven. network of actors – how centralized or densely How these recent policy interventions affect human interconnected they are – influence the functioning behavior and landscape fire risk is unpredictable. New of a governance system and the extent to which it policy does not operate in a vacuum; rather, it is inte- may enable or constrain communication, coordination, grated into the complex, path- dependent wildfire and creative problem solving (Bodin and Crona 2009). www.frontiersinecology.org © The Ecological Society of America AP Fischer et al. Wildfire risk as a socioecological pathology As an example, network analysis was used to map and quantify relationships among a set of organizations involved in forest and wildfire management in Oregon. The analysis indicated that network structure was strongly shaped by the tendency of people to associate with those who possess similar management goals, geographic emphases, and attitudes toward wildfire (Figure 4) (Fischer et al. 2016; Fischer and Jasny in review). In particular, organizations with fire protection and forest restoration goals comprised distinct subnetworks despite a shared concern about the issue of increasing wildfire risk. The lack of cohesion in the overall network could potentially constrain interactions among organizations with diverse information and resources, limiting opportunities for learning and complex Figure 4. A map of actors in a wildfire governance network in Oregon, in which groups problem solving regarding the that interact with each other are closer to each other than to groups that do not interact. wildfire risk pathology. Actors that focus on forest restoration are mainly located in the upper hemisphere of the Network analysis can also inform figure, whereas those that focus on fire protection are largely located in the lower interventions to enhance the struc- hemisphere. This pattern suggests that interaction between actors from the two groups tural characteristics of social net- may be constrained. Policy interventions could create new institutions to bring forest works so as to better support critical restoration and fire protection actors into more frequent and sustained interactions. exchanges of information and resources among key actors (Valente 2012). The Fire Learning Network mentioned earlier is an exam- aged to improve communication, coordination, and ple of a network intervention that has built connectivity joint problem solving. among land management organizations to further restoraOnce social networks are identified, scenario plantion of fire-dependent ecosystems through landscape-scale ning (also referred to as alternative futures modeling) collaborative planning (Butler and Goldstein 2010). offers a systematic method for actors to anticipate Network maps and statistics can reveal highly connected or uncertain future social and ecological conditions resultinfluential organizations whose strategic positions could be ing from potential shifts in social and environmental leveraged to improve communication and cooperation, or trends, or new policies and technologies (Peterson et al. to pinpoint sets of organizations that could benefit from 2003). Scenario planning provides a tool for actors to greater communication and cooperation. Network analysis project social and ecological interactions and outcomes may reveal that conservation groups in the western US are under different scenarios (Hulse et al. 2000; Hulse et al. augmenting the limited capacity of land management agen- in press; Spies et al. in review). Although scenario cies to engage in collaborative landscape planning and planning is not new, emerging stakeholder networks social–ecological thinking by contributing additional and state-of-the-art, spatially-explicit, agent-based labor, skills, and, at times, financial resources. Similarly, models (simulation models that describe autonomous network maps may identify scientists as emerging actors individual agents, eg landowners who make decisions in the wildfire governance system because of their that modify vegetation or built structures) create new increasing role in using, and providing interpretations opportunities for actors to explore socioecological feedof, complex models. Indeed, the analysis of organiza- backs and interactions in real landscapes. Such exertions involved in forest and wildfire management in cises can serve as a discussion aid for actors to collecOregon revealed that several conservation groups and tively identify possible pathways for remedying the academic institutions had much more extensive and wildfire risk pathology. For example, scenario planning heterogeneous networks relative to all other organiza- is facilitating development of more effective and ecotions (Fischer and Jasny in review). The large and logically based forest landscape restoration projects by diverse networks of such organizations could be lever- collaboratives in central Oregon (Figure 5) (Spies et al. © The Ecological Society of America www.frontiersinecology.org 281 Wildfire risk as a socioecological pathology 282 AP Fischer et al. in review). As part of these efforts, stakeholder-generated scenarios are being used with an agent-based model to demonstrate how fuel treatment designs might affect the extent of area burned in the future by high- and mixed-severity fire and the trade-offs among managing for wood, fire risk, and biodiversity. Collaborative groups in central Oregon have shown interest in applying the models to specific landscape-scale projects that help them move beyond forest standscale and short-term perspectives, which can inhibit breaking out of the wildfire risk pathology. Land managers, planners, and other actors in the wildfire governance system can model scenarios that test plausible interventions by exploring uncertainties and risks Figure 5. Representatives of organizational actors within a wildfire governance system associated with implementing alter- in Oregon developing a conceptual map of a wildfire risk scenario. native future policies. These could include using fire to a greater degree as a management tool on public and private lands, shift- sometimes contrary to, the expected effects of land ing responsibility for fire protection from agencies to management actions. They then demonstrated how this homeowners, or zoning land use and development based analysis could be used to anticipate when, where, and on fire risk. Scenario planning can be used to explore the how potentially unexpected fires may burn. Further limits of human adaptation – for instance, to investigate advances in such simulation tools may offer increasingly at what point increasing wildfire risk might compel WUI useful insights into managing the complex feedbacks of residents to move to less fire-prone areas or, alterna- the wildfire risk pathology, and serve as important aids tively, take wildfire management into their own hands. in policy development. Such advanced models may not yet exist, but recent innovations in the implementation of complex agents, J Conclusions social networks, and learning mechanisms may soon bring them within reach. As a case in point, the poten- Although temperate forest regions in the US, southern tial to endow agents with increasingly human character- Australia, and the Mediterranean Basin have different istics (Tweedale et al. 2007) now includes algorithms for landscape histories, their political systems and approaches deliberative reasoning to avoid undesirable situations to fire management all exhibit the socioecological pa(Davidsson 2003; Doniec et al. 2008); proactive, forward- thology of wildfire risk. In Greece, for example, the thinking behavior (So and Sonenberg 2004); and con- decision to shift responsibility for wildfire management founding factors such as spread of misinformation from the Forest Service, located in the Ministry of Agriculture, to the Fire Service, located within the (Acemoglu et al. 2010). The capacity to generate hundreds of spatially explicit Ministry of Public Order – combined with new European alternative futures that explore variability and uncer- Union policies intended to reduce wildfire occurrence – tainty within and among scenario sets can be particu- increased focus on the main symptom of the wildfire larly informative when change is likely to occur outside risk pathology (uncontrollable wildfires) rather than the bounds of historical variability (Hulse et al. in the cause (land-use and population change) (Kalabokidis press). In this vein, Hulse et al. characterized eight alter- et al. 2008). In Australia, post fire disaster recovery native futures for a fire-adapted oak–conifer system has typically included rapid rebuilding, making it difcomposed of multiple sets of contrasting climate, devel- ficult to adapt building practices and landscape design opment, and fire hazard management scenarios and to increasingly fire-prone conditions. In each of these generated simulations of each scenario over a 50-year countries the pathology will continue to be exacerbated period. The authors used the results to explore the by climate change (Flannigan et al. 2013). The need mechanisms through which fires of unprecedented size to adapt is driving rapid policy development, with incould spread through the landscape in response to, and creasing recognition of the importance of collaborative www.frontiersinecology.org © The Ecological Society of America AP Fischer et al. partnerships in some regions. The 2015 decision in Victoria, Australia, to use greater community consultation and partnerships to help identify areas for fuel management to reduce risk, instead of relying on mandated annual targets, is an example of such a shift. As we have demonstrated for the western US, overthrowing all current policies may not be required to mitigate the wildfire risk pathology; revising existing policies could be sufficient. While research, evaluation, and monitoring are required to determine whether policy innovations will be effective and enduring, applying a CNHS framework may help to ensure that policies are well-grounded ecologically and socially. We hypothesize that engaging actors in anticipatory thinking can help reveal how the transformation of maladaptive feedbacks into adaptive feedbacks can come from within the network of actors within a CNHS. As policies are implemented, managers, planners, and other actors can use scenarios and modeling not only to identify social and ecological processes that continue to exacerbate wildfire risk but also to test further strategies to reverse such positive feedbacks. Through adaptive actions and learning, actors in the wildfire governance system can become aware of what parts of the system resist change, and where new policies, networks, or organizations may make a difference. Such a framework may help expand the problemsolving capacity needed to address the pathology of wildfire risk at appropriate spatial, temporal, and social scales. Changing a pathological system is difficult because the conditions and processes that engender the pathology are highly resilient. We caution that even with clear understanding of the wildfire pathology and possible solutions, governance systems may evolve incrementally and in imperfect ways, continuing to resist change even as we learn better ways to manage CNHS. Nevertheless, a fireadapted governance system that engages a wide array of human actors in social networks and planning processes that promote complex thinking about the future offers the best chance of mitigating the wildfire risk pathology, whether in the US or in fire-prone temperate forests elsewhere in the world. J Acknowledgements This paper emerged from a workshop sponsored by the US National Science Foundation’s (NSF’s) Coupled Human and Natural Systems Program (NSF grant CNH1013296), the USDA Forest Service PNW Research Station, and the Joint Fire Science Program (JFSP Project Number 12-5-01-15). We acknowledge support from NSF grants CNH-1013296, CNH-0816475, CNH1313688, GEO-1114898, and DEB-1414041, and we thank A Agrawal and DL Peterson for providing comments on an earlier version of this paper. Any use of trade, firm, or product names is for descriptive purposes © The Ecological Society of America Wildfire risk as a socioecological pathology only and does not imply endorsement by the US Government. J References Acemoglu D, Ozdaglar A, and ParandehGheibi A. 2010. Spread of (mis)information in social networks. Game Econ Behav 70: 194–27. Baker W. 2015. Are high-severity fires burning at much higher rates recently than historically in dry-forest landscapes of the western USA? PLoS ONE 10: e0136147. Bodin Ö and Crona BI. 2009. The role of social networks in natural resource governance: what relational patterns make a difference? Global Environ Chang 19: 366–74. Bowman DMJS, Balch J, Artaxo P, et al. 2011. The human dimension of fire regimes on Earth. J Biogeogr 38: 2223–36. Busby G and Albers H. 2010. Wildfire risk management on a landscape with public and private ownership: who pays for protection? Environ Manage 45: 296–310. Butler WH and Goldstein BE. 2010. The US fire learning network: springing a rigidity trap through multiscalar collaborative networks. Ecol Soc 15: 21. Calkin DE, Cohen JD, Finney MA, et al. 2014. How risk management can prevent future wildfire disasters in the wildland– urban interface. P Natl Acad Sci USA 111: 746–51. Canton-Thompson J, Gebert KM, Thompson B, et al. 2008. External human factors in incident management team decisionmaking and their effect on large fire suppression expenditures. J Forest 106: 416–24. Chapin FS, Trainor SF, Huntington O, et al. 2008. Increasing wildfire in Alaska’s boreal forest: pathways to potential solutions of a wicked problem. BioScience 58: 531–40. Davidsson P. 2003. A framework for preventive state anticipation. In: Butz MV, Sigaud O, and Gerard P (Eds). Anticipatory behavior in adaptive learning systems: foundations, theories, and systems. Berlin, Germany: Springer-Verlag, Berlin. Dennison PE, Brewer SC, Arnold JD, et al. 2014. Large wildfire trends in the western United States, 1984–2011. Geophys Res Lett 41: 2928–33. Doniec A, Mandiau R, Piechowiak S, et al. 2008. Anticipation based on constraint processing in a multi-agent context. Auton Agent Multi-Ag 17: 339–61. Donovan GH and Brown TC. 2007. Be careful what you wish for: the legacy of Smokey Bear. Front Ecol Environ 5: 73–79. Duit A and Galanz V. 2008. Governance and complexity – emerging issues for governance theory. Governance 21: 321–35. Fischer AP and Jasny L. Capacity to adapt to environmental change: evidence from a network of organizations concerned with increasing wildfire risk. Ecol Soc. In review. Fischer AP, Vance-Borland K, Jasny L, et al. 2016. A network approach to assessing social capacity for landscape planning: the case of fire-prone forests in Oregon, USA. Landscape Urban Plan 147: 18–27. Flannigan M, Cantin AS, de Groot WJ, et al. 2013. Global wildland fire season severity in the 21st century. Forest Ecol Manag 294: 54–61. Folke C, Carpenter S, Walker B, et al. 2004. Regime shifts, resilience, and biodiversity in ecosystem management. Annu Rev Ecol Evol S 35: 557–81. Gunderson LH and Pritchard L. 2002. Resilience and the behavior of large-scale systems. Washington, DC: Island Press. Headwaters Economics. 2013. The rising cost of wildfire protection. http://headwaterseconomics.org/wphw/wp-content/ uploads/fire-costs-background-report.pdf. Viewed 02 Feb 2016. Hulse D, Branscomb A, Enright C, et al. Anticipating surprise: using agent-based alternative futures simulation modeling to identify and map surprising fires in the Willamette Valley, Oregon, US. Landscape Urban Plan. In press. www.frontiersinecology.org 283 Wildfire risk as a socioecological pathology 284 Hulse D, Eilers J, Freemark K, et al. 2000. Planning alternative future landscapes in Oregon: evaluating effects on water quality and biodiversity. Landscape 19: 1–19. Johnson KM and Beale CL. 2002. Nonmetro recreation counties: their identification and rapid growth. Sociology Scholarship. Paper 75. http://scholars.unh.edu/soc_facpub/75. Viewed 30 Mar 2016. Jolly WM, Cochrane MA, Freeborn PH, et al. 2015. Climateinduced variations in global wildfire danger from 1979 to 2013. Nat Commun 6: 1–11. Kalabokidis K, Iosifides T, Henderson M, et al. 2008. Wildfire policy and use of science in the context of a socioecological system on the Aegean Archipelago. Environ Sci Policy 11: 408–21. Liu J, Dietz T, Carpenter SR, et al. 2007. Complexity of coupled human and natural systems. Science 317: 1513–16. Maguire LA and Albright EA. 2005. Can behavioral decision theory explain risk-averse fire management decisions? Forest Ecol Manag 211: 47–58. McCaffrey SM. 2004. Thinking of wildfire as a natural hazard. Soc Natur Resour 6: 509–16. McCaffrey SM and Olsen CS. 2012. Research perspectives on the public and fire management: a synthesis of current social science on eight essential questions. General Technical Report NRS-104. Newtown Square, PA: USDA Forest Service, Northern Research Station. Moreira F, Viedma O, Arianoutsou M, et al. 2011. Landscape–wildfire interactions in southern Europe: implications for landscape management. J Environ Manage 92: 2389–402. Moritz MA, Batllori E, Bradstock RA, et al. 2014. Learning to coexist with wildfire. Nature 515: 58–66. NIFC (National Interagency Fire Center). 2015. www.nifc.gov/ fireInfo/fireInfo_statistics.html. Viewed 28 Dec 2015. North MP, Stephens SL, Collins BM, et al. 2015. Reform forest fire management. Science 349: 1280–81. Peterson GD, Cumming GS, and Carpenter SR. 2003. Scenario planning: a tool for conservation in an uncertain world. Conserv Biol 17: 358–66. Reuters. 2013. www.reuters.com/article/us-usa-wildfires-insurance-idUSBRE97D0W620130814. Viewed 28 Dec 2015. Shindler B and Toman E. 2003. Fuel reduction strategies in forest communities: a longitudinal analysis of public support. J Forest 101: 8–15. So R and Sonenberg L. 2004. Agents with initiative: a preliminary report. In: Nickles M, Rovatsos M, and Weiss G (Eds). Agents and computational autonomy: potential, risks, and solutions. Berlin, Germany: Springer-Verlag, Berlin. www.frontiersinecology.org AP Fischer et al. Spies TA, White EM, Ager AA, et al. Using an agent-based model to examine alternative futures for fire and ecosystem services in a multi-ownership landscape in Oregon. Ecol Soc. In review. Spies TA, White EM, Kline JD, et al. 2014. Examining fire-prone forest landscapes as coupled human and natural systems. Ecol Soc 19: 9. Stavros EN, Abatzoglou J, McKenzie D, et al. 2014. Regional projections of the likelihood of very large wildland fires under a changing climate in the contiguous western United States. Climatic Change 126: 455–68. Steelman TA and Burke CA. 2007. Is wildfire policy in the United States sustainable? J Forest 105: 67–72. Stephens SL, Agee JK, Fulé PZ, et al. 2013. Managing forests and fire in changing climates. Science 342: 41–42. Stephens SL, Burrows N, Buyantuyev A, et al. 2014. Temperate and boreal forest mega-fires: characteristics and challenges. Front Ecol Environ 12: 115–22. Theobald DM. 2001. Land-use dynamics beyond the American urban fringe. Geogr Rev 91: 544–64. Tweedale J, Ichalkaranje N, Sioutis C, et al. 2007. Innovations in multi-agent systems. J Netw Comput Appl 30: 1089–115. Valente TW. 2012. Network interventions. Science 337: 49–53. Westerling AL, Hidalgo HG, Cayan DR, et al. 2006. Warming and earlier spring increase western US forest wildfire activity. Science 313: 940–43. Williams J. 2013. Exploring the onset of high-impact mega-fires through a forest land management prism. Forest Ecol Manag 294: 4–10. Wilson RS, Winter PL, Maguire LA, et al. 2011. Managing wildfire events: risk-based decision making among a group of federal fire managers. Risk Anal 31: 805–18. Yoder J and Blatner K. 2004. Incentives and timing of prescribed fire for wildfire risk management. J Forest 102: 38–41. 6 USDA Forest Service, Pendleton, OR; 7University of Colorado– Boulder, Boulder, CO; 8USDA Forest Service, Portland, OR; 9 USDA Forest Service, Davis, CA; 10University of Maine, Orono, ME; 11US Geological Survey, Anchorage, AK; 12King’s College London, London, UK; 13Portland State University, Portland, OR; 14 University of Idaho, Moscow, ID; 15Southern Methodist University, Dallas, TX; 16University of New England, Biddeford, ME; 17 University of Louisville, Louisville, KY; 18USDA Forest Service, Olympia, WA; 19University of Tasmania, Hobart, Australia © The Ecological Society of America