Rethinking the limits of climate change adaptation

Using a case study approach, this research explores the various potential limiting factors of climate change adaptation, based on the experience of four low-lying islands in central Philippines. In the aftermath of the 2013 7.2-magnitude Bohol earthquake, the islands now become flooded even during normal spring tides. Results show that, while development problems can constrain adaptive capacity against climate hazards, the outlook of the affected communities about their future on their home islands may ultimately determine the limits of adaptation. As social factors may play a greater role in adaptation than environmental factors, climate-smart development is needed, as well as proper education regarding climate risks and available adaptation options.

, global mean sea level is likely to increase between 26 centimeters (cm) and 82 cm by the end of the 21st century (Church et al. 2013). Still, more recent work on probabilistic process-based models, which take into account rapid losses in the Antarctic ice sheets, put the potential for sea level rise (SLR) to an upper range of 2.97-3.39 meters for Representative Concentration Pathway 8.5 scenarios (Kopp et al. 2017; Le Bars, Drijfhout, and de Vries 2017). As emissions currently show no signs of abating despite the signing of the Paris Agreement in 2015, SLR could be well within the higher ranges of projections. Further, a special report published by the IPCC describing a 1.5ºC warmer world indicates that more severe climate impacts may occur by the year 2040, unless significant and comprehensive actions to reduce emissions are taken by 2030 (IPCC 2018).
Aside from SLR, tropical cyclone intensification increases the risk of flooding in low-lying coastal communities (Nurse et al. 2014). Moreover, population increase and urbanization could lead to greater losses from climate impacts (Jongman, Ward, and Aerts 2012). With this, many existing studies suggest that climate change will inevitably lead to mass migration or the final abandonment of small low-lying islands (Myers 2002). In particular, it has been hypothesized that, in the long term, SLR coupled with mass coral mortality could render islands uninhabitable due to permanent inundation and land loss (Yamamoto and Esteban 2014). However, more recent studies claim that mass migration could happen even in the short term due to annual flooding, saltwater intrusion, and higher risk of storm surges, causing frequent damage to infrastructure, and disrupting food and water supply (Keener et al. 2013, Storlazzi et al. 2018. Existing studies on mass migration typically frame adaptation to be a problem of tipping points (see Box on page 2), where in situ adaptation is often considered to be possible only up to a certain point, after which retreat becomes inevitable or even necessary (McLeman 2011, Barnett and O'Neill 2011).

Typical Framing of Climate Change Adaptation
Note: In situ and retreat strategies are often portrayed to be mutually exclusive. Source: Authors.
The typical adaptation framing, which portrays in situ and retreat strategies as being mutually exclusive, has at least two problems (see figure on page 1). First, tipping points are often expressed in terms of environmental factors, such as the degree of sea level rise (Nicholls and Cazenave 2010). However, an increasing number of studies indicate that social factors may play a greater role in the final decision to relocate (Perch-Nielsen, Bättig, and Imboden 2008;Arenstam-Gibbons and Nicholls 2006). Second, retreat strategies are shown to be a failure of in situ adaptation (Black et al. 2011). However, presenting these two main options as being mutually exclusive can inadvertently prevent communities from formulating a more integrated approach to climate change adaptation. Further, this false dilemma between retreat and in situ adaptation does not only oversimplify the problem of climate change (Nunn 2009), but also potentially leads to conflicting adaptation priorities and competition for limited financial resources.
This study aims to examine this typical framing of adaptation by exploring how the various potential limiting factors discussed in literature actually apply, or not apply, in the case of four low-lying island communities in central Philippines that currently experience the most extreme tidal flooding in modern history. By better understanding the factors that may play an important role in determining the limits of climate change adaptation, this study also hopes to elucidate the complex relationship between in situ and retreat strategies. This paper is organized as follows. Section II describes the extreme case of tidal flooding in the islands of Tubigon, Bohol considered in this study. Section III outlines resulting problems and available strategies, and section IV discusses factors that could potentially limit adaptation. Section V concludes with remarks on how to reframe our understanding of adaptation, and provides recommendations for future studies.

II. METHODOLOGY
On 15 October 2013, a 7.2-magnitude earthquake-induced land subsidence in four small island communities off the shore of Tubigon, Bohol (namely, Batasan, Ubay, Pangapasan, and Bilangbilangan) (see picture of the aerial view of Bilangbilangan island on page 3), causing them to now become inundated during high tides, especially during spring tides. The level of land subsidence and problems affecting these islands can be used as a proxy to understand how other low-lying islands around the world will adapt to SLR in the future, as they represent one of the most severe cases of tidal flooding in an island environment in modern history. In comparison, some atoll states currently experience severe inundation only about once a year, during king tides.
The four islands of Tubigon have very small land areas (between 1.5 hectares and 6 hectares only), with maximum elevations of about 2 meters above water (Jamero et al. 2017). They are densely populated, with each island having 60-250 households who mainly engage in small-scale fishing for livelihood (Jamero et al. 2018). The islands are impoverished, with the number of households with incomes below the income poverty threshold being significantly higher than the national average (Jamero et al. 2018).
This study uses a case study approach to examine the potential limits of adaptation. Drawing on the results of inundation surveys, site observations, and key informant interviews conducted by the authors from 2015 to 2018, the study compares the theoretical discussions on the limits of adaptation available in literature with the actual adaptation experiences of the island communities of Tubigon.

A. Physical or Biological Limit
The reefs around the islands are built by hard corals, and the islands themselves by the sediments that are produced when hard corals die. Thus, for islands to grow and continue to survive, it is important for corals to remain healthy (Yamamoto and Esteban 2014). However, a number of studies have highlighted the vulnerability of corals to climate change, particularly due to global warming and ocean acidification (Nurse et al. 2014). Following this logic, the very survival of islands is also threatened by climate change, as they face the double problem of SLR and the lack of sediment supply.
While several massive bleaching events have been documented in many places in recent times (Nurse et al. 2014), new studies are beginning to explore the feasibility of coral preservation and transplantation (Yamamoto and Esteban 2014). On the other hand, 4 decades' worth of remotely sensed data regarding the shoreline changes of all 101 islands of Tuvalu show a net increase in land area of the country as a whole, despite the fact that its local SLR was twice as much as the global average within that period (Kench, Ford, and Owen 2018). Similar results have been found by studies conducted in the 29 islands of Funafuti Atoll  as well as 12 different atolls in central and western Pacific . Overall, these studies have found that, while some islands have decreased in size, many have also increased. This lack of uniformity in the islands' response to SLR reflects the highly dynamic nature of these systems, which continually adjust their size, shape, and position within the reef surfaces in response to various environmental changes . Ultimately, these findings challenge the notion that islands will become uninhabitable or even disappear in the face of SLR.
Confirming this point, the population of the island communities of Tubigon remain in their homes, despite coming face-to-face immediately after the earthquake with a degree of SLR not expected in other parts of the planet until the year 2100. Although the islands experience partial or complete flooding every spring tide, their daily lives have continued, with their social and cultural relations not appearing to have deteriorated. Thus, the height and even frequency of tidal flooding may not be sufficient for determining the limits of adaptation. To the contrary, the study found that these factors influence more significantly the level of adaptation, with island communities that are more severely affected also implementing adaptation measures more aggressively (Jamero et al. 2017).
The islands of Batasan and Ubay, which experience a higher degree of tidal flooding across more days within a given year, are widely investing in hard measures such as elevating floors and roads, and building stilted houses. On the other hand, Bilangbilangan and Pangapasan prefer soft measures, such as installing taller furniture at home and elevating personal belongings before tidal flooding. Rather than relocating, all of the islands have changed their evacuation behavior, relocating temporarily to the mainland in response not only to strong typhoons, but also to weak weather disturbances, whenever these coincide with high tides (Jamero et al. 2017).
However, as Batasan and Ubay (see picture on page 7), until recently, have been using coral stones as filling materials, examining the impact of rampant coral mining on the long-term physical viability of the islands is needed. A survey of a mined reef section in Batasan revealed a live hard coral cover of less than 10%, with a further 30% of dead coral with algae, indicating that the area is already severely degraded (Jamero et al. 2017 Also, extreme events such as typhoons and droughts often are perceived to constitute a limit of adaptation, potentially driving mass migration even in the short term (Nakamura et al. 2016, Fletcher and Richmond 2010, Yamamoto and Esteban 2014. However, this study found that the island communities of Tubigon have remained in their homes despite experiencing several typhoons since the 2013 earthquake, and a 4-month long dry spell in 2016. Rather, they have adapted by changing their evacuation behavior and by acquiring more rainwater collectors or, at worst, importing water from the mainland (Jamero et al. 2018). Further, it can be argued that tidal flooding, which began immediately after earthquake-induced land subsidence, can also be considered an abrupt and extreme event. Still, the communities of Tubigon were able to adapt and continue to live on the islands. Overall, these experiences indicate that extreme events also may not be sufficient to breach the limits of adaptation.
Although the impact of high-order extreme events (such as a category 4 or category 5 typhoon) remains to be seen, it is not clear whether this is possible in such sheltered areas of the Philippines.

D. Development Challenges
While the literature on climate change adaptation for small islands still focuses primarily on the physical hazards of climate change (Kelman 2014), the more established literature on disasters has already widely accepted that disaster risk is a function not only of physical hazards from the natural environment, but also of exposure and vulnerability due to human actions (Wisner et al. 2004). In other words, poor socioeconomic development can increase disaster risk by increasing the vulnerability of populations exposed to natural hazards.
Within the context of climate change, climate vulnerability has likewise been found to be mainly socioeconomic in nature Tol 2006, Ayers andForsyth 2009). Given this, the ability to adapt is also being related to the state of development (Smit et al. 2001). In simpler terms, as climate risks are driven by development-related challenges, climate change adaptation must therefore also focus on development. However, this point has not yet been sufficiently recognized in literature. Rather, by focusing primarily on quantifying climate hazards for small islands, many studies on adaptation have highlighted the need for mass migration instead of the need for development.
In line with this, the study found that the small problems that the island communities of Tubigon experience on a daily basis prove more critical. In particular, it is their persistent development problems such as poverty, limited infrastructure, and poor resource management that slowly erodes their capacity to adapt to slow onset environmental changes and extreme events in the long term. This affirms that social factors may have a greater influence on the limits of adaptation, rather than environmental factors alone.

E. Future Outlook of Communities
In Holland Island in Chesapeake Bay, US, where relative SLR accelerated in the mid-19th century, the community initially simply relocated to higher grounds within the island (Arenstam-Gibbons and Nicholls 2006). However, land appropriate for relocation eventually ran out, prompting some families to begin leaving the island. Finally, island abandonment was triggered when the remaining population fell below a level that could support the community. With the closure of important services such as schools, churches, and stores, what was left of the community lost faith in their own future on the island and migrated.
Rethinking the Limits of Climate Change Adaptation | 9 This suggests that the future outlook of the communities themselves may ultimately determine the limits of adaptation. Thus, it is important to guide the communities in understanding fully and evaluating properly the climate risks that they are faced with, as well as the adaptation options that are available to them (including relocation), while giving them the space to contemplate their own future based on their own hopes and their existing social and cultural values.
However, the outlook of island communities does not necessarily have to be negative in order to prompt final abandonment. As envisioned by the Municipal Government of Tubigon, by providing better alternatives to the communities (i.e., potentially better opportunities as a skilled worker earning a stable income in the mainland), they may become inspired to start building their future elsewhere, triggering the natural depopulation of the islands.

V. CONCLUSION
While adaptation is typically understood to be a problem of environmental tipping points, this study highlights the need to reframe adaptation as a matter of choice, based on the affected population's outlook on their own future. At-risk communities should be able to decide for themselves the best way to adapt to climate change, given their own contexts and resources. However, to be able to make an informed choice, the communities need to have full awareness of the various adaptation options that are available to them, and a proper understanding of the risks involved in continuing to live on the islands, as well as in moving to other locations. Still, more importantly, the communities need to be empowered by promoting sustainable development as a means of building adaptive capacity.
With this, in situ strategies that address both the physical hazards posed by climate change and the socioeconomic drivers of vulnerability may be a good starting point for communities to adapt. In particular, in situ strategies should be able to address climate and disaster risks, while promoting socioeconomic development at the same time, so as to increase the overall adaptive capacity. In so doing, in situ strategies also can enable planned relocation, allowing island communities to move if, when, and where they choose to, avoiding the risk of sudden disaster-induced displacement, and minimizing the constraints posed by long-standing socioeconomic issues.

Recommendations for Future Studies
This study highlights the need to reconsider what it means for an island to become "uninhabitable" from both environmental and social perspectives. Based on examples from the Philippines, although environmental thresholds may already have been breached, communities were still able to continue to inhabit their islands. As discussed previously, the outlook of communities is important in determining whether the islands can still be deemed as capable of supporting their own future, and therefore remain habitable.
With few livelihood options, poor education, high poverty incidence, inadequate utilities and infrastructure, and small areas and populations, the island communities of Tubigon would normally be considered to have low adaptive capacities. Despite these constraints, community members and households were still able to implement a good number of adaptation strategies based on their own initiatives. This shows that island communities are far more resilient than otherwise assumed by the mass migration theory. Therefore, scientific experts and policy makers should consider also how to measure adaptive capacity more accurately in order to supplement it more effectively.
Finally, future studies that identify more environment-sensitive strategies and investigate how existing maladaptive strategies may be corrected are also clearly needed to promote the long-term sustainability of coastal areas. While the level and limits of adaptation are potentially determined solely by the affected communities, scientists at least can provide support to understand and prevent maladaptation, and suggest appropriate adaptation strategies based on a given island's characteristics. Thus, this study highlights the need for greater financial and technical support from governments, nongovernment organizations, and academia, where self-managed efforts of island communities fall short. However, at the same time, it is important for these external agencies to view the island communities not only as beneficiaries, but rather as partners in increasing climate resilience.