Small-island communities in the Philippines prefer local measures to relocation in response to sea-level rise


Topographical and inundation survey.

Ground elevation was measured using a laser ranging instrument (IMPULSE, Laser Technology) and a prism during a number of field surveys in March 2016. Starting from the waterline, the elevation data of several transects along the islands were gathered, with the position of each point also determined using GPS instruments. Elevation measurements were corrected for tide. Direct measurements of flood heights and areas throughout the islands were taken during the peak of high tide events on 6 June and 7 June 2016. Based on these direct measurements, the flood threshold was estimated and the number of flooded days in 2016 was counted using tidal software (for the nearby Cebu tide station). Supplementary Fig. 2 shows the number of days with high tides greater than +1.6m in 2016.

Adaptation options and preferred strategies.

One focus group discussion was conducted per island to understand the general impact of tidal flooding on residents, and to identify the communities adaptation options. Each focus group discussion was attended by 6 to 9 local leaders, including: elected officials such as village captains and councillors, and peoples organization leaders; and appointed leaders such as village nutritionists, secretaries and treasurers. A total of 28 local leaders participated in the discussions held in 4 islands, 8 of whom are female and 20 male (see Supplementary Table 1 for summary). All focus group discussions were conducted in Cebuano, the local language, and audio-recorded.

Each discussion lasted 1.5h on average, and was divided into four parts. First, the participants were asked basic questions about the severity and extent of tidal flooding in their respective communities. Specifically, the questions pertained to inundation height, duration and frequency, as well as affected areas and previous flooding experiences. Second, the participants described their typical sources of disaster risk information, their evacuation practice before and after the 2013 earthquake, and their disaster experience during several post-earthquake typhoons. The third part was the main focus of the discussions. In this part, the participants explored the various problems that their communities were experiencing due to tidal flooding, and enumerated the different adaptation strategies they are implementing. Various aspects of community life were covered, including housing, education, livelihood, health and sanitation, and food and water. The participants also expressed their views regarding the relocation programme developed by the Municipal Government of Tubigon, and whether their communities were willing to participate in it. Fourth, the participants shared other problems that their communities were experiencing that did not directly relate to tidal flooding, such as damages from the earthquake, illegal fishing and overfishing.

Key informant interviews with elderly members of the communities were also conducted to gather information about the islands original housing and flooding conditions. One elderly member was interviewed per island, selected on the basis of age (that is, 70 years old and above) and place of birth (that is, on the island). A total of 4 elderly members were interviewed in 4 islands, 3 of whom are male and 1 female. All key informant interviews were conducted in the local language, and audio-recorded.

Each interview session lasted an hour on average, and was facilitated by using a recent satellite image of the relevant island obtained through Google Earth. Using the image as reference, the interviewees described how the islands looked in their childhood. They noted changes in the natural environment, including mangrove clearing and planting, and beach erosion. They also highlighted changes in the built environment, including the number, location and design of the original houses on the islands, as well as the manner in which the communities developed through the years. Finally, the interviewees related the flooding situation in the islands prior to the earthquake, and discussed how a deadly typhoon called Nitang affected their communities in 1984.

Data from focus group discussions with local leaders and key informant interviews with elderly members were translated and transcribed in English, and analysed via qualitative content analysis to uncover dominant and shared messages across the four communities responses.

A self-administered, paper-based household questionnaire survey was conducted in March 2016 to gather information about the residents preferred adaptation strategies on the household level. Since the island communities have small populations (see Supplementary Table 2), the study ran a total population sampling, with households as units.

Any literate member of the household was allowed to participate in the survey, which was written in the local language. The surveys were distributed in the morning, and collected in the afternoon of the same day through the help of village health workers. A cover page was included in the questionnaire to provide information about the researcher and the purpose of the survey, and instructions for respondents to indicate truthful and complete answers.

The questionnaire survey was divided into five sections, and included a total of 70 questions. The respondents answered the questions by encircling their answers from the choices given, or by filling their information in the spaces provided. The first section asked about the basic information of the individual respondent.

The second section was the main focus of the questionnaire. In this section, the respondents were asked about the severity and extent of tidal flooding experienced by their respective households (that is, inundation height, duration and frequency). They were also asked to indicate which adaptation strategies their households are implementing. Various adaptation options were included in the choices, such as migration, hard measures (that is, stilted houses and raised floors) and soft measures (that is, evacuating within the island, elevating belongings and being vigilant during night-time flooding). These options were identified and verified on the basis of the results of the focus group discussions with local leaders. Multiple answers were allowed, as households often implement a combination of adaptation strategies.

In the third section, the respondents were asked about their households sources of disaster risk information, and their evacuation behaviour before and after the earthquake. In the fourth section, the respondents were asked to provide further information about their households, such as their main and alternative livelihoods, number of dependents, savings and government benefits, and access to the mainland. In the fifth section, the respondents were asked about their households relationship with their neighbours and their participation in community activities.

The survey collected valid responses from 476 out of 547 households that reside in the 4 island communities, yielding a margin of error of 2% with a confidence level of 95%. The responses were analysed via descriptive statistics using SPSS. Demographic information about respondents are summarized in Supplementary Fig. 3.

Effectiveness of hard measures.

The actual effectiveness of adaptation strategies, and specifically hard measures (for example, stilted houses, raised floors and raised roads), was also evaluated in 6 and 7 June 2016, during the peak tide. While conducting flood height measurements the heights of the stilts or raised floors of the houses or roads were also measured. The median values of these measurements were then compared with the median of the flood heights.

Results show that not all of the hard measures being implemented, especially in islands with medium flooding severity, are effective against flooding. Assessment during actual high tide events shows that only stilted houses were effective, while raised floors were not.

None of the stilted houses were flooded during high tide events, and on average were significantly higher than the water levels, potentially even providing an allowance for high waves during typhoons and monsoons (Supplementary Table 3). However, they should be properly engineered against strong winds. To this end, the donated stilted houses in Ubay have been designed to have short roof overhangs, although this inadvertently created another problem with rain protection. From an engineering perspective, regular soaking of the walls can cause the houses to quickly deteriorate. Nonetheless, later versions of stilted houses donated by the Bohol Rehabilitation and Recovery Program of the Diocese of Tagbilaran to some households in Batasan, Pangapasan and Bilangbilangan have reverted to standard roof overhangs. Efforts to secure the roof from strong winds, build strong footing, brace walls and strengthen joints have also been made. Still, recipients can implement more steps to reinforce their houses as needed, such as through using guy wires.

The history of disasters in the islands of Tubigon indicates the importance of designing houses to become resilient to multiple hazards. Interviews with elderly residents reveal that the original houses on the islands were also stilted wooden houses, but that these were completely wiped out when typhoon Nitang (Ike) hit them in 1984. Following this experience, many residents have rebuilt their houses using concrete.

In contrast, houses with raised floors were still frequently flooded, especially in islands with medium flooding severity, and on average were hardly higher than the flood level. Raised roads were completely flooded during high tides in all islands except in Ubay, where its median height above former ground level was 70.5cm. However, since it is only +0.28m higher than the median flood height, it could still be easily flooded by high waves.

Environmental impact of coral mining.

Three out of the four islands engaged in coral mining to raise the floors of houses and prevent flooding or to reclaim land to build more houses. To understand the history of this practice, the study conducted key informant interviews with elderly members of the communities.

To measure the environmental impact of coral mining, the authors conducted reef cover assessments of the marine protected areas (MPAs) and mined areas in the reef flats surrounding the islands. Following the point intercept transect method devised by the Reef Check Foundation25, the type of reef cover was noted at every 0.5m along 20m transects. Four replicates were placed 5m apart along a 100m transect, which was positioned parallel to the coastline. The GPS locations of the start and end points of this 100m transect were also noted.

Ten different types of reef cover were considered in the assessment, including hard corals, dead corals and dead corals with algae. The mean percentage cover was calculated by taking the mean of the ratio between the number of points falling under each type and the total number of points per replicate.

The actual height of coral piles along the perimeter of the islands was also surveyed.

Key informant interviews with elderly members explained that, at the start of the twentieth century the first generation of island dwellers (mostly fishermen who migrated from Cebu) began engaging in coral mining, primarily to build coral piles that served as foundations for their houses, roads and community buildings (Supplementary Fig. 4a). Succeeding generations then continued to engage in this activity and further encroached on the sea to build new houses for the expanding community. Thus, although illegal under National Law, coral mining is already part of the islands way of life and, as a long-standing practice, already has its own established local implementation system.

Following this, after the 2013 earthquake coral mining became one of the most readily accessible ways for households to raise their floors (Supplementary Fig. 4b). However, coral mining could be considered as a maladaptive strategy, especially when too many stones are taken at the same time. It can basically damage the reefs that serve as natural barriers against waves, rendering island communities more vulnerable to flooding, especially during monsoons and typhoons.

As mentioned earlier, the use of coral stones to elevate houses has not been effective to prevent them from being flooded during high tides, although it has been more successful in compensating for normal tides. The median heights of coral piles in Ubay, Pangapasan and Batasan are 157.5cm, 166cm and 164cm, respectively, although site observation suggests that coral piles have been implemented far more widely in Batasan.

Compared with coral piles, raised floors were simply not high enough to be able to compensate for flooding during high tides. Several reasons have prevented locals from raising the floors higher by using this technique, including the high costs involved in coral mining and the depletion of corals. One cubic metre of coral stones costs ₱: 200–250 (US$4–5). On the other hand, coral reef cover surveys revealed that live coral cover in the mined areas ranges from only 1% to 3%, while dead coral cover (with or without algae) ranges from 29% to 50% (Supplementary Fig. 5). In comparison, the MPA of Batasan has live coral cover of 47% and dead coral cover (with or without algae) of 30%. However, the MPAs of Ubay and Pangapasan were not as well-maintained and instead have conditions very similar to mined areas, with live coral covers of only 6% and 2%, and dead coral covers (with or without algae) of 48% and 54%, respectively. Live coral covers of less than 10% indicate that the reefs have been seriously damaged 26, which can negatively impact the sediment budget and affect long-term natural island morphological mechanisms27.

Such excessive coral mining highlights the lack of a multiple hazard approach towards adaptation, which can jeopardize the long-term sustainability of island communities. While there is no doubt that even the poorest communities will engage in a variety of adaptation strategies when the impacts of sea-level rise become more apparent in the future, there is a need to examine more carefully the risk of maladaptation.

Ethics statement.

The research was conducted according to the University of Tokyos regulations concerning questionnaire surveys and interviews. The University of Tokyo does not require any specific committee approval for questionnaire surveys of the nature carried out in the present study. Informed consent was obtained from participants for the use of their answers in the research. All photos were taken with the informed consent of the subjects.

Data availability.

Data supporting the findings of this study can be made available by contacting the corresponding author, provided that the request does not infringe on respondents privacy.