1Much recent attention has been given to likely future climate-induced population displacements in Oceania (ejf, 2009; Mimura et al., 2007), yet within the region, forced relocation is not a new phenomenon (Lieber, 1977). In the immediate aftermath of World War Two (wwii), at the time when millions of Europeans were on the move, one tiny island community in the central Pacific was also, reluctantly, undertaking a long-distance journey to a new island home.

3Compared with the many hundreds of thousands of Pacific islanders who face potential relocate in the future owing to climate change (Stern, 2006, Mimura et al., 2007), the historic relocation of the Banabans may be small in scale; however, valuable lessons can be learned from the relocation approach.

4Rarely in the past has an entire island community been forcibly displaced, the process so well documented in the historical archives, and the narrative of the relocation so readily told by the community who experienced the resettlement. The embedment of a narrative of mining exploitation, and potentially of climate, is important in synthesizing the relocation process, but it also needs to be seen as part of a broader history of mobility in the Pacific-island context (Modell, 2002). People within the Pacific have migrated through choice, circumstance or opportunity, and the migration narratives have enriched Pacific traditions and cultures. Mobility is in the blood of Pacific islanders (Hau’Ofa, 1994).

Issue and hypothesis: from an actual case of forced relocation to an hypothetical/prospective comparison

5This article examines the relocation of the Banabans to northern Fiji in 1945 in terms of the decisions and processes made by the colonial authorities, and the experiences of the Banaban community.

6Much of the focus of the work is on the historical process of resettlement, but a current-day perspective is also given in observations related to community reconstruction. Any relocation is an extremely complex process; affected communities are exposed to risks of severe economic, social and environmental hardship, and in the case of mining-induced displacement especially, environmental degradation may be particularly acute (Robinson, 2003).

7Today there is a realisation that very careful planning is required to minimize the trauma of the upheaval for the displaced community. Past development-induced relocations have been criticised for their apparent lack of concern for the social dynamics of the displaced community (Cernea, 1997). In many cases, people were an afterthought in such schemes. Outwardly, the handling of the Banabans by colonial authorities could receive a similar judgment. Beyond forced resettlement and in broader terms of mobility and community building in the Pacific context, Pacific islanders have left the «familiar» and migrated to the «unknown», where they have been successful in creating «moral communities» (Modell, 2002). This heritage of movement, nevertheless, enables Pacific islanders to be still rooted in the ancestral homeland through oral traditions, genealogy and cultural performance (Hau’Ofa, 1994). The work here assesses both the extent to which the needs of the Banaban people were taken into consideration and the success of the community in retaining their traditions and society in the island of Rabi, their new home.

8The remainder of the paper is structured as follows. In the next section, the approach and method is outlined, which includes a short geography of Banaba Island. Then the background to the discovery of phosphate in Banaba is given, followed by a more detailed presentation of the different stages in the relocation process.

9The assessment of the relocation approach includes modern-day observations of the Banabans’ situation. In the comparison section, parallels with current climate change-forced relocations are stated and recommendations presented that will aid future climate-induced community relocations. In the last section overall conclusions are drawn.

Approach and method

A historical extended case study

10This paper examines the relocation of the Banabans to northern Fiji in 1945. It is a case-study approach that uses archive records and other documentation to assess the level of planning, negotiations and co-operation between colonial authorities responsible for the relocation. Much of the focus of the work is on the historical process of resettlement, though the research detailed here also incorporates observations and findings from a recent study-team visit to Rabi Island in April 2011. During the field visit, interviews with key representatives of the community were undertaken, enabling a current-day perspective to be given to the assessment.

11The work examines the extent to which the needs of the Banaban people were considered in the relocation, and also outlines lessons from the relocation that may be applied to future climate change-induced resettlements.

12The author is indebted to the late, Mrs. Makin Corrie, a notable elder of the Banaban community in Suva, who, over a series of meetings, recalled her part in the relocation process. Her oral narrative proved invaluable in enriching the do­cumented history of the Banabans. All the oral histories and personal reflections of the selected Banaban interlocutors are analyzed, edited, and presented here, in conjunction with the archive record. On occasion, the oral history confirms the historical data; however, often, the viewpoint, values and knowledge of an interviewee are treated as unique though not necessarily factual perspectives of the past. With the reliability of the life-history technique highly questionable (Cary, 1999), the Banaban narrative experiences are included here to offer the reader only a construction of the cultural modeling of the relocation reality.

13The study team, consisting of staff from the climate-change unit of the Pacific Conference of Churches, also acknowledge the invitation to attend a meeting of the elders at Banaban House, Suva in March 2011, and are grateful to Talatala Qase Kabong, Divisional Superintendent of the Methodist Church in Rabi, who smoothed the Rabi Island visit and enabled the team members to meet with a wide cross-section of the Banaban community during their stay. The research outlined in this article forms part of the on-going work of the Pacific Conference of Churches in the accompaniment of current-day Pacific islanders facing displacement because of climate change.

A geography of Banaba Island

14Banaba, a small, raised island, is the emerging tip of a submarine mountain (the name «Banaba» means island of rock or «stony»). This barren outcrop is located just south of the equator in the central Pacific, and is very remote. Nauru, its nearest neighbour, is almost 200km to the west, and the administrative centre of Tarawa, the capital of Kirabati (formerly the Gilbert and Ellice Islands Group) 400km away to the east (figure 1).

From phosphate discovery to the resettlement of the Banabans

Discovery of phosphate

15Arthur Ellis, a young geologist at the Pacific Islands Phosphate Company in Sydney, traced a «doorstop» rock back to Banaba, a then «unclaimed» landmass in the central Pacific (Silverman, 1977). At that time about 450 people lived in the island (Sigrah and King, 2006). Ellis arrived in Banaba in May 1900, and immediately found high-grade phosphate among the foreshore rocks. Recalling Banaban oral history, Corrie said that Ellis quickly instructed the locals to collect all the rocks they could find, and then he purchased each heap at 4 shillings a time, before loading them onto his boat, the «Archer» (Corrie, interview, 2^nd February 2011, Suva, Fiji).

16The «Archer» returned again to Banaba just three weeks later on the return leg of her voyage (Shennan and Corrie Tekenimatang, 2005).

17In 1901, to ensure continued exclusive access, the British government annexed the island, and raised the British flag (Sigrah and King, 2001). Banaba Island became part of the British Gilbert and Ellice Island Group. Mining commenced later that year under a more formal, but equally unfair arrangement, of 999 years mining rights for £50 per annum (Macdonald, 2001). The operation grew quickly, and in 1908 bpc was formed.

19The bpc continued to operate a monopoly in Banaba for the next 30 years, until mining operations were brought to a halt by the invasion of Japanese forces in 1942 (Silverman, 1977).

The decision to relocate the Banabans to Rabi Island

20Parallels can be drawn between the mining history of Banaban and another Pacific island, the nation of Nauru. During the xx^th Century, phosphate mining brought the people of Nauru considerable wealth, yet caused the destruction of 80% of the island (Gowdy and McDaniel, 1999; Connell, 2006). Despite the physical devastation of Nauru on a scale similar to that of Banaba, the people of Nauru continued to live in their home island.

21Banaban emigration, however, had been a consideration of the governing officials long before the outbreak of wwii (Silverman, 1962). The Banabans always forcefully rejected such suggestions.

23At that time, the Lever Brothers’ Pacific Plantations Proprietary Limited (pppl) was selling Rabi Island in northern Fiji for £25,000 (Vaskess, 1942). Only a handful of people were already living in the island: an old man from the Solomon Islands, who had been there for many years, and a few copra plantation workers (Benaia 1991). The British authorities, therefore, thinking it an appropriate alternative, suggested the island to the Banabans. The community wanted representatives to visit Rabi to assess its suitability for themselves. However, before the selected members could leave for Fiji, Japanese troops invaded Banaba, and shipping communications with the rest of the colony ceased.

«The object of the purchase is to provide an island for the settlement of the natives of Ocean Island against the time when the phosphate deposits in that island will have been marked out and the island will, in consequence, have become largely uninhabitable.» (Vaskess, 1942)

25The British authorities had acquired Rabi Island without contributing financially to its purchase and also ensured that Banaba Island would be uninhabited when mining recommenced post wwii.

26Meanwhile, the islanders suffered greatly under the Japanese occupation of Banaba. Life was harsh and often brutal. There were severe food shortages, and the community would have probably starved to death had they remained in the island (Vaskess, 1945). Instead the Japanese decided to disperse most of the community to three separate internment camps in Nauru, Tarawa and Kosrae (Holland, 1948; Benaia, 1991).

27A third of the population died during the three-year occupation, some from starvation, others were poisoned, shot, be-headed or died from electrocution. People were killed for relatively minor offences (Maude and Maude, 1994). Their war-time experiences probably left the Banabans questioning themselves, as well as others around them (Colson, 2003).

Table 1. – Profile of those relocated to Rabi Island in 1945 by gender, age and ethnicity

	Men	Women	Children	Total
Banabans	185	200	318	703
Gilbertese	152	97	51	300
Total	337	297	369	1 003

(Source: Silverman, 1971)

28At the end of the conflict only «able» Banabans had survived and they were weakened from their experience, and certainly in no shape to take up the fight against the bpc (MacDonald, 2001). They were gathered by the British government in Tarawa and told that they could not return to Banaba (Kempf, 2003). The authorities announced, untruthfully, that their island home had suffered during the War (Secretary for Fiji Affairs, 1945). The real reason for the relocation was stated in earlier correspondence:

«Many years ago it was realized within a comparatively short time the phosphate deposits in Ocean Island would be worked out, and, as all that would then be left of the island would be a forest of coral pinnacles, the Banaban (the natives of the island) would have to leave and find a home elsewhere. When this was suggested to them, however, they flatly refused to consider it and for some 25 years they stubbornly resisted the idea.» (Vaskess, 1944)

29A year later, Vaskess (1945) made it clear that there was no alternative location for the Banabans in the Gilbert and Ellice Island Group. The assembled survivors had no choice other than to move the 3,200km to northern Fiji. Having already suffered greatly during the preceding years, the Banabans agreed to go to Rabi – an island that they had never seen – on a trial basis (Silverman, 1962).

The resettlement action plan and the arrival of the Banabans

30Two months prior to the arrival of the Banabans in Rabi, a preliminary resettlement plan was circulated by the administrative officer, outlining various socio-economic aspects of the relocation (Kennedy, 1945a). Considerations included the living conditions, housing, camp equipment, cooking, medical care, administration and organisation and status of the Banabans in Rabi and their relation and status within the wider Fiji group. For the first six months, the movement of the Banabans was to be restricted to Rabi only; subsequently, they would be «as free as any other Fijian citizen to roam from island to island» (Cooper, undated).

31The authorities were anticipating 700 arrivals in Rabi, and provisions were calculated accordingly. The awaiting supplies proved inadequate, however, as:

«about 700 people were expected, but latest advices mentioned 900 and they landed 1,003.» (Verrier, 1946)

32At 6.30pm on 15^th December 1945, the 1,003 passengers disembarked the «Triona», a bpc-company ship (Kennedy, 1945b). Two ethnic groups, the Banabans and the Gilbertese, were represented; the Gilbertese were linked to the community either through marriage to individual Banabans, or having previously developed close friendships with particular Banaban families (Benaia, 1991).

33Children made up almost half the Banaban contingency, illustrating just how much the adult Banaban population had suffered during the War (table 1).

34On arrival the Banabans found Rabi Island to be very underdeveloped, especially when compared with the structure of equipment and services in Banaba (Silverman, 1971). According to Hedstrom, the roads in Rabi were merely access tracks to copra plantations, there was no electricity and just a few substantial buildings existed (Hedstrom, Interview, Principal, Rabi High School, Rabi Island, 14^th April 2011). When the Banabans disembarked, Corrie recounted that only tents, household supplies (mosquito nets, blankets and utensils) and basic food rations were waiting for them (Corrie, interview, 2^nd February 2011, Suva, Fiji), a recollection that tallies well with the resettlement action plan equipment roster (Kennedy, 1945a).

Initial stage of resettlement

35The initial camp of tarpaulin tents was on a flat, open field at Nuku, Tabwewa; this location was to be the community’s temporary home for the first few months, as they had been requested by the authorities to remain together at one location for security purposes. Food supplies at the camp were distributed once a week rationed on an individual basis. Children received half the adult allowance (Benaia, 1991).

36The British authorities on arrival made it clear that at the end of the two years, the Banabans would be free to decide either to make Rabi their permanent home or to return to Banaba if, as a community, they so wished (Windrum, 1946).

Transition stage of resettlement

37The Banaban Settlement Act, passed by the Fiji colonial government at the end of 1945, set up the new administrative structure for the Banabans in Rabi (Hindmarsh, 2002), and, after agreement at a meeting of heads of families (Silverman, 1971), an island council was formed as the instrument of local government. It was they who made the key decision to disperse the community from their temporary campsite to the four villages in the island six months after they had first arrived (Silverman, 1977).

39In Rabi, the resettlers were encouraged to move to the village in Rabi that was their respective village in Banaba. Many people, Corrie said, wished to stay in the original settlement of Tabwewa, where they had first been located on their arrival (Corrie, interview, 28^th February 2011, Suva, Fiji).

40In 1946, during what many Banabans believed was their temporary two-year stay in Rabi, Corrie said that representatives of the community returned to Banaba, to assess, as they thought, the rebuilding process (Corrie, interview, 2^nd February 2011, Suva, Fiji). The real purpose for the return of the 149 landowners was for them to register their land boundaries in Banaba in readiness for a pre-mining land survey (Anon, 1946). The group returned to their homeland to find that all buildings had been razed to the ground, no homes remained, and their lands had been cleared in readiness for continued mining activities. Nothing was recognisable, and Corrie noted that it was then that the Banabans realised that it was not the Japanese who had destroyed their home island, it was the undertakings of the bpc (Corrie, interview, 2^nd February 2011, Suva, Fiji).

41Mining recommenced in 1947, the year that the Banaban community elected, in a secret ballot, to remain in Rabi (British Information Services, 1979). All was not smooth, however. The Banabans had swapped the status of landholders, workers and small-scale market producers in Banaba for copra cutters and gardeners in Rabi (Silverman, 1971). A couple of years into their resettlement, the resident governmental adviser called for police protection. In the 1960s the bitterness escalated, and the Banabans elected to dismiss their adviser (Silverman, 1977). It was at this time that the Banabans intensified their claims for financial compensation and wished to further highlight their plight within the international community.

Compensation claim

42The arrival of the phosphate company had introduced the link between land and money to the Banabans (Silverman, 1971), and the community had been in receipt of monies for phosphate extraction for many years. They thought that the monies in the Phosphate Provident Fund were theirs, for their own use; the authorities thought otherwise. Rabi Island was purchased with the monies from the Fund, and the remaining balance was for:

« […] such work and amenities as the sub-division of lands, lay-out of villages, and construction of public utilities such as water supplies, village meeting halls, cooperative store building, wharf, boats, boat-sheds, etc., etc., which may be regarded as part of the settling-in obligations of the Provident Fund.» (Vaskess, 1944)

43With the post-war resumption of mining operations, the newly-relocated community continued to receive royalty payments in the form of bonuses and annuities (Silverman, 1977). The Banabans still felt aggrieved. They were frustrated with constant disputes over land leases on Banaba, and the inadequacy of monies received from the bpc (King and Sigrah, 2004). In the 1960s, disheartened, yet inspired by the political independence, control of phosphate and wealth of neighbouring Nauru (MacDonald, 2001), the resentment towards the British authorities came to a head when, in 1965, the community started legal procedures for compensation in the British courts (Sigrah and King, 2001). The 221-day case was the longest in British legal history (MacDonald, 2001; King and Sigrah, 2004), and the final ruling found the British government guilty of moral negligence, but, in the strict legal sense, not liable for the injustices committed in their name (Binder, 1977).

44A year later, after public and political pressure, the British government offered to set up a trust fund to produce a pension for the Banaban community (Sigrah and King, 2001). In May 1977, the secretary of state for foreign and commonwealth affairs, on behalf of the bpc partner governments, offered the Banabans an ex-gratia payment of Australian $10 million (MacDonald, 2001) – capital taken from the revenues of bpc – on the condition they dropped all further legal action (King and Sigrah, 2004). Four years later, the Banaban finally accepted the Australian $10 million package, plus interest.

«On 13 April 1981, the Rabi council of leaders signed appropriate undertakings bringing these matters to an end. At the same time the British, Australian and New Zealand governments made available, the an ex-gratia basis, 10 million Australian dollars and a sum of a little over 4½ million Australian dollars in lieu of interest thereon for the benefit of the Banaban community as a whole.» (HC Deb, 1981)

45Despite the size of the award, which was inadequate compared with the amount given to their phosphate neighbours in Nauru, the Banabans were only allowed access to the interest from the compensation payment (Hindmarsh, 2002).

From mining narrative to climate-change narrative: a tentative/prospective comparison

46This section of the paper compares the relocation of the Banaban with the contemporary issue of climate change-forced displacement (ccfd). First, development-related relocations are briefly discussed, before climate change-related relocation is examined in terms of definitions, drivers, adaptation strategies and case studies. Finally, a comparison of the relocation of the Banabans with ccfd will be made.

47The relocation of an entire island community for economic gain is rare. Many past forced relocations of communities have been the result of development projects, such as the construction of dams and irrigation schemes (Cernea, 1999; Fujikura et al., 2009), and outcomes have been mixed. Insufficient planning, preparation and implementation result in more failures than successes (Oliver-Smith and de Sherbinin, 2011); experiences from the past suggest that climate-induced displacements risk exposing the people affected to severe economic, social and environmental hardship (Cernea, 1997); communities will be placed in a new context of vulnerability and risks (Birk, 2012).

Definitions, drivers, adaptation strategies and case studies

48Climate change is expected to increase the incidence of community relocations significantly in the future, and much attention has been given to the recent emergence of the issue among policy-makers, academia and the media (Garnaut, 2009; McAdam, 2012; Park, 2011; Warner et al., 2013).

«Relocation of communities will become one of the few practical options (if not the only one) for adaptation to climate change by communities and/or villages.» (Solomon Islands Government, 2008: 86)

49Estimates of the number of people to be affected vary (Lonergan, 2012); mass displacement of communities, however, is likely to result from the combination of sudden-onset, extreme weather events and slow-onset environmental change that will render locations uninhabitable (Bronen, 2014).

50«Sinking» small-island states constitute a particular challenge (Kälin, 2012). In the future peoples of low-lying nations, such as Kiribti, Tuvalu, and the Marshall Islands, may be rendered stateless, if present sea-level rise forecasts prove correct, and their island nations become totally submerged under rising seas (Stern, 2006). To date, international law has failed to adequately address this international dimension (Jarvis, 2010; McAdam, 2011).

51Indeed, there is no accepted definition of a person who moves because of climate change. Renaud et al (2007) identifies three categories of people: «environmental refugees»; «environmentally-forced migrants» and «environmentally-motivated migrants», based on whether or not a «decision» to move is made. Ferris (2012) alternatively offers a categorization of people based on the different environmental circumstances behind the relocation: sudden onset natural disasters, intensified by the effects of climate change; slow onset effects of climate change, creating unviable livelihoods; and, the destruction of part or all of a country from the effects of climate change.

52Planned relocation is an adaptation strategy already being implemented in some states, and movements to date have mainly taken the form of in-country resettlements (Mimura et al., 2007). The Fiji government, for instance, has recognized a national need and prepared planned-relocation guidelines to systemize the relocation process within the nation (Fiji Government, 2014; Wilson, 2014).

53There is still insufficient understanding of the effects of climate changes on relocation, however. Media and environmental groups claim the handful of current community relocations are entirely climate-driven (Vidal, 2005; Morton, 2009). Such uncritical promotion of a direct causal link between climate change and displacement is misplaced, and potentially, alarmist as such cases are scientifically unproven. Accompanying technical assessments of individual locales are seldom performed, and caution needs to be used when discussing the nature of relocation causality - relocation is a highly complex issue (Campbell, 2010) as indeed is the environment (Lonergan, 2012). Isolating the effects of climate change from other drivers of migration and environmental change is difficult (Dun and Gemenne, 2008; Care et al., 2009). Often climate change may merely exacerbate an already perilous economic or social situation (Renaud et al., 2011), and removing environmental processes from the social, economic, political and institutional structures will prove problematic - the environment is a «contextual factor» (Lonergan, 2012). Even when climate-change causality is confirmed, climate-induced migration must not be viewed in isolation from other forms of displacement, rather it needs to be considered in the context of an existing migration system (Hugo, 2010).

54Environmental degradation is a contributory factor in the current «climate-related» relocations used for comparative purposes in this paper (summarized below), but the environment may not necessarily be the primary cause of the relocation. The implications for the community are the same, however – people have to move because their settlement is no longer sustainable:

55Tegue Island, northern Vanuatu: Coastal community in high volcanic island, relocating because of sea-level rise, flooding and accelerated coastal erosion (Nakalevu and Phillips undated; unep, 2005). The area is also seismically and volcanically-active, and prone to earthquakes and tsunamis (Warrick, 2011);

56The Carteret Islands, Papua New Guinea: Low-lying atoll community, relocating inter-island because of coastal inundation and loss of land (Boege, 2011; Bronen, 2014; Edwards, 2013b). Effects in part may also be linked to geological instabilities (Barnett and Campbell, 2010; Weir and Virani, 2011);

57Vunidogoloa village, Vanua Levu, Fiji: Coastal community, relocating to higher ground because of coastal inundation (Wilson, 2014). Villages say conditions are exacerbated by the juxtaposition of the village between Natewa Bay and a local (unnamed) river.

58Narikoso village, Ono Island, Kadavu, Fiji: Coastal community, relocating to higher ground because of coastal inundation (Fiji government, 2014). Villages say conditions are exacerbated by historical decision to block a sea channel between an islet and the village site.

59Parallels between the relocation of the Banabans and climate change-forced displacement ccfd are shown in Table 2. The Banaban case is well documented and learning opportunities are presented from the Banabans’ experiences that may aid future ccfd in the Pacific and elsewhere. Like the Banaban relocation, the effects of climate change can be both narrative and material (Farbotko and Lazrus, 2012), and current-day climate-related relocations will be drawn upon in the discussion.

Populations affected and nature of the displacement

60Island communities have relocated for many reasons in the past: trade, marriage, warfare, families and cultural exchange (Hau’Ofa, 1994; Campbell et al., 2007). The Banabans are in the unenviable position of being an entire island population that has been forcibly relocated from their island home in modern times by outside agents. Low-lying islands are similarly threatened – by submergence from rising sea levels (Mimura et al., 2007) – and island communities across the region face uncertain futures in locations away from their current homes (Stern, 2006). Just as the Banabans were presented with no alternative to relocation, climate change-displaced communities will have little choice, ultimately, other than to leave their homes on a permanent basis.

61Most ccfd relocations will occur within countries, and cross-border relocations will not be the norm (Mimura et al., 2007). A deterrent to any relocation, however, is a lack of land availability.

Table 2. – Comparison between the relocation of the Banabans and climate-change forced displacement (ccfd)

	Banaba Island	Low-lying coastal areas and islands
Population affected	Entire island population	Entire island populations and communities occupying low-lying coastal areas
Nature of displacement	Permanent and international	Likely to be permanent. Mostly internal (within national borders) or may be international.
Direct cause of relocation	Extraction of phosphate from mining activity by foreign private-sector company, leading to extreme degradation and loss of land.	Sea-level rise from human-induced climate change, leading to salination of ground-water lens, reduced agricultural productivity and loss of land.
Indirect cause of relocation	Economic development by industrial nations	Economic development by industrial nations
Geography	Raised island (300m above sea-level at highest point)	Low-lying coral atoll (usually <5m at highest point); low-lying coastal communities in high volcanic islands.
Time frame	Slow-onset event (phosphate extraction)	Slow on-set event (sea-level rise). Long-lead time
Pre-relocation planning	Colonial authorities prepared relocation plan. No community involvement, and Rabi Island purchased without any visit by community representatives to assess its suitability	Authorities will need to be proactive and prepare relocation plans. Community need to be consulted and actively participant in the formation of any plan and in site selection.
Funding	Resettlement costs initially self-funded from phosphate-mining royalties. Thirty years later, compensation from British government following legal action by the community	Resettlement costs initially self-funded or from local / national government. Ability to seek compensation from industrial nations still to be determined
Resettlement location	Alternative Pacific island, with no resident, host community	Unlikely to find an alternative ‘vacant’ Pacific island location
Loss of nationality	Loss of independence (Banaba Island subsumed under Kiribati)	Loss of independence (in worst-case scenario, land territory for some nations may be totally submerged)
Loss of homeland	Population able to visit ‘homeland’ of Banaba Island	Population unable to visit ‘homeland’ in worse-case scenario. Submergence of ancestral land
Post-relocation assistance	No long-term support for livelihood reconstruction	Need for on-going support for livelihood reconstruction

(Source: Author)

«Relocation is problematic when they [communities] do not own land resources on nearby islands thus land tenure and land management systems prohibit any discussion let alone relocate to nearby islands.» (Solomon Islands, 2008: 86)

62During the colonial era, the relocation of the Banabans from Banaba Island to Rabi Island involved the internal movement of a population from one part of a British-ruled colony to another. Opportunities for internal resettlement between colonies, ruled by one common colonial authority, no longer exist (Tabucanon, 2012), and any such move would be reliant on the bilateral or multilateral agreements between sending and receiving countries – the move today would have an international dimension, from Kiribati to Fiji. These two nations have recently completed an agreement that has enabled the Kiribati government to purchase 2,210 hectares of land in Fiji to enhance its food security (Office of the President, Republic of Kiribati, 2014). The bilateral negotiations, that did not include any agreement to resettle people, took more than two years to complete, illustrating the need to assign sufficient time to any future relocation proposal.

Direct and indirect causes of the relocation

63Neither the Banabans previously nor those currently facing ccfd will have contributed greatly to the situation that they face. Fertiliser, the end product of phosphate extraction, was of limited use to the Banabans who, with little soil, grew few crops, and today, small island-developing states account for less than 1% of global greenhouse gas emissions (Julca and Paddison, 2010). The contribution to global climate change of any one low-lying island or at-risk community will be negligible. In both the Banaban case and with ccfd, the root cause of the relocation is economic development elsewhere by people unaware (or uncaring) of the effects of their behaviour on others.

Geography and the importance of land

64People and land in the Pacific have a very special, intuitive relationship (Campbell, 2010). Land to the Banabans was of such immense importance that Silverman (1971) proposed a «blood and mud hypothesis», where people (sharing a common identity) and locality (land rights, residence and sustenance) appear interwoven in the definition of Banaban kinship. Land in Banaba was not merely the ground on which people lived, but a right and medium of exchange between individuals (Teaiwa, 2005). Undoubtedly, the island of Banaba was one of their most powerful cultural symbols (Silverman, 1977; Teaiwa, 2005).

65Mining destroyed many of the Banabans’ sacred water caves (bangabangas) (Sigrah and King, 2001) and removed more than 20 million tons of topsoil (Williams and McDonald, 1985). The island changed beyond recognition. Low-lying islands and coastal areas will similarly lose land under rising seas, yet long before submergence, islands and shorelines will contract from encroaching waves, salt water will contaminate fresh-water sources and populations will be forced to seek relocation elsewhere.

66The slow-onset contamination process has already started. Some low-lying populations complain that ground-water supplies have a salty taste (Locke, 2009). Community response either involves seeking alternative water sources (Weir and Virani, 2011) or, in some atolls such as in Kiribati, it is reported, they mix coconut toddy (the sweet sap from the coconut flower) with fresh water in an attempt to disguise the saline taste (Anterea, interview, 4^th October 2011, Tarawa, Kiribati). When ground-water sources become unreliable and fresh rainwater scarce, communities will be forced to consider relocation. Paradoxically, Banaba Island, the only raised island in the Kiribati group, offers i-Kiribati (the people of Kiribati) little escape from the rising seas, as Banaba records a long history of sustained periods of drought and a scarcity of fresh water (Binder, 1977). Recurring droughts were probably responsible for many people leaving Banaba in the past, that resulted in periodic population declines (Silverman, 1971).

Importance of homeland, identity and a sense of belonging

67The Banabans own two islands, Banaba and Rabi. Resettled on Rabi, they have dual identity: they are Banabans from Banaba, and they express that they are Rabians from Rabi (Kempf and Hermann, 2005).

68Many Banabans articulate a desire to visit Banaba Island during their lifetime. Such a return, though difficult to arrange, would not be impossible for individuals to make (submerged islands leave no options of return). Banabans realise, though, that a collective return to their ancestral homeland is highly unlikely. Reminders of Banaba Island live on in the village names in Rabi, however, and for communal activities, such as intravillage competitions or food production, the villages divide into geographical sections that replicate the old Banaban model (Silverman, 1971). The recently-moved Tegua community, Vanuatu, however, chose to take on the name of their new site, Lirak as their community name; in their tradition names of places or living things that sustain them are used to name settlements or children (Nakalevu and Phillips, undated). Such a focus on the present may assist the community to identify with their new location in the short term.

69In most cases of relocation, however, the pull of the ancestral home remains strong. According to Corrie, a few hundred people have returned to Banaba to act as a retaining population on the island (Corrie interview, 2^nd February 2011, Suva, Fiji), though it is believed the island could not sustain a greater population because of its very limited water supply (McAdam, 2012). The well-publicised relocation of the Carteret Islands community, Papua New Guinea also projects that half the population will remain in the islands by 2020 (Lokani, 2011), and it seems reasonable to assume that others affected by ccfd will make similar requests to retain population in the homeland. By moving some of the people exposed to climate risk away from the affected area, the capacity of the remaining population to adapt to climate change may be enhanced (Hugo, 2010).

70Not only were the Banabans forcibly removed from their island home, but a sizeable portion (20 million tons) of the physical land of Banaba was also shipped elsewhere in the form of phosphate to Australia and New Zealand (Teaiwa, 2005). Teaiwa, therefore questioned, should the Banaban equally yearn after some agricultural field in Australia or New Zealand, fertilized with Banaban topsoil, as they do their island home? (Teaiwa, 2005).

71The disappearance of «sinking» atolls may leave no trace of homeland and may make custody impossible in the longer term; the psychological, social and economic effects of permanently losing one’s homeland are yet to be determined for displaced-atoll islanders.

Culture, identity and integration

72Relocation may result in the potential loss of language and culture (Kelman and West, 2009). The Banabans had a distinctive, rich culture in Banaba, expressed in rituals, dance, marriages, adoptions and the taming of frigate birds. Phosphate mining already threatened the traditions of the Banabans, but further erosion occurred with the relocation of the community to Rabi. On the larger island of Rabi, islanders said that they were no longer able to call out to each other from village to village as they had done on Banaba - distances were too great - and without a coastal road, horseback and sea transport were the only means of communication between settlements (Edwards, 2013a). In contrast, the two current village relocations in Fiji are each moving to higher ground within the original village boundary (Ravula, 2012; Silaitoga, 2012). The communities will remain intact, therefore, and their identity and history maintained. When communities do not have land suitable to host an entire community in one location, they are likely to be dispersed across several locations. In the case of the relocation of the Carteret islanders five different sites on the larger, main island of Bougainville, Papua New Guinea have been identified for resettlement - each given by the Roman Catholic Church - but none are adjacent to one another (Edwards, 2013b).

73While distances between resettlement areas may cause dilution, conversely, isolation may strengthen culture. The Banabans were fortunate to be able to purchase a «vacant» island. The physical geography of an isolated landmass can help retain cultural identity as interactions with neighbouring (host) communities may be limited. Entire islands rarely come to market, however, and opportunities to buy an island, and even freehold land, will be limited in the future. Communities facing relocation may have to accept integration into an already-occupied area with all the associated cultural, social and political issues (Bronen, 2014). Islanders at-risk of relocation may be reliant, therefore, on the good will of others in finding them a new home.

74Much of the culture of the Banabans exists only in modern-day text books, and these are written in English, a language, Corrie notes, that is foreign to the majority of the people they describe (Corrie, interview, 2^nd February 2011, Suva, Fiji). The Banabans say that their day-to-day lives have lost the richness of their unique culture.

75Cultural displays, such as dancing, are now the preserve of marriages and other special occasions (Kempf, 2003), such as the newly-introduced Banaba Day, on the 15^th December each year. The Banabans have feared their eventual assimilation into the Fijian cultural mainstream (Sigrah and King, 2004: 1046), and try to use the performing arts (singing, dancing and storytelling) to spatially anchor, preserve and communicate their history, traditions and culture to others (Kempf, 2003).

Post-relocation support

76The Banabans had little knowledge of farming or of reef fishing, and upon arrival, only had two months of food rations. They also assumed that their stay would be temporary. When relocation is permanent, more attention can be paid to both the conditions and the process by which the resettlement will take place (Ferris, 2011). The villagers of Tegue, Vanuatu were still building houses a year after the new site had been officially opened, with assistance from the wider community and others from neighbouring islands (Nakalevu and Phillips, undated), though with less time spent on maintenance post-relocation, as their new, dryer homes did not rot, the community will have more time to focus on other development issues (Nakalevu and Phillips, undated). Livelihoods need to be restored and communities reconstructed for each future ccfd.

77The Banabans did have the advantage of arriving as one cohort (atoll islanders are unlikely to experience a similar movement of people en-masse), but they arrived in an island without basic infrastructure and no immediate opportunity for formal employment. Little attention was given by the authorities to the group’s mental and socio-economic well-being. Of the first five Carteret Islands families to move to the new relocation site in Bougainville, three returned home within a few months unable to settle into the new environment (Edwards, 2013b).

Funding and compensation

78Future climate injustice will be as much an issue for people affected by ccfd as development-induced injustice was for the Banabans. The Banabans funded their relocation with royalties from their own phosphate fund, and only sought compensation some twenty years later (Sigrah and King, 2001). Most atoll islanders, finding themselves threatened with permanent displacement, will have no means by which to fund their relocation. In Solomon Islands, the napa states:

«Relocation of communities and/or villages will necessarily become the responsibility of the governments at all levels (i.e. community/local, provincial and national).» (Solomon Islands, 2008: 86).

79To date, the Fiji government has financed 75% of relocation projects’ costs, with the community expected to fund the remainder (Pareti, 2013). In the future, however, it is questionable whether with the projected increased incidents of ccfd that national governments will be able to continue to fund similar projects, and monies will need to be sought from elsewhere (Pareti, 2013).

Conclusions

80Human mobility in the Pacific characterizes the region’s past and present and probably will its future. Ever-evolving culture, too, is not immune to movement and mergers (Hau’Ofa, 1994). What makes the Banabans unusual as a displaced people is that they were already displaced from their homeland by the War, prior to their move to Rabi, and they had attained wealth and comfort from the phosphate-mining activity beyond that which they could have generated themselves. Many Pacific journeys of migration are prompted by trade, exchange and opportunity – none of which applies to the Banaban case. The new environment in Rabi was to be very different in every way from their experiences in Banaba and under Japanese occupation.

81The colonial authorities had devised a resettlement action plan, and attempts were made at settling the Banabans into their new home, but the timescale of a few months proved unrealistic. The Banaban people would need self-reliance to make Rabi a success.

82Forced resettlement is now part of the historical narrative of the people of Banaba, one community case study within the wider narrative of mining exploitation in the Pacific; forced resettlement and mining exploitation are one part of the broader history of mobility in the Pacific-Island context, a mobility that includes climate change.

83The re-establishment of a settled community is a long process and can take many years. Lessons to take forward for future ccfd are numerous. The most pressing issues are the need for the involvement of the community at the initial discussions stage, including in the assessment of an alternative home, and for long-term, post-relocation planning, which includes the creation of sustainable livelihoods. Addressing these issues will not be possible without on-going support by external agencies for those displaced, and that support needs to start now.