Introduction

Small-scale fisheries are among the most relied-upon activities in coastal areas globally, yet they are also highly vulnerable to environmental changes induced by climate change (Allison et al. 2009; Muhala et al. 2021; Ruiz-Díaz et al. 2019; World Bank 2013). We recognise the contrary conceptualisation of disasters in the social science literature, where it is argued that the common framing as ‘natural’ fails to capture the deep social impacts of such disruptions, which in fact are central to how disasters are experienced, reported on, and responded to (Calandra 2019). Though this paper focuses on the impact of disasters on people, it does so through analysing the coastal resources available to individuals and communities who depend on them. In addition, ‘natural disaster’ is common vernacular in the region and so we refer to ‘natural disaster’ in this study.

Within this context, the people of Vanuatu (Ni-Vanuatu) are ranked amongst the most vulnerable populations in the world (Aleksandrova et al. 2021). Ni-Vanuatu society has been shaped in many ways by the occurrence of natural disasters (Movono et al. 2021; Ruehr 2022), as reflected in, for example, material culture as well as social practices. Traditional houses in Vanuatu are generally built low to the ground, and with open windows on all sides allowing wind to pass through the house, thereby enhancing structural integrity and reducing the risk of damage during strong winds. Additionally, building materials for traditional housing includes wood and palm-leaf thatch; all easily sourced in and around communities. Post disaster rebuilds thus start as soon as a cyclone has passed, as opposed to waiting for materials from another island or overseas to start the recovery process (Humanitarian Advisory Group and VANGO, 2020; Peters and Bahadur 2014). The different ways land and sea are used after a disaster are also a good example of how social practices in Vanuatu adjust to the continuous occurrence of disasters. Landowners commonly lend land to landless families temporarily when disruptions occur to ensure food security, as observed on the island of Espiritu Santo (Sanma province) during COVID-19 (Movono et al. 2021). Such customary practice is generally arranged through marriage, kin or other social ties. Similarly, tabu areas are often also managed differently during and shortly after a disaster. Community requests to their Chiefs for short (partial) openings of a tabu area are common, as noted in Aneityum (Tafea province). This traditional type of management aligns with customary practices that ensure community’s access to food during challenging times (Neihapi et al. 2023).

Prior to state-building influences from European administrations, the islands of Vanuatu were governed by customary systems (kastom) over land and sea. Today, kastom is recognised by the government as part of the national constitution. It functions in conjunction to newer forms of governance established through both French and British influence, as well as religious groups. Kastom in Vanuatu forms the foundation in communities’ tenancy over land and coastal waters, while also enabling local empowerment in management of their land’s resources. Ni-Vanuatu have a deeply instilled culture of community engagement that is enabled through social networks like wantok systems. This translates in strong cohesion in supporting each other when disaster occur (Eriksson et al. 2017; Johannes 2002; Steenbergen et al. 2020).

We are in a period of unprecedented philanthropic support for conservation and humanitarian aid in the Pacific (Enrici et al. 2023), and the opportunities for engagement for external aid are increasing over time. When a natural disaster occurs, much of the attention focuses on channelling and coordinating international humanitarian aid to fill gaps in capacity, resource national responses, and support national coordination efforts in their transitions from the emergency relief to longer term recovery. Such systems work in a way that tend to fall short of accounting for and building on existing local support systems (i.e., the ways people in rural communities support each other), rendering humanitarian aid less effective. This is particularly the case in the immediate aftermath of disruption, when critical consideration should be given to how central local support systems are for isolated pockets of communities during disaster recovery. Scholars have questioned how this influx of external and/or private funding may affect domestic agendas (Enrici et al. 2023). Though well-intentioned, such efforts have been noted to be less effective because they do not always correspond to what people actually need on the ground and often involve lengthy (bureaucratic) processes. With all the challenges of a post-disaster situation, emergency relief aid often reaches communities days, if not weeks, after the disaster (Humanitarian Advisory Group & VANGO, 2020).

What is still largely ignored in the flurry of humanitarian aid is the dynamics of local support and social safety nets that communities utilise. The innate systems that people have at their disposal to endure these disruptions - rooted in culture and practices - and the evolution of social networks in times of disaster have previously been described in the Pacific and elsewhere (McNamara et al. 2021; Misra et al. 2017; Movono et al. 2021; Steenbergen et al. 2020). The importance of social capital within and between communities, and across larger networks that include the state, private sector, and civil society organisations, are critical (Badjeck et al. 2013; Marín, 2019). For example, Marín (2019)’s analysis of local responses to the 2010 earthquake and tsunami that hit the central-southern coast of Chile, showcased the importance of cross-scale social networks for the recovery of livelihoods among fishing communities as well as their adaptations to new conditions. Yet, strong social networks alone are not enough to ensure adaptability and recovery of fishing communities post-disaster. Other elements such as livelihood flexibility and local ecological knowledge are also key factors determining communities’s ability to face challenges before and after disasters (Marín, 2019). In Vanuatu, the importance of social capital is increasingly recognised, particularly during the Covid-19 pandemic when tropical cyclone Harold hit in 2020 and international borders were closed, and the support and resourcing of national recoveries could not be undertaken (Steenbergen et al. 2020). Previous research has described the different forms of local adaptive capacity that Ni-Vanuatu rural communities draw from, to support and maintain livelihood and food security needs (Steenbergen et al. 2020). Of particular note is the importance of easy access and availability of aquatic foods during and after natural disasters when other food sources (e.g., gardens and imports) fall away. In such cases, aquatic foods often become the principal nutritional food source. Reef fishing and gleaning typically become important, as these activities are readily carried out and do not depend on the use of fishing boats, or canoes (which are often destroyed) (Eriksson et al. 2017).

The differential impacts of disasters across social groups is an emerging focus of research. Though many knowledge gaps remain as to what role(s) women have in disaster resilience, there is evidence that community members adopt different gender roles after a natural disaster. The literature shows that women globally, and in PICTs, are often under-supported by the system, by virtue of their social position, having limited influence over decision making or planning, and restricted access to natural and financial resources (Alam and Rahman 2017; Charan et al. 2016; Misra et al. 2017; Singh et al. 2022). Women have been shown to adopt various roles during natural disasters that enhance community resilience (Alam and Rahman 2017; Charan et al. 2016; Singh et al. 2022). It is increasingly recognised in the literature that disaster risk management should take into account the ways in which gender dynamics influence disaster impacts in any given area before undertaking policy decision-making and design of any project (Wabnitz et al. 2021; World Bank 2021).

Overall, this study seeks to contribute to discussions around improving disaster response in the coastal fisheries space. We highlight local autonomous response capacity, through examples of how people in effected areas respond around a disaster (immediately before, during and after). In that, we examine the external-aid received by communities after the 2023 twin cyclones of Kevin and Judy, the self-organised community-to-community support, and some of the roles adopted by women and the challenges they faced in the aftermath of the twin cyclones. We draw from our findings to suggest (in the current state) where external-aid interventions are most impactful, so as to improve people’s ability to return to normality post disaster.

Vanuatu’s national response system

While there is a recognised need for an effective system that strengthens local community-based responses while facilitating the integration of external aid injections (Humanitarian Advisory Group & VANGO, 2020), the Vanuatu Government, with support from international development partners, has over time developed the institutional machinery to facilitate national-led response system. Depending on the severity of a natural disaster, the Government of Vanuatu may declare a national state of emergency. This activates a cluster-system to marshal a response to the emergency. Under the Ministry of Climate Change sits the National Disaster Committee (NDC) and the National Disaster Management Office (NDMO). The NDMO coordinates the National Emergency Operation Centre which, among other things, is responsible for the activation and coordination of eight different emergency clusters, including: i) Education Cluster; ii) Emergency Telecommunications Cluster; iii) Food Security & Agriculture Cluster (FSAC); iv) Gender & Protection Cluster; v) Health & Nutrition Cluster; vi) Logistics Cluster; vii) Shelter Cluster and viii) Water Sanitation & Hygiene Cluster (OCHA 2023). This system aims to enhance the communication and response between humanitarian and government agencies during natural disasters. The FSAC forms the primary point of engagement for the various departments under the Ministry of Agriculture, Livestock, Forestry, Fisheries and Biosecurity (MALFFB) tasked to ensure primary food production is restored. This is where the Vanuatu Fisheries Department (VFD) engages in the context of coastal fisheries.

Official procedure following a natural disaster first sees priority areas identified based on immediate assessments. This is followed by a “first push”, which involves the distribution of vital supplies to all priority areas, including dry food rations, first aid and drinking water. On completion of the first push, focus shifts towards the “recovery phase”. This phase first involves an assessment of losses and damages in areas hit by the natural disaster, which then informs strategic implementation of recovery efforts.

Methods

Study sites and background

On the 1st and the 3rd of March 2023, Vanuatu was hit by two consecutive category-4 cyclonesFootnote 1, twin cyclones’ Judy and Kevin. Over 250,000 people were affected by the two TCs (i.e., 80% of the country’s population), leaving in its trail significant structural damage (Fig. 1). Both cyclones tracked southwards along the archipelago, delivering greatest destruction to the provinces of Shefa (not including Epi island) and Tafea. These areas of high impact were subsequently designated as Priority Areas 1 and 2 for the post disaster emergency response (Fig. 2).

Fig. 1
figure 1

Damage to road and housing infrastrutcure in Lenakel (Tanna island) after the TCs hit Vanuatu in March 2023

Full size image
Fig. 2
figure 2

Map of Vanuatu showing the paths of TC Kevin and TC Judy, as well as the priority areas designated by NDMO. The red circle indicates Priority Area 1; orange circles indicate Priority Area 2, and yellow circles indicate Priority Area 3 (adapted from OCHA 2023)

Full size image

After TCs Judy and Kevin, a six-month national State of Emergency (SoE) was declared. On the 14th of March 2023, the NDMO cluster-system was activated, and ministerial staff were mobilised and deployed on the ground to complete NDMO’s “first push” activities (e.g., distribution of emergency supplies, such as dry food ratios). The authors engaged with this response through the Vanuatu Fisheries Department (VFD) and formed part of the NDMO Food Security & Agriculture Cluster (hereafter “NDMO/FSAC teams”). This study draws from the experience of the team deployed in the Tafea Province (Fig. 2), particularly on the islands of Tanna and Erromango. The NDMO/FSAC team was allocated to support food distribution for the duration of the “first push” (18th March, to the 13th of April 2023). Following the “first push”, VFD engaged in a loss and damage assessment (hereafter “the loss and damage survey”), aiming to quantify the impacts upon both governmental and private sector assets, to help estimate the requirements and associated costs predicted for the “recovery phase”. This assessment was identified as a priority task to be completed during VFD’s response due to the anticipated impact the cyclones were to have made upon the priority areas.

Utilising census data, VFD estimated that up to 1,096 fishers and 348 aquaculture farmers had been affected with significant degree of damage to boats and fishing gear. The total impacts of the twin cyclones were estimated to reach over 5.2 billion Vatu (~ 680 million AUD) for the fisheries sector (DSPPAC 2023). Predictions indicated that within the fisheries sector, the coastal fisheries sub-sector was most likely to incur the greatest damages from these events. VFD staff were deployed to conduct the loss and damage survey to obtain details on the damages caused by the twin cyclones and inform the recovery phase. Sites were selected according to their proximity to the twin cyclone’s impact zone. The survey sites in the Shefa and Tafea provinces were listed as Priority Area 1 (as per designation by NDMO), while those in the Malampa province were listed as Priority Area 2 since they were located further afield from two cyclone paths (Fig. 2).

Data collection and analysis

The study sought to examine in depth the post disaster collaboration that occurred both within and between communities and the external aid they received, with the aim to clarify how external aid supports such situations. Four dimensions to this questioning guided the research:

  1. i)

    External aid: Individual fishers were asked whether external-aid had been received by the communities affected

  2. ii)

    Collaboration between neighbouring villages: Individual fishers were asked whether collaboration among communities occurred independently of the external-aid received

  3. iii)

    Local practice and kastom: This section included questions from both the village section (which was either completed by the village Chief or a representative of the community when the Chief was absent) and the individual fisher section, and it covered the presence/absence of tabus and tabu rules, as well as whether their village/community has kastoms measures that would help the community post-disaster.

  4. iv)

    Gender perspective: The gender perspective was mainly based on the findings drawn from the qualitative data collected through the NDMO/FSAC deployment in Tanna island.

To do so, two types of datasets were used in this study: i) the loss and damage survey, led by the VFD in May 2023; ii) observational data from the “first push” fieldtrip to the island of Tanna in March–April 2023, as part of the NDMO/FSAC deployment of ministerial staff allocated to the Province of Tafea.

2023 VFD loss and damage survey

The loss and damage survey was conducted from 12 to 29th May 2023 by VFD staff, using tablets with Kobo Toolbox software. This survey assessed the level of fisheries-related damage by the TCs on different aspects (social, economic and ecosystems). It also recorded local practices used by individuals and communities to manage the losses and damages endured. While the loss and damage survey included several fields of enquiry (village profile; individual fisher profile; (aquaculture) farmer profile; and habitats), this study draws from questions within the first three sections of the survey for the five dimensions of enquiry noted above (see SI 1). The survey spanned 17 islands in three provinces: Malampa, Shefa and Tafea (from North to South). Interviews were conducted in Bislama, the lingua franca of Vanuatu, by VFD appointed enumerators. In total, there were 632 participants (71% males and 29% females), of which 62 responded to the village profile section, 551 to the fisher profile section and 19 to the aquaculture farmer section (see Table 1). All the survey data was directly recorded on Kobo Toolbox.

Table 1 Number of respondents (n) of the 2023 VFD loss and damage survey, shown per province and per island
Full size table

Observational data

The experience of being part of the FSAC food distribution deployment team in Tanna Island provided on-the-ground perspectives on opportunities and challenges in delivering food distribution to communities. Detailed notes were taken during the five-week period spent on Tanna. These observations had a particular focus on the external aid brought to Tanna for distribution, organisational challenges and opportunities on the distribution process and gender differences at distribution sites. Upon return from the field, two post-trip briefings where organised by authors to share notes and feedback regarding the food distribution process. The rich observational data provides background on the process at island level and an understanding of challenges faced during on-site response efforts.

Results

Communities within Tafea endured the greatest impacts, with most respondents in this province reporting the TCs as more destructive than TC Pam, in 2015. Survey results indicated that impacts were especially destructive in the small islands of Aniwa and Futuna, where most respondents agreed on this statement (97% and 91.5% respectively). Malampa was the least affected province among the three provinces studied, as expected by the NDMO predictions based on the paths of the twin cyclones.

External aid

The levels of external aid received varied significantly within each province (chi-square test p-value < 0.001). External aid was most reported by Tafea respondents (99% of respondents), followed by Shefa (80%) and Malampa respondents (66%). The descriptions of external-aid received were categorised in three levels: i) National-level aid (including Vanuatu Government help through the NDMO clusters, Vanuatu police, military forces and international aid from New Zealand and Australia, among others); ii) NGOs (including national and international NGOs); iii) Provincial-level aid (including CSOs, Church groups, provincial government authorities such as area councillors and area secretaries, local donors and family members). Overall, national-level aid was most commonly reported (51%), followed by NGO-support (30%) and then provincial-level aid (19%; Fig. 3).

Fig. 3
figure 3

Proportions of different types of help received by communities in each province after the twin cyclones hit Vanuatu in March 2023

Full size image

In Tafea, Tanna received significantly higher levels of external aid (chi-square test p-value < 0.001), followed by the outer islands of Aniwa and Futuna, which received considerable national-level help (68.5% and 54% respectively; Fig. 4). The outer island of Aneityum received the lowest levels of external aid in the province. In fact, very few respondents from the small outer island of Aneityum indicated receiving national-level aid (12.5%). Communities here were mainly supported by NGOs (56%) and provincial-level aid (31%). This was corroborated by the observations collected by the VFD staff deployed in the island of Tanna. The NDMO/FSAC had planned for five staff to service Aneityum, however, owing to adverse weather conditions and to the intensity and magnitude of work needed to complete the distributions in Tanna island (the island with highest population in Tafea) these officers were only able to reach this outer island three weeks after the cyclones.

Fig. 4
figure 4

Map presenting the proportion of each type of external aid received per affected island in the aftermath of the cyclones

Full size image

In Malampa, provincial-aid appeared most important, with most respondents having received help from this level (70%), in contrast with a minority of respondents in Shefa and Tafea reporting to receive aid from provincial bodies (18.6% and 11.6% respectively; Fig. 4).

Overall, 30% of the respondents received help from NGOs, showing that NGOs also played an important role in hel** communities. A significant number of respondents in Tafea and Shefa also reported to have received aid from NGOs (31.5% and 33.7% respectively). All NGOs involved were international organisations, with exception of the Tariwa Organisation; a local NGO that supported the communities on the island of Futuna, in Tafea (Fig. 4).

Local support systems and mechanisms

Collaboration within and between neighbouring villages

Collaboration within and between neighbouring villages was significantly different between provinces (chi-square test p-value < 0.001). Community members from Tafea province most frequently reported collaboration within and/or between neighbouring villages (60.3%), with highest numbers in the outer islands of Tafea (74% in Aneityum, 75% in Aniwa and 62% in Futuna). Residents from Malampa province also reported significant collaboration within and/or between neighbouring villages (56%). This contrasts with Shefa (holding the capital city of Port Vila), where only 16% of respondents mentioned engagement in some form of collaboration among neighbouring communities.

The ways in which communities and community members supported each other were categorised along three main areas: i) Clearing and rebuilding houses and other community assets (e.g., clearing of roads, homes and community assets, as well as sharing tools and materials to do so); ii) Food and water provision activities, which included sharing of aelan kakae (local food) between households, working together to maintain and provide water, distributing relief supplies (including water and food distributed by NDMO), and collective cleaning and replanting of gardens, and iii) Assisting community members, which included ensuring that everyone was accounted for and safe after the TCs (in particular in relation to people living with a disability, elderly, and widows). One community member in Port Resolution (Tanna) noted “the community helped each other by removing fallen trees from roads and sharing food, like yam and banana”.

The three areas in which communities and community members supported each other significantly differed between communities (chi-square test p-value < 0.001). Activities linked to the clearing and rebuilding of houses and other community assets were mentioned the most (Fig. 5), particularly in Tafea (84% of overall responses in the province, and with highest rate of responses in Tanna [92%], Erromango [88%], and Futuna [87%]). These responses correlate with proximity to the path of the cyclones, whereby the category 3–4 winds that hit Shefa and Tafea (priority 1 areas) caused substantially more damage to houses and other structures compared to the damage incurred in Malampa (OCHA 2023).

Fig. 5
figure 5

Map presenting the proportion of the different types of community-to-community collaboration among Ni-Vanuatu populations in the aftermath of the cyclones

Full size image

Food and water provisioning activities were most frequently reported in Malampa and Shefa (43% and 36% of responses respectively). While only appearing in 15% of responses for Tafea (Fig. 5), in the islands of Aneityum and Aniwa specifically, respondents highlighted food and water provisioning as a central way of communities hel** each other after the TCs (33% of responses in both islands). These results were reinforced by observations collected during the NDMO/FSAC distribution process in Tafea. During the distribution processes, one of the main challenges identified to effectively distributing food (other than the weather conditions) was the availability of food rations. Food rations were purchased either in Port Vila or in Santo, and various cargo ships transported the goods to Lenakel (the main port in Tanna). Transport was either directly from Port Vila (Efate) to Lenakel (Tanna) or through the route Luganville (Santo)—Port Vila—Lenakel. These shipments were staggered over a period of four-weeks and food rations were stored in the island of Tanna. As such, rations made available at the inception of distribution were insufficient to service all priority areas within Tanna. This combined with handling and transportation challenges meant that a portion of rations spoiled (tinned fish cartons and rice bags) before they could be distributed. As a result of the limitations in dry-foods availability, food rations had to be distributed in three waves, with knock-on delays in the distribution to the Tafea outer islands of Aneityum and Aniwa.

During the NDMO/FSAC distribution planning, staff were concerned on what the communities of the outer islands would do until the food provisions arrived. Responses from the loss and damage survey reflected this delay, in that some of the outer islands communities were required to find ways of supporting each other by sharing local food, sharing water and replanting gardens. This was in part because they did not receive external aid in the first month after the cyclones, and in some cases when it arrived it was not sufficient.

Finally, assistance to specific individuals or community members was the least mentioned community-led measure in the aftermath of the twin cyclones. Overall, less than 2% of respondents in all provinces reported any collaborations linked with assisting community members (Fig. 5). The few communities that reported assisting widows, elderly and people living with a disability after the cyclones included respondents from the island of Ambrym (Malampa) and respondents from Futuna and Tanna (Tafea). No respondents in the province of Shefa reported to have supported other community members directly.

Kastom practices

Kastom measures were identified as particularly important among respondents, particularly in rural communities. Communities in Tafea showed the highest reliance on kastom measures (64% respondents; chi-square test p-value < 0.001). Records showed utility of kastom in all islands in Tafea. The small islands of Futuna and Aniwa showed the highest numbers of respondents (> 70%) reporting existence of helpful kastom measures after these natural disasters. In Malampa, over 51% of respondents said kastom measures had helped the community after the cyclone, particularly in the small island of Paama (8 km from north to south and only 5 km at its widest point). In Shefa, a smaller proportion of respondents (36%) mentioned the use of kastom during these difficult times (Fig. 6).

Fig. 6
figure 6

Proportions of presence/absence of kastom measures identified by community members in each province after the twin cyclones hit Vanuatu, in March 2023. Note that the question simply asked about the “presence/absence” of kastom measures in their community but it is reflective of the perceived prevalence of kastom, rather than its strict presence or absence

Full size image

Kastom measures mentioned were analysed and classified into three different categories: i) Food preservation and access, and food sharing (46% of total responses), which included opening the tabu area to supply food, food sharing and harvest of local foods, etc.…; ii) Clearing & rebuilding village infrastructure (33% of total responses), including elements such as building shelters, nakamals, clearing roads, etc.; and iii) Community assistance to the less fortunate (20% of total responses), including responses such as assisting widows and people living with a disability. For example, in Tanna, a respondent said “[We] share meals, and kastom medicine treatments”. Some of these kastom practices focused on the preparation for the cyclone, such as food and water preservation, informing the community that a cyclone is coming, and harvesting of local foods pre-cyclones. For example, in the small island of Emae (Shefa) a respondent said the “Chief gathered people and talk to them about cyclone preparation”.

Within the provinces where results denote a greater reliance upon kastom measures, such as Tafea, communities indicated a greater number of people mentioning kastom linked to preservation and availability of local foods (46%) in addition to clearing & rebuilding village (38%). For example, in the island of Aneityum, a man mentioned: “[We use] Traditional measures like not using this water or well which will be used only for cooking and drinking not swimming…”. In Malampa, respondents described kastom practices linked to preservation and availability of local foods (48%), almost as often as community preparation & assistance (41%). For example, a respondent in the island of Ambrym (Malampa) said “The chief opened our tabu eria [closed fishing area] so that people can harvest due to lack of cabbage”.

In fact, we found that in the context of coastal fisheries, “opening the tabu area to supply food” was the most common response (13.4%), after “no kastoms measures used” (45.9%). For example, a resident of Tanna mentioned that after the cyclone they “…opened the conservation eria for a short time”.

The vast majority of the tabu areas mentioned were likely non-formalized, traditional tabu areas (given only a handful are registered with VFD nationally). Temporarily opening tabu areas as a food source has been reported previously by scholars as an approach that supports people in the community after natural disasters (Eriksson et al. 2017). Though respondents in all provinces reported that tabu areas were in place, a significantly greater number of responses that had reported a kastom measure identified tabu areas present in Malampa and Tafea provinces (> 70% responses). Tabu areas seemed to be managed differently in each province. In Malampa, tabu areas were mostly open (68.7%) after the TCs, whereas in the other two provinces they were mainly closed (Tafea, 54.5% and Shefa, 65%). Interestingly, in the Tafea outer islands of Aniwa and Aneytium, 50% or more of the tabu areas were reported to be opened after the twin cyclones. All respondents reported to have tabu area rules, except in Tanna and Erromango (where often no rules were reported).

Finally, a minority of respondents also reported community preparation before the cyclone (such as through building awareness of TCs, forming a disaster response committee and assisting to specific individuals, including widows, people living with a disability and people injured) as kastoms. It is worth noting that these may be instilled structures and practices rather than kastoms.

Women’s role in food distribution

One of the key roles of women that was observed during the NDMO/FSAC food distribution, was the collection of dry goods at collection points on behalf of households. In fact, most recipients at the collection sites were women. In some villages, only women, youth and elderly were present. This rendered the transport of food rations to their households difficult for various reasons. Firstly, the dry food items to be transported were considerably heavy (e.g., each allocation included 18 kg rice bag and 10 kg tinned fish [24 to 48 fish cans per carton] per household). Some communities lived far away from the distribution point and had no means of transport. When living far away, people had to either carry the items home, or organise a car for the community, which was time consuming and often not something women had direct access to. To add to the challenge, women often had (young) children with them, increasing the difficulty of carrying other items. As a result, people at collecting sites often spent the entire day at the site waiting for an opportunity to return home with rations, resulting in them not being able to attend other responsibilities.

Where possible, distribution efforts did include aid agencies that provided targeted assistance for women and youths. A team of the Cooperative for Assistance and Relief Everywhere (i.e., CARE international) was present for some of the NDMO/FSAC food distributions on Tanna (Tafea). Their main role was to ensure that women and children were supported during the process. As most communities did not have electricity, when the distribution was delayed to late in the day, the team of CARE international and the women representing communities agreed to pause the distribution. This allowed time for women to return home, and complete family chores while there was still light. In these instances, communities were required to return without the rations, and come back to the distribution site the following day, prolonging distribution for an additional day and further delaying people’s ability to fulfill pending chores and activities at home.

Discussion

Dynamics of local support and safety nets

Results revealed significant inter- and intra-community collaboration in Vanuatu, both in preparation before cyclones hit and in the aftermath recovery. Below, we expand on this by reflecting on findings along two main streams, namely the variability in ways people respond to disruption depending on their degree of isolation and how people turn to natural assets like fish and marine resources as co** strategy.

Variations in local response capacity based on degree of isolation

Social cohesion is higher in remote areas

Localised support among households was highest in the outer islands of Tafea Province, followed by communities in secondary (islands that do not act as an “urban centre” for outer islands) islands and outer islands of Malampa Province. The lowest report of communities and/or community members supporting each other was recorded in the province of Shefa, particularly on Efate, which includes the capital city of Port Vila. This emphasizes how more geographically isolated communities (small outer islands) supported and engaged significantly more in inter- and intra-community collaboration. A previous study assessing the impact of COVID-19 in Vanuatu similarly describes such trend in forms of collective support to vulnerable individuals and households (Steenbergen et al. 2020).

While such patterns can be associated with the strong values bestowed on social relations amongst kin and households in remote and isolated communities (Steenbergen et al. 2020), it could also arguably be due to necessity when no other options are available. Results from this 2023 study indicated that as geographical isolation increased (e.g. distance to a main city/village) likelihood of delays in the arrival of external aid from the government and other NGOs increased as well. Communities therefore had to act collectively to share the burden of sourcing food and water, and supporting the vulnerable in various ways.

Previous studies have highlighted disaster reduction methods in the Pacific region that support inter-and intra-community cooperation post natural disasters (Campbell 2006, 2014). Findings from this study reiterate Campbell (Campbell 2006)’s reference to community cooperation measures including both, production of surpluses, as well as ceremonies and exchange of goods as part of building ties to face future hardships. We found that collaborative support focused on clearing and rebuilding houses and other community assets, food and water provisioning activities (including sharing of aelan kakae [local foods], as done through kastom – see Box 1), and assisting vulnerable community members. Collective action linked to the clearing and rebuilding of houses and other community assets were mentioned most frequently in Tafea, which also correlates with closer proximity to the path of the cyclones (OCHA 2023).

Box 1 Examining local perceptions on what constitutes kastom versus local collective action.

Respondents often made clear distinctions in what they felt constituted community collaboration measures (i.e., collective action) versus kastom measures, even though analysis of answers from these sections in the survey indicated similar objectives, namely contributing to collective good. For example, an individual from Tanna Island mentioned that the community supported each other by working to repair and clear the road (“cutem lo rod”), however when talking about kastom measures he referred with specificity to the importance of customary practices around the yam harvest (“harvestem lo yam”). This speaks to local broader framing of what kastom entails and potentially also who has legitimate claim to those, understanding that such claims may be contested and that association to kastom comes with local power

It is clear that kastom is deeply embedded in social fabric of communities, however with that then comes the admission that our superficial observations of kastom in this study do not do justice to the depth and complexity of what they represent to the community. Therefore, as much as kastom measures can be seen to provide essential disaster co** strategies through time, they can also complicate externally driven coordination efforts as these interactions can become entwined in subversive power struggles, which can be invisible to the outsider

Food and water provisioning activities, which include activities in preparation to the cyclones were reported to be highest in Malampa and Shefa, while only appearing in a minority of the responses for Tafea. Interestingly, in Aneityum and Aniwa (Tafea), respondents highlighted food and water provisioning as a central way of communities hel** each other after the TCs. Though hel** each other to rebuild the community was still the main intra-community collaboration mentioned, food provisioning was significantly higher in these outer islands than in the rest of Tafea. These results reinforce the observations collected through the NDMO/FSAC distribution process in the province of Tafea, where most efforts focused on the extremely time-consuming and complex process of distributing dry foods to most vulnerable communities in Tanna, while the distribution to the outer islands was delayed due to the weather conditions and the lack of dry food rations available.

Reliance on kastom practice is higher in remote areas

The Pacific way of life has deep connection to customary practices, and this extends to the local ways of governing resources and managing disaster (Eriksson et al. 2017). The use of kastom measures was most pronounced in the outer islands of Tafea Province. The lowest report of communities referring to kastom practices as part of the recovery was recorded in the province of Shefa, which includes the national capital, Port Vila, on the island of Efate. Residents of Shefa province have experienced a greater degree of urbanisation that their rural or small island counterparts further from the capital. This urbanisation has resulted in the change and/or dilution of kastom and traditional cultural influences (Kadfak and Oskarsson 2020). Kastom practices identified as helpful during a natural disaster were in line with the other ways that community members supported each other, for example through food preservation and access (e.g. harvesting and sharing of local produce), clearing and rebuilding villages, and preparation and assistance (hel** the vulnerable in the aftermath). Kastom measures therefore were shown to provide vital support to people before and after a natural disaster. However, as noted in box 1, people’s/groups’ association to local kastom can be socio-politically complex.

Patterns regarding turning to the sea/fish for food

Food security in the Pacific Island Countries and Territories (PICTs) is worsening despite increasing efforts from national and international organisations, development of bilateral projects, national policies and programmes to encourage and support self-sufficiency in food consumption and production (Barnett 2020; Connell 2015; Farrell et al. 2020). Population growth and climate change are increasing pressure on local natural resources and reducing the availability of local foods while exacerbating the reliance of Pacific communities in imported foods (Farrell et al. 2020). Previous studies have identified practical adaptations that could help sustain the important contribution of fisheries to food security (Bell et al. 2018). These arguments become more pertinent considering peoples’ access to natural resources in times of disruption, particularly coastal resources. Our study further emphasizes this, with the most mentioned kastom measure being the short-term opening of closed fishing area (tabu erias) among communities affected by the tropical cyclones. Vanuatu has a long and complex history of community-based marine resource management (Foale et al. 2010; Johannes 2002). This study shows how tabu erias are a central mechanism to support community access to fish-based protein during difficult periods. It furthermore reflects both the significant advancement of marine resource management in rural community life (arguably owing to over three decades of government investment in this area) as well as a binding practice that bridges kastom with fisheries management. We note the latter with caution as we return to the argument in box 1, that appropriation of practice into kastom can also indicate secondary interests such as securing local rights of coastal territories.

Opportunities for engagement for external aid

This study demonstrates NDMO’s central role in supplying relief supplies and dry rations to affected areas. It’s efforts in coordinating external aid covered most affected areas, with the vast majority of communities in Priority Area 1 receiving some form of external aid. This is not to say that nationally-led external aid did not face challenges in reaching communities, particularly those in remote locations and outer islands. The deployment of national teams involved lengthy and expensive processes, with time-consuming coordination of international and national funding. Such processes were compounded by the physically isolated landscape of dispersed, remote islands, where access is often impeded by adverse weather conditions and lack of critical infrastructure (Mileti et al. 1982). Scholars have previously demonstrated the potential role for NGOs in disaster-recovery and management, not only in the provision of dry goods and medical support but also psychological support to individuals and the communities affected (Campbell 2014; Park and Yoon 2022). While all recovery efforts were assumed under the coordinating mandate of NDMO under the national-led program, and so too NGO efforts, this study revealed that not only nationally-led recovery efforts were central, but that NGOs also had an important role in supporting communities post-TCs. This was mainly driven by international NGOs, rather than local NGOs or Civil Society Organisations (CSO). Only one local NGO, the Tariwa Organisation in Futuna, was identified as a localised central part of the external aid received by Futuna community members.

It is increasingly evident that effective disaster recovery hinges on a community development approach, with recovery being conducted at the local level, based on active participation of community members and with maximum reliance on local capacities and expertise (Sullivan 2003). Moreover, much of the capacity to recover from natural disaster depends on having structures in place and agency built in communities over the long term. Vanuatu’s considerable investment to bolster disaster preparedness has resulted in the establishment of mobile phone text-based early warning systems and establishment of community level bodies that aid relief coordination (i.e., community disaster and climate change committees [CDCC]). There remains scope to increase the functionality of the latter through deeper embeddedness in local social institutions and with the recognition of existing (kastom-based) capacities in community (Bubb and Loïc 2022). External aid should be coordinated along with a deeper understanding of diverse local capacities, social resilience and community needs post-natural disaster. Anthropological insights stand to constructively inform such an endeavour, along with national institutions such as the Vanuatu Cultural Centre. Further empowerment and inclusion of local CSOs should be promoted, along with provincial-level aid. While there is certainly a prime role required for centralised national coordination of disaster relief by NDMO, a partial de-centralisation of relief aid in Vanuatu is foreseen to enable provincial and local level CSOs and NGOs to apply their skills, connections with community networks and local resources to support communities in n their post-disaster recovery (Barber 2015; Clarke et al. 2019). As forecasts predict that he Pacific region will be impacted by extreme weather events with increasing frequency due to climate change, further investments in local preparedness and resilience to natural disasters across all primary production sectors is required to ensure rapid local recovery (Zhang and Managi 2020).

Role of women in post disaster

Many knowledge gaps remain as to what role women have in disaster resilience (Alam and Rahman 2017; Charan et al. 2016; Singh et al. 2022). Critical questions to this effect include how gender-based information can feed into policy to enhance community development initiatives; how external-aid programs can adjust their engagement with different social networks; what specific roles women take post-disaster in remote communities versus urban areas and how is that linked to their social position within a community? Various studies have identified different gender roles adopted by community members after a natural disaster in particular communities. Common to these studies is the conclusion that women are generally under-supported in the system, by virtue of their social position, having limited influence over decision making or planning, and restricted access to natural and financial resources (Alam and Rahman 2017; Charan et al. 2016; Misra et al. 2017; Singh et al. 2022). Women have been shown to adopt various roles during natural disasters that enhance community resilience, for example, through enabling communication and enhancing social, economic and ecological resilience (Alam and Rahman 2017; Charan et al. 2016; Singh et al. 2022). It is increasingly recognised that disaster risk management should take into account ways in which gender dynamics influence disaster impacts in any given area before undertaking policy decision-making and design of any project (Wabnitz et al. 2021; World Bank 2021).

These insights apply to the Vanuatu context, where Ni-Vanuatu women play an important role following a natural disaster. One of the key roles identified by this study is that often the recipients at the dry-food collection site were women, and in some villages, only women, youth and elderly were present. In some instances, the predominance of women at some collection sites could be attributed to the prevalence of men spending long periods overseas conducting seasonal work (World Bank 2018). Men may also have been engaged in other community-support activities post cyclone-like rebuilding houses or loading/unloading goods to support distribution of dry foods. In fact, the distribution of relief supplies involves strenuous and physically demanding work, thus mainly young men were recruited as part of the NDMO/FSAC volunteers’ team, potentially leaving a gap in the community for a short period of time. This study has revealed important leads for further research on the role of women in such difficult situations, and highlights the importance of including aspects of inclusion and (in)equity in workload and responsibilities required.

Conclusion

On the 1st and the 3rd of March 2023, Vanuatu was hit by two consecutive category-4 cyclones, TC Judy and TC Kevin. The VFD-led loss and damage survey documented immediate impacts of the two cyclones, on coastal communities in affected zones (including communities of varying remoteness across capital island, secondary island, and outer island sites). Various individual, village and ecosystem level socio-ecological impacts were documented and some of these findings, along with field observations from the NDMO-led first push, were utilised for this study. Results revealed various dynamics of local support systems in Vanuatu and how those relate to external aid efforts, and discussed the role of women in post-disaster periods.

The disaster response system of Vanuatu, during the consecutive category-4 cyclones in 2023, shows how, in the context of limited infrastructure, challenging dispersed island environments, and insufficient capacity to handle sudden surges of funding support – government is limited in its ability to administer multi-scaled, multi-faceted operations. What often goes unrecognised is the extreme working conditions and long working hours that government staff are subject to in these periods. In view of improving an already established disaster response system in Vanuatu, strategic approaches that are cost-effective and geared towards prioritising foundational needs should be prioritised. The study reveals that those needs are not universal, particularly following the cessation of the immediate first push efforts – there are significant variations in disaster responses based on cultural context, location and local organisation.

Disaster preparedness needs to be a central part of all rural development service delivery to communities outside of disaster periods. The VFD through CBFM initiatives stands to increase capacity not only on the technical front of fisheries management but also through establishing connections between governance institutions and community networks. This study also identified important leads for further research on the role of women before and after disaster – to understand aspects of inclusion and differential impacts across marginalized groups, including areas of potential (in)equity. This warrants future research to, for example, unravel the ways women manage such difficult situations and how can they be supported throughout the process, by all including their community, local CSOs and NGOs as well as external-aid. Disaster risk management should take into account ways in which gender dynamics influence disaster impacts in any given area before undertaking policy decision-making and design of any project.

This study illustrated the importance of understanding social systems and their role in enabling community-to-community support, which was found to be particularly critical in the first weeks after a disaster. We argue that, along with external-aid, the social systems discussed here (including social networks, social cohesion, and other types of social capital) are central to disaster response and recovery for Pacific people. Recognising the roles that different support networks, such as external aid and community-to-community support, play at different stages of disaster response and recovery is critical to improving disaster risk management. More broadly speaking, enabling synergies between local and national support systems is key to strengthening community resilience in the face of climate change.