Livelihood resilience of smallholder livestock farmers in the face of climate change in Fiji

Abstract

Smallholder livestock farming in Fiji, a cornerstone of rural livelihoods and food security. However, is increasingly threatened by climate change-induced cyclones, droughts, floods, sea level rise and rising disease pressures. This study assessed the climate resilience of 270 smallholder livestock farmers across six sites representing Fiji’s major agro-ecological zones (Tailevu, Naitasiri, Sigatoka, Rakiraki, Labasa, and Seaqaqa) using the Sustainable Livelihoods Framework (SLF) and a Participatory Resilience Index based on five capital assets: human, social, physical, financial, and natural. Results revealed uniformly low resilience (0.172–0.241), with the lowest scores recorded in the sugarcane belt of Sigatoka, Rakiraki, Labasa, and Seaqaqa, where the collapse of the sugarcane industry has left farmers with low alternative source of income, poor rural roads, limited veterinary services, degraded pastures, and limited market access. Social and physical capitals were the most deficient across all sites, while financial capital performed relatively better in the wetter zones of Tailevu and Naitasiri, where livestock farming thrives as a result of social capital, such as that provided by Fiji Cooperative Dairy Company Ltd (FCDCL), that enable them to access government subsidies, low interest credit services and access market through a milk collection chilling plant that is close to farmers and the potential alternative income from root crop farming such as cassava and Dalo (arrowroots). The Binary Probit Regression revealed that membership in farmers cooperatives, secure fencing, on-farm water harvesting, access to credit, and presence of animal feed were the strongest predictors of both short-term coping (odds ratios 1.88–2.46) and long-term adaptation (odds ratios 1.71–2.33). Farmers ranked animal diseases, absence of improved breeds, low prices of livestock products, and lack of reliable markets as the major barriers to adaptation, while strongly endorsing training, extension services, and climate information as key enablers. Interventions are needed to break the low resilience trap driven by aging and declining farmer population as well as the contraction of alternative income sources such as sugar cane farming. There is an urgent need for balanced, multi-capital interventions particularly organizing farmers in all areas into co-operatives, investing in low-cost physical infrastructure such silvopastoral system and the reduction of price of animal feed through government subsidies. Such interventions are essential for building a climate resilient livestock sector capable of supporting Fiji’s rural livelihoods, food and nutrition security.

1 Introduction

Climate change is emerging as one of the most critical global challenges of the 21st century, with far reaching consequences for natural ecosystems and human livelihoods (IPCC, 2023; Jha and Dev, 2024). Increasing frequency and intensity of extreme weather events such as heatwaves, droughts, tropical cyclones, and floods disrupt agri-food supply chains, slow global agricultural productivity, and threaten the availability of safe, nutritious, and affordable livestock products for a growing global population (Bilotto et al., 2024; Godde et al., 2021). Livestock production systems worldwide are increasingly exposed to rising temperatures, erratic precipitation, water scarcity, and climate-related disasters, all of which contribute to production losses, Increased disease outbreaks and livestock mortalities (FAO, 2019; Thornton et al., 2009). It is estimated that more than 65% of disruptions in livestock production are linked to extreme climatic events (Cottrell et al., 2019).

Globally, the livestock sector supplies approximately 17% of human calorie intake (Thornton et al., 2009) and contributes about 40% of agricultural Gross Domestic Product (GDP) (World Bank, 2022). It creates substantial employment opportunities for rural households. Additionally, livestock are a major source of food, nutritional security, livelihood and income in developing countries (Herrero et al., 2013). Increasing incomes, changing diets, and population growth have led to increased demand and made the livestock sector one of the fastest growing agricultural sub-sectors in middle- and low-income countries (World Bank, 2022). Livestock farming’s social and economic significance underscores the urgency of adapting to climate-related risks. The Intergovernmental Panel on Climate Change (IPCC) warns that continued increases in temperature and extreme weather events may lead to irreversible agro-ecological changes (IPCC, 2023), placing livestock systems and national economic growth at substantial risk in most countries.

In the South Pacific Islands, climate change poses acute challenges to agricultural production. The region is experiencing increasing intensity of tropical cyclones, altered rainfall patterns, prolonged droughts, rising temperatures, and sea-level rise (Fernández-Duque et al., 2025; IPCC, 2023; McIver et al., 2016; Sagero et al., 2025a). These changes reduce pasture quality and availability, increase heat stress on livestock, limit water resources, and facilitate the spread of animal diseases and pests (Magiri et al., 2024). For example, Tropical Cyclone Winston in 2016 caused over FJD$5 million in livestock losses in Fiji, severely affecting cattle, goat, sheep, pig, and poultry producers (Government of Fiji, 2016). Persistent disease challenges, such as bovine brucellosis and tuberculosis (BTB), further constrain productivity and restrict international trade in Fiji (Bundi Magiris et al., 2023).

Fiji’s livestock sector is dominated by traditional production systems (Singh et al., 2023), which are strained by socio-economic limitations and are increasingly unsustainable under changing climatic condition. In 2020, the sector contributed approximately 0.9% to Fiji’s GDP and supported the livelihoods of more than 23,000 rural households (Ministry of Agriculture Fiji, 2020). Despite its importance, domestic production remains insufficient to meet national demand, with Fiji importing over 70% of its beef and mutton and around 50% of its dairy products (Singh et al., 2023; Bundi Magiri et al., 2023). Production shortfalls stem from limited infrastructure, low technological adoption (FAO, 2019), and constraints such as BTB, limited veterinary services, poor-quality breeding stock and the impact of climate change (Magiri et al., 2024). Only poultry production meets local demand (Zindove et al., 2022).

Smallholder agricultural systems often use family labor particularly women, children, and elders in their management (Galhena et al., 2013). For them to contribute and help support national food security these farms must (1) contribute to food production through direct household consumption and reducing dependance on external food systems, therefore, increasing availability, (2) provide income and livelihood, (3) be able to contribute towards diverse diets, and (4) be able to be used as a buffer to market-related shocks to withstand fluctuations in market prices, production losses, and climate-related disruption (HLPE, 2013). As such, smallholder farmers play a significant role in improving the food security status of many households and offer different benefits. Even so, obstacles such as climate change, gender disparity and discrimination related to land and water ownership among others, pose a serious threat to smallholder farms’ ability to produce enough food for their own consumption and for sale (Baylie, 2025).

On the other hand, the resilience of smallholder livestock farmers relies on the capacity of farming systems to anticipate, absorb, recover from, and adapt to adverse climatic conditions while maintaining essential functions and outputs (Meuwissen et al., 2019; Singh et al., 2023) Building resilience requires climate-smart agricultural strategies, including improved pasture management, enhanced animal health systems, diversified livelihood options, climate-resilient infrastructure, and reliable water storage and supply (Meuwissen et al., 2019). Despite increasing impacts of climate change, limited research exists on how livestock farmers in Fiji perceive, experience, and respond to these changes. Most existing studies focus predominantly on crop-based systems (Lata and Nunn, 2012; Singh et al., 2023). This gap highlights the need for a deeper understanding of the vulnerabilities and adaptive capacities of smallholder livestock farmers to inform targeted, context-specific interventions.

Existing resilience research in Fiji demonstrates that climate impacts are mediated not only by hazard exposure but also by social, institutional, and behavioral factors that shape household responses. Gawith et al. (2016) found that communities with stronger resilience characteristics experienced lower climate-related losses during cyclone events in Fiji, particularly where early warning and preparedness enabled anticipatory action. Their study highlighted the importance of social capital, infrastructure, institutional support, and economic resources in reducing disaster losses. Similarly, Brown et al. (2018) showed that exposure to tropical cyclones significantly influenced household risk perceptions, with adaptive responses shaped by informal social protection systems and communal support structures. More recently, Gatiso and Greenhalgh (2025) demonstrated that social capital remains a central determinant of socioeconomic resilience under climate hazards because trust, cooperation, and collective action improve information sharing, resource mobilization, and recovery processes. However, little is known about how these resilience dimensions operate specifically within smallholder livestock systems, where adaptive decisions are shaped by herd management, feed availability, water access, animal health, and market disruptions under recurring climate stress.

Climate stresses extend beyond animals, affecting farmers and entire communities, especially in regions with limited resilience and resources (Brouillet and Sultan, 2023). Resilience of livestock farmers is closely connected to livelihood resilience, which emphasizes the social–ecological capacities households need to cope with and adapt to climatic shocks. According to Tanner et al. (2015), livelihood resilience goes beyond “bouncing back”; it involves the ability to absorb impacts, re-organize livelihood strategies, and transform systems in response to long-term environmental changes. Consequently, understanding the resilience of smallholder livestock farmers is critical for identifying areas that require greater investment and targeted support. Such insights can enhance these farmers’ adaptive capacity and safeguard their livelihoods in the face of increasing climate extremes.

The SLF highlights that resilience varies based on access to five forms of capital: social, financial, human, physical, and natural (United Nations Development Programme, 2017). Applied to livestock farming, this underscores the need for diversified production systems, improved access to resources, strong social networks, climate-smart practices (in feed conservation, water harvesting, improved housing, breed selection and silvopastoral management), and supportive institutional environments including effective veterinary services, secure land tenure, functioning markets, and equitable governance. Thus, livestock resilience is not merely biological but reflects broader household adaptive capacity, livelihood security, and long-term well-being under climate change (Cullen et al., 2021; Thornton et al., 2009).

Several studies have used the SLF to assess resilience in agricultural systems, particularly among smallholder farmers facing climate and socio-economic shocks. Empirical studies consistently demonstrate that resilience in agriculture is shaped by the interaction of human, social, natural, physical, and financial capital. For example, Sun et al. (2023) and Hu and Dong (2023) show that farmers with higher levels of livelihood capital exhibit greater capacity to absorb, adapt to, and recover from shocks, including climate variability and pandemic-related disruptions. Similarly, Ankrah et al. (2023) find that access to human, informational, and financial capital significantly influences farmers’ adoption of climate adaptation strategies in Southern Ghana. (Kwayu et al., 2017) further demonstrate that policy interventions such as payments for ecosystem services can enhance livelihood resilience by strengthening access to natural and financial capital. Studies also highlight the central role of social capital, with strong networks and group participation in improving farmers’ access to resources, information, and collective coping mechanisms (Diehl and Bose, 2023; Wang et al., 2021). Overall, these studies confirm that agricultural resilience is multidimensional and best understood through an integrated livelihood capitals perspective, where deficiencies in one form of capital may constrain resilience despite strengths in others (Ellis, 2000).

This study assesses the resilience of smallholder livestock farmers across five provinces in Fiji, where livestock production remains a cornerstone of rural livelihoods and food security. Drawing on the SLF, it evaluates farmers’ adaptive capacity through the five capitals; human, social, physical, financial, and natural to identify key strengths, vulnerabilities, and systemic constraints that shape their ability to withstand and recover from climate change impacts. By generating robust, context-specific evidence on how the farmers experience and respond to climatic stresses, the research provides critical insights for designing targeted interventions that enhance the long-term resilience, productivity, and sustainability of Fiji’s livestock sector, ultimately informing policymakers, extension services, and development partners in prioritizing strategies that strengthen livelihood security, reduce climate-related risks, and bolster national food and nutrition security.

2 Materials and methods

2.1 Study area

This study was conducted in Fiji, a Pacific Island nation comprising over 300 islands (Figure 1), where the island of Viti Levu, and Vanua Levu have the highest concentration of livestock farming (Kumar & Reddy, 2008). The study was conducted across six sites representing the principal agro-ecological zones and farming systems of Fiji: Tailevu (Central Division, Viti Levu), Naitasiri (Central Division, Viti Levu), Sigatoka (Nadroga-Navosa Province, Western Division, Viti Levu), Rakiraki (Ra Province, Western Division, Viti Levu), Labasa (Macuata Province, Northern Division, Vanua Levu), and Seaqaqa (Macuata Province, Northern Division, Vanua Levu) (Figure 1). These sites were purposively selected in consultation with a multi-stakeholder group comprising the Ministry of Agriculture and Waterways (MoAW) (Animal Health and Production Division), Fiji National University, and the University of the South Pacific, Selection criteria included: (a) high concentration of smallholder livestock farmers (herd size ≥ 5 animal equivalents), (b) representation of major production systems (beef, dairy, goat, sheep, pig and poultry), (c) diversity of agro-ecological conditions (wet-zone, dry-zone), (d) historical exposure to climate extremes (cyclones, droughts, floods), (e) accessibility, and (f) presence of active farmers registered by MoAW Animal Health and Production Division, or a farmers cooperatives; FCDCL.

Figure 1

The selected sites were deliberately chosen to represent Fiji’s agro-ecological and climatic gradient, which shapes livestock production systems and resilience differentially. Tailevu and Naitasiri provinces (Central Division, Viti Levu) are in wetter, windward zones with higher and more reliable rainfall, favoring dairy and mixed livestock systems reliant on pasture based grazing and cooperative structures. In contrast, Sigatoka and Rakiraki (Western Division, Viti Levu) occupy drier leeward areas dominated by sugarcane belts, where drought proneness limits pasture availability and favors beef cattle and small ruminants (goats/sheep) with adaptations such as drought tolerant breeds and water issues is highly valuable. Labasa and Seaqaqa (Northern Division, Vanua Levu) feature flat volcanic soils with intermediate but variable rainfall and exposure to flooding and drought, supporting mixed systems with variable water and disease pressures. These environmental contrasts influence key livelihood factors including pasture quality, water access, fodder management, disease prevalence, and overall resilience. While formal disaggregation by livestock species was not incorporated into the Analytic Hierarchy Process (AHP) weighting due to sample size constraints, this limitation is acknowledged, and future quantitative modeling with larger datasets could enable more granular species-specific analysis.

Fiji’s climate is classified as tropical maritime, with a distinct wet season (November to April) and dry season (May to October) (Sagero, et al., 2024). Rainfall in the Fiji Islands is highly variable and is influenced by the island topography of the larger islands (Viti Levu and Vanua Levu) and the prevailing south-east trade winds (Sharma et al., 2021; Kuleshov et al., 2014; Kumar et al., 2014). These features result in a wet zone on Fiji Islands’ eastern side and a drier zone on the western side (Terry, 2007). The annual rainfall exhibits strong spatial variability: windward (eastern/southeastern) sides receive 3000–6000 mm (wetter zones), while leeward (western/northwestern) sides average 1500–2000 mm (drier zones) (Terry, 2007; Sagero et al., 2024). Moreover, during the wet season an average of four Tropical cyclones per year pass through Fiji’s Exclusive Economic Zone, contributing up to 20% of Fiji’s total seasonal rainfall (Jiang and Zipser 2010; Sharma et al., 2021). Additionally, the Fiji Islands rainfall is highly influenced by global and regional-scale climatic systems such as the El Niño Southern Oscillation (ENSO), South Pacific Convergence Zone (SPCZ), and the Madden–Julian Oscillation (MJO), which influence the spatial temporal variability of rainfall in several areas of Fiji (Deo 2011; Weir et al., 2021; Kumar et al., 2006).

2.2 Data collection

A mixed-methods approach was employed combining quantitative household surveys, focus group discussions (FGDs), key-informant interviews, and farm characterization walks. A household survey was conducted with a representative sample of smallholder livestock farmers. Sample size was calculated using Slovin’s formula (Ryan, 2013) Equation 1 at 95% confidence level.

The 270 households were selected via stratified random sampling from MoAW registries and FCDCL lists of registered farmers (herd ≥5 animals equivalents). Stratification used primary enterprise (beef, dairy, mixed, goats/sheep, pigs, poultry) and ensured approximately 45 households per site (proportional to estimated density). Equation 1 guided overall size from an estimated accessible population of livestock farmers 4,500–5,000 smallholders across sites. A structured questionnaire was administered to 270 livestock farming households across the sites with assistance from local livestock officers. Both iTaukei and Indo-Fijian farmers were included in the survey. The structured tool comprised closed-ended sections on (i) demographics & farm characteristics, (ii) herd composition & production parameters, (iii) shock exposure & recovery, (iv) Likert/perception scales for climate change impacts, (v) self-scoring of SLF capital indicators, and (vi) ranking of barriers/enablers. It included binary items for coping/adaptation outcomes. Six focus group discussions (FGDs) were conducted; one per study site with 8–12 participants each (mixed gender where feasible). Each session lasted 90–120 minutes and followed a Talanoa approach, Pacific Indigenous dialogue method emphasizing relational, open, and empathetic conversation (Farrelly and Nabobo-Baba, 2014).

Semi-structured guides were used to probe participants’ perceptions of climate and other shocks, current adaptation practices, and access to livelihood capitals. Additionally, key informant interviews were carried out (two per site). These targeted relevant experts, including MoAW officers, officers from Yaqara Pastoral Company Limited (YPCL), officers from the Sigatoka Research Station, and officers from the FCDCL. Farm characterization walks were also undertaken on one farm per site (covering both small-scale and large-scale operations). These involved on-site documentation and assessment of key infrastructure elements, including pastures, water sources, fencing, and other facilities. All the interviews were carried out by the principal researchers.

Daily rainfall and temperature data were received from Fiji Meteorological service (FMS), where monthly and seasonal rainfall trends, along with rainfall and temperature extreme indices, were analyzed using the non-parametric Mann-Kendall test (Mann, 1945) for trend detection and Sen’s slope estimator (Sen, 1968) for quantifying the magnitude of change (in mm/year for rainfall and °C/year for temperature) at five meteorological stations located at the study sites (Labasa Field, Lautoka, Nacocolevu, Nausori Airport, and Seaqaqa).

2.3 Resilience assessment of smallholder livestock farming communities

This study employed the SLF as the principal analytical framework for assessing livelihood resilience among smallholder livestock farmers in Fiji because resilience in livestock-dependent households is multidimensional and shaped by access to human, social, natural, physical, and financial livelihood assets (Table 1). Where human capital encompasses the abilities, experience, work skills and good health, that, when combined, allows populations to engage with different livelihood strategies and reach their objectives. Social capital refers to the social resources individuals rely on to achieve certain objectives relating to their livelihoods, including networks and connections, and participation in formal groups. Natural capital is the stock of natural resources from which further resources and services can be developed, which may prove useful to livelihoods. In contrast, physical capital comprises the basic infrastructure and producer goods needed to support livelihoods. Financial capital refers to the financial resources that people use to achieve their livelihood (Ellis, 2000) (Table 1). The SLF has been widely applied in rural resilience research because it captures how households mobilize multiple forms of capital to absorb, adapt to, and recover from environmental shocks while maintaining livelihood functions (Scoones, 1998; Department for International Development, 1999; United Nations Development Programme, 2017). In livestock systems, where climate exposure affects feed resources, water availability, animal health, and market access simultaneously, the framework provides a useful basis for examining resilience beyond income-based indicators alone.

Indicators representing the five livelihood capitals were selected through a combination of literature review, field relevance, and stakeholder validation. First, indicators commonly used in resilience studies of smallholder agricultural systems were identified from previous empirical work (Adger, 2000; Hahn et al., 2009; Adimassu and Kessler, 2016). These were then adapted to reflect livestock production realities in Fiji, including herd size, access to veterinary services, pasture condition, livestock infrastructure, cooperative participation, and off-farm income diversification. To improve contextual validity, the indicator list was discussed with livestock extension officers, farmers representatives, and local stakeholders during field consultations before final inclusion.

Because the selected indicators were measured in different units, all variables were normalized to a common scale ranging from 0 to 1 before aggregation. Positive indicators, where higher values represented stronger resilience, were standardized using min–max normalization, while inverse transformation was applied to indicators negatively associated with resilience, such as distance to veterinary services or exposure to repeated production losses.

This procedure reduced measurement bias and ensured comparability across households. Indicator weighting was undertaken using the Analytic Hierarchy Process (AHP), which is widely used in resilience and multi-criteria assessment because it allows relative importance to be assigned systematically through pairwise comparison (Saaty, 1980). The normalized resilience index for each household was therefore estimated as per Equation 2:

Where represents the resilience score for household i, ​ denotes the normalized weight assigned to indicator j, and represents the standardized indicator value for household i. Livelihood capital scores were then aggregated across each capital domain, and the composite climate resilience index was calculated as the arithmetic mean of the five capital scores. The weights and scores are achieved by pairwise comparisons between all options with each other (Kaspercyzyk & Knickel, 2006). Following established resilience classification thresholds, resilience scores were interpreted as low (0.00–0.33), moderate (0.34–0.66), and high (0.67–1.00) (Hahn et al., 2009).

Given the diversity of livestock systems across the study sites, livestock heterogeneity was explicitly incorporated into the analysis. Farmers were grouped according to dominant livestock production orientation, including dairy cattle, beef cattle, and mixed small livestock systems involving goats, sheep, pig and poultry. This distinction was necessary because different livestock systems face different climate sensitivities and production constraints. Dairy households, for example, depend heavily on regular feed supply, cooling infrastructure, and milk market access, while mixed small-stock systems often display greater flexibility but lower institutional support. Treating livestock categories separately therefore improved interpretation of resilience differences across districts.

2.4 Determinants of coping and adaptation behavior: Binary Probit regression

To identify the key factors influencing smallholder livestock farmers’ ability to (i) cope with climate shocks (e.g., cyclones, droughts, floods) and (ii) adopt long-term adaptation practices (e.g., improved breeds, silvopastoral systems, water harvesting, fodder conservation), two separate binary Probit models were estimated following the method describes by Greene (2018) and Wooldridge (2010).

The dependent variables were constructed as binary outcomes from survey questions:

• Coping = 1 if the household reported having fully recovered livestock numbers and income within 12 months after the most recent major shock (e.g TC Winston, or prolonged drought or extreme floods or disease), 0 otherwise.

• Adaptation = 1 if the household had proactively adopted at least two climate-smart livestock practices in the past five years, 0 otherwise.

The Probit model assumes that farmers’ observed response reflects an unobserved latent variable Y_i^∗ representing the underlying propensity of a household to cope successfully or adopt adaptation strategies under climate-related stress (Equation 3).

Where: , latent resilience decision of household i, β₀​, is intercept, β1​…βk​ are coefficients of explanatory variables, ​… are explanatory variables and is the normally distributed error term. The observed binary outcome is defined as:

Thus, the probability that household i successfully cope with or adapts is expressed in Equation 4:

where Φ(.) is the cumulative distribution function of the standard normal distribution. The theoretical basis and estimation procedure for the Binary Probit model are extensively discussed in Greene (2018).

All continuous independent variables were checked for multicollinearity (VIF < 3.5), and distance variables were log-transformed to reduce skewness. Marginal effects are reported as odds ratios (exponentiated coefficients) for ease of interpretation: an odds ratio > 1 indicates a positive association with successful coping/adaptation, while < 1 indicates a negative association. Site fixed effects were included to control unobserved agro-ecological and provincial differences. Robust standard errors clustered at the village level were used. Model fit was assessed using Pseudo-R², log-likelihood, χ² statistic, and percentage of correctly classified observations. All estimations were performed using R program.

The Binary Probit model was preferred because it provides consistent and efficient parameter estimates for binary dependent variables, while avoiding the limitations of ordinary least squares regression, which may generate predicted probabilities outside the logical 0–1 interval and violate assumptions of homoscedasticity and normality (Wooldridge, 2010). It was further selected because household adaptation decisions are influenced by multiple interacting socio-economic and environmental constraints, making the latent threshold structure of the Probit model appropriate for representing cumulative behavioral responses (Greene, 2018). In agricultural adaptation studies, Probit models are widely applied to estimate adoption and coping decisions because farmer responses often reflect latent utility maximization under climatic uncertainty (Deressa et al., 2009; Di Falco et al., 2011).

Coping capacity was operationalized as a household’s ability to achieve short-term functional recovery following the most severe recent climate-related disturbance affecting livestock production. Rather than defining coping solely as complete restoration of pre-shock livestock numbers or full income recovery within a fixed twelve-month period, the revised measure captures whether households were able to restore core livestock production functions, regain partial productive capacity, and resume normal livelihood operations such as feed access, herd maintenance, and minimum market participation. This approach recognizes that complete herd recovery may remain constrained by repeated climatic shocks, post-disaster livestock supply shortages, market disruptions, livestock disease outbreaks, and delayed access to replacement breeding stock. Coping is therefore interpreted as short-term operational recovery under prevailing livelihood constraints rather than full return to pre-disturbance production levels, consistent with resilience theory in smallholder livestock systems where recovery commonly occurs over multiple production cycles (Adger, 2000).

This specification is consistent with widely used approaches in examining the determinants of households’ coping and adaptation behavior in response to climatic and socio-economic shocks, particularly in agricultural settings. Using survey data, these models estimate the probability of adopting specific coping or adaptation strategies as a function of household characteristics, asset endowments, institutional access, and shock exposure as used by (Adimassu and Kessler, 2016; Das, 2021; Gautam et al., 2021; Wandera et al., 2024). These estimates complement the descriptive resilience index by providing rigorous statistical evidence on which capitals most strongly drive actual coping and adaptation behavior.

Adaptation behavior was also modeled as a binary outcome reflecting proactive climate-responsive farm management. Rather than treating all adaptation as a single undifferentiated response, individual practices were first recorded separately, including feed conservation, improved breeds, water harvesting, silvopastoral practices, housing improvement, and veterinary preparedness. A household was classified as adaptive when at least two climate-responsive livestock practices had been intentionally adopted within the previous five years. This threshold was selected because isolated management actions may reflect routine husbandry rather than deliberate adaptation planning. The binary threshold was then tested through sensitivity analysis using alternative cut-offs of one and three strategies to assess whether major explanatory relationships remain stable.

Sensitivity analysis was conducted to evaluate the robustness of methodological choices. First, resilience scores were recalculated using both AHP-derived weights and equal weights to determine whether sites rankings were sensitive to weighting assumptions. Second, coping models were re-estimated using alternative recovery periods of six, twelve, and eighteen months to account for differing livestock recovery cycles. Third, adaptation thresholds were varied across one, two, and three adopted practices. In all cases, coefficient signs and site-level resilience patterns remained broadly stable, indicating that the principal findings were not driven by a single modeling assumption.

To strengthen interpretation of quantitative findings, qualitative evidence from focus group discussions and key informant interviews was integrated throughout the analysis. Farmer narratives were particularly important for explaining why social capital, physical capital, and natural capital remained consistently weak across several districts, despite moderate variation in financial indicators. This mixed-method integration allowed measured resilience outcomes to be interpreted within the practical realities of livestock production, institutional access, and climate exposure experienced by farmers in Fiji.

2.5 Ethical considerations

Ethical approval was obtained from the University of the South Pacific, environmental research committee (USP-ERCE-16-23).

3 Results

3.1 Socioeconomic and farm characteristics

Most of the respondents were male (87.3%), reflecting the male-dominated nature of livestock management in Fiji, while women’s involvement was largely confined to poultry and goat rearing. Most farmers were over 45 years of age (69%), married (80.3%), and had more than 15 years of livestock farming experience (69%). Household sizes were relatively large, with 47.9% of households having more than six members, thereby providing a substantial family labor force. Education levels varied considerably: 39.4% had completed secondary education, while 33.8% had certificate, diploma, or degree qualifications. Annual household income from livestock and mixed farming was generally low to moderate, with 49.3% earning less than FJD 29,000. Native (iTaukei) land tenure dominated (49.3%), followed by freehold (32.4%) and crown leases (18.3%) (Table 2).

The topography across the project sites varies, with Sigatoka and Rakiraki characterized as undulating to hilly coastal plains, Naitasiri and Tailevu as inland hilly and low laying terrain, and Labasa and Seaqaqa as flat to rolling volcanic soils. Most farms in all study sites are rainfed, relying on natural pastures and limited supplementary feed, which exacerbates vulnerability to droughts, floods and cyclones. Average livestock holdings ranged from 12 to 50 animal equivalents (including goats, sheep, and poultry), with the largest herds in Tailevu and Naitasiri due to dairy farming practices that is largely driven by dairy farmer cooperative, the FCDCL, Sigatoka, Labasa, Rakiraki, and Seaqaqa, were mainly goats and sheep, pig, and poultry. Land ownership patterns showed a higher proportion (up to 80%) were tenants on leased land and often integrated with crop farming.

Livestock farming was not the primary source of household income in most surveyed households. The main household income came from crop sales; sugarcane in Sigatoka, Rakiraki, Labasa and Seaqaqa and root crops and vegetable farming in Naitasiri, Tailevu, Sigatoka and part of Rakiraki. Other off-farm activities were wage labor from tourism (Sigatoka, and Rakiraki) or construction, and transport businesses. Goats were common in Sigatoka, Labasa, Seaqaqa and Rakiraki due the dry weather conditions which is more suitable for the small ruminant production. Farmers have been engaged in livestock rearing for an average of 18–25 years, with 2–4 family members typically involved. Annual income from livestock was generally low, below FJD 5,000 in Sigatoka and Rakiraki, attributed to smaller herd sizes and market access challenges, while Naitasiri and Tailevu reported slightly higher incomes (FJD 20,000–100,000) due to proximity to dairy cooperatives. In Labasa and Seaqaqa, income from livestock ranged between FJD 10,000 and 200,000, with commercial poultry farming and beef cattle dominating the sites.

Table 2 shows spatial differences across the six study sites, reflecting underlying agroecological conditions. The wet zone sites of Naitasiri and Tailevu are dominated by dairy production systems, characterized by larger herd sizes and higher livestock incomes. In contrast, the dry zone sites of Rakiraki and Sigatoka are predominantly oriented toward beef cattle and small ruminants, with significantly lower income levels due to water scarcity and limited pasture productivity. Labasa and Seaqaqa exhibit more diversified livestock systems, particularly poultry and pig enterprises, reflecting adaptive responses to variable climatic and market conditions (Table 2).

3.2 Farmers’ perception about climate change

The results in Table 3 shows a strong and consistent perception of climate change among smallholder livestock farmers across all study sites, with 72.3% of respondents indicating that the climate has changed, compared to 27.7% who perceived changes mainly as climate variability. This perception was slightly higher in Seaqaqa (76.7%) and Naitasiri (75.6%), and lower in the drier sites of Sigatoka (66.7%) and Rakiraki (68.9%), where variability was more frequently reported. Across all locations, the majority of farmers perceived an increase in daytime temperatures (82.2%), with particularly high responses in Seaqaqa (86.7%) and Naitasiri (84.4%), while only small proportions reported decreases (7.6%), no change (5.8%), or uncertainty (4.4%). A similar pattern was observed for nighttime temperatures, with 76.2% reporting an increase, highest in Seaqaqa (82.2%) and Labasa (77.8%), and lower in Sigatoka (71.1%) and Rakiraki (73.3%). Reports of decreasing temperatures (8.2%), no change (7.5%), and uncertainty (8.1%) were minimal. Overall, the findings indicate a clear and widespread perception of rising temperatures across both daytime and nighttime conditions, with some spatial variation, particularly in the drier western regions where climate variability is more commonly perceived.

3.3 Observed rainfall and temperature trend

Analysis of rainfall and temperature trends shows a consistent warming pattern alongside declining or variable rainfall. Temperature trends are highly significant and positive across all sites, with annual Sen’s slopes of 0.02–0.03 °C/year (p<0.001), driven by increases in both summer and winter seasons (Supplementary Table 1). Extreme temperature indices confirm this warming, including more frequent warm nights (TN90p) and warm spell duration (WSDI), alongside reductions in cold nights (TN10p) and cold spell duration (CSDI), indicating a clear intensification of heat extremes (Supplementary Table 2).

Rainfall exhibits predominantly negative trends, particularly at Seaqaqa (significant annual decline of -2.88 mm/year, p<0.01) and during summer/winter periods at several stations (Supplementary Table 3), with widespread decreases in total precipitation (PRCPTOT), very wet days (R95p/R99p), and overall precipitation intensity. While monthly trends vary (e.g., positive in February at multiple sites but declines in other wet months elsewhere), extremes show mixed signals such as slight increases in RX5day at Seaqaqa but reductions in heavy rainfall days and consecutive wet days at most locations pointing to a general drying tendency with regional and seasonal variability across Fiji (Supplementary Table 4).

3.4 Farmers’ perception of the impact of climate change on livestock farming activities

Figure 2 shows that climate change impacts on livestock production are widely perceived as severe across all agroecological zones, with consistently high proportions of respondents rating most impacts as extremely impactful. Pest and disease emerge as the most critical issue, with very high responses across the wet (93.3%), dry (97.8%), and northern (96.7%) zones, indicating a near-universal concern. Reduced pasture and animal health also rank highly, particularly in the dry zone (91.1% and 82.2%, respectively), reflecting the vulnerability of feed resources and livestock conditions in drier environments. Water availability shows the greatest variation, with relatively low extreme impact in the wet zone (36.7%) compared to the dry (62.2%) and northern (47.8%) zones, highlighting spatial differences in water stress. Soil quality impacts are moderately high across all zones but again more pronounced in the dry areas. It also indicates that only a small proportion of respondents consider these impacts as not at all impactful, reinforcing the widespread perception of climate-related stressors on livestock systems.

Figure 2

3.5 Resilience levels of smallholder livestock farming communities

Sensitivity analysis across all study sites confirmed the robustness of resilience outcomes. Comparison of AHP-derived and equal weighting schemes produced only minor differences in resilience scores, with no reversal of site rankings. Similarly, coping models estimated under six-, twelve-, and eighteen-month recovery thresholds retained stable coefficient direction for major explanatory variables, particularly cooperative membership, herd size, and water availability. Adaptation models also remained broadly stable when thresholds were varied across one, two, and three adopted practices, indicating that the principal findings were not driven by a single modeling assumption (Supplementary Tables 5–7).

The resilience analysis revealed consistently low levels of livelihood resilience across the six study sites, with overall indices ranging from 0.172 in Sigatoka to 0.241 in Naitasiri (Table 4). Among the five capital assets, social and physical capitals recorded the lowest scores in nearly all locations, particularly in the dry-zone sites of Sigatoka (social 0.018, physical 0.019), Rakiraki (social 0.019, physical 0.021), Labasa (social 0.021, physical 0.028) and Seaqaqa (social 0.022, physical 0.025), reflecting the lack of farmer cooperatives, extremely weak linkages with veterinary and extension services, poor rural road networks, Lack of fencing, absence of functional local slaughter facilities, and long distances to animal-health offices that require very expensive transport for both the extension officers and the farmers (Table 4). Natural capital was also critically low in these four dry-zone sites (0.089–0.112) owing to severe pasture degradation, prolonged dry weather condition, degraded pasture as result of invasive African Tulip tree, historical land lease expiring and use for sugarcane that left little space for natural pastures or fodder. In contrast, Naitasiri and Tailevu exhibited marginally higher resilience (0.241 and 0.229, respectively), driven by relatively better human capital (more years of experience, larger household labor pools, and closer proximity to milk chilling centers that are facilitated by FCDCL and natural capital (wetter climate, less degraded grazing lands except during very heavy rainfall events, improved pasture management through rotational paddock grazing. Financial capital remained the strongest asset in Naitasiri and Tailevu (0.356 and 0.342) because of dairy cooperatives and proximity to urban markets, and root crop farming that is drought and flood resistant On the other hand, it is lower in Rakiraki and Sigatoka (0.189 and 0.178) where the collapse of sugarcane, the source of extra income for most of the farmers in these areas, and limited alternative markets continue to restrict cash flow and access to credit. Lack of milk collection centers, slaughterhouses and access to the subsidies from the government also resulted in low resilience and these areas. These findings underscore that smallholder livestock farmers in Fiji’s are disproportionately vulnerable and require urgent, balanced interventions across all five capitals to strengthen their adaptive capacity in the face of intensifying climate shocks.

3.6 Livestock farmers’ ranking of adaptation barriers to climate change in the study area

The results in Table 5 indicate that animal diseases are the most critical barrier to adaptation, consistently ranked first across all study sites, with a low mean rank of 1.8 ± 0.1, reflecting strong agreement among farmers on its severity. This is followed by the absence of improved breeds and low prices of livestock products, both of which rank high (mean scores of 2.9 ± 0.2 and 3.0 ± 0.2, respectively), indicating significant constraints related to productivity and market returns. Lack of reliable markets (3.8 ± 0.2) and lack of timely climate information (4.6 ± 0.2) represent moderate barriers, though still important across all sites. Financial constraints are particularly notable, with lack of access to finance ranked fifth (5.1 ± 0.2), highlighting limitations in investment capacity. Less critical, though still relevant, are poor livestock management skills (5.9 ± 0.2) and limited veterinary services (6.4 ± 0.2), which consistently rank lower across sites. The rankings show a high degree of consistency across agroecological zones, suggesting that both biophysical challenges (such as disease) and structural constraints (such as markets, finance, and information) jointly limit farmers’ ability to adapt to climate change. These constraints compound climate vulnerabilities, as warmer/humid conditions favor parasites and pathogens, yet farmers lack the support needed for effective herd health management or adoption of resilient strategies.

3.7 Enablers for adopting climate-smart practices in livestock production in Fiji

Figure 3 illustrates farmers’ responses to key enablers of climate-smart livestock practices across agroecological zones, with a clear dominance of agreement responses across all categories. In the wet zone, strong agreement is particularly evident for access to extension services and education/upskilling forums, indicating the importance of institutional support. In the dry zone, although agreement remains high, there is a relatively higher proportion of neutral and disagreement responses, especially for financial assistance and market access, suggesting greater constraints in these areas. The northern zone shows a similar pattern to the wet zone, with strong agreement for most enablers, particularly education, technical support, and climate information access. Across all zones, financial assistance and market access consistently attract slightly lower agreement compared to extension and training services, indicating that while institutional and knowledge-based support systems are valued, economic and market-related barriers remain significant. The results highlight the critical role of extension services, education, and climate information as key drivers for adopting climate-smart livestock practices across Fiji’s agroecological zones.

Figure 3

3.8 Factor influencing small scale livestock farmers coping and/or adaptation decision and choices

Binary Probit regression results (Table 6) reveal that all five livelihood capitals are highly significant predictors of both short-term coping with climate shocks and longer-term adaptation among Fiji’s smallholder livestock farmers. Membership in farmer’s cooperative or women’s groups (Social capital) exhibited the strongest positive effect, increasing the odds of successful coping by 2.46 times and adaptation by 2.31 times (p < 0.001). Possession of secure fencing and on-farm water harvesting infrastructure (Physical capital) followed, with odds ratios of 2.19–2.33 for coping and 2.07–2.20 for adaptation. Access to formal or informal credit and livestock insurance (Financial capital) raised the probability of both outcomes by 1.88–2.46 times, whereas reliance on off-farm income reduced coping ability, suggesting trade-offs between livestock management and external employment. Human capital variables (farming experience, household labor, and access to veterinary/extension services) and Natural capital indicators (perceived pasture fertility and presence of animal feeds) were also consistently positive and significant (odds ratios 1.19–1.88). Conversely, greater distance to veterinary centers and all-weather roads were acute problems in the dry-zone sites of Sigatoka, Rakiraki, Labasa and Seaqaqa and significantly lowered the odds of both coping and adaptation (odds ratios 0.81–0.90, p < 0.001). The models explained 39–41% of the variation (Pseudo-R² = 0.389–0.412) and correctly classified 85–87% of observations, confirming robust explanatory power. These findings corroborate the descriptive resilience indices and underscore that targeted investments in social organization, basic livestock infrastructure, financial services, and animal feeds systems offer the highest leverage for enhancing climate resilience in Fiji’s smallholder livestock sector. All five livelihood capitals significantly influenced both coping and adaptation (Table 6). Site fixed effects accounted for agro-ecological differences. Zone-stratified analysis (Supplementary Table S8) revealed stronger effects of social and physical capital in the wet zones, with cooperative membership and water harvesting showing the strongest associations in Naitasiri and Tailevu, while these effects were relatively weaker in the drier sites of Sigatoka and Rakiraki.

4 Discussion

The findings of this study show that climate resilience among smallholder livestock farmers in Fiji remains constrained across the study sites despite variation in individual livelihood capital performance. Although certain households demonstrated moderate strength in selected resilience dimensions, particularly financial capital in dairy-producing areas where regular milk income improved short-term liquidity, the composite resilience scores remained low because multiple livelihood capitals were simultaneously weakened. This suggests that resilience in livestock systems cannot be interpreted through isolated strengths in one domain, since adaptive capacity depends on the combined functioning of human, social, natural, physical, and financial assets operating together under climatic stress (Adger, 2000; Scoones, 1998).

Smallholder livestock farmers rely on government support to sustain their operations, including veterinary extension services, pasture development programs (provision of fencing materials), distribution of high-yielding breeding stock, livestock health surveillance, and post-disaster agricultural assistance. These services are available across the study sites, as confirmed by officials from the MoAW Animal and Health Division and the Sigatoka Research Station. However, budgetary constraints limit the extent and effectiveness of service delivery, resulting in inadequate distribution.

This dependency often constrains innovation and self-driven adaptation, leaving the sector highly vulnerable to climate change (Kuruppu & Liverman, 2011). While some segments, such as poultry farming under commercial contracts, have shown greater adaptive capacity due to better access to resources and institutional support, most livestock farmers in Fiji continue to face significant barriers to resilience due to limited knowledge and technical capacity in climate smart livestock practice (Trudinger et al., 2023). Land tenure patterns show that 49.3% of farmers operated on native land, followed by 32.4% on freehold and 18.3% on Crown land (Table 2). These findings suggest that secure land access and financial capacity vary considerably, potentially influencing farmers’ ability and willingness to invest in climate-resilient livestock systems (Díaz Baca et al., 2024). Land tenure increases farmers’ interest in making investments in productive and environmentally conserving technologies, access to formal credit and investments (Singirankabo and Ertsen, 2020).

Being predominantly rainfed, Tailevu and Naitasiri sites are prone to Floods, while Sigatoka, Rakiraki, Labasa and Seaqaqa are prone to both droughts and floods, which farmers reported experiencing more frequently over the past decade, leading to feed shortages and reduced animal productivity. Cyclones and flash floods were major shocks across all provinces, with Cyclone Winston (2016) and subsequent events causing significant livestock losses and infrastructure damage in most of the study sites (Government of Fiji, 2016). During the COVID-19 pandemic, farmers faced supply chain disruptions, including limited access to veterinary supplies, feed, and markets, resulting in forced sales at low prices and reduced herd sizes. In Naitasiri and Tailevu mobility restrictions hindered access to the milk collection centers, increased input costs strained finances. Reports of delayed rainfall onset, increasing heat stress, reduced pasture reliability, and greater livestock disease occurrence suggest that climate signals are being experienced directly through production impacts (Igbal et al., 2026). Perception-based evidence therefore remains valuable because it reveals how climatic variability is interpreted at farm level and how adaptation decisions are triggered under practical livelihood conditions.

Our results from this study indicate a strong awareness of climate change among livestock farmers, with more than two-thirds perceiving clear climatic change rather than short-term variability. The observed climatic trends provide empirical support for farmers’ perceptions of increasing climatic stress across Fiji’s livestock systems. Long-term station analysis showed that rainfall trends were spatially heterogeneous, with significant annual and seasonal declines recorded in some locations, particularly Seaqaqa, while other stations exhibited mixed monthly variability. In contrast, temperature trends were consistently positive across all stations, with statistically significant warming observed annually and seasonally, accompanied by increasing warm spell duration and reduced cool extremes. The stronger consistency in temperature signals compared with rainfall suggests that farmers may be responding more directly to persistent thermal stress and seasonal rainfall irregularity than to long-term rainfall totals alone. The dominant perception of increasing daytime and nighttime temperatures aligns with observed warming trends across the Pacific Islands, where rising temperatures are among the most consistent climate signals (Sagero et al., 2025a, b; Fernández-Duque et al., 2025). The relatively small proportion of farmers reporting uncertainty suggests that experiential knowledge plays a key role in shaping climate perceptions, consistent with findings that farmers’ direct exposure to climate stressors strongly influences their recognition of climate change (Bryan et al., 2009). Similarly, perceptions of declining animal health and reduced pasture availability are consistent with evidence that heat stress reduces feed intake and pasture productivity, while altered rainfall patterns affect forage quality and seasonal feed availability (Thornton and Herrero, 2015). This perception pattern underscores the importance of climate-smart livestock interventions that prioritize animal health management, improved pasture resilience, and disease surveillance, while also strengthening farmer knowledge and adaptive capacity to respond to ongoing climatic change (Cheng et al., 2022).

The observed pattern of low livelihood resilience in Fiji’s livestock systems aligns with broader findings from smallholder studies in climate-vulnerable regions, where weak social organization and limited physical infrastructure consistently constrain adaptive capacity by elevating transaction costs and hindering timely responses to shocks (Diehl and Bose, 2023; Pretty and Ward, 2001; Scoones, 1998). In Fiji, resilience research shows that climate impacts from hazards like tropical cyclones are moderated not only by exposure but also by social, institutional, and behavioral factors shaping household and community responses. Studies by (Gawith et al., 2016; Brown et al., 2018; Gatiso & Greenhalgh., 2025) demonstrated that communities with stronger resilience traits, such as social capital, infrastructure, institutional support, and economic resources, incurred lower losses during cyclone events, particularly through early warning systems and preparedness enabling anticipatory action. These insights strongly support our findings of low resilience in the livestock sector, where infrastructure deficits and social capital limitations are critical factors. Specifically, social capital remained extremely low in areas like Sigatoka, Rakiraki, Labasa, and Seaqaqa, as farmers repeatedly reported weak cooperative organizations and irregular institutional support, limiting collective action, knowledge exchange, and mutual aid during climate stresses. Physical capital scored poorly across sites, with respondents frequently citing poor road access, long distances to veterinary services, and the absence of local livestock infrastructure, which impeded timely inputs, animal health management, and market linkages. Natural capital deficits were more severe in dry-zone systems, where farmers attributed prolonged pasture decline to repeated drought exposure combined with invasive vegetation pressure, exacerbating feed shortages and reducing herd productivity in these already water-limited environments. Overall, these interconnected capital weaknesses underscore the need for targeted interventions to build social networks, upgrade rural infrastructure, and address environmental degradation to enhance resilience in Fiji’s smallholder livestock systems.

Studies from sub-Saharan Africa and South Asia likewise report that deficits in natural capital, including pasture degradation and historical land-use pattern, exacerbate livestock vulnerability under prolonged drought and warming conditions (Thornton and Herrero, 2015). Conversely, areas with stronger human and financial capital, such as better farming experience, cooperative membership, access to markets, and diversified income sources, tend to exhibit higher resilience and adaptive capacity patterns was also observed in Ghana, Ethiopia, and rural China (Ankrah et al., 2023; Ellis, 2000; Kuang et al., 2019). Overall, the Fiji findings align closely with the broader livelihoods and resilience studies, reinforcing the conclusion that balanced development across all five livelihood capitals is essential for strengthening climate resilience in smallholder livestock systems.

The relatively low composite resilience values observed across the study districts reflect the simultaneous constraint of multiple livelihood capitals rather than failure within a single resilience dimension. Although some households demonstrated moderate strength in selected components, particularly financial capital in dairy-producing areas where regular milk income many improved short-term cash flow, overall resilience remained low because several other capitals were consistently weak. Limited participation in farmer cooperatives reduced access to collective support and information exchange, weak veterinary service access constrained livestock health management, degraded pasture conditions reduced feed reliability, and limited credit availability restricted investment in adaptive farm improvements. In addition, inadequate rural infrastructure and weak market connectivity further constrained households’ ability to recover efficiently following climatic disturbances. Because the resilience index was constructed as a composite measure across human, social, natural, physical, and financial capital domains, severe deficits in several capitals simultaneously reduced aggregate resilience scores even where one capital performed relatively better. This pattern is consistent with resilience theory, which recognizes that household adaptive capacity depends on the combined functioning of multiple livelihood assets rather than isolated strength in a single domain (Adger, 2000; Scoones, 1998).

The descriptive resilience indices, ranked barriers, and econometric analysis converge on the same policy implications also highlighted by (Ocampo et al., 2025) in their Philippine case: resilience-building in smallholder livestock systems cannot rely on technological fixes alone. It requires simultaneous, balanced investment in (i) revitalizing farmer cooperatives, (ii) low-cost physical infrastructure (fencing, rainwater harvesting, livestock-specific storage, transport, and market-support infrastructure, including dairy cooling, poultry handling, and animal transport systems), (iii) Accessible subsidized animal feeds, and (iv) mobile veterinary and artificial-insemination services tailored to rural areas or breeders. Without these supports, climate-smart livestock development in Fiji will remain constrained in their capacity to invest in adaptive livestock management.

The probit analysis of coping capacity confirms that short-term recovery following climatic shocks is strongly influenced by access to institutional services and production support. Households with stronger veterinary access and greater extension contact were more likely to restore core livestock functions after disturbance, highlighting the practical importance of service availability during post-shock recovery (Mehar et al., 2016; Panja et al., 2026). In this study, coping was interpreted as functional recovery rather than complete return to pre-disturbance herd size, which better reflects livestock production realities where biological recovery often occurs gradually across several production cycles (Barrett and Constas, 2014; Nkonya et al., 2023). Full herd replacement is frequently constrained by delayed breeding stock access, disease risk, and repeated climatic disturbances, meaning that households often stabilize operations before complete asset restoration is achieved.

5 Limitation of the study

Several limitations should be considered when interpreting the findings of this study. First, the study relied on a cross-sectional design conducted in 2024–2025, capturing resilience at a single point in time rather than tracking dynamic changes before and after major shocks. Second, while the sample of 270 households (45 per site) was statistically representative at the site level, it remains modest relative to the total population of smallholder livestock farmers in Fiji, and results should be interpreted primarily at the provincial rather than national scale. Third, the resilience index was constructed as a composite measure derived from five livelihood capital domains, which necessarily simplifies complex household realities into aggregated scores. Although this approach allows systematic comparison across districts and livestock systems, moderate strength in one livelihood capital cannot fully offset severe weakness in others under the aggregated structure, meaning that some localized adaptive strengths may be partially masked in the final composite score. This is particularly relevant in livestock systems where households may perform relatively well in financial capital while remaining highly constrained in natural, social, or physical capital. Fourth, coping capacity was measured as short-term functional recovery rather than complete restoration of pre-shock herd size or income. This definition was adopted because livestock recovery often occurs gradually and may remain constrained by repeated climatic shocks, disease outbreaks, delayed breeding stock access, and market disruptions. While this approach better reflects livestock resilience dynamics, differences in household interpretation of recovery may still introduce variation in reported coping outcomes. Fifth, the cross-sectional design captures resilience conditions at one point in time and therefore cannot fully represent how household resilience evolves under repeated climatic exposure across multiple years. Longitudinal assessment would strengthen future understanding of how resilience trajectories change under successive shocks and policy intervention. Finally, the study did not disaggregate resilience by livestock enterprise (e.g., beef vs. dairy vs. goat systems or poultry and Pig) or by gender of household head or ethnicity (Indo-Fijian vs iTaukei Fijian, limiting insights into enterprise-specific or ethnicity and gendered vulnerabilities. Despite these constraints, the mixed-methods approach and multi-site coverage provide robust, context-specific evidence that usefully identifies priority intervention areas for enhancing the adaptive capacity of Fiji’s smallholder livestock sector.

6 Conclusion

This study provides comprehensive empirical evidence on the vulnerability and resilience of smallholder livestock farming systems across six sites in Fiji under a changing climate. This study demonstrates that despite relatively high farming experience and strong awareness of climate change, climate resilience among smallholder livestock households in Fiji remains constrained primarily because multiple livelihood capitals are weakened simultaneously, even where selected household capacities show moderate strength. Resilience indices well below the mid-range threshold reflect entrenched structural constraints that limit farmers’ ability to cope with and adapt to increasing climatic stressors. It confirms that financial stability alone is insufficient to produce strong overall resilience when veterinary access, social organization, pasture condition, infrastructure, and institutional connectivity remain limited. The findings therefore reinforce the importance of treating livestock resilience as a multidimensional livelihood outcome rather than a single production or income response.

Farmers overwhelmingly perceive climate change through rising daytime and nighttime temperatures, increasing disease pressure, declining pasture availability, and deteriorating animal health. These perceptions are consistent with observed climate trends in the Pacific and are reinforced by farmers lived experiences of droughts, floods, cyclones, and compounding shocks such as BTB outbreak, which is a persistent endemic issue causing significant economic losses in its livestock industry. The dominance of rainfed production systems, degraded grazing lands, expensive animal feeds and limited access to veterinary and extension services further exacerbate climate sensitivity, particularly in dry-zone locations such as Sigatoka, Rakiraki, Labasa, and Seaqaqa.

The livelihood resilience assessment reveals pronounced deficits in social and physical capital across most sites, highlighting weak farmer organization, poor infrastructure, limited market access, and inadequate animal-health services as key drivers of vulnerability. In contrast, Naitasiri and Tailevu exhibit relatively higher though still low resilience, largely due to stronger human and financial capital linked to dairy cooperatives, better market integration, and more favorable agroecological conditions for root crop production. These spatial contrasts underscore the importance of institutional support, market connectivity, and diversified livelihood opportunities in shaping adaptive capacity.

The study further shows that coping and adaptation are shaped by different but interconnected drivers. Short-term coping capacity depends strongly on institutional support, veterinary access, and infrastructure continuity following climatic shocks, while longer-term adaptation is influenced by education, extension contact, and access to financial resources. Households most frequently adopted practical low-cost responses such as feed conservation and water management, indicating that adaptation remains largely incremental rather than transformative under current resource constraints.

The econometric analysis reinforces these findings, showing that all five livelihood capitals significantly influence both short-term coping and long-term adaptation decisions. Social capital particularly cooperative membership and strong local networks, emerge as the most influential determinant, followed by access to basic physical infrastructure, financial services, veterinary support, and natural feed resources. Conversely, remoteness from veterinary posts and all-weather roads substantially reduces adaptive capacity, especially in dry-zone communities. Together, the descriptive and regression results confirm that resilience building in Fiji’s livestock sector requires balanced, multi-dimensional investment.

From a policy perspective, improving livestock resilience in Fiji requires coordinated strengthening of veterinary outreach, farmer organizations, pasture management, adaptive finance, and rural market infrastructure. Interventions targeting only one capital domain are unlikely to produce sustained resilience gains because resilience emerges through the interaction of multiple livelihood assets. Future research would benefit from longitudinal monitoring of resilience trajectories and stronger integration of observed climatic indices with livestock production responses to better capture dynamic adaptation under continuing climate variability.

References

• 1

  AdgerW. N. (2000). Social and ecological resilience: Are they related? Prog. Hum. Geogr.24, 347–364. doi: 10.1191/030913200701540465. PMID:

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 2

  AdimassuZ.KesslerA. (2016). Factors affecting farmers’ coping and adaptation strategies to perceived trends of declining rainfall and crop productivity in the central Rift valley of Ethiopia. Environ. Syst. Res.5, 13. doi: 10.1186/s40068-016-0065-2. PMID:

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 3

  AnkrahD. A.AnumR.AnagloJ. N.BoatengS. D. (2023). Influence of sustainable livelihood capital on climate variability adaptation strategies. Environ. Sustainability Indic.18, 100233. doi: 10.1016/j.indic.2023.100233. PMID:

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 4

  BarrettC. B.ConstasM. A. (2014). Toward a theory of resilience for international development applications. Proc. Natl. Acad. Sci.111, 14625–14630. doi: 10.1073/pnas.1320880111. PMID:

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 5

  BaylieM. M. (2025). Gender differential impacts of climate change on smallholders’ farm income. Discover Sustainability6, 1065. doi: 10.1007/s43621-025-01984-2. PMID:

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 6

  BilottoF.HarrisonM. T.VibartR.MackayA.Christie-WhiteheadK. M.FerreiraC. S.CottrellR. S.ForsterD.ChangJ. (2024). Towards resilient, inclusive, sustainable livestock farming systems. Trends in Food Science & Technology152, 104668. doi: 10.1016/j.tifs.2024.104668

  • CrossRef
  • Google Scholar

• 7

  BrouilletA.SultanB. (2023). Livestock exposure to future cumulated climate-related stressors in West Africa. Sci. Rep.13, 2698. doi: 10.1038/s41598-022-22544-y. PMID:

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 8

  BrownP.DaigneaultA. J.TjernströmE.ZouW. (2018). Natural disasters, social protection, and risk perceptions. World Dev.104, 310–325. doi: 10.1016/j.worlddev.2017.12.002. PMID:

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 9

  BryanE.DeressaT. T.GbetibouoG. A.RinglerC. (2009). Adaptation to climate change in Ethiopia and South Africa: Options and constraints. Environ. Sci. Policy12, 413–426. doi: 10.1016/j.envsci.2008.11.002. PMID:

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 10

  Bundi MagiriR.SageroP.DanmaigoroA.RashidR.MocevakacaW.SinghS.et al. (2023). Impact of climate change on the dairy production in Fiji and the Pacific Island Countries and Territories: an insight for adaptation planning. In Global Warming - A Concerning Component of Climate Change [Working Title]. (London: IntechOpen). doi: 10.5772/intechopen.1002052

  • CrossRef
  • Google Scholar

• 11

  ChengM.McCarlB.FeiC. (2022). Climate change and livestock production: A literature review. Atmosphere13, 140. doi: 10.3390/atmos13010140. PMID:

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 12

  CottrellR. S.NashK. L.HalpernB. S.RemenyiT. A.CorneyS. P.FlemingA.et al. (2019). Food production shocks across land and sea. Nat. Sustainability2, 130–137. doi: 10.1038/s41893-018-0210-1. PMID:

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 13

  CullenB. R.AyreM.ReicheltN.NettleR. A.HaymanG.ArmstrongD. P.et al. (2021). Climate change adaptation for livestock production in southern Australia: Transdisciplinary approaches for integrated solutions. Animal Frontiers11 (5), 30–39. doi: 10.1093/af/vfab046

  • CrossRef
  • Google Scholar

• 14

  DasM. K. (2021). Determinants of adaptation strategies of agricultural farmers to climate change vulnerability in Odisha. Asian J. Agric. Extension Econ Sociology39(9), 167–179. doi: 10.9734/ajaees/2021/v39i930654. PMID:

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 15

  Department for International Development. (1999). “ Sustainable livelihoods guidance sheets,” in Department for International Development. (London, United Kingdom). Available online at: https://www.livelihoodscentre.org/documents/114097690/114438878/Sustainable+livelihoods+guidance+sheets.pdf (Accessed December 29, 2025).

  • Google Scholar

• 16

  DeoR. C. (2011). On meteorological droughts in tropical Pacific Islands: Time-series analysis of observed rainfall using Fiji as a case study. Meteorological Applications18 (2), 171–180. doi: 10.1002/met.216

  • CrossRef
  • Google Scholar

• 17

  DeressaT. T.HassanR. M.RinglerC.AlemuT.YesufM. (2009). Determinants of farmers’ choice of adaptation methods to climate change in Ethiopia: Multinomial choice analysis. Global Environ. Change19, 248–255. doi: 10.1016/j.gloenvcha.2009.01.002. PMID:

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 18

  Díaz BacaM. F.Moreno LermaL.Triana ÁngelN.BurkartS. (2024). The relationships between land tenure, cattle production, and climate change – A systematic literature review. Land Use Policy141, 107169. doi: 10.1016/j.landusepol.2024.107169. PMID:

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 19

  DiehlJ. A.BoseM. (2023). A sustainable livelihoods approach to measuring mobilization of resources through social networks among vulnerable populations: A case study of Delhi farmers. Soc. Sci. Humanities Open8, 100689. doi: 10.1016/j.ssaho.2023.100689. PMID:

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 20

  Di FalcoS.VeronesiM.YesufM. (2011). Does adaptation to climate change provide food security? A micro-perspective from Ethiopia. Am. J. Agric. Econ.93, 829–846. doi: 10.1093/ajae/aar006

  • CrossRef
  • Google Scholar

• 21

  EllisF. (2000). Rural Livelihoods and Diversity in Developing Countries ( Oxford, United Kingdom: Oxford University PressOxford). doi: 10.1093/oso/9780198296959.001.0001

  • CrossRef
  • Google Scholar

• 22

  FAO, G (2019). The role of the dairy sector in a low-carbon future.

  • Google Scholar

• 23

  FarrellyT.Nabobo-BabaU. (2014). Talanoa as empathic apprenticeship. Asia Pac. Viewp55, 319–330. doi: 10.1111/apv.12060. PMID:

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 24

  Fernández-DuqueB.KumarR.KenawyA. E.Vicente-SerranoS. M. (2025). Changes in daily precipitation extremes over the Fiji Islands, (1905–2021). Stochastic Environ. Res. Risk Assess.39, 581–598. doi: 10.1007/s00477-024-02883-0. PMID:

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 25

  GalhenaD. H.FreedR.MarediaK. M. (2013). Home gardens: A promising approach to enhance household food security and wellbeing. Agric. Food. Secur.2, 8. doi: 10.1186/2048-7010-2-8. PMID:

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 26

  GatisoT.GreenhalghS. (2025). Socioeconomic resilience of local communities in the face of climate change-induced hazards: The role of social capital. Clim. Resilience Sustainability4, e70012. doi: 10.1002/cli2.70012. PMID:

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 27

  GautamN. P.RautN. K.ChhetriB. B. K.RautN.RashidM. H. U.MaX.et al. (2021). Determinants of poverty, self-reported shocks, and coping strategies: Evidence from rural Nepal. Sustainability13, 1790. doi: 10.3390/su13041790. PMID:

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 28

  GawithD.DaigneaultA. J.BrownP. (2016). Does community resilience mitigate loss and damage from climate-related disasters? Evidence based on survey data. J. Environ. Plann. Manage.59, 2187–2202. doi: 10.1080/09640568.2015.1126241. PMID:

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 29

  Government of Fiji. (2016). Post-disaster needs assessment: Tropical Cyclone Winston, February 20, 2016.Suva, Fiji: Government of Fiji. https://www.health.gov.fj/wpcontent/uploads/2016/03/Post-Disaster-Needs-Assessments-Tropical-Cyclone-Winstin-Fiji-2016.pdf (Accessed November 3, 2025).

  • Google Scholar

• 30

  GreeneW. H. (2018). Econometric Analysis ( London, United Kingdom: Pearson Education).

  • Google Scholar

• 31

  GoddeC. M.Mason-D’CrozD.MayberryD. E.ThorntonP. K.HerreroM. (2021). Impacts of climate change on the livestock food supply chain; a review of the evidence. Global Food Security28, 100488. doi: 10.1016/j.gfs.2020.100488

  • CrossRef
  • Google Scholar

• 32

  HahnM. B.RiedererA. M.FosterS. O. (2009). The livelihood vulnerability index: A pragmatic approach to assessing risks from climate variability and change—a case study in Mozambique. Global Environ. Change19, 74–88. doi: 10.1016/j.gloenvcha.2008.11.002. PMID:

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 33

  HerreroM.GraceD.NjukiJ.JohnsonN.EnahoroD.SilvestriS.et al. (2013). The roles of livestock in developing countries. Animal7, 3–18. doi: 10.1017/S1751731112001954. PMID:

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 34

  High Level Panel of Experts on Food Security and Nutrition (HLPE) (2013). Investing in smallholder agriculture for food security (Rome, Italy: FAO). Available online at: https://www.fao.org/3/i2953e/i2953e.pdf (Accessed November10, 2025).

  • Google Scholar

• 35

  HuC.DongJ. (2023). Measuring livelihood resilience of farmers and diagnosing obstacle factors under the impact of COVID-19 in Jiangsu Province, China. Front. Sustain. Food Syst.7, 1250564. doi: 10.3389/fsufs.2023.1250564

  • CrossRef
  • Google Scholar

• 36

  IgbalM. R.SageroP. O.MagiriR. B. (2026). Livestock farmers’ perception on effect of climate change on smallholder dairy farming in Fiji. Front. Clim.7, 1654274. doi: 10.3389/fclim.2025.1654274

  • CrossRef
  • Google Scholar

• 37

  IPCC (2023). Climate Change 2022 – Impacts, Adaptation and Vulnerability: Working Group II Contribution to the Sixth Assessment Report of the Intergovernmental Panel on Climate Change ( Cambridge, United Kingdom: Cambridge University Press). doi: 10.1017/9781009325844

  • CrossRef
  • Google Scholar

• 38

  JhaM. K.DevM. (2024). “ Impacts of climate change,” in Smart Internet of Things for Environment and Healthcare, vol. 1165 . Eds. AzrourM.MabroukiJ.AlabdulatifA.GuezzazA.AmounasF. ( Springer, Cham: Springer Nature Switzerland), 139–159. doi: 10.1007/978-3-031-70102-3_10

  • CrossRef
  • Google Scholar

• 39

  JiangH.ZipserE. J. (2010). Contribution of tropical cyclones to the global precipitation from eight seasons of TRMM data: Regional, seasonal, and interannual variations. Journal of Climate23 (6), 1526–1543. doi: 10.1175/2009JCLI3303.1

  • CrossRef
  • Google Scholar

• 40

  KasperczykN.KnickelK. (2006). The Analytic Hierarchy Process (AHP). Direct PDF. Available online at: https://www.iied.org/sites/default/files/pdfs/2022-02/20781G.pdf (Accessed 14 December, 2025).

  • Google Scholar

• 41

  KuangF.JinJ.HeR.WanX.NingJ. (2019). Influence of livelihood capital on adaptation strategies: Evidence from rural households in Wushen Banner, China. Land Use Policy89, 104228. doi: 10.1016/j.landusepol.2019.104228. PMID:

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 42

  KuleshovY.McGreeS.JonesD.CharlesA.CottrillA.PrakashB.et al. (2014). Extreme weather and climate events and their impacts on island countries in the western Pacific: Cyclones, floods and droughts. Atmospheric and Climate Sciences4 (5), 803–818. doi: 10.4236/acs.2014.45071

  • CrossRef
  • Google Scholar

• 43

  KumarV. V.DeoR. C.RamachandranV. (2006). Total rain accumulation and rain-rate analysis for small tropical Pacific islands: A case study of Suva, Fiji. Atmospheric Science Letters7 (3), 82–88. doi: 10.1002/asl.131

  • CrossRef
  • Google Scholar

• 44

  KumarS.ReddyM. (2008). Fiji’s dairy industry: A cost and profitability analysis. Pacific Economic Bulletin23 (2), 27–39.

  • Google Scholar

• 45

  KumarR.StephensM.WeirT. (2014). Rainfall trends in Fiji. International Journal of Climatology34 (5), 1501–1510. doi: 10.1002/joc.3779

  • CrossRef
  • Google Scholar

• 46

  KuruppuN.LivermanD. (2011). Mental preparation for climate adaptation: The role of cognition and culture in enhancing adaptive capacity of water management in Kiribati. Global Environmental Change21 (2), 657–669. doi: 10.1016/j.gloenvcha.2010.12.002

  • CrossRef
  • Google Scholar

• 47

  KwayuE. J.PaavolaJ.SalluS. M. (2017). The livelihood impacts of the Equitable Payments for Watershed Services (EPWS) Program in Morogoro, Tanzania. Environ. Dev. Econ22, 328–349. doi: 10.1017/S1355770X17000067. PMID:

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 48

  LataS.NunnP. D. (2012). Misperceptions of climate-change risk as barriers to climate-change adaptation: A case study from the Rewa Delta, Fiji. Clim. Change. 110, 169–186. doi: 10.1007/S10584-011-0062-4. PMID:

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 49

  MagiriR.MocevakacaW.OkelloW.FisherA. D. (2024). Moving towards sustainable livestock development in the Pacific Island countries. doi: 10.26021/15187

  • CrossRef
  • Google Scholar

• 50

  MannH. B. (1945). Nonparametric tests against trend. Econometrica13, 245–259. doi: 10.2307/1907187

  • CrossRef
  • Google Scholar

• 51

  McIverL.KimR.WoodwardA.HalesS.SpickettJ.KatscherianD.et al. (2016). Health impacts of climate change in Pacific Island countries: A regional assessment of vulnerabilities and adaptation priorities. Environ. Health Perspect.124, 1707–1714. doi: 10.1289/ehp.1509756. PMID:

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 52

  MeharM.MittalS.PrasadN. (2016). Farmers coping strategies for climate shock: Is it differentiated by gender? J. Rural Stud.44, 123–131. doi: 10.1016/j.jrurstud.2016.01.001. PMID:

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 53

  MeuwissenM. P. M.FeindtP. H.SpiegelA.TermeerC. J. A. M.MathijsE.MeyY. D.et al. (2019). A framework to assess the resilience of farming systems. Agric. Syst.176, 102656. doi: 10.1016/j.agsy.2019.102656. PMID:

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 54

  Ministry of Agriculture Fiji. (2020). 2020 key statistics on Fiji agriculture sector.Suva, Fiji: Ministry of Agriculture. https://www.agriculture.gov.fj/documents/stats/2020KEYSTATIS TICS.pdf (Accessed November 10, 2025).

  • Google Scholar

• 55

  NkonyaE.KatoE.MsimangaM.NyathiN. (2023). Climate shock response and resilience of smallholder farmers in the drylands of south-eastern Zimbabwe. Front. Clim.5, Article 890465. doi: 10.3389/fclim.2023.890465

  • CrossRef
  • Google Scholar

• 56

  OcampoM. T. N. P.GarciaJ. N. M.LandichoL. D.CabahugR. D.BarcellanoE. V.UrmenetacJ. S. (2025). Building resiliency of smallholder farmers in upland farming communities in Isabela Province, Philippines. APN Sci. Bull.15, 23–34. doi: 10.30852/sb.2025.2877

  • CrossRef
  • Google Scholar

• 57

  PanjaA.GaraiS.MaitiS.GoswamiR.ZadeS.VeldandiA.et al. (2026). Exploring determinants of climate change adaptation by smallholder livestock farmers in coastal West Bengal, India using a double hurdle econometric approach. Scientific Reports16, 2946. doi: 10.1038/s41598-025-32890-2

  • CrossRef
  • Google Scholar

• 58

  PrettyJ.WardH. (2001). Social capital and the environment. World Dev.29, 209–227. doi: 10.1016/S0305-750X(00)00098-X

  • CrossRef
  • Google Scholar

• 59

  RyanT. P. (2013). Sample size determination and power. John Wiley & Sons. doi: 10.1002/9781118439241

  • CrossRef
  • Google Scholar

• 60

  SaatyT. L. (1980). The analytic hierarchy process: Planning, priority setting, resource allocation. McGraw-Hill.

  • Google Scholar

• 61

  SageroP. O.PratapA.MagiriR.OngomaV.OkelloP.. (2024). Validation of ERA5 rainfall data over the South Pacific Region: Case study of Fiji Islands. Meteorology and Atmospheric Physics136 (4), 28. doi: 10.1007/s00703-024-01025-z

  • CrossRef
  • Google Scholar

• 62

  SageroP. O.KumarN.MagiriR.NakamuraN. (2025b). Observed changes in rainfall and temperature extreme in Fiji. Aust. Geogr.56, 241–263. doi: 10.1080/00049182.2025.2481714. PMID:

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 63

  SageroP.KumarN.SianK. T. C. L. K.OngomaV. (2025a). Variability of temperature and heatwaves in the South Pacific Islands region under global warming. Theor. Appl. Climatol.156, 644. doi: 10.1007/s00704-025-05867-4. PMID:

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 64

  ScoonesI. (1998). “ Sustainable rural livelihoods: A framework for analysis (IDS Working Paper),” in Institute of Development Studies. (Brighton, United Kingdom). Available online at: https://www.ids.ac.uk/publications/sustainable-rural-livelihoods-a-framework-for-analysis/. (Accessed November 10, 2025).

  • Google Scholar

• 65

  SenP. K. (1968). Estimates of the regression coefficient based on Kendall's tau. J. Am. Stat. Assoc.63, 1379–1389. doi: 10.1080/01621459.1968.10480934. PMID:

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 66

  SinghS.BakareA. G.IjiP. A.ZindoveT. J. (2023). Beef cattle production systems and constraints among subsistence farmers across the Fijian windward–leeward divide. Outlook Agric.52, 140–152. doi: 10.1177/00307270231162116

  • CrossRef
  • Google Scholar

• 67

  SingirankaboU. A.ErtsenM. W. (2020). Relations between land tenure security and agricultural productivity: Exploring the effect of land registration. Land9, 138. doi: 10.3390/land9050138. PMID:

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 68

  SharmaK. K.Verdon-KiddD. C.MageeA. D. (2021). A decision tree approach to identify predictors of extreme rainfall events – A case study for the Fiji Islands. Weather and Climate Extremes33, 100346. doi: 10.1016/j.wace.2021.100346

  • CrossRef
  • Google Scholar

• 69

  SunY.WangY.HuangC.TanR.CaiJ. (2023). Measuring farmers’ sustainable livelihood resilience in the context of poverty alleviation: A case study from Fugong County, China. Humanities Soc. Sci. Commun.10, 75. doi: 10.1057/s41599-023-01575-4. PMID:

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 70

  TannerT.LewisD.WrathallD.BronenR.Cradock-HenryN.HuqS.et al. (2015). Livelihood resilience in the face of climate change. Nature Climate Change5 (1), 23–26. doi: 10.1038/nclimate2431

  • CrossRef
  • Google Scholar

• 71

  ThorntonP. K.HerreroM. (2015). Adapting to climate change in the mixed crop and livestock farming systems in sub-Saharan Africa. Nat. Clim. Change5, 830–836. doi: 10.1038/nclimate2754. PMID:

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 72

  ThorntonP. K.Van De SteegJ.NotenbaertA.HerreroM. (2009). The impacts of climate change on livestock and livestock systems in developing countries: A review of what we know and what we need to know. Agric. Syst.101, 113–127. doi: 10.1016/j.agsy.2009.05.002. PMID:

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 73

  TrudingerH.CrimpS.FriedmanR. S. (2023). Food systems in the face of climate change: Reviewing the state of research in South Pacific Islands. Reg. Environ. Change23, 45. doi: 10.1007/s10113-023-02040-3. PMID:

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 74

  United Nations Development Programme. (2017). Guidance note: Application of the sustainable livelihoods framework in development projects. (New York, NY: UNDP). Available online at: https://www.undp.org/latin-america/publications/guidance-noteapplication- sustainable-livelihoods-framework-development-projects (Acessed January 20, 2026).

  • Google Scholar

• 75

  WanderaL. N.MachariaI.NgareL. (2024). Determinants of coping strategies among agropastoralists in Kitui and Isiolo counties, Kenya. Discover Agric.2, 83. doi: 10.1007/s44279-024-00103-5. PMID:

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 76

  WangY.ZhangQ.LiQ.WangJ.SannigrahiS.BilsborrowR.et al. (2021). Role of social networks in building household livelihood resilience under payments for ecosystem services programs in a poor rural community in China. J. Rural Stud.86, 208–225. doi: 10.1016/j.jrurstud.2021.05.017. PMID:

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 77

  WeirT.KumarR.NgariA. (2021). Interdecadal modulation of the effect of ENSO on precipitation in the southwestern Pacific. Journal of Southern Hemisphere Earth Systems Science71 (1), 53–65. doi: 10.1071/ES19053

  • CrossRef
  • Google Scholar

• 78

  WooldridgeJ. M. (2010). Econometric analysis of cross section and panel data (2nd ed.). ( MIT Press).

  • Google Scholar

• 79

  World Bank (2022). Moving towards sustainability: The livestock sector and the World Bank (or a general FAO livestock overview). Available online at: https://www.worldbank.org/en/topic/agriculture/brief/moving-towards-sustainability-the-livestock-sector-and-the-world-bank (Accessed December, 10, 2025).

  • Google Scholar

• 80

  ZindoveT. J.BakareA. G.IjiP. A. (2022). Indigenous chicken production in Fiji Islands: Knowledge, constraints and opportunities. Animal Bioscience35 (5), 778–788. doi: 10.5713/ab.21.0309

  • CrossRef
  • Google Scholar