Our planet is changing rapidly. Climate change is more punctuated and severe than at any time in recorded history (Foley et al. 2005, Steffen et al. 2005). Climate change affects the lives of people across the globe, who often respond by adapting their livelihoods (Osbahr et al. 2010). Although adaptation is a common response to changing climates, not everyone adapts successfully, leading to a range of potential negative consequences (Adger et al. 2005, Vincent 2007, Jones and Boyd 2011, Evans et al. 2016). Understanding how responses to climate change vary is critical to minimizing its potential negative impacts.
Across contexts, successful adaptation to climate change is often influenced by the availability and quality of capital (Scoones 1998, Bebbington 1999, Pretty and Ward 2001). Capital is often classified into five broad types: physical, financial, social, natural, and human. Physical capital refers to access to and quality of local infrastructure and physical assets (de Sherbinin et al. 2008; Vincent and Cull, unpublished manuscript: http://kulima.com/wordpress/wp-content/uploads/2010/12/PEGNet-conference-2010-_Vincent-and-Cull_-climate-and-development-panel-160810.pdf). Physical capital is positively associated with food security and agricultural adaptation (Mbukwa 2014). Financial capital includes income, financial assets, and access to money (Abel et al. 2006, de Sherbinin et al. 2008). Households with greater financial capital are often less risk averse and have greater access to information and opportunities, enabling adaptation (Franzel 1999, Deressa et al. 2009). Social capital refers to integration into broad social groups, networks, and organizations, and access to resources via these groups (Putnam et al. 1993, Bebbington 1997). Social capital supports adaptation via collective action, resource networks, and establishment of support systems during times of stress (Adger 2003, Deressa et al. 2009). Natural capital encompasses access to and quality of natural resources and the services they provide, such as farmland, water, and forests (Costanza et al. 1997, Nawrotzki et al. 2012). High quality natural capital supports resilience in the face of environmental stressors and is often associated with more social, financial, and human capital (McDowell and Hess 2012, Nawrotzki et al. 2012). Finally, human capital is the skills, knowledge, experience, health, and education an individual has (Scoones 1998, Nawrotzki et al. 2012). Human capital supports the competence and capabilities necessary for adaptation and minimizes barriers to many adaptation options (Paavola 2008, Cassidy and Barnes 2012). The five types of capital provide the resources, opportunities, and skills necessary to adapt to changing conditions. They are so strongly linked to adaptation that they are often used as a measure of adaptive capacity, i.e., the ability to adapt (Yohe and Tol 2002, Hinkel 2011, Bryan et al. 2015).
All five types of capital are important for successful adaptation, but under certain circumstances, one type of capital may be more important than the others in determining adaptation success (Bebbington 1997, Katz 2000, Adger 2003, Nhuan et al. 2016). One type of capital may increase access to other types of capital or play a disproportionate role in adaptation success. Social capital is often associated with more diversified livelihoods, which are better able to withstand stresses and allow greater access to other types of capital (Cassidy and Barnes 2012). Further, adaptations created through the use of social capital are often more feasible and successful than adaptations driven by other types of capital (Tibesigwa et al. 2014). Along with social capital, natural capital also heavily influences adaptation. Declines in both quality and quantity of natural resources (soil erosion, water pollution, forest fragmentation, access to land, etc.) limit adaptation options and increase negative impacts associated with climatic variability (McDowell and Hess 2012). Human capital influences livelihood diversification, access to natural capital, access to social networks, and total financial capital (Gunderson and Holling 2002; A. Masae, unpublished manuscript). The role that each type of capital plays in adaptation is highly context dependent (Adger et al. 2005, Smit and Wandel 2006). Accordingly, it is necessary to investigate relationships among types of capital and the strength of each type in different contexts to improve our understanding of climate adaptation (Kiem and Austin 2013). Despite the large body of research on climate change adaptation, there is still a need to identify limits to adaptation and the role of capital to facilitate effective adaptation strategies (Esteve et al. 2018).
Here, we examine adaptive capacity and household adaptation to climate stress in situ. We measure adaptation at the household level because decisions made at this scale are measurable and directly influence local resource use, human health, and survival (Ellis 1998, Pelling 2003). Our goal was to evaluate relationships between the different types of capital and identify which types of capital most strongly influence climate adaptation. We investigated these relationships in communities in eastern rural Eswatini, an area that experienced a severe drought between 2015 and 2016. We hypothesized that social and natural capital would most strongly influence adaptation because social capital influences access to resources and opportunities related to adaptation, and natural capital affects the sensitivity to environmental change (Katz 2000, Sseguya 2009, Paul et al. 2016).
Drought has disrupted livelihoods in Eswatini for centuries and is expected to worsen as climate change continues (Mabuza et al. 2009, Moore and Daday 2010, Oseni and Masarirambi 2011). We conducted this research from 2015 to 2016, when Eswatini experienced one of the most extreme droughts in recent memory (Mabuza 2016). Based on initial information from focus groups, we created a household survey to collect data on household demographics, measures of the five types of capital, and adaptation. We translated the survey and focus group instruments from English to SiSwati and then back-translated them to English to ensure that they were understandable and locally appropriate. We conducted surveys and focus groups in SiSwati with a translator present who was fluent in English and SiSwati.
Eswatini is a small subtropical country with a population of approximately 1.4 million and an area of 17,565 km² (Goudie and Price Williams 1983; Central Intelligence Agency World Factbook: https://www.cia.gov/library/publications/the-world-factbook/geos/wz.html). It varies in altitude from 160 to 1860 m above sea level, has a mean annual rainfall between 500 and 1500 mm, and ranges in average temperature from 16 to 22°C (Goudie and Price Williams 1983). Eswatini has a wet and a dry season and is characterized by thunderstorms in the summer and decreased precipitation in the winter. Swazi homesteads are traditionally made up of a group of buildings (devoted to sleeping, cooking, storage, etc.) and land dedicated to farming (Russell 1983). Approximately 70–80% of the population is either directly engaged in or associated with homestead-based farming (Boudreau 2010). Communal grazing land is traditionally used for cattle. This dependence on crops and cattle is embedded in the social, cultural, and economic landscapes of Eswatini and is perceived as critical to the country’s success and development (Forster and Nsibande 2000).
We conducted our research in the Lubombo district of eastern Eswatini (Fig. 1). This district comprises a mosaic of intensive agriculture, protected areas, subsistence cropland, grazing land, and human settlements. Sugarcane plantations are the main commercial agriculture, whereas maize is the primary subsistence crop (Department of Water Affairs and Forestry, Republic of South Africa 2002). The Lubombo district contains two main geographic regions: the Lowveld and Lubombo Mountains. The Lowveld receives the least rain in Eswatini, at 500–600 mm/yr, and has a combination of basaltic clay soils in the more eastern regions and sandstone shale soils in the west (Cleverly 1979, Rached et al. 1996, Manyatsi et al. 2015). The Lubombo Mountains range in altitude from 250–600 m, receive 600–900 mm/yr of rainfall, and have varied soils (Cleverly 1979, Rached et al. 1996, Manyatsi et al. 2015). We conducted research in six communities: three in the Lowveld (Mpaka, Matsetsa, and Lonhlupekho) and three in the Lubombo Mountains (Shewula, Mhlumeni, and Luketseni; Fig. 1). We selected these communities by working with local informants to reflect the variation in infrastructure, agro-ecology, and population size in the region. Within each of the communities, we conducted surveys in 50 homesteads. We first digitized the boundaries of each of the six communities based on information from local informants. We then generated 50 random geographic points and selected the homestead nearest to each point to be surveyed.
We conducted focus groups to understand the history and current challenges faced by each community (Appendix 1, section A1.1). We conducted at least two focus groups per community, with an average of seven participants (5–20) in each group. These groups included community members who had knowledge of past community events and were selected with the help of local informants. All focus group members were between 20 and 80 years old. We separated male and female participants into different groups. We asked open-ended questions to prompt broad discussion. Focus groups typically lasted for 1–2 h.
In the first section of the focus group, we asked participants about major changes in the community. Sources of major changes included infrastructure, disease, livelihood strategies, leadership, access to education, natural resources, healthcare, and jobs. We also asked about adaptation options, past experiences with drought, and major stresses and shocks experienced within the community. Further, we asked open-ended questions to gather general information on daily activities, resource collection, employment options, social interactions, and health. We also used focus groups to identify the specific set of adaptations most common in the region in response to drought.
Each household survey took approximately 1 h and included six thematic sections: (1) demographics and human capital; (2) natural capital, farming, and food resources; (3) financial capital: (4) physical capital: (5) social capital; and (6) adaptations. We conducted the survey with the primary household decision maker when possible. If the primary decision maker was unavailable, we conducted the survey with the oldest household member present.
First, we asked questions about household demographics and human capital (Appendix 1, section A1.2). We collected data on age, highest level of education, employment status (including self-employment), marital status, and time at the homestead for those living at the homestead. We used this information to quantify dependency ratios (ratio of children under 16 and adults over 65 to the total household population aged 16 to 65), household jobs and businesses, average education, and time in the community.
For the second section of our survey, we asked questions about natural capital, food sources, and crops (Appendix 1, section A1.2). To quantify natural capital, we asked questions about access to and use of natural resources (de Sherbinin et al. 2008, Below et al. 2012, Cassidy and Barnes 2012, Tibesigwa et al. 2014). For example, to determine the use of wild food, we asked, “What proportion of your household’s diet comes from wild foods?” (We provided examples of common wild foods). Additionally, we asked, “How many kilometers do you travel to collect wild foods?” To determine farming behavior, we asked the respondent to describe the most recent farming activities, including time, crops planted, proportion of fields used, and yield. We also asked the respondent to compare their most recent farming activities to past farming activities. We also asked about ownership of livestock. We asked about the number and type of livestock owned and if any had been sold or died recently.
In the third section, we asked question pertaining to financial capital (Appendix 1, section A1.2). Recognizing the variety of livelihood strategies available, we asked questions accounting for multiple sources of income. We asked about total income from salaried employment, small businesses, and exchange of goods and services. Additionally, we asked for descriptions of small business activities, costs of materials, and how often sales or exchanges are made, on average. Further, because people are often hesitant to report total income (Moore et al. 2000, United Nations Department of Economic and Social Affairs 2005, Meyer et al. 2015), we asked questions that could serve as a proxy for relative income. We asked, “On average, how much money does your household spend on groceries each month?” and “On average, how much money does your household spend on school fees each year?” There are a range of costs associated with schooling beyond the seventh grade in Eswatini. Additionally, some schools charge more than others, and many households are unable to afford to send their children to school at all, although most acknowledge the value of advanced education. Thus, school fees serve as an additional proxy for household wealth for those families with children. Based on our preliminary surveys, this information was well known to heads of households and was correlated with income.
The fourth section of our survey was used to ascertain physical capital, including relative access to physical resources and infrastructure. We asked respondents to describe the location, means of travel, and most recent visit to hospitals, schools, and markets. We also asked about access to water and electricity. Specifically, we asked: “What is your household’s main source of drinking water?”, “How far do you travel to this water source?”, and, “How many months out of the year is this water source available?” We asked if the household had access to electricity and about any costs associated with electricity and water access. Last, we asked about ownership of physical assets (Below et al. 2012, Jain et al. 2015), including vehicles (cars, tractors, etc.), farming tools (hoe plow), and electronics (refrigerator, stove, television, computer, tablet, cell phone; Appendix 1, section A1.2).
The fifth section of our survey asked participants about their social capital. We collected data on participation in community groups, attendance at community meetings, hiring or working for neighbors, and trade or exchange with neighbors. We also asked if anyone in the homestead held a leadership position in the community. Such positions could include those in schools or churches, official community titles such as chief or induna (local leader), working with community police, or serving on a leadership council, etc.
The last section of our survey focused on exposure to drought and past adaptations. Based on preliminary research, we identified nine primary adaptation strategies typically employed in response to drought conditions: planting drought- or heat-resistant crops, conservation farming to minimize soil erosion and nutrient loss, beekeeping, chicken husbandry, selling natural resources (primarily firewood), selling handicrafts, looking for off-farm employment, and participating in training or activities with aid organizations. For each adaptation, we asked if they had ever performed the adaptation strategy and if they had performed it within the past five years. We did not distinguish between households that specifically chose the behavior as an adaptation to the current drought and those who attempted it for other reasons because climate adaptation often happens in conjunction with other factors (Adger et al. 2005). In addition, for adaptations that the respondent did not perform, we asked what the barriers were (e.g., time, labor, money, knowledge, interest, social norms, etc.; Appendix 1, section A1.2). To determine if an adaptation was successful, we asked about the outcomes of the adaptation. Specifically, we asked if the adaptation, in the respondent’s opinion, led to a change in food consumption, income, or agricultural output. We categorized any adaptation that led to an increase in any of the three possibilities as successful. We also asked if, in the respondent’s view, the adaptation was successful in minimizing the negative impacts of the current drought. Such a definition of adaptation is subjective to the performer and operates on a relatively short timescale. We believe that this definition of success is relevant in determining future adaptation options and perceived well-being. If a household believes that an adaptation improved conditions during this drought, they are more likely to engage in that adaptation in the future.
The initial survey included > 60 questions related to the five types of capital. We first conducted a multicollinearity analysis using Pearson’s correlation coefficient to identify highly correlated survey variables, which we removed from the analysis. To create a more parsimonious model and avoid overfitting of the data, we conducted nonlinear principal components analysis (PCA) on the variables from the remaining questions for each capital in the package GIFI (de Leeuw et al. 2009). Nonlinear PCA accommodates nonparametric data, including ordinal and binomial data such as those collected in our research (Linting et al. 2007). From the nonlinear PCA, we obtained two sets of variables that we could include in our models. The first set was a principal component for each of the five types of capital. The second set included PCA loadings for the individual survey variables that accounted for the most variation in the data for each type of capital (hereafter “capital indicators”).
We created several response variables to assess the relationship between adaptation and different types of capital using generalized linear models. We fitted models to the total number of adaptations attempted by a household during the past five years (this includes households that did not attempt any adaptations), a binary measure of adaptation success, and the ratio of successful adaptations to attempted adaptations (this excludes households that did not attempt any adaptations). Additionally, we fit models to binary response variables (attempted, 0 or 1) for each of the adaptations (e.g., conservation farming, alternative crops, selling chickens) that were attempted by at least 15% of the survey population.
For each of the three response variables, we created two sets of competing models (models 1–3). The first set included a global model with all five types of capital (using the PCA metrics), competing capital models, and a null model (model subset a). The second set of competing models included a global model with all capital indicators, competing capital indicator models, and a null model (model subset b). We selected the best models using the Aikaike Information Criterion (AIC). We ranked all models according to AICc values and took their model-averaged coefficients (Burnham and Anderson 2002). We considered models with ΔAICc values ≤ 2 as candidates for the best model. We completed our analysis using the packages MASS and MuMIn in RStudio with the R platform (Venables and Ripley 2002, Barton 2015, RStudio Team 2015, R Core Team 2016).
The 13 community focus groups identified drought as the biggest change affecting their communities. Respondents reported that drought affected food availability, food quality, jobs, interpersonal relationships, infrastructure, development, and a host of other aspects of their lives. The most commonly cited type of response to drought was changing farming strategies, which included farming different types of crops, conservation farming, farming a smaller proportion of available land, or temporarily stopping farming activities.
We completed a total of 307 household surveys. In terms of human capital, the average household size was 6.7 (standard deviation [SD] = 3.5), and the average proportion of household members with jobs was 0.31 (SD = 0.32). For natural capital, most households travel 3–5 km to collect fuelwood, 56% of households did not plant crops during the previous wet season, and 30% of households eat wild food collected from nearby natural resources (although this accounts for < 10% of the total diet for all households). Additionally, there was an average of 1.4 cows, 8.3 chickens, and 2.6 goats per household. In terms of financial capital, 57% of households earn < 1000 Emalangeni (USD $74.50) per month, and households spend an average of 674 Emalangeni (USD $50) on groceries each month. In terms of physical capital, there is an average of 3.7 (SD = 2.4) buildings per household, 64% of households have electricity, and 45% of households get water from a tap within the homestead, whereas 25% of households get water from a borehole, well, or dam. In addition, 50% of households own a refrigerator, 90% own a cell phone, and 19% own a vehicle. In terms of social capital, 42% of households work for or hire their neighbors, 37% of households are involved in a community organization, and 22% of households have a family member with a leadership position in the community.
We included 19 capital indicators in the final PCA and indices (Fig. 2). PCA variable loadings represent the variation in the data explained by each variable (Abdi and Williams 2010). The indicators with the largest principal component loadings for each capital were amount of land available to farm (natural capital), participating in a community organization (social capital), average money spent on groceries (financial capital), ratio of adults to dependents (human capital), and number of buildings in the homestead (physical capital).
Of the households surveyed, 78% attempted one or more adaptations. Of those that attempted to adapt, 57% reported at least one successful adaptation (Table 1). The most common adaptations attempted were planting alternative crops that were understood to be heat or drought resistant. Fifty-two percent of households planted alternative crops, and it was reported as successful 15% of the time. The second most common adaptation was implementation of conservation farming practices that are taught by government officials, aid organizations, or community members. Thirty percent of households attempted conservation farming, and it was reported as successful 10% of the time. Other common adaptations included raising chickens to sell (23% of households), making handicrafts to sell (19% of households), collecting natural resources to sell (18.5% of households), and looking for work off-farm, typically in Swazi cities or in South Africa (24% of households; Table 1). Twenty-two percent of households did not attempt any adaptation, and 55% of households attempted more than one adaptation.
The top competing models for total adaptations attempted per household included variables for social, natural, financial, and human capital (Table 2, model 1a; see also Appendix 2). The model-averaged beta estimates indicate that social capital (β = 0.19 ± 0.048) was most strongly correlated with attempted adaptations (Fig. 3). Working for or hiring neighbors (β = 0.16 ± 0.09), participating in a community organization (β = 0.33 ± 0.1), number of chickens (β = 0.02 ± 0.01) and goats (β = 0.02 ± 0.01) owned, area farmed (β = −0.05 ± 0.03), and distance to savanna were in the top competing capital indicator models (Table 2, model 1b).
The top competing models for the ratio of successful adaptations to attempted adaptations included variables for natural, social, financial, and physical capital (Table 2, model 2a; see also Appendix 2). The model-averaged beta estimates indicate that natural capital (β = 0.32 ± 0.22) was most strongly correlated with adaptation success ratio (Fig. 4). The number of buildings in the homestead (β = −0.12 ± 0.05), working for or hiring neighbors (β = 0.39 ± 0.18), the average homestead harvest (β = 0.13 ± 0.07), and household income (β = −0.2 ± 0.1) were included in the top competing models (Table 2, model 2b).
The top competing models for successful adaptation included social, natural, physical, and financial capital (Table 2, model 3a; see also Appendix 2). The model-averaged beta estimates indicated that social capital (β = 0.59 ± 0.4) was most strongly correlated with a successful adaptation (Fig. 5). Participation in a community organization (β = 1.22 ± 0.35), working for or hiring neighbors (β = 0.34 ± 0.29), owning goats (β = 0.08 ± 0.04) or chickens (β = 0.04 ± 0.03), and distance to savanna (β = 0.11 ± 0.14) were in the competing models (Table 2, model 3b).
While overall adaptation behavior was strongly correlated with social and natural capital, individual adaptations varied more with individual survey variables. The decision to engage in conservation farming was positively correlated with social and natural capital, and specifically, with working for or hiring neighbors, land available to farm, and distance to savanna patches available for wood collection (Table 3). The decision to plant alternate crops, however, was most strongly correlated with participation in a community organization, cow ownership, and distance to water (Table 3). A household’s decision to sell chickens was positively correlated with social and natural capital, specifically, working for neighbors and consuming wild foods (Table 3). The decision to sell natural resources (primarily firewood) was negatively correlated with household electricity and education, and positively correlated with working for and hiring neighbors (Table 3). Finally, the decision to sell handicrafts was positively correlated with social and natural capital variables, including participation in a community organization, consuming wild foods, and ownership of goats (Table 3).
As climate change continues to threaten environmentally dependent livelihoods in the developing world, there is a need to identify the resources and strategies that allow livelihoods to persist across contexts. To contribute to that effort, we studied adaptation to drought-related stress among rural farmers in Eswatini. We found that social capital and natural capital were the most important types of capital in predicting successful adaptation to drought. Our data highlight the critical link between social and natural capital and adaptation observed elsewhere (Bebbington 1997, Woolcock 1998, Nawrotzki et al. 2012, Tibesigwa et al. 2014) and furthers the understanding of pathways through which capital operates to support livelihoods.
Social capital and community networks operate in diverse ways to serve as critical resources for adaptation to drought across contexts. For instance, the decision to engage in an adaptation is often driven by encouragement and information gained via trusted information sources and contacts within social networks (Lo 2013, Udmale et al. 2014). In contrast, younger, more socially isolated households may be less likely to adapt effectively and more likely to experience greater stress and loss as a result of drought (Austin et al. 2018, Neef et al. 2018). Examining the role of social capital in a household and community context in Eswatini enables us to understand better how place influences the role of social capital and adaptation (Pelling and High 2005). In our study area, households reported increased knowledge of adaptation options and support for adaptation via social networks, particularly involvement with community organizations. Gains in social capital are also positively associated with increases in other forms of capital and greater access to resources that support climate adaptation (Narayan and Pritchett 1999, Guiso et al. 2004). We found that households participating in community-led organizations reported increased human capital via training in farming strategies, small business management, and trade skills. Similarly, households participating in community organizations and those that worked for neighbors reported increased financial security and access to short-term financial resources and information on adaptation strategies, and increased access to otherwise inaccessible physical capital such as tractors. In Eswatini, as elsewhere, social capital operates along multiple pathways to enable households to obtain additional capital and resources, allowing them to engage in more adaptation during drought.
Natural capital also affects adaptation in multiple complex pathways. First, households that are heavily dependent on natural resources are more likely to be affected by climate shocks and stresses such as drought than those that are decoupled from the environment (Osbahr et al. 2008, Nawrotzki et al. 2012, Blignaut et al. 2014, Guerry et al. 2015). In Eswatini, many households reported a lack of alternative livelihood strategies and strong cultural connections to natural resources, increasing their vulnerability to drought. Second, natural capital also supports adaptation by providing critical resources that are often readily available and do not require extensive knowledge or training to exploit (Osbahr et al. 2008, Belay et al. 2017). Respondents in our study reported that local natural capital, with which they were familiar, supported adaptations, including shifting to drought- or heat-resistant crops, making and selling crafts, and collecting and selling firewood, fruit, and thatch grass. Finally, natural capital, depending on the context, can also serve as a form of social and financial capital (Kerven 1992, Turner 2009). Specifically, livestock and communal rangelands serve important cultural and economic functions across southern Africa (Cousins 1999, Carter and May 1999). Respondents in the communities we surveyed reported liquidating livestock, using livestock as collateral, and strengthening social connections through livestock during the drought, enabling increased adaptation success.
While social and natural capital were the best predictors of overall adaptation, we also found evidence that a diverse pool of capital enables diverse adaptation and a greater likelihood of success. Adaptation strategies such as planting drought-resistant crops, for example, require access to physical (plow), natural (land, seeds), and social or human capital (education on alternative crops). To sell firewood, a household must have access to savannas or forests to collect firewood (natural capital), a network of households to buy firewood (social capital), and demand driven by a lack of affordable access to electricity (limitations on physical capital). Households with greater capital diversity, i.e., a more even spread across the five types of capital, were more likely to have successful adaptations. Diverse capital also supports diverse adaptations and livelihood strategies, which are especially valuable when environmental stresses are extreme (Berman et al. 2015, Huynh and Stringer 2018).
It is worth noting that these data were collected during the drought. Perceptions of the success of an adaptation may have changed following the end of the drought. Adaptations operate across broad spatial and temporal scales. Adaptations that benefit the adopter now may have negative effects in the future (Paavola 2008, Barnett and O’Neill 2010). Similarly, adaptations that benefit one household or community may have negative effects on other households or communities (Adger et al. 2005, Holler 2014). This is a limitation of our study, and there is a need for longitudinal research assessing perceptions of adaptation effectiveness in Eswatini and elsewhere.
Our findings suggest that maintenance and growth of social and natural capital are important ways to improve the ability of households to respond to drought and other changes. Evidence of this trend has been observed elsewhere, where the establishment of local community organizations and programs led to increased economic activity and increased resilience to natural disasters (Westerman et al. 2012). Our work provides further evidence of the complex ways in which social capital operates and additional evidence for relevant indicators of social capital at household and community levels (Pelling and High 2005). As our data suggest, social capital enables adaptation largely because it creates opportunities for tangible exchange of goods, services, information, and trust, which allow for successful adaptation. Broad networks of resources exchange across spatial, temporal, and structural scales have enabled transformational adaptation, which increased resilience to climate change (Dowd et al. 2014). The importance of social capital is now being incorporated into community-based disaster preparedness strategies to strengthen community relationships and leverage assets to improve community-level responses to natural disasters (Luna 2001). As our research demonstrates, social capital is critical to adaptation in our changing world and needs to be at the forefront of development and climate adaptation policy. Communities, governments, and organizations should establish and foster programs that allow households to connect in meaningful ways, to ensure they have access to the resources and opportunities necessary for long-term adaptation and resilience.
We thank all of the communities in the Lowveld of Eswatini who shared their lives and struggles with us. We also thank all the community leaders and translators who helped make this research possible. Finally, we thank All Out Africa, the Ford Foundation, and the Bill and Melinda Gates Foundation for financial and logistical support.
Abdi, H., and L. J. Williams. 2010. Principal component analysis. Wiley Interdisciplinary Reviews: Computational Statistics 2(4):433-459. https://doi.org/10.1002/wics.101
Abel, N., D. H. M. Cumming, and J. M. Anderies. 2006. Collapse and reorganization in social-ecological systems: questions, some ideas, and policy implications. Ecology and Society 11(1):17. https://doi.org/10.5751/ES-01593-110117
Adger, W. N. 2003. Social capital, collective action, and adaptation to climate change. Economic Geography 79(4):387-404. https://doi.org/10.1111/j.1944-8287.2003.tb00220.x
Adger, W. N., N. W. Arnell, and E. L. Tompkins. 2005. Successful adaptation to climate change across scales. Global Environmental Change 15(2):77-86. https://doi.org/10.1016/j.gloenvcha.2004.12.005
Austin, E. K., T. Handley, A. S. Kiem, J. L. Rich, T. J. Lewin, H. H. Askland, S. S. Askarimarnani, D. A. Perkins, and B. J. Kelly. 2018. Drought-related stress among farmers: findings from the Australian rural mental health study. Medical Journal of Australia 209(4):159-165. https://doi.org/10.5694/mja17.01200
Barnett, J., and S. O’Neill. 2010. Maladaptation. Global Environmental Change 20(2):211-213. https://doi.org/10.1016/j.gloenvcha.2009.11.004
Barton, K. 2015. Package ‘MuMIn’: model selection and model averaging based on information criteria. R package version 1.9.13. R Foundation for Statistical Computing, Vienna, Austria. [online] URL: https://cran.r-project.org/web/packages/MuMIn/index.html
Bebbington, A. 1997. Social capital and rural intensification: local organizations and islands of sustainability in the rural Andes. Geographical Journal 163(2):189-197. https://doi.org/10.2307/3060182
Bebbington, A. 1999. Capitals and capabilities: a framework for analyzing peasant viability, rural livelihoods and poverty. World Development 27(12):2021-2044. https://doi.org/10.1016/S0305-750X(99)00104-7
Belay, A., J. W. Recha, T. Woldeamanuel, and J. F. Morton. 2017. Smallholder farmers’ adaptation to climate change and determinants of their adaptation decisions in the Central Rift Valley of Ethiopia. Agriculture and Food Security 6:24. https://doi.org/10.1186/s40066-017-0100-1
Below, T. B., K. D. Mutabazi, D. Kirschke, C. Franke, S. Sieber, R. Siebert, and K. Tscherning. 2012. Can farmers’ adaptation to climate change be explained by socio-economic household-level variables? Global Environmental Change 22(1):223-235. https://doi.org/10.1016/j.gloenvcha.2011.11.012
Berman, R. J., C. H. Quinn, and J. Paavola. 2015. Identifying drivers of household coping strategies to multiple climatic hazards in Western Uganda: implications for adapting to future climate change. Climate and Development 7(1):71-84. https://doi.org/10.1080/17565529.2014.902355
Blignaut, J., J. Aronson, and R. de Groot. 2014. Restoration of natural capital: a key strategy on the path to sustainability. Ecological Engineering 65:54-61. https://doi.org/10.1016/j.ecoleng.2013.09.003
Boudreau, T. 2010. LIFT Swaziland livelihood data analysis. Food Economy Group, Potten End, UK. [online] URL: https://www.heacod.org/en-gb/Published%20Reports/FEG%20Pre-assessment%20livelihood%20analysis%20FINAL_Nov%202010.pdf
Bryan, B. A., J. Huai, J. Connor, L. Gao, D. King, J. Kandulu, and G. Zhao. 2015. What actually confers adaptive capacity? Insights from agro-climatic vulnerability of Australian wheat. Plos One 10:e0117600. https://doi.org/10.1371/journal.pone.0117600
Burnham, K. P., and D. R. Anderson. 2002. Model selection and multimodel inference: a practical information-theoretic approach. Second edition. Springer, New York, New York, USA. http://dx.doi.org/10.1007/b97636
Carter, M. R. and J. May. 1999. Poverty, livelihood and class in rural South Africa. World Development 27(1):1-20. https://doi.org/10.1016/S0305-750X(98)00129-6
Cassidy, L., and G. D. Barnes. 2012. Understanding household connectivity and resilience in marginal rural communities through social network analysis in the village of Habu, Botswana. Ecology and Society 17(4):11. https://doi.org/10.5751/ES-04963-170411
Cleverly, R. W. 1979. The volcanic geology of the Lubombo monocline in Swaziland. Transvaal Geological Society of South Africa 82:343-348.
Costanza, R., R. d’Arge, R. de Groot, S. Farber, M. Grasso, B. Hannon, K. Limburg, S. Naeem, R. V. O’Neill, J. Paruelo, R. G. Raskin, P. Sutton, and M. van den Belt. 1997. The value of the world’s ecosystem services and natural capital. Nature 387(6630):253-260. https://doi.org/10.1038/387253a0
Cousins, B. 1999. Invisible capital: the contribution of communal rangelands to rural livelihoods in South Africa. Development Southern Africa 16(2):299-318. https://doi.org/10.1080/03768359908440079
de Leeuw, J., P. Mair, and W. W. Wien. 2009. Gifi methods for optimal scaling in R: the package homals. Journal of Statistical Software 31(4):1-30. http://dx.doi.org/10.18637/jss.v031.i04
de Sherbinin, A., L. K. VanWey, K. McSweeney, R. Aggarwal, A. Barbieri, S. Henry, L. M. Hunter, W. Twine, and R. Walker. 2008. Rural household demographics, livelihoods and the environment. Global Environmental Change 18(1):38-53. https://doi.org/10.1016/j.gloenvcha.2007.05.005
Department of Water Affairs and Forestry, Republic of South Africa. 2002. Report on the social aspects on the Usuthu River: Swaziland and Mozambique. MNM Consultants, KwaNgwanase, South Africa. [online] URL: http://www.dwaf.gov.za/sfra/sea/usutu-mhlathuze%20wma/Social%20%20Component/Final%20Draft%20Mozambique%20and%20Swaziland%20Social%20Report.pdf
Deressa, T. T., R. M. Hassan, C. Ringler, T. Alemu, and M. Yesuf. 2009. Determinants of farmers’ choice of adaptation methods to climate change in the Nile Basin of Ethiopia. Global Environmental Change 19(2):248-255. https://doi.org/10.1016/j.gloenvcha.2009.01.002
Dowd, A.-M., N. Marshall, A. Fleming, E. Jakku, E. Gaillard, and M. Howden. 2014. The role of networks in transforming Australian agriculture. Nature Climate Change 4(7):558-563. https://doi.org/10.1038/nclimate2275
Ellis, F. 1998. Household strategies and rural livelihood diversification. Journal of Development Studies 35(1):1-38. https://doi.org/10.1080/00220389808422553
Esteve, P., C. Varela-Ortega, and T. E. Downing. 2018. A stakeholder-based assessment of barriers to climate change adaptation in a water-scarce basin in Spain. Regional Environmental Change 18(8):2505-2517. https://doi.org/10.1007/s10113-018-1366-y
Evans, L. S., C. C. Hicks, W. N. Adger, J. Barnett, A. L. Perry, P. Fidelman, and R. Tobin. 2016. Structural and psycho-social limits to climate change adaptation in the Great Barrier Reef region. Plos One 11(3):e0150575. https://doi.org/10.1371/journal.pone.0150575
Foley, J. A., R. DeFries, G. P. Asner, C. Barford, G. Bonan, S. R. Carpenter, F. S. Chapin, M. T. Coe, G. C. Daily, H. K. Gibbs, J. H. Helkowski, T. Holloway, E. A. Howard, C. J. Kucharik, C. Monfreda, J. A. Patz, I. C. Prentice, N. Ramankutty, and P. K. Snyder. 2005. Global consequences of land use. Science 309(5734):570-574. https://doi.org/10.1126/science.1111772
Forster, P. G., and B. J. Nsibande. 2000. Swaziland: contemporary social and economic issues. Routledge, London, UK. https://doi.org/10.4324/9781315190822
Franzel, S. 1999. Socioeconomic factors affecting the adoption potential of improved tree fallows in Africa. Agroforestry Systems 47(1-3):305-321. https://doi.org/10.1023/A:1006292119954
Goudie, A., and D. Price Williams. 1983. The Atlas of Swaziland. Swaziland National Trust Commission, Lobamba, Swaziland.
Guerry, A. D., S. Polasky, J. Lubchenco, R. Chaplin-Kramer, G. C. Daily, R. Griffin, M. Ruckelshaus, I. J. Bateman, A. Duraiappah, T. Elmqvist, M. W. Feldman, C. Folke, J. Hoekstra, P. M. Kareiva, B. L. Keeler, S. Li, E. McKenzie, Z. Ouyang, B. Reyers, T. H. Ricketts, J. Rockström, H. Tallis, and B. Vira. 2015. Natural capital and ecosystem services informing decisions: from promise to practice. Proceedings of the National Academy of Sciences 112(24):7348-7355. https://doi.org/10.1073/pnas.1503751112
Guiso, L., P. Sapienza, and L. Zingales. 2004. The role of social capital in financial development. American Economic Review 94(3):526-556. http://dx.doi.org/10.1257/0002828041464498
Gunderson, L. H., and C. S. Holling. 2002. Panarchy: understanding transformations in human and natural systems. Island Press, Washington, D.C., USA.
Hinkel, J. 2011. “Indicators of vulnerability and adaptive capacity”: towards a clarification of the science-policy interface. Global Environmental Change 21(1):198-208. https://doi.org/10.1016/j.gloenvcha.2010.08.002
Holler, J. 2014. Is sustainable adaptation possible? Determinants of adaptation on Mount Kilimanjaro. Professional Geographer 66:526-537. https://doi.org/10.1080/00330124.2014.922015
Huynh, L. T. M., and L. C. Stringer. 2018. Multi-scale assessment of social vulnerability to climate change: an empirical study in coastal Vietnam. Climate Risk Management 20:165-180. https://doi.org/10.1016/j.crm.2018.02.003
Jain, M., S. Naeem, B. Orlove, V. Modi, and R. S. DeFries. 2015. Understanding the causes and consequences of differential decision-making in adaptation research: adapting to a delayed monsoon onset in Gujarat, India. Global Environmental Change 31:98-109. https://doi.org/10.1016/j.gloenvcha.2014.12.008
Jones, L., and E. Boyd. 2011. Exploring social barriers to adaptation: insights from western Nepal. Global Environmental Change 21(4):1262-1274. https://doi.org/10.1016/j.gloenvcha.2011.06.002
Katz, E. G. 2000. Social capital and natural capital: a comparative analysis of land tenure and natural resource management in Guatemala. Land Economics 76(1):114-132. https://doi.org/10.2307/3147261
Kerven, C. 1992. Customary commerce: a historical reassessment of pastoral livestock marketing in Africa. ODI Agricultural Occasional Paper 15. Overseas Development Institute, London, UK. [online] URL: https://www.odi.org/publications/1706-customary-commerce-historical-reassessment-pastoral-livestock-marketing-africa
Kiem, A. S., and E. K. Austin. 2013. Drought and the future of rural communities: opportunities and challenges for climate change adaptation in regional Victoria, Australia. Global Environmental Change 23(5):1307-1316. https://doi.org/10.1016/j.gloenvcha.2013.06.003
Linting, M., J. J. Meulman, P. J. F. Groenen, and A. J. van der Koojj. 2007. Nonlinear principal components analysis: introduction and application. Psychological Methods 12(3):336-358. https://doi.org/10.1037/1082-989X.12.3.336
Lo, A. Y. 2013. The role of social norms in climate adaptation: mediating risk perception and flood insurance purchase. Global Environmental Change 23(5):1249-1257. https://doi.org/10.1016/j.gloenvcha.2013.07.019
Luna, E. M. 2001. Disaster mitigation and preparedness: the case of NGOs in the Philippines. Disasters 25(3):216-226. https://doi.org/10.1111/1467-7717.00173
Mabuza, K. 2016. Swaziland rapid assessment summary report 2016. Deputy Prime Minister’s Office, Mbabane, Swaziland. [online] URL: https://www.humanitarianresponse.info/sites/www.humanitarianresponse.info/
Mabuza, M. L., S. L. Hendriks, G. F. Ortmann, and M. M. Sithole. 2009. The impact of food aid on maize prices and production in Swaziland 48(1):85-105. https://doi.org/10.1080/03031853.2009.9523818
Manyatsi, A. M., N. Zwane, and M. Dlamini. 2015. Evaluation of satellite rainfall estimates for Swaziland. American Journal of Agriculture and Forestry 3(3):93-98. https://doi.org/10.11648/j.ajaf.20150303.15
Mbukwa, J. N. 2014. Some aspects of correlation of physical capital and infrastructures on household food security: evidence from rural Tanzania. Journal of Economics and Sustainabile Development 5(9):26-34.
McDowell, J. Z., and J. J. Hess. 2012. Accessing adaptation: multiple stressors on livelihoods in the Bolivian highlands under a changing climate. Global Environmental Change 22(2):342-352. https://doi.org/10.1016/j.gloenvcha.2011.11.002
Meyer, B. D., W. K. C. Mok, and J. X. Sullivan. 2015. Household surveys in crisis. Journal of Economic Perspectives 29(4):199-226. http://dx.doi.org/10.1257/jep.29.4.199
Moore, J. C., L. L. Stinson, and E. J. Welniak Jr. 2000. Income measurement error in surveys: a review. Journal of Official Statistics 16(4):331-361.
Moore, M. D., and J. Daday. 2010. Barriers to human capital development: case studies in Swaziland, Cameroon and Kenya. Africa Education Review 7(2):283-304. https://doi.org/10.1080/18146627.2010.515418
Narayan, D., and L. Pritchett. 1999. Cents and sociability: household income and social capital in rural Tanzania. Economic Development and Cultural Change 47(4):871-897. https://doi.org/10.1086/452436
Nawrotzki, R. J., L. M. Hunter, and T. W. Dickinson. 2012. Rural livelihoods and access to natural capital: differences between migrants and non-migrants in Madagascar. Demographic Research 26:661-700. [online] URL: https://www.demographic-research.org/volumes/vol26/24/
Nhuan, M. T., N. T. Tue, N. T. H. Hue, T. D. Quy, and T. M. Lieu. 2016. An indicator-based approach to quantifying the adaptive capacity of urban households: the case of Da Nang city, Central Vietnam. Urban Climate 15:60-69. https://doi.org/10.1016/j.uclim.2016.01.002
Neef, A., L. Benge, B. Boruff, N. Pauli, E. Weber, and R. Varea. 2018. Climate adaptation strategies in Fiji: the role of social norms and cultural values. World Development 107:125-137. https://doi.org/10.1016/j.worlddev.2018.02.029
Osbahr, H., C. Twyman, W. N. Adger, and D. S. G. Thomas. 2008. Effective livelihood adaptation to climate change disturbance: scale dimensions of practice in Mozambique. Geoforum 39(6):1951-1964. https://doi.org/10.1016/j.geoforum.2008.07.010
Osbahr, H., C. Twyman, W. N. Adger, and D. S. G. Thomas. 2010. Evaluating successful livelihood adaptation to climate variability and change in Southern Africa. Ecology and Society 15(2):27. https://doi.org/10.5751/ES-03388-150227
Oseni, T. O., and M. T. Masarirambi. 2011. Effect of climate change on maize (Zea mays) production and food security in Swaziland. American-Eurasian Journal of Agriculture and Environmental Science 11(3):385-391. [online] URL: http://hpccc.gov.in/PDF/Agriculture/Effect%20of%20Climate%20Change%20on%20Maize%20(Zea%20mays).pdf
Paavola, J. 2008. Livelihoods, vulnerability and adaptation to climate change in Morogoro, Tanzania. Environmental Science and Policy 11(7):642-654. https://doi.org/10.1016/j.envsci.2008.06.002
Paul, C. J., E. S. Weinthal, M. F. Bellemare, and M. A. Jeuland. 2016. Social capital, trust, and adaptation to climate change: evidence from rural Ethiopia. Global Environmental Change 36:124-138. https://doi.org/10.1016/j.gloenvcha.2015.12.003
Pelling, M. 2003. The vulnerability of cities: natural disasters and social resilience. Routledge, London, UK.
Pelling, M., and C. High. 2005. Understanding adaptation: What can social capital offer assessments of adaptive capacity? Global Environmental Change 15(4):308-319. https://doi.org/10.1016/j.gloenvcha.2005.02.001
Pretty, J., and H. Ward. 2001. Social capital and the environment. World Development 29(2):209-227. https://doi.org/10.1016/S0305-750X(00)00098-X
Putnam, R. D., R. Leonardi, and R. Y. Nonetti. 1993. Making democracy work: civic traditions in modern Italy. Princeton University Press, Princeton, New Jersey, USA. https://doi.org/10.2307/j.ctt7s8r7
R Core Team. 2017. R: a language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. [online] URL: https://www.R-project.org/
Rached, E., E. Rathgeber, and D. B. Brooks. 1996. Water management in Africa and the Middle East: challenges and opportunities. International Development Research Centre, Ottawa, Canada. [online] URL: https://www.idrc.ca/en/book/water-management-africa-and-middle-east-challenges-and-opportunities
RStudio Team. 2015. RStudio: integrated development for R. RStudio, Boston, Massachusetts, USA. [online] URL: http://www.rstudio.com/
Russell, M. 1983. Boundaries and structures in the Swaziland homestead. University of Swaziland, Kwaluseni, Eswatini.
Scoones, I. 1998. Sustainable rural livelihoods: a framework for analysis. IDS Working Paper 72. Institute of Development Studies, Brighton, UK. [online] URL: https://www.ids.ac.uk/publications/sustainable-rural-livelihoods-a-framework-for-analysis/
Smit, B., and J. Wandel. 2006. Adaptation, adaptive capacity and vulnerability. Global Environmental Change 16(3):282-292. https://doi.org/10.1016/j.gloenvcha.2006.03.008
Sseguya, H. 2009. Impact of social capital on food security in southeast Uganda. Dissertation. Iowa State University, Ames, Iowa, USA. [online] URL: https://doi.org/10.31274/etd-180810-2327
Steffen, W., A. Sanderson, P. D. Tyson, J. Jäger, P. A. Matson, B. Moore III, F. Oldfield, K. Richardson, H. J. Schellnhuber, B. L. Turner II, and R. J. Wasson. 2005. Global change and the Earth system: a planet under pressure. Springer, Berlin, Germany. http://dx.doi.org/10.1007/b137870
Tibesigwa, B., M. Visser, and W. Twine. 2014. Investigating the sensitivity of household food security to agriculture-related shocks and the implication of informal social capital and natural resource capital: the case of rural households in Mpumalanga, South Africa. Working Paper 470. Economic Research Southern Africa, Claremont, South Africa. [online] URL: https://econrsa.org/node/965
Turner, M. D. 2009. Capital on the move: the changing relation between livestock and labor in Mali, West Africa. Geoforum 40(5):746-755. https://doi.org/10.1016/j.geoforum.2009.04.002
Udmale, P. D., Y. Ichikawa, A. S. Kiem, and S. N. Panda. 2014. Drought impacts and adaptation strategies for agriculture and rural livelihood in the Maharashtra State of India. Open Agriculture Journal 8:41-47. https://doi.org/10.2174/1874331501408010041
United Nations Department of Economic and Social Affairs. 2005. Household sample surveys in developing and transition countries. ST/ESA/STAT/SER.F/96. United Nations, New York, New York, USA. [online] URL: https://unstats.un.org/unsd/hhsurveys/pdf/Household_surveys.pdf
Venables, W. N., and B. D. Ripley. 2002. Modern applied statistics with S. Fourth edition. Springer, New York, New York, USA. https://doi.org/10.1007/978-0-387-21706-2
Vincent, K. 2007. Uncertainty in adaptive capacity and the importance of scale. Global Environmental Change 17(1):12-24. https://doi.org/10.1016/j.gloenvcha.2006.11.009
Westerman, K., K. L. L. Oleson, and A. R. Harris. 2012. Building socio-ecological resilience to climate change through community-based coastal conservation and development: experiences in southern Madagascar. Western Indian Ocean Journal of Marine Science 11(1):87-97.
Woolcock, M. 1998. Social capital and economic development: toward a theoretical synthesis and policy framework. Theory and Society 27(2):151-208. https://doi.org/10.1023/A:1006884930135
Yohe, G., and R. S. J. Tol. 2002. Indicators for social and economic coping capacity—moving toward a working definition of adaptive capacity. Global Environmental Change 12(1):25-40. https://doi.org/10.1016/S0959-3780(01)00026-7