EN中文

Journal of Resources and Ecology >

2024 , Vol. 15 >Issue 4: 1015 - 1026

DOI: https://doi.org/10.5814/j.issn.1674-764x.2024.04.021

Resource Economy

Assessing and Comparing Smallholders’ Vulnerability to Climate Change among Countries in the Pan-Third Pole Region

  • XU Xiangbo , 1, 2 ,
  • XU Ce , 3, 4, * ,
  • LI Chang , 1, 2, 5, * ,
  • FU Chao 1, 2 ,
  • ZHOU Yunqiao 6
Expand
  • 1. Key Laboratory of Ecosystem Network Observation and Modeling, Institute of Geographic Sciences and Natural Resources Research, Chinese Academy of Sciences, Beijing 100101, China
  • 2. International Ecosystem Management Partnership, United Nations Environment Programme, Beijing 100101, China
  • 3. Land Consolidation and Rehabilitation Center, Ministry of Natural Resources, Beijing 100035, China
  • 4. Technology Innovation Center for Land Engineering, Ministry of Natural Resources, Beijing 100035, China
  • 5. University of Chinese Academy of Sciences, Beijing 100049, China
  • 6. State Key Laboratory of Tibetan Plateau Earth System, Resources and Environment, Institute of Tibetan Plateau Research, Chinese Academy of Sciences, Beijing 100101, China
* XU Ce, E-mail: ; LI Chang, E-mail:

XU Xiangbo, E-mail:

Received date: 2023-12-18

  Accepted date: 2024-02-20

  Online published: 2024-07-25

Supported by

The National Natural Science Foundation of China(7231101308)

The National Natural Science Foundation of China(72374190)

The National Natural Science Foundation of China(31861143015)

The Strategic Priority Research Program of the Chinese Academy of Sciences(XDA20010303)

Fold

Abstract

A comprehensive assessment of climate change vulnerability is imperative for formulating effective adaptation strategies and advancing sustainable development goals. As one of the most climate-vulnerable regions globally, the Pan-Third Pole area lacks transnational vulnerability assessments, which poses a significant obstacle to efficient climate adaptation. This study conducted transnational comparisons based on primary micro-survey data collected uniformly across Nepal, Cambodia, Thailand, and Myanmar, which are all located in the Pan-Third Pole region. Evaluating and comparing the vulnerabilities employed an extended framework of climate change indicators. The findings reveal substantial variations in vulnerability among the countries, with Cambodia exhibiting the highest vulnerability, followed by Thailand, Myanmar, and Nepal in descending order, primarily due to differences in exposure. Household exposure to climate change also varied significantly. Sensitivity scores decreased in the order of Nepal > Cambodia > Thailand > Myanmar, with demographic factors, tap water accessibility, and land being the major contributors and sources of differentiation among the countries. Regarding adaptability, Thailand demonstrated the highest adaptability, with human and financial capital as the key differentiators. The outcomes underscore the need for tailored policy measures addressing the diverse vulnerabilities, including enhancing household disaster prevention and capital protection. Furthermore, targeted international investments are crucial for improving adaptability among smallholders in this unique region.

Key words： climate change; vulnerability; smallholder; sustainable development; Pan-Third Pole region

Cite this article

XU Xiangbo , XU Ce , LI Chang , FU Chao , ZHOU Yunqiao . Assessing and Comparing Smallholders’ Vulnerability to Climate Change among Countries in the Pan-Third Pole Region[J]. Journal of Resources and Ecology, 2024 , 15(4) : 1015 -1026 . DOI: 10.5814/j.issn.1674-764x.2024.04.021

1 Introduction

Climate change poses a significant threat to both ecology and the global economy, with far-reaching impacts observed worldwide (Thornton et al., 2014). These effects are projected to escalate in the coming decades, posing substantial challenges to the achievement of sustainable development goals (SDGs) (Casado-Asensio and Steurer, 2014; Jiang et al., 2021). The repercussions of climate change extend to people's livelihoods, thereby diminishing the sustainability of household economies (Pandey et al., 2017). Empirical studies have revealed that extreme weather conditions contribute to reduced freshwater availability and ecosystem supply, as well as increased food shortages (Pandey et al., 2017; Nef et al., 2021). Furthermore, climate change is linked to resource depletion, energy shortages, migration, and social unrest (Liu et al., 2020; Williams et al., 2020). Vulnerable ecological regions that house some of the world’s most impoverished populations, such as the Third Pole region, are disproportionately affected by climate risks compared to other regions (Pandey et al., 2017). Due to their fragile economic and environmental foundations, these regions face heightened risks and experience more severe disasters than the global averages (Gentle and Maraseni, 2012). Countries within these areas are particularly susceptible to climate change effects, with significant variations observed between nations (Gentle and Maraseni, 2012; Sun et al., 2021). Thus, it is important to understand the climate change vulnerability of smallholder farmers in these countries, consider their spatial differences among the countries, and develop effective adaptation strategies.
Understanding livelihood vulnerability is crucial for updating climate change adaptation strategies (Shah et al., 2013). Vulnerability assessments are integral to evaluating climate-induced risks on society-nature relations and play a pivotal role in enhancing the resilience of rural communities to climate change (Xu et al., 2020). The Intergovernmental Panel on Climate Change (IPCC) defines vulnerability as “the degree to which a system is susceptible to and unable to cope with adverse effects of climate change, including climate variability and extremes”. A typical vulnerability assessment comprises three dimensions: exposure, sensitivity, and adaptation (Thomas et al., 2019). Exposure is a central dimension that refers to “the nature and degree to which a system is exposed to significant climatic variations” (McCarthy et al., 2001; Adger and Agnew, 2004). Sensitivity is the extent to which a system is adversely or favorably affected by climate-related stimuli (McCarthy et al., 2001). Adaptive capacity assessment is generally defined as a function of society, economy, and human capabilities (McCarthy et al., 2001).
While most vulnerability assessment studies reference the IPCC framework, their quantitative assessment methods are not consistent (Reed et al., 2013; Xu et al., 2020). Quantitative assessments often rely on index systems, such as the Livelihood Vulnerability Index (LVI) developed by Hahn et al. (2009) for Mozambique and the LVI-IPCC method used by Mukherjee et al. (2019) for Sagar Island, India. Yang et al. (2021) applied an index system referencing LVI-IPCC to analyze livelihood vulnerability and adaptation strategies in Sichuan Province, China. Zhang and Fang (2020) used LVI-IPCC and the Sustainable Livelihood Index (SLI) to assess vulnerability and sustainability in rural villages in Nepal. However, the lack of standardization in index selection impedes widespread application (Li et al., 2019). The sustainable livelihood framework is based on five dimensions (financial, human, social, physical, and natural capital) and provides a systematic approach for constructing indicator systems (Pandey et al., 2017; Natarajan et al., 2022). For example, Xu et al. (2020) addressed this issue by integrating the sustainable livelihood framework into the vulnerability assessment framework.
The Third Pole region centered around the Himalayas is globally recognized as an economically and ecologically vulnerable area that is threatened by climate change. Himalayan glaciers are melting at an alarming rate, with the region experiencing a warming rate that is double the global average (Yao et al., 2017). This has led to an increased frequency of extreme events, including floods, landslides, water stress, and diseases, significantly impacting the local communities’ natural environment and economy (Adhikari et al., 2020). Climate change has adverse effects on ecosystem functions and the critical goods and services provided by ecosystems (Hurteau et al., 2014). Most of the Third Pole region’s population relies on agriculture, with smallholders being particularly vulnerable to the impacts of climate change (Paul et al., 2019; Nef et al., 2021). Factors such as small scale, low capitalization, and low technological level exacerbate their vulnerability (Shah et al., 2013; Pandey et al., 2017). Additionally, geographical location, topography, and economic, political, and cultural characteristics further influence climate change vulnerability, with variations between countries, districts, and communities (Thornton et al., 2014; Zhang and Fang, 2020).
Several studies in the Third Pole region have assessed vulnerability to climate change and adaptation potential at different scales, considering both perceived and actual responses to climate change (Pandey et al., 2017; Paul et al., 2019; Adhikari et al., 2020). For example, a quantitative assessment of capital-based vulnerability across three districts in Nepal revealed diverse susceptibilities to climate-induced disasters, resulting in significant economic losses (Zhang and Fang, 2020). In the delta region of Myanmar, an assessment found that climate change adaptation measures adopted by farmers are crucial for limiting vulnerability (Oo et al., 2018). Similarly, a study in Vietnam constructed the livelihood vulnerability index for small-scale tea farming households in Lam Dong Province (Thuy et al., 2021).
Most of the existing vulnerability assessment frameworks are tailored to local conditions, hindering cross-country comparisons of smallholders’ vulnerability to climate change. Additionally, many studies evaluating national vulnerability rely on secondary data, which limits their applicability, especially in the diverse climatic zones of the Third Pole region. To address these gaps, this study evaluated and compared the vulnerabilities of Nepal, Cambodia, Thailand and Myanmar in the Pan-Third Pole region based on primary micro-survey data, by applying an extended framework of climate change vulnerability to construct the indicator system. The key factors affecting climate change vulnerability were identified through an analysis of spatial variations in vulnerabilities and policy implications are proposed for coping with climate change in these regions.

2 Data and methods

2.1 Study regions

This study was conducted in the rural regions of four countries in the Third Pole region: Nepal, Cambodia, Thailand, and Myanmar (Fig. 1). These countries contain populations of 29.1 million, 16.0 million, 69.9 million and 55.0 million people, respectively, with agricultural populations accounting for approximately 80%, 85%, 80%, and 60% of their total populations, respectively (Sun et al., 2021). Nepal has a monsoon climate with four main seasons. The terrain is high in the north and low in the south, with mountains to the east, west, and north. Most of the territory is hilly, and land more than 1000 m above sea level accounts for nearly half of the total area. Cambodia has a tropical climate and is characterized by hot weather year-round, with the rainy season occurring from May to mid-November and the dry season from mid-November to April. It is low-lying. Most of the area is occupied by the central Plains, with mountain ranges and highlands around it. Thailand’s landscapes vary from low mountain ranges to fertile alluvial plains, accompanied by extensive beaches. Located in the monsoon region of Asia, Myanmar is a tropical country which slopes from north to south. However, the temperatures are not consistently high throughout the year.
Fig. 1 Locations of the study regions in Nepal, Cambodia, Thailand and Myanmar

2.2 Data collection

The primary data for this study were collected by the United Nations Environment Programme-International Ecosystem Management Partnership (UNEP-IEMP) at the household level in 2019, with the base year set as 2018. Stratified and random sampling methods were employed to select 37 sample villages from 15 counties across the four countries, considering their economic development levels. A total of 1060 households were randomly chosen from the rosters of these villages, and in-person interviews were conducted using a pre-tested, structured questionnaire.
The interviews aimed to collect comprehensive information on the livelihood situations of the participants, typically involving the head of the household as the primary respondent. In cases where the head of the household was unavailable, other family members who are relatively knowledgeable about the household were consulted as alternative respondents. The interviews were conducted by postgraduates who were trained by researchers through professional courses. Before the field survey, simulation tests were conducted to ensure the interviewers’ proficiency in data collection.
To enhance data quality and reliability, each questionnaire was meticulously reviewed by three different interviewers after its completion. This rigorous verification process was implemented to minimize errors and maintain the accuracy of the collected information.

2.3 Methods

2.3.1 Climate change vulnerability assessment framework

In this study, we employed an extended climate change vulnerability assessment framework at the household level, which was established by integrating the sustainable livelihood theory with the IPCC vulnerability assessment method. This framework, originally proposed by Xu et al. (2020), combines the strengths of the sustainable livelihood framework and the IPCC method. To adapt it to the specific characteristics of our study region, we modified the framework and indicator system, allowing for the quantification of vulnerability across different countries and a comprehensive analysis of spatial variations.
To gauge the exposure level to climate change for each household, we used the economic losses incurred due to extreme climate events. Drawing on the sustainable livelihood framework advocated by the Department for International Development (DFID, 1999), the components of family capital were categorized into five types: human capital, social capital, physical capital, natural capital, and financial capital. Natural capital was specifically employed to evaluate household sensitivity to climate change, aligning with the approach proposed by Hahn et al. (2009). Additionally, we selected 24 factors from the other four capital types to evaluate household adaptive capability.

2.3.2 Construction of the vulnerability assessment indicator system

The vulnerability assessment indicator system is comprehensively detailed in Table 1. In this study, the evaluation of exposure centered on household losses attributed to extreme climate events, encompassing eight sub-components. Sensi-tivity, encompassing ten indicators, was assessed considering that livelihoods in developing countries are intimately tied to resource availability. Water and energy availability to farmers were selected as crucial indicators that reflect their reliance on these resources. Additionally, indicators related to the status of land, livestock, and houses of smallholders were incorporated to capture the vulnerability of these capital components to climate change. Demographic characteristics, such as medical expenditure, dependency ratio (ratio of non-working age population to working-age population), and the ratio of individuals older than 65 years, were also considered due to their significant influence on household sensitivity.
Table 1 Classification and weights of factors in the vulnerability assessment
Component Sub-component Factor Functional relationship with vulnerability wi wi’
Exposure Hazardous events Drought +
Flood/heavy rainfall +
Windstorm +
Extreme heat/heat wave +
Extreme cold/cold damage +
Hail +
Mudslide +
Insect pest +
Sensitivity Demographic factors Medical expenditure - 0.025 0.052
Dependency ratio + 0.032 0.066
The elder ratio + 0.018 0.037
Water Tap water + 0.016 0.033
Energy Energy expenditure + 0.025 0.051
Land Land area + 0.025 0.052
Land type 0.034 0.071
Livestock Animal power + 0.076 0.157
Poultry + 0.144 0.301
House Number of houses + 0.087 0.181
Adaptive capacity Human capital Education - 0.035 0.067
Non-agricultural employment - 0.022 0.043
Medical insurance - 0.017 0.032
Endowment insurance (over 16 years old) - 0.016 0.031
Agricultural skills training - 0.030 0.059
Family decision making - 0.019 0.037
Social capital Government - 0.025 0.047
University - 0.025 0.048
Hospital - 0.025 0.048
Enterprise - 0.025 0.047
Physical capital Road - 0.038 0.073
Transportation - 0.034 0.065
Cellphone - 0.034 0.066
Food self-sufficiency - 0.021 0.040
Agricultural species diversity - 0.037 0.071
Financial capital Income - 0.025 0.048
Income resources - 0.056 0.108
Household savings - 0.019 0.037
Agricultural insurance - 0.017 0.033

Note: The weights in this table were calculated based on our survey data. The sources and detailed explanations of the variables are in Table S1 in the supplementary material, which is available on the www.jorae.cn or can be obtained by contacting the author directly.

To evaluate adaptive capacity, the human, social, physical, and financial capital of households were examined. Human capital, considered the bedrock of sustainable livelihoods, was assessed using factors such as household members’ education levels and physical condition. Social capital was gauged based on the number of people employed in critical career areas, recognizing their substantial influence on smallholders’ livelihoods in these countries. The physical capital of smallholders, which is indicative of the level of convenience in their lives and agricultural production activities, was evaluated through factors such as accessibility of roads, transportation, diversity of agricultural production, and possession of communication equipment. Financial capital was represented by the quantity and sources of income, the state of deposits, and the purchase of agricultural insurance. Detailed descriptions of these factors can be found in Table S1.

2.3.3 Methods for vulnerability assessment

To construct a comprehensive vulnerability index, all indicators were normalized to values between 0 and 1. Household exposure intensity at each case point was obtained based on the mean value of the loss of all households in each type of climate change event. Then, the exposure intensities of the various climate change events at each case were summed. Finally, the exposure intensities of these cases were normalized. The calculation process is:
${{E}_{p}}=\left( \frac{\underset{i=1}{\overset{m}{\mathop{\mathop{\sum }^{}}}}\,\underset{j=1}{\overset{n}{\mathop{\mathop{\sum }^{}}}}\,{{E}_{pij}}}{n} \right)/\max \left( \frac{\underset{i=1}{\overset{m}{\mathop{\mathop{\sum }^{}}}}\,\underset{j=1}{\overset{n}{\mathop{\mathop{\sum }^{}}}}\,{{E}_{ij}}}{n} \right)$
where Ep represents the exposure of the p case; Epij represents the intensity for the j household and the i extreme climate event of the p case; m represents the number of different types of disasters; and n represents the number of households in the p case.
A value of 0 or 1 was assigned for each binary variable. For the ratio indicators, such as dependency ratios and the proportion of non-agricultural workers in the family, normalization was carried out as follows:
$N{{S}_{i,j}}=\left\{ \begin{matrix} \frac{{{a}_{i,j}}-\min \left( {{a}_{i}} \right)}{\max \left( {{a}_{i}} \right)-\min \left( {{a}_{i}} \right)},\ \ \text{if}\ N{{S}_{i,j}}\ \text{is}\ \text{a}\ \text{positive}\ \text{indicator} \\ \frac{\max \left( {{a}_{i}} \right)-{{a}_{i,j}}}{\max \left( {{a}_{i}} \right)-\min \left( {{a}_{i}} \right)},\ \ \text{if}\ N{{S}_{i,j}}\ \text{is}\ \text{a}\ \text{negative}\ \text{indicator} \\ \end{matrix} \right.$
where NSi,j represents the final normalized score of the j household’s i indicator; ai,j represents the original score of the j household’s i indicator; and ai represents the scores of the i indicator among all the households.
For continuous variables such as land area, income and other similar variables, 0 was assigned to these variables if their original values were 0. For values greater than 0, they were put in ascending order and then evenly divided into four groups which were assigned new values of 0.25, 0.5, 0.75, and 1, respectively. This transformation was done to eliminate the influence of extreme values and make the different types of variables comparable.
Each indicator was weighted to calculate a comprehensive index after normalization (Table 1). After analyzing the advantages and shortcomings of existing methods for determining the index weights, the inverse variance method (Iyengar and Sudarshan, 1982), which was described in our previous paper (Xu et al., 2020), was used as the ideal optimization scheme in this study.
The weights of indicators were calculated as follows:
${{w}_{i}}=\frac{c}{\sqrt{\operatorname{var}\left( {{x}_{i}} \right)}}$
where wi represents the weight of the indicator xi; and the indicator weights of sensitivity and adaptive capacity were calculated, respectively. c is the constant after standardization, which was calculated as:
$c={{\left[ \underset{i=1}{\overset{m}{\mathop{\mathop{\sum }^{}}}}\,\frac{1}{\sqrt{\operatorname{var}\left( {{x}_{i}} \right)}} \right]}^{-1}}$
where the c values of sensitivity and adaptive capacity were calculated, respectively; and m represents the indicator number.
The weight of each indicator was then divided by the sum of weights of all the indicators within the same subcomponent to determine a new weight for the indicator to obtain the score of each subcomponent:
${{{w}'}_{i}}=\frac{{{w}_{i}}}{\underset{i=1}{\overset{m}{\mathop{\mathop{\sum }^{}}}}\,{{w}_{i}}}$
where ${{{w}'}_{i}}$ represents the new weight of the indicator xi that is used to calculate the score of the subcomponents. The wi values of sensitivity and adaptive capacity were calculated, and are indicated as ${{{w}'}_{si}}$ and ${{{w}'}_{ACi}}$, respectively.
The sensitivity (S) and adaptive capacity score (AC) were calculated as follows:
${{S}_{j}}=\underset{i=1}{\overset{m}{\mathop{\mathop{\sum }^{}}}}\,{{{w}'}_{si}}N{{S}_{i,j}}$
$A{{C}_{j}}=\underset{i=1}{\overset{m}{\mathop{\mathop{\sum }^{}}}}\,{{{w}'}_{ACi}}N{{S}_{i,j}}$
where Sj represents the sensitivity score of the j household; and ACj represents the adaptive capacity score of the j household.
The scores of the subcomponents within each component were summed to obtain the exposure, sensitivity and adaptive capacity scores. The average vulnerability score of all sample households in a country was used to represent the vulnerability of that country. The vulnerability (V) of each household was calculated as:
V=E+S–AC
where, E represents the exposure index; S represents the sensitivity index; and AC represents the adaptive capacity index.

2.4 Data analysis

All statistical analyses for this study were conducted using Microsoft Excel and Stata 15.0. The figures were generated using ArcGIS 10.0 and Origin 2022. This combination of software tools ensured a robust and comprehensive approach to data analysis and visualization in this study.

3 Results

3.1 Climate change vulnerability of the four countries

Clear differences in vulnerability were found (Fig. 2), which depended mainly on the levels of exposure, sensitivity, and adaptive capacity in each of the four countries. Notably, Cambodia exhibited the highest vulnerability (0.73), surpassing Thailand (0.165), Myanmar (0.056), and Nepal (0.009). The variations in the exposure, sensitivity, and adaptive capacity scores were significant and attributed to regional disparities in climate, geological landforms, infrastructure construction, and agricultural production levels.
Fig. 2 Climate change vulnerability assessment and composition for each of the four countries
In terms of exposure, Cambodia obtained the highest score (0.79), which was markedly higher than Thailand (0.36), Nepal (0.16), and Myanmar (0.13). Sensitivity scores followed the order of Nepal (0.20) > Cambodia (0.19) > Thailand (0.18) > Myanmar (0.16). The adaptive capacity varied between 0.25 and 0.37, with Thailand leading (0.37), followed by Nepal (0.34), Myanmar (0.29), and Cambodia (0.25).
These results underscore the diverse vulnerabilities across the sampled countries, emphasizing the need for targeted and context-specific adaptation strategies that consider the unique challenges posed by exposure, sensitivity, and adaptive capacity in each nation.

3.2 Exposure assessment of the four countries

The primary factors contributing to domestic exposure exhibited significant variations among the countries (Fig. 3). The complexity of factors influencing the overall exposure level was notably pronounced in Nepal, where a diverse range of hazards had occurred, including droughts, floods, wind, extreme heat/heat waves, extreme cold, hail, mudslides, and insect pests. In Cambodia, exposure was primarily attributed to drought and insect plagues, along with extreme heat/heat waves, floods, and wind. In Thailand and Myanmar, exposure was linked to droughts and floods, and mudslides, respectively.
Fig. 3 Intensities of extreme climate events in each of the four countries
From a broader perspective, insect pests were a prevalent hazard faced by all four countries, with Cambodia experiencing a significantly higher severity (0.3385) compared to Nepal (0.0225), Thailand (0.0114), and Myanmar (0.0021). Droughts and floods played pivotal roles in the exposure levels across the studied regions, excluding Myanmar. Drought conditions were more severe in Cambodia (0.2765) and Thailand (0.1975) than in Nepal (0.0476), while floods were more common in Thailand (0.1542) and Cambodia (0.0568). Additionally, mudslides emerged as the predominant hazard contributing to exposure in Myanmar (0.1271).
In summary, while smallholders in the Pan-Third Pole region face risks such as flooding and mudslides resulting from shifting snowlines and increases in river runoff due to climate change, they also confront significant losses due to drought conditions. This underscores the multifaceted nature of the climate-related risks faced by smallholders in the region.

3.3 Sensitivity assessment of the four countries

The sensitivity assessment was conducted based on the six sub-components of water, demographic factors, houses, livestock, energy, and land, which encompass 10 specific factors (Fig. 4 and Table 1). The overall sensitivity score exhibited a decreasing order of Nepal, Cambodia, Thailand, and Myanmar, with only marginal differences among the countries, typically around 0.180 (ranging from 0.164 to 0.195). However, significant heterogeneity was observed in the scores of the individual sensitivity factors among the four countries, as shown in Fig. 4.
Fig. 4 Compositions of sensitivity in the four countries

Note: Scale units represent the sensitivity score of the sampled countries.

The most substantial difference was identified in the demographic factors, including medical expenditure, dependency ratio, and the ratio of elderly individuals over 65 years old. Cambodia obtained the highest score (0.073), with Myanmar following closely at 0.066. Noteworthy intercountry disparities were also evident in land (total area of land owned by households) and tap water accessibility. For land, Nepal (0.055) and Thailand (0.044) secured the top two positions, while Cambodia (0.038) slightly exceeded Myanmar (0.031). Regarding demographic factors, Cambodia (0.073) and Myanmar (0.066) exhibited similar scores that were significantly higher than those of Thailand (0.031) and Nepal (0.025). The scores for houses, livestock, and energy in all four countries were relatively similar, hovering around 0.040, 0.012, and 0.014, respectively. The accessibility of tap water and owned land emerged as two primary sources of sensitivity in Thailand and Nepal, whereas the greater vulnerability in demographic factors in Cambodia and Myanmar minimized the overall intercountry differences.

3.4 Adaptive capacity assessment of the four countries

3.4.1 Differences in adaptive capacity at the subcomponent level

The assessment of adaptive capabilities at the household level used the sustainable livelihood framework, with indicators categorized into four subcomponents: human capital, social capital, physical capital, and financial capital (Fig. 5). Substantial differences were observed among the four countries at the subcomponent level, with the major factors contributing to overall differences identified as financial capital and human capital, which exhibited larger intercountry variations than physical and social capital.
Fig. 5 Sources of adaptive capacity among the four countries
Thailand (0.131) attained the highest score for human capital, followed by Nepal (0.099), Cambodia (0.065), and Myanmar (0.062). Conversely, the financial capital score decreased in the order of Thailand (0.072), Nepal (0.066), Myanmar (0.061), and Cambodia (0.036). Comparatively minor differences were noted in physical capital between Nepal (0.128), Thailand (0.124), Myanmar (0.112), and Cambodia (0.101). Social capital exhibited no notable differences among the countries, with all scores around approximately 0.049.
In summary, Thailand and Nepal demonstrated relatively higher adaptive capabilities among the four countries, while Myanmar and Cambodia exhibited lower capabilities for adapting to climate change. These findings emphasize the importance of considering financial and human capital as the key factors influencing adaptive capacity, thereby highlighting potential areas for targeted interventions and policy measures to enhance climate resilience.

3.4.2 Differences in adaptive capacity at the factor level

An assessment of the individual factors was conducted to further analyze the differences in the adaptive capacities of households, which revealed significant variations at the factor level (Fig. 6). Notably, Thailand stood out as superior to the other countries in terms of human capital, boasting considerably higher scores for endowment insurance (0.112) and medical insurance (0.118) (Fig. 6a). In contrast, Cambodia (0.009 and 0.004) and Myanmar (0 and 0) displayed low proportions of family members with medical and endowment insurance, coupled with noticeable disadvantages in education, resulting in substantially lower scores in human capital.
Fig. 6 Radar maps of the subcomponents of adaptive capacity in the four countries

Note: (a), (b), (c), and (d) show the inter-country differences in human capital, social capital, physical capital, and financial capital, respectively.

The analysis of physical capital indicated that the highest scores (ranging from 0.172 to 0.210) were found in the use of cellphones, with similar scores across all four countries (Fig. 6c). However, variations were observed in other factors, including roads, agricultural species diversity, food self-sufficiency, and transportation.
For financial capital, the households in Nepal, Cambodia, Thailand, and Myanmar demonstrated multiple income resources, with distinct differences between the countries (Fig. 6d). Factors other than income were similar across these four countries, but also exhibited variations by country. On the other hand, the social capital of the four countries showed high homogeneity (Fig. 6b).

4 Discussion

Assessments of livelihood vulnerability are crucial for decision-making authorities who strive to enhance regional adaptation to climate change and achieve the Sustainable Development Goals (SDGs) (Adger and Agnew, 2004). Multivariate analyses can provide a comprehensive understanding of the primary factors influencing vulnerability in a region. The vulnerability of households to climate change is generally determined by exposure, sensitivity, and adaptive capacity (Reed et al., 2013).
Our results show that the highest vulnerability of households to climate change is in Cambodia, mainly due to a high exposure level and the lowest adaptive capacity. Additionally, the relatively higher exposure levels make Thailand more vulnerable than Nepal and Myanmar, with the differences in sensitivity and adaptive capability between these three countries being relatively small. Therefore, exposure plays a significant role in the variations of vulnerability among the four countries. Regional differences are common in vulnerability assessments, but the contribution of exposure varies depending on the actual situation (Zhang and Fang, 2020). Changes in exposure often become the main factor influencing vulnerability when there are considerable differences in terrain, landforms, and climates between study regions (Das et al., 2020; Xu et al., 2020).
Household losses in ecologically and economically vulnerable areas can be influenced by a myriad of climatic hazards, and the high uncertainty in the behavior of extreme climatic events is a global phenomenon with limited measures for mitigation (Sun et al., 2021). This study identified droughts, floods, rainfall, windstorms, heatwaves, extreme cold damage, hail, mudslides, and insect pests as major hazards resulting from different climate events. These hazards, which are highly dependent on geography and regional climate, can impose stress on rural livelihoods by diminishing the existing livelihood options (Yang et al., 2021). Warm and humid environments are favorable for insects, which pose a significant threat to agricultural production (Babendreier et al., 2019). Additionally, the floods in Thailand are unpredictable in quantity and frequency, and have been exacerbated by deforestation and land conversion to farming (Singkran, 2017). Myanmar features a mountain plateau and basin with precipitation reaching 3000-5000 mm, and it experiences frequent mudslides (Fig. 3). Nepal faces various environmental challenges, but their intensities appear mild due to its unique location along the southern slopes of the Himalayan ranges, so it is relatively more suitable for daily living (Gentle et al., 2012). Non-climatic factors, such as the availability of irrigation facilities, can indirectly cause household losses along with climatic factors (Yang et al., 2021). The over-reliance of Thailand and Cambodia on water resources for agriculture significantly impacts their water budgets (Costanza et al., 2011). Diminishing natural resources, limited irrigation, and the expansion of agricultural systems further aggravate the impacts of drought (Thilakarathne and Sridhar, 2017).
In this study, sensitivity was defined as the extent to which households in the study area are sensitive to climate change. These hazards directly impact livelihoods, primarily by affecting the physical and natural capital of livelihoods (Zhang et al., 2020). Agricultural production is crucial for smallholders’ livelihoods, but it has become increasingly uncertain due to climatic hazards, especially in rain-fed agriculture and subsistence areas (Gentle and Maraseni, 2012). Smallholders with more assets, such as land and houses, are becoming more sensitive to climate hazards, whereas this scenario is less applicable to some large farmers who can compensate for losses through various means (Xu et al., 2020). Advances in demographic factors in Cambodia and Myanmar compensated for their lower scores in land and tap water access, minimizing the intercountry differences in this study. High accessibility of tap water and dependence on energy can increase households' susceptibility to natural hazards in less developed regions, making them more vulnerable to hazard-induced interruptions in the tap water or energy supplies (Xu et al., 2020). Almost all households in Thailand and approximately 95% in Nepal now have access to improved water sources (Thilakarathne and Sridhar, 2017). In Cambodia, approximately 13% of the population lacks access to tap water, and most people in Myanmar do not depend on tap water but on private wells, according to World Bank data.
Households with more private land, livestock, and houses would suffer greater losses when disasters occur, making them more sensitive to climate change and related hazards (Reed et al., 2013). The areas of cultivated land account for 18.1%, 20.6%, 35.2%, and 16.7% of the total in Nepal, Cambodia, Thailand, and Myanmar, respectively, according to World Bank data. Factors such as medical expenditure, dependency ratio, and the ratio of people over 65 years of age are domestic factors commonly perceived to contribute to the increase in household sensitivity (Reed et al., 2013). The proportions of people older than 65 years were 4.4%, 3.8%, 9.2%, and 5% in Nepal, Cambodia, Thailand, and Myanmar by 2017, respectively, according to World Bank data. The percentage of people with social insurance was above 95% for Thailand and about 50% for Nepal. However, Cambodia had an insurance penetration rate of only approximately 1.04% of the population, and it was less than 5% for Myanmar. These demographic differences provide evidence that supports our results, indicating that demographic factors, land, and tap water accessibility are the major contributors to the differences in sensitivity among these countries.
Adaptive capacity is closely tied to the availability and accessibility of infrastructure and public services (Hahn et al., 2009). Human capital and natural capital play significant roles in the adaptive capacity of households to climate change (Pandey et al., 2017). Employability, particularly for youth, depends on language, technical, and vocational skills, and poorly educated laborers are less likely to find employment (Zhang et al., 2020). The advantages of financial capital, especially human capital in Thailand and Nepal, help to moderate the impacts on farmers’ livelihoods. Cambodia and Myanmar are more vulnerable due to their relatively low human capital, which can be attributed to disadvantages in medical insurance, agricultural skills training, and education in rural areas (Fig. 6a). The inadequate educational system in rural areas of these countries results in less than 10% of adolescents receiving higher levels of education.
Inter-country differences are also significant in terms of the four sub-components of financial capital. Financial capital plays a crucial role in determining the livelihood adaptive capacities in a changing climate (Pandey et al., 2017). It includes income, income sources, household savings, and agricultural insurance, which are crucial for achieving people’s livelihood objectives and responding to extreme climatic events. High incomes, substantial savings, and diverse income sources can help households overcome disasters (Gentle et al., 2012). Agricultural insurance offsets part of the household losses caused by disasters (Das et al., 2020). However, most households in Cambodia lack agricultural insurance or savings (Fig. 6d), which makes Cambodia the most vulnerable to financial capital. This conclusion aligns with the findings of previous studies (Mendoza et al., 2014).
The differences at the country level are relatively small in terms of the five aspects of physical capital and the four aspects of social capital, according to our assessment. Physical capital refers to the basic infrastructure and resources required to maintain livelihoods (Pandey et al., 2017). Roads, transportation, and communication tools are fundamental for household production and living. Food insecurity throughout the year is a major impact of climate change in these regions (Pandey et al., 2017). Smallholders' most common strategies for coping with the negative impacts of various climatic and non-climatic factors include reducing food consumption or changing crop varieties (Mendoza et al., 2014). Differences in agricultural policies, infrastructure construction, and market development between regions contribute to each country's unique advantages, except for cellphones, which may be attributed to the relatively uniform level of communication facility construction in the rural areas of this region (Mendoza et al., 2014; Sarker et al., 2019).
Social capital is the social resources that people leverage to pursue their livelihood goals within the context of a sustainable livelihood framework (Xu et al., 2020). Having more friends and relatives working in key areas is widely believed to provide more support, as social networks are necessary not only for agricultural inputs and selling agricultural commodities but also for sharing climate and market information (Oo et al., 2018). Vulnerable social groups have less access to social organizations and community institutions compared to non-vulnerable groups. The small differences in social capital observed in the selected countries could be attributed to the gaps between urban and rural areas, which limit employment opportunities in critical career areas for the rural population (Bosworth and Venhorst, 2018).
It is worth noting that numerous challenges, such as organizational and coordination issues, hinder data collection in multinational areas, and these challenges were especially exacerbated in this study by the impact of COVID-19. Consequently, our data may not have promptly captured the adverse effects of climate change on livelihoods in the sampled countries.
Given that this study only sampled four countries in the Third Pole region for the identification and comparison of smallholders’ vulnerability to climate change, it cannot fully depict the status throughout this expansive region. Enhancing the representativeness of the study will require the inclusion of more countries and additional samples.
There are limitations in measuring the exposure level in this study. The exposure level of smallholders can be gauged through either meteorological monitoring or questionnaire surveys. While meteorological monitoring provides more objective results based on environmental data, it only captures the average exposure level of a region while questionnaire surveys excel at reflecting individual differences. Achieving a more comprehensive analysis of smallholders’ exposure to climate change calls for the synthesis of these two measurement methods in future studies.

5 Conclusions and policy implication

The set of indicators for household vulnerability analysis structured around the five livelihood capitals has proven to be a powerful tool for understanding inter-country differences in exposure, sensitivity, and adaptive capability at the household level. The overall vulnerabilities of Nepal, Cambodia, Thailand, and Myanmar significantly differed across the sustainable livelihood components and the LVI-IPCC criteria. Exposure scores revealed large inter-country differences among the four countries, and inter-country differences in sensitivity and adaptive capability at the sub-component level cannot be ignored. The heterogeneity observed in human capital and financial capital-related factors underlines the comprehensive analysis from the perspective of sustainable livelihood to determine household adaptive capacity. The regional differences in environmental, socioeconomic, and demographic drivers identified using these assessments may allow policymakers in the Third Pole region and other vulnerable areas to develop effective solutions for common vulnerability problems based on the experiences of other countries, contributing further to achieving SDGs (including SDG 1, 2, 3, 4, 5, 6, 10, and 11) in vulnerable areas worldwide.
We propose several specific policy measures to reduce vulnerability by improving household capital protection in the different countries. In countries with relatively higher exposure levels, such as Cambodia and Thailand, strengthening existing extension activities that provide climate change information to rural populations and enhancing specific disaster prevention and protection capabilities of local communities are needed. To effectively reduce sensitivity, targeted policy measures are also necessary, such as consummating the social security systems in Cambodia and Myanmar, although this is less urgent in Nepal and Thailand. Considering the high heterogeneity in the dominant factors of the four capitals that determine smallholders’ adaptive capacity, local governments should adjust their policies for household capital protection and improvement according to their specific conditions.
In addition to the recommendations for domestic policy, there is a call for targeted external investments from international organizations, including UNDP, UNEP, and the World Bank, given that Nepal, Cambodia, Thailand, and Myanmar are developing countries located in the most vulnerable regions. International aid focusing on human, social, physical, and financial capitals is highly recommended. External investments in infrastructure construction, public health, public education, and technical assistance should be helpful for improving the adaptability of smallholders in this special region.
We believe that this study makes a significant contribution to climate change vulnerability assessment, and inter- national comparisons are essential for effective climate adaptation strategies and achieving sustainable development goals.

Availability of data and materials

The data were collected by the authors, and it has been uploaded to National Tibetan Plateau Data Center. The data can be provided upon request. Anyone can search and apply for the data through the website of the Data Center (http://data.tpdc.ac.cn) following the underlying information below:
1. Zhang L X, Bai Y L. 2021. Dataset of sustainable livelihood: Demographic, human capital, and employment (2018). National Tibetan Plateau Data Center, DOI: 10.11888/Socioeco.tpdc.271139. CSTR: 18406.11.Socioeco.tpdc.271139.
2. Zhang L X, Bai Y L. 2021. Dataset of sustainable livelihood: Land endowment (2018). National Tibetan Plateau Data Center, DOI: 10.11888/Socioeco.tpdc.271141. CSTR: 18406.11.Socioeco.tpdc.271141.
3. National Tibetan Plateau Data Center. 2018. Dataset of sustainable livelihood: Social capital (2018). National Tibetan Plateau Data Center, DOI: 10.11888/Socioeco.tpdc.271142.CSTR:18406.11.Socioeco.tpdc.271142.
4. Zhang L X, Bai Y L. 2021. Dataset of sustainable livelihood: Income (2018). National Tibetan Plateau Data Center, DOI: 10.11888/Socioeco. tpdc.271144. CSTR: 18406.11.Socioeco.tpdc.271144.
5. Zhang L X, Bai Y L. 2021. Dataset of sustainable livelihood: Financial capital (2018). National Tibetan Plateau Data Center, DOI: 10.11888/ Socioeco.tpdc.271143. CSTR: 18406.11.Socioeco.tpdc.271143.
6. Zhang L X, Bai Y L. 2021. Dataset of sustainable livelihood-Public infrastructure (2018). National Tibetan Plateau Data Center, DOI: 10.11888/Socioeco.tpdc.271140. CSTR: 18406.11.Socioeco.tpdc.271140.

References

Publishing order | Descend order by publishing year | Descend order by cited within
[1]
Adger W N, Agnew M. 2004. New indicators of vulnerability and adaptive capacity. Tyndall project IT1.11: July 2001 to June 2003. Final project report. Norwich, England: University of East Anglia.

[2]
Adhikari S, Dhungana N, Upadhaya S. 2020. Watershed communities’ livelihood vulnerability to climate change in the Himalayas. Climatic Change, 162(3): 1307-1321.

[3]
Babendreier D, Wan M, Tang R, et al. 2019. Impact assessment of biological control-based integrated pest management in rice and maize in the greater Mekong subregion. Insects, 10(8): 226. DOI: 10.3390/insects10 080226.

[4]
Bosworth G, Venhorst V. 2018. Economic linkages between urban and rural regions—What’s in it for the rural? Regional Studies, 52(8): 1075-1085.

[5]
Casado-Asensio J, Steurer R. 2014. Integrated strategies on sustainable development, climate change mitigation and adaptation in Western Europe: Communication rather than coordination. Journal of Public Policy, 34(3): 437-473.

[6]
Costanza R, Kubiszewski I, Paquet P, et al. 2011. Planning approaches for hydropower development in the lower Mekong Basin. Portland, USA: Institute for Sustainable Solutions, Portland State University. http://www.pdx.edu/sustainability/lower-mekong-report.

[7]
Das M, Das A, Momin S, et al. 2020. Mapping the effect of climate change on community livelihood vulnerability in the riparian region of Gangatic Plain, India. Ecological Indicator, 119: 106815. DOI: 10.1016/j.ecolind.2020.106815.

[8]
Gentle P, Maraseni T N. 2012. Climate change, poverty, and livelihoods: Adaptation practices by rural mountain communities in Nepal. Environmental Science & Policy, 21: 24-34.

[9]
Hahn M B, Riederer A M, Foster S O. 2009. The Livelihood Vulnerability Index: A pragmatic approach to assessing risks from climate variability and change—A case study in Mozambique. Global Environmental Change, 19(1): 74-88.

[10]
Hurteau M D, Bradford J B, FuléP Z. 2014. Climate change, fire management, and ecological services in the southwestern US. Forest Ecology and Management, 327: 280-289.

[11]
Iyengar N S, Sudarshan P A. 1982. A method of classifying regions from multivariate data. Economics & Politics, 17: 2047-2052.

[12]
Jiang H, Yu Y, Chen M M, et al. 2021. The climate change vulnerability of China: Spatial evolution and driving factors. Environmental Science and Pollution Research, 28: 39757-39768.

[13]
Kumar P, Geneletti D, Nagendra H. 2016. Spatial assessment of climate change vulnerability at city scale: A study in Bangalore, India. Land Use Policy, 58: 514-532.

[14]
Li L, Cao R, Wei K, et al. 2019. Adapting climate change challenge: A new vulnerability assessment framework from the global perspective. Journal of Cleaner Production, 217: 216-224.

[15]
Liu Y, Li N, Zhang Z, et al. 2020. Climate change effects on agricultural production: The regional and sectoral economic consequences in China. Earths Future, 8(9): e2020EF001617. DOI: 10.1029/2020EF001617.

[16]
McCarthy J J, Canziani O F, Leary N A, et al. 2001. Climate change 2001: Impacts, adaptation, and vulnerability:Contribution of Working Group II to the third assessment report of the Intergovernmental Panel on Climate Change (Vol. 2). Cambridge, UK: Cambridge University Press.

[17]
Mendoza M E, Naret H, Ballaran V Jr, et al. 2014. Assessing vulnerability to climate change impacts in Cambodia, the Philippines and Vietnam: An analysis at the commune and household level. Journal of Environmental Science and Management, 17(2): 78-91.

[18]
Mukherjee N, Siddique G, Basak A, et al. 2019. Climate change and livelihood vulnerability of the local population on Sagar Island, India. Chinese Geographical Science, 29(3): 417-436.

DOI

[19]
Natarajan N, Newsham A, Rigg J, et al. 2022. A sustainable livelihoods framework for the 21st century. World Development, 155: 105898. DOI: 10.1016/j.worlddev.2022.105898.

[20]
Nef D P, Neneth D, Dini P, et al. 2021. How local communities attribute livelihood vulnerabilities to climate change and other causes: A case study in North Vanuatu. Climatic Change, 168(3): 1-20.

[21]
Oo A T, Huylenbroeck G V, Speelman S. 2018. Assessment of climate change vulnerability of farm households in Pyapon District, a delta region in Myanmar. International Journal of Disaster Risk Reduction, 28(6): 10-21.

[22]
Pandey R, Jha S K, Alatalo J M, et al. 2017. Sustainable livelihood framework-based indicators for assessing climate change vulnerability and adaptation for Himalayan communities. Ecological Indicators, 79: 338-346.

[23]
Paul A, Deka J, Gujre N, et al. 2019. Does nature of livelihood regulate the urban community’s vulnerability to climate change? Guwahati City: A case study from North East India. Journal of Environmental Management, 251: 109591. DOI: 10.1016/j.jenvman.2019.109591.

[24]
Reed M S, G Podesta, I Fazey, et al. 2013. Combining analytical frameworks to assess livelihood vulnerability to climate change and analyse adaptation options. Ecological Economics, 94: 66-77.

PMID

[25]
Sarker M N I, Wu M, Alam G M, et al. 2019. Livelihood vulnerability of riverine-island dwellers in the face of natural disasters in Bangladesh. Sustainability, 11(6): 1-23.

[26]
Shah K U, Dulal H B, Johnson C, et al. 2013. Understanding livelihood vulnerability to climate change: Applying the livelihood vulnerability index in Trinidad and Tobago. Geoforum, 47: 125-137.

[27]
Sharma J, Ravindranath N H. 2019. Applying IPCC 2014 framework for hazard-specific vulnerability assessment under climate change. Environmental Research Communications, (1): 051004. DOI: 10.1088/2515-7620/ab24ed.

[28]
Singkran N. 2017. Flood risk management in Thailand: Shifting from a passive to a progressive paradigm. International Journal of Disaster Risk Reduction, 25: 92-100.

[29]
Sun M, Xu X, Wang L, et al. 2021. Stable energy, energy inequality, and climate change vulnerability in Third pole regions: Empirical analysis in cross-national rural areas. Renewable & Sustainable Energy Reviews, 147. DOI: 10.1016/j.rser.2021.111197.

[30]
Thilakarathne M, and Sridhar V. 2017. Characterization of future drought conditions in the Lower Mekong River Basin. Weather and Climate Extremes, 17: 47-58.

[31]
Thomas K, Hardy R D, Lazrus H, et al. 2019. Explaining differential vulnerability to climate change: A social science review. Wiley Interdisciplinary Reviews: Climate Change, (2). DOI: 10.1002/WCC.565.

[32]
Thornton P K, Ericksen P J, Herrero M, et al. 2014. Climate variability and vulnerability to climate change: A review. Global Change Biology, 20(11): 3313-3328.

DOI PMID

[33]
Thuy N N, Anh H H. 2021. Determinants of livelihood vulnerability of smallholder tea farmers in Lam Dong Province, Vietnam. IOP Conference Series: Earth and Environmental Science, 653(1): 012094. DOI: 10.1088/1755-1315/653/1/012094.

[34]
Williams D S, Celliers L, Unverzagt K. 2020. A method for enhancing capacity of local governance for climate change adaptation. Earths Future, 8(7). DOI: 10.1029/2020EF001506.

[35]
Xu X, Wang L, Sun M, et al. 2020. Climate change vulnerability assessment for Smallholders in China: An extended framework. Journal of Environmental Management, 276: 111315. DOI: 10.1016/j.jenvman.2020.111315.

[36]
Yang X, Guo S, Deng X, et al. 2021. Study on livelihood vulnerability and adaptation strategies of farmers in areas threatened by different disaster types under climate change. Agriculture, 11(11): 1088. DOI: 10.3390/agriculture11111088.

[37]
Yao T D, Chen F H, Cui P, et al. 2017. From Tibetan Plateau to Third Pole and Pan-Third Pole. Bulletin of Chinese Academy of Sciences, 32(9): 924-931. (in Chinese)

[38]
Zhang C, Fang Y. 2020. Application of capital-based approach in the measurement of livelihood sustainability: A case study from the Koshi River basin community in Nepal. Ecological Indicators, 116(3): 106474. DOI: 10.1016/j.ecolind.2020.106474.

Options
Outlines

/

〈 〉