Farmers’ livelihood adaptation to environmental change in an arid region: A case study of the Minqin Oasis, northwestern China

Ecological Indicators 93 (2018) 411–423

Contents lists available at ScienceDirect

Ecological Indicators journal homepage: www.elsevier.com/locate/ecolind

Original Articles

Farmers’ livelihood adaptation to environmental change in an arid region: A case study of the Minqin Oasis, northwestern China Jia Chena,b, Sha Yina, Hans Gebhardtc, Xinjun Yanga,b,

T

⁎

a

College of Urban and Environmental Sciences, Northwest University, Xue fu Ave.1, Xi’an 710127, China Shaanxi Key Laboratory of Earth Surface System and Environmental Carrying Capacity, Northwest University, Xi’an 710127, China c Department of Geography, University of Heidelberg, Berliner Str. 48, Heidelberg 69120, Germany b

A R T I C LE I N FO

A B S T R A C T

Keywords: Adaptation strategies Livelihood Government policy Adaptive capacity Environmental change Adaptation outcome

Adaption to global environmental change is a focus of sustainability research. Farmers face multiple environmental and social pressures due to global environmental change. Effective livelihood changes must be taken to decrease asset losses and to adapt to current or future environmental challenges. However, there are few studies that systematically understand and assess farmers’ livelihood adaptation. We developed an integrated analytical framework for livelihood adaptation and explored the relationships between adaptive capacity, adaptation outcomes, and farmers’ adaptation strategies. We applied this framework to a case study of the Minqin Oasis in China and assessed the livelihood adaptation of farmers facing environmental change and water scarcity. Household surveys and semi-structured interviews were used for data collection. We found that (1) farmers’ livelihood adaptation choices were limited due to current government policies and their own resources and (2) livelihood adaptive capacity (such as land, water resources, and social networks) and policy reform (water resource fees, and cultivated land compression) had a key impact on farmers’ adaptation. The factors representing a poor livelihood strategy and adaptation outcomes of the farmer include the following: (1) a low level of livelihood awareness among farmers (such as passive farmers), (2) a lack of livelihood assets, (3) government focus on environmental recovery, and (4) a weakened role of the market. To improve the adaptation of farmers’ livelihoods to environmental change, these measures must balance the relationship between environmental restoration and farmers’ livelihoods, consider a variety of key forces, and guide farmers to adopt effective strategies. This study facilitates the development of livelihood adaptation analysis methods for global change studies. Case-based research results can be used to improve local decision-making and can provide an assessment reference method for farmer adaptation to regional and global environmental change.

1. Introduction The impact of global climate change has drawn the attention of the international community over the past decade (Adger et al., 2005; IPCC, 2014; John Smithers, 1997; Turner et al., 2003). The aggravation of drought and the scarcity of water resources have become primary reasons to restrict the sustainable development of populations, especially in arid and semi-arid regions. Water scarcity is expected to be a major challenge for most people in Asia in the future (IPCC, 2014). Northwestern China is a sensitive area in the global ecological environment and is susceptible to climate change. The frequency of extreme drought events has increased significantly in recent years (Zhang et al., 2015). With the arid geographical environment and the scarcity of resources, the rural area of northwestern China has become a region that is highly vulnerable to climate change. At the same time, the

⁎

activities of human beings (such as excessive reclamation, excessive extraction of groundwater, etc.) in the past have exacerbated environmental degradation in arid regions of northwestern China (Danfeng et al., 2006; Feike et al., 2017) and have increased desert expansion, groundwater depletion, water quality degradation, deforestation, biodiversity loss, and natural disasters (Zhang et al., 2010). In the process of environmental degradation, land and water resources were reduced, resulting in negative impacts on the agricultural economy. Rural development has been seriously hampered, especially threatening farmers’ livelihoods and sustainable growth. Adapting to arid environments and a lack of waters resources have become a core issue in many areas of developing countries, including northwestern China, which is also related to the sustainable development of local rural communities. Livelihood issues have been central to rural development and

Corresponding author at: College of Urban and Environmental Sciences, Northwest University, Xue fu Ave.1, Xi’an 710127, China. E-mail address: [email protected] (X. Yang).

https://doi.org/10.1016/j.ecolind.2018.05.017 Received 30 June 2017; Received in revised form 5 May 2018; Accepted 7 May 2018 1470-160X/ © 2018 Elsevier Ltd. All rights reserved.

Ecological Indicators 93 (2018) 411–423

J. Chen et al.

interaction of the coupled human-environment system from the perspective of vulnerability, resilience and adaptation of systems (Adger, 2006; Holling, 2001; Smit and Wandel, 2006). Among them, combined study of resilience and adaptation shows that the promotion of adaptive capacity in a system enhances the adaptation of the system subject confronting external changes and then promotes the resilience of the damaged system amidst the disaster or environmental risk (Nelson, 2011). Therefore, adaptation (including adaptive capacity) plays an important role in buffering the human system coping with the vulnerability to climatic and environmental changes. International institutions and scholars (e.g. Chen et al., 2014; Grothmann and Patt, 2005; Shinn, 2016; Butler et al., 2014; Li et al., 2014; Perry et al., 2010) have carried out much research and practices. Analyses of adaptation to changing climatic conditions have been undertaken for a variety of purposes, such as recovery from natural disasters, policies responding to climate change in a region and society, resource adaptation management, etc. (McCubbin et al., 2015; Pandey et al., 2011; Turner et al., 2003). With the accumulation of research and practical observations of adaptation, a series of methods and analytical frameworks for adaptation have risen. However, due to the dependence of climate change adaptation on the theory of resilience and vulnerability and differences in the definition of adaptation, it is difficult to separate and form a unified research framework. Existing research methods and analytical frameworks (see Dessai et al., 2005; Parry and Carter, 1998) have been fundamentally considered from complex human-environmental systems or based on a top-down approach to analyze the potential risk impacts and to develop adaptive measures through downscaling model calculations. Owing to the systematic analysis guide of the impact of climate change, the research scale has focused on the global, national, regional level, and complex integrated systems. In recent years, there have also been a series of different adaptive analyses frameworks and methods in agricultural systems, resource systems, and disaster assessment research based on the background of climate change (Chen et al., 2016; Kalaugher et al., 2013; Pandey et al., 2011; Warrick et al., 2016), while inadequate attention has been paid to the micro social systems adapting to the risk. Actor-based analysis looks at adaptation as the decision-making process (Nelson et al., 2007). The ultimate goal of adaptation research is to determine how humans (such as households, inhabitants or farmers, etc.) adapt to changes in order to improve resilience to the environment and disasters. Accordingly, research into the adaptive behavior and decision-making characteristics of humans needs to be further strengthened, especially the adaptive capacity of the community and on the farmers’ scale. Table A.1 in Appendix A presents a summary of previous studies on livelihood (human) adaptation. The initial study of human activity adaptation focused on proposing a framework for understanding human adaptation to climate events or risk disasters (John Smithers, 1997; Mortimore and Adams, 2001; Smit et al., 1999), but the quantitative analysis of livelihood adaptation is not very operational. Subsequent research on adaptation also focused on the impact of social factors on the farmers’ livelihoods adaptation (Carr, 2008; Osbahr et al., 2008; Thomas and Twyman, 2005). However, many studies lack systemic thinking and focus adaptation analysis only on the social structural elements. At present, under the background of environmental (climate) change, research on farmers' adaptation is mainly based on the adaptive behavior and strategy selection (Alauddin and Sarker, 2014; Reed et al., 2013). Researchers tend to use qualitative research to understand the adaptive behavior of households and analyze influencing factors (Hoque et al., 2017; Nicholas and Durham, 2012), ignoring the inherent logical relationship between adaptive capacity and adaptive outcomes. Many studies still follow the Sustainable Livelihoods Framework established by the DFID (Department for International Development) (2000) (Khayyati and Aazami, 2016; Motsholapheko et al., 2011). Therefore, current studies lack a comprehensive research framework, combined with livelihood options, in

practice over the past decade (Khayyati and Aazami, 2016; Oberlack et al., 2016; Scoones, 2009). Rural areas occupy the vast majority of arid regions in northwestern China. As the basic livelihood unit in rural society, farmers bear multiple risks resulting from climatic (environmental) changes and socio-economic policies. Drought, land desertification and other risks of disturbance will undoubtedly increase the vulnerability of farmers’ livelihoods. In rural communities, where access to income is limited, various agriculture-related activities, which strongly depend on soil and water, can contribute significantly to livelihood security (Khayyati and Aazami, 2016). Therefore, drought and water resources scarcity become the main influencing factors restricting the livelihood maintenance of farmers and threatening social welfare (Alam, 2015). Due to its location in the desert and the temperate continental arid climate, Minqin County is a typical arid region in China. With the disruption of global climate change and human activities, the runoff flowing into the Minqin Region decreased. Because water resources are scarce and the groundwater is massively pumped, a large area of natural vegetation has been destroyed causing the desert to extend to the oasis (Yin et al., 2016). The original fragile arid environment has sharply deteriorated, which seriously threatens local food production, water security, public health, natural resources, and biological diversity. More than 70 thousand people and 120 thousand livestock have great difficulty accessing drinking water. More than 200 million m2 of farmland have already been abandoned, leading to severe challenges to the survival and livelihood security of farmers; furthermore, some farmers have become “ecological refugees” in the region (Zhang et al., 2011; Zhao et al., 2015). Therefore, it is of urgent practical significance to mitigate the risks of climatic and environmental change impacting farmers' livelihoods. Although current studies focus on the adaptation and vulnerability of climate change in developing countries, most of them concentrate on the field of global impacts, human adaptation, and socio-ecological vulnerability of climate change (Adger et al., 2003; Fazey et al., 2010; Polsky et al., 2007; Snorek et al., 2014). Research on livelihood adaptation in the face of environmental change is still limited at the household level (Abid et al., 2016; Khayyati and Aazami, 2016) (especially the contradiction between the restoration of the ecological environment and the maintenance of farmers’ livelihoods in arid environments). Global climate change not only has a direct impact on the biological environment and natural resources but also indirectly derives a series of social problems, such as livelihood poverty, social inequality and competition for resources. Therefore, this study combines sustainable livelihoods with adaptation to propose a conceptual framework. This framework uses case studies to explain the adaptation of farmers facing risks in complex livelihood systems. It will focus on the livelihood adaptation of farmers to environmental change and water resource scarcity; the findings of these typical case studies may provide valuable references to mitigate farmers' livelihood risks and to promote scientific policy-making by decision-makers. The research objectives can be divided into four areas:

• What are the impacts of environmental change on farmers, and which adaptive strategies have been adopted by farmers? • What are differences in the adaptive capacity of farmers for different adaptation strategies? • What are the outcomes and perceptions of farmers’ livelihood adaptation? • What are the impacts (effects) of the adaptive capacity of farmers and local government policy actions on farmers’ adaptation results?

2. Analytical frameworks 2.1. Restrictions of current framework In the 1980s, sustainable science began to understand the 412

Ecological Indicators 93 (2018) 411–423

J. Chen et al.

economic indicators such as income, welfare and so on (Haglund et al., 2011;Nguyen et al., 2015), it is difficult to reflect satisfaction and personal cognition of adaptation subject to the livelihood adaptation results, which make analyses of livelihood adaptive outcomes biased. Therefore, in this framework, we improve the evaluation of outcomes of livelihood adaptation and use livelihood outputs (such as living standards and livelihood stability), as well as livelihood perceptions (such as degree of satisfaction and farmer cognition) to express livelihood adaptive outcomes.

the assessment of adaptive capacity and adaptive degree. In addition, there is a common feature between the past research framework and the methodology of livelihood adaptation, and it does not understand livelihood adaptation as an organic systematic process. Since this research always accompanies a choice of the subject adaptive strategies, they are often closely related to livelihoods for farmers scale. Choosing an adaptive livelihood strategy is a key to ensuring the survival of farmers amidst the risk of environmental and climatic changes. However, in different livelihood adaptation strategies, the available resources for farmers to address risks must be different, and their livelihood adaptation outcomes will be affected. Therefore, adaptation strategy, adaptive capacity, and adaptation outcomes are organic closely linked, constituting a complete adaptation process. Current studies lack a comprehensive research framework, combined with livelihood options, in the assessment of adaptive capacity and adaptive degree. This study argues that in order to fully understand the livelihood adaptation of environmental change risk and to solve the challenges of existing research, a framework combining the adaptation concept and sustainable livelihood approach as well as uniting qualitative and quantitative research methods is necessary.

3. Materials and methods 3.1. Study areas The study region is located in Minqin County (ranging from 101°49′ to 104°12′E and from 38°03′ to 39°28′N), Gansu Province, in northwestern China, which has a typical arid environment (Fig. 2). The estimated area of the research region is 15,870 km2, including desert, the Gobi, denuded mountains and saline land areas, which account for 91%, and an oasis area that accounts for only 9%. The west, north and east of Minqin County are surrounded by the Badan Jilin Desert and the Tengger Desert. The middle segment is a long oasis formed by the Shiyang River, where the population is mainly distributed. Minqin County is a temperate continental arid climate zone, with large temperature differences, less precipitation, and strong evaporation. The annual average precipitation is 113.6 mm. The annual average evaporation is 24 times the precipitation. There are no surface water resources in Minqin County, and water resources are composed only of surface runoff in the Shiyang River and groundwater. As the down reach of the Shiyang River, the water resources in Minqin are severely affected by upstream water consumption. The county's per capita water resource availability is only 520 m3, accounting for one-fifth of the average level in China. Wind and dust, drought and water resource scarcity are the region’s main natural features. Due to the surrounding area being a desert and the community population being concentrated in the oasis irrigation region, this research selected the core irrigated region as the study sample area, with the exception of Caiqi, Changning, as well as Chongxing in Minqin Oasis, using survey and interview methods to obtain data.

2.2. Integrated analytical framework for livelihood adaptation In the climate change field, adaptation is viewed as ecological, social or economic system responses to realistic or projected climate change and its effects, designed to mitigate hazards or to develop favorable opportunities adjusting to their behaviors (IPCC, 2001, 2014). Among them, the cycle of the adaptive process is an important inherent property in the system (Holling, 2001); it is intimately associated with the concepts of adaptive capacity (Smit and Wandel, 2006). For example, the livelihood system is an open, dynamic, balanced system where people obtain income through livelihood activities and maintain family life and reproduction through consumption (Zhang et al., 2016). In a livelihood system, households have the ability to adjust their behaviors and adopt the necessary strategies to adapt to disturbance from external pressures. However, their livelihood adaptive behaviors are inevitably affected by the local natural environment and social policy. The degree of adaptation by farmers is also affected by their own assets, external forces and livelihood output. Therefore, this paper defines livelihood adaptation as a process wherein a system copes with the pressure and risk from the external environment by using its own resources (assets), which are close to or reserved to adapt dynamically and to maintain the appropriate state, and which proposes a framework of livelihood adaptation that places adaptive strategies, adaptive capacity, and adaptive outcome at its core (Fig. 1). In this framework, we improved the single livelihood assessment method (i.e., the livelihood strategy and the livelihood outcome analysis) of previous farmers’ adaptation risk and constructed a comprehensive livelihood adaptation process by combining the livelihood adaptation strategy, the livelihood adaptive capacity and the livelihood adaptation outcome. This approach emphasizes the importance of livelihood adaptive capacity to cope with pressure within different livelihood strategies, and advocates logical analysis of an “External pressure-Adaptation process-Adaptedness”. The process of livelihood adaptation comes from adaptation of the main body of choice, including the reactive or autonomous adaptation by farmers and guidance from the government or an organization. Adaptive capacity is represented by a set of resources to respond to disturbances and to cope with current or future risks, which is the core of farmers’ livelihood resilience. From a livelihood perspective, capacity is portrayed by livelihood capitals and their dynamics reflect buffer capacity; furthermore, self-organization and learning ability within individual livelihoods is crucial for building resilience (Speranza et al., 2014; Nelson et al., 2007). Therefore, assessment of adaptive capacity includes three elements, including buffering capacity, learning ability, and self-organization. Since previous livelihood outcomes are quantified by

3.2. Sampling and data collection The research data were collected using a field survey. We conducted a pre-survey in April 2015 in Minqin County, China, which is to understand local natural resources, environmental changes and regional differences in agricultural cultivation. Based on the pre-survey, a multistage sampling technique was used to select the study sites and to sample farm households in the study area. Due to desert coverage around Minqin County, the township is concentrated in the core area of the oasis, therefore during the first stage, we selected all townships (13 townships) in the core area of the oasis as the survey sites. In the second stage, 2–4 administrative villages were selected from each township, totaling 35 administrative villages (see Appendix A. Fig. A.1) based on characteristics of the different township areas, the population, and the number of administrative villages. In the third stage, 10 households were randomly selected in the targeted administrative villages for surveying. Based on the questionnaire, we obtained basic data (the basic characteristics of the farmer household, livelihood capital, adaptation behavior, perceptions of environmental change, and policies, etc.) about the farmers’ livelihood adaptation strategies and capacities. First, we designed a questionnaire based on the pre-survey in April 2015. Second, in July 2015, a pretesting of the questionnaire was sent out and the problems existing in the questionnaire were revised in time to avoid missing any important information or errors. Third, before a formal questionnaire was performed, the investigators were trained and guided 413

Ecological Indicators 93 (2018) 411–423

J. Chen et al.

External pressure Pressure condition: Environment change, Climate change, Economic risk, Policy reform, Resource scarcity ……

Adaptive processes Livelihood adaptation strategies

Livelihood adaptive capacity

Structures subject Self -reaction of farmers Government policies Social aid ……

Three elements Buffering capacity Learning ability Self-organization

Livelihood adaptive outcome

Livelihood outputs Living standard Livelihood stability Livelihood perceptions Degree of satisfaction Farmer cognition

Adaptedness Livelihood hardship Choice of adaptation strategy afresh

Livelihood sustainable

Fig. 1. Conceptual framework for household livelihood adaptation (adaptedness is a state in which a system is effective in relating to the environment and meets the normative goals of stakeholders).

Fig. 2. Location of the study area. 414

415

Self-organization

Learning ability

Per capita actual cultivation area Water resources irrigation Agricultural change

Buffering capacity

Neighborhood trust Involvement opportunities

Organizational potential Social network

Technology study and application Experience learning exchange

Policy awareness

Per capita water expenditures Funding opportunities Labor dependency ratio The proportion of disabled and compromised groups Education level

Annual per capita income

Cultivated land quality Livestock and plants

Material assets

Index

Dimension

The ratio of government members of household family members Number of households in labor exchange and relational network support. Relational network support: species of support type, 0.25 = one, 0.5 = two, 0.75 = three, 1 = four. The degree of mutual trust in the neighborhood. 0 = very little, 0.3 = a small part, 0.6 = vast majority, 1 = whole. Opportunities for participation in policy and collective meetings of village or community groups.0 = very little, 0.25 = less, 0.5 = ordinary, 0.75 = much more, 1 = quite a lot.

Understanding of government's environmental managements and economic adaptation policies, 0 = a complete lack of understanding, 0.25 = little understanding, 0.5 = common understanding, 0.75 = more understanding, 1 = very understanding. Agricultural science and technology training opportunities and application. Training opportunities: training times (value of 0–9). Technology application: 0 = no, 1 = yes. Farmer learning about experiences of coping with risks, 0 = very little, 0.25 = less, 0.5 = ordinary, 0.75 = much more, 1 = quite a lot.

The education level of male domestic labor. Education level: 0 = illiteracy, 0.25 = primary school, 0.5 = junior middle school or technical secondary school, 0.75 = high school or junior college, 1 = university and above.

+ +

0.644 0.336

0.175 12.021

0.485

+ + +

1.467

0.533

+ +

0.433

113.700 0.834 0.754 0.588

1681.400

+

− + − −

+

0.550 15.166

0.393

+

+ +

0.898 0.112

+ +

Actual effective irrigation degree.0.25 = a small portion, 0.5 = about half, 0.75 = vast majority, 1 = whole. Whether or not there is an area of greenhouses and breeding brooders or other characteristic forestry and fruit industry practices in the household.0 = no, 1 = yes Household owned fixed assets, including daily appliances and housing. Housing types: 0 = thatched cottage, 0.25 = civil building, 0.5 = clay and brick building, 0.75 = brick building, 1 = concrete structure. Per capita living space: the ratio of total housing area and total number of households. Actual situation of cultivated land quality: 0.25 = poor soil, 0.5 = general soil, 0.75 = good soil, 1 = excellent soil. The scales of livestock and crops. Livestock scale = cattle × 1 + donkeys × 0.8 + sheep × 0.6 + pigs × 0.4 + chickens × 0.2. Crops scale: value of 0–8. The ratio of household agricultural, migrant workers and other income (including government subsidies) and the total number of households in the last year. Units: US dollars The ratio of the total expenditure of household irrigation and living water and the total number of households. Units: US dollars Credit opportunities and the number of borrowers. Credit opportunities: 0 = no, 1 = yes The number of households with non-labor force/The number of households’ labor force The proportion of persons with disabilities, illness, age greater than 65 and those less than 5

3.755

Mean

+

Anticipated impact

The ratio of actual cultivated area and total number of households.

Index description and definition

Table 1 Household-level adaptive capacity variables, descriptive statistics and anticipated impact on adaptive capacity.

0.162 0.248

0.516 11.174

0.272

1.767

0.199

0.224

105.096 0.485 0.737 0.222

1016.920

0.138 10.546

0.125

0.158 0.105

6.865

Standard deviation

J. Chen et al.

Ecological Indicators 93 (2018) 411–423

Ecological Indicators 93 (2018) 411–423

J. Chen et al.

3.3.2. Adaptive capacity indicators According to the conceptual framework, adaptive capacity includes the three dimensions of buffering capacity, learning ability, and selforganization. We reviewed the existing literature in different disciplines (anthropology, biology, geography, etc.) to identify the most important indicators for assessing adaptive capacity (Table 1). Speranza et al. (2014) found that, from an actor and livelihood perspective, livelihood capital reflects buffering capacity. Therefore, the farmer livelihood capital is taken as a buffer capacity indicator. Following Li et al. (2008), in the construct of learning ability, we used policy information sharing, the application or exchange of knowledge technology to portray farmer learning ability; self-organization was expressed by the indicators related to farmer social networks and community relations. Finally, based on comprehensive field investigation, we selected adaptive capacity indicator variables (e.g., agricultural transformation, water resources expenditure, etc.) that have a close relationship to local conditions.

Table 2 Farmers’ adaptive action. Adaptive action

Household (n = 360)

Percentage (%)

Adjustment of traditional agriculture saving and reducing water changing types of crop planting changing types of breeding

305 252 77

84.72 70.00 21.39

Development of facility agriculture and animal husbandry greenhouse 70 developing breeding brooders 234

19.44 65.00

Social assistance seeking help from relatives and friends bank loans taking part in social insurance

168 158 36

46.67 43.89 10.00

Livelihood diversification expanding income sources working outside farming contracting land

150 210 61

41.67 58.33 16.94

Passive behaviors Reduced consumption immigration or other

76 4

21.11 1.11

3.3.3. Adaptive outcome indicators The sustainable livelihood framework of DFID defines a livelihood outcome as more income, increased well-being, decreased vulnerability, food security, and more sustainable natural resources (DFID, 2001). For farmers to adapt to environmental changes and seek survival and development, this means to maintain a certain income and living standard. Because income influences living standards, we use a per capita net income to measure farmers’ livelihood adaptive outcomes, with higher per capita net income indicating better adaption. Livelihood stability means that farmers are less vulnerable to risks and maintain a stable livelihood, which has been represented by livelihood diversity indexes in some research (Yu et al., 2013). While livelihood diversity just indicates that the farmers’ livelihood behaviors are various, not all livelihoods behaviors have stable and continuous income. Therefore, the index of income diversity was used to measure farmers’ livelihood stability, with a high index representing a good adaptive outcome. Moreover, farmers’ perceptions can provide supplementation for an objective index and should not be neglected when estimating their adaptive outcome. Therefore, we use interviews data which is about the farmers’ perceptions of livelihood adaptation strategies representing livelihood perceptions.

Notes: Expanding income sources include odd jobs, mining herbs, business, etc.

in the field. Finally, from August 7th to 24th, 2015, the investigators officially distributed questionnaires to 35 administrative villages in 13 townships of the sample area to collect farmer information and capital data. A total of 370 questionnaires were distributed and 362 valid questionnaires returned. The perceptual data acquisition of farmers’ livelihood outcome was based on semi-structured interviews. The research group conducted an additional survey on the study area from August 2nd to August 16th, 2016; the villages that were not covered in the sample areas last year were covered. The interviewees included officials (village chiefs, clerks, accountants, etc.) and senior farmers. Through semi-structured interviews, each participant was interviewed for 30–50 min to obtain perceptions about government policy and farmers’ livelihood adaptation.

3.3. Indicators and quantitative analysis 3.3.4. Quantitative analysis To measure adaptive capacity, we followed Pandey (2011), the adaptive capacity index (ACI) is calculated as an aggregate of the adaptive capacity dimensions indexes, which was calculated as the weighted sum of each indicators. The weight to each of the indicators was calculated using the entropy value method (Shannon, 1948; Shuiabi et al., 2005). The index was calculated as the sum of the weighted adaptive capacity indicators of a farmer:

3.3.1. Classification of adaptation strategies In order to identify farmers’ livelihood adaptation strategies, firstly, we conducted a statistical analysis for the adaptive actions of farmers. According to statistical analysis (Table 2), we summarized farmers' adaptive behaviors as follows: adjustment of traditional agriculture, the development of facility agriculture and animal husbandry, social assistance, livelihood diversification and passive behaviors. Secondly, based on the classification of adaptive behaviors and considering the high number of adaptive behaviors of working outside farming, farmers’ livelihood adaptation strategies were organized into the following six categories: agricultural transformation, traditional agricultural adjustment, work equivalent to farming, advanced work outside of farming, diverse adaptation, and passive adaptation (Table 3).

m

ACIj = BIj + LIj + SIj =

∑

n

∑

Wmi Xmj +

j=1

z

Wni Xnj +

j=m+1

∑

Wzi Xzj

j=n+1

where ACIj is the adaptive capacity index of household j, BIj, LIj and SIj are the buffering capacity, learning capacity, self-organization index of household j, W1i…Wzi are the weight to adaptive capacity indicators,

Table 3 Farmers’ livelihood adaptation strategies and classification description. Livelihood adaptation strategies

Classification description

Household (n = 360)

Percentage (%)

Agricultural transformation

Select a variety of adaptation methods to adjust traditional agriculture and develop facility agriculture and animal husbandry Only select adaptation method involving traditional agricultural adjustment Select a variety of adaptation methods toward agricultural transformation and work outside of farming Select adaptation method of work outside of farming as the dominant Select a variety of adaptation methods equal to or more than five species Select adaptation methods of social assistance and passive behavior as the primary approach

53

14.72%

32 70

8.89% 19.44%

20 171 14

5.56% 47.50% 3.89%

Traditional agricultural adjustment Work equivalent to farming Advanced work outside of farming Diverse adaptation Passive adaptation

416

Ecological Indicators 93 (2018) 411–423

J. Chen et al.

Table 4 Indexes of adaptive capacity, buffering capacity, self-organization and learning capacity. Livelihood adaptation strategies

Agricultural transformation Traditional agricultural adjustment Work equivalent to farming Advanced work outside of farming Diverse adaptation Passive adaptation

Aggregate adaptive capacity index

Dimensions’ adaptive capacity index

Mean value

Median

Maximum

Minimum

Buffering capacity Mean value

Self-organization Mean value

Learning capacity Mean value

0.179 0.137 0.154 0.134 0.193 0.128

0.159 0.104 0.148 0.128 0.180 0.126

0.435 0.372 0.360 0.222 0.524 0.214

0.056 0.051 0.052 0.064 0.074 0.081

0.076 0.054 0.063 0.054 0.080 0.058

0.052 0.039 0.053 0.045 0.054 0.036

0.050 0.044 0.037 0.035 0.058 0.034

and X1j…Xzj are the standardized value of adaptive capacity indicators (value of 0–1). The basic assumption was that a farmer’s livelihood adaptive outcome depends linearly on their livelihood adaptive capacity. To conduct this analysis, a multiple linear regression analysis (OLS model) was used. Model 1 and Model 2 (see Table 6) aimed at determining how adaptive capacity factors influenced the farmers’ index of income diversity and per capita net income at statistical significance of 5% probability level. The OLS model is specified as:

implementation of government environmental policies having a serious impact on local development and farmers’ livelihoods. Water decreased, water quality deteriorated. Due to climate change and the development of the upper reaches of the oasis, inland runoff dropped sharply in Minqin. The runoff was 542 million m3 in the 1950s and decreased to only 100 million m3 in 2006 (Fig. 3). An internal oasis has formed due to the extremely serious water shortage. To solve the problems of industrial and agricultural production and the need for human and livestock drinking water, people have opted to dig wells and extract groundwater. From the 1970s to 2006, the annual excess groundwater was 460 million m3, which caused the groundwater level to drop by 0.3–0.8 meters per year (Yin et al., 2016). The county mineralization of shallow water increased by 0.2–0.35 g/liter per year, and the deep-water mineralization annual average increased by 0.24 g/ liter (Fig. 4). Due to the deterioration of shallow water quality in the lake area and in Quanshan Town of Minqin County, the water cannot be used for irrigation in some areas. It is difficult to obtain human and livestock drinking water. Simultaneously, the quality of cultivated land has seriously decreased. Desertification menace, increasing natural disaster. Because of climate drought and lack of water resources, large tracts of forest, grassland, and desert plants have deteriorated and declined due to dehydration causing desertification of farmers' cultivated land. At the same time, there are more than 164,736 acres of quicksand area in the peripheral Minqin Oasis that need urgent governance. Located on the edge of the desert, farmland is easy to desertify. Meanwhile, agricultural production is often affected by wind and sand. With the intensification of desertification, sand storms and other natural disasters occur frequently. In 2001, there were 16 times more than 8-Level Winds and 14 times sand storms (NBSC, 2002). Severe dust storms not only cause direct economic losses to agricultural production but also have a serious impact on the health of farmers. Reduction of cultivated land resources, changes in agricultural structure. Due to ecological damage and drought environmental stress, the situation is becoming increasingly severe. Since 2007, the local government and the state have increased the ecological management investments in Minqin, on the basis of increasing inland runoff, whereas the

E= β 0 + β1A1 + β 2A2 + β3A3 + …+β pAp + ε where E is the dependent variable, β0 is the constant term, βp denote parameter estimates, Ap are the explanatory variables, ε is the random error. Because of the large number of adaptive capacity indicators, there may be multi-colinearity problems, so a principal component analysis (PCA) was used to solve this problem (Below et al., 2012). The KOM and Bartlett test (KOM-0.65, LR–792.494, Sig-0.00) indicates that we can carry out a principal component analysis. According to the PCA results (see Appendix A. Table A.2), the cumulative variance contribution rate (57.308%) indicates the strong explanatory power of seven principal components. Considering the possible endogeneity issue, we used the seven principal components as a variable to carry out multiple regression analysis. We also calculated the Variance Inflation Factor (VIF) for each of the principal components variables (Monirul Alam et al., 2016). The VIFs range from 1.00 to 1.054, which is below the conventional thresholds of 10 or higher used in regression diagnosis. 4. Results and discussion 4.1. The impacts of environmental change and farmers’ adaptation strategies 4.1.1. The impacts of environmental change and government policy Due to the influence of climate change and human activities, the environment has changed significantly, combined with the Table 5 Farmers’ livelihood adaptive outcome and adaptedness. Livelihood adaptation strategies

Livelihood outputs Living standard

Livelihood stability

Agricultural transformation Traditional agricultural adjustment

−0.004 0.147

0.368 0.310

Work equivalent to farming

0.185

0.398

Advanced work outside of farming

0.398

0.344

Diverse adaptation

0.164

0.392

Passive adaptation

0.072

0.277

Livelihood perceptions

Adaptedness

New agricultural income, but large input costs, unstable benefits. Cash crops and incomes increase, but high agricultural costs, especially water costs, cultivated lands decrease. Applying surplus labor and time increasing family incomes with agricultural income promotion. Working outside farming can earn more money than agriculture but it is unstable, daily costs are high. Diverse livelihoods avoid risks of single livelihood, applying family resources and labor properly. Due to poverty and a weak foundation, can only live by support and savings.

Unsustainable Unsustainable

417

Sustainable Unsustainable Sustainable Unsustainable

Ecological Indicators 93 (2018) 411–423

J. Chen et al.

Table 6 Regression estimation results showing the impact of adaptive capacity on the farmer's livelihood outputs Component variables

Model 1 Index of income diversity

Model 2 Per capita net income

coef

t

sig.

coef

t

sig.

Agricultural transformation Human assets

0.107** 0.346*

2.229 7.341

0.026 0.000

−0.263* 0.251*

−5.431 5.243

0.000 0.000

Natural and financial assets

0.286*

5.963

0.000

0.244*

5.020

0.000

*

4.999

0.000

Water resources and social networks

0.236

a.

Including adaptive capacity indicators

Adaptive capacity dimensions

Agricultural change, husbandry and planting The proportion of disabled and compromised group, education level Per capita actual cultivation area, annual per capita income, per capita water expenditure Water resources irrigation, experience learning exchange, social network

Buffering capacity Buffering capacity Learning ability Buffering capacity Buffering capacity Self-organization

a. Model 1: dependent variable is the index of income diversity, adjusted R2 = 0.22 (p < 0.01); Model 2: dependent variable is per capita net income, adjusted R2 = 0.20 (p < 0.01); *p < 0.01 **p < 0.05.

Unit:1 20

109m3 Runoff in Shiyang River Basin

18

Runoff into Minqin

16 14 12 10 8 6

4 2 0 1955

1965

1975

1985

1995

2000

2005

2008

2011

Fig. 3. Changes in runoff over the past 60 years.

of the environment and the livelihood interests of farmers. The government has implemented a policy of “shutting down the wells and reducing the farmland”, which makes the land resources of farmers who have a large area of arable land decrease sharply in the lake area and along the edge of the desert area. Furthermore, part of the environmental restoration policy constrains farmers’ livelihood activities. This includes paying for water resources for irrigation and daily life, limiting

use of a series of controlled environmental restoration measures have been attempted, such as “shutting down the wells and reducing the farmland”, “sand control”, water resource charging, and water consumption control. The government puts forward the development model of “facility agriculture and animal husbandry” and “characteristic horticulture” through a policy of adapting farmers’ economic development. However, there are contradictions between the restoration

Unit: (g/liter) 3 Mineralization of groundwater Depth of groundwater level 2.5

1.51

1.5

0.5

2.398 2.052

1.92

2

1

Unit: (m) 25 2.559 18.796

2.554 19.509

17.135

13.634

12.98

15

1.14 0.95

2.7

20

10

7.65 3.6

5

0

0

1965

1975

1985

1995

2000

2005

2008

Fig. 4. Variation in groundwater quality and depth over 50 Years. 418

2011

Ecological Indicators 93 (2018) 411–423

J. Chen et al.

Fig. 5. Adaptive capacity dimensions matrix.

breeding brooder (65%, n = 234), and working outside farming (58.33%, n = 210) are also important livelihood adaptive selections made by farmers. Furthermore, due to government policy support, farmers can obtain bank loans (the premise is having an age less than 60 and having the ability to repay). To maintain family life, some farmers seek increasing other income, such as by pressuring sand, gathering herbs and selling, as well as operating small shops. However, when economic difficulties are encountered by the family, people will seek help from friends and relatives or pursue bank loans. Therefore, bank loans, seeking help from relatives and friends and expanding income sources are three adaptive methods that have a proportion of nearly 50%. A total of 16.94% of farmers choose to contract land (idle and arable land owned by farmers who choose to work outside in a full year) to further expand the area of farming and increase agricultural income. In the rest of the adaptive methods, farmers select the adaptive methods of changing the types of breeding and reducing consumption, which only accounted for approximately 20%. Unstable or low incomes among farmers facing the current situation said that they just reduced their own consumption. Taking part in social insurance as an adaptive method involved only 36 households, and immigration or other approaches by the surveyed farmer was only 4 households, accounting for only 1.11%. In general, agricultural transformation and working outside farming have become the main behaviors of farmers seeking to adapt to environmental changes. The adaptive behaviors of farmers are diversified. Facing risks, people do not choose only one livelihood adaptive behavior but rather a combination of many adaptive behaviors as an adaptation strategy to enhance their adaptive capacity. Among the farmers’ adaptation strategies (see Table 3), there were 53 farmer households that adopted agricultural transformation. Farmers adopting traditional agricultural

the planting of high water consumption crops (some of the crops with good economic benefits), forced planting of economic crops (some of the economic crops without a market), etc. Therefore, farmers choose the kind of livelihood strategy in the face of environmental change and the adaptive capacity of the farmer's livelihood strategy under government policy, which become the keys to sustainable development of local farmers’ livelihood systems. 4.1.2. Farmers’ adaptation strategies Reactive or autonomous adaptation (Engle, 2011) is a kind of response by a population (human) that is inherent to external environmental adaptability. In addition to reactive adaptation, humans have the capability to take proactive adaptation measures to mitigate perceived negative impacts from these future events, which is anticipatory or planned adaptation (Engle, 2011; Fankhauser et al., 1999). Reactive or autonomous adaptation is mostly adaptive selection among individuals, while anticipatory or planned adaptation is related to decision-making by subjects such as institutions and the government. Among the farmers’ adaptive behaviors (see Table 2), the development of facility agriculture and animal husbandry and immigration are advocated and supported by the government. Bank loans are supported by the government and financial institutions. Other adaptive methods are independent selections made by farmers according to their experience and market information. Table 2 indicates that the most adaptive method chosen by farmers is saving and reducing water (84.72%, n = 305). In the arid environment of Minqin County, water resources are very scarce. Therefore, farmers have to save water and reduce water consumption in order to cut down the cost of agricultural production and living water. Then, changing the types of crop planting (70%, n = 252), developing the 419

Ecological Indicators 93 (2018) 411–423

J. Chen et al.

groundwater, except for Minqin; areas near Minqin do not charge water resource fees. Before 2007, we did not pay for water, so our agricultural cost was not as high. But now water is too expensive, the cost of farming is all dependent on water resource fees.” Farmers who are characterized as work equivalent to farming and diverse adaptation types think that working outside farming and supplementing with agriculture can help the surplus labor to increase other income to effectively maintain the family livelihood and improve the living standard. Farmers of advanced work outside of farming are more likely to have a higher income than with jobs in farming, but the cost of living in urban areas is high. Once they lose their jobs, family livelihoods are not guaranteed. Based on the comprehensive results of livelihood outputs and livelihood perceptions, we can conclude that farmers of work equivalent to farming and diverse adaptation types are sustainable; agricultural transformation, traditional agricultural adjustment and passive adaptation are unsustainable. Farmers think a risk of unemployment and income instability exists in advanced work outside of farming, therefore advanced work outside of farming can also be considered unsustainable.

adjustment included only 32 households. Second, work equivalent to farming and diverse adaptive strategies were dominant. Among them, diverse adaptive strategies included up to 171 households, whereas the proportion of advanced work outside of farming and passive adaptation strategies was relatively low. The results described above illustrate that the adaptive strategy of farmers in Minqin tended toward a diverse livelihood approach, which mainly involved positive adaption and mixing farming with working. The difference in the choice of farmers’ adaptive strategies was derived from differentiation of farmers’ selfadaption and planned adaptive selection guided by the government. 4.2. Farmers’ livelihood adaptive capacity for different adaptation strategies At the farmers’ adaptive strategies level, there were significant differences in the adaptive capacity dimensions among the different adaptation strategies. From buffering capacity and learning ability perspective, the farmer’s agricultural transformation and diverse adaptation had a strong buffering capacity and learning ability, and the other three types of adaptation strategies were relatively weak. In selforganization, three types of adaptation strategies consisting of agricultural transformation, work equivalent to farming, and diverse adaptation, were high. On the other hand, the two-factor matrix (Fig. 5) indicated that the buffer capacity-self-organizing matrix was the highest among the farmers in the categories of agricultural transformation and diversification adaptation; the self-organization-learning ability matrix was the highest of diverse adaptation, followed by traditional agricultural adjustment, agricultural transformation and work equivalent to farming; the learning ability-buffer capacity matrix was the highest diverse adaptation; the lowest was passive adaptation. On the whole, among the adaptive capacity dimensions, the diverse adaptation and the agricultural transformation strategies were the highest and the passive adaptation strategies were the lowest. Moreover, According to a descriptive statistical analysis (see Table 4), farmer households adopt six kinds of livelihood adaptation strategies, which are basically balanced in the adaptive capacity index; however, there are internal differences. The adaptive capacity of diverse adaptation strategies was strongest, and the mean value of the adaptive capacity index was 0.193, followed by agricultural transformation and work equivalent to farming, with a mean value of 0.179 and 0.154. The lowest adaptive capacity of the famer households was passive adaptation strategy; this adaptation strategy group mean value was 0.128; the maximum adaptive capacity index was only located at 0.214 (Table 4).

4.4. Influencing factors on farmer's livelihood adaptation There are differences between farmers’ adaptation strategies and their adaptive capacity to environmental change, which will inevitably affect the farmers’ livelihood adaptation outcomes. What is the reason for the differences in adaptedness, and how does this difference occur? This is what we are going to analyze and discuss. The multiple regression analysis and the farmer interview data were used to explain the reasons for the differences. 4.4.1. Role of adaptive capacity Adaptive capacity varies between different contexts and systems and is not equally distributed (Adger et al., 2007); it directly affect the farmers' livelihood adaptation effect. Table 6 shows the result for the impact of adaptive capacity on the farmer's livelihood outputs, and the adaptive capacity indicators and dimensions representing the component variables. The results of model (1) and (2) demonstrate that agricultural transformation, human assets, cultivated land and financial assets and water resources and social assets have significant effects on livelihood adaptation, and the four principal components had positive effects on farmers’ livelihood stability (Table 6). However, the principal component of agricultural transformation had a negative effect on livelihoods, which is not conducive to improving income levels. According to the matrix of rotation components (see Appendix A. Table A.2), it can be seen that the actual per capita cultivated land and annual income (cultivated land and financial assets), educational level (human assets), experiential learning communication, and social network (social assets) play a positive role in livelihood adaptation output (adaptation results). It shows that based on the natural and financial assets (resources), cognitive ability and social network are beneficial for improving the adaptive capacity of farmers to environmental changes, thus ensuring the farmers' effective livelihood output. However, the factors of agricultural change, husbandry and planting (agricultural transformation) as well as water resource expenditures (negative correlation with financial assets) have negative effects on the livelihoods of farmers indicating that the adaptation effect of agricultural transformation or traditional agricultural adjustment strategies is unsustainable. Therefore, we know that government promotion of agricultural transformation measures is favorable to helping farmers adapt to the adverse effects of environmental change, and livelihood stability has improved, but in recent years, household income has not been ideal. Resource (asset) accessibility influences farmers’ adaptive capacity and a loss of resources will reduce a farmer’s ability to resist risks. In the case of traditional agricultural transformation and passive adaptation, a series of policies of “shutting down wells and reducing the farmland”

4.3. Farmers’ livelihood adaptive outcome and perceptions Based on the conceptual framework, farmers’ livelihood adaptive outcomes include livelihood outputs and livelihood perceptions. In terms of livelihood outputs (see Table 5), the advanced work outside of farming type represents the best result, with the highest living standard index and livelihood stability index. Diverse adaptation and work equivalent to farming types follow the best, and the passive adaptation type is relatively weak. Famers’ livelihood perceptions indicate that farmers believe that agricultural transformation has a certain role in adapting to lack of water resources and to deterioration of land quality. However, agricultural transformation requires considerable investment (capital, manpower, materials, etc.), and the market price of agricultural and animal husbandry products (greenhouse vegetables and sheep, etc.) will fluctuate, which makes income unstable. Farmers exhibiting a traditional agricultural adjustment type still take farm land for a living, mainly in the cultivation of cash crops. This type of farmer thought that, although the income from traditional grain crops was greatly improved, the cultivated land required many water resources to irrigate, water resources expenditures were too high, and pure agricultural income is low. Farmers said, “there is no place charging for surface water or 420

Ecological Indicators 93 (2018) 411–423

J. Chen et al.

and “returning farmland to forest” (Government starting point is ecological restoration) cause farmers the loss or reduction of cultivated land resources and the limitation of irrigation water resources; furthermore fees cause a sharp increase in agricultural costs and therefore farmers become weak in buffer capacity. Because of the extreme lack of living assets (such as cultivated land, economic assets, educational opportunities, etc.), farmers using passive adaptation cannot adapt to environmental change and drought stress; they have to focus on less consumption, and seek social assistance and government protection. However, farmers’ perceptions (such as education level and social network) are different. Farmers in advanced work outside of farming think that it is difficult for them to survive and improve family life through agriculture because of a loss of cultivated land resources or an increase in agricultural production costs; thus, seeking new livelihood adaptation is necessary. One farmer explained that “water resources control is conducive to ecological restoration, but it still has some impacts on us. Due to the high water costs and lack of land, you have to do something else in order to live, and it is better to do business or go out to work.” Comparatively speaking, owing to the choice of reasonable and diversified livelihoods, it can make up a deficiency in resource reduction, face environmental change risks and avoid traditional agricultural decline and instability in the process of agricultural transformation; therefore, diverse adaptation and work equivalent to farming show better adaptive outcomes. In addition, an interesting result of the OLS model calculation is that the principal components of technology application and social relations have no significant effect on the livelihood output of farmers, which indicates that these adaptability factors (technical learning and application, participation opportunities, organizational potential and neighborhood trust) did have not a direct impact on the farmers’ livelihood adaptive outcome. This is contrary to the findings of some previous studies stated (Monirul Alam et al., 2016; Quiroga et al., 2015) that the adoption of technology and the social relations structure have a stimulatory effect on the livelihood of farmers. The possible reasons are as follows: the technical training promoted by the government is mainly based on crop cultivation techniques used for agricultural transformation in study areas. Due to the instability in the agricultural transformation market and high input costs, farmers reduce the use of some technologies. Second, the existing agricultural organization recognizes farmers as the basic unit of agricultural production in China. The correlation among farmers is weak. Farmers’ livelihood strategies are dominated by independent decision-making or government planning arrangements. Therefore, social relations hardly play a role in the livelihood adaptation of farmers. Accordingly, it is important to identify what builds adaptive capacity or, similarly, what functions as barriers or limits to adaptations (Adger et al., 2008).

which occupy the lands, do not grow within 3 or 4 years. It is not a wise choice to plant fruits. Besides, there is no market for grapes and dates. It seems that we can get more benefits from cultivating grain crops. Even when you get harvests, there is no market to sell. The large quantity of greenhouses nowadays make the market more competitive. We can only earn 366 dollars per greenhouse, while with the serious pest and disease and large input costs, not much money is left. Furthermore, as the government advocates water conservation in order to restore the deteriorating ecology, a series of adjustments and controlling policies for water resources, planting structure and arable land resources have been applied. For the purpose of ecological restoration, the government has adopted coercive measures, such as prohibiting water-consuming crops, agricultural water consumption and water available times set by the government and so on. However, farmers’ agricultural incomes rely on the market price, which is not connected with the policy. Some farmers said that “we believe that restoration policy is good but that the government should not force us to plant what we do not want to plant. We should live self-determined and we know what kind of crops can bring us more benefits.” The government policy is impractical and unreasonable allocation of water resources make farmers miss the best times to market agricultural products, thus leading to a decrease in agricultural income. Therefore, the impractical policy causes lost incomes for farmers. A farmer explained this as follows: “when you need water, it is not available, when you don’t need it, water comes and you have to pay whether watered or not.” We know why the adaptation outcome of agricultural transformation and traditional agricultural adjustment is poor. The reason is that government policies did not work, leading to mal-adaptation. Furthermore, government intervention and irrational guidance had a negative impact, and in the process of farmers’ livelihood adaptation, the market leverage was weakened by government action; farmers could not adapt to the market to maintain the livelihood interests. Engle (2011) found that complementary to reactive or autonomous adaptation is anticipatory or planned adaptation. The adaptation strategy of diverse adaptation and work equivalent to farming is in connection with reactive or autonomous adaptation and anticipatory or planned adaptation. Although government anticipatory or planned adaptation strategies exist, the government’s policy is not very influential and farmers can depend on market information or experience to reduce unfavorable adaptation outcomes. Farmers of advanced work outside of farming can choose almost anything they would like to do, which totally avoids agricultural policy limitations, so farmer income can improve even with the problem of income instability. 4.5. Policy implications 4.5.1. Livelihood challenges and uncertainties The results show that diversified livelihood adaptation (diverse adaptation, working equal to farming) is conducive to avoiding the vulnerability of livelihood caused by the risk of environmental change. However, farmers’ livelihood diversification in the study area still faces many challenges. First, the lack of available resources due to droughts and fragile ecosystems makes it impossible for some traditional livelihood options (such as animal husbandry and forestry). Second, some mandatory ecological restoration policies implemented by the government have resulted in a limited choice of farmers’ diversified livelihood options (such as grazing prohibition, prohibiting water-consuming crops). In addition, the relocation of migrants (young labor force, student education immigrants) driven by degraded natural environment and the inability of aging local laborers to continue farming have all contributed to simplifying the livelihoods of local farmers in the future (such as purely migrant workers). Therefore, the existing problems, such as the relationship between the restriction of livelihood options and the interests of farmers and the uncertainty of future livelihoods emphasize the importance of government’s effective guidance in

4.4.2. Role of government policy and market Adaptation to environmental change is a complex problem, and anticipatory or planned adaptation does not always end well. Barnett and O’Neill (2010) found that some decisions (strategies) may lead to adaptation failure, or even increase the vulnerability of the system. This problem of increasing risks from adaptation is often termed ‘maladaptation’. (Burton, 1997; Barnett and O’Neill, 2010). The original government policy and adaptation guidance aimed at mitigating the livelihood hazards caused by environmental change. However, the policy-guided adaptation strategies do not always play a positive role. Taking agricultural transformation adaptation strategy as an example, farmers generally believe that the government policy of developing a characteristic forest and fruit industry is a failure, because there is no market; and some farmers of facilities agriculture and animal husbandry do not think that it is generally beneficial. Some farmers and village leaders explained this as follows: Fruit trees need a lot of water to survive, but water is so precious in this place. We also lack cultivated lands, but some fruits, such as grapes, 421

Ecological Indicators 93 (2018) 411–423

J. Chen et al.

systematic evaluation and constructs a method for indexing farmers’ livelihood adaptation research. The objectives of this study were to analyze the behavior of farmer adaption to environmental change and to understand farmers’ livelihood adaptation outcomes and their influencing factors in the case of Minqin. This study reveals that farmers adapting to environmental change are diversified. Agricultural adjustment and work is the main behavioral choice of farmers, but adaptation strategies mainly involve diverse adaptation and work equivalent to farming. There are also differences in the adaptive capacity of each adaptation strategy. Single or passive livelihood adaptation strategies, such as agricultural transformation and traditional agricultural adjustment, are not sustainable livelihoods. Adaptive capacity is weakened due to a shortage of resources (reduction in cultivated land, expenditure on water resources and so on) and on government policies, which leads to poor adaptation of farmers’ livelihoods. To a certain extent, government policy avoids the adverse effect of environmental change is conducive to ecological restoration, and guides farmer initiatives to adapt to environmental changes, which is beneficial for farmers to change their original livelihoods. However, some unreasonable policy actions ignored farmer demands during the process of ecological environmental restoration, which led to agricultural transformation, and the traditional agriculture adjustment of farmers’ livelihood interests are lost. Furthermore, diversification of farmers’ livelihood incomes and the adaptability of employment help to avoid the risk of agricultural transformation. Social network resources and education level are conducive to enhancing farmers’ livelihood adaptive capacity and livelihood cognition. These factors are the main reasons for the diverse adaptation and work equivalent to farming that allow farmers to maintain livelihood sustainability.

information and the farmers’ own cognition. The establishment of social organizations or cooperatives among farmers will help to circumvent the existing single livelihood risks of farmers and play a collective adaptive role to ensure the diversity and continuity of livelihood adaptation strategies. 4.5.2. Policy and action Unsuccessful adaptation does not necessarily mean that adaptation has significantly increased vulnerability - it may simply mean that an action does not work (Barnett and O’Neill, 2010). This study highlighted that balancing interests between ecological environmental restoration policy and farmers’ livelihoods had the importance of farmer sustainable livelihoods. For arid regions with environmental change (the same as other developing countries), the ultimate goal of promoting environmental recovery is still to maintain the well-being of farmers’ livelihood. Model results suggest that resource (asset) accessibility, social network and cognitive ability are the key factors that affect adaptation results. This underscores that under the premise of ensuring ecological restoration, it should reduce the cost of agricultural production and increase the livelihood resources (such as arable land, water source, etc.) for farmers. Moreover, in addition to studying agricultural planting techniques, the government should strengthen the cognitive training for farmers on re-employment skills and market information and should encourage them to establish networks of risk communication and resource assistance. Results also indicate that government policy disturbance (agricultural restructuring, water control) and lack of market information have hindered the livelihood of farmers. Therefore, the government or administrative agencies adopting polices (or advocating adaptive strategies) should be aware of famers’ needs in the study area. For instance, with respect to water resources management, measures must be taken according to circumstances and to allocate water rationally in line with crop requirements. Additionally, with regard to water distribution in the area, measures must be taken according to local conditions. The rationing of water resources should be appropriate to the region, which is at the edge of water source areas or on sandy land. Second, the premise of adjusting agricultural structure should pay attention to the role of the market. The aim should be to plant crops that have low water requirements, which is the premise of ecological restoration, while not ignoring the economic benefits of crops. It is our task to seek a balance between ecological restoration and water consumption of crops, open up the market and protect farmer agricultural benefits. However, in the long-term, shifting the traditional agricultural development model and promoting the development of the modern agricultural industry in the study area will reduce the dependency of farmers' livelihoods on ecological resources (cultivated land and water resources) and help the farmers succeed in adaptation.

Acknowledgements This study was supported by the National Natural Science Foundation of China (No. 41571163), Northwest University Doctorate Dissertation of Excellence Funds (No. YYB17016) and Humanities and Social Science Talent Plan in Shaanxi Province, China (HSSTP). We would like to thank Kongsen Wu, Qian Liu, Lili Xia, Ruixuan Yu, Lingxin Zhai and Bo Li for contributing to the household survey and data collection. Furthermore, we are grateful for the effective coordination of the Minqin local government and the support from local farmers in our survey. The authors would like to express their appreciation to anonymous reviewers for the insightful comments that improved this manuscript. Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.ecolind.2018.05.017.

5. Conclusions

References

Based on an understanding of previous research and the improvement of the analytical framework, this article established a framework for livelihood adaptation evaluation under the background of climate (environmental) change, using case studies to validate the improvement of the analytical framework. The results of the study reveal that using only the sustainable livelihoods framework and econometric methods, we may conclude that the adaptation effect of households with high-livelihood outcomes is good (Table 5); however, when the analysis framework incorporates livelihood adaptive perceptions, we found that the adaptation strategy of high-livelihood outcomes is not necessarily sustainable, because the quantification of livelihood output hides the perceived livelihood risk and the livelihood satisfaction of farmers, which may result in biased evaluation of livelihood outcomes. Second, the analysis framework in this study, which explains the logical relationship and their impact among farmers' livelihood adaptation strategies, adaptive capacity and adaptation outcomes, provides a

Abid, M., Schilling, J., Scheffran, J., Zulfiqar, F., 2016. Climate change vulnerability, adaptation and risk perceptions at farm level in Punjab, Pakistan. Sci. Total Environ. 547, 447–460. Adger, W.N., 2006. Vulnerability. Global Environ. Change 16, 268–281. Adger, W.N., Arnell, N.W., Tompkins, E.L., 2005. Successful adaptation to climate change across scales. Global Environ. Change 15, 77–86. Adger, W.N.S., Katrina, H., B. te al.,, 2003. Adaptation to climate change in the developing world. Progr. Dev. Stud. 3, 179–195. Adger, W.N., Agrawala, S., Mirza, M.M.Q., Conde, C., O’Brien, K., Pulhin, J., Pulwarty, R., Smit, B., Takahashi, K., 2007. In: Parry, M.L., Canziani, O.F., Palutikof, J.P., van der Linden, P.J.,Hanson, C.E. (Eds.), Assessment of adaptation practices, options, constraints and capacity. Climate Change 2007: Impacts, Adaptation and Vulnerability. In: Contribution of Working Group II to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change. Cambridge University Press, Cambridge, UK. Adger, W.N., Dessai, S., Goulden, M., Hulme, M., Lorenzoni, I., Nelson, D.R., Naess, L.O., Wolf, J., Wreford, A., 2008. Are there social limits to adaptation to climate change? Clim. Change 93, 335–354.

422

Ecological Indicators 93 (2018) 411–423

J. Chen et al.

NBSC (National Bureau of Statistics of China), 2002. China Statistical Yearbook: Minqin County. China Statistical Press. Nelson, D.R., 2011. Adaptation and resilience: responding to a changing climate. Wiley Interdiscip. Rev. Clim. Change 2, 113–120. Nelson, D.R., Adger, W.N., Brown, K., 2007. Adaptation to environmental change: contributions of a resilience framework. Annu. Rev. Environ. Resour. 32, 395–419. Nguyen, T.T., Do, T.L., Bühler, D., Hartje, R., Grote, U., 2015. Rural livelihoods and environmental resource dependence in Cambodia. Ecol. Econ. 120, 282–295. Nicholas, K.A., Durham, W.H., 2012. Farm-scale adaptation and vulnerability to environmental stresses: insights from winegrowing in Northern California. Global Environ. Change 22, 483–494. Oberlack, C., Tejada, L., Messerli, P., Rist, S., Giger, M., 2016. Sustainable livelihoods in the global land rush? Archetypes of livelihood vulnerability and sustainability potentials. Global Environ. Change 41, 153–171. Osbahr, H., Twyman, C., Neil Adger, W., Thomas, D.S.G., 2008. Effective livelihood adaptation to climate change disturbance: scale dimensions of practice in Mozambique. Geoforum 39, 1951–1964. Pandey, V.P., Babel, M.S., Shrestha, S., Kazama, F., 2011. A framework to assess adaptive capacity of the water resources system in Nepalese river basins. Ecol. Ind. 480–488. Parry, M., Carter, T., 1998. Climate Impact and Adaptation Assessment: A Guide to the IPCC Approach. Earthscan publications Ltd. Perry, R.I., Ommer, R.E., Barange, M., Werner, F., 2010. The challenge of adapting marine social–ecological systems to the additional stress of climate change. Curr. Opin. Environ. Sustain. 2, 356–363. Polsky, C., Neff, R., Yarnal, B., 2007. Building comparable global change vulnerability assessments: the vulnerability scoping diagram. Global Environ. Change 17, 472–485. Reed, M.S., Podesta, G., Fazey, I., Geeson, N., Hessel, R., Hubacek, K., Letson, D., Nainggolan, D., Prell, C., Rickenbach, M.G., Ritsema, C., Schwilch, G., Stringer, L.C., Thomas, A.D., 2013. Combining analytical frameworks to assess livelihood vulnerability to climate change and analyse adaptation options. Ecol. Econ. 94, 66–77. Scoones, I., 2009. Livelihoods perspectives and rural development. J. Peasant Stud. 36, 171–196. Shannon, C.E., 1948. A mathematical theory of communication. Bell Syst. Tech. J. 27, 379–423. Shinn, J.E., 2016. Adaptive environmental governance of changing social-ecological systems: empirical insights from the Okavango Delta, Botswana. Global Environ. Change 40, 50–59. Shuiabi, E., Thomson, V., Bhuiyan, N., 2005. Entropy as a measure of operational flexibility. Eur. J. Oper. Res. 165, 696–707. Quiroga, Sonia, Suárez, Cristina, Solís, Juan Diego, 2015. Exploring coffee farmers' awareness about climate change and water needs: smallholders' perceptions of adaptive capacity. Environ. Sci. Policy 45, 53–66. Smit, B., Burton, I., Klein, R.J.T., et al., 1999. The science of a adaptation-a framework for assessment. Mitig. Adapt. Strat. Glob. Change 4, 199–213. Smit, B., Wandel, J., 2006. Adaptation, adaptive capacity and vulnerability. Global Environ. Change 16, 282–292. Snorek, J., Renaud, F.G., Kloos, J., 2014. Divergent adaptation to climate variability: a case study of pastoral and agricultural societies in Niger. Global Environ. Change 29, 371–386. Speranza, I.C., Wiesmann, U., Rist, S., 2014. An indicator framework for assessing livelihood resilience in the context of social–ecological dynamics. Global Environ. Change 28, 109–119. Thomas, D.S.G., Twyman, C., 2005. Equity and justice in climate change adaptation amongst natural-resource-dependent societies. Global Environ. Change 15, 115–124. Turner 2nd, B.L., Matson, P.A., McCarthy, J.J., Corell, R.W., Christensen, L., Eckley, N., Hovelsrud-Broda, G.K., Kasperson, J.X., Kasperson, R.E., Luers, A., Martello, M.L., Mathiesen, S., Naylor, R., Polsky, C., Pulsipher, A., Schiller, A., Selin, H., Tyler, N., 2003. Illustrating the coupled human-environment system for vulnerability analysis: three case studies. PNAS 100, 8080–8085. Warrick, O., Aalbersberg, W., Dumaru, P., McNaught, R., Teperman, K., 2016. The ‘Pacific Adaptive Capacity Analysis Framework’: guiding the assessment of adaptive capacity in Pacific island communities. Reg. Environ. Change 17, 1039–1051. Yin, S., Chen, J.Wu., Yang, K.S., X J.,, 2016. Adaptation of farming households under drought stress: based on a survey in the Minqin Oasis. Progr. Geogr. 35, 644–654. Yu, Z.L. Yang, Yang, X.J., T.,, 2013. Exploring conditions, determinants and mechanisms of rural households' adaptability to tourism development: a case study of Jinsixia in Qinling Mountains. Acta Geogr. Sin. 68, 143–1156. Zhang, J.Y., Dai, M.H., Wang, L.H., et al., 2016. Household livelihood change the rocky desertification control project in karst areas, Southwest China. Land Use Policy 56. Zhang, K., Feng, Q., Lü, Y.Q., et al., 2011. Study on spatial heterogeneity of soil water contents in Oasis-desert Belt of Minqin. J. Desert Res. 31, 1149–1155. Zhang, Q., Yao, Y.B., Li, Y.H., et al., 2015. Research progress and prospect on the monitoring and early warning and mitigation technology of meteorological drought disaster in Northwest China. Adv. Earth Sci. 30, 196–213. Zhang, T.R., Zhang, Y.F., Chai, X.M., et al., 2010. Impact of human activities on sandy desertification of northwestern China and countermeasures analysis. J. Desert Res. 30. Zhao, X.Y.Z., Liu, H.L., C F.,, 2015. The farmers' livelihood risk and their coping strategy in the downstream of Shiyang River: a case of Minqin Oasis. Geogr. Res. 34, 922–932.

Alam, K., 2015. Farmers’ adaptation to water scarcity in drought-prone environments: a case study of Rajshahi District, Bangladesh. Agric. Water Manage. 148, 196–206. Alauddin, M., Sarker, M.A.R., 2014. Climate change and farm-level adaptation decisions and strategies in drought-prone and groundwater-depleted areas of Bangladesh: an empirical investigation. Ecol. Econ. 106, 204–213. Barnett, J., O’Neill, S., 2010. Maladaptation. Global Environ. Change 20, 211–213. Below, T.B., Mutabazi, K.D., Kirschke, D., Franke, C., Sieber, S., Siebert, R., Tscherning, K., 2012. Can farmers’ adaptation to climate change be explained by socio-economic household-level variables? Global Environ. Change 22, 223–235. Burton, I., 1997. Vulnerability and adaptive response in the context of climate and climate change. Clim. Change 36, 185–196. Butler, J.R.A., Suadnya, W., Puspadi, K., Sutaryono, Y., Wise, R.M., Skewes, T.D., Kirono, D., Bohensky, E.L., Handayani, T., Habibi, P., Kisman, M., Suharto, I., Hanartani, Supartarningsih, S., Ripaldi, A., Fachry, A., Yanuartati, Y., Abbas, G., Duggan, K., Ash, A., 2014. Framing the application of adaptation pathways for rural livelihoods and global change in eastern Indonesian islands. Global Environ. Change 28, 368–382. Carr, E.R., 2008. Between structure and agency: Livelihoods and adaptation in Ghana’s Central Region. Global Environ. Change 18, 689–699. Chen, C., Doherty, M., Coffee, J., Wong, T., Hellmann, J., 2016. Measuring the adaptation gap: a framework for evaluating climate hazards and opportunities in urban areas. Environ. Sci. Policy 66, 403–419. Chen, H., Wang, J., Huang, J., 2014. Policy support, social capital, and farmers’ adaptation to drought in China. Global Environ. Change 24, 193–202. Danfeng, S., Dawson, R., Baoguo, L., 2006. Agricultural causes of desertification risk in Minqin, China. J. Environ. Manage. 79, 348–356. Dessai, S., Lu, X., Risbey, J.S., 2005. On the role of climate scenarios for adaptation planning. Global Environ. Change 15, 87–97. DFID, 2001. In: Sustainable Livelihoods Guidance Sheets. Department for International Development, London, pp. 68–125. Engle, N.L., 2011. Adaptive capacity and its assessment. Global Environ. Change 21, 647–656. Fankhauser, S., Smith, J.B., Tol, R.S.J., 1999. Weathering climate change: some simple rules to guide adaptation decisions. Ecol. Econ. 30, 67–78. Fazey, I., Kesby, M., Evely, A., Latham, I., Wagatora, D., Hagasua, J.-E., Reed, M.S., Christie, M., 2010. A three-tiered approach to participatory vulnerability assessment in the Solomon Islands. Global Environ. Change 20, 713–728. Feike, T., Khor, L.Y., Mamitimin, Y., Ha, N., Li, L., Abdusalih, N., Xiao, H., Doluschitz, R., 2017. Determinants of cotton farmers’ irrigation water management in arid Northwestern China. Agric. Water Manage. 187, 1–10. Grothmann, T., Patt, A., 2005. Adaptive capacity and human cognition: the process of individual adaptation to climate change. Global Environ. Change 15, 199–213. Haglund, E., Ndjeunga, J., Snook, L., Pasternak, D., 2011. Dry land tree management for improved household livelihoods: farmer managed natural regeneration in Niger. J. Environ. Manage. 92, 1696–1705. Hoque, S.F., Quinn, C., Sallu, S., 2017. Differential livelihood adaptation to social-ecological change in coastal Bangladesh. Reg. Environ. Change. Holling, C.S., 2001. Understanding the complexity of economic, ecological, and social systems. Ecosystems 4, 390–405. IPCC, 2001. Climate Change 2001: Impacts-Adaptation-and-Vulnerability. In: Contribution of Working Group II to the Third Assessment Report of the Intergovernmental Panel on Climate Change. Cambridge University Press. IPCC, 2014. Climate Change 2014: Impacts, Adaptation, and Vulnerability Part B: Regional Aspects. In: Working Group II Contribution to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change. Cambridge University Press. John Smithers, B.S., 1997. Human adaptation to climatic variability and change. Global Environ. Change 7, 129–146. Kalaugher, E., Bornman, J.F., Clark, A., Beukes, P., 2013. An integrated biophysical and socio-economic framework for analysis of climate change adaptation strategies: the case of a New Zealand dairy farming system. Environ. Modell. Software 39, 176–187. Khayyati, M., Aazami, M., 2016. Drought impact assessment on rural livelihood systems in Iran. Ecol. Ind. 69, 850–858. Li, F., Liang, J., Clarke, K., Li, M., Liu, Y., Huang, Q., 2014. Urban land growth in eastern China: a general analytical framework based on the role of urban micro-agents’ adaptive behavior. Reg. Environ. Change 15, 695–707. Li, X., Kunnathur, A., Ragu-Nathan, T., Jitpaiboon, T., 2008. Development and validation of learning capability construct in IOS supply chain network context, Decision Sciences Institute 2008 Annual Meeting. Baltimore. http://www. decisionsciences. org/Proceedings/DSI2008/docs/253-9820.pdf (accessed 20.10. 12). McCubbin, S., Smit, B., Pearce, T., 2015. Where does climate fit? Vulnerability to climate change in the context of multiple stressors in Funafuti, Tuvalu. Global Environ. Change 30, 43–55. Mortimore, Michael J., Adams, W.M., 2001. Farmer adaptation, change and 'crisis' in the Sahel. Global Environ. Change 11, 49–57. Monirul Alam, G.M., Alam, K., Shahbaz, M., 2016. Influence of institutional access and social capital on adaptation choices: empirical evidence from vulnerable rural households in Bangladesh. Ecol. Econ. 130, 243–251. Motsholapheko, M.R., Kgathi, D.L., Vanderpost, C., 2011. Rural livelihoods and household adaptation to extreme flooding in the Okavango Delta, Botswana. Phys. Chem. Earth, Parts A/B/C 36, 984–995.

423