Linking ecosystem services supply, social demand and human well-being in a typical mountain–oasis–desert area, Xinjiang, China

Ecosystem Services 31 (2018) 44–57

Contents lists available at ScienceDirect

Ecosystem Services journal homepage: www.elsevier.com/locate/ecoser

Linking ecosystem services supply, social demand and human well-being in a typical mountain–oasis–desert area, Xinjiang, China Hejie Wei a,b, Huiming Liu c, Zihan Xu a,b, Jiahui Ren a,b, Nachuan Lu a,b, Weiguo Fan a,b, Peng Zhang a,b, Xiaobin Dong a,b,d,⇑ a

State Key Laboratory of Earth Surface Processes and Resource Ecology, Faculty of Geographical Science, Beijing Normal University, Beijing 100875, People’s Republic of China College of Resources Science and Technology, Faculty of Geographical Science, Beijing Normal University, Beijing 100875, People’s Republic of China Satellite Environment Center, Ministry of Environmental Protection, Beijing 100094, People’s Republic of China d Joint Center for Global Change and China Green Development, Beijing Normal University, Beijing 100875, People’s Republic of China b c

a r t i c l e

i n f o

Article history: Received 14 February 2018 Accepted 19 March 2018

Keywords: Ecosystem services supply Ecosystem services social demand Supply–demand mismatches Human well-being Mountain–oasis–desert area

a b s t r a c t Identifying the links among ecosystem services (ES) supply, social demand and human well-being is important to realize sustainability, especially in mountain–oasis–desert (MOD) areas, which are facing an intense conflict between socioeconomic development and ecological conservation. Using a biophysical model, we mapped six ES in the Manas River Basin, which is a typical MOD area. A questionnaire survey was employed to evaluate social demand for ES and human well-being in four different regional units (i.e., high mountain, low hills, oasis and desert) in our site. Spider diagrams were applied to identify the links among ES supply, social demand and human well-being. The results showed that a high supply of provisioning services occurred in the oasis, while a high supply of regulating services existed in the high mountain region. The ES social demand was not completely accordant with the biophysical supply in spatial distribution, and the factors from the supply side and demand side could both cause ES supply– demand mismatches. The total well-being level of all indicators was higher in the oasis and desert than in the upstream areas (i.e., the high mountain region and low hills region), but some indicators (e.g., water consumption) were the inverse. The supply–demand mismatches in provisioning services had a strong impact on human well-being, while the supply–demand mismatches in regulating services had a low impact on human well-being. This can be explained by the ES social demand questionnaire results, which showed that the level of social importance was higher for provisioning services than for regulating services at our site. In accordance with our results, we recommended several policies to promote ecological conservation and improve human well-being in the Manas River Basin, and these policies could also be applied in other MOD areas. Ó 2018 Published by Elsevier B.V.

1. Introduction Ecosystem services (ES) refer to the direct and indirect contributions to human well-being that originate from ecosystems (de Groot et al., 2010); therefore, a strong connection exists between ES and human well-being. Understanding the relationships between ES and human well-being is important not only for the purpose of scientific research but also to inform policy and practice (Alkemade et al., 2014; Bennett et al., 2015; Daily et al., 2009; Geijzendorffer et al., 2017). Since the Millennium Ecosystem Assessment (MA) proposed a conceptual framework about the

⇑ Corresponding author at: Faculty of Geographical Science, Beijing Normal University, Beijing 100875, People’s Republic of China. E-mail address: [email protected] (X. Dong). https://doi.org/10.1016/j.ecoser.2018.03.012 2212-0416/Ó 2018 Published by Elsevier B.V.

relationships between ES and human well-being on a global scale (Millennium Ecosystem Assessment, 2005), the number of studies integrating ES and human well-being has gradually increased (Bennett et al., 2015; Costanza et al., 2007; Haines-Young and Potschin, 2010; Horcea-Milcu et al., 2016; Hossain et al., 2017; Santos-Martin et al., 2013; Yang et al., 2010). Haines-Young and Potschin (2010) proposed the cascade framework from biodiversity to human well-being. Santos-Martin et al. (2013) used the DPSIR (Driver-Pressure-State-Impact-Response) framework to unravel the relationships between ES and human well-being. HorceaMilcu et al. (2016) constructed the conceptual model of mediating factors to disaggregate the contributions of ES to human wellbeing. In a case study, Yang et al. (2010) quantified the change in ES and farmers’ well-being in the city of Guyuan in the Loess Plateau and analyzed the relationships between them. Abunge et al.

H. Wei et al. / Ecosystem Services 31 (2018) 44–57

(2013) applied a participatory well-being assessment in Kenya to connect marine ES and fishermen’s well-being. Wang et al. (2014) used cluster analysis to investigate different patterns of relationships between ES and human well-being in the upstream watershed of the Miyun Reservoir. Ciftcioglu (2017) applied a socio-cultural preference method to evaluate the relative values between ES and human well-being based on local people’s perceptions in the Lefke Region in North Cyprus. However, most case studies have applied the single supply-side or demand-side to assess ES and have then analyzed the relationships between ES and human well-being. The relationships between ES supply–demand match states and human well-being are neglected and still unclear. Wei et al. (2017) noted that ES supply–demand mismatches may impact human well-being by causing unsatisfied demand. Studying the relationships between ES supply and social demand is important to understand the interactions between ES and human well-being, especially on a small spatial scale. In recent years, several studies have attempted to integrate the supply and social demand in ES assessment. Kroll et al. (2012) quantified the supply–demand ratios of food, energy and water from 1999 to 2007 in the Leipzig–Halle region of Germany to reveal the relationships between land use change and ES. Burkhard et al. (2012) proposed a matrix approach based on land cover types to map imbalances between ES supply and demand. Schulp et al. (2014) mapped the supply and demand of agricultural pollination services in Europe and determined that the demand area was larger than the supply area. Castro et al. (2014) applied different value-dimensions to evaluate ES and established that high mountains and coastal platforms exhibited the greatest differences between ES biophysical supply and social preference in the province of Almería, Spain. Baró et al. (2015) identified mismatches between regulating services supply and demand based on environmental quality standards in five European cities. However, these case studies almost always conclude at the supply–demand relationship stage and lack further analysis about the impact of the mismatches on human well-being. We need to fully understand the ES supply-side and demand-side to successfully link ES and human well-being and to formulate scientific decisions (Geijzendorffer et al., 2017; Mensah et al., 2017; Wang et al., 2017a). To resolve the problem, employing interdisciplinary approaches (e.g., biophysical and sociological approaches) to assess ES supply, social demand and human well-being is greatly needed but severely limited. Linking ES and human well-being has attracted much attention, but case studies in a mountain–oasis–desert (MOD) area are limited (Fu et al., 2017; Xu et al., 2016). The combination of a mountain, oasis and desert has been defined as an MOD system (Zhang, 2001). The MOD system is a coupled socio-ecological system that faces conflict between socioeconomic development and ecological conservation. The Manas River Basin (MRB) is a typical representative of MODs in the arid region. Extending from upstream to downstream in the MRB, natural and socioeconomic conditions have considerable differences, and the relationships between ES and human well-being remain to be elucidated. In this study, we aim to: (i) clarify the spatial patterns of ES supply, (ii) quantify ES social demand and human well-being in four different regional units (i.e., high mountain, low hills, oasis and desert) and (iii) identify the links among ES supply, social demand and human well-being for policy decisions (i.e., the mismatches between ES supply and social demand and the relationships between ES and human well-being).

45

and the Gurbantunggut Desert (Yuan et al., 1995). The basin is in an arid area, and the landscape is dominated by mountains, plains and sand dunes. Following the construction of a key water-control project and reservoir in the plain, the Manasi River and adjacent rivers have been linked by crisscrossing channels, so the scope of the MRB often includes the Manasi River (the longest river) and three adjacent rivers, the Bayingou River, Jingou River and Taxi River (Cheng et al., 2006) (Fig. 1). The annual average runoff volume of all the rivers is 2.3  109 m3 (Cheng et al., 2006). The MRB covers an area of approximately 2.29  104 km2. The longest distance from east to west is 198.7 km, and the longest distance from north to south is 260.8 km. In the MRB, grasslands, croplands and sand are the major land use types and account for 2.85%, 23.11% and 22.06% of the area, respectively. The administrative districts in the MRB include the Shihezi reclamation area, Shawan County and Manasi County (Yuan et al., 1995). The population of our site was 5.91  104 in 1949 and increased to approximately one million in 2015. During the past sixty years, along with the desert turning into an oasis, the MRB has been a rapid economic development area. Our site is also the grain, oil and cotton production base for the Xinjiang Uygur Autonomous Region (Zhang et al., 2012). Mountains, oases and deserts are the primary geographic landforms in inland arid regions in China (Xu et al., 2016). From upstream to downstream in the MRB, geographic landforms successively occur as mountain, oasis and desert. According to the differences in natural ecological and socioeconomic conditions, the MRB is divided into four regional units, including the high mountain, low hills, oasis and desert (Fig. 1 and Table 1). Human wellbeing and ES social demand are involved in our study, and there must be a certain population in the regional units, so the transitional zone between the oasis and desert is included in the scope of the desert area. The dominant physical geography and socioeconomic characteristics in different regional units are listed in Table 1. In the high mountain region and low hills region in the upstream watershed, the population is sparse and composed primarily of minorities, and socioeconomic development is relatively backward. Moreover, soil fertility is low, which is not conducive to crop farming. The oasis region in the middle of the watershed supports approximately 76.65% of the total population and has the highest urbanization rate with the largest city, Shihezi. Additionally, agriculture and industry are both relatively developed, and economic output accounts for approximately 85.11% of the total gross domestic product. The desert region in the downstream watershed contains approximately 17.74% of the total population, and the agricultural population accounts for 60.00% of the total population. Agriculture has reached a relatively high level, but the desert region has low rainfall, which is a restriction on agricultural development. Land use change (e.g., desert turning into oasis) has promoted the socioeconomic development in recent decades in the MRB, but it has also caused ecological problems (e.g., grass degradation and soil loss). Aiming to address the ecological problems caused by socioeconomic development, a series of ecological conservation policies have been implemented, such as closing hillsides to ban grazing and returning grazing land to pasture (Li et al., 2015; Liu and Zheng, 2014). However, determining how to realize a win-win between promoting ecological protection and improving human well-being is still a difficult problem.

3. Methods 2. Study area

3.1. Research framework

Located in northern Xinjiang, China (43°270 -45°210 N and 85°0 10 -86°320 E) (Fig. 1), the MRB is adjacent to Tianshan Mountain

Considering both the supply-side and demand-side ES, our study aimed to identify the links among ES supply, social demand

46

H. Wei et al. / Ecosystem Services 31 (2018) 44–57

Fig. 1. Map of the study area: (a) location, (b) regional units and sample points and (c) land use types. Note: Sample points indicate the approximate locations of the questionnaire survey.

Table 1 Main characteristics of different regional units in the Manas River Basin. Regional units

High mountain

Low hills

Oasis

Desert

Data sources/references

Location Area (103 km2) Mean altitude (m) Main land use types

Upstream 5.34 3212.12 Grassland, woodland and glaciers

Upstream 4.17 1061.45 Grassland and cropland

Midstream 5.76 418.42 Cropland, grassland and urban

Downstream 7.68 337.57 Sand and cropland

Population density (inhab./km2) Urbanization rate (%)

1.67

11.63

136.18

23.63

19.99

24.79

67.71

40.87

3.35

4.31

8.37

5.01

5.19 382.59

5.26 269.99

7.67 202.31

7.14 126.59

– Cheng et al. (2006) Geospatial Data Cloud (http://www.gscloud.cn/) Data Center for Resources and Environmental Sciences, Chinese Academy of Sciences (http://www.resdc.cn) Statistical data in 2015 (Statistics Bureau of Xinjiang Uygur Autonomous Region, 2016) Statistical data in 2015 (Statistics Bureau of Xinjiang Uygur Autonomous Region, 2016) Statistical data in 2015 (Statistics Bureau of Xinjiang Uygur Autonomous Region, 2016) National Climatic Bureau (http://data.cma.cn) National Climatic Bureau (http://data.cma.cn)

4

Per capita GDP (10 RMB/inhab.) Temperature (°C)* Precipitation (mm)* *

The reference year for the temperature and precipitation is 2015.

47

H. Wei et al. / Ecosystem Services 31 (2018) 44–57

Fig. 2. The framework to link ecosystem services supply, social demand and human well-being in the Manas River Basin.

and human well-being across different regional units based on a framework (Fig. 2). We selected six ES, including cultivated crops, soil conservation, water regulation, sand fixation, climate regulation and habitat, and each basis for selection was listed in Table 2. To link ES supply, social demand and human well-being, we first modeled the biophysical supply of six ES and identified the highsupply areas for ES. Second, using the face-to-face questionnaire, we evaluated the social demand for ES by analyzing the social preferences of stakeholders in different regional units. The stakeholders were primarily local inhabitants, including farmers and civilians. Third, we designed the indicator system of human wellbeing at our site and assessed the well-being level in four different regional units with a questionnaire. Lastly, we compared the levels of ES supply and social demand to identify the supply–demand mismatches across different regional units. The mismatches

reflected the unequal status between ES supply and social demand and were relative results. Additionally, we applied spider diagrams and Spearman’s rank correlations to analyze and examine the spatial correspondence among ES supply, social demand and human well-being. 3.2. Ecosystem services supply: modeling and mapping ES supply includes the components of a given ecosystem based on biophysical properties, ecological functions, and social properties in a particular area, according to most studies (Burkhard et al., 2012; Crossman et al., 2013; Villamagna et al., 2013). We applied different indicators to quantify and map six ES supplies (Table 2). The crop production index was chosen to quantify the supply of cultivated crops and to consider the linear relationships

Table 2 Selecting a basis for ecosystem services (ES) and quantification models for the ES supply indicators. ES Provisioning services Cultivated crops

Regulating services Soil conservation

Water regulation

Sand fixation Climate regulation

Habitat

Selection basis

ES supply indicators

Quantification models

Crop production is directly related to local inhabitants’ well-being, and the spatial supply in different regional units has clear differences.

Crop production index (calculation by net primary production)

Carnegie Ames Stanford Approach (CASA) (Field et al., 1995; Potter et al., 1993)

The altitude rapidly declines from upstream to middle and downstream areas in our site, and the soil erosion is considerable. Water is the lifeblood of arid regions, and the Manas River Basin is located in an arid region. Our site is adjacent to the desert, and sand erosion by wind is severe. Our site is located in an arid area, and its ecosystems are very sensitive to climate change. The desert and forest ecosystems in our site are the key areas for biodiversity conservation.

Soil retention amount per unit area (t/km2)

Revised universal soil loss equation (RUSLE) (Renard et al., 1991)

Water yields (mm)

Water balance equation (Budyko, 1974; Potter et al., 2005)

Sand fixation amount per unit area (t/km2) Carbon sequestration of noncrop plants (t/km2)

Revised wind erosion equation (RWEQ) (Fryrear et al., 2000) Carnegie Ames Stanford Approach (CASA) (Field et al., 1995; Potter et al., 1993) Integrate valuation of ES and tradeoffs (InVEST) (Tallis et al., 2013)

Habitat quality index

48

H. Wei et al. / Ecosystem Services 31 (2018) 44–57

between crop production and net primary production (NPP), in which NPP was standardized to reflect the crop production index. The Carnegie Ames Stanford Approach (CASA) (Field et al., 1995; Potter et al., 1993) was employed to estimate the NPP. We used the soil retention amount to quantify the supply of soil conservation, and the revised universal soil loss equation (RUSLE) (Renard et al., 1991) was employed to estimate the soil loss and retention. Water yields were applied to quantify the supply of water regulation and estimated with the water balance equation (Budyko, 1974). The supply of sand fixation was reflected by the sand fixation amount, which was modeled by the revised wind erosion equation (RWEQ) (Fryrear et al., 2000). Because the carbon sequestration of crop plants would return to the atmosphere, we applied the carbon sequestration of non-crop plants to quantify the supply of climate regulation (Raudsepp-Hearne et al., 2010). The Integrated Valuation of ES and Tradeoffs (InVEST) (Tallis et al., 2013) was employed to map the habitat quality index, which was used to quantify the supply of habitat. 3.2.1. Net primary production CASA inverted NPP belongs to a typical model of light use efficiency (Potter et al., 1993). The approach is relatively simple, and most parameters can be obtained from remote sensing data. The complete equation for the NPP calculation is as follows (Field et al., 1995):

NPP ¼ 0:5  Rs  FPAR  f 1  f 2  W  emax

ð1Þ

where the constant 0.5 is the ratio of effective solar radiation for vegetation use accounting for solar radiation; Rs is the solar radiation (106 MJ/km2); FPAR is the fraction of photosynthetically active radiation absorbed by the vegetation canopy; f1 and f2 reflect the stress of extreme temperatures on plant photosynthesis; T reflects the stress of moisture on plant photosynthesis; and emax is the maximum light use efficiency of vegetation growth (106 t C/MJ), which was set as the simulation results from Zhu et al. (2006). The detailed calculation processes for each parameter refer to Wei et al. (2016). 3.2.2. Soil erosion and retention The soil retention amount can be calculated by the difference between the potential amount and the actual amount of soil erosion, and the key equations are as follows (Renard et al., 1991):

Ap ¼ R  K sl  LS

ð2Þ

Ar ¼ R  K sl  LS  C sl  P

ð3Þ

Ac ¼ Ap  Ar

ð4Þ

where Ap is the potential amount of soil erosion (102 t/km2); R reflects the rainfall erosion effect (102 MJ mm/km2 h); Ksl is the soil erodibility factor (t km2 h/km2 MJ mm); LS is the topographic factor; Csl is the vegetation cover factor; P reflects the factor of soil and water conservation measures and is closely linked to land use types; and Ac is the actual amount of soil erosion (102 t/km2). The detailed calculation processes for each parameter refer to Ouyang et al. (2016). 3.2.3. Water yields In arid areas, water yields primarily depend on precipitation and evapotranspiration and can be calculated by the difference between them (Budyko, 1974; Potter et al., 2005). Evapotranspiration can be estimated by precipitation, potential evapotranspiration and the leaf area index (Sun et al., 2011). The key formulas for water yield calculation are as follows:

WR ¼ PPT  ET

ð5Þ

ET ¼ c þ b  PPT  PET þ a  PPT  LAI

ð6Þ

where WR is the water yield (mm); PPT is the precipitation (mm); ET is the evapotranspiration (mm); PET is the potential evapotranspiration (mm); and LAI is the leaf area index. The calculation of each parameter refers to Jia et al. (2014). 3.2.4. Sand erosion and fixation The sand fixation amount can be calculated by the difference between the potential amount and the actual amount of sand erosion by wind. The key equations for the actual amount of sand erosion are as follows (Fryrear et al., 2000):

SF r ¼

2 2z  Q max  eðz=Sr Þ Sr

ð7Þ

Sr ¼ 150:71  ðWF  EF  SCF  K  CÞ0:3711

ð8Þ

Q max ¼ 109:8  ðWF  EF  SCF  K  CÞ

ð9Þ 2

where SFr is the actual amount of sand erosion (kg/m ); Z is the distance from the upwind edge of the field (m); Sr is the critical field length (m); Qmax is the maximum transport capacity (kg/m); WF reflects the influence of a climate condition on sand erosion by wind; EF is the soil erodible factor; SCF is the soil crusting factor; K is the surface roughness factor; and C is the vegetation factor. The detailed calculations for each parameter refer to Ouyang et al. (2016). 3.2.5. Habitat quality Based on the land use types, the InVEST habitat quality model combines the habitat suitability for land use, intensity of threatening factors and sensitivity of land use types to threatening factors (Tallis et al., 2013). According to the local characteristics of the MRB, our study selected several threatening factors, including settlement, farmland, river, roads and population. The specific settings for the parameters (i.e., the effect intensity of each threatening factor, habitat suitability of each land use type and sensitivity degree of land use types to threatening factors) refer to the detailed description of the InVEST tool (Tallis et al., 2013) and to the actual conditions in the MRB. 3.2.6. Data sources The land use datasets from 2015 were interpreted by Landsat OLI/TIRS (30 m) and were provided by the Data Center for Resources and Environmental Sciences, Chinese Academy of Sciences (RESDC) (http://www.resdc.cn). Meteorological data for stations (Table 3) were collected from the National Climatic Bureau (http://data.cma.cn) and the local weather bureau, including daily temperature, daily precipitation, daily wind and daily solar radiation in 2015. Kriging interpolation was applied to produce raster maps of meteorological data. The 1 km MODIS (Moderate Resolution Imaging Spectroradiometer) NDVI (normalized difference vegetation index) data were obtained from NASA’s Earth data (https:// lpdaac.usgs.gov/). Soil attribute data were derived from the Soil Survey Office of the Xinjiang Uygur Autonomous Region, including the content of sand, silt, clay and organic carbon. The 90 m DEM (digital elevation model) was obtained from the Geospatial Data Cloud (http://www.gscloud.cn/). 3.3. Ecosystem services social demand and human well-being questionnaire ES social demand was defined as the importance or preference for ES specific attributes by stakeholders or society (MartínLópez et al., 2012; Villamagna et al., 2013). Human well-being was defined as the valuable activities and status based on the experiences of the inhabitants (Millennium Ecosystem Assessment,

49

H. Wei et al. / Ecosystem Services 31 (2018) 44–57 Table 3 Information from weather stations within and near the Manas River Basin. Name

Latitude (° N)

Longitude (° E)

Altitude (m)

Name

Latitude (° N)

Longitude (° E)

Altitude (m)

Shihezi Manasi Mosuowan Paotai Shawan Shimenzi Kensiwate Fuhai

44.32 44.32 45.02 44.85 44.33 43.87 43.97 47.12

86.05 86.20 86.10 85.25 85.62 86.23 85.95 87.47

444 472 347 338 523 1358 861 501

Hebukesaier Tuoli Kelamayi Wusu Caijiahu Dabancheng Wulumuqi Bayinbuluke

46.78 45.93 45.62 44.43 44.20 43.35 43.78 43.03

85.72 83.60 84.85 84.67 87.53 88.32 87.65 84.15

1292 1078 450 479 441 1104 935 2458

2005). ES social demand and human well-being in different regional units in the MRB were evaluated by a participatory approach. We conducted face-to-face interviews and questionnaires with 900 inhabitants in 19 randomly sampled points (Fig. 1b) from the upstream to the downstream areas and obtained 815 valid questionnaires. The timing of the questionnaire survey was in August and September 2017, and the respondents were mainly local inhabitants (e.g., farmers and civilians). The structural proportion of the respondents was in accordance with the proportion of urban and rural inhabitants, as the proportion could have influenced the results. Within 815 completed questionnaires, the high mountain, low hills, oasis and desert regions had 18, 77, 505 and 202 respondents, respectively. The contents of the questionnaire were composed of three main parts: (1) the general information of the respondents, including sex, age, education, location, occupation and basic family details (Table 4); (2) the ES social demand of the respondents, including the ES perception and the importance of ES to their well-being; and (3) the subjective well-being level of the respondents. 3.3.1. Ecosystem services social demand assessment The social demand for ES can be formed at different spatial scales (Geijzendorffer and Roche, 2014). In this paper, in order to focus on the local inhabitants, we only quantified the local demand for ES in the MRB. The specific process of evaluating ES social demand included the following steps. First, we explained the six ES to improve the understanding of the population interviewed. Second, the respondents were asked whether they perceived the importance of ES in their social lives. Lastly, referencing the method by Iniesta-Arandia et al. (2014), we compelled the respondents to select the most important ES for their well-being or the population’s well-being (no more than four) from the six ES. We calculated the percentage of people who viewed a particular ecosystem service as important, and the social demand level was quantified by the percentage (Castro et al., 2014). 3.3.2. Human well-being assessment Several studies (MA, 2005; Smith et al., 2013; Wang et al., 2017a; Wang et al., 2017b) have shown that income, material needs, health and security are the important elements for human well-being. We established a set of 14 indicators from 3 domains (i.e., the basic requirements for a high quality of life, health and

security) to assess human well-being. The determination of evaluation indicators was required to follow the principles of relative independence, difference, stability and operability. The selection of 14 indicators combined the expert opinions and the local characteristics of the MRB, which included income, food consumption, water consumption, housing, energy consumption, traffic, physical health, mental health, food variation, medical facilities, public security, water quality, air quality, and avoidance of natural hazards. The mark method was employed to obtain the weight of each indicator, and all the respondents were asked to mark the indicators (Wang et al., 2017a). Subjective satisfaction was applied to obtain the well-being level for each indicator. A Likert scale was widely used in the survey research and was employed to evaluate the satisfaction degree of the respondents to the well-being indicators (i.e., strongly disagree, disagree, neutral, agree, and strongly agree). The responses of the population interviewed were coded in order from the lowest score of 0.2 (strongly disagree) to the highest score of 1.0 (strongly agree). The higher the score, the higher the satisfaction degree of the respondents to each indicator. The well-being levels in four different regional units were quantified by the mean level of the satisfaction degree of all respondents within the area of the regional unit. 3.4. Data processing and analysis The evaluation values for ES social demand and human wellbeing were in the range of 0–1, and we normalized the values of ES supply maps to the range of 0–1 so they could be compared to ES social demand and human well-being. The minimum–maximum approach was employed to normalize ES supply maps (Willemen et al., 2010). Using spatial analysis techniques in ArcGIS, the level of ES supply in different regional units was calculated. We identified the ES supply–demand match status across different regional units, and Spearman’s rank correlations were applied to test the correlation between the ES supply and social demand in each regional unit. In the four regional units, we explored the relationships between ES and human well-being by comparing five indicators, including the supply of provisioning services, supply of regulating services, social demand for provisioning services, social demand for regulating services and human well-being. Spider diagrams were

Table 4 Main characteristics of the interviewees. Characteristics

Category/number (proportion)

Sex Age Education Location Years of residency Family members Occupation

Male/462 (56.69%); Female/353 (43.31%) <20 years/26 (3.19%); 20–39 years/254 (31.17%); 40–59 years/505 (61.96%); 60 years/30 (3.68%) None/10 (1.23%); Primary/272 (33.37%); Secondary/269 (33.01%); University/264 (32.39%) High mountain/18 (2.21%); Low hills/77 (9.45%); Oasis/505 (61.96%); Desert/202 (24.79%) <5 years/11 (1.35%); 5–19 years/111 (13.62%); 20–39 years/354 (43.44%); 40 years/339 (41.60%) 2 people/67 (8.22%); 3–4 people/597 (73.25%); 5 people/85 (10.43%); 6 people/66 (8.10%) Farming/347 (42.58%); Grazing/69 (8.47%); Enterprise or public institution/312 (38.28); Other/87 (10.67%)

50

H. Wei et al. / Ecosystem Services 31 (2018) 44–57

created to compare the interactions between ES and human wellbeing in different regional units. Spearman’s rank correlations were applied to examine the spatial correspondence between ES and human well-being. 4. Results 4.1. Mapping the supply of ecosystem services Affected by the spatial differences in natural ecological and socioeconomic characteristics, the supply of ES had substantial

spatial variation in our site. A high supply of provisioning services was distributed in the oasis and in the south of the desert (Fig. 3a), while a high supply of regulating services was concentrated in the north of the high mountain region (Fig. 3i). The croplands were almost all located in the oasis and desert regions (Fig. 1c), and only a few were in the low hills, so the supply of crops was mainly from these regions. The crop production was highest in the oasis (59.04% of the total production in the MRB) and was dominated by grains (e.g., corn and wheat), fruit (e.g., grapes and peaches) and vegetables (e.g., tomatoes). Crop production in the desert region accounted for 33.99% of the total produc-

Fig. 3. Spatial distribution of the supply of ecosystem services in our site in 2015: (a) cultivated crops (crop production index), (b) soil erosion by water, (c) soil conservation (soil retention), (d) water regulation (water yields), (e) sand erosion by wind, (f) sand fixation, (g) climate regulation (carbon sequestration of non-crop plants), (h) habitat (habitat quality) and (i) integrated regulating services (integrated regulating services index).

H. Wei et al. / Ecosystem Services 31 (2018) 44–57

tion and prioritized cotton and fruit. Crop production in the high mountain region and low hills region only accounted for 6.48% and 0.48% of the total production, respectively, and mainly consisted of grain (e.g., corn). The spatial distribution in soil erosion showed substantial differences (Fig. 3b), and the average erosion modulus was 2322 t/ km2 at our site. Soil erosion mainly occurred in the upstream watershed, and the erosion modulus exceeded 5000 t/km2, owing to intensive erosion according to the grading standard (Ministry of Water Resource of China, 2008). Soil erosion in the oasis and desert regions was low and less than 1000 t/km2, owing to micro erosion. The average soil retention modulus was 1838 t/km2 in the entire study area. Similar to the spatial distribution for soil erosion, soil retention also occurred mainly in the upstream watershed (Fig. 3c). In the high mountain region, the mean value of the soil retention modulus exceeded 7000 t/km2, and the soil retention amount accounted for 86.04% of the total amount in the basin. In the low hills region, the mean value of the soil retention modulus was 1051 t/km2, and the soil retention amount only accounted for 9.97% of the total. The soil retention modulus in the oasis and desert region was low and less than 150 t/km2. Located in an arid area, our site is a water shortage region, and water yields were low at less than 10 mm. A substantial spatial variation in water yields was detected (Fig. 3d). The high mountain region and low hills region were the main contributing areas of runoff generation, and the mean values of water yields in the two regions were 171 mm and 25 mm, respectively. The water yields in the oasis and desert regions were negative because a large number of crops were planted in these areas, which caused huge water consumption levels. Sand erosion by wind in the desert was the most severe (Fig. 3e), and the erosion modulus was 3264 t/km2, owing to middle degree erosion according to the grading standard (Ministry of Water Resource of China, 2008). Sand erosion in the oasis and the low hills regions was relatively low, and the erosion moduli were 318 t/km2 and 463 t/km2, respectively. The high mountain region had limited sand erosion, and the modulus was only 33 t/ km2. The average sand fixation modulus was 670 t/km2 in the entire study area, and the spatial distribution had considerable differences (Fig. 3f). Sand fixation mainly occurred in the oasis and desert regions, and the sand fixation quantity in the two regions accounted for 42.60% and 50.42% of the total quantity, respectively. The soil retention modulus was higher in the oasis and desert regions (1155 t/km2 and 1025 t/km2) than in the high mountain and low hills regions (less than 250 t/km2). The average carbon sequestration for non-crop plants was 52 g C/m2, and a spatial difference existed from the upstream to the downstream areas (Fig. 3 g). In the high mountain region and low hills region, the carbon sequestration was 52 g C/m2 and 81 g C/m2 and accounted for 23% and 28% of the total amount, respectively. In the oasis and desert regions, the carbon sequestration was 41 g C/m2 and 46 g C/m2 and accounted for 20% and 29% of the total amount, respectively. The total carbon sequestration amount had few differences in the different regional units. The habitat quality had high levels in the upstream and downstream areas but low value in the midstream watershed in our site (Fig. 3h). The spruce forests in the upstream watershed and the deserts in the downstream watershed were the important biological habitats in the MRB and biodiversity reserves in China. There were few threatening factors and human disturbances in the high mountain, low hills and desert regions, so the habitat quality was relatively high, and the indices were 0.50, 0.44 and 0.42, respectively. The oasis had many anthropogenic threatening factors (e.g., roads, settlement, cropland, and canals), so the habitat quality was poor, and the index was only 0.03.

51

4.2. Social demand for ecosystem services The social perception of ES showed that 86.68% of the respondents believed that the MRB provided important ES to society. In the entire study area, compared with regulating services, the respondents preferred provisioning services (Fig. 4), which was in accordance with the level of socioeconomic development. The MRB is located in western China, and compared with eastern China, its socioeconomic development level was relatively low, so the social demand of inhabitants for ES prioritized provisioning services. Fig. 4 shows that 80.00% of the respondents viewed cultivated crops as an important ES, while only 32.15% of the respondents viewed habitat as an important ES, which was the least valued ES. Additionally, corresponding with the arid region, water regulation was viewed as the most important regulating service. In different regional units in our site, the level of social demand for ES showed different patterns (Fig. 5b). Cultivated crops were considered a relatively important ES in all regional units. Soil conservation was viewed as more important in the high mountain region and low hills region, where soil erosion was severe (Fig. 3b). Similar to soil conservation, sand fixation was viewed as more important in the oasis and desert regions, where sand erosion was severe (Fig. 3e). Compared with other regional units, climate regulation and habitat were viewed as more important in the oasis. Compared with the upstream areas, water regulation was viewed as more important in the oasis and desert regions (midstream and downstream).

4.3. The mismatches between ecosystem services supply and social demand In the different regional units in our site, the biophysical supply of ES was not completely accordant with the social demand for ES (Fig. 5). For example, according to the Spearman’s rank correlations between ES supply and social demand in the four regional units, a

Fig. 4. The evaluation of the social demand for ecosystem services based on the importance of ecosystem services to the inhabitants’ well-being.

52

H. Wei et al. / Ecosystem Services 31 (2018) 44–57

Fig. 5. Ecosystem services supply and social demand in the different regional units. Standardization is based on a linear rescaling of values in the 0–1 range on the basis of the minimum and maximum in the ES supply maps (Fig. 3).

negative correlation was identified in the low hills region (Rho = 0.829; p < 0.05). Fig. 5 shows that in the high mountain region, the supply–demand mismatches mainly occurred with cultivated crops and habitat. Cultivated crops was viewed as an important ES, but its biophysical supply was the lowest. The supply of habitat was the highest, but its social perception level was lowest. In the low hills region, the biophysical supply of cultivated crops, soil conservation and sand fixation was at a relatively low level, but the social preference for these ES was relatively high. In the midstream and downstream areas, the serious supply–demand mismatches mainly occurred with regulating services. Water and climate regulation were viewed as important ES, but the biophysical supply of the two ES was low in the oasis and desert regions. Affected by the dense population, the oasis had the lowest supply of habitat; however, the social preference level of the respondents was the highest of the four regional units. Although the desert was an important biological habitat, it was underappreciated by local inhabitants.

4.4. Human well-being Of all the indicators of human well-being, income had the highest weight (Fig. 6a), which was in accordance with the relatively low level of socioeconomic development. Additionally, the weights of water consumption, housing, physical health, mental health and avoidance of natural hazards were also relatively high. In the entire study area, compared with the other indicators, food consumption, energy consumption and traffic had higher indicators of well-being (Fig. 6b), but their weight was low (Fig. 6a). Moreover, the satisfaction degree for mental health and food variation was relatively high, and the scores were greater than 0.6. Compared with the other indicators, water consumption, water quality and air quality had lower indicators of well-being (Fig. 6b), and the scores were less than 0.4. The well-being for other indicators (i.e., income, housing, physical health, medical facilities, public security and hazard avoidance) was in the middle, and the scores were in the range of 0.4–0.6.

H. Wei et al. / Ecosystem Services 31 (2018) 44–57

Across the different regional units, the well-being for each indicator had differences (Fig. 6c). For example, the well-being levels for income, food consumption, housing, energy consumption, traffic, food variation, medical facilities, and avoidance of natural hazards were highest in the oasis, followed by the desert, low hills and high mountain regions. However, the well-being levels for water consumption, water quality and air quality were highest in the high mountain region, followed by the low hills, oasis and desert regions.

53

the lowest well-being and the highest supply of regulating services. Spearman’s rank correlations also showed that human well-being had a negative correlation with the supply of regulating services in the different regional units (Rho = 1.00; p < 0.01). Moreover, across the different regional units, human well-being had a weak spatial correspondence with the supply–demand mismatches in regulating services.

5. Discussion 4.5. The relationships between ecosystem services and human wellbeing By comparing the five indicators related to ES and human wellbeing, spider diagrams showed the differences between ES and human well-being in the four regional units. High human wellbeing was linked with a high supply of provisioning services (Fig. 7). For example, both the subjective well-being and the supply of provisioning services were highest in the oasis but lowest in the high mountain region. Spearman’s rank correlations also showed that human well-being had a positive correlation with the supply of provisioning services in different regional units (Rho = 1.00; p < 0.01). Across the different regional units, human well-being had a strong spatial correspondence with the supply–demand mismatches in provisioning services (Fig. 7). For example, compared with the oasis and desert regions, the high mountain and low hills regions had relatively serious mismatches in provisioning services, and correspondingly, the human well-being in the two regions was poor. Poor human well-being was linked to the high supply of regulating services (Fig. 7). For example, the high mountain region had

5.1. Linking ecosystem services supply, social demand and human well-being Identifying the link among ES supply, social demand and human well-being has been a considerable challenge in realizing sustainability (Baró et al., 2015; Bennett et al., 2015; Millennium Ecosystem Assessment, 2005). In the past several years, some conceptual models have been proposed that integrate the ES supplyside and demand-side, such as the ecosystem properties, potentials, services and benefits framework (Bastian et al., 2013), the ES delivery chain (Villamagna et al., 2013) and the framework for the integrated assessment of ES supply and demand (Wei et al., 2016). Some indicators, such as the supply–demand ratio of food, energy, and carbon (Kroll et al., 2012; Zhao and Sander, 2015), and some approaches, such as the matrix (Burkhard et al., 2012) and participatory approach, were applied to identify the mismatches between ES supply and social demand. There have been some conceptual models attempting to link ES and human wellbeing, such as the cascade model (Haines-Young and Potschin,

Fig. 6. The weight and level of well-being indicators. Note: IN: Incomes; FC: Food consumption; WC: Water consumption; HO: Housing; EC: Energy consumption; TR: Traffic; PH: Physical health; MH: Mental health; VF: Variety in food; MF: Medical facilities; PS: Public security; WQ: Water quality; AQ: Air quality; ANH: Avoidance of natural hazards.

54

H. Wei et al. / Ecosystem Services 31 (2018) 44–57

Fig. 7. Spider diagrams that show the relationship between ecosystem services and human well-being. The value of the supply of ecosystem services in different regional units was calculated from Fig. 3a and i.

2010), the DPSIR model (Hou et al., 2014) and the social-ecological conceptual model for rural populations (Delgado and Marín, 2016). Cluster analysis (Raudsepp-Hearne et al., 2010; Wang et al., 2017a) and correlation analysis (Delgado and Marín, 2016; Wang et al., 2014) were employed to identify the relationships between ES and human well-being. Our framework (Fig. 2) has linked ES supply, social demand and human well-being by integrating multiple approaches and indicators, such as an ecological model (biophysical indicators), a participatory approach (social preference indicators) and statistical analysis (Spearman’s rank correlations). Subsequently, in the four regional units in our site, we identified the mismatches between ES supply and social demand and the relationships between ES and human well-being. Our results have showed that the ES biophysical supply was not completely accordant with the ES social demand in the spatial distribution. The factors from the supply-side and demand-side could both cause ES supply–demand mismatches. From the supply side, a weak ecosystem function could cause mismatches between ES supply and social demand. For example, in the low hills region, with low grassland coverage, the function of protecting soil and sand to avoid erosion by water and wind was weak, which became a dominant factor in causing the supply–demand mismatches in soil conservation and sand fixation. In another example, in the high mountain region and low hills region, soil fertility was poor, and the area of arable land was limited, which caused the supply–demand mismatches for cultivated crops. Several other studies also reached the same conclusions (Baró et al., 2015; Quintas-Soriano et al., 2014; Zhao and Sander, 2015). For example, because of the limited number of urban trees, carbon sequestration by trees could not compensate for carbon emissions in urban areas (Baró et al.,

2015; Zhao and Sander, 2015). On the demand side, high social demand for ES could also cause mismatches between ES supply and social demand. The high demand for ES may be related to economic development, population growth or land use change. For example, in the past several decades, population growth and economic development have caused a rapid expansion of the oasis in the midstream and downstream areas of our site, and agricultural irrigation has required a considerable quantity of water. Although most water yields in the upstream watershed were consumed by agricultural irrigation in the oasis, the water demand could still not be met. Several other studies observed similar situations (Gopalakrishnan et al., 2016; Jie et al., 2015; Kroll et al., 2012; Stürck et al., 2015). For example, Jie et al. (2015) noted that there was very high water demand in the Changsha–Zhuzhou–Xi angtan city cluster because of economic development and population growth and that the water yields from the nearby valley could not satisfy the city’s demand. Our results have shown that human well-being had a positive correlation with the supply of provisioning services but had a negative correlation with the supply of regulating services in the different regional units in our site. Several other studies have presented similar results (Millennium Ecosystem Assessment, 2005; Wang et al., 2014; Wang et al., 2017a). According to the MA (Millennium Ecosystem Assessment, 2005), in the past one hundred years, provisioning services and human well-being have increased, but the supply of most regulating services has been declining. The relationships between ES supply–demand mismatches and human well-being are complex (Fig. 7). Our case study has revealed that the mismatches between provisioning services supply and demand have a relatively strong impact on the

H. Wei et al. / Ecosystem Services 31 (2018) 44–57

total well-being. For example, in the high mountain region and low hills region, the supply–demand mismatch in cultivated crops could directly influence well-being through food variation (Fig. 6c). Moreover, because the local ecosystem could not provide several provisioning services (e.g., cotton and vegetables), the inhabitants in the high mountain region and low hills region may pay for those services in the markets; thus, the supply–demand mismatches in cultivated crops could have an indirect impact on income, which is the most important component of human wellbeing in our site. Our case study has also revealed that mismatches between regulating services supply and demand have a relatively low impact on the total well-being. These conclusions can be explained by the results of the questionnaire survey on ES social demand. In our site, the social preference was higher for provisioning services than for regulating services (Fig. 4). However, we cannot neglect the maintenance of regulating services, because they underlie the production of provisioning services (Rodríguez et al., 2006) and have an effect on human well-being in the long term (Wang et al., 2017a). Through the spider diagrams, we also determined that compared with the region with poor human well-being (e.g., the high mountain region), the region with high human wellbeing (e.g., the oasis) placed more importance on regulating services (Fig. 7). A possible explanation is as follows. Higher education levels and income often link a region with high human well-being (Martín-López et al., 2012), and the inhabitants in these areas may have higher ecological predispositions and awareness, so they viewed the regulating services as more important. 5.2. Policy implications Identifying the links among ES supply, social demand and human well-being is important not only for the purpose of scientific research but also to inform policy and practice (e.g., land use management and ecological compensation) (Alkemade et al., 2014; Daily et al., 2009; Geijzendorffer et al., 2017; Zheng et al., 2016). Our study results can provide valuable information to assist policy-makers or practitioners to maintain a particular ES or to improve human well-being. In the high mountain region, import policies can be employed to relieve the supply–demand mismatches in provisioning services. However, the well-being level for traffic is also poor (Fig. 6b), so improving the traffic conditions is very important not only for satisfying the social demand for provisioning services but also to improve the inhabitants’ well-being in the high mountain region. To overcome the supply–demand mismatch in soil conservation and sand fixation in the low hills region, multiple ecological conservation policies should continue to be implemented, including closing hillsides to ban grazing and returning grazing land to pasture (Li et al., 2015; Liu and Zheng, 2014). For relieving the supply–demand mismatches in water regulation in the oasis and desert regions, water-saving irrigation techniques (e.g., drip irrigation technology) should be popularized and can realize the efficient use of water resources. In the desert region, the habitat awareness of local inhabitants is low, which may further cause declines in habitat. The establishment of biodiversity reserves can be an effective measure in the key biodiversity regions of the desert. Compared with the oasis, the total well-being in the high mountain region and low hills region is lower, which is mainly due to low well-being levels for income, housing, traffic and medical facilities (Fig. 6c); thus, determining how to improve the well-being of these indicators is crucial. Although the income level in the upstream watershed is low, the ecological resources are of high quality, which can be used to promote eco-tourism to improve the income level of local inhabitants (Wang et al., 2017a). The high supply of regulating services in the upstream areas has benefited the midstream and downstream areas in the MRB. Ecological com-

55

pensation to the upstream watershed from the midstream and downstream watershed can improve the well-being of the inhabitants in the upstream watershed, which can realize poverty alleviation by ES. Morri et al. (2014) also considered that the midstream and downstream reaches of the Marecchia River Basin should pay a ‘‘debt” to the upstream area, because the high ES supply (e.g., carbon sequestration) of the latter benefited the former. In addition to monetary payments, the channels for ecological compensation should be further expanded. For example, when multiple nature conservation policies (e.g., herbage production in enclosures) are performed in the upstream area, the governments can implement affordable housing projects for the herdsmen and build roads to connect the midstream and upstream areas, which can both improve the well-being level of inhabitants in the upstream area.

6. Conclusions In the four different regional units from the upstream to the downstream areas in the MRB, biophysical models were applied to map the supply of six ES, and then, participatory approaches were employed to evaluate the social demand for six ES and human well-being. Our results showed that the biophysical supply of ES was not completely accordant with the social demand for ES in different regional units. The factors for the supply-side and demand-side could both cause ES supply–demand mismatches. The mismatches between provisioning services supply and demand had a strong impact on human well-being, while the mismatches between regulating services supply and demand had a low impact on human well-being. Across the four regional units, human well-being had a positive correlation with the supply of provisioning services but a negative correlation with the supply of regulating services. These conclusions can be explained with the questionnaire results of ES social demand, which showed that the social preference level in the MRB was higher for provisioning services than for regulating services. According to our results, we recommend several policy suggestions; for example, when multiple ecological conservation measures are performed in the upstream area, the midstream and downstream areas can implement ecological compensation through multiple channels, including improvement of the traffic, housing and medical conditions to enhance human well-being in the upstream area. Our results can provide valuable information to maintain a particular ES and improve human well-being not only for the MRB but also in other typical MOD areas.

Conflicts of interest The authors declare no conflict of interest.

Acknowledgements The authors are grateful to the National Natural Science Foundation of China (41671531, 41271549), the Key Project of the National Societal Science Foundation of China (15ZDB163), and the China Science & Technology Supporting Program (2017YFE0100400) for support of this research.

Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at https://doi.org/10.1016/j.ecoser.2018.03.012.

56

H. Wei et al. / Ecosystem Services 31 (2018) 44–57

References Abunge, C., Coulthard, S., Daw, T.M., 2013. Connecting marine ecosystem services to human well-being: insights from participatory well-being assessment in Kenya. Ambio 42, 1010–1021. https://doi.org/10.1007/s13280-013-0456-9. Alkemade, R., Burkhard, B., Crossman, N.D., Nedkov, S., Petz, K., 2014. Quantifying ecosystem services and indicators for science, policy and practice. Ecol. Indic. 37, 161–162. Baró, F., Haase, D., Gómez-Baggethun, E., Frantzeskaki, N., 2015. Mismatches between ecosystem services supply and demand in urban areas: a quantitative assessment in five European cities. Ecol. Indic. 55, 146–158. Bastian, O., Syrbe, R.U., Rosenberg, M., Rahe, D., Grunewald, K., 2013. The five pillar EPPS framework for quantifying, mapping and managing ecosystem services. Ecosyst. Serv. 4, 15–24. Bennett, E.M., Cramer, W., Begossi, A., Cundill, G., Diaz, S., Egoh, B.N., Geijzendorffer, I.R., Krug, C.B., Lavorel, S., Lazos, E., Lebel, L., Martin-Lopez, B., Meyfroidt, P., Mooney, H.A., Nel, J.L., Pascual, U., Payet, K., Harguindeguy, N.P., Peterson, G.D., Prieur-Richard, A.H.N., Reyers, B., Roebeling, P., Seppelt, R., Solan, M., Tschakert, P., Tscharntke, T., Turner, B.L., Verburg, P.H., Viglizzo, E.F., White, P.C.L., Woodward, G., 2015. Linking biodiversity, ecosystem services, and human well-being: three challenges for designing research for sustainability. Curr. Opin. Environ. Sustain. 14, 76–85. https://doi.org/10.1016/j.cosust.2015.03.007. Budyko, M.I., 1974. Climate and Life. Academy of Sciences, Main Observatory of Leningrad. Academic Press, New York. Burkhard, B., Kroll, F., Nedkov, S., Müller, F., 2012. Mapping ecosystem service supply, demand and budgets. Ecol. Indic. 21, 17–29. Cheng, W., Zhou, C., Liu, H., Zhang, Y., Jiang, Y., Zhang, Y., Yao, Y., 2006. The oasis expansion and eco-environment change over the last 50 years in Manas River Valley, Xinjiang. Sci. China Ser. D 49, 163–175. https://doi.org/10.1007/s11430004-5348-1. Ciftcioglu, G.C., 2017. Assessment of the relationship between ecosystem services and human wellbeing in the social-ecological landscapes of Lefke Region in North Cyprus. Landsc. Ecol. 32, 897–913. https://doi.org/10.1007/s10980-0170494-y. Costanza, R., Fisher, B., Ali, S., Beer, C., Bond, L., Boumans, R., Danigelis, N.L., Dickinson, J., Elliott, C., Farley, J., Gayer, D.E., Glenn, L.M., Hudspeth, T., Mahoney, D., McCahill, L., McIntosh, B., Reed, B., Rizvi, S.A.T., Rizzo, D.M., Simpatico, T., Snapp, R., 2007. Quality of life: an approach integrating opportunities, human needs, and subjective well-being. Ecol. Econ. 61, 267– 276. https://doi.org/10.1016/j.ecolecon.2006.02.023. Crossman, N.D., Burkhard, B., Nedkov, S., Willemen, L., Petz, K., Palomo, I., Drakou, E. G., Martín-Lopez, B., McPhearson, T., Boyanova, K., Alkemade, R., Egoh, B., Dunbar, M.B., Maes, J., 2013. A blueprint for mapping and modelling ecosystem services. Ecosyst. Serv. 4, 4–14. Daily, G.C., Polasky, S., Goldstein, J., Kareiva, P.M., Mooney, H.A., Pejchar, L., Ricketts, T.H., Salzman, J., Shallenberger, R., 2009. Ecosystem services in decision making: time to deliver. Front. Ecol. Environ. 7, 21–28. https://doi.org/10.1890/080025. de Groot, R.S., Alkemade, R., Braat, L., Hein, L., Willemen, L., 2010. Challenges in integrating the concept of ecosystem services and values in landscape planning, management and decision making. Ecol. Complex. 7, 260–272. https://doi.org/ 10.1016/j.ecocom.2009.10.006. Delgado, L.E., Marín, V.H., 2016. Well-being and the use of ecosystem services by rural households of the Río Cruces watershed, southern Chile. Ecosyst. Serv. 21, 81–91. https://doi.org/10.1016/j.ecoser.2016.07.017. Field, C.B., Randerson, J.T., Malmström, C.M., 1995. Global net primary production: combining ecology and remote sensing. Remote Sens. Environ. 51, 74–88. Fryrear, D.W., Bilbro, J.D., Saleh, A., Schomberg, H., Stout, J.E., Zobeck, T.M., 2000. RWEQ: improved wind erosion technology. J. Soil Water Conserv. 55 (2), 183– 189. Fu, Q., Li, B., Hou, Y., Bi, X., Zhang, X., 2017. Effects of land use and climate change on ecosystem services in Central Asia’s arid regions: a case study in Altay Prefecture, China. Sci. Total Environ. 607–608, 633–646. https://doi.org/ 10.1016/j.scitotenv.2017.06.241. Geijzendorffer, I.R., Roche, P.K., 2014. The relevant scales of ecosystem services demand. Ecosyst. Serv. 10, 49–51. Geijzendorffer, I.R., Cohen-Shacham, E., Cord, A.F., Cramer, W., Guerra, C., MartínLópez, B., 2017. Ecosystem services in global sustainability policies. Environ. Sci. Policy 74, 40–48. https://doi.org/10.1016/j.envsci.2017.04.017. Gopalakrishnan, V., Bakshi, B.R., Ziv, G., 2016. Assessing the capacity of local ecosystems to meet industrial demand for ecosystem services. AIChE J. 62, 3319–3333. Haines-Young, R., Potschin, M., 2010. The links between biodiversity, ecosystem services and human well-being. In: Raffaelli, D., Frid, C. (Eds.), Ecosystem Ecology: A New Systhsis. Cambridge University Press, Cambridge, UK. Horcea-Milcu, A., Leventon, J., Hanspach, J., Fischer, J., 2016. Disaggregated contributions of ecosystem services to human well-being: a case study from Eastern Europe. Reg. Environ. Change 16, 1779–1791. https://doi.org/10.1007/ s10113-016-0926-2. Hossain, M.S., Eigenbrod, F., Amoako Johnson, F., Dearing, J.A., 2017. Unravelling the interrelationships between ecosystem services and human wellbeing in the Bangladesh delta. Int. J. Sustain. Dev. World Ecol. 24, 120–134. https://doi.org/ 10.1080/13504509.2016.1182087. Hou, Y., Zhou, S., Burkhard, B., Müller, F., 2014. Socioeconomic influences on biodiversity, ecosystem services and human well-being: a quantitative

application of the DPSIR model in Jiangsu China. Sci. Total Environ. 490, 1012–1028. https://doi.org/10.1016/j.scitotenv.2014.05.071. Iniesta-Arandia, I., García-Llorente, M., Aguilera, P.A., Montes, C., Martín-López, B., 2014. Socio-cultural valuation of ecosystem services: uncovering the links between values, drivers of change, and human well-being. Ecol. Econ. 108, 36– 48. https://doi.org/10.1016/j.ecolecon.2014.09.028. Jia, X., Fu, B., Feng, X., Hou, G., Liu, Y., Wang, X., 2014. The tradeoff and synergy between ecosystem services in the Grain-for-Green areas in Northern Shaanxi, China. Ecol. Indic. 43, 103–113. https://doi.org/10.1016/j.ecolind.2014.02.028. Jie, X., Yu, X., Na, L., Hao, W., 2015. Spatial and temporal patterns of supply and demand balance of water supply services in the Dongjiang Lake Basin and its beneficiary areas. J. Resour. Ecol. 6, 386–396. Kroll, F., Müller, F., Haase, D., Fohrer, N., 2012. Rural–urban gradient analysis of ecosystem services supply and demand dynamics. Land Use Policy 29, 521–535. Li, Q., Pu, C., Liu, P., Wei, Z., Liu, C., 2015. A Survey on policy of transforming grazing lands to pastures and its influential factors interviewed with herdsmen in Xinjiang Uygur Autonomous Region. Bull. Soil Water Conserv. 35 (6), 281–284. Liu, Z., Zheng, Y., 2014. Analysis of herdsmen willingness to accept grassland ecosystem compensation in Xinjiang. Acta Agrect. Sin. 22 (6), 1212–1215. Martín-López, B., Iniesta-Arandia, I., Garcia-Llorente, M., Palomo, I., Casado-Arzuaga, I., Del Amo, D.G., Gomez-Baggethun, E., Oteros-Rozas, E., Palacios-Agundez, I., Willaarts, B., Gonzalez, J.A., Santos-Martin, F., Onaindia, M., Lopez-Santiago, C., Montes, C., 2012. Uncovering ecosystem service bundles through social preferences. PLoS One 7, e38970. Mensah, S., Veldtman, R., Assogbadjo, A.E., Ham, C., Kakaï, R.G., Seifert, T., 2017. Ecosystem service importance and use vary with socio-environmental factors: a study from household-surveys in local communities of South Africa. Ecosyst. Serv. 23, 1–8. Millennium Ecosystem Assessment (MA), 2005. Ecosystems and Human Wellbeing. Synthesis. Island Press, Washington, DC. Ministry of Water Resource of China, 2008. SL 190-2007 Standards for classification and gradation of soil erosion. China Water and Power Press, Beijing, China (in Chinese). Morri, E., Pruscini, F., Scolozzi, R., Santolini, R., 2014. A forest ecosystem services evaluation at the river basin scale: supply and demand between coastal areas and upstream lands (Italy). Ecol. Indic. 37, 210–219. Ouyang, Z., Zheng, H., Xiao, Y., Polasky, S., Liu, J., Xu, W., Wang, Q., Zhang, L., Xiao, Y., Rao, E., Jiang, L., Lu, F., Wang, X., Yang, G., Gong, S., Wu, B., Zeng, Y., Yang, W., Daily, G.C., 2016. Improvements in ecosystem services from investments in natural capital. Science 352, 1455–1459. Potter, C.S., Randerson, J.T., Field, C.B., Matson, P.A., Vitousek, P.M., Mooney, H.A., Klooster, S.A., 1993. Terrestrial ecosystem production: a process model based on global satellite and surface data. Glob. Biogeochem. Cycles 7, 811–841. Potter, N.J., Zhang, L., Milly, P.C.D., McMahon, T.A., Jakeman, A.J., 2005. Effects of rainfall seasonality and soil moisture capacity on mean annual water balance for Australian catchments. Water Resour. Res. 41, W06007. Quintas-Soriano, C., Castro, A.J., García-Llorente, M., Cabello, J., Castro, H., 2014. From supply to social demand: a landscape-scale analysis of the water regulation service. Landscape Ecol. 29, 1069–1082. https://doi.org/10.1007/ s10980-014-0032-0. Raudsepp-Hearne, C., Peterson, G.D., Bennett, E.M., 2010. Ecosystem service bundles for analyzing tradeoffs in diverse landscapes. Proc. Natl. Acad. Sci. U.S.A. 107, 5242–5247. https://doi.org/10.1073/pnas.0907284107. Renard, K.G., Foster, G.R., Weesies, G.A., Porter, J.P., 1991. RUSLE: revised universal soil loss equation. J. Soil Water Conserv. 46 (1), 30–33. Rodríguez, J.P., Beard, T.D., Bennett, E.M., Cumming, G.S., Cork, S.J., Agard, J., et al., 2006. Trade-offs across space, time, and ecosystem services. Ecol. Soc. 11, 28. https://doi.org/10.2307/2390206. Santos-Martin, F., Martin-Lopez, B., Garcia-Llorente, M., Aguado, M., Benayas, J., Montes, C., 2013. Unraveling the relationships between ecosystems and human wellbeing in Spain. PLoS One 8, e73249. https://doi.org/10.1371/journal. pone.0073249. Schulp, C.J.E., Lautenbach, S., Verburg, P.H., 2014. Quantifying and mapping ecosystem services: demand and supply of pollination in the European Union. Ecol. Indic. 36, 131–141. Smith, L.M., Case, J.L., Smith, H.M., Harwell, L.C., Summers, J.K., 2013. Relating ecoystem services to domains of human well-being: foundation for a U.S. index. Ecol. Indic. 28, 79–90. https://doi.org/10.1016/j.ecolind.2012.02.032. Statistics Bureau of Xinjiang Uygur Autonomous Region, 2016. Xinjiang Statistic Yearbook. China Statistic Press, Beijing, China. Stürck, J., Schulp, C.J.E., Verburg, P.H., 2015. Spatio-temporal dynamics of regulating ecosystem services in Europe – The role of past and future land use change. Appl. Geogr. 63, 121–135. Sun, G., Alstad, K., Chen, J.Q., Chen, S.P., Ford, C.R., Lin, G.H., Liu, C.F., Lu, N., McNulty, S.G., Miao, H.X., Noormets, A., Vose, J.M., Wilske, B., Zeppel, M., Zhang, Y., Zhang, Z.Q., 2011. A general predictive model for estimating monthly ecosystem evapotranspiration. Ecohydrology 4, 245–255. Tallis, H.T., Ricketts, T., Guerry, A.D., Wood, S.A., Sharp, R., Nelson, E., Ennaanay, D., Wolny, S., Olwero, N., Vigerstol, K., 2013. InVEST 3.3.3 User’s Guide. The natural capital project. https://www.naturalcapitalproject.org/invest/. Villamagna, A.M., Angermeier, P.L., Bennett, E.M., 2013. Capacity, pressure, demand, and flow: a conceptual framework for analyzing ecosystem service provision and delivery. Ecol. Complex. 15, 114–121. Wang, D.S., Li, Y.F., Zheng, H., Ouyang, Z.Y., 2014. Ecosystem services’ spatial characteristics and their relationships with residents’ well-being in Miyun Reservoir watershed. Acta Ecol. Sin. 34 (1), 70–81 (in Chinese).

H. Wei et al. / Ecosystem Services 31 (2018) 44–57 Wang, B., Tang, H., Xu, Y., 2017a. Integrating ecosystem services and human wellbeing into management practices: insights from a mountain-basin area, China. Ecosyst. Serv. 27, 58–69. https://doi.org/10.1016/j.ecoser.2017.07.018. Wang, X., Dong, X., Liu, H., Wei, H., Fan, W., Lu, N., Xu, Z., Ren, J., Xing, K., 2017b. Linking land use change, ecosystem services and human well-being: a case study of the Manas River Basin of Xinjiang, China. Ecosyst. Serv. 27, 113–123. https://doi.org/10.1016/j.ecoser.2017.08.013. Wei, H.J., Zhang, Y.F., Dong, X.B., Lu, N.C., Wang, X.C., 2016. Estimating the spatiotemporal characteristic of vegetation water use efficiency over Weihe River Basin. J. Nat. Resour. 31 (8), 1275–1288 (in Chinese). Wei, H., Fan, W., Wang, X., Lu, N., Dong, X., Zhao, Y., Ya, X., Zhao, Y., 2017. Integrating supply and social demand in ecosystem services assessment: A review. Ecosyst. Serv. 25, 15–27. https://doi.org/10.1016/j.ecoser.2017.03.017. Willemen, L., Hein, L., Verburg, P.H., 2010. Evaluating the impact of regional development policies on future landscape services. Ecol. Econ. 69, 2244–2254. https://doi.org/10.1016/j.ecolecon.2010.06.012. Xu, Y., Tang, H., Wang, B., Chen, J., 2016. Effects of land-use intensity on ecosystem services and human well-being: a case study in Huailai County, China. Environ. Earth Sci. 75, 1–11. https://doi.org/10.1007/s12665-015-5103-2.

57

Yang, L., Zhen, L., Wei, Y., Jiang, L., Cao, X., Long, X., 2010. Impacts of ecosystem services change on human well-being in the Loess Plateau. Resour. Sci. 32 (5), 849–855 (in Chinese). Yuan, G.Y., Kuchler, J., Li, J.S., 1995. Conservation and Rational Use of the Agriecosystem Resources in the Manas River Valley. Publishing House of Science, Technique and Sanitation of XinjiangUrumqi, China. Zhang, X., 2001. Ecological restoration and sustainable agricultural paradigm of Mountain-Oasis-Ecotone-Desert system in the north of the Tianshan Mountains. J. Integr. Plant Biol. 43, 1294–1299. Zhang, Q., Xu, H., Fan, Z., 2012. Potential risk assessment of ecological problems of Manas River Basin. J. Nat. Disaster 21 (3), 70–78. Zhao, C., Sander, H.A., 2015. Quantifying and mapping the supply of and demand for carbon storage and sequestration service from urban trees. PLoS One 10, e0136392. Zheng, H., Li, Y., Robinson, B.E., Liu, G., Ma, D., Wang, F., Lu, F., Ouyang, Z., Daily, G.C., 2016. Using ecosystem service trade-offs to inform water conservation policies and management practices. Front. Ecol. Environ. 14, 527–532. Zhu, W., Pan, Y., He, H., Yu, D., Hu, H., 2006. Simulation of maximum light use efficiency for some typical vegetation types in China. Chin. Sci. Bull. 51 (4), 457–463.