Land Use Policy 82 (2019) 120–129

Contents lists available at ScienceDirect

Land Use Policy

journal homepage: www.elsevier.com/locate/landusepol

Herder stocking rate and household income under the Grassland Ecological Protection Award Policy in northern T ⁎ Yantin Yina,1, Yulu Houb,1, Colin Langfordc, Haihua Baia, Xianyang Houa, a Institute of Grassland Research, Chinese Academy of Agricultural Sciences, Huhhot, , 010010, China b Agriculture Information Institute, Chinese Academy of Agricultural Sciences, Beijing, 100081, China c Graham Centre for Innovation, Charles Sturt University, Wagga Wagga, NSW, 2800, Australia

ARTICLE INFO ABSTRACT

Keywords: The Grassland Ecological Protection Award Policy (GEPAP) is the largest payment for ecosystem services (PES) Grassland degradation program targeting grasslands in China. It subsidizes households to reduce livestock numbers or ban the grazing Overstocking of livestock to restrict the large-scale degradation of grasslands. While the GEPAP has drawn attention to these Payment for ecosystem services issues, questions regarding the performance of the GEPAP have still not been clearly answered. This research Conservation program used a balanced dataset of 726 surveyed households from 5 regions of Inner Mongolia to assess the impacts of the Livelihoods policy on stocking rate and household income. Results indicated that contrary to the aim of the GEPAP, the China overall stocking rate marginally significantly increased. Net household income extremely significantly decreased even though total income increased. Income from animals still formed the major proportion of household in- come, and off-farm income only played a complementary role in household income even though there was an increase in the amount and its proportion in total household income. Regression results indicated that the households with a lower subsidy level tended to have higher stocking rates and incomes. Stocking rate was unrelated to the policy or market price for livestock, while the household income was positively affected by livestock price. Results from this research have implications for the design, implementation and enforcement of conservation programs of grasslands in China and other developing countries.

1. Introduction Protection Award Policy. The Grassland Ecological Protection Award Policy (GEPAP) was first Grasslands have experienced large-scale degradation to different implemented in 2011 to protect grasslands, guarantee the supply of extents in China (Akiyama and Kawamura, 2007; Harris, 2010; Hua and livestock products (e.g., local beef and mutton) and improve household Squires, 2015; Li et al., 2012, 2007; Wen et al., 2013), which adversely income for herders (called liang bao yi cu jin in Chinese language). The affects ecological environment services (Tang et al., 2015; Wen et al., targeted areas cover Inner Mongolia, Xinjiang, Tibet and au- 2013) and livelihoods in pastoral regions (Cao et al., 2010; Reynolds tonomous regions, and Qinghai, Sichuan, and Yunnan provinces et al., 2005; Waldron et al., 2010). Grassland degradation has therefore (that is, 8 autonomous regions/provinces) and the Xinjiang Production become the focus of Chinese rangeland management. Overstocking is and Construction Corps. The GEPAP primarily introduced a grazing ban considered as the principal cause of grasslands degradation by policy policy and a targeted grass-animal balance policy. GEPAP policies are makers and many researchers (Kemp et al., 2011a, 2011b; Li et al., implemented in grassland regions where there are different degrees of 2012; Sneath, 2000; Squires, 2014; Wang et al., 2008). Reducing grassland degradation. stocking rate is proposed as an effective tool to rehabilitate degraded Where grasslands have been severely degraded GEPAP requires that grasslands. For these purposes, the Chinese government has im- grasslands grazing by livestock is totally forbidden or only a very small plemented several national conservation programs, including the Eco- number of livestock (e.g., 50 adult sheep in the desert region) are logical Migration Program, Returning Grazing Land to Grassland pro- permitted to be grazed in the grassland. After 5 years of subsidy, ject, the Beijing–Tianjin Sandstorm Source Controlling project, the grazing will be continually forbidden or the grassland will be managed Conversion of Grain for Green Program and the Grassland Ecological under the grass-animal balance policy depending on the condition of

⁎ Corresponding author. E-mail address: [email protected] (X. Hou). 1 Two authors are the first-coauthors of this manuscript. https://doi.org/10.1016/j.landusepol.2018.11.037 Received 5 April 2018; Received in revised form 20 November 2018; Accepted 20 November 2018 Available online 10 December 2018 0264-8377/ © 2018 Published by Elsevier Ltd. Y. Yin et al. Land Use Policy 82 (2019) 120–129 the grassland. some grasslands and declining in some other regions. Some researchers Where grasslands are determined to fit for grazing and the agree- have suggested the methods on how to reduce livestock, and at the ment stipulates that the household may only breed a specified number same time improve household income and some have suggested that of livestock in comparison with the owned grassland based on the more investment should be given to vocational and adult education for carrying capacity of grassland (the standard stocking rate given by the herders with the Internet becoming a new primary source for learning policy) and if households comply with the policy, they can receive the information (Ma, 2017). Other studies have also shown that the live- subsidy. This policy is aimed at guiding and encouraging herder stock husbandry production system (e.g., the warm sheds, new varieties households to seasonally and rotationally graze their grasslands using a of livestock and livestock fattening practices) should be improved (Mai balance between grass supply and animal needs with the long term aim et al., 2016; Wang et al., 2016). Other researchers have suggested that of establishing a long-term rational use of grassland. the GEPAP should be improved, by using differential standards for According to the GEPAP, households can receive 90 yuan/ha/year different households (Hu, 2016b; Kong et al., 2016; Pan, 2015; Wei and (US$ 14/ha/year) if they comply with the grazing ban policy, and 22.5 Qi, 2017a), and identifying all the stakeholders and determining the yuan/ha/year (US$ 3.5/ha/year) if they comply with the grass-animal roles they will play during in the implementing GEPAP (Pan, 2015). balance policy. Each household can get an additional 500-yuan-per- However, the common limitations of previous studies are the small year (US$ 78/year) comprehensive subsidy for means of production. sample size, and/or the narrow geographic coverage, and/or few time The first round of GEPAP expired in 2015 and the second phase was points. For instance, the study by Wang et al. (2016) was based on a implemented in 2016 and will run until 2020. In the second round, the dataset of only 38 households from one county () in Inner geographic scope has been increased to an additional 6 provinces/re- Mongolia and prior to the expiration of the first round program while gions, including Hebei, Shanxi, Liaoning, Jilin and Heilongjiang pro- Chen and Xiao (2013) surveyed 395 households but only in 2012. Other vinces (5 provinces) and Heilongjiang Land Reclamation Bureau. studies were conducted using the secondary evidence (Chai, 2016; Yin, Subsides were also increased to 112.5 yuan/ha/year (US$ 17.6/ha/ 2017) or only government statistical data (Yang et al., 2016) to analyze year) and 37.5 yuan/ha/year (US$ 5.9/ha/year) for the grazing ban the impacts of the GEPAP. Consequently the problems in relation to the and the grass-animal balance policies, respectively. performance of the GEPAP were not clearly understood before the be- Based on the standards given by central government, the specific ginning of Phase two. Questions such as the following were not an- subsidies per ha per year to households will be adjusted in different swered: were the stocking rates reduced, did household income in- regions (but the final subsidies for households must be higher than or crease, did the households diversify their income sources, did the equal to the government standards). These adjustments will be depen- herders comply with the GEPAP policy, what were the perceptions of dent on the productivity of different types of grasslands. For example, herders’ on reducing livestock numbers and increasing incomes. The in Inner Mongolia the grassland productivity decreases from east to answers to these questions are needed because they are related to the west, hence subsidies are the highest in the east meadow steppe region, effectiveness of the GEPAP but also the development of further efforts approximately 206 yuan/ha/year (US$ 32/ha/year) and 69 yuan/ha/ for grassland protection in China. Also to our knowledge, there have year (US$ 10.8/ha/year) in response to the grazing ban and the grass- been no papers about the impact of GEPAP published in any interna- animal balance polices in the second-round GEPAP, respectively. tional journal, even though GEPAP has been the focus of international Subsidies in the west sandy steppe region are equal to the government attention. standards. Because of its broad geographic coverage (13 provinces/ This paper shows how through the performance of a comprehensive autonomous regions, and the Xinjiang Production and Construction analysis were the impacts of the GEPAP on stocking rates and house- Corps, and Heilongjiang Land Reclamation Bureau) and subsidy con- hold incomes in Inner Mongolia. A unique balanced panel dataset from tents (subsidies for grazing ban, grass-livestock balance, livestock hus- two rounds of household surveys was used for the analysis. The dataset bandry production and improved varieties of livestock and forage), and contains more than 700 herder households from 15 counties in Inner the huge public investment (over 13 billion yuan (US$ 2 billion) per Mongolia-the largest and most representative grassland region in China. year, total 77.4 billion yuan (US$ 12 billion) in the first round, and The dataset covers two important time points: 2010 (before the first more than 110 billion yuan (US$ 17 billion) so far), the GEPAP is by far round GEPAP) and 2015 (the year when the first round GEPAP fin- the largest payment for ecosystem services (PES) program targeting ished). The grassland type and institutional changes were also taken grasslands in China. into consideration. The 15 counties cover all 5 types of grasslands (3 The ultimate aim of the GEPAP is to reduce grassland degradation counties per type) in Inner Mongolia, and also cover the grazing ban with the additional aim of reducing stocking rates, while at the same areas and the grass-animal balance areas. The program sought to time increasing herder household incomes. Small-scale households quantify changes in the stocking rate and household incomes since from 268 pastoral and agro-pastoral counties nationwide have been GEPAP was implemented and herder opinions on the GEPAP. This will involved in this national program. improve our understanding of the impacts of the GEPAP and the ana- The importance of the GEPAP in reducing the degradation of lysis will also make a contribution to the appropriate design and im- grasslands, the impacts on herders stocking rates and household income plementation of conservation programs in China. This work also will have been studied by many Chinese researchers (Huarui, 2013; Jin, enable China’s lessons and experiences of degraded ecosystems re- 2016; Li et al., 2017; Shi, 2017; Wang et al., 2016; Yang et al., 2016; storation to a broader international significance. Yin, 2017; Zhang, 2015). These studies have shown that stocking rates This reminder paper is organized as follows: Sections 2 described increased in most regions contrary to the policy aim, and the contra- the study areas and the household surveys. The results were presented dictory results on household incomes, which increased in some regions in Section 3. The discussion was provided in Section 4. In the final but decreased in some regions. section, the conclusions were given. Research examined the willingness of herder households to parti- cipate in the program (Jiang et al., 2014; Wei and Qi, 2017b) and their 2. Study area satisfaction levels (Baosaodubilige, 2015; Lei, 2016) which showed that households who had more labor, larger and more degraded grasslands 2.1. Study sites were more inclined to participate in the GEPAP. However, the majority of households were not satisfi ed with the policy due to perceived low Inner Mongolia was selected for the study area because it is the subsidies. The ecological performance of the GEPAP have also been dominant area of Chinese northern grasslands. Inner Mongolia grass- assessed by some studies (Huang, 2015; Li and Zeng, 2016) which have lands are ecologically divided into five types: meadow steppe, typical shown grassland condition increasing in some regions, no change in steppe, desert steppe, sandy steppe and desert regions. The five regions

121 Y. Yin et al. Land Use Policy 82 (2019) 120–129

Fig. 1. The distribution of surveyed counties in Inner Mongolia, China.

Table 1 management in northern grassland areas supported by National Natural List of grassland types and counties surveyed in 2010 & 2015. Science Foundation of China (NSFC). Three counties were selected in – Rangeland type Counties each grassland region. In each county, 5 15 households were randomly selected from 5 Gachas (village) which belong to 2–3 Sumus (township) Meadow steppe Chenbarhu Banner, Xin Barag Left Banner, Ewenki in proportion to its total household numbers with at least 60 households Autonomous Banner interviewed. At least 180 households in each region and a total of 902 Typical steppe East Ujimqin Banner, Xilinhot City, Xianghuang Banner households were interviewed in 2010. In 2015, the survey was repeated Desert steppe Sunite Left Banner, Sunite Right Banner, Siziwang Banner Sandy steppe Etuoke Banner, Hangjin Banner, Wushen Banner after the first round of GEPAP had concluded. In this way we were able Desert Alxa Left Banner, Alxa Right Banner, Wulate Rear Banner to collect a first-hand panel data covering before and after the im- plementation of the GEPAP. In the second survey, the persons collecting the data were asked to find and interview the households who had been are distributed from east to west with gradually decreasing rainfall and interviewed in 2010 prior to collecting data from new households. A grassland productivity. Three counties in each region were chosen as total of 905 households were surveyed in 2015. In each survey, the sample areas (Fig. 1)(Table 1). Grasslands have been allocated to each household head or a family member who was familiar with the household under the Household Responsibility System (HRS) since the household’s characteristics was interviewed. Households that were in- 1980s. The main land use activity was smallholder animal husbandry. terviewed in 2010 but not in 2015 were deleted from the data in this Under the GEPAP, the grass-animal balance policy is mainly im- study. Finally, a balanced panel data set of 755 households was ob- fi plemented in rst four regions and the grazing ban policy is im- tained. 29 households were deleted due to the issues of data quality, plemented in the desert region. resulting in a balanced panel data of 726 households for this study. The dataset contains the information: (1) the socio-demographic backgrounds of households (including the characteristics of households 2.2. Data and the household head); (2) grassland information: contract area, grazing area, hay-cutting area and the details of grassland transfer (the The data used in this study was derived from two rounds of face-to- area and its use if rented from other households); (3) the livestock face household surveys based on a stratified random sampling strategy. conditions, including type and numbers, newborn animals and the sale A pilot survey was conducted to improve the questionnaire before each details of livestock; (4) household income and expenditure details: total round of formal survey began. The first survey was conducted in 2010 household income consists of the livestock income from the sale of live as a part of a previous project assessing climatic change and adaptive

122 Y. Yin et al. Land Use Policy 82 (2019) 120–129 animals and the livestock product, off-farm income mainly generated in Inner Mongolia (p = 0.055), which was contrary to the objective of from the labor payment during the harvest, wage employment and self- GEPAP. The stocking rates increased with different significant degrees employment, selling the planted crops (e.g., corn and Medicago Sativa in the meadow steppe (p = 0.007), desert steppe (p = 0.033) and sandy Linn in the sandy steppe region), trees and Chinese herbs (e.g., Cistanche steppe regions (p = 0.020). The meadow steppe region experienced the deserticola Ma and Cynomorium songaricum Rupr. in the desert region), most significant and largest increase in stocking rates of nearly two-fold subsidies from the GEPAP and other income which refers to other (from 2.83 sheep units/ha in 2010 to 4.54 sheep units/ha in 2015). welfare transfer from government, cash support and gifts from relatives Only in the desert region where the grazing ban policy was im- and friends. Household expenditure includes livestock production plemented, the stocking rate was reduced by approximately 63% (from (mainly including the costs of purchasing forage and fodder, the pro- 1.04 sheep units/ha in 2010 to 0.39 sheep unit/ha in 2015 at the 0.001 duction machines and vehicles), fuel and living expenditures (mainly extremely significant level. Such performance was obviously far from including education, medicine, and the costs of social relations, buying the aim of GEPAP. the vehicles for living convenience, and building/buying the houses/ Mongolian gers). Net household income equals the total household cost 3.3. Household income subtracted from the total household income. Net livestock income was obtained by total livestock cost subtracted from total livestock income. Fig. 3 shows the changes in household income during the program All the monetary variables were converted to the 2010 constant yuan period. The total household income experienced an extremely sig- using local consumer price index. (5) Herder attitude and opinions of nificant increase at the 0.001 level in all regions except the typical the GEPAP. steppe region, where the total income showed little increase during the first-round GEPAP period, and was not statistically significant (p 2.3. Data analysis = 0.305). Turning to the net livestock income, there was, overall, more than a 50 percent decrease in comparison with 2010 (54,839 to 26,185 After a day’s survey, all the finished questionnaires of that day were yuan (US$ 8461 to 4040) at 2010 constant price), which is extremely checked again to ensure the integrity and accuracy of data. After the significant at the 1% level. It even became negative in 2015 in the whole survey, data were transcribed from the returned questionnaires sandy steppe region (-30652 yuan, that is approximately -US$ 4729). to create an electronic Excel dataset. Whenever needed, the Excel data There was also an exception in the meadow steppe region, where the were also compared to the original questionnaires to furthermore check net livestock income increased but not significantly (p = 0.249). the accuracy of the data in case of data entry errors. Overall, the off-farm income didn’t show a significant increase (p Data was analyzed in SPSS version 21. First, we conducted a de- = 0.181) and even decreased in the sandy steppe region but not at the scriptive analysis by frequencies and percentages to report household significant level (p = 0.752), except in the desert region, where the off- demographics and livestock assets. Second, paired t-tests were applied farm income rose approximately twofold (p = 0.024). In terms of the to compare the differences between herd sizes, stocking rates, cate- net household income, Fig. 3 (d) indicated that it showed an overall gories of incomes and the shares of livestock income in total household extremely significantly decrease at the 1% level. The average net income before and after the first-round GEPAP, respectively. household income was even negative in desert steppe and sandy steppe We also performed a multiple regression with panel data of 726 regions. The households with negative net incomes accounted for up to households in Stata version 11 to explore the factors that influenced the 60% after the first-round GEPAP in comparison to 37% before it. herder stocking rate and household income. Based on the literature, summary statistic results and what we found through the surveys, the 3.4. Proportion of total household income from livestock independent variables that are likely to influence herder households’ stocking rates and incomes were defined in Table S1, including the To determine the changes in household on grassland before and proportion of subsidy, grassland type, characteristics of the household after the first-round GEPAP, we calculated the proportion of total head (gender, age, education level and whether they work in Gacha), household income from livestock (Fig. 4), which equaled total house- distance to township, household labor, grassland area, herd size and the hold income divided by total livestock income. Between 2010 and price. 2015, the proportion of total income from livestock decreased at the Then a first run of the multiregression with household panel data 0.01 extremely significant level in each region, which indicated that was performed with all variables. A second model with significant herder households relied much less on livestock husbandry as the main variables obtained from the first run model was then developed. Then income source. It is also worthy to note that the proportion of total the interaction terms were created after mean centering. The final income by livestock income was still very high after the first-round multiple regression results with significant variables and interaction GEPAP. The mean proportion of total income from livestock in 2015 items were given. was 0.71, and the highest was 76% in the meadow steppe region, and even the lowest was also up to 62% in the desert region. This indicated 3. Results that livestock income was still a dominant part of total household in- come. 3.1. Background information of surveyed households 3.5. Multivariate regression results for stocking rate and household income The background information of interviewed households was sum- marized in Table 2. The medians of age, education level, family size, Using the balanced panel data, four multivariate liner regressions household labor and herd size were calculated. Overall, herd size didn’t were conducted to estimate the factors which influenced the stocking significantly change (p = 0. 556) (whether to increase or decrease). rate and household income. In the four models: stocking rate, total However, it significantly increased in the desert steppe region income, livestock income and off-farm income were dependent vari- (p = 0.034) in 2015 compared with 2010. In the desert region, the herd ables, respectively, and the variables were used as independent ones size even showed an extremely significant decrease (p = 0.000). (see Table S1). The regression results obtained from the final models with significant variables and interactions items were summarized in 3.2. Stocking rates Table S2-S5. Overall, the proportion of subsidy was significantly negatively Fig. 2 shows that the overall average stocking rate marginally sig- correlated with livestock income (p = 0.000) and positively related to nificantly increased from 2010 to 2015 (from 1.35 to 1.56 sheep units) off-farm income (p = 0.000), which means that the households with a

123 Y. Yin et al. Land Use Policy 82 (2019) 120–129

Table 2 Summary statistics of information of households.

Meadow steppe Typical steppe Desert steppe Sandy steppe Desert Inner Mongolia

Herder characteristics Percent male heads of households in 2010 (%) 94 98 95 95 86 94 Percent male heads of households in 2015 (%) 90 97 91 92 89 92 Age in 2010 45 41 44 47 46 45 Age in 2015 48 47 48 52 50 49 Education level in 2010 2 2 2 2 2 2 Education level in 2015 2 2 1 2 2 2 To work in Gacha in 2010 (%) 8 17 5 1 6 8 To work in Gacha in 2015 (%) 4 15 6 13 11 10 Household characteristics Family size in 2010 4 4 4 4 4 4 Family size in 2015 4 4 4 4 4 4 Household labor in 2010 2 2 4 2 2 2 Household labor in 2015 2 2 2 2 2 2 Household assets Herd size in 2010 (sheep units) 514 373 319 189 345 317 Herd size in 2015 (sheep units) 493 435 379 184 296 320

(Gender: male = 1, female = 0; Education level: 0 = No formal education. 1 = primary school; 2 = junior high school; 3 = senior high school, 4 = College).

Fig. 2. The stocking rates changes between 2010 and 2015. Fig. 4. Livestock income dependence dynamics.

Fig. 3. Household income changes.

124 Y. Yin et al. Land Use Policy 82 (2019) 120–129 lower proportion of subsidy over total income had higher livestock kept increasing, more than 50% of interviewed herders said that the incomes while the households who had higher proportion of subsidy subsidies from the GEPAP should be increased as a whole (Table 3). were more likely to have higher off-farm income. It implied that for Regarding the market, 16% (113 herders) said that the prices of live- these herders that the subsidy was lower compared to their livestock stock products should be increased, and 1% (6 herders) suggested that income, they tended to not comply with the GEPAP. It implied that the government should set the lowest prices for livestock products, subsidy had no expected impact on household income by the policy. which would ensure the stability of incomes of households and then the This observation was replenished by the result above regarding the security of their livelihoods. There were also 10 herders who stated that proportion of livestock income in total income. The livestock income more sales channels for livestock products should be developed as most still accounted for a dominant part of total household income, which herders were located remotely from the town (the market) and had influenced herder decision regarding whether or not to continue to little access to these markets but mainly depended on the itinerant breed animals. traders, who often undercut the prices of livestock products and ad- Grassland type had a significantly negative impact on stocking rate. versely affected the incomes of households. There were also 6% (40 The grassland productivity reduced gradually with the decreasing herders) and 3% (24 herders) of respondents respectively said that the rainfall from east to west, which correspondingly resulted in stocking varieties of livestock should be improved and more access to livestock rates becoming lower gradually. It was also significantly negatively husbandry technologies (e.g., intensive livestock fattening) be pro- related with the livestock and total income, while its effect on off-farm vided. Some herders (4%, 27 ones) also said that better infrastructure income was significantly positive and large, which indicates that the (e.g., the warm sheds and wells) were needed to provide a basis for households in the west regions had lower livestock income but higher optimized livestock and grassland management. Nearly 1% of herders off-farm income. In combination, the net effect on total income was hoped that they could have more access to off-farm employment, which negative and small. would increase household income through transferring labor to non- In term of the household characteristics, the gender had a positive farm industries. relationship with livestock and total income. This means that the households with male heads were more likely to have higher livestock 4. Discussion and total income. Age had a significantly positive relationship with off- farm income, with older herders having higher off-farm incomes. Chinese government and researchers are all very interested in the In regard to household assets/resources, grassland area and herd impacts of the GEPAP, the largest PES program for grasslands in China. size were all significantly positively correlated with livestock and total We used a household-level balanced panel data consisting of 726 income, which shows that the households with larger grasslands and households that covered all 5 types of grasslands areas in Inner number of livestock tended to have higher livestock and total incomes. Mongolia, to assess the impacts of the first round GEPAP on stocking Those two variables were negatively and positively related to the rate and household income and quantify herder opinions on the GEPAP. stocking rates, respectively. The larger grasslands and less livestock the Through the detailed analysis it has been shown that overall stocking households had, the lower stocking rates there were. The distance to rates increased and the net household income decreased significantly, township had a significantly negative relationship with off-farm in- contrary to the aim of the GEPAP. Some suggestions were provided to come. The households who lived closer to township usually had higher improve the assessment and implementation of the conservation po- off-farm incomes. licies in China and even other developing countries. Price was significantly and positively correlated with livestock and Our results suggest that at least in the short term, the GEPAP failed total income. The higher the price per sheep unit was, the higher li- to reduce stocking rates in Inner Mongolia grasslands as a whole (The vestock and total incomes the households had. It can infer that the overall stocking rate increased from 1.35 sheep units/ha to 1.56 sheep market played an important role in the household income. But it had no units/ha). The stocking rates increased in 4 of 5 types of regions, which significant impact on the stocking rate in all regions. has also been observed in other studies (Chen et al., 2013; Hao, 2016; Hu, 2016b; Jin, 2016; Li et al., 2017; Ye et al., 2015). The stocking rates 3.6. Herder perception of the GEPAP after the first round GEPAP were higher than the policy standards in each region. For instance, in the meadow steppe region the actual We also interviewed herders’ opinions of the GEPAP. Approximately average stocking rate in 2010 was 4.54 sheep units/ha, far higher than 96% of interviewed households held a positive attitude towards the the local 0.83 sheep unit/ha standard of policy. GEPAP. However, notably, their positive attitude didn’t mean that they However, this does not mean the GEPAP should be discontinued due complied with the GEPAP. In contrast, it was just because that they got to a short-term failure. Survey data indicated that the desert region subsidies from the government for doing nothing. The majority of produced an extremely significant 14 percent reduction in livestock herders did not reduce their livestock numbers or ban the grazing of number (435–373 sheep units) from 2010 to 2015 at the 0.001 level animals in their grasslands at all according to the policy. One herder (see supplementary Fig. 1). According to our survey, the grass-animal even said that they thanked the government for the subsidy, and be- balance policy has been implemented in the other 4 types of grassland cause of it they could breed more livestock. There were also less than regions (the meadow steppe, typical steppe, desert steppe and sandy 4% of herders who said that the GEPAP was unpopular and several ones steppe regions) where more than half of households increased the who didn’t even care about the GEPAP at all. These herders considered stocking rates, and there was almost no monitoring activity about li- the present subsidies to be too low to offset the loss brought by reducing vestock number or punishment from local governments due to in- livestock, and believed that they still needed to depend on livestock for creasing livestock numbers. The herders themselves decided how many their livelihoods. livestock to breed. In contrast, in the west desert region, the grazing ban Herders also expressed the concern that the subsidy may be elimi- policy was more strictly implemented, and the grazing was totally nated if the GEPAP policy expired. If yes, their livelihoods were bound forbidden or only a small number of livestock (e.g., 50 adult sheep in to be negatively affected if they reduced livestock numbers or sold all of Alxa Right Banner) were allowed to graze the grasslands, no matter their livestock according to the policy, which further complemented our how large the grasslands were. And there would be punishment if results about the role of subsidy in household income. This helps to households didn’t comply with the policy (e.g., in Alxa Left Banner, the explain why a majority of households were not willing to reduce live- excess animals would be directly taken away by staff of local govern- stock numbers. mental monitoring institutions for selling or the herder must pay 3000 When asked what measures should be taken to ensure that herders yuan (about US$ 440) fi ne). Hence, it suggests that the local govern- could reduce livestock while at the same time the household income ance needs to be strengthened in pastoral regions, especially in terms of

125 Y. Yin et al. Land Use Policy 82 (2019) 120–129

Table 3 Herders’ opinions about two aims of reducing livestock number and increasing household income.

Meadow steppe Typical steppe Desert steppe Sandy steppe Desert Inner Mongolia

Household number N = 112 N = 171 N = 140 N = 144 N = 159 N = 726 Households who increased stocking rates 69 (62%) 95(56%) 96 (69%) 83 (58%) 29(18%) 372 (51%) Higher subsidies 60(54%) 123(72%) 73(52%) 58(43%) 69(43%) 383(53%) Higher livestock product prices 27(24%) 19(11%) 42(30%)) 14(10%) 11(7%) 113(16%) More sales channels 2(2%) 1(1%) 0(0) 0(0) 7(4%) 10(1%) Improved livestock 3(3%) 3(2%) 3(2%) 5(3%) 26(16%) 40(6%) More access to technologies 4(4%) 1(1%) 3(2%) 4(3%) 12(8%) 24(3%) Improved infrastructure 6(5%) 9(5%) 4(3%) 3(2%) 5(3%) 27(4%) More off-farm chances 1(1%) 3(2%) 0(0) 0(0) 2(1%) 6(1%)

(Number outside the parentheses is the number of herders, and the percent is in parentheses). local participation in monitoring and management (Farhad et al., 2015; Chinese herbs (Cistanche deserticola Ma and Cynomorium songaricum Li et al., 2017; Paavola and Hubacek, 2013; Yin et al., 2014). Rupr). Successful monitoring and management of the GEPAP at the local However, the off-farm income still only played a complementary area will need to be carefully considered especially regarding the costs role in household income (Chai, 2016; Huarui, 2013; Jinhua, 2015), of improving oversight of the program without placing a financial and animal husbandry was still the dominant source of household in- burden or even budget deficits on local government as the im- come in all regions, which was consistent with many studies (Chai, plementation and monitoring costs would increase with the aims of PES 2016; Chen et al., 2013; Huarui, 2013; Lei, 2016). It implied that there programs getting larger (Xu et al., 2006), and the contracts/require- is a substantial amount of work to be done to educate herders on the ments being tailored to individual users (Jack et al., 2008). In Chinese value of PES schemes and how they can be implemented on farm and pastoral regions, the GEPAP has both ecological and economic aims, not just be a way for the herders to get extra money without changing and has been adapted for individual households who cover large areas their farming practices. There is a still a long way to go to broaden as the herders are working in their contracted grasslands which can be herder households’ income channels and invest more in the training on quite large (there are many households whose grazing lands are as large off-farm employment in order to stabilize the transferred agriculture as 1000 ha in Inner Mongolia), which would definitely increase the labor force’s employment in case of herders going back to grazing. It costs of improving oversight. will directly contribute to herders successfully transferring to non- It needs to be recognized that herders are direct decision-makers of agricultural industries and reducing their livelihoods’ dependence on livestock practices and the on-ground stewards of grassland resources grasslands and animal husbandry. who play a very important role in the efficient implementation of the The subsidy from the GEPAP has already become the second most conservation programs and the improvement of grassland ecosystems important source of household income (Jin, 2016; Wang et al., 2016). (Cencetti, 2011; Cong and Guo, 2010; Squires and Yang, 2009). The data showed that the overall proportion of subsidy compared to Therefore it is also imperative that the herders actively participate in total income was 15% with the highest being 21% in the desert steppe the GEPAP and willingly reduce their stocking rates. The present policy region (see supplementary Fig. 2). Unfortunately, the subsidy con- has been implemented top-down in China and the herders have had no tributed little to reducing livestock numbers as expected by the GEPAP, any saying but to accept it, which was another factor that negatively despite improving herder households’ livelihoods (Lei, 2016; Li et al., influenced herders’ attitudes towards the policy (Dalintai, 2011; Hou 2017; Yin, 2017). The subsidy issue has caused a new concern about the et al., 2014; Yin et al., 2013; Zhou and Liu, 2009). This is different from GEPAP. Many PES programs in China, including the GEPAP, were many countries, for example, Australia, where government and farmers usually first implemented for 5 years, and then the programs expired or supported and worked together on several natural resource manage- there would also be an adjustment if further implemented (Wenming ment projects (Hajkowicz, 2009; Page and Bellotti, 2015) and many et al., 2013), which increased uncertainty around the PES programs. No farmers actively undertaken some form of practice aimed towards im- one was sure whether the programs would continue or not after the first proving the natural resource management (Page and Bellotti, 2015), phase, or the subsidy would be reduced even if there was the second which contributed greatly to the success of the PES schemes. phase. Such uncertainty of PES programs themselves was considered as The overall net livestock and household income decreased, which a most frequently cited factor hindering farmers from active partici- was also contrary to the aim of the GEPAP. This was similar to the pation in the conservation schemes (Page and Bellotti, 2015). That is findings of some researches (Jin, 2016; Wang et al., 2016; Wenming also the herders’ main concerns about the GEPAP, in conjunction with et al., 2013), which demonstrated that the GEPAP had negative impacts their habitual overstocking caused by their desirable stocking rates on the livestock and household income in some regions. However, the (Wei and Hou, 2015), which negatively affected their willingness to overall off-farm income in 2015 (29,310 yuan, that is, approximately comply with the policy to reduce livestock numbers. Hence, it needs a US$4288) showed an increase of 32% over that of 2010 (22,246 yuan, longer time frame, more than 5 years, to reduce herder dependence on that is, about US$ 3255) (Fig. 3(c)) and its proportion in total income livestock husbandry. Many researchers have advocated that long-term also increased by 2% (from 6% in 2010 to 8% in 2015) (see supple- compensation mechanism for the grassland ecosystems should be es- mentary Fig. 2). These findings were supported by some researchers tablished (Huarui, 2013; Jinhua, 2015; Wang et al., 2016; Wei and Hou, (Chai, 2016 ; Huarui, 2013), and more importantly, also in line with the 2015; Wenming et al., 2013; Yi et al., 2015). aim of the GEPAP, which was designed to help herder households have It should be noted that the second-round GEPAP has already be multiple channels of increasing incomes rather than depend solely on implemented in 2016 and subsidies for the grazing ban and grass-an- livestock husbandry. Our survey data also showed that 28% of house- imal balance policies were also increased to 112.5 yuan/ha/year (US$ holds (205 households) had off-farm incomes in 2015 compared with 17.6/ha/year) and 37.5 yuan/ha/year (US$ 5.9/ha/year), respectively. 24% (177 households) in 2010 (χ2 = 2.785, p = 0.095). In the west It is hoped that through the long-term compensation the degradation of desert region with the grazing ban policy strictly implemented, the grasslands canbe restrained, and herder households develop a new way redundant labor force due to the reduction in livestock number were of living for ecologically and economically sustainable development in already induced to seek off-farm employment such as planting trees and Chinese pastoral regions.

126 Y. Yin et al. Land Use Policy 82 (2019) 120–129

Market (price per sheep unit) had no statistically significant impact ignored. It could be due to the fact that different regions have diverse on the stocking rates, which was consistent with the finding of Dean situations and local specificities (Yin et al., 2014). For example, the and Macdonald (1994) in South Africa. It implied that it was im- grasslands conditions become worse from east to west along with the possible, at least for a short time, to reduce the stocking rates of decreasing productivity in the study region, which would diversely households through increasing the market prices. Some empirical stu- affect the performance of the PES programs. In addition, local economy dies showed that households could get the same or even more incomes (especially the prices of livestock products) was considered as more if they reduced the livestock numbers (lower stocking rates), and at the important to the local performance of the PES programs in some aspects same time bred animals by modern management practices (warm sheds, (Lin and Yao, 2014), compared to the economic conditions at national stall-feeding and fattening) and sold them with higher market prices or provincial scales. Therefore it is not surprising to see some divergent (Kemp et al., 2011a, 2011b; Kemp et al., 2011a, 2011b; Michalk et al., results in previous studies (Hu, 2016a; Shi, 2017; Yang et al., 2016). 2011; Zheng et al., 2011). It suggests that the modern management Some factors also worked to different degrees or in opposite directions practices should be improved in Chinese pastoral regions, where the in different regions. Hence it calls for caution when analyzing the ag- majority of households still breed animals at present by the traditional gregated data and require that the same questions in different regions/ production mode. counties may need to be adapted for a local area in separate case studies Unfortunately, they ignored a fact that households have live under (Yin et al., 2010). the unexpected natural environment. Once they encounter a natural disaster such as a heavy snow, there would usually be a great loss of 5. Conclusions livestock. Under such condition, it may take households several years to or even they would never recover from the disaster if they had less In this paper, the impacts of the GEPAP on herder stocking rate and livestock. Their livelihoods would be adversely affected. After all, an- household income were assessed using a balanced data of 726 surveyed imal husbandry was still their main source of income. In contrast, if households in Inner Mongolia. Results indicated that the GEPAP failed they had more livestock, their animal production and livings would to reduce the stocking rate as a whole. The net household income de- recover relatively quickly, especially when there was a good year with clined, despite an increase in the total income. Livestock husbandry was grasslands having high productivity. It implies that one premise is that still the dominant source of household income. Off-farm income still households’ incomes and livings keep stabilized or even higher/better if played a complementary role though its amount and proportion of total herder households willingly reduce livestock numbers. income increased. The subsidy from the GEPAP became the second However, it must be admitted that high market prices contribute to important source of household income. Regression results indicated increasing household incomes of herders (Wang et al., 2013; Wei and that GEPAP had no significant impact on the stocking rates as a whole. Hou, 2015), which was also proved by our analysis (Tables S3-S4). The households with lower proportion of subsidy in total income Hence, it suggests that the natural environmental changes, especially tended to have higher incomes, which reflected the less importance of the natural disasters (e.g., the heavy snow and drought) should be taken subsidy to herder households compared with livestock and thus influ- into consideration in the improvement of market. The livestock in- enced herder decision making. The stocking rate was statistically un- surance should be introduced into the pastoral regions, where the in- related to market (the price per sheep unit) while the household income surance market was effectively absent. Besides, the governments should was positively affected by it. The households whose heads were male set the lowest prices of livestock products and provide the price subsidy tended to have higher household incomes. under natural disaster conditions to motivate households to sell live- Our results imply that the restoration of grassland ecosystems was a stock in order to reduce the losses and ensure the stability of their complex process that required a large amount of financial resources by living, because the longer the disaster lasts, the greater the losses, in- governments subsidized to millions of herder households over a long cluding the livestock and economic losses, which may lead to a poverty time. For the long-term success of PES programs, effective monitoring trap. For instance, our recent survey found that in and management should be strengthened, and differential subsidies of typical steppe region, to help herders deal with the drought from should be given to herder households who comply with the programs to 2015 to 2017, the local government stabilized the livestock prices and diff erent extents in different regions to encourage their active partici- advocated herders selling livestock at reasonable prices as early as pation. It also suggests that the natural environmental changes, espe- possible in 2017, and at the time, gave 60–80 yuan/sheep subsidy (US$ cially the extreme climate events, should be taken into consideration in 8.7–11.5/sheep). Under such policy many local households willingly the enhancements of market of pastoral regions. In addition, many sold their excess livestock before their conditions deteriorated. It con- auxiliary measures, such as providing more access to improved tech- tributed greatly to reducing livestock numbers and stabilizing the in- nologies, improving the infrastructure (e.g., roads, sheds), and pro- comes of households under the threat of a natural disaster. viding more access to the training of off-farm employment and jobs The subsidy was the same for all households in one region who should be taken to work together to ensure that households could still belong to the grass-animal balance and grazing ban areas, respectively. increase incomes even if reducing livestock, and transfer more labor to No consideration was given to whether the herder would consider re- non-farm industries, and finally the grassland degradation could be ducing livestock numbers and no monitoring was carried out to de- restrained. The results also imply that herder decision making played a termine if the households had reduced their livestock or ban the live- significant role in the performance and success of conservation pro- stock grazing according to the policy. This negatively influenced herder grams which needs to be further researched. The GEPAP performance in households’ initiatives to reduce livestock numbers and became another China provides an example for the implementation of a PES program important impediment to households complying with the policy as all where useful lessons have been learned. We hope that this study will the households received the same subsidy in a specific region. Hence provide useful insights for the design and implementation of con- differential payments should be given to households who comply with servation programs. the policy to different extents and in different landscape types (Addison and Greiner, 2016; Fu et al., 2014; Hu et al., 2015; Newton et al., 2012; Declarations of interest Qi et al., 2016; Wei and Qi, 2017a) in order to give reasonable pay- ments to the households who reduced numbers of livestock so as to None. motivate them to reduce livestock willingly, and improve the efficiency of the GEPAP. Acknowledgements There were also some exceptions contrary to the whole results re- garding the GEPAP performance in some regions that cannot be This paper was funded by National Natural Science Foundation of

127 Y. Yin et al. Land Use Policy 82 (2019) 120–129

China (NSFC) (71774162), Science and Technology Innovation Project University, Lanzhou. (in Chinese) Master thesis. of Chinese Academy of Agricultural Sciences (CAAS-ASTIP-IGR2015- Jinhua, 2015. The effect of Grassland Ecological Protection Subsidy Policies on herders and the relevant measures. Chin. Livestock Poultry Breeding 11, 48 (in Chinese). 05) and National Basic Research Program of China (2014CB138806). Michalk, D., Kemp, D., Han, G.D., Hua, L.M., Nan, Z.B., Takahashi, T., Wu, J., Xu, Z., We thank all other participants for their efforts in the survey work. We 2011. Re-designing livestock strategies to reduce stocking rates and improve incomes are also grateful for the herders to give us much time to allow us to ask on western China’s grasslands. In: Kemp, D.R., Michalk, D.L. (Eds.), Development of Sustainable Livestock Systems on Grasslands in North-Western China’. ACIAR them so many questions. Proceedings 134, Australian Centre for International Agricultural Research, Canberra, pp. 140–151. Appendix A. Supplementary data Kemp, D., Han, G.D., Junk, G., Langford, C., Michalk, D., Millar, G., Takahashi, T., Xu, Z., Zhao, M.L., 2011a. Demonstration farms to improve grasslands and household in- comes in western China: what can be done what are we learning. In: Kemp, D.R., Supplementary material related to this article can be found, in the Michalk, D.L. (Eds.), Development of Sustainable Livestock Systems on Grasslands in online version, at doi:https://doi.org/10.1016/j.landusepol.2018.11. North-Western China’. ACIAR Proceedings 134, Australian Centre for International – 037. Agricultural Research, Canberra, pp. 152 161. Kemp, D.R., Brown, C., Han, G., Michalk, D., Nan, Z., Wu, J., Xu, Z., 2011b. Chinese grasslands: problems, dilemmas, and finding solutions. In: Kemp, D.R., Michalk, D.L. References (Eds.), Sustainable Development of Livestock Systems on Grasslands in North- Western China. ACIAR Proceedings 134, Australian Centre for International Agricultural Research, Canberra, pp. 12–24. – Addison, J., Greiner, R., 2016. Applying the social ecological systems framework to the Kong, D., Hu, Z., Jin, L., 2016. Research on the allocation model for grassland eco- evaluation and design of payment for ecosystem service schemes in the Eurasian compensation fund: based on the empirical analysis of 34 Gacha in Inner Mongolia. J. – steppe. Anglais 25, 2421 2440. Arid Land Resour. Environ. 30, 1–6 (in Chinese). Akiyama, T., Kawamura, K., 2007. Grassland degradation in China: methods of mon- Lei, W., 2016. An Empirical Study on the Implementation Effect of Grassland Ecological – itoring, management and restoration. Grassl. Sci. 53, 1 17. Compensation Policy. Inner Mongolia agricultural University, Master thesis. Baosaodubilige, 2015. Study on Satisfaction to the Grassland Ecology Award Mechanism Huhhot. (in Chinese). . fl and the In uncing Factors of the Nomads-take Xin Barag Right Banner As an Li, P., Sun, X., Zhang, J., Zhang, X., Ren, W., 2017. Problems and suggestions on the Example. Inner Mongolia Agricultural University, Huhhot. (in Chinese) Master thesis. grassland eco-compensation policy. Chin. J. Grassland 39, 1–6 (in Chinese). Cao, S., Wang, X., Song, Y., Chen, L., Feng, Q., 2010. Impacts of the Natural Forest Li, S., Verburg, P.H., Lv, S., Wu, J., Li, X., 2012. Spatial analysis of the driving factors of Conservation Program on the livelihoods of residents of northwestern China: per- grassland degradation under conditions of climate change and intensive use in Inner ff – ceptions of residents a ected by the program. Ecol. Econ. 69, 1454 1462. Mongolia, China. Regional Environ. Change 12, 461–474. ’ Cencetti, E., 2011. Tibetan Plateau grassland protection: Tibetan herders ecological Li, W., Ali, S.H., Zhang, Q., 2007. Property rights and grassland degradation: a study of conception versus state policies. HIMALAYA J. Assoc. Nepal Himalayan Stud. 30, 12. the Xilingol Pasture, Inner Mongolia, China. J. Environ. Manage. 85, 461–470. ff Chai, G., 2016. Analysis of the e ect of Grassland Ecological Protection Subsidy Policies- Li, X., Zeng, K., 2016. Ecological benefits of grassland ecological subsidy incentives taking Sunan Yugu Autonomous county of Gansu province as an example. J. Anim. policy-take in Luhuo county, Ganzi Tibetan Autonomous Prefecture, Sichuan as an – Sci. Vet. Med. 35, 113 116 (in Chinese). example. China Agric. Inf. 65–67. Chen, Y., Liu, Y., Amula, 2013. An investigation report of Grassland Ecological Protection Lin, Y., Yao, S., 2014. Impact of the Sloping Land Conversion Program on rural household – Subsidy Incentives in Inner Mongolia. Inner Mongolia Prataculture 25, 15 18 (in income: An integrated estimation. Land Use Policy 40, 56–63. Chinese). Ma, N., 2017. Research on human resources development in pastoral area of Xilingol Chen, H., Xiao, H., 2013. Herders' evaluation and expectation of the Grassland Ecological League in the background of Grassland Ecological Subsidy System. Inner Mongolia Protection Award Policy- based on an investigation in the context of sustainable Sci. Technol. Econ. 15–17 (in Chinese). – development. Modern Econ. Res. 8, 42 46 (in Chinese). Mai, X., Cao, Y., Zhou, Z., Huang, L., Zhou, Q., Chen, Y., 2016. Design and contruction of Cong, Z., Guo, J., 2010. Strategy for behavior of the herders in the process of grassland solar greenhouse for livestock in Tibetan regions of Sichuan province in the context of ecological performance in Inner Mongolia autonomous region. J. Inner Mongolia Grassland Eco-compensation Policy. J. Sourthwest Univ. Nationalities (Nat. Sci. Ed.) – Univ. (Phil. Soc. Sci.) 42, 5 10 (in Chinese). 42, 24–33 (in Chinese). fl Dalintai, Narengawa, 2011. Re ection on theory and system of overgrazing in Inner Newton, P., Nichols, E.S., Endo, W., Peres, C.A., 2012. Consequences of actor level live- – Mongolia. Northern Econ. 6, 32 35 (in Chinese). lihood heterogeneity for additionality in a tropical forest payment for environmental Dean, W.R.J., Macdonald, I.A.W., 1994. Historical changes in stocking rates of domestic services programme with an undifferentiated reward structure. Glob. Environ. Chang. livestock as a measure of semi-arid and arid rangeland degradation in the Cape Part A 22, 127–136. – Province, South Africa. J. Arid Environ. 26 (3), 281 298. Paavola, J., Hubacek, K., 2013. Ecosystem services, governance, and stakeholder parti- – Farhad, S., Gual, M.A., Ruiz-Ballesteros, E., 2015. Land Use Policy 46, 91 102. cipation: an introduction. Ecol. Soc. 18, 42. https://doi.org/10.5751/ES-06019- Fu, B., Wang, Y.K., Xu, P., Yan, K., Li, M., 2014. Value of ecosystem hydropower service 180442. and its impact on the payment for ecosystem services. Sci. Total Environ. 472, Page, G., Bellotti, B., 2015. Farmers value on-farm ecosystem services as important, but – 338 346. what are the impediments to participation in PES schemes? Sci. Total Environ. 515, – Hajkowicz, S., 2009. Land Use Policy 26, 471 478. 12–19. ff Hao, T., 2016. Inner Mongolia Grassland Ecological Protection Subsidy Incentives E ects Pan, J., 2015. Subject and relationship of rights and obligations connotation for grassland Evaluation-the Ewenke District As an Example. Inner Mongolia Agricultural eco-compensation—analysis of Grassland Ecological Protection Subsidy and Reward University, Huhhot. (in Chinese) Master thesis. Program in Tianzhu countty of Gansu province. J. HIT (Soc. Sci. Ed.) 17, 37–44. Harris, R.B., 2010. Rangeland degradation on the Qinghai-Tibetan Plateau: a review of Qi, X., Gao, B., Wang, H., Zhou, J., Qiao, G., 2016. The study on the compensation and – the evidence of its magnitude and causes. J. Arid Environ. 74, 1 12. award standards for forage-livestock balance and grazing prohibition based on ’ Hou, X., Yin, Y., Michalk, D., Yun, X., Ding, Y., Li, X., Ren, J., 2014. Herders opinions herders’ perspective of grassland ecological protection subsidies and incentives po- about desirable stocking rates and overstocking in the rangelands of northern China. licies-take as an example. J. Arid Land Res. Environ. 30, 30–35 (in – Rangel. J. 36, 601 610. Chinese). Hu, X., 2016a. Achievement and problems of applying Ecological Protection and Reynolds, S., Batello, C., Baas, S., Mack, S., 2005. Grassland and forage to improve li- – Compensation Policy in Altay Prefecture. China. Environ. Sci. Manage. 41, 148 151. velihoods and reduce poverty. McGilloway, D. (Ed.), Grassland: a Global Hu, Z., 2016b. China Grassland Eco-compensation Mechanism: Empirical Research in Resource’,Proceedings of the XXth International Grassland Congress 323–338. Inner Mongolia and Gansu. China Agricultural University, Beijing. (in Chinese) Ph.D. Shi, W., 2017. Implementation effect evaluation of Grassland Ecological Protection thesis. Award in Xinjiang. Qinghai Prataculture 26, 40–43 (in Chinese). ff “ Hu, Z., Kong, D., Jin, L., 2015. Grassland eco-compensation: rate di erentiations of re- Sneath, D., 2000. Changing Inner Mongolia, pastoral Mongolian Society and the Chinese ” ward for Balanced Grazing and its reasons. China Popul. Resour. Environ. 25, State. Oxford University Press, UK. – 152 159 (in Chinese). Squires, D., 2014. Biodiversity conservation in Asia. Asia Pac. Policy Stud. 1, 144–159. ’ Hua, L., Squires, V.R., 2015. Managing China s pastoral lands: current problems and fu- Squires, V.R., Yang, Y., 2009. 2 historical degradation episodes in China: socio-economic – ture prospects. Land Use Policy 43, 129 137. forces and their interaction with rangeland grazing systems since the 1950s. In: ff Huang, A., 2015. Evaluation of the Ecological E ects of Grassland Ecological Subsidy Squires, V.R., Lu, X., Lu, Q., Wang, T., Yang, Y. (Eds.), Rangeland Degradation and Incentives Policy in Northwest of Sichuan. Lanzhou University, Lanzhou. (in Recovery in China’S Pastoral Lands. MPG Books Group, Londong, UK, pp. 15–30. Chinese). Tang, L., Dong, S., Sherman, R., Liu, S., Liu, Q., Wang, X., Su, X., Zhang, Y., Li, Y., Wu, Y., ff Huarui, 2013. The e ect of Grassland Ecological Protection Subsidy Policies on herders. Zhao, H., Zhao, C., Wu, X., 2015. Changes in vegetation composition and plant di- – Inner Mongolia Sci. Technol. Econ. 3 7 (in Chinese). versity with rangeland degradation in the alpine region of Qinghai-Tibet Plateau. Jack, B.K., Kousky, C., Sims, K.R., 2008. Designing payments for ecosystem services: Rangel. J. 37, 107–115. lessons from previous experience with incentive-based mechanisms. Proc. Natl. Acad. Waldron, S., Brown, C., Longworth, J., 2010. Grassland degradation and livelihoods in – Sci. U. S. A. 105, 9465 9470. China’s western pastoral region: a framework for understanding and refining China’s ’ Jiang, D., Sarula, Wang, L., 2014. Willingness of herders participation in grassland recent policy responses. China Agric. Econ. Rev. 2, 298–320. ecological protection subsidies reward system. J. Inner Mongolia Agric. Univ. (Soc. Wang, J., Brown, D.G., Agrawal, A., 2013. Climate adaptation, local institutions, and – Sci. Ed.) 16, 13 17 (in Chinese). rural livelihoods: a comparative study of herder communities in Mongolia and Inner ff Jin, C., 2016. Economic E ect Analysis and Policy Proposal of Grassland Ecological Mongolia, China. Glob. Environ. Change 23, 1673–1683. — Compensation Policy a Case Study in Alasan Region, Inner Mongolia. Lanzhou Wang, H., Zhou, X., Wan, C., Fu, H., Zhang, F., Ren, J., 2008. Eco-environmental

128 Y. Yin et al. Land Use Policy 82 (2019) 120–129

degradation in the northeastern margin of the Qinghai–Tibetan Plateau and com- Xilinhot city in Inner Mongolia as an example. Anim. Husbandry Feed Sci. 36, 81–85 prehensive ecological protection planning. Environ. Geol. 55, 1135–1147. (in Chinese). Wang, J., Wang, Z., Xu, L., Ding, Y., 2016. Problems and countermeasures in the im- Yi, F., Lv, W., Ye, R., Wang, X., Saixiyala, Sun, H., Liu, Y., 2015. The performance of plementation of Grassland Ecological Grant-premium Mechanism based on in- Grassland Ecological Protection Subsidy and Reward Policy in Hulunbeir, Inner vestigation of Household in Xilinhot. Chin. J. Grassland 38, 1–7 12. (in Chinese). Mongonlia. Inner Mongolia Sci. Technol. Econ. 3–5 (in Chinese). Wei, H., Qi, Y., 2017a. Analysis of grassland eco-compensation standard based the dif- Yin, R., Yin, G., Li, L., 2010. Assessing China’s ecological restoration programs: what’s ferentiation of the opportunity losses caused by reducing livestock. J. China Agric. been done and what remains to be done? Environ. Manage. 45, 442–453. Univ. 22, 199–207 (in Chinese). Yin, R.S., Liu, C., Zhao, M., Yao, S.B., Liu, H., 2014. The implementation and impacts of Wei, H., Qi, Y., 2017b. The analysis of herdsmen’s willingness to accept the reducing- China’s largest payment for ecosystem services program as revealed by longitudinal livestock policy based on the CVM. J. Arid Land Resour. Environ. 31, 45–50 (in household data. Land Use Policy 40, 45–55. Chinese). Yin, X., 2017. Implementation performance and suggestions of grassland eco-compensa- Wei, Q., Hou, X., 2015. Reflection of establishing a long-term mechanism of grassland tion policies: based on Urat Back Banner, Inner Mongolia. Ecol. Econ. 33, 39–45 (in ecological compensation in China. Scientia Agric. Sinica 48, 3719–3726 (in Chinese). Chinese). Wen, L., Dong, S., Li, Y., Li, X., Shi, J., Wang, Y., Liu, D., Ma, Y., 2013. Effect of de- Yin, Y., Hou, X., Michalk, D., Yun, X., Ding, Y., Li, X., Ren, J., 2013. Herder mental gradation intensity on grassland ecosystem services in the alpine region of Qinghai- stocking rate in the rangeland regions of northern China. In: Michalk, D.L., Millar, Tibetan Plateau, China. PloS one 8 e58432. G.D., Badgery, W.B., Broadfoot, K.M. (Eds.), Revitalising Grasslands to Sustain Our Wenming, Tuya, Eerdunwuritu, Han, C., 2013. Investigation of the implementation of Communities: Proceedings 22nd International Grassland Congress, India, pp. grassland ecological protection subsidy and reward system in some areas of Inner 1833–1836. Mongolia. J. Inner Mongolia Agric. Univ. (Soc. Sci. Ed.) 15, 16–19 (in Chinese). Zhang, H., 2015. Effect of the Grassland Ecological Protection Subsidy and Reward Xu, J., Yin, R., Li, Z., Liu, C., 2006. China’s ecological rehabilitation: unprecedented ef- System on poverty-a Case Study in Alashan Left Banner. Inner Mongolia. Xuehai. pp. forts, dramatic impacts, and requisite policies. Ecol. Econ. 57, 595–607. 50–56 (in Chinese). Yang, C., Meng, Z., Yang, X., 2016. Livestock production and pastoralists’ income in Zheng, Y., Xu, Z., Kemp, D., Jones, R., 2011. Modeling optimal grazing management for pastoral and semi-pastoral area under Grassland Ecological Protection Award and grassland rehabilitation on the typical steppe: a case study in Taipusi Banner, Inner Subsidy Policy. Chin. Agric. Sci. Bull. 32, 8–12 (in Chinese). Mongolia, China. Philippine Agric. Sci. 93, 420–428. Ye, R., Lv, W., Yi, F., Liu, Y., Chang, H., Sun, H., Liu, Y., 2015. Research on the im- Zhou, S.K., Liu, J., 2009. Grassland degradation: herders’ cognition and their counter- plementation effect of Grassland Ecological Protection Subsidy Award Policy-taking measure. Agric. Econ. 9, 24–26 (in Chinese).

129