Soil organic carbon dynamics in Xilingol grassland of northern China induced by the Beijing-Tianjin Sand Source Control Program

Abstract

To mitigate impacts of sandstorms on northern China, the Chinese government launched the Beijing–Tianjin Sand Source Control Program (BTSSCP) in 2000. The associated practices (i.e., cultivation, enclosure, and aerial seeding) were expected to greatly enhance grassland carbon sequestration. However, the BTSSCP-induced soil organic carbon (SOC) dynamics remain elusive at a regional level. Using the Xilingol League in Inner Mongolia for a case study, we examined the impacts from 2000 to 2006 of the BTSSCP on SOC stocks using the IPCC carbon budget inventory method. Results indicated that over all practices SOC storage increased by 1.7%, but there were large differences between practices. SOC increased most rapidly at the rate of 0.3 Mg C·ha⁻¹·yr⁻¹ under cultivation, but decreased significantly under aerial seeding with moderate or heavy grazing (0.3 vs.0.6 Mg C·ha⁻¹·yr⁻¹). SOC increases varied slightly for grassland types, ranging from 0.10 Mg C·ha⁻¹·yr⁻¹ for temperate desert steppe to 0.16 Mg C·ha⁻¹·yr⁻¹ for temperate meadow steppe and lowland meadow. The overall economic benefits of the SOC sink were estimated to be 4.0 million CNY. Aerial seeding with no grazing was found to be the most cost-effective practice. Finally, we indicated that at least 55.5 years (shortest for cultivation) were needed for the grasslands to reach their potential carbon stocks. Our findings highlight the importance and effectiveness of BTSSCP in promoting terrestrial carbon sequestration which may help mitigate climate change, and further stress the need for more attention to the effectiveness of specific practices.

Keywords

grassland carbon sequestration / ecological restoration / Beijing-Tianjin Sand Source Control Program (BTSSCP) / IPCC carbon budget inventory method

Introduction

Grasslands cover 41.7% of China’s total land area and mostly occur in the northern arid and semiarid regions of China (DAHV and CISNR, 1996). They are of key value in servicing local ecological functions and socio-economics (Kang et al., 2007). However, about 90% of the grasslands in the arid and semiarid regions are seriously degraded due to intensive human activities and climate change (Chen and Wang, 2000; Runnström, 2000; Ni, 2003). Deterioration of these grasslands’ biodiversity together with ecosystem functions and services have created serious desertification and sandstorms that can affect broad areas beyond their local area (Zhang et al., 2004; Liu, 2006; Wu et al., 2012). Beijing, the capital of China, has long suffered from sandstorms arising from the grasslands, and these seriously affect industrial and agricultural activities, transportation, and daily life (Goudie and Middleton, 1992; Wang et al., 1999; Gao et al., 2002).

To mitigate the impacts of sandstorms and improve the ecological condition of the northern grasslands, the Chinese government launched the Beijing-Tianjin Sand Source Control Program (BTSSCP) in 2000. It covers 75 counties (banners, cities, or districts) of Beijing, Tianjin, Hebei, Shanxi, and Inner Mongolia, with a combined area of 46 million ha (Wu et al., 2012; http://www.gov.cn). The BTSSCP includes multiple practices such as grassland management (i.e., cultivation, enclosure, and aerial seeding), returning cropland to forest, and reforestation/afforestation (Wu et al., 2013). About 56.2 billion CNY (at present, $1U.S.=6.2 CNY) was invested in the program from 2000 to 2010 (http://www.gov.cn). The ecological effects of the program have attracted considerable interest by scientists and decision makers (Zhang et al., 2012; Huang et al., 2013; Wu et al., 2013, 2014b; Zeng et al., 2014).

Although carbon sequestration was not a goal of the BTSSCP, such a large-scale project has great potential for enhancing terrestrial carbon sequestration and thus may help mitigate climate change (Houghton et al., 1999; Ogle et al., 2004; IPCC, 2006; Wang et al., 2011; Liu et al., 2013; Zhou et al., 2014). As the largest organic carbon pool of terrestrial ecosystems (Schimel, 1995), soil organic carbon (SOC) can be considered as one of the key indicators for evaluating the effectiveness of the BTSSCP (Zeng et al., 2014). However, the impacts of the BTSSCP on the grassland SOC pools remains poorly understood at a regional scale. Relatively limited efforts have focused only on vegetation cover and site-scale observations (Su et al., 2003, 2004, 2005; He et al., 2008, 2009; Li et al., 2011; Wang et al., 2011; Li et al., 2012; Liu et al., 2012; Wu et al., 2013, 2014a,b; Hao et al., 2014; Hu et al., 2015). For example, Wu et al. (2013, 2014b) found that the normalized difference vegetation index (NDVI) increased significantly over 45.9% of the BTSSCP region. He et al. (2009) demonstrated that grazing exclusion practices increased the SOC density of the 0−10 cm soil layer by 73.3% in the sand fraction in Inner Mongolia. Thus a systematic evaluation of SOC dynamics is urgently needed for better understanding of the carbon sequestration potential and the effectiveness of BTSSCP.

Because SOC cannot be directly observed at the regional or global scale, extrapolating site-scale measurements by a statistical approach has emerged as the common approach for predicting SOC over large areas (Wang et al., 2003; Wu et al., 2003; Holmes et al., 2006; Xie et al., 2007; Meersmans et al., 2008; Piao et al., 2009; Yang et al., 2010). For example, Holmes et al. (2006) examined SOC patterns in the southwestern Amazon basin by using the ordinary kriging interpolation method via a large soil profile database. Piao et al. (2009) scaled up soil carbon inventory data obtained in the 1980s to a national level for China with an empirical regression method. Yang et al. (2010) estimated the SOC stock of northern grasslands in China during 2000s via the relationship between SOC densities and environmental factors. However, the soil carbon inventories extrapolation method is not appropriate to study the rapid SOC dynamics induced by the BTSSCP since the soil inventories are time-consuming and costly. Instead, the IPCC carbon budget inventory method (IPCC, 1997, 2006) can more accurately evaluate the effects of managements on the SOC stocks at a regional scale. It has been successfully used to assess management-induced SOC changes in the United States, Europe, and Russia (IPCC, 1997, 2006; Subak, 2000; Houghton, 2001; Ogle et al., 2004; Somogyi et al., 2007; Dolman et al., 2012).

In this study, we assessed the impacts of the BTSSCP on the grassland SOC pools from 2000 to 2006 in the Xilingol league of Inner Mongolia using the IPCC carbon budget inventory method (IPCC, 2006). The Xilingol grassland is well known for its large area and representativeness of the northern arid and semiarid grassland ecosystem (Bai et al., 2004, 2008; Li et al., 2012). Due to their location and serious environmental problems, all the counties (banners) of Xilingol grassland were set up as pilot areas for the BTSSCP. Assessing the grassland SOC dynamics in Xilingol can therefore provide important scientific and practical bases for the sustainable implementation of the BTSSCP in the arid/semiarid grassland areas of China.

The objectives of this study were to (i) quantify the impacts of the BTSSCP on SOC stocks, (ii) assess the economic benefits of SOC sequestration, and (iii) investigate the time needed to reach the potential SOC density of the grassland for the three most commonly grassland management practices (i.e., cultivation, enclosure, and aerial seeding).

Materials and methods

Study area

The Xilingol grassland (41°32′−46°41′N, 111°6′− 120°10′E) is located in the central part of Inner Mongolia, China (Fig. 1). It covers a total area of 20×10⁶ ha and has a typical semiarid climate, with a mean annual temperature of 2.2°C (varying between −2.3°C and 5.6°C) and a mean annual precipitation of 278 mm (varying between 135 mm and 433 mm). Precipitation and soil fertility decrease gradually from east to west, resulting in three zonal vegetation types of temperate meadow steppe, temperate steppe, and temperate desert steppe (Liu, 1960; Li, 1962, 1979; Wu and Loucks, 1992). The temperate meadow steppe is dominated by Leymus chinensis and Stipa baicalensis, the temperate steppe by Leymus chinensis, Stipa grandis, Stipa krylovii, Artemisia frigida, and the temperate desert steppe by Stipa klemenzii and Cleistogenes songorica. The “non-zonal” grassland type of lowland meadow (dominated by many species including Calamagrostis epigejos, Sanguisorba officinalis, and Potentilla anserina) also occurs in Xilingol. In addition to these four dominant grassland types, small areas of mountain meadow and temperate steppe desert are found in the east and west margins of the region, respectively.

The Xilingol grassland was set up as the pilot area for the BTSSCP in 2000, and included the three main management practices of cultivation, enclosure, and aerial seeding (http://www.gov.cn). Cultivation includes planting productive forage (i.e., Medicago sativa), irrigation, fertilization, and fencing the planting area to keep animals out. The enclosure practice involves enclosure of areas of grassland by post and wire fencing. Two grazing policies were implemented under this practice: seasonal controlled grazing and no grazing (Li et al., 2013). The aerial seeding practice involves spreading seeds of forage species over grassland from a plane. By the end of 2006, the management areas covered by the cultivation, enclosure, and aerial seeding practices were 40.2, 669.3, and 34.3 ×10³ ha, respectively (Table 1).

Methods and data

SOC stock estimation

We assumed that the cultivation, enclosure, and aerial seeding practices were implemented averagely on the four main grassland types (i.e., temperate meadow steppe, temperate steppe, temperate desert steppe, and lowland meadow) because the spatial distribution of these practices was not available for this study. The SOC stock (soil depth of 40 cm) at the t^th year in the entire BTSSCP area (C_t) was calculated as follow:

(1)C_t={\displaystyle \sum _{i=\mathrm{1}}^{\mathrm{3}}{C}_{it}},

where i represents the grassland management practice (i.e., i=1 is cultivation, i=2 is enclosure, and i=3 is aerial seeding), and C_it is the SOC stock for the i^th practice at the t^th year. C_it was calculated using the IPCC carbon budget inventory method (IPCC, 1997, 2006). This approach has been successfully used to calculate the management-induced changes of SOC pools in the United States, Europe, and Russia (IPCC, 1997, 2006; Subak, 2000; Houghton, 2001; Ogle et al., 2004; Somogyi et al., 2007; Dolman et al., 2012). The SOC stock over the first 20 years under certain practices (C_i20) was estimated as follow:

(2)C_{i\mathrm{20}}=D_{soc}\times S_i\times F_i,

where D_soc is the SOC density at the initial year (Mg C·ha⁻¹), S_i is the area covered by the i^th practice (ha), and F_i is the stock change factor. Since our study period (i.e., 2000 to 2006) was shorter than 20 years, and as the area covered by the i^th practice increased each year, due to data limitation we assumed the carbon increased equally each year and calculated C_it as follow:

(3)C_{it}=D_{soc}\times \Delta S_{it}+D_{soc}\times S_{it}\times (F_i-\mathrm{1})/\mathrm{20},

where DS_it and S_it are the increased and total areas (ha) covered by the i^th practice in the t^th year, respectively. The values of DS_it and S_it were provided by the Inner Mongolia Rangeland Survey Institute (Table 1). The values of D_soc for lowland meadow, temperate meadow steppe, temperate steppe, and temperate desert steppe were set to (51.78, 51.11, 40.78, and 31.72) Mg C·ha⁻¹, respectively, which were obtained using China’s national soil inventory data (Wang et al., 2003) and the grassland distribution map (ECAGRC, 1993, http://www.geodata.cn/). The value of F_i (Table 2) was established according to “2006 IPCC Guidelines for National Greenhouse Gas Inventories” (Ogle et al., 2004; IPCC, 2006) and previous syntheses of the management-induced SOC changes in China’s grasslands (Shi et al., 2009; Lu, 2009; Wang et al., 2011).

Since the grassland under aerial seeding practice was not fenced, we further classified the practice into four types based on the grazing intensities in Xilingol (Liu, 2006): no grazing (AS), light grazing (AS-LG), moderate grazing (AS-MG), and heavy grazing (AS-HG). The SOC stock for aerial seeding under a certain grazing intensity was estimated as

(4)C_{3jt}=D_{soc}\times \Delta S_{\mathrm{3}jt}+D_{soc}\times (F_{\mathrm{3}j}-\mathrm{1})\times S_{\mathrm{3}jt}/\mathrm{20},

where j represents grazing activity (i.e., j=1 is no grazing, j=2 is light grazing, j=3 is moderate grazing, and j=4 is heavy grazing), DS_3jt and S_3jt are calculated by the area of aerial seeding and the area ratio of j^th grazing activity (Liu, 2006), and F_3j was obtained by multiplying the stock change factor of aerial seeding by that of the j^th grazing intensity.

Economic benefits of SOC sequestration

The economic benefit of SOC sequestration (B_c) was calculated as

(5)B_c=V_c-C_p,

where V_c is the economic value of SOC sequestration, and C_p is the cost of the management practices. We calculated V_c as

(6)V_c={\displaystyle \sum _{i=\mathrm{1}}^{\mathrm{3}}(C_{i\mathrm{2006}}-C_{i\mathrm{2000}})\times P_c},

where C_i2006 and C_i2000 represent the SOC stocks for the i^th practice in 2006 and 2000, respectively, and P_c is the carbon sink price (CNY/t CO₂). The value of P_c was set to 136.5 CNY/t CO₂ according to the world’s largest emissions trading scheme – the European Union Emission Trading Scheme (Convery and Redmond, 2007).

C_p was calculated as follows:

(7)C_p={\displaystyle \sum _{i=\mathrm{1}}^{\mathrm{3}}(U_i\times S_i)},

where U_i is the unit cost for the i^th practice, and S_i is the total area covered by the i^th practice. The value of U_i was set to 2100, 300, and 210 CNY·ha⁻¹ for the cultivation, the enclosure, and the aerial seeding practices, respectively, according to the financial investment of BTSSCP and local practices (http://www.ndrc.gov.cn and http://www.forestry.gov.cn).

Moreover, we calculated the cost of SOC sequestration for the ith practice (C_si) in order to identify the most cost-effective practice:

(8)C_{si}={\displaystyle (V_{ci}-C_{pi})}/(C_{i2006}-C_{i2000})\mathrm{.}

Calculation of the time needed to reach the potential SOC density

The time needed to reach the potential SOC density under a certain practice (T_i) was calculated as:

(9)T_i=(D_{soc\max }-D_{soc\mathrm{2006}})/\Delta D{}_i,

where D_socmax is the potential SOC density (assumed soil depth was 40 cm), D_soc2006 is the SOC density in 2006, and DD_i is the annual increase of SOC density. Using the space-for-time substitution method, D_socmax was calculated as the average value of the top 10% SOC density for particular grassland types based on the soil inventory data (Wang et al., 2003) and the grassland distribution map (ECAGRC, 1993, http://www.geodata.cn/) of the study area. DD_i was calculated as

(10)\Delta D_i=D_{soc\mathrm{2006}}\times S_{i\mathrm{2006}}\times (F_i-\mathrm{1})/\mathrm{20.}

We assumed that there were two grazing scenarios under aAerial seeding after 2006: no grazing (scenario 1) and grazing at 2000 to 2006 intensity (scenario 2). Consequently, the time needed to reach D_socmax under aerial seeding (T₃) was calculated separately for the two scenarios after 2006. It was calculated according to Eq. (9) in scenario 1 and as follows in scenario 2:

(11)T_{3j}=(D_{soc\max }-D_{soc\mathrm{2009}})/\Delta D_{3j},

where T_3j and DD_3j represent the time needed to reach D_socmax and the annual increase of SOC density for aerial seeding under the j^th grazing intensity, respectively.

Results

Changes of SOC

The SOC stock increased by 1.7% (32.0 Tg C in 2006 and 31.4 Tg C in 2000) after the implementation of the BTSSCP, with a carbon sequestration of 0.1 Mg C·ha⁻²·yr⁻¹ in the study period. Changes of the SOC stocks and densities varied substantially across management practices (Table 3). Specifically, the increase of SOC stock under enclosure (i.e., 0.53 Tg C) was greatest and significantly larger than that under the other practices (less than 0.07 Tg C). SOC density also rapidly increased under cultivation (0.3 Mg C·ha⁻¹·yr⁻¹), followed by AS. Notably, SOC storage and density were obviously reduced under AS-MG and AS-HG.

Further, changes of SOC stock and density differed for grassland types (Table 4). The temperate steppe showed the largest increase in SOC stock of 0.33 Tg C, more than 2.3 times that for the other three grassland types. Interestingly, the increases of SOC density overall varied slightly across grassland types, with the least for temperate desert steppe (0.10 Mg C·ha⁻¹·yr⁻¹), and the largest for both lowland meadow and temperate meadow steppe (i.e., 0.16 Mg C·ha⁻¹·yr⁻¹).

Economic benefits of SOC sequestration

The total economic benefits of the grassland SOC sequestration was relatively small from 2000 to 2006 (i.e., 4.0×10⁶ CNY), with the economic value and cost of 296.6 and 292.6 ×10⁶ CNY, respectively (Fig. 2). The economic benefits varied greatly across the management practices. Enclosure had the largest benefits at 63.4 ×10⁶ CNY, while cultivation had a loss of 49.7×10⁶ CNY. The economic balances for the other practices were relatively small (2.0−8.4×10⁶ CNY). Nevertheless, AS had the lowest cost of carbon sequestration (i.e., 0.2 ×10³ CNY Mg C⁻¹) while cultivation had the highest cost of carbon sequestration (i.e., 1.2×10³ CNY Mg C⁻¹).

Time needed to reach the potential SOC density

The time needed to reach the potential SOC density varied across the practices. If there was no grazing after 2006 (i.e., scenario 1), the time was shortest under cultivation (55.5 years), followed by AS (65.6 years), AS-LG (68.1years), AS-MG (81.5 years), and enclosure (86.4 years) (Fig. 3). However, if grazing after 2006 remained at the same level as in 2000 to 2006 (i.e., scenario 2), the time was longest under AS-LG (114.9 years). The SOC storage could not be restored for the AS-MG and AS-HG in scenario 2, as they had no SOC gain.

Discussion

Although carbon sequestration was not a goal of the BTSSCP, the grassland management practices adopted by the Program can increase SOC storage (Conant et al., 2001; Schuman et al., 2002) and have been proposed as an effective way to sequester carbon and therefore help to mitigate climate change (Follett et al., 2001). To our knowledge, no study has been conducted to explore the effects of the BTSSCP on the SOC pools at a regional scale. Our results show that the SOC stock increased significantly after the initiation of the BTSSCP and were highly consistent with site-scale observations (Wu et al., 2008; He et al., 2008, 2009). The increased carbon sequestration can be mainly attributed to the multiple improvements brought about by the management practices of irrigation, fertilization, seeding, and grazing exclusion (Conant et al., 2001; Ogle et al., 2004). These improvements may increase SOC storage by modifying net primary production, root turnover, and carbon allocation between roots and shoots (Post and Kwon, 2000; Conant et al., 2001; Ogle et al., 2004; Steffens et al., 2008). For example, cultivation practices can increase vegetation production by planting more productive grasses, irrigation, and fertilization (Russell and Williams, 1982; Fisher et al., 1994; Crawford et al., 1996). Enclosure practices can elevate SOC inputs from vegetation biomass and litter accumulation by decreasing that removed by grazing (Su et al., 2005; Bilotta et al., 2007).

We showed that the SOC density increased by 0.1 Mg C·ha⁻² per year in the Xilingol grasslands, which was much higher than the average rate for the grasslands in northern parts of China or the whole of China during the 1980s and 2000s (i.e., 0.04 Mg C·ha⁻²·yr⁻¹ vs. 0.018 Mg C·ha⁻²·yr⁻¹) (Piao et al., 2009; Yang et al., 2010). At the same time, the evidence is that the BTSSCP can significantly enhance grassland biomass inclusive of both above-ground and below-ground biomass (Wu et al., 2013, 2014b; Hao et al., 2014). For example, Hao et al. (2014) found that the mean above-ground biomass increased by 0.1−1.1 Mg C·ha⁻¹ from 2004 to 2011 in the Xilingol grasslands due to grazing exclusion. Together these observations suggest that the BTSSCP may play an important role in promoting terrestrial carbon sequestration and therefore help mitigate climate change.

Further, we show that the changes of SOC density varied substantially across management practices with the largest increase under cultivation (Table 3). This can be attributed to the greater management inputs (i.e., seedling, irrigation, and fertilization) made for cultivation than for the other practices (Arnold et al., 1976; Ogle et al., 2004). For example, Ogle et al. (2004) found that the increasing rate of SOC storage under multiple improvements was 1.11 times that made with a single improvement. Notably, consistent with the previous site-scale observations (He et al., 2008; Li et al., 2008; Wu et al., 2008; Wang et al., 2011), SOC decreased significantly under the AS-MG and AS-HG practices (Table 3). This phenomenon can be largely explained by the decrease of SOC input from aboveground biomass and roots and the enhancement of soil organic matter (SOM) decomposition caused by consumption of biomass by grazing (Johnson and Matchett, 2001; Neff et al., 2005).

Changes of SOC density differed slightly for grassland types. Relatively larger increases were observed in lowland meadows and temperate meadow steppes (Table 4), mainly due to the higher precipitation and background SOC storage of these grassland types (DAHV and CISNR, 1996). Hu et al. (2015) concluded that precipitation was the major driver of SOC changes after grazing exclusion in northern China. Also, SOC change was reported to be positively related to background SOC content (Holmes et al., 2006; Yang et al., 2010). This implies that the implementation of the BTSSCP will be more effective in regions with high precipitation and original carbon storage if viewed in the terms of carbon sequestration.

In general the economic benefits of the SOC sequestration were small but varied significantly with management practices (Fig. 2). Though having the highest increase of SOC density (Table 3), cultivation had the largest benefit loss because its cost was much higher than its economic value (84.6 vs. 34.9 million yuan), and thus was not recommended as the optimal practice (Zhang and Liu, 2005; MacLeod and McIvor, 2006). Aerial seeding was the most cost-effective and economic practice in promoting SOC sequestration.

Importantly, we found that a long time is needed for the degraded grasslands to reach their potential SOC density, with the shortest time of 55.5 years with cultivation (Fig. 3). This was comparable to the estimate by Burke et al. (1995) in Colorado, USA that more than 50 years were required for the semiarid grassland to reach their potential SOC stocks. This value, however, was larger than the estimate (above 20 years) by He et al. (2008) for lightly degraded grasslands in Inner Mongolia to return to their natural state. This is indicative of the more serious grassland degradation of our study area.

There remain uncertainties in our study. First, the SOC stock change factor was established according to global and national field observations (IPCC, 2006; Lu, 2009; Shi et al., 2009; Wang et al., 2011), and these values may be slightly different from the actual situation in the BTSSCP region. Secondly, due to the data limitations, we assumed that the cultivation, enclosure, and aerial seeding practices are distributed evenly over the four grassland types. Further studies are needed to ascertain the location of the BTSSCP-induced SOC sink. Thirdly, we analyzed the SOC dynamics in the first seven years of the implementation of the BTSSCP and assumed that the SOC increased equally each year. Nevertheless, the most recent synthesis indicated that the rate of SOC increase declined linearly with the time from the implementation of grassland management (Hu et al., 2015), suggesting that the soil carbon sink may be under-estimated in our study. We do not consider climatic effects in our analysis, although these effects are documented to be able to alter biomass substantially in the Xilingol grassland (Li et al., 2012; Wu et al., 2013, 2014b; Hao et al., 2014; Tian et al., 2015) and therefore may impact on soil carbon sequestration. For example, Wu et al. (2013, 2014b) indicated that NDVI decreased by more than 20% over most of the Xilingol grasslands in the years of 2007 and 2009 because of summer drought. These considerations stress the necessity for more effort towards understanding the ongoing SOC consequences of the BTSSCP by the combined use of direct observations, remote sensing, inventory data, and ecosystem modelling.

References

[1] Arnold P, Hunter F, Gonzalez Fernandez P (1976). Long-term grassland experiments at Cockle Park [Great Britain]. Annales Agronomiques (France)

[2] Bai Y, Han X, Wu J, Chen Z, Li L (2004). Ecosystem stability and compensatory effects in the Inner Mongolia grassland. Nature, 431(7005): 181–184

[3] Bai Y, Wu J, Xing Q, Pan Q, Huang J, Yang D, Han X (2008). Primary production and rain use efficiency across a precipitation gradient on the Mongolia plateau. Ecology, 89(8): 2140–2153

[4] Bilotta G, Brazier R, Haygarth P (2007). The impacts of grazing animals on the quality of soils, vegetation, and surface waters in intensively managed grasslands. Adv Agron, 94: 237–280

[5] Burke I C, Lauenroth W K, Coffin D P (1995). Soil organic matter recovery in semiarid grasslands: implications for the conservation reserve program. Ecol Appl, 5(3): 793–801

[6] Chen Z Z, Wang S P (2000). Typical Steppe Ecosystem of China.Beijing: Science Press (in Chinese)

[7] Conant R T, Paustian K, Elliott E T (2001). Grassland management and conversion into grassland: effects on soil carbon. Ecol Appl, 11(2): 343–355

[8] Convery F J, Redmond L (2007). Market and price developments in the European Union emissions trading scheme. Rev Environ Econ Policy, 1(1): 88–111

[9] Crawford M, Grace P, Bellotti W, Oades J (1996). Below-ground inputs of carbon by crops and pastures. In: Proceedings of the 8th Australian Agronomy Conference Queensland, Australia, 172–175

[10] DAHV (Department of Animal Husbandry and Veterinary, Institute of Grasslands, Chinese Academy of Agricultural Sciences),CISNR (Commission for Integrated Survey of Natural Resources, Chinese Academy of Sciences) (1996). Rangeland resources of China. Beijing: Agricultural Science and Technology, 1–346 (in Chinese)

[11] Dolman A, Shvidenko A, Schepaschenko D, Ciais P, Tchebakova N, Chen T, van der Molen M K, Belelli Marchesini L, Maximov T C, Maksyutov S, Schulze E D (2012). An estimate of the terrestrial carbon budget of Russia using inventory-based, eddy covariance and inversion methods. Biogeosciences, 9(12): 5323–5340

[12] ECAGRC (Editorial Committee of Atlas of Grassland Resources of China) (1993). Atlas of grassland resources of China (1:1M).Beijing: China Atlas (in Chinese)

[13] Fisher M J, Rao I M, Ayarza M A, Lascano C E, Sanz J, Thomas R J, Vera R R (1994). Carbon storage by introduced deep-rooted grasses in the South American savannas. Nature, 371(6494): 236–238

[14] Follett R, Kimble J, Lal R (2001). The Potential of US Grazing Lands to Sequester Soil Carbon. Boca Raton, Florida: Lewis Publishers

[15] Gao Q X, Su F Q, Ren Z H (2002). Dust weather and its influences of Beijing. China. Environ Sci, 22: 468–471 (in Chinese)

[16] Goudie A S, Middleton N J (1992). The changing frequency of dust storms through time. Clim Change, 20(3): 197–225

[17] Hao L, Sun G, Liu Y, Gao Z, He J, Shi T, Wu B (2014). Effects of precipitation on grassland ecosystem restoration under grazing exclusion in Inner Mongolia, China. Landscape Ecol, 29(10): 1657–1673

[18] He N, Wu L, Wang Y, Han X (2009). Changes in carbon and nitrogen in soil particle-size fractions along a grassland restoration chronosequence in northern China. Geoderma, 150(3−4): 302–308

[19] He N, Yu Q, Wu L, Wang Y, Han X (2008). Carbon and nitrogen store and storage potential as affected by land-use in a Leymus chinensis grassland of northern China. Soil Biol Biochem, 40(12): 2952–2959

[20] Holmes K W, Chadwick O A, Kyriakidis P C, Silva de Filho E P, Soares J V, Roberts D A (2006). Large-area spatially explicit estimates of tropical soil carbon stocks and response to land-cover change. Global Biogeochem Cycles, 20(3): GB3004

[21] Houghton R (2001). Counting terrestrial sources and sinks of carbon. Clim Change, 48(4): 525–534

[22] Houghton R A, Hackler J L, Lawrence K (1999). The US carbon budget: contributions from land-use change. Science, 285(5427): 574–578

[23] Hu Z, Li S, Guo Q, Niu S, He N, Li L, Yu G (2015). A synthesis of the effect of grazing exclusion on carbon dynamics in grasslands in China. Glob Change Biol, 

[24] Huang L, Xiao T, Zhao Z, Sun C, Liu J, Shao Q, Fan J, Wang J (2013). Effects of grassland restoration programs on ecosystems in arid and semiarid China. J Environ Manage, 117: 268–275

[25] IPCC (1997). Revised 1996 IPCC guidelines for national greenhouse gas inventories: Reference Manual (Volume 3) (Houghton JT, Meira LG, Filho LG, Lim B, Treanton K, Mamaty I, Bonduki Y, Griggs DJ, Callender BA, eds.) Intergovernmental Panel on Climate Change

[26] IPCC (2006). 2006 IPCC guidelines for national greenhouse gas inventories Volume 4: Agriculture, Forestry and Other Land Uses (AFOLLI) Hayama, Japan: IPCC /IGES, 2006.8

[27] Johnson L C, Matchett J R (2001). Fire and grazing regulate belowground processes in tallgrass prairie. Ecology, 82(12): 3377–3389

[28] Kang L, Han X, Zhang Z, Sun O J (2007). Grassland ecosystems in China: review of current knowledge and research advancement. Philos Trans R Soc Lond B Biol Sci, 362(1482): 997–1008

[29] Li A, Wu J, Huang J (2012). Distinguishing between human-induced and climate-driven vegetation changes: a critical application of RESTREND in inner Mongolia. Landscape Ecol, 27(7): 969–982

[30] Li B (1962). Basic types and eco-geographic distribution of zonal vegetation in Inner Mongolia.  Acta Sci Nat Univ Neimongol, (4): 42–72 (in Chinese)

[31] Li B (1979). General characteristics of the steppe vegetation in China. China’s Grasslands, 1: 2–12 (in Chinese)

[32] Li C, Hao X, Zhao M, Han G, Willms W D (2008). Influence of historic sheep grazing on vegetation and soil properties of a Desert Steppe in Inner Mongolia. Agric Ecosyst Environ, 128(1−2): 109–116

[33] Li S L, Yu F H, Werger M J, Dong M, Ramula S, Zuidema P A (2013). Understanding the effects of a new grazing policy: the impact of seasonal grazing on shrub demography in the Inner Mongolian steppe. J Appl Ecol, 50(6): 1377–1386

[34] Li Y, Zhao H, Zhao X, Zhang T, Li Y, Cui J (2011). Effects of grazing and livestock exclusion on soil physical and chemical properties in desertified sandy grassland, Inner Mongolia, northern China. Environmental Earth Sciences, 63(4): 771–783

[35] Liu A J (2006). The Study of Monitoring Method for ANPP and Dynamic Balance between Supply and Demand Using Remote Sensing. Dissertation for Ph.D degree. Beijing: China Agricultural University (in Chinese)

[36] Liu J, Wu J, Wu Z, Lu A, Yue J (2012). Interactions between climate change and human activities in dryland degradation in Beijing-Tianjin sand-storm source region, China. In: Geoscience and Remote Sensing Symposium (IGARSS) IEEE International, 2012. IEEE, 6416–6419

[37] Liu X, Zhang W, Cao J, Shen H, Zeng X, Yu Z, Zhao X (2013). Carbon storages in plantation ecosystems in sand source areas of north Beijing, China. PLoS ONE, 8(12): e82208

[38] Liu Z (1960). General vegetation pattern of the Inner Mongolia Steppe region. Acta Sci Nat Univ Neimongol, 2: 47–74 (in Chinese)

[39] Lu H L (2009). LULUCF inventory and LULUCF activities effect in China based upon improved IPCC methodology. Dissertation for Ph.D degree. Beijing Institute of Geographic Sciences and Natural Resources Research, Chinese Academy of Sciences (in Chinese)

[40] MacLeod N, McIvor J G (2006). Reconciling economic and ecological conflicts for sustained management of grazing lands. Ecol Econ, 56(3): 386–401

[41] Meersmans J, De Ridder F, Canters F, De Baets S, Van Molle M (2008). A multiple regression approach to assess the spatial distribution of soil organic carbon (SOC) at the regional scale (Flanders, Belgium). Geoderma, 143(1−2): 1–13

[42] Neff J, Reynolds R, Belnap J, Lamothe P (2005). Multi-decadal impacts of grazing on soil physical and biogeochemical properties in southeast Utah. Ecol Appl, 15(1): 87–95

[43] Ni J (2003). Plant functional types and climate along a precipitation gradient in temperate grasslands, north-east China and south-east Mongolia. J Arid Environ, 53(4): 501–516

[44] Ogle S M, Conant R T, Paustian K (2004). Deriving grassland management factors for a carbon accounting method developed by the intergovernmental panel on climate change. Environ Manage, 33(4): 474–484

[45] Piao S, Fang J, Ciais P, Peylin P, Huang Y, Sitch S, Wang T (2009). The carbon balance of terrestrial ecosystems in China. Nature, 458(7241): 1009–1013

[46] Post W M, Kwon K C (2000). Soil carbon sequestration and land-use change: processes and potential. Glob Change Biol, 6(3): 317–327

[47] Runnström M C (2000). Is northern China winning the battle against desertification? Satellite remote sensing as a tool to study biomass trends on the Ordos Plateau in semiarid China.AMBIO: A Journal of the Human Environment, 29: 468–476

[48] Russell J S, Williams C H (1982). Biogeochemical interactions of carbon, nitrogen, sulfur and phosphorus in Australian agroecosystems. In: Freney J R, Galbally I E, eds. Cycling of carbon, nitrogen, sulfur and phosphorus in terrestrial and aquatic ecosystems. Berlin: Springer-Verlag, 61–75

[49] Schimel D S (1995). Terrestrial ecosystems and the carbon cycle. Glob Change Biol, 1(1): 77–91

[50] Schuman G, Janzen H, Herrick J (2002). Soil carbon dynamics and potential carbon sequestration by rangelands. Environ Pollut, 116(3): 391–396

[51] Shi F, Li Y E, Gao Q Z (2009). Effect of management on grassland soil organic carbon. Pratacultural Science, 26: 9–15 (in Chinese)

[52] Somogyi Z, Cienciala E, Mäkipä R, Muukkonen P, Lehtonen A, Weiss P (2007). Indirect methods of large-scale forest biomass estimation. Eur J For Res, 126(2): 197–207

[53] Steffens M, Kölbl A, Totsche K U, Kögel-Knabner I (2008). Grazing effects on soil chemical and physical properties in a semiarid steppe of Inner Mongolia (PR China). Geoderma, 143(1−2): 63–72

[54] Su Y Z, Li Y L, Cui J Y, Zhao W Z (2005). Influences of continuous grazing and livestock exclusion on soil properties in a degraded sandy grassland, Inner Mongolia, northern China. Catena, 59(3): 267–278

[55] Su Y Z, Zhao H L, Zhang T H (2003). Influences of grazing and exclosure on carbon sequestration in degraded sandy grassland, Inner Mongolia, north China. N Z J Agric Res, 46(4): 321–328

[56] Su Y Z, Zhao H L, Zhang T H, Zhao X Y (2004). Soil properties following cultivation and non-grazing of a semi-arid sandy grassland in northern China. Soil Tillage Res, 75(1): 27–36

[57] Subak S (2000). Agricultural soil carbon accumulation in North America: considerations for climate policy. Glob Environ Change, 10(3): 185–195

[58] Tian H, Cao C, Chen W, Bao S, Yang B, Myneni R B (2015). Response of vegetation activity dynamic to climatic change and ecological restoration programs in Inner Mongolia from 2000 to 2012. Ecol Eng, 82: 276–289

[59] Wang S, Tian H, Liu J, Pan S (2003). Pattern and change of soil organic carbon storage in China: 1960s–1980s. Tellus B Chem Phys Meterol, 55(2): 416–427

[60] Wang S, Wilkes A, Zhang Z, Chang X, Lang R, Wang Y, Niu H (2011). Management and land use change effects on soil carbon in northern China's grasslands: a synthesis. Agric Ecosyst Environ, 142(3−4): 329–340

[61] Wang S G, Yang M, Qi B, Xin C L, Yang M F (1999). Influence of sand-dust storms occurring over the Gansu Hexi district on the air pollution in Lanzhou city. J Desert Res, 19: 354–358 (in Chinese)

[62] Wu H, Guo Z, Peng C (2003). Distribution and storage of soil organic carbon in China. Global Biogeochem Cycles, 17(2): GB1048

[63] Wu J, Loucks O L (1992). Xilingele (The Xilingol Grassland) In: The US National Research Council, ed. Grasslands and Grassland Sciences in Northern China.Washington DC: National Academy Press, 67–84

[64] Wu J, Zhao L, Zheng Y, Lü A (2012). Regional differences in the relationship between climatic factors, vegetation, land surface conditions, and dust weather in China’s Beijing-Tianjin Sand Source Region. Nat Hazards, 62(1): 31–44

[65] Wu L, He N, Wang Y, Han X (2008). Storage and dynamics of carbon and nitrogen in soil after grazing exclusion in grasslands of northern China. J Environ Qual, 37(2): 663–668

[66] Wu X, Li Z, Fu B, Zhou W, Liu H, Liu G (2014a). Restoration of ecosystem carbon and nitrogen storage and microbial biomass after grazing exclusion in semi-arid grasslands of Inner Mongolia. Ecol Eng, 73: 395–403

[67] Wu Z, Wu J, He B, Liu J, Wang Q, Zhang H, Liu Y (2014b). Drought offset ecological restoration program-induced increase in vegetation activity in the Beijing-Tianjin Sand Source Region, China. Environ Sci Technol, 48(20): 12108–12117

[68] Wu Z, Wu J, Liu J, He B, Lei T, Wang Q (2013). Increasing terrestrial vegetation activity of ecological restoration program in the Beijing–Tianjin Sand Source Region of China. Ecol Eng, 52: 37–50

[69] Xie Z, Zhu J, Liu G, Cadisch G, Hasegawa T, Chen C, Sun H, Tang H, Zeng Q (2007). Soil organic carbon stocks in China and changes from 1980s to 2000s. Glob Change Biol, 13(9): 1989–2007

[70] Yang Y, Fang J, Ma W, Smith P, Mohammat A, Wang S, Wang W (2010). Soil carbon stock and its changes in northern China’s grasslands from 1980s to 2000s. Glob Change Biol, 16(11): 3036–3047

[71] Zeng X, Zhang W, Cao J, Liu X, Shen H, Zhao X (2014). Changes in soil organic carbon, nitrogen, phosphorus, and bulk density after afforestation of the “Beijing–Tianjin Sandstorm Source Control” program in China. Catena, 118: 186–194

[72] Zhang G, Dong J, Xiao X, Hu Z, Sheldon S (2012). Effectiveness of ecological restoration projects in Horqin Sandy Land, China based on SPOT-VGT NDVI data. Ecol Eng, 38(1): 20–29

[73] Zhang T H, Zhao H L, Li S G, Li F R, Shirato Y, Ohkuro T, Taniyama I (2004). A comparison of different measures for stabilizing moving sand dunes in the Horqin Sandy Land of Inner Mongolia, China. J Arid Environ, 58(2): 203–214

[74] Zhang W, Liu C (2005). The impact of environmental policy on household income and activity choice: evidence from sandstorm source control program in North China. In: American Agricultural Economics Association Annual Meeting, Providence, RI, 24–27

[75] Zhou D, Zhao S, Liu S, Zhang L (2014). Modeling the effects of the Sloping Land Conversion Program on terrestrial ecosystem carbon dynamics in the Loess Plateau: a case study with Ansai County, Shaanxi province, China. Ecol Modell, 288: 47–54

RIGHTS & PERMISSIONS

Higher Education Press and Springer-Verlag Berlin Heidelberg