ARTÍCULOS ORIGINALES

Effects of grazing on plant species diversity and carbon partitioning in semiarid rangelands of northeastern China

Efectos del pastoreo en la diversidad de especies vegetales y particionamiento del carbono en áreas semiáridas de vegetación natural en el noreste de China

 

Hu FL^1,2, B Liu^1,3, ZM Liu^1,3, YT Fang^1,3, CA Busso⁴

¹ Institute of Applied Ecology, Chinese Academy of Science, Shenyang 110016, P R China.
² University of Chinese Academy of Sciences, Beijing 100039, P R China.
³ State Key Laboratory of Forest and Soil Ecology, Institute of Applied Ecology, Chinese Academy of Sciences, Shenyang 110164, P. R. China.
⁴ Departamento de Agronomía-CERZOS [Consejo Nacional de Investigaciones Científicas y Técnicas (CONICET)], Universidad Nacional del Sur, 8000 Bahía Blanca, Pcia. Buenos Aires, Argentina.
Address Correspondence to: Dr. Zhimin Liu, e-mail: zmliu@iae.ac.cn

Recibido / Received 16.IV.2014 .
Aceptado / Accepted 26.IV.2014.

 

---

Abstract. Grasslands are one of the most widespread landscapes worldwide, covering approximately one-fifth of the world's land surface, where grazing is a common practice. How carbon storage responds to grazing in steppes remains poorly understood. We quantified the effects of grazing on community composition and species diversity, and carbon storage in two typical grasslands of northeastern China, one in Horqin and the other one in Hulunbeier. In both grasslands, grazing did not infuence plant species diversity. However, it substantially decreased aboveground carbon by 31% and 54% in Horqin and Hulunbeier, respectively. Fenced and grazing treatments showed a similar belowground carbon at both locations. The predominant carbon pool in the study grassland ecosystem was found in the upper 100 cm soil depth, from 98.2 to 99.1% of the total carbon storage. There were no significant effects of grazing on soil carbon neither in the whole profile nor in the uppermost 20 cm soil depth in the two study grasslands. Studies on the efects of varying rangeland management, such as region disparity and grazing systems, may have important consequences on species diversity and carbon partitioning, and thus on rangeland stability and ecosystem functioning.

Keywords: Species diversity; Carbon allocation; Carbon stock; Optimal partitioning; Community heterogeneity; Soil depth.

Resumen. Las áreas de vegetación natural son unos de los paisajes más ampliamente distribuidos en el mundo, cubriendo aproximadamente un quinto de la superficie terrestre mundial, donde al pastoreo es una práctica común. Se conoce poco en relación a como el pastoreo modifica el almacenamiento de carbono en las estepas. En este estudio se cuantificaron los efectos del pastoreo en la composición y diversidad de especies en la comunidad, y el almacenamiento de carbono en dos áreas típicas de vegetación natural del noreste de China, una en Horqin y la otra en Hulunbeier. En ambas áreas, el pastoreo no afectó la diversidad de especies vegetales. Sin embargo, este redujo substancialmente el carbono en la parte aérea en un 31 y 54% en Horqin y Hulunbeier, respectivamente. El carbono en la parte subterránea fue similar en clausuras y sitios pastoreados en ambos lugares. La mayor cantidad de carbono en las áreas estudiadas se encontró en los primeros 100 cm de profundidad del suelo, desde 98,2 a 99,1% del almacenamiento total de carbono. No hubo efectos significativos del pastoreo en el carbono de todo el perfil y de los primeros 20 cm de profundidad del suelo. Variaciones en el manejo de las áreas de vegetación natural, como resultado de distintas áreas de estudio o sistemas de pastoreo, pueden tener importantes consecuencias en la diversidad y el particionamiento de carbono de las especies, y así en la estabilidad y funcionamiento del ecosistema en dichas áreas.

Palabras clave: Diversidad de especies; Distribución del carbono; Cantidad de carbono; Particionamiento óptimo; Heterogeneidad de la comunidad; Profundidad de suelo.

---

 

INTRODUCTION

Grasslands are one of the most widespread landscapes worldwide, covering approximately 20% of the world's land surface (Scurlock & Hall, 1998). China's grasslands cover 6 to 8% of that surface area and have from 9 to 16% of the total carbon in the world grasslands (Ni, 2002). Because of the importance in regulating the nutrient and carbon cycles, much emphasis has been placed in estimating biomass of the world's grasslands (Klaus et al., 2012; Flanagan et al., 2013). However, overgrazing has contributed to desertification and degradation of more than two-thirds of these fragile ecosystems because of human activities (Karl & Trenberth, 2003; Li et al., 2012). When grasslands are either grazed by livestock or fenced to prevent long-term grazing, above- (AGB) and belowground (BGB) biomasses, and soil organic carbon (SOC), may respond diferently, thus influencing the global carbon cycle (Cox et al., 2000; Amthor, 2006). Therefore, understanding the effects of grazing on carbon storage of grasslands is crucial to help making predictions on global C budgets.
There has been some progress in monitoring and understanding the factors influencing carbon partitioning (Bultynck & Lambers, 2004). Nevertheless, the effects of species composition and diversity in afecting carbon partitioning remain unclear. Many studies have shown that grazing might increase (Manley et al., 1995) or decrease (Wiesmeier et al., 2012) soil carbon accumulation compared to adjacent, fenced soils. The fact that plant and soil properties have not shown a consistent increase or decrease response to grazing has been attributed to the fact that a certain threshold level has been reached (Baron et al., 2002; Han et al., 2008; Silveira et al., 2013). Some studies have been conducted on the Tibetan Plateau or Loess Plateau of China to assess carbon sequestration (Zhao et al., 2006; Feng et al., 2013). However, little has been done to address the influences of grazing on species diversity and carbon allocation in China's grasslands.
We determined how plant species composition and diversity, and ecosystem carbon storage, responded to grazing in the Horqin and Hulunbeier grasslands of northeastern China. Grazing-induced changes in community composition determine in turn changes in aboveground net primary production (ANPP). This study was carried out on both fenced and grazed grasslands in two regions of northeastern China to determine (1) to what extent grazing affects plant community and ecosystem carbon storages, (2) if plant carbon storage differs between the two study regions.

MATERIALS AND METHODS

Study area. This study was conducted in two grasslands in Inner Mongolia, China (Fig.1). One of them was in southwestern Horqin Sandy Land (42° 56′ N - 43° 20′ N, 119° 29′E - 119° 41′ E; 428 to 494 m.a.s.l.). This is characterized by sand dunes alternating with interdune lowlands, where overgrazing and extensive deforestation are the major factors leading to its desertification (Zhang et al., 2004). This area has a combination of temperate, semiarid, and continental monsoonal climate. Average annual precipitation is 340 mm, with 70-80% of which occurring between May and September. The mean annual open-pan evaporation is approximately 1935 mm. Annual temperature averages approximately 6.4 °C, with a mean minimum monthly temperature of -13.1 °C in January, and a maximum of 23.7 °C in July. The dominant species are sagebrush (Salsola ruthenica lljin) and wheat grasses (Cyperus rotundus L.). Horqin steppe is also exposed to severe winds (Liu et al., 2007).

Fig. 1. Schematic diagram indicating the study sites (Horqin and Hulunbeier) in northern China, and the experimental design. Half of each 500 x 500 m plot (n=6) was grazed while the other half remained ungrazed (i.e., fenced). Sampling was conducted within each of fve 1 x 1 m subplots within each plot.
Fig. 1. Diagrama indicando los sitios de estudio (Horqin y Hulunbeier) en el norte de China, y el diseño experimental. La mitad de cada parcela de 500 x 500 m (n=6) fue pastoreada mientras que la otra mitad permaneció sin pastorear (es decir, alambrada). El muestreo se condujo dentro de cada una de cinco subparcelas de 1 m x 1m dentro de cada parcela.

The other grassland was located in central Hulunbeier (47° 20′ N - 50° 12′ N, 118° 31′ E - 121° 03′ E; 599 to 618 m.a.s.l.). The mean annual precipitation is 308 mm; about 70% of annual precipitation occurs in summer and autumn. Total annual evaporation can reach 1371 mm. The mean annual wind velocity is 3.5 m/s, and soils are classified as sandy chestnut. The dominant plant species on the typical steppe are Stipa grandis and Leymus chinensis. Recently, agriculture has resulted in heavier grazing pressure and farm cultivation leading to an apparent deterioration in rangeland quality and productivity (Tang et al., 2004).

Experimental design. On the two study sites, sampling was carried out in mid-July 2010 on 5 to 6 year-old enclosures to grazing by domestic herbivores. At each sampling area of the Horqin and Hulunbeier regions, six plots (500 m × 500 m) were randomly established within the grazed and fenced treatments (Fig. 1). Plots were at least 2 km apart. In each plot, five 1 m × 1 m subplots were set (Fig. 1). Subplots were 15 m away one from each other. In each subplot (Fig.1), aboveground biomass and litter were harvested, and species richness and density were recorded. After litter was collected, three replicates of soil samples were taken at 0-10, 10-20, 20-40, 40-60, 60-80, and 80-100 cm soil depth using a 7 cm-diameter soil auger. Roots were also collected at each soil depth at locations 0.5 to 1 m from places where the soil samples were obtained.
Roots were thoroughly washed manually free from soil in deionized water using a 35-mesh screen (Williams & Baker, 1957). Aboveground plant litter and roots were first dried at 75 °C to a constant weight in an oven, and weighed thereafter. Soil samples were air-dried, and then sieved (2-mm mesh) to remove roots and other debris. Before air-drying, one subsample was obtained from each soil sample, which was first weighed and then oven-dried at 105 °C to determine gravimetric soil moisture following Brown (1995). Another subsample from each soil sample, after air-drying, was ground to pass a 0.15-mm sieve; this subsample was analyzed for soil organic carbon by the dichromate oxidation method of Walkey and Black (Nelson & Sommers, 1996).

Species diversity was calculated using the Simpson's Diversity Index (D) as:

where P_i is the proportion of the i^th species in the community containing s species (Simpson, 1949). This index represents the probability that two individuals randomly selected from a sample will belong to different species. The proportion (i.e., percentage) of each species biomass within each sample plot was first calculated with respect to the total biomass contribution of all species within that plot. This proportion was then expressed as a fraction (i.e., P) to use it in the formula of the Simpson´s Diversity Index.

Carbon storage in plants was expressed as:

B' = B x 45%

where B' is carbon in the plant biomasses of the aboveground, belowground and litter components (g C/cm²), and B is the dry weight of the biomass samples (Crutzen & Andreae, 1990).
Soil organic carbon (SOC) was calculated following Tian et al. (2009) as:

SOC_t = ∑[C_iD_iE_i(1 - G_i)] x 100

where SOC_t is the total value of soil organic carbon storage (g C/m²); C_i is the concentration of soil organic carbon at the soil depth i (%); D_i is the soil bulk density (g/cm³); E_i is the thickness of the soil layer (cm), and G_i is the portion of gravel whose diameter is greater than 2 mm (%).

Statistical analysis. Detrended correspondence analysis (DCA) of different plots was performed by Canoco (version 4.5, Canoco, Netherlands). Data analysis was performed using ANOVA on a windows-based SPSS (version 13, SPSS, Chicago, IL, USA) with significance at p<0.05. When F tests were significant, means were compared using the LSD test of Fisher (Steel & Torrie, 1960). Regression analysis was used to compare the relationship between AG- and BGC in both grazing systems and sites following Neter et al. (1985).

RESULTS

Species composition and diversity. Species composition and density were quite different in the study regions (Tables 1 and 2). The numbers of species in the Horqin and Hulunbeier regions were 53 and 25, respectively (Tables 1 and 2). However, plant density appeared to be greater at the Hulunbeier (e.g., Cyperus fuscus, Agropyron cristatus) than at the Horqin site (Tables 1 and 2). Fenced and grazed grasslands showed some common species in each region (Tables 1 and 2). Some of them were Corispermum stauntonii Moq., Salsola ruthenica Iljin and Setaria viridis (L.) P. Beauv.

Table 1. Species composition and density (# individuals/m²) on fenced and grazed plots in Horqin region. Each value is the mean ± 1 SE of n=6.
Tabla 1. Densidad (número de individuos/m²) y composición de especies en clausuras y parcelas pastoreadas en la región de Horqin. Cada valor es el promedio ± 1 E.E. de n=6.

Table 2. Species composition and density (# individuals/m²) on fenced and grazed plots in Hulunbeier region. Each value is the mean ± 1 SE of n=6.
Tabla 2. Densidad (número de individuos/m²) y composición de especies en clausuras y parcelas pastoreadas en la región de Hulunbeier. Cada valor es el promedio ± 1 E.E. de n=6.

Detrended corresponded analysis on the study species supported that species composition contributed a wider range in the Horqin than in the Hulunbeier grassland. A similar trend between grazed and fenced grasslands was shown in each region (Fig. 2). However, plant species diversity was similar (p>0.05) between fenced and grazed grasslands in either the Horqin or Hulunbeier region (Fig. 3).

Fig. 2. Detrended correspondence analysis (DCA) of different plots, where "×" represents grazed grasslands in Horqin (Hor-G),"▢" represents fenced grasslands in Horqin (Hor-F), "♊" represents grazed grasslands in Hulunbeier (Hul-G), and "○" represents fenced grasslands in Hulunbeier (Hul-F).
Fig. 2. Análisis de correspondencia de diferentes parcelas, donde ‘'x'' representa los pastizales pastoreados en Horqin (Hor-G), ‘'▢" representa los pastizales alambrados en Horqin (Hor-F), "♊" representa los pastizales pastoreados en Hulunbeier (Hul-G), y "○" representa los pastizales alambrados en Hulunbeier (Hul-F).

Fig. 3. Variation of the Simpson´s index of diversity between different sites or grazing treatments. Each histogram is the mean ± 1 SE of n=6. Equal letters above histograms indicate not significant differences at p<0.05.
Fig. 3. Variación del índice de diversidad de Simpson entre diferentes sitios o tratamientos de pastoreo. Cada histograma es el promedio ± 1 EE de n=6. Letras iguales sobre los histogramas indican no diferencias signifcativas a p<0,05.

Proportion of aboveground- and belowground carbon. In Horqin, grazing significantly decreased (p<0.05) aboveground carbon, while it had no effect (p>0.05) on belowground carbon (Table 3). Aboveground carbon in the grazed plots was 36% that in the fenced plots (Table 3). Litter carbon was also lower (p<0.05) in the grazed than fenced plots (Table 3). A similar pattern was found in the Hulunbeier region, where aboveground carbon on the grazed plots was 38% of that in the fenced grasslands. Furthermore, the proportion of BGC in Hul-G reached 75.1% (p>0.05) of that in Hul-F (Table 3). Once again, LC was higher (p<0.05) in the fenced than grazed plots (Table 3).

Table 3. The contribution of each vegetation component (ABC= aboveground carbon, BGC= belowground carbon, LC= litter carbon) and the soil (SOC= soil organic carbon) to carbon storage (g C/m²) on fenced or grazed grasslands at the Horqin or Hulunbeier sites in northern China. Hor-F: Horqin-fenced; Hor-G: Horqin-grazed. Each value is the mean ± 1 SE of n=6. Within each column and study site, different letters indicate significant differences between grazing treatments at p<0.05. Percentage refers to the contribution of each vegetation component and the soil to total carbon storage.
Tabla 3. Contribución de cada componente de la vegetación (ABC= carbono en la parte aérea; BGC = carbono en la parte subterránea; LC= carbono en la broza) y el suelo (SOC= carbono orgánico del suelo) al almacenamiento de carbono (g C/m²) en áreas de vegetación natural pastoreadas o clausuradas al pastoreo en Horqin o Hulunbeier en el norte de China. Hor-F: clausura en Horqin; Hor-G: pastoreo en Horqin. Cada valor es el promedio ± 1 EE de n=6. Dentro de cada columna y sitio de estudio, letras diferentes indican diferencias significativas entre tratamientos de pastoreo a p<0,05. El porcentaje se refiere a la contribución de cada componente de la vegetación y el suelo al almacenamiento total de carbono.

While the percentage of aboveground carbon was greater (p<0.05) at the grazed than at the fenced area in Horqin, no differences (p>0.05) were found in that percentage between the grazed and fenced areas in Hulunbeier (Fig. 4).

Fig. 4. Proportion of AGC and BGC in the grazed and fenced grasslands at each of the two study sites. "Hor_F" represents fenced grasslands in Horqin; "Hor_G" represents grazed grasslands in Horqin; "Hul_F" represents fenced grasslands in Hulunbeier, and "Hul_G" represents grazed grasslands in Hulunbeier. Different capital letters within each site represent significant differences (p<0.05) in aboveground carbon. Histograms are the mean ± 1 SE of n=6.
Fig. 4. Proporción de AGC y BGC en los pastizales pastoreados o clausurados al pastoreo en cada uno de los dos sitios de estudio. "Hor_F" representa pastizales alambrados en Horqin; "Hor_G" representa pastizales pastoreados en Horqin; "Hul_F" representa pastizales alambrados en Hulunbeier, y "Hul_G" representa pastizales pastoreados en Hulunbeier. Diferentes letras mayúsculas dentro de cada sitio representan diferencias significativas (p<0,05) en el carbono de la parte aérea. Los histogramas son el promedio ± 1 E.E. de n=6.

Relationship between AG- and BGC at the study regions. At the Horqin site, there was a significant decrease in BGC as AGC increased at both the grazed (p=0.092) and the fenced (p=0.005) sites (Fig. 5). In the Hulunbeier region, however, even though it appeared to be a positive tendency at both study grazing treatments, the relationship was only signifcant at p=0.073 in the fenced area (Fig. 5).

Fig. 5. The relationship of AGC/BGC at the either grazed (HorG) or fenced (Hor-F) grassland in Horqin or the grazed (Hul-G) of fenced (Hul-F) grassland in Hulunbeier. Each symbol represents individual values within 1 x1 m subplots distributed at random within a 500 x 500 m area. Half of this area was grazed while the other half remained ungrazed.
Fig. 5. Relación AGC/BGC en pastizales pastoreados (Hor-G) o alambrados (Hor-F) en Horqin o pastoreados (Hul-G) o alambrados (Hul-F) en Hulunbeier. Cada símbolo representa valores individuales dentro de subparcelas de 1 x 1 m distribuídas al azar dentro de un área de 500 x 500 m. La mitad de esta área fue pastoreada mientras que la otra mitad permaneció sin pastorear.

Soil organic carbon (SOC). Tere were no significant differences (p>0.05) in SOC storage in the whole soil profile between grazed and fenced grasslands in either the Horqin or Hulunbeier grassland (Table 3). Further analysis showed that in three of six paired plots (plot 3, plot 4 and plot 6) of Horqin region (Fig. 6), soil organic C storage was significantly lower (p<0.05) in the grazed (by 32-95%) than in the fenced plots. Such difference along the soil profile, however, was not observed in the Hulunbeier region (Fig. 7).

Fig. 6. SOC distribution at different soil depths at either grazed or fenced grasslands in Horqin region. Each symbol is the mean ± 1 SE of n=5 subplots within each study plot. Notice the change of scale in Plot 4.
Fig. 6. Distribución del SOC a diferentes profundidades del suelo en pastizales pastoreados o alambrados en Horqin. Cada símbolo es el promedio ± 1 E.E. de n=5 subparcelas dentro de cada parcela estudiada. Note el cambio de escala en la Parcela 4.

Fig. 7. SOC distribution at different soil depths at either grazed or fenced grasslands in Hulunbeier region. Each symbol is the mean ± 1 SE of n=5 subplots within each study plot. Notice the change of scale in the study plots.
Fig. 7. Distribución del SOC a diferentes profundidades del suelo en pastizales pastoreados o alambrados en Hulunbeier. Cada símbolo es el promedio ± 1 E.E. de n=5 subparcelas dentro de cada parcela estudiada. Note el cambio de escala en las parcelas estudiadas.

Soil at 20-40 cm soil depth contributed the highest carbon storage most of the times among the different soil depths at both study regions and grazing treatments (Figs. 6 and 7). In the Horquin region, the 20-40 cm soil depth produced 70.86% and 71.39% of the total soil carbon storage at the fenced and grazed grasslands, respectively. Similarly, in the Hulunbeier region, that soil depth produced 72.23% and 75.35% of the total soil carbon storage at the fenced and grazed grasslands, respectively (Figs. 6 and 7).

The contribution of soil to the total carbon pool. The greatest (p<0.05) carbon pool in both study regions was in the soil, where it ranged from 98.2 to 99.08% at the Horqin and Hulunbeier study sites (Table 3). In the Horqin region, soil contributed 8672.09 g C/m² which represented approximately 98.78% of the total carbon in fenced grasslands (Table 3). At the same site, AGC + LC contributed 0.83% which was twice that of the BGC (Table 3). Compared with measurements at the fenced grasslands, AGC + LC in the grazed grasslands contributed a relatively lower percentage (0.49% in Horqin and 0.62% in Hulunbeier, respectively), while BGC contributed 0.43% at the Horqin, and 0.38% at the Hulunbeier study sites, respectively.

DISCUSSION

Effects of grazing on community composition. Grazing can afect the abundance, richness, and spatial heterogeneity of vegetation (Hilbert et al., 1981; Adler et al., 2001; Fuhlendorf & Engle, 2004). However, the effects of grazing at different regions may depend on its intensity and history (Austrheim & Eriksson, 2001; Díaz et al., 2001). Some studies have indicated that heavy grazing caused degradation of the alpine meadows (Zhou et al., 2005). Meanwhile, other studies have shown that moderate to heavy grazing by a combination of herbivore species resulted in similar plant communities (Allred et al., 2012; Wang et al., 2012). In our study, the plant species of the study regions showed indeed differences by DCA. However, the plant communities were similar between grazed and fenced grasslands within the same region; species diversity, expressed by the Simpson index, also showed no significant (p<0.05) differences between grazed versus fenced grasslands in the same region. Dominant species were Setaria viridis (L.) P. Beauv. and Arthraxon hispidus (Tunb.) Makino in Horqin, and Cyperus fuscus L., Agropyron cristatum (L.) Gaertn. and Stipa capillata L. in Hulunbeier. Thus, grazing showed little influence on plant species diversity at the two study locations.

Effects of grazing on BGC and its relationship with AGC. Grazing can stimulate the productivity of vegetation in a wide range of ecosystems (McNaughton, 1985; Hadar et al., 2009; Smith et al., 2012), such as in forests and grasslands (Lohmann et al., 2012). However, the effects of grazing on BGC are rather elusive. Some studies have showed that grazing negatively impacted BGC (Rodriguez et al., 2007). Other findings, however, revealed that BGC increased under grazing as compared to control plants (Huhta et al., 2009). Our results indicated that grazing significantly decreased AGC, while BGC remained similar on grazed versus ungrazed sites in both study regions (Table 3).
Optimal partitioning and isometric allocation are two important hypotheses in biomass allocation of plants (Gedroc et al., 1996; Yang et al., 2009; Chan, 2012). The optimal partitioning hypothesis suggests that plants respond to environmental conditions by allocating biomass among various organs which contributes to capture nutrients, water, and light to maximize their growth rate (Kobe et al., 2010; Nair et al., 2013). In annual species, AGB/BGB partitioning is partially consistent with the optimal partitioning theory, but it might be also highly ontogenetically constrained (Gedroc et at, 1996). By comparison, the isometric allocation hypothesis predicts that the relationship between AGB and BGB is isometric across different plant species (Worley & Harder, 1996; Gray et al., 2006).
In our study, as AGC increased, BGC decreased at the Horqin region, which is consistent with the theory of optimal partitioning (Fig. 5). If a plant allocates more biomass belowground to acquire more moisture and/or nutrients, it may decline its resource partitioning to aboveground plant parts. Meanwhile, if a plant develops more biomass aboveground to contribute capturing of more light, resource partitioning to roots may decline. On the contrary, the AG- and the BGC showed a positive relationship in the Hulunbeier region. Different community composition may cause different growth strategies to generate different AG- and BGC relationships at both study sites. In our study, plant species showed a greater difference at the Hulunbeier than at the Horqin region (Fig.2). In addition, different thermal conditions may cause different plant growth levels (Lambers et al., 1998). Horqin (42° 56' - 43° 20' N, 119° 29' E - 119° 41' E) and Hulunbeier (47° 20' - 50° 12' N, 118° 31' E - 121° 03' E) lied along a similar longitude but a different latitude, indicating a similar hydrological constraint but different thermal conditions. The mean annual open-pan evaporation was 1935 mm in Horqin, and 1371 mm in Hulunbeier. A lower evaporation in Hulunbeier grasslands could maintain higher AGC values when BGC increased. Thus, differences in plant species composition and mean annual evaporation could help explain why plant and soil variables did not show a consistent response to grazing in both the Horqin and the Hulunbeier regions.

Effects of grazing on SOC. In our study, grazing exclusion for 8-10 years did not change the total mass of C in the plant-soil system compared to values obtained under grazing conditions both at the Horqin and Hulunbeier regions. However, it did change the distribution of C among the plant components in those systems. This result is consistent with many studies that soils and ecosystems will not accumulate sufficient C when an area is fenced to herbivory grazing for a short time period (Schuman et al., 1999). Other studies have found that after a 20-year grazing exclusion, grasslands in northern China were stable in their productivity (Kawamura et al, 2005; Pei et al, 2008), spatial pattern of soil properties (Su et al, 2006) and C storage (Wu et al, 2008; He et al, 2008).
Our study also demonstrated that soil organic C was significantly lower in grazed than fenced plots (by 32-95%) in three out of six paired plots in Horqin region. At the same time, no differences were observed along the soil profile between fenced and grazed plots in the Hulunbeier region (Fig. 6 and Fig. 7). Also, soil carbon in the topsoil (0-20 cm) approached 1000 g C/m² in Horqin, and 1500 g C/m² in Hulunbeier. Several authors reported that variations in SOC levels can be due to climate, soil or vegetation type differences (Jonasson et al., 1986; Dec et al., 2010; Crowther & Bear, 2012). Variations in SOC could also be due to differences in grazing history or in the history of grazing intensity before and after grazing exclusion (Eckstut et al., 2011). A major problem in our study was the community heterogeneity between the two study regions. In Horqin, habitats were more fragile than those in Hulunbeier, with higher evaporation and less vegetation, which contributed to make a greater difference on carbon storage along the various study soil depths. The soil, however, was the largest "carbon pool" in the study grassland ecosystems, where it contributed with more than 98% of the total carbon. This is consistent with previous studies (Conant et al, 2001; Chen et al, 2012). Fenced grasslands in both Horqin and Hulunbeier steppes provided twice as much litter carbon as grazed ones, which would be transformed into soil organic matter and make the soil more productive.

MANAGEMENT IMPLICATIONS

The rangeland management profession has clearly advanced the natural resource conservation worldwide (James et al, 2013). Numerous experimental studies have demonstrated that grazing intensity has been the most important principle of grazing management (Hickman et al, 2004; Briske et al, 2008). However, the potential contribution to it from different regions has largely been ignored. While grazing decreased the proportion of aboveground carbon, it showed little influence on plant species diversity in both the Horqin and Hulunbeier regions. Therefore, for the restoration of plant species diversity at the study sites, it appears unnecessary to take into account the relationship between grazing practice and regional disparity.
Relationships between precipitation versus soil nutrient or regional disparity in relevant decision-making approaches have yielded different for rangeland management (Derner & Schuman, 2007; Gillingham et al, 2012). Our results showed that as aboveground carbon increased, belowground carbon decreased in Horqin, while positive correlations between the two plant parts were found in the Hulunbeier region. This suggests that plant photosynthesis might be enhanced in Hulunbeier to produce more aboveground carbon, while more soil nutrients should be provided for rangeland restoration in Horqin. Management plans for conservation and restoration of grasslands under grazing should therefore focus on the specifc site conditions, and take several abiotic and biotic factors into account.

ACKNOWLEDGEMENTS

We thank Mr. Q.-L. Zhou, Mrs. X.-H. Li, Mr. Y.-M. Luo and Mr. C.-Y. Chao for fieldwork assistance. This research was financially supported by the National 973 Project (2013CB429903). C.A. Busso thanks (1) the sabbatical leave given by Universidad Nacional del Sur and the Consejo Nacional de Investigaciones Científicas y Técnicas de la República Argentina (CONICET), (2) the Associateship awarded by the Third World Academy of Sciences (TWAS)-UNESCO, and (3) housing, facilities and financial support from the Institute of Applied Ecology, Chinese Academy of Sciences, Shenyang, China.

REFERENCES

1. Adler, P., D. Raff & W. Lauenroth (2001). The effect of grazing on the spatial heterogeneity of vegetation. Oecologia 128: 465-479.         [ Links ]

2. Allred, B.W., S.D. Fuhlendorf, F.E. Smeins & C.A. Taylor (2012). Herbivore species and grazing intensity regulate community composition and an encroaching woody plant in semi-arid rangeland. Basic and Applied Ecology 13: 149-158.         [ Links ]

3. Amthor, J.S. (2006). Terrestrial higher‐plant response to increasing atmospheric [CO₂] in relation to the global carbon cycle. Global Change Biology 1: 243-274.         [ Links ]

4. Austrheim, G. & O. Eriksson (2001). Plant species diversity and grazing in the Scandinavian mountains‐patterns and processes at different spatial scales. Ecography 24: 683-695.         [ Links ]

5. Baron, V., E. Mapfumo, A. Dick, M. Naeth, E. Okine & D. Chanasyk (2002). Grazing intensity impacts on pasture carbon and nitrogen fow. Journal of Range Management 55: 535-541.         [ Links ]

6. Briske, D.D., J. Derner, J. Brown, S. Fuhlendorf, W. Teague, K. Havstad, R. Gillen, A. J. Ash & W. Willms (2008). Rotational grazing on rangelands: reconciliation of perception and experimental evidence. Rangeland Ecology and Management 61: 3-17.         [ Links ]

7. Brown, R.W. (1995). The water relations of range plants: Adaptations to water deficit. In: Bedunah, D.J., Sosebee, R.E. (eds.). p. 291-413. Wildland Plants. Physiological Ecology and Developmental Morphology. Society for Range Management, Denver, CO. 710 p.         [ Links ]

8. Bultynck, L. & H. Lambers (2004). Effects of applied gibberellic acid and paclobutrazol on leaf expansion and biomass allocation in two Aegilops species with contrasting leaf elongation rates. Physiologia Plantarum 122: 143-151.         [ Links ]

9. Chan, T.M. (2012). Optimal partition trees. Discrete and Computational Geometry: 1-30.         [ Links ]

10. Chen, Y., Y. Li, X. Zhao, T. Awada, W. Shang & J. Han (2012). Effects of Grazing Exclusion on Soil Properties and on Ecosystem Carbon and Nitrogen Storage in a Sandy Rangeland of Inner Mongolia, Northern China. Environmental Management 50: 622-632.         [ Links ]

11. Conant, R.T., K. Paustian & E.T. Elliott (2001). Grassland management and conversion into grassland: efects on soil carbon. Ecological Applications 11: 343-355.         [ Links ]

12. Cox, P.M., R.A. Betts, C.D. Jones, S.A. Spall & I.J. Totterdell (2000). Acceleration of global warming due to carbon-cycle feedbacks in a coupled climate model. Nature 408: 184-187.         [ Links ]

13. Crowther, T.W. & A.D. A'Bear (2012). Impacts of grazing soil fauna on decomposer fungi are species-specific and density-dependent. Fungal Ecology 5: 277-281.         [ Links ]

14. Crutzen, P.J. & M.O. Andreae (1990). Biomass burning in the tropics: Impact on atmospheric chemistry and biogeochemical cycles. Science 250: 1669-1678.         [ Links ]

15. Dec, D., J. Dörner, O. Balocchi, R. Gilkes & N. Prakongkep (2010). Spatial and temporal changes of soil physical properties of an Andisol in southern Chile as a consequence of grazing and wetting and drying cycles. 19th World Congress of Soil Science, Soil Solutions for a Changing World. Available online: http://www.iuss. org/19th% 20WCSS/symposium/pdf/2235. pdf[Links].         [ Links ]

16. Derner, J. & G. Schuman (2007). Carbon sequestration and rangelands: a synthesis of land management and precipitation effects. Journal of Soil and Water Conservation 62: 77-85.         [ Links ]

17. Díaz, S., I. Noy‐Meir & M. Cabido (2001). Can grazing response of herbaceous plants be predicted from simple vegetative traits? Journal of Applied Ecology 38: 497-508.         [ Links ]

18. Eckstut, M.E., C.D. McMahan, B.I. Crother, J.M. Ancheta, D.A. McLennan & D.R. Brooks (2011). PACT in practice: comparative historical biogeographic patterns and species-area relationships of the Greater Antillean and Hawaiian Island terrestrial biotas. Global Ecology and Biogeography 20: 545-557.         [ Links ]

19. Feng, X., B. Fu, N. Lu, Y. Zeng & B. Wu (2013). How ecological restoration alters ecosystem services: an analysis of carbon sequestration in China's Loess Plateau. Scientific Reports 3.         [ Links ]

20. Flanagan, L.B., E.J. Sharp & M.G. Letts (2013). Response of plant biomass and soil respiration to experimental warming and precipitation manipulation in a Northern Great Plains grassland. Agricultural and Forest Meteorology 173: 40-52.         [ Links ]

21. Fuhlendorf, S. & D. Engle (2004). Application of the fire-grazing interaction to restore a shifting mosaic on tallgrass prairie. Journal of Applied Ecology 41: 604-614.         [ Links ]

22. Gedroc, J., K. McConnaughay & J. Coleman (1996). Plasticity in root/shoot partitioning: optimal, ontogenetic, or both? Functional Ecology 10: 44-50.         [ Links ]

23. Gillingham, A., S. Ledgard, S. Saggar, K. Cameron, H. Di, C. De Klein & M. Aspin (2012). Initial evaluation of the effects of dicyandiamide (DCD) on nitrous oxide emissions, nitrate leaching and dry matter production from dairy pastures in a range of locations within New Zealand. In: L.D. Currie and C.L. Christensen (Eds.). Advanced nutrient management: gains from the pastgoals for the future. http://frc.massey.ac.nz/publications.html. Occasional Report.         [ Links ]

24. Gray, W.D., C.R. Sims, W.T. Fu & M.J. Schoelles (2006). The soft constraint hypothesis: a rational analysis approach to resource allocation for interactive behavior. Psychological Review 113: 461.         [ Links ]

25. Hadar, L., I. Noy‐Meir & A. Perevolotsky (2009). The effect of shrub clearing and grazing on the composition of a Mediterranean plant community: functional groups versus species. Journal of Vegetation Science 10: 673-682.         [ Links ]

26. Han, G., X. Hao, M. Zhao, M. Wang, B. H. Ellert, W. Willms & M. Wang (2008). Effect of grazing intensity on carbon and nitrogen in soil and vegetation in a meadow steppe in Inner Mongolia. Agriculture, Ecosystems and Environment 125: 21-32.         [ Links ]

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

28. Hickman, K.R., D.C. Hartnett, R.C. Cochran & C.E. Owensby (2004). Grazing management effects on plant species diversity in tallgrass prairie. Rangeland Ecology and Management 57: 58-65.         [ Links ]

29. Hilbert, D., D. Swift, J. Detling & M. Dyer (1981). Relative growth rates and the grazing optimization hypothesis. Oecologia 51: 14-18.         [ Links ]

30. Huhta, A., P. Rautio, K. Hellström, M. Saari & J. Tuomi (2009). Tolerance of a perennial herb, Pimpinella saxifraga, to simulated fower herbivory and grazing: immediate repair of injury or postponed reproduction? Plant Ecology 201: 599-609.         [ Links ]

31. James, J.J., C.S. Boyd & Svejcar (2013). Seed and Seedling Ecology Research to Enhance Restoration Outcomes. Rangeland Ecology and Management 66: 115-116.         [ Links ]

32. Jonasson, S., J.P. Bryant, F.S. Chapin III & M. Andersson (1986). Plant phenols and nutrients in relation to variations in climate and rodent grazing. American Naturalist 394-408.         [ Links ]

33. Karl, T.R. & K.E. Trenberth (2003). Modern global climate change. Science 302: 1719-1723.         [ Links ]

34. Kawamura, K., T. Akiyama, H.-o. Yokota, M. Tsutsumi, T. Yasuda, O. Watanabe & S. Wang (2005). Quantifying grazing intensities using geographic information systems and satellite remote sensing in the Xilingol steppe region, Inner Mongolia, China. Agriculture, Ecosystems and Environment 107: 83-93.         [ Links ]

35. Klaus, V.H., T. Kleinebecker, S. Boch, J. Müller, S.A. Socher, D. Prati, M. Fischer & N. Hölzel (2012). NIRS meets Ellenberg's indicator values: Prediction of moisture and nitrogen values of agricultural grassland vegetation by means of near-infrared spectral characteristics. Ecological Indicators 14: 82-86.         [ Links ]

36. Kobe, R.K., M. Iyer & M.B. Walters (2010). Optimal partitioning theory revisited: Nonstructural carbohydrates dominate root mass responses to nitrogen. Ecology 91: 166-179.         [ Links ]

37. Lambers, H., F.S. Chapin III & T.L. Pons (1998). Plant Physiological Ecology. Springer-Verlag. New York, New York. 538 p.         [ Links ]

38. Li, Y., X. Zhou, J.R. Brandle, T. Zhang, Y. Chen & J. Han (2012). Temporal progress in improving carbon and nitrogen storage by grazing exclosure practice in a degraded land area of China's Horqin Sandy Grassland. Agriculture, Ecosystems and Environment 159: 55-61.         [ Links ]

39. Liu, Z., Q. Yan, X. Li, J. Ma & X. Ling (2007). Seed mass and shape, germination and plant abundance in a desertified grassland in northeastern Inner Mongolia, China. Journal of Arid Environments 69: 198-211.         [ Links ]

40. Lohmann, D., B. Tietjen, N. Blaum, D.F. Joubert & F. Jeltsch (2012). Shifting thresholds and changing degradation patterns: climate change effects on the simulated long‐term response of a semiarid savanna to grazing. Journal of Applied Ecology 49: 814-823.         [ Links ].

41. Manley, J., G. Schuman, J. Reeder & Hart (1995). Rangeland soil carbon and nitrogen responses to grazing. Journal of Soil and Water Conservation 50: 294-298.         [ Links ]

42. McNaughton, S. (1985). Ecology of a grazing ecosystem: the Serengeti. Ecological Monographs 55: 259-294.         [ Links ]

43. Medina-Roldán, E., J. Paz-Ferreiro & R.D. Bardgett (2012). Grazing exclusion affects soil and plant communities, but has no impact on soil carbon storage in an upland grassland. Agriculture, Ecosystems and Environment 149: 118-123.         [ Links ]

44. Nair, S., S. Maas, C. Wang & S. Mauget (2013). Optimal Field Partitioning for Center-Pivot-Irrigated Cotton in the Texas High Plains. Agronomy Journal 105: 124-133.         [ Links ]

45. Nelson, D.W., L.E. Sommers, D. Sparks, A. Page, P. Helmke, R. Loeppert, P. Soltanpour, M. Tabatabai, C. Johnston & M. Sumner (1996). Total carbon, organic carbon, and organic matter. Methods of soil analysis. Part 3-chemical methods. pp. 961-1010.         [ Links ]

46. Netter, J., W. Wasserman & M.H. Kuther (1985). Applied linear statistical models: Regression, analysis of variance, and experimental designs. Irwin RD, Homewood, ILL. 1127 p.         [ Links ]

47. Ni, J. (2002). Carbon storage in grasslands of China. Journal of Arid Environments 50: 205-218.         [ Links ]

48. Pei, S., H. Fu & C. Wan (2008). Changes in soil properties and vegetation following exclosure and grazing in degraded Alxa desert steppe of Inner Mongolia, China. Agriculture, Ecosystems and Environment 124: 33-39.         [ Links ]

49. Rodriguez, M., M. Bertiller & C. Sain (2007). Spatial patterns and chemical characteristics of root biomass in ecosystems of the Patagonian Monte disturbed by grazing. Journal of Arid Environments 70: 137-151.         [ Links ]

50. Schuman, G., J. Reeder, J. Manley, R. Hart & W. Manley (1999). Impact of grazing management on the carbon and nitrogen balance of a mixed-grass rangeland. Ecological Applications 9: 65-71.         [ Links ]

51. Scurlock, J. & D. Hall (1998). The global carbon sink: a grassland perspective. Global Change Biology 4: 229-233.         [ Links ]

52. Silveira, M.L., K. Liu, L.E. Sollenberger, R.F. Follett & J. Vendramini (2013). Short-term effects of grazing intensity and nitrogen fertilization on soil organic carbon pools under perennial grass pastures in the southeastern USA. Soil Biology and Biochemistry 58: 42-49.         [ Links ]

53. Simpson, E.H. (1949). Measurement of diversity. Nature 163: 688.         [ Links ]

54. Smith, M.D., A. Knapp, S.L. Collins, N. Govender, K. Kirkman, R.W.S. Fynn, D.E. Burkepile, N. Hagenah, K. Matchett & D. Thompson (2012). Divergence in savanna grassland community responses to fire and grazing in North America and South Africa.         [ Links ]

55. Steel, R.G.D. & J.H. Torrie (1960). Principles and Procedures of Statistics. (With special Reference to the Biological Sciences.) Toronto: McGraw-Hill Book Company.         [ Links ]

56. Su, Y.Z., Y.L. Li & H.L. Zhao (2006). Soil properties and their spatial pattern in a degraded sandy grassland under post-grazing restoration, Inner Mongolia, northern China. Biogeochemistry 79: 297-314.         [ Links ]

57. Tang, C., Y. Quian, Y. Kang, G. Cui, H. Lu, L. Shu, Y. Wang & L. Tang (2004). Study on the ecological distribution of alveolar Echinococcus in Hulunbeier Pasture of Inner Mongolia, China. Parasitology 128: 187-194.         [ Links ]

58. Tian, G., T. Granato, A. Cox, R. Pietz, C. Carlson & Z. Abedin (2009). Soil carbon sequestration resulting from long-term application of biosolids for land reclamation. Journal of Environmental Quality 38: 61-74.         [ Links ]

59. Wang, S., J. Duan, G. Xu, Y. Wang, Z. Zhang, Y. Rui, C. Luo, B. Xu, X. Zhu & X. Chang (2012). Effects of warming and grazing on soil N availability, species composition, and ANPP in an alpine meadow. Ecology 93: 2365-2376.         [ Links ]

60. Wiesmeier, M., M. Steffens, A. Kölbl & I. Kögel-Knabner (2012). Soil degradation in semi-arid grasslands due to intensive grazing in Northern China. EGU General Assembly Conference Abstracts. 270 P.         [ Links ]

61. Williams, T.E. & H.K. Baker (1957). Studies on the root development of herbage plants. I. Techniques of herbage root investigations. Journal of the British Grassland Society 12: 49-55.         [ Links ]

62. Worley, A.C. & L.D. Harder (1996). Size-dependent resource allocation and costs of reproduction in Pinguicula vulgaris (Lentibulariaceae). Journal of Ecology 84: 195-206.         [ Links ]

63. Wu, L., N. He, Y. Wang & X. Han (2008). Storage and Dynamics of Carbon and Nitrogen in Soil after Grazing Exclusion in Grasslands of Northern China. Journal of Environmental Quality 37: 663-668.         [ Links ]

64. Yang, Y., J. Fang, C. Ji & W. Han (2009). Above‐and belowground biomass allocation in Tibetan grasslands. Journal of Vegetation Science 20: 177-184.         [ Links ]

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

66. Zhao, L., Y. Li, S. Xu, H. Zhou, S. Gu, G. Yu & X. Zhao (2006). Diurnal, seasonal and annual variation in net ecosystem CO₂ exchange of an alpine shrubland on Qinghai‐Tibetan plateau. Global Change Biology 12: 1940-1953.         [ Links ]

67. Zhou, H., X. Zhao, Y. Tang, S. Gu & L. Zhou (2005). Alpine grassland degradation and its control in the source region of the Yangtze and Yellow Rivers, China. Grassland Science 51: 191-203.         [ Links ]