Introduction

Soils are the largest carbon pool in terrestrial ecosystems, containing more than 1,500 Pg C (Raich and Schlesinger 1992; Eswaran et al. 1993). Grasslands, one of the terrestrial ecosystems, account for 40 % of the total land area in China. However, by the end of the 20th century, 90 % of this grassland experienced degradation as a consequence of a rapid expansion of livestock numbers and economic reforms initiated in the 1980s (Wang et al. 2004). As a result of reclamation, overgrazing of grasslands, wind erosion, and desertification, the soil carbon of China has become a carbon source (Li 2000). Soil organic material [SOM, which is 1.724 times the soil organic carbon (SOC)] plays a major role in nutrient cycling and soil quality, and, ultimately, has critical implications for the management of ecosystems and physical stability of soil erosion resistance (Su and Zhao 2003; Zhao et al. 2006).

Lowery et al. (1995) and López (1998) noted that soil coarseness is caused by selective erosion on soil clays and very fine sands. Even if supplemental carbon is added, soil mineralization becomes very weak with the occurrence of desertification because the soil substrate on which soil microorganisms depend has deteriorated (Su et al. 2004). As a result, the soil becomes susceptible to wind erosion (Arianoutsou-Faraggitaki 1985). Serious wind erosion leads to loss of surface soil, exposing the calcium carbonate-rich subsurface to the air. As the carbon oxidizes, CO2 can be released (Lal 2003).

Studies by McPherson (1995) and Wondzell and Ludwing (1995) indicate that livestock overgrazing and increasing aridity are the major causes of grassland desertification, although the complex interactions of many factors may contribute to desertification (Humphrey 1958; Reynolds et al. 1999). Desertification not only results in soil degradation and severe losses to land productivity (Gad and Abdel-Samie 2000), but fosters atmospheric emissions of soil C and N as greenhouse gases (Tan and Lal 2005). Since SOM can be associated with different soil chemical, physical, and biological processes, it is considered one of the most important indicators of environmental quality. Soil C and N have been used to assess the effects that different management options have on soil quality (Silveira et al. 2009). The C/N ratio can be used not only as a sensitive index, but can also affect the cycle of SOC and N (Ren et al. 2006). Soil organic material serves as an important storehouse of nutrients, drives the nutrient cycle, maintains soil structural stability, aids the infiltration of air and water, promotes water retention, and reduces soil erosion (Gregorich et al. 1994). For these reasons, the effect of desertification on soil C and N have become a concern in recent years (Johnson and Curtis 2001; Breuer et al. 2006; Qi et al. 2008). The maintenance of adequate levels of SOM should be an integral component of soil management strategy, because SOM is essential to ecosystem productivity and regeneration (Chan et al. 2002).

During the recent decades, high altitude soil has attracted more attention in the debate on the potential impact of environmental change on the global C cycle (Beniston et al. 1997; Diaz and Bradley 1997; Goulden et al. 1998; Christensen et al. 1999; Oechel et al. 2000). With an average elevation above 4,000 m, the Qinghai–Tibet Plateau is called the roof of the world. Wang et al. (2002) and Zeng et al. (2003) have shown that high altitude soil in alpine meadow (AM) ecosystems on the Qinghai–Tibet Plateau has exhibited significant degradation due to natural and human disturbance factors. Decreased vegetation has led to a lower supply of organic material and a reduction in SOM, decreased infiltration capacity, increased runoff, and accelerated soil erosion (Gamougoun et al. 1984; Evans 1998; Trimble and Mendel 1995).

The current study examined spatial and temporal variability in SOC and other related soil properties near Maduo County (33°50′–35°40′N, 96°50′–99°20′E) on the Qinghai–Tibet Plateau, and Maduo County (Fig. 1) is the source region of the Yellow River in the northeastern region of the Qinghai–Tibet Plateau, where desertification has restricted the development of livestock husbandry. The objectives of this study were to (1) evaluate differences in soil properties within and among sites that transition from hilltops to level areas, and (2) determine the effects of desertification on the physical and chemical properties of soil. This work studied the changes to SOC, total nitrogen (TN), pH, and soil bulk density (SBD) from the upslope areas to level areas at the bottom of slopes due to wind erosion, and documented the relationship between soil properties and desertification of AM. The working hypothesis is that, as desertification advances, SOC and TN will decrease, and pH and SBD will increase.

Fig. 1
figure 1

Study area

Full size image

Materials and methods

Study area

The area of Maduo County is about 26,267 km2, and its elevation ranges from approximately 4,200 to 4,800 m above sea level. The region is characterized by low mountains, numerous lakes, and wide valleys. A warm season occurs from June to September and a long cold season runs from October to May. The annual average air temperature is −4.1 °C and there is no absolute frost-free season. The annual precipitation is about 303.9 mm, with 86 % of the precipitation mainly occurring from May to September, in the form of snow or heavy rain. The annual potential evaporation is about 1,264 mm, and the maximum wind speed is 26 m/s. Vegetation consists largely of high-cold meadow and high-cold steppe species.

Experimental design

An experimental plot was selected on which livestock grazing occurred during the cold season (from October to May). Different sub-plots were selected at the top, middle, and bottom of an approximately 30° slope, and a level area at the base of the slope with altitudes of 4,320, 4,275, 4,238 and 4,227 m, respectively. Land desertification was classified based on the criteria developed by Xue et al. (2009) as shown in Table 1: lightly desertified land (LDL), moderately desertified land (MDL), seriously desertified land (SDL), and very seriously desertified land (VSDL). In addition, a non-degraded alpine meadow was also selected as a control (Xue et al. 2009).

Table 1 Grassland grading for different desertification levels in Maduo County in Qinghai Province, China
Full size table

For each land classification, soil samples were collected from each plot in 10 cm increments, representing 0–10, 10–20, and 20–30 cm depths. Each sample was oven-dried at 105 °C for over 24 h, weighed on an electronic balance (±0.01 g), and analyzed for TN using the Walkley–Black method (Nelson and Sommers 1982). The Walkley–Black method (Allison 1965) was also used to determine SOC, except that a 0.33 M solution of K2Cr2O7 was used instead of a 0.17 M solution, enabling a larger subsample to be used in the analyses. SBD and pH were also measured for each sample. Soil pH was determined in a suspension of soil and water (soil:water ratio was 1:5) using a glass electrode. The results of each sample from all three depths were averaged. Reported results represent averages of three replicates from each site.

Statistical analysis

The calculation of C and storage in the soil was based on the method of Li et al. (2005a, b) as follows: SOC or TN in soil = (soil area) × (soil depth) × (average SBD) × (average SOC or TN content). All statistical analyses were performed using SPSS 10.0 (SPSS for Windows, Version 10.0, Chicago, IL, USA).

Results

Soil properties

In the 0–30 cm depth range, SOC was the highest in the AM soils (26.61 g/kg), followed by LDL (25.65 g/kg), MDL (25.08 g/kg), SDL (15.51 g/kg) and VSDL soils (10.90 g/kg) (Fig. 2).

Fig. 2
figure 2

SOC in different desertified lands

Full size image

Soil organic carbon generally decreased as desertification increased upslope and downslope; the decrease in SOC was most pronounced at the bottom of the slope. SOC generally increased as desertification increased from AM to MDL, then decreased from MDL to VSDL (Fig. 2). As desertification increased, SOC decreased significantly (p < 0.05) (Table 2).

Table 2 T test among soil properties in the different desertified lands
Full size table

In the 0–30 cm depth range, TN was highest in the AM soils (2.641 g/kg), followed by LDL (2.54 g/kg), MDL (2.49 g/kg), SDL (1.66 g/kg) and VSDL soils (1.29 g/kg).

Total nitrogen decreased with increasing desertification at the upper and lower portions of the slope, with this tendency being more pronounced at the bottom. TN increased with increasing desertification from AM to MDL with this tendency being more pronounced in the level areas. TN decreased from MDL to VSDL (Fig. 3).

Fig. 3
figure 3

TN levels in different desertified lands

Full size image

The pH levels (Fig. 4) were the lowest in the MDL soils (8.01), followed by SDL (8.05), AM (8.09), LDL (8.09), and VSDL soils (8.26).

Fig. 4
figure 4

pH levels in different desertified lands

Full size image

With increasing desertification, pH may increase or decrease from AM and LDL to MDL, but it always increases from MDL and SDL to VSDL regardless the position on the slope. As desertification proceeds from AM to VSDL, pH tends to increase. For example, when compared with AM soils, the pH in VSDL increased 2.80, 0.25, 2.20 and 3.16 %, respectively at the top, middle, bottom, and level areas, with the pH values being highest at the bottom, followed by level, middle, and top areas (Fig. 4).

As shown in Fig. 5, SBD increased significantly with increased desertification. The mean SBD increased in downslope samples and increased from June to September, SBD was the lowest in the AM soils (1.03 g/cm3), followed by LDL (1.07 g/cm3), MDL (1.14 g/cm3), SDL (1.27 g/cm3) and VSDL soils (1.33 g/cm3). Generally, the changes were greater in the level area than in the other positions of slope. SBD increased with increasing desertification regardless of where it was located on the slope (Fig. 5).

Fig. 5
figure 5

SBD in different desertified lands

Full size image

T tests show that samples from AM, LDL, and MDL are not significantly different. Similarly, samples for SDL and VSDL are not significantly different, regardless of SOC, TN, and SBD. There are significant differences between samples from AM, LDL, and MDL and samples for SDL and VSDL for SOC, TN, and SBD. There are no significant differences in pH (Table 2).

C/N ratios

There was positive correlation between SOC and TN at all slope positions (Table 3).

Table 3 Relationship between SOC and TN (n = 16)
Full size table

As desertification increases from AM to VSDL, the C/N ratio (the actual ratios are 10 times of the ratios which are shown in the figures) increases at the top, middle, bottom of an approximately 30° slope, and a level area at the base of the slope. Moreover, the land deteriorates from AM to LDL, the C/N ratio increases slowly. With increasing degradation from MDL to SDL, and from SDL to VSDL, the C/N ratio increases rapidly.

Generally, if the decomposition rate of C is faster than that of N, the C/N ratio will be lower. Not much change in the C/N ratio is evident from AM and LDL to MDL; however, the more pronounced change in the ratio from MDL to SDL and from SDL to VSDL indicates that N losses are greatest in SDL and VSDL (Fig. 6).

Fig. 6
figure 6

C/N ratios in different desertified lands

Full size image

Changes to the coefficient of variation

At all slope positions, the largest coefficient of variation was for SOC followed by TN, SBD, and pH (Table 4).

Table 4 Coefficient of variation for various slope locations during the field study season
Full size table

The largest coefficient of variation for SOC illustrates C losses with increases in desertification and the differences were quite pronounced (p < 0.05) (Table 2). The coefficient of variation for SOC is generally highest in September, except for that in the middle of the slope in August. The coefficient of variation is high for SOC and TN in the middle of the slope and there was no pronounced change for pH and SBD at any position on the slope (Table 4).

Discussion

Effects of desertification on soil properties

Soil chemical and physical properties decline under heavy grazing as desertification progresses from AM to VSDL. In particular, SOC and TN levels decrease and SBD and pH levels increase significantly. This indicates that soil properties are severely affected by desertification, and is consistent with the results of Zhao et al. (2007) and Su and Zhao (2003). Increases in SBD and decreases in SOC enhance the likelihood of wind erosion on soils (López 1998), lowering land productivity as soils become coarser and leaner. This ultimately leads to vegetative degradation and land desertification.

Soil erosion by wind is the main cause for changes in soil particle composition (He et al. 2004); in addition, trampling by animals promotes the loss of top soil (Huang et al. 2007). During the transformation of the soil from AM to VSDL, SOC content generally decreases (Fig. 2). Huang et al. (2007) state that overgrazing affects soil nutrient losses by (1) reducing vegetative growth and exposing the soil surface to erosion which leads to direct soil nutrient losses; and (2) reducing the return of litter to the soil. During the process of desertification from AM to MDL, some positions on the slope, especially the level base areas, show a tendency to accumulate SOC (Fig. 2). Moderate grazing can contribute to this accumulation. Productivity and soil fertility improve as a result of residual litter and livestock excrement being returned to soil (Wang et al. 2008); however, as a desertification progress, SOC decreases are evident. Fu and Chen (2004) and Dormaar et al. (1990) claim that an important manifestation of heavy grazing is the reduction of SOC. The soil environment, including both soil quality and soil stability of AM, gradually degrades (Wang et al. 2008). In addition, while clays can retain SOC, the decreases in SOC were affected by the development of bare surfaces and wind erosion, which reduce the clay content of the soil (Lynch and Cotnoir 1956).

The TN content of the soil also decreases significantly with the development of desertification (Fig. 3). Under heavy grazing, the output of N and inorganic N increases due to frequent livestock feeding, which means more N is transported out of the system by plants (Li et al. 2005a, b). This leads to N decrease in the soil (Fig. 3), similar to results found by Guan et al. (1997). Although the decrease in TN is not pronounced, it is likely that soil properties are weakened by soil buffering capacity (Wang et al. 1998).

Owing to the dominating upward (i.e., toward the soil surface) direction of water movement in the soil profile, soluble salts in the groundwater move upward and accumulate in topsoil. The desert soil in northwestern China is naturally saline (Institute of Soil Science, Academia Sinica 1978), a fact that was confirmed in the present work. Figure 4 shows that pH increases with greater desertification from 7.21 to 8.71. The increase in pH may be due to an increase in the ion circulation rate that the results from the soils receiving a large amount of livestock wastes and/or an increase in salt accumulation caused by the higher evaporation of soil surface water that resulted from decreased vegetative coverage and floor litter (Zhou et al. 2005). The increases in soil pH may affect the bioavailability of nutrients, thus influencing the growth, development, and absorbing functions of plants and resulting in aggravated desertification.

During the study, SBD increased from June to September. Although the experimental plots were subjected to grazing only in winter, lower rainfall may have led to increased soil compaction, which causes SBD to increase. From AM to VSDL, SBD decreased as desertification increased (Fig. 5). The SBD increase is usually accompanied by a reduction in porosity during the process of grassland desertification (Wang et al. 2004). Therefore, SBD can be used as an indicator to monitor soil degradation and to estimate the degree of soil desertification. This phenomenon is similar to that noted by Greenwood et al. (1997) and Rong et al. (2001), who stated that SBD had a cumulative effect.

The C/N ratio of the soil is an indicator for assessing the C and N nutrition balance and is significant in that it reflects the carbon and nitrogen cycling of the soils (Qi et al. 2008). The C/N ratio of soils decreased from 10.14 to 8.36 as desertification increased because the soil carbon declined more rapidly relative to the decrease in nitrogen (Fig. 6). As desertification increases, the decrease in the C/N ratio results in enhanced microbial activity that accelerates the rate of decomposition and mineralization of the SOM and TN, and reduces the capacity of the soils to fix organic carbon. As such, more inorganic N is released from the soil, negatively affecting the environment. Moreover, additional inorganic N is released by the mineralization of organic matter (Li et al. 2006). Inorganic N does not accumulate readily, in part because of leaching and denitrification (Franzluebbers et al. 1996; Vanlauwe et al. 1996; Lupwayi and Haque 1998), which may also contribute to the pollution of groundwater and surface water.

T tests on SOC, TN, and SBD in the sample plots showed significant differences (Table 2) and allow differentiation between the different grazing intensities on AM lands and the state of vegetative growth. The T test can be used as an index to evaluate the health of AM lands (Zhou et al. 2005). Based on the changes to the coefficient of variance (Table 4), the middle portions of the slope are more easily desertified, assuming similar grazing conditions, thus the middle portion of the slope should receive additional attention.

Soil carbon and nitrogen storage

Using satellite images of desertified areas from (Xue et al. 2009), the SOC and TN contents of the soil in Maduo County were calculated for depths of 0–30 cm (Table 5).

Table 5 Areas of desertified land types and alpine meadow in Maduo County (km2)
Full size table

The storage of SOC and N for the different land classifications are shown in Table 6. The soil profile (0–30 cm depth) in this region stored 5.04 × 1011 kg C and 5.07 × 1010 kg N. The AM soil had the highest SOC and N stocks, 1.78 × 1011 kg C and 1.77 × 1010 kg N, respectively, which was more than 20 times that of the VSDL SOC stocks and 1.6 times that of the VSDL N stocks. The results given by Xue et al. (2009) showed that desertification of AM land had expanded from 12,827.6 km2 in 1987 to 12,977.1 km2 in 2006 (Table 5). Determining the total stocks of SOC and TN for each land classification provides an estimate of the stock lost due to desertification (Table 6). The difference in SOC and TN stock noted for AM and VSDL lands indicate that the process of desertification caused the loss of 130.34 g/m2 of SOC and 10.14 g/m2 of TN.

Table 6 SOC and TN storage in desertified land types and alpine meadow in Maduo County (2006) at depth 0–30 cm
Full size table

Large losses of SOC and TN indicate substantial environmental degradation. As C and N are lost from the soil, land productivity deteriorates, and the atmospheric release of greenhouse gases contributes to global climate change (Duan et al. 2001). Saggar et al. (2001) and Su et al. (2002) have indicated that land use/cover and land management are two primary factors influencing soil nutrient content. These results show that land use/cover had significant effects, not only on the soil SOC and TN contents, but also on the amounts of these nutrients lost in the land desertification process. On the one hand, high quality grassland, which not only had higher moisture content and levels of nutrients, but also suffer less from wind erosion. On the other hand, soils subjected to overgrazing and trampling by livestock not only decrease vegetation height and cover, but also destroy the soil crust and intensify wind-induced soil erosion (Zhao et al. 2007). Thus, AM lands begin with better soil quality and nutrient content and have more SOC and TN available for loss through desertification. In addition, SOC and N storage capacity were lost more rapidly during the later stages of desertification than in the initial stages. More than likely, the loss of nutrients led to a decline in vegetative coverage, and desertified land expanded quickly due to overgrazing and wind erosion in the later stages of desertification.

Conclusions

Desertification of AM has increased on the Qinghai–Tibet Plateau during the last century, mainly because of climatic changes and anthropogenic activities. The natural climate effects of the AM on the growth, but unreasonable human activities aggravate the AM to the desertification development.

Through the experimental plot was selected and then different sub-plots were decided at the top, middle, and bottom of an approximately 30° slope, and a level area at the base of the slope. Therefore, land desertification was classified based on the criteria. This study concentrates on SOC, TN, SBD and pH at different desertified alpine meadow. Analysis of the results of this work show that (1) AM desertification in the Qinghai–Tibet plateau is driven by overgrazing and has resulted in significantly increased SBD and pH and significantly decreased SOC and TN; (2) under similar grazing conditions, SOC and TN was more easily lost in the middle portion of a 30 % slope and, accordingly, the middle portion of the slope should be given priority for protection; (3) desertification affected SOC much more than it affected TN, pH, and SBD; (4) the C/N ratio increased and the rate of N decomposition increased as desertification increased, which led to environmental pollution and river eutrophication; and (5) soil carbon storage and nitrogen storage were lost more rapidly during the later stages of desertification than in the initial stages.

Soil bulk density has been shown to be closely related to the degree of soil degradation. This study has shown that soil properties and vegetation succession changes significantly as AM go through a desertification process as a result of overgrazing. AM desertification was found to be accompanied by severe soil erosion, declining soil nutrition, and losses of species diversity. This information can provide a useful baseline to better understand the AM desertification process in Qinghai–Tibet Plateau.

Currently, overgrazing is one of the primary causes of AM degradation and desertification in the Qinghai–Tibet Plateau. Therefore, reduce grazing intensity, which is the most cost-effective method of preventing AM from desertification, is to relieve grazing intensity and implement reasonable grazing systems, give vegetation some time for rehabilitation and then recovering the update and propagate abilities. First, reduce grazing intensity. The coverage and height of the AM and biomass of the upper and under-ground increased significantly after taking this method. Increasing the height and coverage of AM can make ground more rough. It not only reduces wind erosion, but also deposits quicksand. Meanwhile, the underground biomass increasing enhances the ability of consolidating soils by grass roots to play a great role in combating desertification. The second take reasonable grazing systems. Rotational grazing can give AM a short-term rehabilitation time to ensure its update and better growth. It is very necessary to carry out fencing policy rapidly to control the spread of sand disasters before moderate desertification.

Above all, conservative grazing methods can alleviate pressures on AM grasslands and lead to a healthier AM on the Qinghai–Tibet Plateau.