Soil Science Society of America Journal 66:1648-1655 (2002)
© 2002 Soil Science Society of America
DIVISION S-6—SOIL & WATER MANAGEMENT & CONSERVATION
Effect of Land Use on Soil Degradation in Alpine Grassland Soil, China
Ronggui Wu and
H. Tiessen*
Dep. of Soil Science, University of Saskatchewan, SK, Canada S7N 5A8
* Corresponding author (tiessen{at}sask.usask.ca)
 |
ABSTRACT
|
|---|
Grassland soils in Northern China are being seriously degraded under cultivation and grazing. This study investigated the impacts of land use on soil erosion and soil fertility in alpine grassland of China. Land uses included three levels of pasture degradation, lightly (LDP), moderately (MDP), and heavily degraded pasture (HDP) classified based on vegetation cover, and cultivated fields ranging from 1 to 50 yr of cultivation. Soil samples were collected from 18 sites at seven locations in Chernozemic soils between elevations of 2600 to 3000 m. Soil erosion was estimated by 137Cs radioactivity. When pasture was heavily degraded, 137Cs activity was significantly reduced, and organic C (OC), total N, and cation-exchange capacity declined by 33, 28, and 18% respectively. Cultivation of grassland worsened soil erosion, and after 8-, 16-, and 41-yr cultivation soil OC decreased by 25, 39, and 55%, respectively. Regionally, 59% of OC was lost within 30 to 50 yr of cultivation. There were concomitant losses of total N and exchange capacity. On cultivated soils, soil erosion and mineralization were equally responsible for organic C losses. Pasture degradation and cultivation also caused changes in soil P. Mineralization of organic P, incorporation of subsoil by tillage following erosion, and fertilization increased levels of Ca-P in cultivated fields. This study indicated that grassland degradation and cultivation caused not only severe soil erosion, but also fertility decline and chemical changes of P dynamics.
Abbreviations: Bicarb-P, soil P fractions extracted by NaHCO3 • C1-10, cultivation length between 1 and 10 yr • C11-30, cultivation length between 11 and 30 years • C31-50, cultivation length between 31 and 50 yr • Ca-Pi, Ca bound inorganic P extracted by diluted HCl • Cult-8, cropped fields in Year 8 of cultivation • Cult-16, cropped fields in Year 16 of cultivation • Cult-41, cropped fields in Year 41 of cultivation, Cult-48, cropped fields in Year 48 of cultivation • EC, electrical conductivity with water/soil ratio of 2:1 • ECEC, effective cation-exchangeable capacity • HDP, heavily degraded pasture • HHCl-P, soil P fractions extracted by hot concentrated HCl • LDP, lightly degraded pasture • MDP, moderately degraded pasture • NaOH-P, soil P fractions extracted by NaOH • OC, organic C • Res-P, soil P fractions extracted by resin
|
INTRODUCTION
|
|---|
SOIL DEGRADATION is a reduction of the current or future capacity of soil to produce (Dregne, 1987; Higgins, 1988). It can be because of erosion, decline in fertility, changes in aeration and moisture content, salinization, or a change in soil flora or fauna (Barrow, 1991). Scherr (1999) indicated that water erosion caused the most degradation, followed by wind erosion, soil nutrient depletion, and salinization. Erosion and soil fertility decline contribute about 84 and 7%, respectively, to world soil degradation (Oldeman, 1993). In China about 280 million ha (30% of the total land area) have been affected by degradation (Scherr, 1999). The provinces with largest eroded areas are Sichuan with 24.9 million ha, Inner Mongolia with 18.6 million ha, and Gansu with 17.2 million ha (Liu, 1997). Among the total of 45.4 million ha of land in Gansu, 
38% is considered eroded as a result of incorrect land use (Soil Survey Office of Gansu Province, 1993). Underlying causes for increasing soil degradation are changes in soil cover because of overgrazing, deforestation, and agricultural activity. An expanding population requires more food and has increased pressure to expand agricultural areas (Glantz, 1994) into marginal land and steep hillsides.
In the alpine region of Gansu province, where the present study was conducted, arable land use has advanced to an elevation of 3000 m, where traditional land use has been extensive grazing of sheep (Ovis aries) and yak (Bos grunniens) . At this agricultural frontier, grazing has intensified leading to wide-spread grassland degradation, and arable land use has been expanding around villages during the past 40 yr. Despite low yields and precarious harvests farmers crop alpine barley (Hordeum vulgare), rapeseed (Brassica napus), and oat (Avena sativa) to supplement food and fodder supplies. Grazing and arable lands account for about 85 and 3.5%, respectively, of the total 2.3 million ha between 2600 to 3000 m elevation. About half the grasslands are moderately to severely degraded because vast areas of pastures are grazed at stocking rates well above carrying capacity (Wu, 2001).
In the past, most Chinese research on soil degradation or soil erosion has focused on cropland in south or central China. Studies in the northern grassland regions have focused on rangeland resources (Zhu, 1993; Liu et al., 1994; Han et al., 1994; Wei and Liu, 1994), forage and livestock production (Li et al., 1994), plant communities (Sun, 1989; Nie and Yu, 1993; Shen et al., 1994; Nan et al., 1994; Niu, 1994; Xiu, 1996), and nutrient status of different grasses (Zhang et al., 1996). Causes of range degradation have been identified as overgrazing (Zhang, 1995; She and Sun, 1995), overexploitation of native plants (Yang, 1993; Yang et al., 1995), and rodent or insect effects (Wan and Wang, 1990; Zhu, 1993; Chen, 1994), resulting in declines in grass productivity (Sun, 1989; Xu, 1990), and reductions of plant cover (Xu, 1990). The relationships between range and soil degradation including erosion and soil fertility decline have not been well documented (Zhang, 1995).
Alpine grasslands are also under pressure from grazing and cultivation in other regions. In Switzerland, it is recommended that sheep grazing be limited to elevations below 2400 m, and that grazing above 2000 m be limited to a very short season in late summer (Blankenhorn, 1999). First signs of inappropriate grazing in the Alps are a reduction in plant diversity, followed by reduced cover (Blankenhorn, 1999), but changes in fertility have also been observed (Andrighetto et al., 1993). Grazing is the main cause of degradation in Alpine grasslands in many regions (Hall et al., 1999), but arable agriculture is also important where it has advanced to high-altitude fragile ecosystems. Traditional mountain societies have developed appropriate conservation techniques, such as practiced in the Andes (Dehn, 1995), but it is also recognized that mountain agriculture is on the margins of ecological sustainability (Landwirtschaftliche Forschungskomission, 1984). In northwest China, where high altitude arable agriculture is a recent development and grazing pressures are increasing, the land resource is at risk. Therefore, the present study evaluates soil degradation associated with typical land use and range degradation in the alpine grasslands of Gansu province.
|
MATERIALS AND METHODS
|
|---|
Two land use patterns were selected, grazed pasture, and adjacent cultivated fields. Pastures with different degrees of degradation, and cropped fields with known length of cultivation were chosen. Pasture degradation was classified in three categories based on plant cover (Zhang, 1995), plant diversity, and rodent (Myospalax bailey and Ochotona curzoniae) infestation (Wu, 2001): (i) LDP was characterized by a near 100% ground cover, high plant diversities, and few or no occurrences of rodent disturbance; (ii) MDP with about 80% plant cover; and (iii) HDP which had a <60% plant cover, low plant diversity, and common disturbances by rodents.
Soil degradation was studied in detail using sample transects at and near the Tianzhu Grassland Station of Gansu Agricultural University (37°11'48'' N lat., 102°47'3'' E long., elevation 2940 m, mean annual temperature -0.3°C, mean annual precipitation 416 mm), in August of 1997. Soil was classified as a chernozemic alpine meadow soil in Chinese Soil Classification (Xi et al., 1998), which corresponds to a typic Cryrendoll. Results from a second detailed study conducted at Shandan on the same soil type were similar to those in Tianzhu and since the regional scale sampling (see below) included the soils from Shandan, these results are not discussed separately.
To verify and extrapolate the findings of this local scale sampling to the region, 18 sites in seven locations throughout Gansu province with elevations between 2600 and 3000 m above mean sea level were studied. These sites were sampled using quadrats, in which pasture degradation was characterized and soils were sampled in July through August 1999.
Transect Sampling
At the Tianzhu grassland station, soil samples were taken along three parallel 165-m long transects from southeast to northwest. The transects were 10 m apart and soil samples were taken at 5-m intervals along each transect. At each sampling point, three cores (5-cm diam.) were randomly taken within 50 cm of each other to a depth of 15 cm, corresponding to the depth of the plough layer of the cultivated fields. About 500-g composite soil sample was obtained after combining three cores at each point. Each transect crossed three land uses, LDP, an oat field, and MDP. Both pastures were fenced. The oat field had been cultivated for 8 yr (Cult-8) in a rotation with rapeseed. In each transect, 34 soil samples were collected, six in LDP, 14 in the oat field, and 14 in MDP.
Two additional transects were sampled about 300 m northeast on a gentle slope of <5% crossing three land uses: an oat field cultivated for 16 yr (Cult-16), an oat/rapeseed field cultivated for 41 yr (Cult-41), and a fenced MDP. A total of 35 samples were taken along each transect, 14 in Cult-41, seven in Cult-16, and 14 in MDP. In a centrally located profile pit, a soil sample was taken from the B horizon between 25 and 50 cm, and also from the C horizon which extended to between 85 and 110 cm.
A further transect was sampled about 4 km southeast of the Tianzhu Grassland Station on a similar Cryrendoll. Two land uses were traversed, a nonfenced HDP, and a field cultivated for 48 yr. Only one transect from northwest to southeast with an average slope of 37% was taken with 10-m sampling intervals, resulting in a total of 40 soil samples, six from HDP, 34 from an alpine barley field. The first 12 samples in the barley field were located on a shoulder slope, the following 11 on a midslope and the last 11 on the footslope. The vegetation at both pasture sites consisted of Elymus nutans Grisb., Poa pratensis L., and Kobresia royleana.
Regional Scale Sampling
The seven locations of the regional study (Luchang, Shandan, Huangcheng, Tianzhu, Ganjia, Sangke, and Nayi) were all located where grazing and cropping coexist at the high elevation agricultural frontier. The region borders the Chinese loess plateaux to the east and the Gobi dessert to the North and all soils are formed from a similar mix of aeolian deposits that may have undergone some local alluvial and colluvial sorting (Soil Survey Office of Gansu Province, 1993). In the altitudinal belt of this study, all soils are therefore formed on similar parent material under the same grassland vegetation and are classified as typic Cryrendolls.
Quadrat (50 by 50 cm) sampling (Mueller-Dombois and Ellenberg, 1974) was used to identify plant species and estimate plant cover to categorize pasture degradation. Three cores of soil samples were collected inside each quadrat to a depth of 15 cm after counting the numbers of plant species and individual plants. The cores were mixed to obtain a composite soil sample representing the soil corresponding to the state of pasture degradation in the quadrat. Two separate soil samples were taken to a depth of 15 cm from adjacent cultivated fields. The year of first cultivation and fertilization history were recorded. Little 137Cs is found below the top 10 cm of undisturbed soils (VandenBygaart et al., 1998). We took five samples from a location at Shandan where the sod layer (top 10 cm) had been stripped and found no 137Cs activity. At the regional sites, the plough layer was no deeper than 15 cm. Because the 137Cs distribution in cultivated soils is uniform throughout the plough layer (VandenBygaart et al., 1998), both pasture and cultivated fields were sampled to a 15-cm depth in this study.
Six land use categories were distinguished in regional sampling: LDP, MDP, HDP, and different ranges of cultivation length between 1 and 10, 11 and 30, and 31 and 50 yr (C1-10, C11-30, and C31-50). The numbers of (composite) soil samples collected in the regional sampling were LDP, 11; MDP, 11; HDP, 17; C1-10, 13; C11-30, 11; and C31-50, 9.
Chemical and Physical Analyses
All soil samples were crushed and sieved to 2 mm after air-drying. On transect samples, soil pH and electrical conductivity (EC) were determined using a 2:1 water/soil ratio (Henry et al., 1987). Effective cation-exchangeable capacity (ECEC) was determined as the sum of exchangeable Ca, Mg, K, and Na (1.0 M NH4Cl extraction). Inorganic C and OC were analyzed by dry combustion (LECO, 1987). After thoroughly mixing the 2-mm sample, about a 60-g subsample was ground to 0.15 mm for total N and P analyses (H2O2 + H2SO4 digestion, Thomas et al., 1967), and sequential P fractionation (Tiessen and Moir, 1993). Soil pH, EC (at 5:1 ratio), OC, Total N, and total P from regional samples were analyzed in China based on the methods of Nanjing Soil Institute (1978). Selected samples from individual degraded pasture and cultivated fields were analyzed for particle size using the pipette method (Gee and Bauder, 1986). All data are presented on a mass basis.
Soil bulk density was determined at Tianzhu between 10 and 15 cm using a 100 cm3 cutting ring (5.00 cm in diameter) and (oven-dry) weighing, based on the method of the Nanjing Soil Institute (1978). The bulk density was used to establish 137Cs budgets and estimate soil erosion at Tianzhu. Volume- and area-base results were only used for the 137Cs data because there is a natural cut-off within the sampled layer. For any other soil property, converting data from a mass basis to an area budget would require that changes in bulk density be compensated by increased sampling depth. Without sampling to equivalent mass, an error is introduced which is avoided in our presentation.
Analyses of 137Cesium and Erosion
To determine impact of land use on soil erosion, the 137Cs content of soil samples was analyzed by high efficiency gamma spectroscopy (de Jong et al., 1982; Kachanoski et al., 1992). Cesium-137 is a unique tracer for soil erosion because it was evenly distributed on a regional scale during a known deposition period of atmospheric H-bomb testing; once deposited, it is strongly sorbed (Owens et al., 1997). As a result, Cs movement in a landscape is a measure of soil particle movement. The year (1963) of the maximum Cs input is also the mid point of the period during which 95% of Cs inputs occurred. Therefore, we used 1963 as starting point to estimate the soil erosion, following the example of Wu et al. (1994) and Yang et al. (1999). This gave a total of 35 yr for the calculation of Cs and soil loss for our study. All soil samples were taken and measured at about the same time and soil losses were estimated on the relative basis. The radioactivity of 137Cs in LDP was used as a reference to calculate relative soil loss from other land. Although some erosion probably also occurred in LDP despite its 100% plant cover (Zhang et al., 1993; Wu et al., 1994), it is the least eroded reference available in the region. A minimum estimate of soil loss from the other pastures was therefore calculated based on the following equation (de Jong and Kachanoski, 1988):
 | [1] |
where Er equals the erosion rate expressed as percentage of topsoil loss per year, 137Csu represents the specific activity of 137Cs in LDP (kBq m-2),137Cse is the specific activity of 137Cs in eroded pasture (kBq m-2), and n equals 35 yr since 137Cs fallout peak. Since all the soil samples in a location were taken at the same time and were counted within weeks of each other, radioactive decay of 137Cs was not considered.
On cultivated land 137Cs-free subsoil is brought up by tillage as soil erosion removes topsoil. As a result, 137Cs in the plough layer is diluted while it is being lost. Therefore, Eq. [1] is modified according to Kachanoski (1993) as:
 | [2] |
where E equals the annual erosion rate (kg m-2), M represents the specific soil mass of the plough layer in which the 137Cs is distributed (kg m-2), R is the ratio of the 137Cs concentration in the eroding sediment to that in the plough layer (1.1 was used, Kachanoski and de Jong, 1984),Tn denotes the total 137Cs (Bq m-2) present in the soil after n years of erosion, To is the total 137Cs (Bq m-2) present in the soil of uneroded landuse pattern, that is LDP, and n = number of cultivation years (up to a maximum of 35 yr).
Statistical Analysis
Transects were examined for the absence of trends not related to field divisions. All other statistical analyses were done using SPSS software (version 10.0). Differences between land use patterns were assessed using Tukey's honest significant difference (HSD) test (Lilienfein et al., 2000) at 0.05 level. Because large variations in soil properties existed between different locations, paired sample standard T-test was used on samples from individual locations from the regional sampling.
|
RESULTS AND DISCUSSION
|
|---|
Effect of Land Use on Soil Erosion
The average 137Cs activity under LDP at Tianzhu, which was used as the baseline soil, was 2.66 kBq m-2 (Table 1), close to the baseline of 2.64 kBq m-2 measured by Wu et al. (1994) in the same region (N 35.8°, E 107.6°). Loss of 137Cs in MDP was about 12% compared with LDP (Table 1). The losses of 137Cs from cultivated fields were considerably higher. Less than half of the baseline 137Cs activities were observed in Cult-8, and one-third in Cult-16 and Cult-41. The 137Cs load significantly decreased with increasing inorganic C (r = -0.93**) and pH (r = -0.68**), and was highly and positively correlated with OC (r = 0.78**), total N (r = 0.77**), and ECEC (r = 0.59**).
The highest annual erosion rate of soil was observed in recently cultivated fields. Erosion averaged 9.5 kg m-2 yr-1 after 8 yr of cultivation, and only 3.1 kg m-2 yr-1 over the 35 yr of Cs erosion in Cult-41 (Table 1). Reasons for the lower apparent erosion rate in Cult-41 may be: (i) 10% of the 137Cs load in China was deposited in an isolated event around 1986 (Yang et al., 1999). The error introduced by this late addition is greater for the more eroded sites and will cause an underestimation of erosion. (ii) The large variation of 137Cs activity in grassland (Fig. 1)
implied that initial levels of 137Cs activity in the pastures varied depending on the degree of pasture degradation and local redeposition of top soils (resulting in spots with higher 137Cs activity than the reference value in some cases) before breaking the land. The uncertainty produced by this is greater for younger cultivated sites. (iii) The area is adjacent to the Gobi desert, and regional dust storms may replenish 137Cs on a regular basis, causing a relatively low but constant long-term 137Cs radioactivity level. A similar situation has been documented in an erosion study in Niger (Chappell et al., 1998). (iv) Most importantly, to accelerate nutrient release in the initial cultivation phase, local farmers practice a soil fertility management when breaking the land that involves stripping the sod, piling, and burning it and then redistributing the ashes as fertilizer. This practice is equivalent to the use of burnt biomass that supports shifting cultivation systems in the tropics. As a result, a significant proportion of topsoil 137Cs may be removed from portions of fields and concentrated in spots were the sod is burned.
Soil losses from the steeper slope (37%) outside the station were much higher than those in the station. Average erosion over 35 yr reached 4.9 kg m-2 yr-1 on the midslope (Table 1). The 137Cs activity in HDP decreased by 46% compared with the reference LDP, indicating that severe soil erosion also occurred when grass cover was reduced.
Sand content was significantly negatively correlated with 137Cs activity (r = -0.75**, Fig. 2)
. Chen (1994) also reported that fine particle (<0.01 mm) content decreased from 30 to 36% to 12 to 17% after degradation occurred in a similar grassland soil. The B horizon of the soils at Tianzhu is sandier than the top soil (Table 2). Equation 2 estimates top soil loss based on the incorporation of 137Cs-free subsoil into the topsoil by tillage. Using the same equation, the differences of sand content (kg m-2) between the top and lower soil and with length of cultivation (Table 3) can be used as a tracer. Inserting the difference in sand content into Eq. [2], and using top soil sand content of the eroded sites to estimate erosion, losses of 5.8 kg m-2 yr-1 and 3.3 kg m-2 yr-1 can be calculated for the Cult-16 and Cult-41 fields, respectively. These values are quite comparable with those calculated based on the 137Cs activity.
Results from regional scale sampling showed a more linear loss of soil with length of cultivation than the local study in Tianzhu (Table 4). Because no bulk density was measured, 137Cs radioactivity was calculated based on weight exclusively. Activities of 137Cs declined significantly from an average of 20.2 Bq kg-1 in LDP to 10.0 Bq kg-1 in HDP, and then to 4.7 Bq kg-1 in crop fields with more than 30-yr cultivation. Short-term cultivation C1-10 and C11-30 had a similar 137Cs concentration as HDP. Comparison of these values with those in Table 1 indicates that long-term erosion losses are quite similar across the region, and that the detailed study reflects regional processes of degradation.
Effect of Land Use on Soil Organic Matter and Fertility
When native pasture was brought into crop production, OC declined by 22, 37, and 55% after 8-, 16-, and 41-yr cultivation, respectively relative to the MDP (Table 5). Soil OC loss can be because of soil erosion (Wu, 2001) and mineralization (Tiessen et al., 1992). To determine the relative importance of the two processes, the following estimation was made. About 92% of topsoil (0–15 cm) was lost after 16-yr cultivation (Cult-16) in Tianzhu based on 137Cs measurements and Eq. [2]. The OC concentration from 15 to 30 cm was taken as a constant mean value of 65 g kg-1. This was calculated by interpolating between the average of the 0- to 15-cm sample (Table 5) and of the B horizon (25–50 cm) sample (Table 2). Chen et al. (1995) reported a comparable concentration of organic matter in 0- to 30-cm depth of 125 g kg-1. If 92% of the topsoil were replaced by subsoil, as suggested by the 137Cs data, it would account for a decrease of OC to
assuming pasture was converted to cropland when it was moderately degraded. The actual OC concentration in Cult-16 was 51 g kg-1 (Table 5), mineralization of organic matter should be responsible for the difference of 66 - 51 = 15 g kg-1 OC. Therefore, of the total OC lost (81 - 51 = 30 g kg-1), soil erosion and mineralization each accounted for 15 mg kg-1 or half the loss.
The effects of land use on soil total N followed the same trend as on OC. No significant differences were obtained in total N between LDP and MDP, but the 41-yr cultivation resulted in about a 50% decline in total N. A slight decrease in C/N ratio can be attributed to the preferential loss of C-rich recent plant litter that is abundant in pasture soils. At the steeper site outside the station, losses were much greater, and significant loss also occurred from HDP.
In addition to soil organic matter losses, other indications of soil degradation were evident. Effective cation-exchange capacity significantly decreased with time of cultivation, as is expected from the trend toward sandier texture and organic matter loss. Soil ECEC was highly correlated with soil clay content (r = 0.97**) and soil OC (r = 0.90**).
Soil electrical conductivity decreased significantly when pasture was cultivated. Compared with LDP, EC dropped by 6, 16, and 20% on average after 8-, 16-, and 41-yr cultivation, respectively (Table 5). The variance of EC from MDP was significantly higher than that from LDP (F = 3.46, > F0.05), indicating pasture degradation may develop patchy salinity. This was confirmed in the detailed transect study at Shandan which is drier than Tianzhu, where maximum EC values of 0.83 were observed in the degraded pastures (data not presented). The salinity problem resolved under cultivation, probably because of better infiltration.
Soil total P increased with cultivation. Based on estimated fertilizer P additions (Wu, 2001), about 32 mg P kg-1 soil was attributable to P added as fertilizer in Tianzhu during 16 yr of cultivation. The actual difference of total P between pasture and 16 yr of cultivation was 70 mg kg-1 (Table 5), assuming pasture was converted into crop fields when it was moderately degraded. This meant that only about one-half of the 70 mg kg-1 P was accounted for by fertilizers. The remainder could be from animal wastes that have been used in the region at rates of 15 to 20 Mg ha-1 (Xu, personal communication, 1997), potentially supplying about 100 to 130 kg P ha-1 (Nanjing Soil Institute, 1978) per year. Since there were no records of how often and how much animal wastes were applied, it was impossible to give an accurate estimation for the contribution of animal wastes.
On the steep slope, soil ECEC, OC, and total N decreased significantly after pasture was converted to a cropped field (Table 5). Soil pH and total P increased with cultivation. Slope position also had an impact on soil chemical properties. The footslopes had a significantly higher ECEC than the other two positions, this was most likely due to topsoil eroded from the shoulder and midslopes and deposited in the footslopes. Soil OC and total N concentration also changed with slope positions. Shoulders are significantly lower in total N than the footslopes, while midslopes lay in between.
Table 6 shows the impacts of land use patterns on soil chemical properties on a regional scale. Soil OC declined significantly after pasture was put into crop production. Because variations of soil properties existed among locations (Table 7), most of soil fertility-related parameters were not significantly different between degraded pastures when compared across all sites (Table 6). Therefore soil variables were analyzed using paired sample statistics, resulting in significant differences for most of soil variables between LDP and HDP, and between MDP and HDP (Table 8). No significant differences were found between LDP and MDP.
The P fractionation indicated that bioavailable Pi (resin-P + bicarb-Pi) from Cult-41 was significantly higher than those from Cult-8, Cult-16, and pasture (Table 9). Inorganic P extracted by 0.1 M NaOH similarly increased with longer cultivation time (Cult-41). Calcium-associated Pi extracted by dilute HCl increased significantly with cultivation years from Cult-8 to Cult-16 and continuing to increase up to Cult-41. These increases in Pi were likely due to fertilization (Lilienfein et al., 2000) or to mineralization of organic P (Po). An additional cause for an increase in acid-extractable Ca-Pi may have been the incorporation of primary P from the subsoil by tillage. To test this hypothesis, the P budget in Table 10 was made: (i) about 92% of topsoil was lost after 16-yr cultivation. The subsoil contained 209 mg kg-1 of Ca-Pi (using a constant concentration of Ca-Pi from 15–30 cm). Chen et al. (1995) had shown that there were no differences in total P between 15 and 45 cm at this site. Replacing 92% of the top 15 cm by subsoil would account for an increase of Ca-Pi to
in the topsoil after 16-yr cultivation; (ii) fertilizer P was estimated to contribute 32 mg P kg-1 during 16-yr cultivation (Wu, 2001), mainly in Ca-Pi form; (iii) Po mineralization amounted to 60 mg kg-1 from MDP to Cult-16 (Table 9); (iv) Pi increases in other fractions than Ca-Pi were 16 mg kg-1 (Table 9). This gives a balance of Ca-Pi as shown in Table 10. Possible sources of higher Ca-Pi in Cult-16 were therefore (i) addition of P fertilizer 24%, (ii) incorporation of subsoil 31%, and (iii) mineralization of Po 45%.
Organic P decreased significantly from pasture to crop fields that had been cultivated for 16 yr or more. About 18 and 35% of the original Po was lost after the 16- and 41-yr cultivation, respectively. The 18% loss of Po and the increases in inorganic P (Table 9) in Cult-16 relative to MDP can be used to check whether the data on 137Cs, C , and P are internally consistent. Using a constant P concentration for the 15- to 30-cm layer of 770 mg kg-1, incorporation of P upon erosion of top soil following the cultivation of MDP accounted for (760 - 647) x 92% = 104 mg kg-1. The measured increase in inorganic P was 135 mg kg-1 (Table 9). Therefore 31 mg kg-1 should have been contributed from Po mineralization. The loss of Po amounted to 60 mg kg-1 (Table 10). Therefore, organic P which is bound to OC was depleted equally by mineralization and erosion, following the pattern seen for OC.
Results from regional scale sampling were quite similar to those from local scale sampling (Table 11). Cultivated fields had higher available Pi (resin-P + bicarb-Pi) levels than those in pasture soils. No significant differences in NaOH-Pi were observed among land uses. The Ca-associated Pi increased when pasture was cultivated, suggesting that most applied P, as well as mineralized P, were transformed to secondary Ca-associated Pi, reflecting chemical changes during soil degradation.
|
CONCLUSION
|
|---|
There are two major processes of soil degradation in the research region: soil erosion and organic matter mineralization. Erosion was the main cause of soil degradation under pasture. Cultivation of grassland worsened soil erosion and caused mineralization of OC and P.
This study used 137Cs levels to estimate soil erosion and the contributions of soil erosion to soil OC loss and to changes in soil P composition. About half the OC and organic P losses in a 16-yr-old cultivated site were due to erosional top soil loss; the the other half was attributable to mineralization and reduced organic matter inputs under cultivation. Erosion and incorporation of subsoil into the plow layer also accounted for about half the increases in Ca-bound P, with the remainder attributed to the mineralization of organic P and some fertilizer additions. Changes observed in soil P fractions and soil OC thus permitted the same conclusions, corroborating the evidence provided by different analytical approaches.
In the profile examined at one site, sand content was greater in the B horizon, than at the surface. This provided the opportunity to test the results obtained from decreasing 137Cs activities with those based on increasing sand content as erosion removed top soil. Again the conclusions from these two analytical approaches corroborated each other well. Limitations of the 137Cs method are seen in the older sites, where little activity is left and errors from drift-in of 137Cs containing dust and from small recent 137Cs releases become relatively more important.
To prevent further degradation of alpine grasslands, sloping croplands may have to be set aside to pasture, and stocking densities on pastures may have to be reduced. The visible degradation of the plant cover under pasture is a good indicator of underlying soil degradation and fertility loss.
|
ACKNOWLEDGMENTS
|
|---|
Financial support from Potash and Phosphate Institute, Potash & Phosphate Institute of Canada is gratefully acknowledged. The authors thank Dr. de Jong, the Department of Soil Science, University of Saskatchewan, for the 137Cs radioactivity determination.
Received for publication October 29, 2001.
|
REFERENCES
|
|---|
- Andrighetto, I., G. Cozzi, P. Berzaghi, and M. Zancan. 1993. Avoidance of degradation of Alpine pasture through grazing management: investigations of change in vegetation nutrition characteristics as a consequence of sheep grazing at different periods of the growing season. Land-Degradation-Rehabil. 4(1):37–43.
- Barrow, C.J. 1991. Land Degradation: Development and breakdown of terrestrial environments. Cambridge University Press, New York.
- Blankenhorn, H.-J. 1999. Bulletin 3/99 - Artikel 11. (In German.) Bundesamt für Umwelt, Wald und Landschaft •CH-3003 Bern.
- Chappell, A., A. Warren, M.A. Oliver, and M. Charlton. 1998. The utility of 137Cs for measuring soil redistribution rates in southwest Niger. Geoderma 81:313–337.
- Chen, Y., Z. Wu, X. Zhu, and Z. Hu. 1995. Phosphorus cycling in Grass - Kobresia alpine grassland (In Chinese). Acta Prataculturae Sinica 4(2):75–81.
- Chen, Z. 1994. Several hot topics on grassland ecology (In Chinese). Pratacultural Science 11(1):42–45.
- Dehn, M. 1995. An evaluation of soil conservation techniques in the Ecuadorian Andes. Mountain research development. 15(2):175–182.
- de Jong, E., and R.G. Kachanoski. 1988. The importance of erosion in the carbon balance of prairie soils. Can. J. Soil Sci. 68:111–119.
- de Jong, E., H. Villar, and J.R. Bettany. 1982. Preliminary investigation on the use of 137Cs to estimate erosion in Saskatchewan. Can. J. Soil Sci. 62:673–683.
- Dregne, H.E. 1987. Soil erosion: Cause and effect. Land Use Policy 4:412–418.
- Gee, G.W., and J.W. Bauder. 1986. Particle size analysis. p. 383–411. In A. Klute (ed.) Methods of Soil Analysis. Part 1. 2nd ed. Agron. Monogr. 9. ASA and SSSA, Madison, WI.
- Glantz, M.H. 1994. Drought follows the plow: cultivating marginal areas. Cambridge University Press, Cambridge, UK.
- Hall, K., J. Boelhouwers, and K. Driscoll. 1999. Animals as erosion agents in the alpine zone: Some data and observations from Canada, Lesotho, and Tibet. Arct-antarct-alp-res. 31(4):436–446.
- Han, G., D. Liu, Q. Liu, and Y. Meng. 1994. Evaluation of grassland resources in major pastoral areas of China. p. 203–208. In B. Li (ed.) Proceedings of the International Symposium on Grassland Resources. Hohhot, China. 15–20 Aug. 1993. China Agriculture Scientech Press, Beijing.
- Henry, L., B. Harron, and D. Flaten. 1987. The nature and management of salt-affected land in Saskatchewan. Agdex 518, January, 1987.
- Higgins, G.M. 1988. Soil degradation and its control in Africa. Paper to First all-African Soil Science Conference, Kampala, Uganda. December 5–10, 1988.
- Kachanoski, R.G. 1993. Estimating soil loss from changes in soil 137Cs. Can. J. Soil Sci. 73:629–632.
- Kachanoski, R.G., and E. de Jong. 1984. Predicting the temporal relationship between soil cesium-137 and erosion rate. J. Environ. Qual. 13:301–304.[Abstract/Free Full Text]
- Kachanoski, R.G., M.H. Miller, D.A. Lobb, E.G. Gregorich, and R. Drotz. 1992. Management of farm field variability, III- Effects of tillage systems on soil and phosphorous loss. SWEEP Report #45, University of Guelph, Guelph, ON.
- LECO. 1987. Total and organic carbon in siltstone. LECO Corp. Application Bulletin. Form no. 203-601-140. LECO Corp., St. Joseph, MI.
- Landwirtschaftliche Forschungskommission. 1984. Mountain agriculture in the struggle between economy and ecology. Die Berg-landwirtschaft im Spannungsfeld zwischen Oekonomie und Oekologie: Kolloquium MAB Bern, 23. November 1984. Wissenschaftliche Beilage zum Landwirtschaftlichen Jahrbuch der Schweiz. Bern: Bundesamt fur Landwirtschaft.
- Li, B., Z. Ren, Q. Yuan, P. Shi, J. Chen, Z. Li, and J. Li. 1994. Experiments on dynamics monitoring of balance between forage and livestock in Northern China. p. 35–50. In B. Li (ed.) Proceedings of the International Symposium on Grassland Resources. Hohhot, China. 15–20 Aug. 1993. China Agriculture Scientech Press, Beijing.
- Lilienfein, J., W. Wilcke, M.A. Ayarza, L. Vilela, S.C. Lima, and W. Zech. 2000. Chemical fraction of phosphorus, sulfur, and molybdenum in Brazilian savannah Oxisols under different land use. Geoderma 96:31–46.
- Liu, J. 1997. China Agriculture Yearbook. China Agriculture Press (English version). Beijing.
- Liu, Q., X. Wu, and L. Bao. 1994. Natural grassland resources of China and their characteristics. p. 163–169. In B. Li (ed.) Proceedings of the International Symposium on Grassland Resources. Hohhot, China. 15–20 Aug. 1993. China Agriculture Scientech Press, Beijing.
- Mueller-Dombois, D., and H. Ellenberg. 1974. Aims and Methods of Vegetation Ecology. John Wiley & Son, New York.
- Nan, Y., L. Xu, Z. Kou, and J. Wei. 1994. Study on stability of shrub community for forage and dune arrest. p. 135–140. In B. Li (ed.) Proceedings of the International Symposium on Grassland Resources. Hohhot, China. 15–20 Aug. 1993. (In Chinese.) China Agriculture Scientech Press, Beijing.
- Nanjing Soil Institute. 1978. Soil Physical and Chemical Analyses Handbook. Shanghai Scientech Press, Shanghai.
- Nie, Z., and L. Yu. 1993. Several aspects of grazing ecological research (In Chinese). Grassland of China 6:64–69.
- Niu, J. 1994. The analysis of spatial distribution and succession of main sandland plant communities on the Ordos plateau, Inner Mongolia. p. 277–282. In B. Li (ed.) Proceedings of the International Symposium on Grassland Resources. Hohhot, China. 15–20 Aug. 1993. China Agriculture Scientech Press, Beijing.
- Oldeman, L.R. 1993. An international methodology for a world soil and terrain database and a global assessment of soil degradation. Expert Consultation of the Asian Network on Problem Soils, Bangkok, Thailand.
- Owens, P.N., D.E., Walling, Q. He, S. Jo, and I.D.L. Foster. 1997. The use of cesium-137 measurement to establish a sediment budget for the Start catchment, Devon, UK. Hydrol. Sci. 42:405–423.
- Scherr, S.J. 1999. Soil Degradation—A threat to developing country food security by 2020. Food, Agriculture, and the Environment Discussion Paper 27, February of 1999. International Food Policy Research Institute, Washington, DC.
- She, F., and G. Sun. 1995. Grassland resource estimation and the way of improving grassland productivity for Jingtai, Gansu province (In Chinese). Pratacultural Sci. 12(2):9–12.
- Shen, Y., S. Yan, and X. Zhu. 1994. The distribution and soil environmental properties of the main plant communities in a salinized meadow land of Hexi Corridor (In Chinese). Acta Phytoecologica Sinica 18(1):95–102.
- Soil Survey Office of Gansu Province. 1993. Soils of Gansu. China Agriculture Press, Beijing.
- Sun, Z. 1989. Causes of grassland deterioration and management strategies in Wengniute Banner (In Chinese). Pratacultural Sci. 6(2):60–64.
- Thomas, R.L., R.W. Sheard, and J.R. Moyer. 1967. Comparison of conventional and automated procedures for N, P and K analysis of plant material using a single digestion. Agronomy J. 59:240–247.[Abstract/Free Full Text]
- Tiessen, H., and J.O. Moir. 1993. Characterization of available P by sequential extraction. p. 75–86. In M. R. Carter (ed.) Soil Sampling and Methods of Analysis. Lewis Publishers, Chelsea, MI.
- Tiessen, H., I.H. Salcedo, and E.V.S.B. Sampaio. 1992. Nutrient and soil organic matter dynamics under shifting cultivation in semi-arid northeastern Brazil. Agric. Ecosyst. Environ. 38:139–151.
- VandenBygaart, A.J., R. Protz, and D.C. McCabe. 1998. Distribution of natural radionuclides and 137Cs in soils of southwestern Ontario. Can. J. Soil Sci. 79:161–171.
- Wan, Y., and L. Wang. 1990. The present situation of degraded grassland of Qinghai province and recovering ways (In Chinese). Grassland of China 6:58–60.
- Wei, S., and X. Liu. 1994. The resources characteristics of meadow-steppe in Hulunber of China. p. 415–420. In B. Li (ed.) Proceedings of the International Symposium on Grassland Resources. Hohhot, China. 15–20 Aug. 1993. China Agriculture Scientech Press, Beijing.
- Wu, R. 2001. The effect of land use on soil fertility and P dynamics in alpine grassland soils of Gansu, China. Ph.D. thesis, Dep. of Soil Science, University of Saskatchewan, Saskatoon, Saskatchewan.
- Wu, Y., Z. Li, X. Zhang, and Y. Wang. 1994. Study on 137Cs method used in soil erosion and sediment delivery on sloping fields in gully area on loessal plateau. (In Chinese.) Bull. Soil Water Conserv. 14(2):22–26.
- Xi, C., K. Zhu, and G. Du. 1998. Soils of China. China Agriculture Press, Beijing (In Chinese).
- Xiu, Y. 1996. The reason and control strategy of grassland degradation in Weng Niu Te Qi of Inner Mongolia (In Chinese). Grassland of China 5:63–65.
- Xu, Z. 1990. Controlling stocking rate, the key of maintaining grassland ecological balance (In Chinese). Pratacultural Sci. 7(5):1–6
- Yang, F. 1993. A primary approach on the grassland desertification status and controlling measures in Ningxia (In Chinese). Pratacultural Sci. 10(1):55–57.
- Yang H., M. Du, Q. Zhang, and J. Yang. 1999. Quantitative model considering surface enrichment to estimate soil erosion rates using 137Cs (In Chinese). J. Soil Erosion Soil Water Conserv. 5(1):42–48.
- Yang, S., B. Yu, W. Xu, D. Gao, Z. Wang, and F. Gao. 1995. Causes of grassland degradation and related control measures in Songnen (In Chinese). Pratacultural Sci. 12(4):5–9.
- Zhang, X., S. Li, T.A. Quine, and D.E. Walling. 1993. Effect of cultivation on soil erosion estimated by 137Cs measurement (In Chinese). Chinese Sci. Bull. 38:2072–2076.
- Zhang, Y., G. Yang, Z. Jin, and L. Gu. 1996. A study on the nutritive trends of the six species of main herbages on sub-alpine meadows (In Chinese). Pratacultural Sci. 13(5):12–16.
- Zhang, Z. 1995. The effect and causes of grassland degeneration (In Chinese). Pratacultural Sci. 12(6):1–5.
- Zhu, T. 1993. Grasslands of China. p. 61–82. In R. T. Coupland (ed.) Ecosystems of the World 8B: Natural grasslands, eastern hemisphere and résumé. Elsevier, Amersterdam, the Netherlands.
TOP
ABSTRACT
INTRODUCTION
