Effects of erosion patterns on nutrient loss following deforestation on the Loess Plateau of China

Agriculture, Ecosystems and Environment 108 (2005) 85–97 www.elsevier.com/locate/agee

Effects of erosion patterns on nutrient loss following deforestation on the Loess Plateau of China Fenli Zheng a, Xiubin He b,*, Xuetian Gao a, Chang-e Zhang c, Keli Tang a a

National Laboratory of Soil Erosion and Dryland Farming on Loess Plateau, Institute of Soil and Water Conservation, CAS and MRS, Northwestern Sci-Tech University of Agriculture and Forestry, 26 Xinong Rd., Yangling, Shaanxi 712100, PR China b Institute of Mountain Hazards and Environment, Ministry of Water Conservance, CAS, No. 9, Block 4, South Renmin Road, Chengdu, Sichuan 610041, PR China c Chinese Academy of Agricultural Science, No.12, Zhongguancun South Street, Beijing 100081, PR China Received 24 July 2003; received in revised form 30 November 2004; accepted 13 December 2004

Abstract Soil degradation caused by deforestation is one of main environmental problems all over the world. The objective of this paper was to quantitatively evaluate the effects of erosion patterns on organic matter, nitrogen and phosphorus (P) losses, by monitoring soil erosion and nutrient loss in newly deforested lands in the Ziwuling region on the Loess Plateau of China from 1986 to 1996. Eight field runoff plots, with various sizes to enable documentation different combinations of dominant erosion processes, were established on a hillslope. Results showed that the nutrient loss was dramatically affected by erosion patterns and erosion intensity. Seven years after deforestation, organic matter, total nitrogen, ammonium nitrogen and available P reduced by 69, 46.7, 65.6 and 86.6%, respectively. The most severe soil erosion and nutrient loss occurred in the ephemeral gully channels. Organic matter, total nitrogen, ammonium nitrogen and available P had different nutrient enrichment ratios in eroded sediment. The available P had the highest enrichment ratios, followed by ammonium nitrogen. The nutrient enrichment ratios were closely related to sediment concentration and erosion patterns. In sheet erosion zone, the nutrient enrichment ratios in eroded sediment decreased with an increase of sediment concentration; in the combined erosion zones of sheet, rill and shallow gully, the nutrient enrichment ratios in sediment initially decreased and then increased with an increase of sediment concentration. The nutrient enrichment in eroded sediment also greatly affects by rainfall characteristics such as rainfall amount and intensity. # 2005 Elsevier B.V. All rights reserved. Keywords: Erosion patterns; Nutrient loss; Nutrient enrichment; Deforestation; Rainfall; The Loess Plateau of China

1. Introduction * Corresponding author. Tel.: +86 28 85232105; fax: +86 28 85222258. E-mail address: [email protected] (X. He).

Soil erosion, in addition to causing on-site loss of topsoil and reducing the productivity of the land, brings about major off-site environmental effects such

0167-8809/$ – see front matter # 2005 Elsevier B.V. All rights reserved. doi:10.1016/j.agee.2004.12.009

86

F. Zheng et al. / Agriculture, Ecosystems and Environment 108 (2005) 85–97

as water body pollution and eutrophication (e.g., Lal, 1976; Morgan, 1995; He et al., 2003). There are many studies concerning the prediction of soil losses based on the universal soil loss equation (USLE) (e.g., Wischmeier and Smith, 1978) or the water erosion prediction project (WEPP) (e.g., Nearing et al., 1989). Meanwhile, a large number of studies have investigated eutrophic components (e.g., Sharpley et al., 1991; Sharpley and Halvorson, 1994), pesticides, herbicides (Miguel and Clara, 2003) and agricultural chemicals transfer (e.g., Johnson and Baker, 1982, 1984; Smolikowski et al., 2001). The later has recently been an interesting research area for both soil and environmental scientists. A number of studies have been designed to determine the effects of factors, such as rainfall intensity, runoff rate, slope, clay content and storm pattern on nutrient losses in the Loess Plateau. Bai et al. (1991) reported that nutrient loss increased with increasing slope steepness under simulated rainstorm conditions in a 5-m long and 1.5-m wide soil box. Monke et al. (1977) studied relationships between runoff, erosion and nutrient movement in the interrill areas, and pointed out that the proportion of clay in sediment decreased with an increasing runoff rate while trends in nutrient contents showed an opposite trend. They also showed that enrichment ratios in sediment ranged from 1.2 to 2.3. Flanagen and Foster (1989) studied storm pattern effects on nitrogen (N) and phosphorus losses, and found that N and P in eroded sediment were significantly enriched. Young et al. (1986) showed that nutrient enrichment in eroded sediment decreased with increasing runoff rate and suspended sediment concentration. Alberts and Moldenhauer (1981) and Alberts et al. (1981) studied the transport of N and P with sediment in runoff passing through cornstalk residue strips. They reported that particles less than 0.002 mm in diameter had highest total N and P concentrations with concentrations decreasing in larger size classes. Sharpley et al. (1991) measured soil erosion and associated N and P losses with sediment under natural rainfall as affected by land management. They also found that both N and P contents in sediment were strongly related to sediment concentration of individual runoff events (Eqs. (1) and (2)): N s ¼ 0:03X s 0:68

(1)

Ps ¼ 0:72X s 0:30

(2)

where Ns is content of N in sediment (g/kg), Ps content of P in sediment (g/kg) and Xs is sediment concentration (g/L). Similarly, Johnson and Baker (1982, 1984) reported on the relationship between Kjeldahl-N and total P in sediment in over 650 individual runoff samples collected from corn and soybean fields over a 5-year period. They also reported a strong relation between N and P contents in sediment and sediment concentration in runoff events, respectively (Eqs. (3) and (4)). N s ¼ 6675X s 0:11

(3)

0:18

(4)

Ps ¼ 4315X s

where Ns is content of N in sediment (mg/kg), Ps content of P in sediment (mg/kg) and Xs is sediment concentration (g/L). Regardless of the exponent differences in the above equations, the research results indicated that nutrient content in eroded sediment decreased with an increase of sediment concentration. On the Loess Plateau of China, heavy rainfall during summer and earlier autumn causes severe soil erosion. Soil erosion patterns change from sheet, rill to shallow gully (like ephemeral gully) erosion from top to bottom along loessial hillslopes (Fig. 1) (Zheng and Huang, 2002). These different soil erosion patterns have different soil detachment and transport capacities, therefore, they have different impacts on soil nutrient loss. Numerous studies have been conducted on soil erosion on the Loess Plateau, and sediment issues in the Yellow River basin (e.g., Huang, 1953; Zhu, 1956, 1981; Chen et al., 1988; Tang, 1991; Jing et al., 1993; Chen, 1993; Wang and Jiao, 1996; Cai et al., 1998; Xu, 1999; Zheng and Gao, 2000; Laflen et al., 2000). These research findings have increased the understanding of soil erosion issues and associated environmental problems. Since the 1980s, there have been a few papers that have also investigated the relationship of soil erosion to nutrients loss on the Loess Plateau (Tang et al., 1987; Peng and Wang, 1995; Hamilton and Luk, 1993). However, little attention has been given to quantify the effects of the soil erosion process on nutrient loss, particularly, the effects of erosion patterns on nutrient loss after conversion of forestlands into croplands.

F. Zheng et al. / Agriculture, Ecosystems and Environment 108 (2005) 85–97

87

Fig. 1. Erosion zone distribution in a loess hillslope.

The present study was designed to monitor soil erosion and nutrient loss on field runoff plots in Ziwuling forest region on the Loess Plateau of China, to quantify the effects of erosion patterns on organic matter (OM), total nitrogen (TN), ammonium–nitrogen (NH4–N) and available phosphorus losses, and to study the relationship between soil erosion and soil quality degradation after forest vegetation removal.

2. Methods 2.1. Study site description Natural vegetation in most of regions on the Loess Plateau has been destroyed over the past more than 100 years except in the Ziwuling area. The Ziwuling forest area, which occupied about 23,000 km2, is the sole secondary forest region remaining on the Loess plateau. It is situated in the hilly region between Dongzhi Plateau and Luochuan Plateau (latitude 338500 –368500 N and longitude 1078300 –1098400 E). About 130 years ago, the eco-environment of the Ziwuling region was similar to most of regions of the hilly-gully region in the Loess Plateau today, i.e., natural vegetation was completely destroyed, and soil

erosion was severe. Around 130 years ago, a war resulted in population decrease and out-migration. Since that time, the secondary forest vegetation has been developing. The field experiment was conducted at the Fuxian Observatory for Soil Erosion and Eco-environment, which was established in 1989 (Tang et al., 1993). This Observatory is located at the eastern slope of the Ziwuling secondary forest region. The landforms are characterized by low mountains and hills covered by loess with elevation ranging from 920 to 1683 m with a relative height of 100–150 m and a gully density of 4.5 km/km2. The mean annual temperature is 6–10 8C, precipitation ranged from 600 to 700 mm, of which 60% falls from June to September. The maximum precipitation in a month is equivalent 25–40% of the annual total, and the maximum daily rainfall is 87 mm. In the forest area, the main tree species are oak (Quercus liaotungensis Koidz), poplar (Populus davidiana Doze) and birch (Betula platyphyua Sukats). The canopy density of timber forest is more than 70%. The soil surface in the forestland is covered by a 2–5-cm deep litter layer. The soil is calcic loess with the particle size distribution of 6.7% sand, 72.1% silt and 21.2% clay (He and Huang, 2001). According

88

F. Zheng et al. / Agriculture, Ecosystems and Environment 108 (2005) 85–97

Fig. 2. Land use, topography and sizes of natural runoff plots and their locations at hillslope and distribution of soil erosion zones within the plots.

to the Chinese soil classification, the soil is a gray forest soil. It has an obvious organic horizon, welldeveloped soil aggregates and a dense rooting system. However, argillic horizons are not obvious (He et al., 2002). 2.2. Site selection and field plots design Topography in the hilly-gully region of the Loess Plateau is complex and slope is very steep. Along with slope length, slope gradient increased, i.e., slope shape is convex. Slope gradients change from 38 to 128 at the sheet erosion dominant zone (top slope), 12–258 at the rill erosion dominant zone (middle slope) to 18–358 at the shallow gully erosion dominant zone (down slope). To identify replicable slope sections to establish runoff plots, contour maps and aerial photos were analysed to obtain typical information about topographic features, such as slope gradients, lengths and shapes in different erosion zones as well as the dimension of shallow gullies. Later on, a field survey was made to select a representative hillslope to establish runoff plots according to the typical topographical features. Eight

runoff plots were constructed on a hillslope (Fig. 2) in the summer of 1989. Each plot was surrounded with concrete block borders (extending 10 cm above the surface and 10 cm below the surface). A trough for runoff collection was installed at the outlet of each plot. Plots 1 and 2, with a size of 2.5-m width and 15-m length, were set up on top slope with slope degree of 5– 128 to monitor sheet erosion; Plots 3 and 4 for the combined erosion zones of sheet and rill were 2.5-m wide and 40.4-m long with 5–268 slope degree; and Plots 6 and 7 for the combined erosion zones of sheet, rill and shallow gully were 13.6-m wide and 73-m long with a 5–358 slope. All those experimental plots had been a forestland, and were cleared and plowed by handy tools such as shovels and dibblers in the summer of 1989. They have been fallowed since 1989, but plowed in each spring. As controlling plots, Plots 5 and 8 were set up on the undisturbed forestland. The Plot 5, located in the combined erosion zones of sheet and rill, was neighbored to the Plot 4 with similar size and micro-topography to it; Plot 8, in the combined erosion zones of sheet, rill and shallow gully, was neighbored to the Plot 7 with similar size and micro-topography to it.

4.58 4.64 5.14 2.36 2.26 2.13 118.3 116.0 103.8 105.2 93.6 71.0 1.14 1.11 1.38 1.22 1.28 0.96 19.02 18.78 22.36 18.76 15.25 12.56 5.10 5.03 5.91 2.60 2.53 2.71 131.0 130.4 114.0 116.0 105.4 103.6 1.21 1.16 1.47 1.32 1.34 1.20 21.46 20.28 28.79 17.7 17.99 15.71 8.78 8.45 15.60 6.83 9.73 15.94 183.6 181.9 192.3 176.3 188.6 206.4 1.31 1.29 1.75 1.47 1.61 1.80

OM: organic matter.

Sheet Sheet Rill Sheet Rill Rill and shallow gully Sheet erosion (zone 1) Combined zone of sheet and rill (zone 1 + zone 2) Combined zones of sheet, rill and Combined zones of rill and shallow gully erosion (zone 1 + zone 2 + zone 3)

a

NH4–N (mg/kg) Total N (g/kg) OM (g/kg)

a

89

21.45 23.55 37.74 29.74 28.25 40.46

P (mg/kg) NH4-N (mg/kg) Total N (g/kg) OMa (g/kg) OM (g/kg)

Total N (g/kg)

NH4–N (mg/kg)

P (mg/kg)

7 years

P (mg/kg)

a

3 years

After forest land being disturbed Control (forestland)

Dominated erosion patterns

Storage containers were used to collect runoff volume for each runoff plot for each runoff event. Multi-slot divisors were installed in case of that one collective container is not enough to hold all runoff in an extreme rainfall event. The multi-slot divisor has one hole with 5-cm diameter in the back to collect runoff from the trough at the plot outlet, and 11 holes with 3-cm diameter in the front at the 65-cm height. Runoff from the middle hole channeled into the next multi-slot divisor or container and the rest from other 10 holes was drained. One collective container without multi-slot divisors was used to collect runoff from Plots 1, 2, 5 and 8, one multi-slot divisor and one collective container for both Plots 3 and 4, and two multi-slot divisors and one collective container for both Plots 6 and 7. During observation periods from 1989 to 1996, prior to the raining season (in April), the soil in the runoff plots was turned over by a shovel to a depth of 20 cm, according to locally traditional cultivation. Soil samples for nutrient test were taken by core samplers twice a year, i.e., 3 days after tillage in April and after the rain season in the end of October. According to the topography and erosion patterns, five sampling sites at the points of 1.0, 3.5, 7.0, 10.5 and 14 m along slope length in the sheet erosion zone, seven at 1.0, 4.0, 7.5, 12.0, 16.5, 21 and 25 m along slope length in the rill erosion zone, 18 sampling sites at 1, 5, 12, 19, 27, 34 m along slope length in the shallow gully erosion zone (six in the shallow gully channel and 12 at both sides of shallow gully channel) were chosen. In addition, soil samples from woodlands (Plots 5 and 8) were collected according to corresponding sample locations of deforested lands. After each runoff event, water level in each container was measured to calculate runoff volume, and four sediment 1-l bottle runoff samples were collected from each container after mixing round with a muddler. After more than 24-h sedimentation, poured out the clear water from bottles (one-third or half of the water may be poured out at the first time, and waiting again for the next time) (Tang et al., 1993). The rest samples were dried by air or sunlight. Then, the air-dried soil samples were transported to the Institute for laboratory analysis.

Erosion zones

2.3. Data collection

Table 1 Nutrient status before and after the removal of forest vegetation in combined zones of sheet and rill erosion, and combined zones of sheet, rill and shallow gully erosion (the data are average mean values of the years)

F. Zheng et al. / Agriculture, Ecosystems and Environment 108 (2005) 85–97

90

F. Zheng et al. / Agriculture, Ecosystems and Environment 108 (2005) 85–97

2.4. Laboratory analysis

3. Results and discussions

After each runoff event, dry sediment samples were immediately transported to the Institute for chemical analysis. Organic matter was determined using the method of Walkley and Black, total nitrogen (Kjeldahl digest) and NH4–N (2.0 M KCl extractant) were determined using an Automated Ion analyzer, available phosphorus (0.5 M NaHCO3 extractant) was analyzed with colorimetry.

3.1. Effects of soil erosion patterns on nutrient loss

2.5. Calculation The values of soil OM, TN, NH4–N or available P content of the sampling sites in each runoff plot in each April were taken as the reference values to calculate the enrichment ratio in sediment of that year. The differences of the annual values in each runoff plot between April and October of each year were calculated to assess nutrients loss. In the newly deforested land on the Loess Plateau of China, nutrient content in topsoil greatly varied with erosive events and time due to severe soil loss. In addition, the runoff plots with different sizes had different nutrient contents due to differences in erosion rates and erosion patterns. In this study, nutrient enrichment ratio in eroded sediment from each runoff plot in each observation year was defined as a special nutrient content in eroded sediments to the corresponding the soil reference value in April of that year for the same plot. These nutrient enrichment ratios were used to quantify nutrient loss associated with different erosion intensities and patterns.

Sheet erosion, rill erosion and shallow gully erosion have different soil detachment and transport capacities, therefore, they potentially have different impact on soil nutrient loss. Table 1 shows the reduction of nutrients in the topsoil (20 cm) in sheet, rill and shallow gully zones after 3 and 7 years after forest vegetation removal. In sheet erosion zone, average annual soil erosion rate was 6700 t km2 year1 (Table 2). Consequently, after 3 years of converting forestland into farmland, OM decreased by 13.9%, TN by 9.6%, NH4–N 28.3%, and available P by 40.4%. Seven years following deforestation, OM decreased by 20%, TN decreased by 13.6%, NH4–N 36.2%, and available P by 45.1%. As erosion rate increases of 13,200 t km2 year1 in the combined erosion zones of sheet and rill to 18,500 t km2 year1 in the combined erosion zones of sheet, rill and shallow gully, soil nutrient contents of OM, TN, NH4–N and available P decreased accordingly (Table 1). In the most severely eroded shallow gully channel, soil degradation was the most severe. Three years after forest vegetation removal, reductions in OM, TN, NH4–N and available P were 61.2, 49.8 and 83.1%, respectively. After 7 years, the reductions in OM, TN, NH4–N and available P were 69.0, 46.7, 65.6 and 86.6%, respectively. These results indicated that nutrient loss occurred rapidly following deforestation. The degree of nutrient loss on hillslopes increased as sheet erosion shifted to rill erosion or shallow gully erosion. The content of available P had

Table 2 Average erosion rates and standard deviations (in parentheses) from sheet erosion zone, the combined erosion zones of sheet and rill, and the combined erosion zones of sheet, rill and shallow gully erosion during the observation periods Erosion zone

Dominated erosion patterns

Mean erosion rate (S.D.) (t km2 year1)

Ratios of erosion rates to that in the sheet erosion zone

Sheet (zone1) The combined erosion zones of sheet and rill (zone 1 + zone 2) The combined erosion zones of sheet, rill and shallow gully (zone 1 + zone 2 + zone 3)

Sheet Rill

6775 (1345) 13221 (2369)

1.0 1.95

Rill and shallow gully

18579 (3826)

2.74

F. Zheng et al. / Agriculture, Ecosystems and Environment 108 (2005) 85–97

91

Table 3 Nutrient contents in eroded sediment and their enrichment ratios from sheet erosion during single rainstorm Rainfall parameters Amount (mm)

I30(mm h1)

PI30 (mm2 h1)

24.2 13.6 18.2 27.9 108.0 75.2 11.5

46.8 27.0 7.8 27.0 69.0 16.2 69.0

1132.56 367.20 141.96 753.30 7452.00 1218.24 793.50

Erosion rate (t km2)

Sediment concentration (kg/m3)

Nutrient content OM (g/kg)

Total N (g/kg)

NH4–N (mg/kg)

P (mg/kg)

Enrichment ratios OM

Total N

NH4–N

P

2418 160 24 95 6526 690 255

254 68 11 50 127 27 19

23.99 27.16 31.08 23.54 21.19 25.48 24.36

1.15 1.28 1.54 1.28 1.17 1.34 1.37

128.2 146.2 184.8 148.8 123.1 152.8 153.9

10.19 13.38

1.08 1.22 1.40 1.29 1.16 1.40 1.34

0.94 1.05 1.26 1.09 1.00 1.14 1.16

1.11 1.26 1.60 1.45 1.20 1.49 1.50

1.80 2.36

11.36 8.66 12.32 13.47

2.48 1.89 2.69 2.94

I30 means maximum 30-min rainfall intensity during a single rainstorm.

the highest reduction, then followed by NH4–N and OM, after clearance of the forested area for all erosion patterns. This could be attributed to phosphorous, ammonium–nitrogen and organic matter were strongly bounded by clay of soil and transported with eroded sediment, whereas N is associated with OM (Peng, 1999; Jun et al., 2003).

sediment were from 1.08 to 1.40, and 0.89 to 1.26, respectively. The enrichment ratios were similar to the results of Monke et al. (1977), regardless of different geographical locations. But the enrichment ratios in sheet erosion zone from this study were higher than those obtained from long-term cultivated areas in the severely eroded region of Loess Plateau, indicating accelerated nutrient loss in the beginning period of forest vegetation removal (Li, 1990). The nutrient enrichment in eroded sediment also greatly affects by rainfall amount and rainfall intensity (Tables 3–5). This can explain that rainfall amount and rainfall intensity significantly influence runoff detachment and transport capacity, and further affect erosion rate. Jiang et al. (1990) reported that PI30 (rainfall amount by maximum 30-min rainfall intensity), like EI30 (rainfall energy by maximum 30-min rainfall intensity) could be taken as an indicator to express rainfall characteristic. Statistical analysis showed that erosion rate had a close relation with PI30 (Fig. 4).

3.2. Nutrient enrichment in sediment Several researchers reported that N and P in eroded sediment were significantly enriched (Monke et al., 1977; Flanagen and Foster, 1989; Young et al., 1986). Our data also showed that OM, TN, available P and NH4–N were greatly enriched in eroded sediment (Tables 3–5). The data in Tables 3–5 and Fig. 3 indicated that available P had the highest enrichment ratio with values ranging from 1.32 to 3.04, then followed NH4–N with values ranging from 1.11 to 1.94. The enrichment ratios of OM and TN in

Table 4 Nutrient contents in eroded sediments and their enrichment ratios from the combined erosion zones of sheet and rill during single rainstorm Rainfall parameters Amount (mm)

I30 (mm h1)

PI30 (mm2 h1)

24.2 13.6 18.2 27.9 108.0 75.2 11.5

46.8 27.0 7.8 27.0 69.0 16.2 69.0

1132.56 367.20 141.96 753.30 7452.00 1218.24 793.50

Erosion rate (t km2)

Sediment concentration (kg/m3)

Nutrient content OM (g/kg)

Total N (g/kg)

NH4-N (mg/kg)

P (mg/kg)

OM

Total N

NH4–N

P

6771 359 75 424 11218 1169 213

455 136 11 50 219 82 46

21.47 18.92 19.30 21.50 20.04 18.13 16.43

1.42 1.12 1.37 1.21 1.08 1.03 1.05

134.2 104.0 127.5 108.6 94.4 103.8 112.5

11.64 7.62 15.53 12.81 7.27 10.52 12.09

1.33 1.17 1.2 1.38 1.29 1.17 1.06

1.30 1.01 1.24 1.21 1.08 1.03 1.05

1.60 1.24 1.52 1.38 1.20 1.32 1.43

2.28 1.49 3.04 2.92 1.66 2.40 2.75

I30 means maximum 30-min rainfall intensity during a single rainstorm.

Enrichment ratios

92

F. Zheng et al. / Agriculture, Ecosystems and Environment 108 (2005) 85–97

Table 5 Organic matter, total N and NH4–N contents in eroded sediments and their enrichment ratios from the combined erosion zones of sheet, rill and shallow gully Rainfall parameters Amount (mm)

I30 (mm h1)

PI30 (mm2 h1)

64.3 59.6 10.4 38.8 82.8 15.8 18.0 17.0 13.7 48.8 54.8 28.7 24.2 13.6 18.2

27.6 76.8 60.0 46.8 30.0 47.4 43.2 51.0 13.2 48.0 76.8 56.4 46.8 27.0 7.8

1774.68 4577.28 624.00 1815.84 2484.00 748.92 777.60 867.00 180.84 2342.40 4208.64 1618.68 1132.56 367.20 141.96

Erosion rate (t km2)

Sediment concentration (kg/m3)

Nutrient content OM (g/kg)

Total N (g/kg)

NH4–N (mg/kg)

OM

Total N

NH4–N

2049 8337 2060 4271 3735 1033 1182 3310 161 7853 4589 7235 1981 776 97

592.4 686.0 802.0 441.8 130.3 912.0 766.3 32.7 170.9 305.0 912.0 527.0 183.5 67.0 484.4

21.59 21.42 21.20 20.14 21.93 22.90 17.78 21.07 19.56 19.34 19.84 16.57 15.97 18.16 17.86

1.33 1.37 1.35 1.20 1.43 1.48 1.21 1.35 1.07 1.11 1.45 0.97 1.10 1.23 1.17

177.1 169.8 136.1 147.4 160.9 163.1 117.7 165.5 120.3 113.4 137.3 107.2 134.7 126.1 108.7

1.12 1.08 1.13 1.12 1.22 1.22 1.11 1.32 1.22 1.21 1.25 1.16 1.29 1.35 1.23

1.01 1.04 1.09 0.97 1.15 1.19 1.03 1.15 0.91 0.95 1.23 0.97 1.10 1.23 0.89

1.21 1.16 1.21 1.31 1.43 1.45 1.38 1.94 1.41 1.33 1.61 1.36 1.71 1.60 1.38

Enrichment ratios

I30 means maximum 30-min rainfall intensity during a single rainstorm.

Regardless of different erosion zones, the relation between erosion rate and PI30 can be expressed as exponent function. The nutrient content and enrichment ratios in sediment greatly changed with the deforested year. Fig. 3 showed that available P contents constantly decreased from 23.84 mg/kg in the first year of 1989 to

9.34 mg/kg in 1996, while P enrichment ratios increased from 1.56 in 1989 to 2.14 in 1996. This could mainly attribute to severe topsoil nutrient loss in the early stage of deforestation (Table 2), and was partly caused by a great increase of soil erosion rate per unit of rainfall erosivity with the deforested year (Fig. 3).

Fig. 3. Soil erosion rate per unit of rainfall erosivity, available P contents in eroded sediments and enrichment ratios from the combined erosion zones of sheet, rill, and shallow gully (nutrient data in years of 1991 and 1994 were not available).

F. Zheng et al. / Agriculture, Ecosystems and Environment 108 (2005) 85–97

93

Fig. 4. Relationship between rainfall and erosion rate. (a) In zone 1; (b) in zone 1 + 2; (c) in zone 1 + 2 + 3 (data observed between 1989 and 1992); (d) in zone 1 + 2 + 3 (data observed between 1993 and 1996).

Fig. 5. Nutrient enrichment ratios in sediment at sheet erosion zones (data from the individual rainfall, see Tables 3 and 4 for rainfall characters).

94

F. Zheng et al. / Agriculture, Ecosystems and Environment 108 (2005) 85–97

Fig. 6. Nutrient enrichment ratios in sediment at combined erosion zones of sheet, rill and shallow gully (data from the individual rainfall, see Tables 3 and 4).

3.3. The relationship between sediment concentration and nutrient enrichment ratios in eroded sediment Young et al. (1986) pointed out that the nutrient enrichment ratio in eroded sediment decreased with an increase of sediment concentration. Our data showed the similar trend in the sheet erosion zone where sheet erosion was dominant (Tables 3 and 4, Fig. 5). However, in the combined erosion zones of sheet, rill and shallow gully where rill erosion and shallow gully erosion were dominant, the nutrient enrichment ratios in sediment initially decreased and then increased with an increase of sediment concentration (Fig. 6). The nutrient enrichment rations of OM, TN, NH4–N and available P reached their minimum when sediment concentration was between 400 and 600 kg/m3. For a given soil, particle or aggregate size in eroded sediment depends on the detachment and transport capacities of the runoff. In the sheet erosion zone, sheet flow has lower transport capacity compared to concentrated flow in the combined erosion zones. Lower rainfall amount and intensities (rainfall amount

15 mm or I30  10 mm/h) generated lower runoff discharge as well as the corresponding transport capacity (Table 6), resulting in lower sediment yield and erosion rate. Consequently, particles in eroded sediment were finer, and finer particles contain higher nutrient contents (Alberts and Moldenhauer, 1981), leading to the higher nutrient enrichment. As rainfall amount and intensity (rainfall amount >15 mm or I30 > 10 mm/h) increased, sheet flow rate and its corresponding transport capacity increased, resulting in greater sediment concentration and delivery. The fraction of coarse particles, i.e., silt or sand in eroded sediment increased. Thus, nutrient enrichment ratios in eroded sediment were lower. Table 6 showed that runoff rate, sediment concentration and erosion rate increased with rainfall amount and intensity. Fig. 7 demonstrated that fine particle content (<0.001 mm) in sediment decreased with erosion rate, while coarse particle content (>0.005) contrarily increased. Therefore, nutrient enrichment ratios in the eroded sediment decrease with an increase of sediment concentration in the sheet erosion zone. The similar result was also reported by Monke et al. (1977) despite differences in geographic locations.

F. Zheng et al. / Agriculture, Ecosystems and Environment 108 (2005) 85–97

95

Table 6 Runoff and erosion rates from selected rainfall events Erosion rate (t/km2)

Erosion zone

Rainfall amount (mm)

I30a (mm/h)

Runoff rate (mm)

Sediment concentration (kg/m3)

Sheet erosion

108.0 18.2

69.0 7.8

58.7 4.9

127.0 11.0

The combined of sheet and rill erosion zone

108.0 27.9

69.0 27.0

51.1 8.9

219 50.0

11218 424

The combined erosion zones of sheet, rill and shallow gully

108.0 75.2

69.0 16.2

43.7 31.1

790.4 64.0

14430 1989

a

6526 75.0

I30 means maximum 30-min rainfall intensity during a single rainstorm.

In the combined erosion zones of sheet, rill and shallow gully, lower rainfall amount and intensity (rainfall amount 15 mm or I30  10 mm/h) generated lower runoff discharge and its corresponding lower transport capacity. Therefore, particles in eroded sediment were mainly fine, resulting in higher nutrient contents and enrichment ratios in sediment. As rainfall amount and intensity (15 mm < rainfall amount  40 mm or 10 mm/h < I30  50 mm/h) increased, concentration flow, i.e., rill flow and shallow gully flow, which had greater transport capacity, was predominated. The fraction of coarse particles in the eroded sediment increases greatly, which caused the reduction of nutrient contents and enrichment ratios in eroded sediment. Furthermore, as rainfall amount and intensity (rainfall amount

>40 mm and I30 > 50 mm/h) further increased, rill flow, especially shallow gully flow transported soil clods, and large aggregates which contained rich aggregates. Fig. 8 showed that aggregate content in eroded sediment increased with erosion rate. This explained why nutrient contents and enrichment ratios in eroded sediment increased with an increase of sediment concentration (Tables 3–5). A mudflow was observed during large storm events at the combined erosion zones. When mudflow occurred, sediment concentration could reach 800– 900 kg/m3 and fine particle was greatly increased, particles of less than 0.001 mm accounting for 57.8% of total particle size, this might caused an increase of nutrient contents and enrichment ratios with increases of sediment concentration or erosion rates

Fig. 8. Comparison of sediment aggregate distribution between different erosion intensity (erosion rates are at 2551 kg/km2 during the single rainfall event in August 1, 1996 and at 11430 kg/km2 during the single rainfall event in July 29, 1995) and the topsoil in combined erosion zone of sheet, rill and shallow gully.

Fig. 7. Comparison of sediment particle distribution between different erosion intensity (erosion rates are at 404 kg/km2 during the single rainfall event in August 10, 1994 and at 2852 kg/km2 during the single rainfall event in June, 23, 1994) and the topsoil in sheet erosion zone.

96

F. Zheng et al. / Agriculture, Ecosystems and Environment 108 (2005) 85–97

(Tables 3–5). It could be seen from Fig. 6 that when sediment concentration was greater than 800 kg/m3, the nutrient enrichment ratio increased more quickly.

4. Conclusions This paper presents the field study of erosion patterns effects on nutrient loss in the newly deforested lands of the Loess Plateau, China, using monitoring data from field runoff plots. The results showed that after 7 years of deforestation, OM, TN, NH4–N and available P reduced by were 69, 46.7, 65.6 and 86.6%, respectively. OM, TN, NH4–N, and available P in eroded sediment had different nutrient enrichment. Available P had the highest enrichment ratios with values ranging from 1.32 to 3.04, followed by NH4–N with values ranging from 1.11 to 1.94. The results of this study further us to understand interaction deforestation, soil erosion process, soil degradation and environmental quality deterioration. The results of this study also challenge the current monitoring method of soil erosion using field runoff plots at hillslope scales on the Loess plateau. Due to the small size of the field runoff plots (normally size is 5 m  20 m), the soil loss observed from the plots only includes sheet erosion and rill erosion, does not cover ephemeral gully erosion. This could be a main reason for under-prediction estimated by current erosion prediction model. Therefore, research on ephemeral gully erosion and its effects on nutrient loss will provide valuable information for quantifying relationship soil erosion and land degradation. Although extensive efforts have been made to determine effects of factors on nutrient loss, the specific effects of erosion patterns have not been previously quantified. The understanding of erosion patterns effects on nutrient loss in the newly deforested lands may contribute to more effective control measures that can minimize the soil erosion and chemical loading to surface runoff at both hillslope and watershed scales.

Acknowledgements This study was funded by the CAS Knowledge Innovation (KZCX3-SW-422), National 973 Program

(2002CB111502) and National Natural Science Foundation of China (90302001 and 40335050).

References Alberts, E.E., Moldenhauer, W.C., 1981. Nitrogen and phosphorus transported by eroded soil aggregates. Soil Sci. Soc. Am. J. 45, 391–396. Alberts, E.E., Neibling, W.H., Moldenhauer, W.C., 1981. Transport of sediment nitrogen and phosphorus in runoff through cornstalk residue strips. Soil Sci. Soc. Am. J. 45, 1177–1184. Bai, H., Tang, K., Cha, X., 1991. Soil erosion and nutrient loss processes on slope farming land. Bull. Soil Water Conserv. 11 (3), 14–19 (in Chinese). Cai, Q., Wang, G., Cheng, Y., 1998. Erosion Process and Modeling in the Small Watershed of the Loess Plateau. Sciences Press, Beijing (in Chinese). Chen, Y., Jing, K., Cai, Q., 1988. Modern Erosion on the Loess Plateau and its control. Sciences Press, Beijing (in Chinese). Chen, Y., 1993. Hillslopes and Channel Erosion and Sediment Research. Meteorological Press, Beijing (in Chinese). Flanagen, D.C., Foster, G.R., 1989. Storm pattern effect on nitrogen and phosphorus losses in surface runoff. Trans. ASAE 32 (2), 535–543. Hamilton, H., Luk, S.H., 1993. Nitrogen transfer in a rapidly eroding agroeco-systems: Loess Plateau. China. J. Env. Qual. 22, 133– 140. He, X., Huang, Z., 2001. Zeolite application for enhancing water infiltration and retention in loess soil resources. Conserv. Recycl. 34, 45–52. He, X., Li, Z., Hao, M., Tang, K., Zheng, F., 2003. Down-scale analysis for water scarcity in response to soil-water conservation on Loess Plateau of China. Agricult. Ecosys. Environ. 94 (3), 355–361. He, X., Tang, K., Tian, J., Matthews, J.A., 2002. Paleopedological investigation in three agricultural loess soils on the Loess Plateau of China. Soil Sci. 167 (7), 478–491. Huang, B., 1953. The affecting soil erosion factors and patterns in the loess region of Shaanxi Province. Sci. Bull. 9, 57–63 (in Chinese). Jiang, Z., Jia, Z., Liu, Z., 1990. Rainfall and topography and soil erosion at hillslopes. J. Soil Water Conserv. 4, 76–86 (in Chinese). Jing, K., Chen, Y., Li, F., 1993. Sediment and Environment of the Yellow River. Sciences Press, Beijing (in Chinese). Johnson, H.P., Baker, J.L. 1982. Field-to-Stream Transport of Agricultural Chemicals and Sediment in an Iowa Watershed: Part I. Data Base for Model Testing (1976–1978). Report no. EAP-600/S3-82-032. Environmental Research Laboratory, Athens, GA. Johnson, H.P., Baker, J.L. 1984. Field-to-Stream Transport of Agricultural Chemicals and Sediment in an Iowa Watershed: Part II. Data Base for Model Testing (1979–1980). Report no. EAP-600/S3-84-055. Environmental Research Laboratory, Athens, GA.

F. Zheng et al. / Agriculture, Ecosystems and Environment 108 (2005) 85–97 Jun, W., Bojie, F., Yang, Q., Liding, C., 2003. Analysis on soil nutrient characteristics for sustainable land use in Danangou catchment of the Loess Plateau China. CATENA 54, 17–29. Laflen, J.M., Tian, J., Huang, C., 2000. Soil Erosion and Dryland Farming. CRC Press, Boca Raton, FL. Lal, R., 1976. Soil erosion on alfisols in Western Nigeria: IV. Nutrient element losses in runoff and eroded sediments. Geoderma 16, 403–417. Li, S., 1990. Relation between soil nutrient and soil loss. Yellow River 12 (2), 16–19 (in Chinese). Miguel, A.A., Clara, I.N., 2003. Soil fertility management and insect pests: harmonizing soil and plant health in agroecosystems. Soil Tillage Res. 72, 203–211. Monke, E.J., Marelli, H.J., Meyer, L.D., Dejong, J.F., 1977. Runoff, erosion and nutrient movement from intrerrill areas. Trans. ASAE 20 (3), 58–61. Morgan, R.P.C., 1995. Soil Erosion and Conservation. Longman, London. Nearing, M.A., Foster, G.R., Lane, L.J., Finkner, S.C., 1989. A process-based soil erosion model for USDA-water erosion prediction project technology. Trans. ASAE 32, 1587–1593. Peng, L., Wang, J., 1995. Soil nitrogen absorption, utilization and loss at dryland farming systems. J. Soil Erosion Soil Water Conserv. 1 (2), 9–16 (in Chinese). Peng, L., 1999. The progressive process, prospects and distribution of fertilizer use and grain production in China. Acta Agriculturae Boreali-occidentalis Sinica 8, 1–5 (in Chinese). Sharpley, A.N., Simth, S.J., Williams, J.R., Jones, O.R., Coleman, G.A., 1991. Water quality impacts associated with sorghum culture in the Southern Plains. J. Environ. Qual. 20, 239–244. Sharpley, A.N., Halvorson, A.D., 1994. The management of soil phosphorus availability and its impact on surface water quality. In: Lal, R., Stewart, B.A. (Eds.), Soil Processes and Water Quality. Lewis Publishers, FL, pp. 7–90.

97

Smolikowski, B., Puig, H., Roose, E., 2001. Influence of soil protection techniques on runoff, erosion and plant production on semi-arid hillsides of Cabo Verde, Agriculture. Ecosys. Environ. 87, 67–80. Tang, K., 1991. Soil Erosion on the Loess Plateau: Its Regional Distribution and Control. China Sciences and Technology Press, Beijing (in Chinese). Tang, K., Zhang, Z., Kong, X., Shi, R., Huang, Y., 1987. Soil erosion and soil degradation on the Loess Plateau. Bull. Soil Water Conserv. 7 (6) (in Chinese). Tang, K., Zheng, Z., Zhang, K., Wang, B., Cai, Q., Wang, W., 1993. Research methods on relationship between soil erosion and ecoenvironment in the Ziwuling forest area. Memoir of Northwestern Institute of Soil and Water Conservation, vol. 17. Chinese Academy of Sciences, pp. 3–10. Wang, W., Jiao, J., 1996. Rainfall, Erosion and Sediment of the Loess Plateau and Sediment Delivery of the Yellow River Basin. Sciences Press, Beijing (in Chinese). Wischmeier, W.H., Smith, D.D., 1978. Predicting rainfall erosion losses. Agricultural Handbook 537, USDA. Xu, J., 1999. Erosion caused by hyper-concentrated flow on the Loess Plateau. CATENA 36, 1–19. Young, R.A., Onless, A.E., Mutchler, C.K., 1986. Chemical and physical enrichments of sediment from cropland. Trans. ASAE 29 (1), 165–169. Zheng, F., Gao, X., 2000. Soil Erosion Processes and Modeling at Loessial Hillslope. Shaanxi People’s press, Xi’an (in Chinese). Zheng, F., Huang, C., 2002. Gully erosion. In: Lal, R. (Ed.), Encyclopedia of Soil Science. Marcel Dekker, ING, New York, U.S.A., pp. 630–634. Zhu, X., 1956. Soil erosion classification at loess region. J. Soil Sci. 4 (2), 99–116 (in Chinese). Zhu, X., 1981. Water erosion patterns and their affecting factors. Bull. Soil Water Conserv. 1 (3), 1–9 (in Chinese).