Soil moisture variation and land use on a small catchment

of the loess plateau, China

 

  Bojie Fu, Jun Wang, Yang Qiu, Liding Chen

Department of systems ecology, Research Center for Eco-Environmental Sciences, Chinese Academy of Sciences, P.O.BOX 2871, Beijing 100085, China

 

 

Abstract

 

    Soil moisture plays a critical role in both plant growth and vegetation restoration in semi-arid environments, its spatial and temporal variability results from topography, soils, vegetation and land use. However, a little knowledge exists about land use structure (pattern) on soil moisture variability. In order to analyze soil moisture variations in relation to land uses and its pattern, five land use structures and seven land use types were selected to monitor a programme for soil moisture in this paper. The soil particle size distribution and bulk density of seven land use types for additional information were also determined. Soil moisture measurements were performed biweekly at 26 locations in a small catchment (3.5km2) in the loess plateau of China from May to October 1998. The measurements were taken using Time Domain Reflectometry (TDR) at five depths of soil profile (0-5cm, 15-20cm, 25-30cm, 45-50cm and 70-75cm). The data were analyzed for soil moisture variations in time and space for seven land use types. Three-peak and three-valley for the variations of soil moisture during growing season was found. The influence of shrubland on mean soil moisture within 0-70cm was significant difference, comparison with cropland, orchard and intercropping land. Three types of soil moisture changes in profile were classified, increasing, decreasing and fluctuant types. An analysis of differences in soil moisture for five land use structures indicated that the influence of land use patterns on soil moisture were complex. The study provides insightful implications for hydrological modeling and runoff and erosion control in this area.

 

 

Keywords: soil moisture; land use structure; runoff and erosion control; loess plateau of China

 

 

Introduction

 

Soil moisture plays a critical role in plant growth and vegetation restoration in semi-arid environments. However, soil moisture exhibits highly variable in space and time. Consequently, high resolution ground-based monitoring is required to characterize these variations. At present, rapid and reliable measurements of soil moisture are possible with Time Domain Reflectometry (TDR), enabling us to carry out detailed measurement campaigns for spatial and temporal pattern of soil moisture in small areas (Dasberg and Dalton, 1985). This variability of soil moisture results from the differences in topography (Burt and Butcher, 1985), soils (Hawley et al., 1983), vegetation (Le Roux et al., 1995) and land use. A better understanding of the characteristics of soil moisture variability is important for improving hydrological models (Grayson et al., 1992) and land management in runoff and erosion control (Fitzjohn et al., 1998). Due to the importance of soil moisture, there have been a number of papers indicating the soil moisture variability (e.g. Anderson and Kneale, 1980; Sala et al., 1992; Bárdossy and Lehmann, 1998; Zhao et al., 1999). These studies seek and evaluate the factors controlling soil moisture, determine the significance in ecosystem processes and predict soil moisture in catchment or large scale. However, a little attention just is paid to influence of land use structure (pattern) on soil moisture. Evaluating the effects of land use and its pattern on soil moisture is difficult, because the differences in land uses which produce a change in the soil properties and evapotranspiration are likely to increase soil moisture variability across the landscape (Andrew et al., 1998).

The loess plateau of China has the highest rate of erosion in the world. The average and maximum erosion rates are 150 and 390 Mg ha-1 yr-1, respectively (Chen and Luk, 1989), which are equivalent to surface lowering of 1.2-3.1 cm yr-1. One of the main reasons for soil erosion is irrational land use (Fu, 1989; Fu and Gulinck, 1994). Soil moisture is an important factor for plant growth and erosion control, but it is short to indicate its spatial-temporal behaviors for different land uses in the loess plateau. Furthermore, because land uses can give rise to variations in soil physical and hydrological properties in relation to soil moisture, the creation of a mosaic pattern of land use may be advantageous in runoff and erosion control in this region. The purpose of this paper is to analyze the relationships between land use and soil moisture by means of intensive monitoring in space and time. In particular, our work focused on the following aspects:

(1)     To determine temporal variation of soil moisture for different land uses.

(2)     To analyze differences in soil moisture of profile for different land uses.

(3)     To study the spatial variations of soil moisture in five transects which reflect typical land uses in the study area.

(4)     To discuss the implications of soil moisture in relation to land use for the hydrological models, runoff and erosion control.

 

Materials and Methods

 

2.1. Study area

    The Da Nangou catchment (3653N, 10917E) is situated on the middle part of the loess plateau in northern Shaanxi province in China. The catchment has an area of 3.5km2 and an altitude between 1000-1350m. There are significant topographic variations with typical loess hills and gully landforms within the study area. A Digital Elevation Model (DEM) map from topographical map provides data on slope angles and relative elevation (Fig.1). Due to long-term human activity, natural vegetation has been destroyed. Land use types are slope cropland, fallow land, grassland, shrubland, orchardland and woodland consisting of mosaic patterns. Crops are mainly potatoes (Solanum tuberosum), beans (Phaseolus valgaris), maize (Zea mays L.) and millet (Panicum miliaceum). The forest, artificial woods, is dominated by locust trees (Robinia pseudoacacia L.). The grassland is mainly covered by annuals such as sweet wormwood (Artemisia annua L.), annual fleabane (Erigeron annuus Pers.) and sandy needlegrass(Stipa glareosa p. Smirn). Littleleaf peashrub (caragana microphylla) in shrubland and apple tree (Malus pumila mill) in orchard are present. Fallow land slowly came into being after cultivated plots were abandoned two and three years ago.

The region has a semiarid continental climate with an average annual temperature of 8.8. Monthly mean temperatures range from 22.5 in July to -7 in January. The average annual precipitation is 562mm with great interannual variability and 60 percent of the rainfall falls between July and September. There are 159 frost-free days and an average of 2415 hours of sunshine each year.

The soils, developing on wind-accumulated loess parent material, are thick at an average of 50-80m. There is no distinct B horizon of the soil profile developed. The active moisture change occurs in deeper layer due to relatively high permeability of this soil. Thus, from a runoff process perspective, soil moisture in the deeper layers has also important influence. The most common soil in the catchment is loessal with texture of silt ranging from 64% to 73% and clay varying from 17% to 20% (Table 1). It is weekly resistant to erosion. The erosion rate is extreme serious at about 10000-12000 tons km-2 yr-1 (Song et al., 1989).

 

 

2.2. Methods

According to different topographical position and land use, total 26 sample points were selected for measuring soil water content in the catchment (see Fig.1). Five transects, typical

 

 

Table 1.  Soil particle size distribution and bulk density in seven land use types

Land use type

2-0.05mm

(%)

0.05-0.002mm

(%)

<0.002mm

(%)

Ratio(sand/clay)

(%)

Bulk density

(g/cm3)

Cropland

14.80

65.62

19.58

75.55

1.28

Fallow land

13.44

67.73

18.84

71.33

1.25

Grassland

16.35

64.78

18.88

86.58

1.22

Woodland

10.79

70.79

18.43

58.53

1.25

Orchard

10.25

72.10

17.65

58.07

1.27

Intercropping land

13.50

68.57

17.93

75.28

1.21

Shrubland

13.70

69.13

17.17

79.81

1.24

 

 

land use structures in existence on the hillslope, were selected. The land use structure combinations from the top to foot of the hillslope were: cropland - cropland - cropland (north-west facing), cropland - woodland - orchard (south facing), fallow land - grassland – cropland  (north-east facing), fallow land - shrubland - intercropping land - woodland (north facing) and fallow land - shrubland - cropland - orchard (north facing).

    Measuring the soil moisture content using time domain reflectometry (TDR) is well proved and documented (Whalley, 1993). However, a calibration is necessary because the soils apparent dielectric constant depends not only on the water content, but also on factors such as bulk density, porosity and chemical composition of the soil water for a specific soil. In this study, the application of portable TDR (Eijkelkamp Agrisearch Equipment, The Netherlands) was confirmed by calibration through comparison with gravimetric water content measurements from saturated soil water content to permanent wilting point. These soil moisture contents were determined from 19 mixed soil samples with desired quality of water in the laboratory. The regression equation is

Y=0.8243X-0.0454, R2=0.9585,

where X-soil moisture content using TDR, Y-soil moisture content using oven-dry method multiplied by soil bulk density (Fig.2). The original moisture data obtained in the field were converted using this equation in the remainder analysis.

Soil moisture was measured using the portable TDR on 10 occasions during the growing season from May to October 1998 at approximately biweekly intervals. When measurements to use soil auger to the anticipated depth, four parallel steel rods (length 6cm, diameter 0.3cm, and spacing 2.5-3cm) were inserted vertically into the soil, and remained in position until the value displayed TDR was stable. Owing to destructive nature of the soil auger method, five random locations within a 2 m circle around each sample point, were taken to measure moisture content at five depths: 0-5cm, 15-20cm, 25-30cm, 45-50cm and 70-75cm. The mean moisture content


Digital Elevation Model of the Study Area

 



Spatial distribution of sample sites

 


Figure 1  Digital elevation model and spatial distribution of sample sites in the Da Nangou catchment


 


Figure 2.  Calibration of the TDR using regression analysis (solid line = regression line) variance

 

 

for five locations was computed as soil moisture of the sample site. Rainfall, total 465.42mm in study period, was recorded by an automatic datalogged raingauge located in the catchment. In order to evaluate the soil moisture response to rain event, the daily rainfall during observation

period was calculated in terms of the records of the raingauge (Fig.3).

Soil samples for determining particle size distribution were collected according to the horizons of soil profile at 17 locations. Particle size distribution were measured using traditional sieving methods to quantify the coarse grains (gravel) and then using hydrometer method to determine the particle fractions. The bulk density (g/cm3) for 20-25cm depth of each of the 26 sampling points using ring (diameter 5cm and height 5cm) was computed as the ratio of the mass of dry soil (g) to the volume of the sample (cm3) (Editorial Committee, 1996).


 

 


Figure 3  Temporal variations of mean soil moisture within 0-70cm in seven land use types. Also shown are SD value (%) and daily rainfall (mm)

 

 

Results

 

3.1. Temporal variations of soil moisture in different land uses

 

3.1.1. Temporal variations of mean soil moisture within 0-70cm

The temporal variations of mean soil water content within 0-70cm under seven land use types are shown in Fig.3. Also shown are the daily rainfall and the variance over time. The seasonal trends in the mean soil moisture are apparent. First, as expected, an increase and decrease in soil moisture corresponded to high and low rainfall, and its changes was characterized by three-peak and three-valley (Fig.3). Second, the mean moisture content reached a peak following a heavy rain event (on 20 and 21 May) and decreased thereafter. Moreover, a dry sequence appeared. Although several small rain events occurred between 30 May and 1 July, they did not interrupt the dry trend, unless moisture measurements were taken immediately after rain event. High evapotranspiration seems likely to the cause.

In general, the moisture content reaching peak value corresponded to the depth of precipitation, with higher mean moisture contents appearing after heavier rain. However, the differences in response to the rain of land use existed. For example, the peak in mean moisture content for woodland and intercropping land showed lag effect following a rain event (Fig.3). The main reason may be interception of tree crown and buffering influence of groundcover. Also, the difference in soil physical properties such as particle distribution and bulk density (Table 1) may contribute to the difference. Further close inspection, the mean soil moisture in shrubland was lower than in other land uses from 4 June to 2 September and between 1 and 15 October. There is more likely to the fact that caragana microphylla has deep and enormous roots for soil water intake to survive in dry environments (Wang and Li, 1989). The influence of land use types on soil moisture was further analyzed using analysis of (ANOVA) whose result showed significance under a=0.1 level. The results of multiple comparison indicated that the impact of shrubland on mean moisture was significant difference between cropland, orchard and intercropping land during study period (Table 2).

 

Table 2 Results of multiple comparison for mean soil moisture content of seven land use types during observation period (the same letter for two land use types represents significance under a=0.05 level between them)

Land use type

Mean soil moisture content (%, v/v)

Cropland

13.08 (A)

Fallow land

11.57

Grassland

11.60

Woodland

12.02

Orchard

13.12 (B)

Intercropping land

13.71 (C)

Shrubland

9.67 (A,B,C)

 

 

The temporal dynamics of the variance of mean soil moisture for 26 sampling points were difficult to characterize. The spring and summer months experienced more frequent rain events so that the mean moisture content was relatively similar to each of sample sites before 19 August. As a result, the variance exhibited low level. Compared with previous months, the variability was evident during September and October. Moreover, it was found that middle rain event occurred may result in higher variability under dry conditions (Fig.3). This result was not well consistent with the finding that heavier rains and higher mean moisture content were often associated with higher variability (Bell et al., 1980; Famiglietti et al., 1998). A possible explanation is that land use reported by them is uniform, but multiple land uses in our study. An alternative explanation is that surface soil moisture is studied in their papers, while mean soil moisture within 0-70cm in this paper.

 

3.1.2. Temporal variations of soil moisture for profile

The temporal variations of soil moisture for profile under seven land use types are shown in Table 3. Also shown is the variance of soil moisture within five depths during study period. Our results indicated that the variance (stand deviation SD) of soil moisture within upper layers (5cm, 15cm, 25cm) exhibited high level as reported by many authors (e.g. Anderson and Burt, 1978; Barling et al., 1994), partly due to the variability of meteorological conditions (rainfall, radiation, temperature). However, soil moisture in deeper layer with seasonal changes became relatively stable and greatly influenced by land use. The differences in the variability of soil moisture along depth existed among seven land uses. For example, the variance in moisture under cropland, fallow land and shrubland decreased with depth (Table 3). However, it showed fluctuation for five depths under the other four land uses comparison with above land uses. For example, woodland exhibited the highest value (SD is 5.7) in the variability of soil moisture under 70cm. This is quite likely due to the fact that trees have deeper roots to result in difference in the consumption for moisture.

Because of the temporal and profile sampling in our work, it is helpful to elucidate the temporal evolution of soil moisture for five depths under land uses, which enhances our ability to understand the influence of land use on soil moisture. Although the cropland, fallow land and shrubland shared the same characteristics in the variance of soil moisture with depth, soil moisture contents under cropland and fallow land increased with depth. And it in shrubland except for 5cm exhibited decrease trend with depth between 4 August and 15 October (Table 3). In addition, soil water content within 70cm in woodland reached to the peak value (23.12%) in 4 June and later gradually decreased. Comparison with woodland, although soil water content in orchard was also higher value (15.88%) within 70cm in 4 June, it did not show clear decreasing trend thereafter (Table 3). Rainfall and the difference in distribution of roots may contribute to the difference.

 

3.2. Profile variations of soil moisture in different land uses

 

The mean soil moisture contents of five depths under cropland, fallow land, grassland, woodland, orchard, intercropping land and shrubland during observation period were calculated during study period. Figure 4 provides the profile variations of soil moisture for seven land use types. Important differences in soil moisture along depth are apparent. Three types are classified based on their differences in soil moisture along depth. One is increasing type, and its water gradient (the value for soil moisture of lower layer minus upper layer) is greater zero. If the value is great zero which implies lower down infiltration trend for soil moisture, and vice versa. This type includes cropland, fallow land, intercropping land and grassland. Soil water content within 5cm was 8.6% in fallow land, 10.25% in cropland, 10.23% in intercropping land and 10.45% in grassland. Soil moisture in fallow land and cropland increased at the similar range with depth and almost showed two parallel curves in Fig.4. One possible explanation was that fallow land had similar soil physical properties to cropland (Table 1).

While soil moisture content in intercropping land gradually increased with depth from 10.23% within 5cm to 11.08% within 15cm. Then, it was raised greatly and came to 17.6% within 70cm. This revealed reciprocal advantages of intercropping system for improving soil moisture (Kiepe, 1995). Soil moiture content in grassland increased at relatively small range with depth (Fig. 4). Another type is fluctuant type whose water gradient value (>0 or <0) depends on soil moisture of two adjacent layers, consisting of woodland and orchard. Their soil moisture presented high-low-high-low-high change trend in profile. There were two potential explanations. First, the feedback effect of root distribution on soil moisture may contribute to the difference (Sala et al., 1992). Second, woods possibly transforme soil physical properties such as soil bulk density, physical composition and porosity (Zhu, 1993; José et al., 1995). These changes, in turn, influence on infiltration rate, storage and redistribution of soil water (Jiang, 1997; Kang et al., 1996). The decreasing type, third group, includes only one land use - shrubland. Its water gradient is less than zero. Soil water content declined with depth. It reduced from 11.26% within 5cm to 8.01% within 70cm.

Results of above analysis suggest that soil physical properties such as particle distribution and bulk density that varies jointly land use result in the differences in moisture for profile. Unfortunately, we only had the bulk density data within 20-25 cm and were unable to obtain the data along depth. However, several authors reported that the bulk density increased generally with depth in this area (e.g. Wu et al., 1991; Hou et al., 1995; Jiang, 1997). Some of them also observed that particle distribution did not exhibit systematical trend with depth. Our result is consistent with the finding (data not shown). These differences in soil physical properties along depth can be expected to yield variations in infiltration and water disposition qualities of the soil. However, their exact roles are more difficult to characterize. For example, cropland and fallow land exhibited similar soil physical attributes (Table 1), whereas soil moisture along depth in cropland showed high level. Did too orchard and woodland. Comparison with them, intercropping land with the lowest bulk density and middle sand/clay ratio showed higher soil moisture content below 25cm. So the exact role of the influence of soil physical properties on

Table 3 Temporal variations of soil moisture for profile in seven land use types

Land use

Depth

(cm)

5/23/98  6/4/98  6/19/98  7/1/98   8/4/98  8/19/98  8/30/98  9/15/98 10/1/98 10/15/98  SD soil   water  content  (%, v/v)

 

 

cropland

 

 

5

14.81

11.71

5.03

8.61

12.95

6.45

11.43

11.01

11.38

9.07

2.97

15

15.49

12.04

7.91

9.15

14.24

9.19

12.76

13.90

13.57

10.50

2.55

25

16.89

13.59

10.71

10.42

14.82

11.02

12.59

14.73

14.78

11.31

2.20

45

19.07

15.94

13.58

12.74

15.11

13.05

13.08

13.83

14.59

12.08

2.04

70

19.00

17.85

15.82

15.83

16.49

13.99

14.40

15.17

17.16

12.99

1.83

 

fallow

land

5

13.35

12.01

4.62

6.30

11.51

4.54

8.34

9.47

8.33

7.56

3.02

15

15.18

12.31

7.35

6.53

12.88

7.44

10.52

10.02

10.84

8.45

2.77

25

16.45

13.79

9.22

8.18

13.48

9.55

11.57

11.01

12.87

9.46

2.58

45

17.09

15.81

11.91

10.37

14.25

12.11

12.81

14.61

12.99

9.91

2.28

70

16.86

16.70

13.87

14.28

13.43

13.58

13.07

15.63

13.29

12.99

1.48

 

 

Grassland

5

16.73

17.49

8.70

7.54

15.61

6.23

7.41

4.15

9.67

10.97

4.65

15

17.30

15.70

8.02

6.74

15.74

8.28

8.99

5.61

10.84

10.63

4.10

25

16.15

15.51

9.63

7.69

14.30

9.32

8.59

7.55

11.95

10.87

3.19

45

15.99

15.14

11.80

11.31

13.68

11.31

11.25

10.04

12.22

11.20

1.92

70

16.31

14.75

15.90

12.23

12.36

9.80

11.45

14.95

13.43

10.67

2.24

 

 

Woodland

5

17.26

18.62

9.03

12.96

14.62

8.72

11.93

5.34

13.98

9.56

4.12

15

16.56

16.89

10.82

7.05

14.71

8.85

10.73

5.27

11.25

9.09

3.89

25

15.99

17.63

11.59

10.26

13.72

10.00

11.33

6.09

11.85

8.57

3.40

45

16.93

19.16

11.21

14.12

12.23

10.45

10.45

5.56

11.80

8.29

3.96

70

17.51

23.12

20.44

14.46

13.10

9.39

10.01

7.95

6.67

8.03

5.70

 

 

Orchard

5

16.34

14.77

6.86

12.33

18.27

8.98

14.06

7.08

15.62

13.12

3.93

15

16.30

13.87

8.27

9.95

17.41

8.52

13.52

8.58

15.53

12.97

3.44

25

17.21

14.89

10.51

11.12

14.61

10.69

12.70

10.06

16.83

14.64

2.67

45

16.67

15.33

11.25

12.20

12.17

11.93

10.52

12.74

14.50

14.96

2.01

70

14.06

15.88

13.92

14.17

13.24

11.12

10.32

17.70

11.56

16.34

2.38

 

intercrop-

ping land

5

10.46

12.69

7.23

7.96

11.70

10.46

8.34

7.74

15.95

9.79

2.69

15

11.29

12.28

7.05

6.90

13.59

10.54

11.95

8.20

17.44

11.53

3.19

25

11.37

14.34

10.08

9.14

14.25

16.83

15.00

11.29

23.47

13.23

4.12

45

15.41

14.83

13.40

11.70

15.00

19.99

18.13

12.11

20.80

16.12

3.09

70

17.88

16.73

14.17

14.91

15.08

21.59

21.59

16.40

17.72

20.02

2.69

 

 

Shrubland

5

18.54

8.90

2.25

5.01

12.52

4.33

17.11

22.41

12.11

9.44

6.59

15

17.72

10.71

4.69

4.03

11.12

4.86

9.60

21.39

12.77

8.01

5.68

25

17.88

10.96

5.27

4.20

10.87

3.79

7.74

19.00

13.43

6.11

5.49

45

18.05

12.69

7.00

4.49

8.90

4.03

7.28

14.29

4.12

5.78

4.81

70

16.31

14.34

9.64

5.76

8.57

3.29

7.00

9.72

1.31

4.12

4.75

Rainfall

(mm)

 

114.82

29.452

15.721

10.945

59.302

47.163

43.78

19.104

0.985

15.127

 

 

 

 

 

 

 

 

 

 

soil moisture under land uses need further investigation.

We classified soil moisture of profile for seven land uses into three types and indicated their water gradient. If water gradient value is less than zero, this implies a high potential for down infiltration. It is practical significance because the main runoff generating mechanism results from rainfall intensity exceeding the soil infiltration capacity in this area (e.g. Wang and Jiao, 1996; Liu and Kang, 1999). The decreasing type whose water gradient less than zero bears its implications for minimizing the frequency of runoff occurrence.

 


Figure 4.  Profile variations of mean soil moisture in seven land use types

 

 


3.3. Variations of soil moisture in land use structure

 

Comparison of changes in the mean soil water content for five land use structures leads to a number of considerations concerning the influence of them on soil moisture (Fig.5). According to soil water flow principle, soil moisture increases from the top to foot of hill slope in a homogeneous sloping plane (Hawley et al., 1983). Soil moisture change of one land use type (cropland) in a transect presented a stable increase from the top to foot of hill slope (Fig.5.A). Anderson and Kneal (1980) and Knapp et al. (1993) observed the similar pattern. The mean soil moisture content is 10.92% in this land use structure. However, soil moisture change trend in other land use structures was different from this result. In land use structure of cropland-woodland-orchard, soil water contents in upper-slope for cropland and down-slope for orchard were higher than that in middle slope for woodland (Fig.5. B). It was 12.03% in cropland, 10.57% in orchard and 7.07% in woodland. Mean soil moisture in woodland was low during growing period. It may be relative to higher potential evapotranspiration than that in cropland and orchard on south facing slope. Fig.5.C shows the change in soil moisture of land use pattern for fallow land-grassland-cropland. Although the variation show high-low-high trend from the top to foot of hillslope, 11.09 for fallow land, 10.82 for grassland and 11.1 for cropland, it is almost uniform across the slope. Soil moisture is correlative with slope degree and position (Ried, 1973). Fallow land and cropland with low degree locate in the top and foot part of hill slope, their soil water contents are relatively high. Due to grassland in middle part of slope, the down impulsive force of runoff from fallow land to grassland becomes stronger to result in infiltrating less water into soil, soil water content in grassland is low. In land use structure of fallow land – shrubland – intercropping land (terrace) – woodland (Fig.5.D), intercropping land had relatively high soil water content for 13.71%. It may be result from benching terrace for gentle slope and intercropping system for improvement of soil physical properties in favor of keeping soil moisture. While soil moisture content was further raised in the foot of hillslope, and was 16.98% in woodland. In fallow land – shrubland – cropland (terrace) – orchard structure (Fig.5.E), terrace softened slope degree and encouraged infiltration to make increase in soil moisture, whose average value was 15.2%. Although orchard located in down-slope, its soil water content was relatively low, 12.67%. It is quite likely that fruit trees require more water for growth, florescence and fruiting. Soil water contents were 12.21% and 9.67% in fallow land and shrubland, respectively.

We further explored the relationships between soil physical properties and moisture variability along slope by means of particle size distribution and bulk density. The particle distribution is highly variable along slope, and the slope position does not determine the particle distribution at 0-20 cm and 20-50 cm depth (data not shown). Bulk density is also variable and does not show any relationship with the slope (see Fig.5). These results inconsistent with previous finding by Famiglietti et al. (1998) suggest that land use produce a change soil physical properties which vary topography. This change, in turn, influences the soil moisture. For example, although cropland located in the foot of slope and showed higher bulk density (1.32 g/cm3), its soil moisture content is close to cropland with low bulk density (1.23 g/cm3) lied in upslope (Fig.5 C). Again, due to lower bulk density (1.20 g/cm3) in favour of water infiltration into soil, woodland in downslope position exhibited the highest soil moisture content (Fig.5.D). Because of the effects of land use, soil physical properties do not exhibit systematic trend along slope, distinguishing between the relative roles of topographic and soil properties in influencing soil moisture variability would require more intensive field monitoring or plot experiments that is beyond the scope of this work.

 


 



 

Figure 5.  The distribution of soil moisture contents on five land use structures. Also shown is the bulk density (g/cm3)

 

 

Discussions

 

4.1. Mechanism for controls on soil moisture variability at Na Nangou catchment

 

Since seven land use types were selected to monitor soil moisture in the catchment, their influences on soil moisture variations may be over or underestimated with respect to one land use across the catchment. Additionally, that more rain events occurred in spring during our study year is different from normal year when rains often occur in summer. With these bear in mind, we summarize the mechanistic controls on soil moisture variability within Da Nangou catchment.

Because there is no any other source to add soil moisture except for precipitation (Wu, et al., 1991; Li et al., 1998), precipitation and evapotranspiration, two opposite and complementary processes to each other, jointly control the level of soil moisture. Moreover, their seasonal characteristics determine the temporal trend of soil moisture. Land use can give rise to the differences in evapotranspiration and soil physical properties (Table 1) which influence the soil moisture dependent upon rainfall. Therefore, land uses exhibited in differences in seasonal and profile trend of soil moisture. The seasonal dynamics in mean soil moisture within 0-70cm for seven land uses loosely mimicked precipitation depth, and can be classified three-peak and three-valley. During May, June, July and August, although mean soil moisture content exhibited fluctuation, the spatial variance was low (Fig.3), and soil moisture was relatively uniform across the catchment. There is likely due to the fact that differences in evapotranspiration and soil properties of land uses were limited by frequent rains. Conversely, high spatial variance in mean soil moisture existed during September and October. In addition, the difference in peak value of soil moisture between woodland and intercropping land and other four land uses may be explained by differences in soil physical attributes and the buffer effect of ground cover.

Land use types together with rainfall controlled seasonal variability in mean soil moisture of profile. Higher variability with upper layers (0-25cm) is partly due to meteorological condition (Table 3). However, land use has main impact on soil moisture with depth. Clearly, differences in distribution of roots and soil physical properties will also contribute to the difference in temporal variability of soil moisture for profile.

Along the slope, soil properties and topography continue to jointly influence moisture, and the impact of topography on soil moisture is apparent, especially for one land use distributed in slope (Fig.5 A). However, multiple land uses disturb this influence of topography (Fig. 5, B, C, D, and E). Results of the particle distribution and bulk density indicate that they do not exhibit systematically trend along slope under the effects of land use. These would result in a fluctuation in soil moisture from the top to foot of slope.

 

4.2. Implications for modeling

 

Although this analysis is preliminary, it provides comprehensive implications for modeling. Soil moisture content is a state variable which is either simulated or required as input for some hydrological models such as distributed model. It is important that watershed is subdivided into spatial elements as homogeneous areas in hydrological response as possible. These factors influencing on soil moisture are the major reference in determining hydrologic response. Hawley et al. (1983) reported that knowledge of the topographic variation of surface (especially in relative elevation) might be used as substitute for soil moisture in determining areas of homogeneous hydrological response. But application to multiple land uses in slope and catchment scale is problematic, because multiple land uses can increase soil moisture variability. Our results in the loess plateau of China demonstrated that soil moisture changes differed in five land use structures (Fig.5), and that land use was relatively great contributors to profile change of soil moisture observed in this study (Fig.4). Comparison with cropland, orchard and intercropping land, the difference in mean soil moisture within shrubland was significant during study period (Table 2). Again, soil moisture for profile of seven land uses exhibited three types (Fig. 4) which implied different potential for down infiltration. So determining homogenous hydrological units is not only considered the topographical variations, but also for land use and its patterns. In addition, due to land use disturbance of the systematic trend of soil physical properties along slope, the effects of land use pattern on soil moisture distribution is complex (Fig.5). The average soil moisture content on a slope ranged from 10.43% (Fig.5.B) to 13.62% (Fig.5.E). Averaging soil moisture values over the slope or neglecting the profile feature of land use for modeling may lead to error.

 

4.3. Implications for land management

 

Land management primarily aims to erosion control and improvement in soil moisture for plant in semi-arid environments. In loess plateau of China, two distinct characteristics, severe erosion and deficit of soil moisture, seriously restrict the productivity of land. How to control on them is the major target in this region.

The principal runoff generating mechanism is rainfall intensity exceeding the soil infiltration capacity in this region. This infiltration capacity seems to be controlled by soil moisture and soil properties (Hou et al., 1995; Jiang, 1997; Boix-Fayos et al., 1998), since the rate of infiltration into the soil surface is primarily a function of near-surface soil properties and antecedent moisture (Philip, 1957). However, land uses can produce changes in soil physical attributes and evapotranspiration to result in soil moisture change. Nevertheless, our results indicated that seven land use types exhibited the differences in temporal dynamics and profile feature of soil moisture (Fig.3, 4, and 5, and Table 2).

Therefore, homogeneous areas in hydrological response termed of hydrological response unit (Flugel, 1995) may be divided based on land use type and its location at the slope. For a specific hydrological response unit, it may exhibit a threshold value determined by the infiltration intensity (Liu and Kang, 1999) for runoff to occur. Overland flow from a unit will only occur when the infiltration threshold is exceeded for a specific storm. At slope scale, the potential surface runoff areas are dependent on their location and on their arrangement. When there are differences in initial soil moisture and infiltration rate between these close areas, the time of surface runoff beginning (Zhang and Liang, 1995) is different and runoff producing areas are spatially isolated. So surface runoff of upper slope will be re-absorbed by the surrounding drier or higher infiltration rate areas which act as sinks for overland flow and transported sediments, and the runoff from the top to bottom of slope will not occur. At the catchment scale, widespread runoff and erosion must be overcome the spatial arrangement and threshold values of hydrological response units at all small scales and require prolonged or larger magnitude storms. Therefore, creating a mosaic pattern of areas with contrasting hydrological response may be an effective management strategy in runoff and erosion control (Fitzjohn et al., 1998; Fu et al., 1998). Mosaic patterns can be achieved by arrangement of land use.

For a heavy storm, the rainfall intensity exceeds the majority of infiltration threshold of the individual hydrological response units, larger areas will be contributing to surface runoff regardless of the spatial distribution of land use. In this situation, the differences in initial soil moisture and infiltration rate may not be essential. However, initial soil moisture (Zhang and Liang, 1995; Andrew and Rodger, 1998) and infiltration rate (Liu and Kang, 1999) are important in the process producing runoff in many situations. This suggests that the differences in soil moisture and infiltration affected by land use and its pattern may also be advantageous in runoff and erosion control. The similar land use along a slope showed the similar infiltration rate (Jiang, 1997), increase in soil moisture from the top to foot of the slope (Fig.5.A, C) and increase within profile (Fig. 4) may easily give rise to source areas producing runoff connected and continuous hydrological pathways. This would result in the possibility for widespread runoff and erosion over slope. In addition, due to similar harvesting period for crops, the soils exposed are weakly resistant to water and wind, and are often highly erodible with severe erosion occurring over very short distances. We can establish buffer zones (different vegetation strips) for re-absorbing the runoff and trapping sediments from upslope (Morgan, 1992) in this land use structure. However, soil moisture variations in other three land use structures are different from above two ones (Fig. 5). Contrasted hydrological response along slope, resulting from initial soil moisture including profile (Fig.4) and infiltration rate (Li et al., 1995; Kang et al., 1996), makes the differences in time producing runoff in different land uses. There is no connectivity between runoff producing areas on a slope to minimize runoff and erosion on many situations. It is quite likely that land use mosaic can create a self-regulating system in controlling runoff and erosion. For example, spatial variations in erosion intensity is generally controlled by topography and exhibits an increase trend of vertical zonal distribution from the top to bottom of slope (Tang, 1999) in this area. However, several authors found by means of field survey and modeling that rational arrangement of land use destroyed this law and showed patch distribution of erosion intensity in relation to land use in this area (e.g. Jiang et al., 1996; Dong et al., 1998). Consequently, the more attention should be paid to the selection and arrangement of land use on a slope and catchment scale (Wu and Yang, 1998) based on spatial soil

moisture pattern and ecological properties of plant species. Seeding drought-tolerated shrubs and grasses at upslope, benching terrace for crops and intercropping system in middle slope and planting trees (fruit trees) in the down-slope and bottom of gully, may be a more appropriate land use structure to soil and water conservation in the study area.

Soil moisture is one of primary limiting factors for plant growth in semi-arid areas. Some measures such as terrace aim to encourage infiltration for increasing soil water content. Our results indicated that the mean soil water content in terrace for cropland (15.2%) was higher than that (11.1%) in slope cropland. While intercropping system, exerting mutual benefits of tree and crop, is also a better tillage system in semi-arid areas (Kiepe, 1995). In addition, soil moisture with 0-25cm with great change is strongly affected by precipitation and evaporation (Table 3), so selecting growing period of crops similar to rainy season can improve availability of rainfall.

 

 

Conclusions

 

Variability in seasonal dynamics and profile feature of soil moisture in relation to land use and its pattern was studied at 26 locations at Da Nangou catchment in the loess plateau of China. Land use and its pattern also contribute to the soil moisture variations except for rainfall, topography and soil properties. The temporal dynamics for seven land use types showed three-peak and three-valley during growing season. Lag influence on soil moisture was found in woodland and intercropping land. The influence of shrubland on mean soil moisture content during study period was significant difference comparison with cropland, orchard and intercropping land. Three types for soil moisture variations in profile in seven land use types were classified. Increasing type included cropland, fallow land, intercropping land and grassland. Fluctuant type had woodland and orchard. Shrubland was present in decreasing type. Soil moisture variability differed in five land use structures, which indicated that the influences of land use pattern on soil moisture were complex. Although the spatial scale of the study is small, this research has implications for a range of issues in hydrological modeling and land management. First, a thorough knowledge of hillslope-scale soil moisture variability will provide a foundation for better understanding hillslope hydrological, ecological and biogeochemical processes, many of which are nonlineraly related to soil moisture content. Second, since hillslopes are fundamental landscape units, this work will provide a basis for characterizing soil moisture variations at larger scales. Third, understanding the relationships between soil moisture and land use will be helpful to improve the spatial arrangement of land use and erosion control in the loess area of China.

 

 

Acknowledgments

 

The project was supported by the National Natural Science Foundation of China (contract No. 49725101 and INCO-DC of European Commissioncontract No.ERBIC18CT970158. The authors wish to acknowledge the members of project team for measuring the soil moisture in the field together. Gratitude is expressed to two anonymous reviewers for their useful comments, which did much to improve the original manuscript.

 

 

References

 

Anderson, M.G., Burt, T.P., 1978. Toward a more detailed field monitoring of variable source areas. Water Resources Research 14, 1123-1131.

Anderson, M.G., Kneale, P.E., 1980. Topography and hillslope soil water relationships in a catchment of low relief. Journal of Hydrology 47, 115-128.

Andrew, W.W., Günter, B., Rodger, B.G., 1998. Geostatistical characterisation of soil moisture patterns in the Tarrawarra catchmetn. Journal of Hydrology 205, 20-37.

Bárdossy, A., Lehmann, W., 1998. Spatial distribution of soil moisture in a small catchment. Part 1:

geostatistical analysis. Journal of Hydrology 206, 1-15.

Barling, R.D., Moore, I.D., Grayson, R.B., 1994. A quasi-dynamic wetness index for characterizing the spatial distribution of zones of surface saturation and soil water content. Water Resources Research 30, 1029-1044.

Bell, K.R., Blanchard, B.J., Schmugge, T.J., Witczak, M.W., 1980. Analysis of surface moisture variations within large field sites. Water Resources Research 16(4),796-810.

Boix-Fayos, C., Calvo-Cases, A., Imeson, A., Soriano-Soto, M., Tiemessen, I., 1998. Spatial and short-term termporal variations in runoff, soil aggregation and other soil properties along a mediterrranean climatological gradient. Catena 33, 123-138.

Burt, T.P., Butcher, D.P., 1985. Topographic controls of soil moisture distribution. Journal of Soil Science 36, 469-486.

Chen, Y.Z., Luck, S.H., 1989. Sediment sources and recent changes in the sediment load of yellow River, China. In: Rindwanich, S. (Ed.), Land conservation for Future Generations. Ministry of Agriculture, Bangkok, pp. 313-323.

Dasberg, S., Dalton, F.N., 1985. Time domain reflectometry field measurements of soil water content and electrical conductivity. Soil Science Society of America Journal 49, 293-297.

Dong, R., Zhu, X., He, Z., Wan, T., Wang, X., 1998. Laws of soil erosion in loess hilly and Gully region of Dingxi prefecture. Bulletin of Soil and Water Conservation 18(3), 1-15, in Chinese.

Editorial Committee 1996 Soil Physical and Chemical Analysis & Description of Soil Profiles. Standards Press of China, Beijing, pp. 5-151 (in Chinese).

Famiglietti, J.S., Rudnicki, J.W., Rodell, M., 1998. Variability in surface soil content along a hillslope transect: Rattlesnake Hill, Texas. Journal of Hydrology 210, 259-281.

Fitzjohn C., Ternan J.L., Williams, A. G., 1998. Soil moisture variability in a semi-arid gully catchment: implications for runoff and erosion control. Catena 32, 55-70.

Flugel, W.A., 1995. Delineating hydrological response units by geographical information system analyses for regional hydrological modelling using PRMS/MMS in the drainage basin of the river Brol, Germany. Hydrological Processes 9, 423-436.

Fu, B., 1989. Soil erosion risk and its control in the loess plateau of China. Soil Use and Management 5, 76-82.

Fu, B., Gulinck H., 1994. Land evaluation in area of severe erosion: The loess plateau of China. Land Degradation & Rehabilitation 5(1), 33-40.

Fu, B., Ma, K., Zhou, H., Chen, L., 1999. The effect of land use structure on the distribution of soil nutrients in the hilly area of the loess plateau, China. Chinese Science Bulletin 44, 732-736.

Grayson, R.B., Moore, I.D., McMahon, T.A., 1992. Physically based hydrologic modeling. 1. A terrain-based model for investigative purposes. Water Resources Research 28, 2639-2658.

Hawley, M.E., Jackson, T.J., McCuen, R.H., 1983. Surface soil moisture variation on small agricultural watersheds. Journal of Hydrology 62, 179-200.

Hou, X., Bai, G., Cao, Q., 1995. Contrast study on soil infiltration capacity and anti-scourability in Robinia Pseudoacacia, Caragana Microphylla and Hippophae Rhamnoides woodlands. Journal of Soil and Water Conservation. 9(3), 90-95, in Chinese.

Jiang, Z., Wang, Z., Liu, Z., 1996. Quantitative study on spatial variation of soil erosion in a small watershed in the loess hilly region. Journal of Soil Erosion and Soil and Water Conservation 2(1), 1-9, in Chinese.

Jiang, D. 1997. Soil erosion and control models in the loess plateau. Water Resources Press, Beijing (in Chinese).

José, M.F., Francisco, L., Julia, M., Asunción, R., 1995. Land use and soil-vegetation relationships in a Mediterranean ecosystem: El Ardal, Murcia, Spain. Catena 25,153-167.

Kang, S., Zhang, S., Nie, G., Shi, S., Gou, Z., Qi, Z., Cui, X., 1996. Research on soil infiltration distribution of Aobao water basin in Inner Mongolia. Journal of Soil Erosion and Soil and Water Conservation 2(2), 38-46, in Chinese.

Kiepe, P., 1995. No runoff, no soil loss: soil and water conservation in hedgerow barrier systems. Wageningen, pp. 1-42.

Knapp, A.K., Fahnestock, J.T., Hamburg, S.P., Statland, L.B., Seastedt, T.R., Schimel, D.S., 1993. Landscape patterns in soil-water relations and primary production in tallgrass prairie. Ecology 74, 549-540.

Le Roux, X., Bariac, T., Mariotti, A., 1995. Spatial partitioning of the soil water resource between grass and shrub components in a West African humid savanna. Oecologia 104, 147-155.

Li, G., Luk, S.H., Cai, Q.G., 1995. Topographic zonation of infiltration in the hilly loess region, North China. Hydrological processes 9, 227-235.

Li, H., Wang, M., Cai, B., 1998. Study on characteristics of soil water of planted forest and its relation to precipitation in northwestern Shannxi. Journal of Soil Erosion and Soil and Water Conservation 4(4), 60-65, in Chinese.

Liu, X., Kang, S., 1999. Some developments and review of rainfall-infiltration-runoff yield research. Bulletin of Soil and Water Conservation 19(2), 57-62, in Chinese.

Morgan, R.P.C., 1992. Soil conservation options and in the U.K. Soil Use Manage 8, 176-180.

Philip, J.R., 1957. The theory of infiltration, 5. the influence of the initial soil moisture content. Soil Science 83, 329-339.

Ried, I., 1973. The influence of slope orientation upon the soil moisture regime and its hydrogeomorphological significance. Journal of Hydrology 19, 309-321.

Sala, O.E., Lauenroth, W.K., Parton, W.J., 1992. Long-term soil water dynamics in the shortgrass steppe. Ecology 73(4), 1175-1181.

Song, G.Q., Li, L.T., Guo, F.G., Zhao, M.L. 1989. Land classification of experiment and exemplary areas on the loess plateau. Memoir of Northwestern Institute of Soil and Water Conservation, Academia Sinica and Ministry of Water resources 10, 1-13, in Chinese.

Tang, K., 1999. Characteristics and perspectives on scientific discipline of soil erosion and soil and water conservation in China. Research of Soil and Water Conservation 6(2), 2-7, in Chinese.

Wang, M.B., Li, H.J., 1989. Study on soil water ecological environment of artificial Caragana Korshinskii bushwood. Memoir of Northwestern Institute of Soil and Water Conservation, Academia Sinica and Ministry of Water resources 10, 155-160, in Chinese.

Wang, W., Jiao, J., 1996. Statistic analysis on variation of rainfall and runoff-sediment yield process on slope surface in loess plateau region. Bulletin of Soil and Water Conservation. 16(5), 21-28, in Chinese.

Whalley, W R. 1993. Considerations of the use of time-domain reflectometry (TDR) measuring soil water content. Journal of Soil Science 44,1-9.

Wu, Q., Liu, X., Zhao, H., 1991. Soil physical properties and its water characteristics in mountain poplar forest land. Memoir of Northwestern Institute of Soil and Water Conservation, Academia Sinica and Ministry of Water resources 14, 79-95, in Chinese.

Wu, Q., Yang, W. (Eds.), 1998. Vegetation construction and sustainable development in the loess plateau of China. Science Press, Beijing, pp. 1-68 (in Chinese).

Zhang, G., Liang, Y., 1995. Study on runoff beginning time of artificial grassland in loess hilly region. Journal of Soil and Water Conservation 9(3), 78-83, in Chinese.

Zhao, X., Wu, F., Liu, B., Liu, S., 1999. Effects of primary factors on soil moisture in cultivated slopeland on loess plateau. Bulletin of Soil and Water Conservation 19(1), 10-14, in Chinese.

Zhu, Z.C., 1993. The main characteristics of the vegetation and its impact on the soil essence in the loess plateau of northern Shaanxi province. Acta Phytoecologica et Geobotanica Sinica 17(3), 280-286, in Chinese.