Geoderma Regional 7 (2016) 251–258
Contents lists available at ScienceDirect
Geoderma Regional journal homepage: www.elsevier.com/locate/geodrs
Particle size distribution of soils (0–500 cm) in the Loess Plateau, China Chunlei Zhao a,c, Ming'an Shao a,c,⁎, Xiaoxu Jia b,c,⁎⁎, Chencheng Zhang c a b c
College of Natural Resources and Environment, Northwest A&F University, Yangling 712100, China Key Laboratory of Ecosystem Network Observation and Modeling, Institute of Geographical Sciences and Natural Resources Research, Chinese Academy of Sciences, Beijing 100101, China State Key Laboratory of Soil Erosion and Dryland Farming on the Loess Plateau, Northwest A&F University, Yangling 712100, China
a r t i c l e
i n f o
Article history: Received 22 December 2015 Received in revised form 21 May 2016 Accepted 25 May 2016 Available online 26 May 2016 Keywords: Particle size distribution Fractal dimension Spatial variation Deep loess soil Typical Loess Plateau
a b s t r a c t To investigate the spatial variability of soil particle size distribution (PSD) in the Loess Plateau (LP) region of China, 2673 disturbed soil samples were collected in the 243 soil profiles (0–500 cm) across the typical loess zone. The PSD of soil samples were determined using the laser diffraction technique and the regional spatial distribution patterns of PSD were analyzed through classical statistical and geo-statistical methods. The results showed that silt loam was the dominant soil texture (92.6%) at the 0–500 cm soil layer. Sand, silt and clay contents varied slightly with increasing soil depth, suggesting that soil texture was almost homogeneous across the soil profile. Soils were overall sandy in the north and clayey in the south, but soil texture variation uneven with increasing latitude. PSD pattern across the typical LP region depicted latitudinal zonality. Fractal analysis showed a strong relationship between fractal dimension (D) and clay content (R2 = 0.98), demonstrating that D was controlled by the clay content rather than the coarse particles at the regional scale. The limited changes of PSD in the soil profile and the moderate variation of soil texture across the LP will provide a reference to regional scale hydrological, erosion and ecological models in the Loess Plateau of China. © 2016 Elsevier B.V. All rights reserved.
1. Introduction Soil particle size distribution (PSD) is an important soil physical parameter that affects soil conservation, water and nutrient movement, vegetation productivity and thus ecological restoration (Pan et al., 2010). PSD is widely used in soil classification and in estimating soil hydraulic properties such as soil water retention curve, soil hydraulic conductivity and soil bulk density (Filgueira et al., 2006; Hwang et al., 2011; Hollis et al., 2012; Antinoro et al., 2014). Besides, different PSD-driven sorption properties of soil could affect the mineralization of decoupled carbon and nitrogen and the activity of invertase and xylanase during organic matter decomposition (Stemmer et al., 1999; Bimueller et al., 2014; Zhou et al., 2015). PSD is therefore important for understanding the physical and chemical processes of soil development (e.g., soil water and nutrient cycles) in terrestrial ecosystems. Fractal geometry is a tool widely used to build insight into complex natural phenomena such as soil/rock mechanics and physics (Millan et al., 2003). Also, fractal dimension (D) is extensively used in explaining the scaling domains of PSD (Tyler and Wheatcraft, 1989, 1992; Filgueira et al., 2006; Zhao et al., 2009). Soil fractal analysis can be
⁎ Correspondence to: M. Shao, College of Natural Resources and Environment, Northwest A&F University, Yangling 712100, China. ⁎⁎ Correspondence to: X. Jia, Key Laboratory of Ecosystem Network Observation and Modeling, Institute of Geographical Sciences and Natural Resources Research, Chinese Academy of Sciences, Beijing 100101, China. E-mail addresses:
[email protected] (M. Shao),
[email protected] (X. Jia).
http://dx.doi.org/10.1016/j.geodrs.2016.05.003 2352-0094/© 2016 Elsevier B.V. All rights reserved.
used to define PSD which reflects the degree of uniformity of soil texture. For example, Su et al. (2004) noted that D can be used to describe PSD characteristics and its relationship with land desertification. Wang et al. (2008) showed that multi-fractal dimensions reflect soil physical properties as well as soil quality, and also significantly correlated with land use. Zhao et al. (2009) concluded that fractal analysis can be used to evaluate the effect of check-dam on soil texture deposited on farmland soils. The study demonstrated the feasibility of using fractal dimension of PSD in the analysis of soil detachment, soil erosion or even soil desertification. The Loess Plateau (LP) region of China (with an area of 62 × 104 km2) has unique landscape over deep loess deposits with intensive soil erosion since ancient times (Chen et al., 2007; Zhao and Xu, 2013). The intensive soil erosion is increasingly limiting the land productivity, inducing environmental degradation and the rise of riverbeds in the lower reaches of the Yellow River (Shi and Shao, 2000). This has increased the need for reliable PSD analysis as a measure to present erosion and/or land degradation in the region. To control soil erosion and improve land quality, China initiated a state-funded program called “Grain-for-Green” in the LP region in 1999 (Chen et al., 2007). Increasing vegetation cover by planting trees and/or establishing grassland is an effective measure of reducing soil erosion in the region. However, vegetation restoration vis-à-vis soil erosion control drives are limited by water movement and soil water content in both the shallow and deep soil layers (Mengistu, 2009). The need for information on fresh PSD is further evident at regional-scale hydrological modeling of the adaptability and sustainability of vegetation
252
C. Zhao et al. / Geoderma Regional 7 (2016) 251–258
restoration drive such as the “Grain-for-Green” project funded by the state. Moreover, the evaluation of soil water availability, water holding capacity, deep soil desiccation and soil bulk density or its pedo-transfer operations requires PSD in diverse soil layers (Sun and Yang, 2013; Wang et al., 2014). Although critical for successful and sustainable restoration of vegetation, fresh regional-scale data on PSD and its fractal features are scarce for the LP region. Therefore, the objectives of this study were to: 1) collect profile PSD data for the 0–500 cm soil layer in the LP region in China; 2) investigate regional spatial variability of PSD across a typical LP region; and 3) characterize fractal features of PSD for various soil layers in the LP study area.
2. Materials and methods 2.1. Study area The LP region (34°00′–45°05′N, 101°00′–114°33′E) is the largest loess zone (total area of 62 × 104 km2) in the world and it is located in North Central China (Fig. 1). The LP study area stretches across an altitude range of 200–3000 m, usually over a sediment layer of thickness of 30–80 m (sometimes 150–180 m). While annual mean precipitation (MAP) range from northwest to southeast is 150–800 mm, mean annual temperature (MAT) range from northwest to southeast is 3.6–14.3 °C (Fig. 1). The order of vegetation zones along the southeast-northwest transect of the LP study area is forest, forest-steppe, typical-steppe, desert-steppe and steppe-desert. The order of distribution of dominant soil types from the south to the north of the Loess Plateau is Haplic Luvisols, Terric Anthrosols, Calcic Chernozems, Aridic Leptosols, Calcaric Regosols, Calcic Kastanozems and Aridic Arenosols (FAO/ISRIC/ISSS, 1998). Further details about the study area are document by Shi and Shao (2000) and He et al. (2003). This study was conducted in a typical LP region that is about 2/3 (43 × 104 km2) of the Loess Plateau (Fig. 1). Specifically, the typical LP region is the spatial location covering Shanxi, Shaanxi, Gansu, Ningxia, Henan and Inner Mongolia. The distribution of loess is most continuous (in horizontal and vertical space), thickest and has vast loess geomorphic landforms in this region. Land degradation and soil fertility loss are common in most of the LP study area due to severe soil erosion and water scarcity. The main geomorphic landforms are Yuans (large flat surfaces with little or no erosion), ridges, hills and gullies.
2.2. Soil sampling To accurately determine PSD that is representative of the typical LP region, we devised an intensive soil sampling scheme close to the grid-cell level of the entire study area. For more convenient access to the sampling sites, we chose sampling routes by considering the road transportation systems in the LP study area. The distances between adjacent sites were approximately 40 km. But for better representation of areas with complex landscape and geomorphology, we reduced the sampling distance by half to include at least one additional randomly selected site along the 40 km sampling distance. Each randomly selected sampling point represented the main land use, soil type and topography within the range of sight. To reduce the effect of the roads on the collected data, sampling points were located at least 150 m away from the road. Thus a total 243 sampling sites were selected and correctly located using a GPS receiver (5 m precision). The locations of the 243 sampling sites across the LP study area are plotted in Fig. 1. For each site, disturbed soil samples were collected using a soil auger (5 cm in diameter) and 11 soil samples collected in different soil layers along the soil transect to the depth of 500 cm. The profile soil samples were collected at the 0–10, 10–20, 20–40, 40–60, 60–80, 80–100, 100–150, 150–200, 200–300, 300–400 and 400–500 cm soil layers. From June to October 2012, we visited 243 sampling sites in the typical LP region and a total of 2673 disturbed soil samples were collected for laboratory analyses. 2.3. Laboratory analyses The disturbed soil samples were air-dried and passed through a 2mm mesh for PSD analysis. Hydrogen peroxide was used to remove organic matter and sodium hexametaphosphate used to disperse the samples. After the processing, PSD (volume fraction) of each sample was measured by laser diffraction using the Mastersizer 2000 (Malvern Instruments, Malvern, England) as described by Jia et al. (2013) and Wu and Lu (2012). PSD was classified based on clay (b0.002 mm), silt (0.002–0.05 mm) and sand (0.05–2 mm) contents as in the soil taxonomy developed by U.S. Department of Agriculture (USDA). 2.4. Soil fractal theory According to Katz and Thompson (1985) and Tyler and Wheatcraft (1989, 1992), when a soil particle size is bigger than Ri (i.e.,
Fig. 1. Map of the 243 sampling locations across the typical Loess Plateau region in central China and the distributions of mean annual temperature (MAT) and mean annual precipitation (MAP) in the region.
C. Zhao et al. / Geoderma Regional 7 (2016) 251–258
Ri N Ri + 1, i = 1, 2, 3,…), the cumulative volume (V) of the particles is calculated as:
253
3. Results and discussions 3.1. Overall spatial variation of profile PSD
V ðrNRi Þ ¼ C v ½1−ðRi =λv Þ3−D
ð1Þ
where Ri is the characteristic scale, D is fractal dimension of PSD, and Cv and λv are constants which describe the shape and scale of the soil particles, respectively. When Ri = 0 in Eq. (1), then V is the total volume (VT) of all the particles and Cv = VT can be solved. Then when Ri = Rmax in Eq. (1) (maximum soil particle diameter, i.e., 2 mm in this study), λv = Rmax can be solved. Subsequently, Eq. (1) is re-expressed as: R 3−D V ðrNRi Þ ¼ 1− i R max VT
ð2Þ
To more conveniently calculate D, a logarithmic expression of Eq. (2) is proposed by Peng et al. (2014) and Tyler and Wheatcraft (1992) as: V ðrbRi Þ ¼ ð3−DÞ log Ri log R V
max
T
ð3Þ
2.5. Geostatistical analysis A combination of classical and geo-statistical analysis was used to determine the spatial variability of PSD and fractal dimension properties in the study area. Geo-statistical methods can be used to quantitatively determine the magnitude of variation and spatial structure. In this study, semi-variogram analysis (Matheron, 1963) was used to determine the spatial structures of both fractal dimension and PSD. Generally, spatial dependence was classified as strong for nugget-sill ratio b 25%, moderate for nugget-sill ratio range of 26–75% and weak for nugget-sill ratio N 75% (Cambardella et al., 1994). SPSS 17.0 was used to calculate descriptive statistical parameters, Microsoft EXCEL 2013 was used calculate CV and D. Then the experimental semi-variogram model construction along with PSD interpolation was done in GIS (Version: ESRI® ArcMap™ 9.2).
In the 0–500 cm soil profile, PSD varied slightly with increasing soil depth. Silt had the highest particle-size fraction (64.2–68.0%), followed by sand (19.7–24.0%) and then clay (11.8–12.7%) in the different soil layers (Fig. 2a). Thus silt was the most dominant soil particle in the 0–500 cm soil profile in the typical LP study area in China. The 0–5 m soil layer constituted mainly of Malan loess, with a small fraction of the upper layer consisting of Lishi loess. Using several soil profiles in the study area, Liu (1965) concluded that PSD of Malan loess layer was similar to that of upper Lishi loess layer. Zhao et al. (2011) concluded that the most erodible particlesize range in deposited farmlands was 0.25–0.5 mm, followed by 0.2–0.25 mm. In this study, the maximum sand and minimum clay and silt contents were in the surface (0–10 cm) soil layer. This was attributed mainly to the high degree of erosion of the top soil layer (Su et al., 2004). Based on the collected samples, the soil types in the study area included Haplic Luvisols, Terric Anthrosols, Calcic Chernozems, Aridic Leptosols, Calcaric Regosols, Calcic Kastanozems and Aridic Arenosols (FAO/ISRIC/ISSS, 1998), accounting for 14%, 9%, 21%, 43%, 7%, 2% and 4% of the 243 sampling sites, respectively. The tally of different soil textures in the collected soil samples at the various soil depths is listed in Table 1. There are 7 distinct soil textures in the 0–500 cm soil depth in the typical LP study area. Sand, loamy-sand and sandy-loam are coarse texture, while silt-loam, silt, loam and silty-clay-loam are medium texture (Table 1). The soil texture types were different at different soil depths. Only three types of soil texture were in the 0–20 cm soil layer. However, this number increased with increasing soil depth up to the maximum of seven types at the 100–150 cm soil layer. The soil texture type decreased beyond the 150 cm soil depth (Table 1). The soil texture type was higher in the shallow soil layers (0–100 cm) than in the deep soil layers (100–500 cm). Silt-loam (which accounted for 92.6% of the entire samples) was the most dominant soil texture in the various soil layers, followed by sandy-loam, loam, loamy-sand, sand, silt and then silty-clay-loam.
Fig. 2. Coefficient of variation (CV) and the mean values of Clay, Silt, Sand and D at different soil depths in the typical Loess Plateau region of China.
254
C. Zhao et al. / Geoderma Regional 7 (2016) 251–258
Table 1 Soil texture characteristics for different soil layers in the 0–500 cm profile in the typical Loess Plateau region of China. Soil depth (cm)
0–10 10–20 20–40 40–60 60–80 80–100 100–150 150–200 200–300 300–400 400–500 0–500
Medium texture
Coarse texture
Silty loam
Loam Silt Silty clay loam
Sand Sandy loam
Loamy sand
213 216 220 219 214 213 215 217 222 219 220 2476
11 9 4 6 7 7 4 6 2 4 2 62
0 0 0 1 1 3 1 0 1 0 0
0 0 2 1 2 0 2 2 0 1 1 11
0 0 0 0 0 1 2 2 1 1 0 7
0 0 0 0 0 0 1 0 0 0 1 2
11 10 9 8 11 11 10 8 9 10 11 108
The average CV for the different soil layers in the study area was 25.2–31.0% for clay, 16.4–76.6% for silt and 55.0–76.6% for sand (Fig. 2b). This suggested that the degree of variation in sand was higher of than that in clay and silt in the study area. CV was highest for the 100 cm depth (Fig. 2b), suggesting the highest variability of PSD was at that soil depth. This may be attributed to soil water movement both during precipitation recharge and through evapotranspiration discharge (Fu et al., 2013). 3.2. Geostatistical analysis of PSD At each of the 243 sites, a total of 11 soil layers (0–10, 10–20, 20–40, 40–60, 60–80, 80–100, 100–150, 150–200, 200–300, 300–400 and 400– 500 cm soil layers) were investigated for PSD pattern. The PSD in each layer was normally distributed and the data was directly used in geostatistical analysis (Regalado and Ritter, 2006). Before semi-variogram analysis, any spatial trend of the first- or second-order in the dataset were removed (Wang et al., 2007). The best-fit geostatistical model was determined through ranking of R2 (Coefficient of determination) and SD (standard deviation) derived from validation analysis. The results showed that the Gaussian model had the best PSD fit for the
various soil layers, relatively with the highest R2 (N0.64) and smallest SD (b 0.07) — please see Fig. 3. The Gaussian model parameters (Nugget ratio, Range, Nugget and Sill) and the spatial structure indicators are plotted in Fig. 4. The trends in the Gaussian model parameters were similar for silt and sand, both of which increased with increasing soil depth. The range of the Nugget values for PSD was 2.3–80.3, suggesting a shorter heterogeneity than the ~ 40 km sampling gap (Fig. 4c). The range of the Sill values, representing total spatial variation, was 32.8–48.7. The Gaussian model Nugget, which estimated of the contribution of random measurement to total variation in PSD (Wang et al., 2012a), was positive (Fig. 4d). The Nugget ratio for the different soil layers was 1.85–32.65% (Fig. 4a). The Nugget ratio trend generally increased with increasing soil depth, except for the 100–200 cm soil layer. The surface soil layer (0– 10 cm) had the minimum Nugget ratio for the three soil particle textures (b 2.00% for sand and silt and b 16.96% for clay), indicating a strong spatial dependence among the soil particles. This was because the surface soil layer was exposed to several processes after deposition, including erosion, agriculture and weathering (Cambardella et al., 1994; Montagne et al., 2008). Compared with the surface soil, the deeper soil layers were much closer to the state at the time of deposition (Wang et al., 2009). Thus spatial dependence of the surface soil was stronger than that of the other soil layers. Soils can also be affected by climatic factors, which are different across the typical LP region study area. This further increased the degree of spatial dependence of the surface soil layer (Lotspeich and Everhart, 1962). The range of the Gaussian model parameters for clay, silt and sand was 474.34–530.00 km (Fig. 4b), confirming that the 40 km interval was appropriate for spatial variability analysis of PSD in the study area. This was critical for optimizing research plan and improving accuracy level of regional soil water analysis. PSD values for any two locations were only spatially correlated if the separation distance was smaller than the range (Zeleke and Si, 2005). 3.3. Spatial patterns of PSD PSD maps were generated for each depth using Universal Kriging interpolation (Arc Map™ 9.2) and plotted in Fig. 5 and Figs. S1–2.
Fig. 3. Box and whisker plots depicting the range, 25th, 50th and 75th percentiles along with mean coefficient of determination (R2) and standard deviation (SD) of cross validation.
C. Zhao et al. / Geoderma Regional 7 (2016) 251–258
255
Fig. 4. The Gaussian model parameters for clay, silt, sand and the mean value (D) for different soil layers in the typical Loess Plateau region of China.
Generally, there existed a trend in PSD for all the soil depths for sandy soil in the north and more clayish soil in the south followed by a trend for decreasing elevation and increasing precipitation along the northsouth transect of the typical LP region study area. Loess originates from the deposition of wind-blown dust and debris from the northwest of the LP region. As the northwest LP region is closer to dust alimentation zone, the deposited debris are coarser as the weight shortens the deposition distance. However, for the southeast LP region (which is far from the alimentation zone), only finer debris can be transported that long due the light weight (Su et al., 1959; Liu, 1965; Wang et al., 2012b). After deposition, the loess transforms during soil formation as driven by various factors such as temperature, humidity, vegetation,
etc. The climate in the southeast of the LP region is warmer and wetter than that in the northwest, thus the loess deposits undergo intense weathering (Liu, 1985). Over time, soils in the southeast of the LP region become finer, which in turn increase clay content of the original deposits. Platy structures exist in some areas, disrupting the continuing trend in PSD. Platy structures in the study area are the combined effect of topography, geomorphology, sedimentation, etc. (Smalley et al., 2009). Generally, PSD in the study area had a significant banding structure, with deeper trends (denser isolines) in the north for silt and sand than for clay. On the other hand, the trend in clay gradually decline with increasing latitude. This was driven by the so-called Mu Us sand belt in
256
C. Zhao et al. / Geoderma Regional 7 (2016) 251–258
Fig. 5. Plots of the distributions of soil clay content for the various soil layers in the typical Loess Plateau region of China.
north of the study area. Sandy soils are common in the north and medium-texture soils common in the south (Wang et al., 2012a). The denser isolines demarcate the transition zone between the typical LP region and the Mu Us sand belt. Thus the degree of spatial variation in PSD was different for different regions. It was stronger in the north for silt and sand, and in the south for clay. The observed spatial distribution pattern could be useful in simulations of regional scale hydrological or vegetation analysis. 3.4. Fractal dimensions of PSD The range of D for PSD calculated by Eq. (3) was 2.13–2.69, well within the expected range of 0–3. The range of CV for D was 1.75– 2.30%, with a weak variability for the 0–500 cm soil profile of the study area (Fig. 2b). Based on fractal dimension analysis of PSDs, Su et al. (2004) observed a simple linear correlation between D and soil texture. Based on regression analysis of 649 PSDs in three representative dam farmlands in the north of the LP study area, Zhao et al. (2009) obtained a good linear fit between D and sand, silt and clay. At regional scale, different results were obtained for the 2673 PSDs. Clay and D had the best fit (R2 = 0.98), followed by sand (R2 = 0.79) and then silt (R2 = 0.63) — see Fig. 6a–c. The fits for soils with 0–60% sand, 40– 80% silt and 5–20% clay were better than for other sand, silt and clay ratios. This suggested that D was a possible indicator for changes in soil texture. Based on geo-statistical analysis, there was also more similarity between D and clay than between D and sand or silt. Gaussian model produced the best fit in terms of semi-variogram of D. Fig. 4(c–d) depicts
the Nugget ratio and range of D for the different investigated soil layers in the study area. The range was common (400–500 km) for the soil layers, with D curves similar for clay. This indicated that the spatial distribution pattern and variation trend of D were similar for clay. It further confirmed the feasibility of applying D in evaluating clay and the related soil properties. The study suggested that fractal analysis was applicable in evaluating the evolution of fine particles and also a good indicator for fine particle fractions as a driving factor of desertification in typical loess regions. There was a good linear regression fit between D and sand for the 0–60% range. This suggested that D reflected changes in sand particles and could therefore be used as an indicator for desertification. Equally, the spatial distribution pattern of fractal dimensions reflected the spatial heterogeneity of fine fractions and fine particles with a dominant expansion-contraction, absorbability or mechanical properties (Dixon, 1991). 4. Conclusions The regional distribution pattern and fractal features of PSD across a typical LP region in China were investigated. The study showed that: 1) Soil clay, silt and sand contents in various soil layers varied moderately in spaces across the study area. This suggested that change in soil texture with depth was not significant, a trend that could influenced by data availability or experimental conditions. 2) Soils were overall sandy in the north and clayey in the south, but soil texture variation uneven with increasing latitude. Change in silt and sand was greater in the north than in the south of the typical LP study
C. Zhao et al. / Geoderma Regional 7 (2016) 251–258
257
Farming on the Loess Plateau (No. A314021402-1503). The authors are indebted to the editors and reviewers for the constructive comments and suggestions during the review phase of the manuscript. We acknowledge the inputs of Dr. Liu ZP, Dr. Yang XL and Mr. Yang LQ during the research design and field data collection. Map. KMZ file containing the Google maps of the most important areas described in this article. Appendix A. Supplementary data Supplementary data associated with this article can be found in the online version, at http://dx.doi.org/10.1016/j.geodrs.2016.05.003. These data include the Google maps of the most important areas described in this article. References
Fig. 6. Plots of the relationships between fractal dimension of Particle Size Distribution (PSD) and clay, silt and sand content in the 0–500 cm soil profile in the typical Loess Plateau region of China (n = 2673).
area. This was useful for determining the parameters for large-scale soil water simulation in the study area. The trend in clay was relatively even and it decreased with increasing latitude for all the soil layers. 3) Fractal analysis showed that fractal demotions (D) was controlled by the clay content, suggesting that D might be a good indicator for soil erosion in the study area. Geostatistical analysis showed that the spatial variability of clay was similar to that of fractal dimension. This further demonstrated that a strong correlation existed between clay content and fractal dimension in the study area. Acknowledgement This study was supported by the National Natural Science Foundation of China (No. 41530854, 41390461 and 41501233) and the Opening Foundation of State Key Laboratory of Soil Erosion and Dry land
Antinoro, C., Bagarello, V., Ferro, V., Giordano, G., Iovino, M., 2014. A simplified approach to estimate water retention for Sicilian soils by the Arya-Paris model. Geoderma 213 (1), 226–234. Bimueller, C., Mueller, C.W., von Luetzow, M., Kreyling, O., Koelbl, A., Haug, S., Schloter, M., Koegel-Knabner, I., 2014. Decoupled carbon and nitrogen mineralization in soil particle size fractions of a forest topsoil. Soil Biology & Biochemistry 78 (8), 263–273. Cambardella, C.A., Moorman, T.B., Novak, J.M., Parkin, T.B., Karlen, D.L., Turco, R.F., Konopka, A.E., 1994. Field-scale variability of soil properties in Central Iowa soils. Soil Sci Soc Am J 58 (5), 1501–1511. Chen, L., Wei, W., Fu, B., Lu, Y., 2007. Soil and water conservation on the Loess Plateau in China: review and perspective. Progress in Physical Geography 31 (4), 389–403. Dixon, J.B., 1991. Roles of clays in soils. Applied Clay Science 5 (5–6), 489–503. Filgueira, R.R., Fournier, L.L., Cerisola, C.I., Gelati, P., Garcia, M.G., 2006. Particle-size distribution in soils: a critical study of the fractal model validation. Geoderma 134 (3–4), 327–334. Fu, X., Shao, M., Wei, X., Wang, H., Chen, Z., 2013. Effects of monovegetation restoration types on soil water distribution and balance on a hillslope in northern Loess Plateau of China. Journal of Hydrologic Engineering 18 (4), 413–421. He, X.B., Li, Z.B., Hao, M.D., Tang, K.L., Zheng, F.L., 2003. Down-scale analysis for water scarcity in response to soil-water conservation on Loess Plateau of China. Agriculture Ecosystems & Environment 94 (3), 355–361. Hollis, J.M., Hannam, J., Bellamy, P.H., 2012. Empirically-derived pedotransfer functions for predicting bulk density in European soils. European Journal of Soil Science 63 (1), 96–109. Hwang, S.I., Yun, E.Y., Ro, H.M., 2011. Estimation of soil water retention function based on asymmetry between particle- and pore-size distributions. European Journal of Soil Science 62 (2), 195–205. ISSS/ISRIC/FAO, 1998. World Reference Base for Soil Resources. World Soil Rec. Rep 84. FAO, Rome. Jia, X., Shao, M.A., Wei, X., Wang, Y., 2013. Hillslope scale temporal stability of soil water storage in diverse soil layers. J Hydrol 498 (18), 254–264. Katz, A.J., Thompson, A.H., 1985. Fractal sandstone pores: implications for conductivity and pore formation. Phys Rev Lett 54 (12), 1325–1328. Liu, D.S., 1965. The Loess Accumulation in China. Science Press, Beijing, China (in Chinese). Liu, D.S., 1985. Loess and Environment. Science Press, Beijing, China (in Chinese). Lotspeich, F.B., Everhart, M.E., 1962. Climate and vegetation as soil forming factors on the Llano Estacado. J Range Manage 15 (3), 134–141. Matheron, G., 1963. Principles of geostatistics. Econ Geol 58 (8), 1246–1266. Mengistu, D.K., 2009. The influence of soil water deficit imposed during various developmental phases on physiological processes of tef (Eragrostis tef). Agriculture Ecosystems & Environment 132 (3–4), 283–289. Millan, H., Gonzalez-Posada, M., Aguilar, M., Dominguez, J., Cespedes, L., 2003. On the fractal scaling of soil data. Particle-size distributions. Geoderma 117 (1–2), 117–128. Montagne, D., Cornu, S., Forestier, L.L., Hardy, M., Josière, O., Caner, L., Cousin, I., 2008. Impact of drainage on soil-forming mechanisms in a French Albeluvisol: input of mineralogical data in mass-balance modelling. Geoderma 145 (3–4), 426–438. Pan, F.F., Peters-Lidard, C.D., King, A.W., 2010. Inverse method for estimating the spatial variability of soil particle size distribution from observed soil moisture. Journal of Hydrologic Engineering 15 (11), 931–938. Peng, G., Xiang, N., Lv, S.Q., Zhang, G.C., 2014. Fractal characterization of soil particle-size distribution under different land-use patterns in the Yellow River Delta wetland in China. Journal of Soils and Sediments 14 (6), 1116–1122. Regalado, C.M., Ritter, A., 2006. Geostatistical tools for characterizing the spatial variability of soil water repellency parameters in a laurel forest watershed. Soil Sci Soc Am J 70 (4), 1071–1081. Shi, H., Shao, M.A., 2000. Soil and water loss from the Loess Plateau in China. J Arid Environ 45 (1), 9–20. Smalley, I., O'Hara-Dhand, K., Wint, J., Machalett, B., Jary, Z., Jefferson, I., 2009. Rivers and loess: the significance of long river transportation in the complex event-sequence approach to loess deposit formation. Quaternary International 198 (s1–2), 7–18. Stemmer, M., Gerzabek, M.H., Kandeler, E., 1999. Invertase and xylanase activity of bulk soil and particle-size fractions during maize straw decomposition. Soil Biology & Biochemistry 31 (1), 9–18.
258
C. Zhao et al. / Geoderma Regional 7 (2016) 251–258
Su, L.Y., Bao, Y.Y., Tian, G.G., Qiu, Y.J., Xie, H.X., 1959. Analysis of the loess particles in the middle reaches of the Yellow River. Chinese Journal of Geology 8, 251–255 (in Chinese). Su, Y.Z., Zhao, H.L., Zhao, W.Z., Zhang, T.H., 2004. Fractal features of soil particle size distribution and the implication for indicating desertification. Geoderma 122 (1), 43–49. Sun, H., Yang, J., 2013. Modified numerical approach to estimate field capacity. Journal of Hydrologic Engineering 18 (4), 431–438. Tyler, S.W., Wheatcraft, S.W., 1989. Application of fractal mathematics to soil water retention estimation. Soil Sci Soc Am J 53 (4), 987–996. Tyler, S.W., Wheatcraft, S.W., 1992. Fractal scaling of soil particle-size distributions: analysis and limitations. Soil Sci Soc Am J 56 (2), 362–369. Wang, D., Fu, B., Zhao, W., Hu, H., Wang, Y., 2008. Multifractal characteristics of soil particle size distribution under different land-use types on the Loess Plateau, China. Catena 72 (1), 29–36. Wang, L., Mou, P.P., Huang, J., Wang, J., 2007. Spatial heterogeneity of soil nitrogen in a subtropical forest in China. Plant & Soil 295 (1–2), 137–150. Wang, Y., Shao, M.A., Liu, Z., Warrington, D.N., 2012a. Investigation of factors controlling the regional-scale distribution of dried soil layers under forestland on the Loess Plateau, China. Surv Geophys 33 (2), 311–330. Wang, Y., Shao, M.A., Liu, Z., Warrington, D.N., 2012b. Regional spatial pattern of deep soil water content and its influencing factors. Hydrol Sci J 57 (2), 265–281.
Wang, Y., Shao, M.A., Liu, Z., Zhang, C., 2014. Prediction of bulk density of soils in the Loess Plateau region of China. Surv Geophys 35 (2), 395–413. Wang, Y., Zhang, X., Huang, C., 2009. Spatial variability of soil total nitrogen and soil total phosphorus under different land uses in a small watershed on the Loess Plateau, China. Geoderma 150 (1–2), 141–149. Wu, T., Lu, G., 2012. Climatic sub-cycles recorded by the fourth paleosol layer at Luochuan on the lLoess Plateau. Environmental Earth Sciences 66 (5), 1329–1335. Zeleke, T.B., Si, B.C., 2005. Scaling relationships between saturated hydraulic conductivity and soil physical properties. Soil Sci Soc Am J 69 (6), 1691–1702. Zhao, Y., Xu, M., 2013. Runoff and soil loss from revegetated grasslands in the hilly Loess Plateau region, China: influence of biocrust patches and plant canopies. Journal of Hydrologic Engineering 18 (4), 387–393. Zhao, P., Shao, M., A, Omran, W., Amer, M., Abdel, M., 2011. Effects of erosion and deposition on particle size distribution of deposited farmland soils on the Chinese Loess Plateau. Rev Bras Ciênc Solo 35(6), 2135–2144. Zhao, P., Shao, M.A., Zhuang, J., 2009. Fractal features of particle size redistributions of deposited soils on the dam farmlands. Soil Sci 174 (7), 403–407. Zhou, B., Lv, J., Wang, Q., 2015. Chloride transport in an aggregated clay soil column. Journal of Hydrologic Engineering 20 (11) http://dx.doi.org/10.1061/(ASCE)HE.19435584.0001227.
