Soil hydro-physical characteristics and water retention function of typical shrubbery stands in the Yellow River Delta of China

Catena 156 (2017) 315–324

Contents lists available at ScienceDirect

Catena journal homepage: www.elsevier.com/locate/catena

Soil hydro-physical characteristics and water retention function of typical shrubbery stands in the Yellow River Delta of China

MARK

Jiangbao Xia⁎, Ziguo Zhao, Ying Fang Binzhou University, Shandong Provincial Key Laboratory of Eco-Environmental Science for the Yellow River Delta, Binzhou 256603, PR China

A R T I C L E I N F O

A B S T R A C T

Keywords: Soil water storage capacity Soil water retention Fractal dimension Soil infiltration Vegetation Shell sand

The shell ridge worldwide that presents a combination of old and new shell ridges is located in the Yellow River Delta of China. The water storage and retention capacities of the shell sand soil are key factors for the growth and development of vegetation and act as important indicators in evaluations of vegetation-related soil and water conservation. We investigated the hydrological and physical characteristics of the soil and the interactions of these characteristics for the following dominant shrubbery stands on the shell ridge: Tamarix chinensis Lour, Periploca sepium Bunge and Ziziphus jujuba var. Spinosa. The different shrubbery stands presented significantly different physical shell sand parameters, such as water storage capacity, permeability and water-holding capacity. Compared with the bare soil, the shrubbery-covered shell sand soils showed significantly reduced bulk densities; increased porosities, water storage capacities, infiltration rates and water-holding capacities; and reduced particle sizes. The shrubbery increased the amount of soil water-stable aggregates and reduced the amount of soil air-dried aggregates, and the fractal dimension of the soil water-stable aggregates was significantly higher than that of the air-dried aggregates. T. chinensis shrub growth yielded a better soil aggregate structure compared with that of Z. jujuba, whereas the P. sepium shrub yielded the poorest structure. The soil porosity and the mean weight diameter of the soil aggregates were the major factors that affected the water retention capability of the shell sand. The best water retention capability was observed with the T. chinensis shrub stand followed by the Z. jujuba and P. sepium stands, whereas the bare soil exhibited the poorest results. Our results indicate that T. chinensis seedlings are the best choice for planting to improve soil porosity and soil aggregates on the shell ridges of the Yellow River Delta.

1. Introduction Shell ridges are mainly formed by the shells of shelled organisms that live in the intertidal zone and shell debris that is transported by waves and accumulates near the high-tide line. Shell ridges are distinctive shell sand deposits that lie on the upper surface of tidal flats, where shellfish grow in abundance (Xie et al., 2012). Two shell ridges that are roughly parallel with the coast are located in the Yellow River Delta of China. These are the largest and only ridges with cooccurring old and new shell ridges in the world; thus, they have great scientific value and significance for research on China's marine geology, biodiversity conservation, and ecosystem types. Vegetation on the shell ridges is affected by natural factors, such as droughts and human interference, and often exhibits varying degrees of degradation, which exacerbates soil erosion. Revegetation on shell ridges provides windbreaking and sand-fixing functions and plays a role in soil and water conservation, and all of these features are important for improving the

regional ecological environment and maintaining the local ecosystem stability (Tian et al., 2011; Xia et al., 2013). However, the choice of shrub type based on the objectives of improving the soil physical and hydrological properties and yielding benefits, such as soil and water conservation, remain to be resolved. Several studies conducted on shell ridges (Meldahl, 1995; Saito et al., 2000; Xie et al., 2012) have primarily focused on the coastal geomorphology, sea level and climate changes, coastal ecosystem evaluations, and vegetative distributions (Saito et al., 2000; Xie et al., 2012). The ability of soil to store and retain moisture has become a topic of great interest in ecological and hydrological studies (Shi et al., 2016; Antinoro et al., 2017). Most current investigations of the water storage capacity of soil under different vegetation covers have been conducted via static comparative analyses of the level of soil moisture (Vogelmann et al., 2013; Zhang et al., 2016). However, studies of the water storage functions of soil are mainly performed by investigating single factors, such as the soil pore characteristics, permeability and aggregates (Xia

Abbreviations: BD, soil bulk density; BDLLWR, BD from the least limiting water range; D, soil particle fractal dimension; DC, degree of compactness; MWD, mean weight diameter ⁎ Corresponding author. E-mail address: [email protected] (J. Xia). http://dx.doi.org/10.1016/j.catena.2017.04.022 Received 25 October 2016; Received in revised form 5 April 2017; Accepted 21 April 2017 Available online 05 May 2017 0341-8162/ © 2017 Elsevier B.V. All rights reserved.

Catena 156 (2017) 315–324

J. Xia et al.

et al., 2012; Vogelmann et al., 2013; Xia et al., 2013; Zhang et al., 2016), and these studies usually focus on hilly areas (Zhang et al., 2006; Liu et al., 2009; Jiang et al., 2013) and rarely consider the soil improvements and soil and water conservation of the coastal shelterbelt (Liu et al., 2015). Because of the lack of data on the physical and hydrological functions of soil under different vegetation covers, the water storage and water retention properties of soil have not been properly characterized and the factors that influence shell sand soil with different vegetation types in muddy coastal zones are poorly understood. To an extent, these limitations have hindered the optimization of configurable vegetation patterns in such regions and prevented the efficient use of the available soil moisture. Serving as a transition zone between water and land interactions, muddy coastal shell ridges exhibit transitional characteristics in the soil environment formation, water storage, water transport and other ecological processes that are dominated by water. In addition, the special shell sand imparts unusual hydrological and physical properties to the shell ridge that result in different water storage and regulatory functions from those of common terrestrial ecosystems (Tian et al., 2011; Xia et al., 2013). Therefore, to clarify the water-holding capacity of the shell soil and the factors that influence shell ridges with different shrub types, three typical shrubs that grow in the shell ridge of the Yellow River Delta, Tamarix chinensis Lour, Periploca sepium Bunge and Ziziphus jujuba var. Spinosa, were used to investigate the bulk density, porosity, particle composition, aggregate size, particle fractal dimension, storage capacity, infiltration characteristics, soil moisture characteristic curves and other parameters of the shell sand soils under different shrubbery stands. We hypothesized that the water storage and retention capacity of the shell sand would be closely related to the vegetation types and great differences would be observed in the hydrology-physical characteristics of the shell sand under the different shrubbery stands. One goal of this study was to understand the role of the different shrub stands in improving the physical and hydrological structure of the soil and clarify ecological characteristics of soil moisture and the factors influencing the soil moisture parameters on the shell ridges under different shrubbery stands. An additional goal was to comprehensively evaluate the soil water-holding capacity under different shrubbery stands using a principal component analysis (PCA) and the fuzzy membership function method. The results will provide a theoretical basis and technical reference for selecting the optimal shrub type for vegetation recovery of the shell ridges of the Yellow River Delta.

Fig. 1. The geographic location map of the experimental site on the shell ridges in the Yellow River Delta. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

annua L., and Zoysis macrostachya Merr. 2.2. Sample site In July 2015, three monospecific shrubs, T. chinensis, P. sepium and Z. jujuba, growing on similar habitats in shell ridges were selected (Fig. 2). The sampling sites had the following general characteristics: the average age of the dominant shrubs was 8 years; the average heights of the shrubs T. chinensis, P. sepium and Z. jujuba were 1.95 m, 1.58 m and 1.68 m, respectively; and the average ground diameters were 2.36 cm, 1.33 cm and 1.78 cm, respectively. T. chinensis, P. sepium and Z. jujuba had canopy densities of 0.82, 0.75 and 0.78, respectively, and vegetation coverage of 75%, 64% and 70%, respectively. Six experimental observation standard plots for each shrub measuring 10.0 m × 10.0 m were selected, and bare shell sand plots in the same area were used as controls. In each observation plot, five test sites were chosen according to the S-type sampling method, and undisturbed soil sampling and soil parameter measurements were mainly conducted from the 0–20 cm soil layer. Soil aggregates were collected from undisturbed soil samples and then air dried indoors, and the samples were treated with extra caution to minimize the degree of disturbance during their acquisition and transport to avoid damage to the soil aggregates. All the analyses in the present work have been done on undisturbed samples.

2. Materials and methods 2.1. Study area The study sites were located in Wangzidao (N38°14′30″, E117°54′38″), which is in the central and eastern coastal lowlands of Wudi County, Binzhou City, Shandong Province, China. The experimental site is located in the buffer zone of the Binzhou National Shell Ridge and Wetland Nature Reserve, the geogaphic location map of the research area is shown in the Fig.1, and it has a coverage area of 80,480 hm2. This area is in the East Asian monsoon semi-humid continental climate zone in the warm temperate region, and it has an average annual rainfall of 550 mm, an average annual potential evaporation of 2431 mm, an average annual temperature of 12.36 °C, an average annual sunshine duration of 2849 h, and an average annual frost-free period of 205 d. The soil types are mainly shell sand soil and coastal saline soil; the areas on the seaward and landward sides are dominated by coastal salinized soil; and the beach ridge area is dominated by shell sand soil with an average thickness of 1.0–2.5 m (up to 3.0–4.0 m at certain locations). The soil parent material is composed of aeolian sediment and calcareous shell soil. The main shrub species in this area are T. chinensis, P. sepium, and Z. jujuba var. spinosa, and the primary herbaceous plants include Limonium bicolor, Setaria viridis, Artemisia

2.3. Analysis of soil capacity properties 2.3.1. Determinations of the basic soil physical properties and size of aggregates Soil moisture was determined using the drying method, and indicators such as the bulk density (BD) and porosity were determined using the soil cutting ring soaking method. The soil capillary water storage, non-capillary water storage and saturated water storage were all determined at a soil depth of 0.2 m (Institute of Soil Science, 1978). The soil water storage capacity formulas used in this study are as follows: 316

Catena 156 (2017) 315–324

J. Xia et al.

T. chinensis Lour (TC)

P. sepium Bunge (PS)

Z. jujuba var. spinosa (ZJ)

Bare land as control check (CK)

Fig. 2. The photo of different shrubbery stands on the shell ridges in the Yellow River Delta. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

Wt = 10000 × Pt × h

(1)

Wc = 10000 × Pc × h

(2)

D=3−

−2

In these equations, Wt is soil saturated water storage (t hm ); Wc is soil capillary water storage (t hm− 2); Wnc is soil non-capillary water storage (t hm− 2); PC is soil capillary pore (%); Pt is soil total porosity (%); h is soil depth (m). The degree-of-compactness (DC) was calculated as follows:

DC =

BD BD LLWR

2.3.3. Determination of the soil infiltration capability The soil infiltration rates at different times were determined using the double-loop method (INW, USA), and the infiltration process curve was generated. Double-loop method for the determination of the main steps can be described as follows, 1) Putting the two loops at the test site, and pressing them evenly into the soil; 2) Using the Markova bottle to supply water to two loops at the same time, maintaining a constant water level; 3) Recording test time with a seconds counter, and recording the corresponding water level change in Markov bottle at the same time. The double-loop method was used for in situ measurement, and samples were undisturbed soil. Two models were used to simulate the soil infiltration process.

(4)

where DC is the degree-of-compactness (%); BD is the bulk density (mg m− 3); and BDLLWR is the BD from the least limiting water range (BDLLWR = − 0.00078% clay + 1.83803) according to Reichert et al. (2009). The contents of air-dried and water-stable aggregates of the different soil size fractions were determined using the dry-sieving and wet-sieving methods, respectively (Institute of Soil Science, 1978). The mean weight diameter (MWD) (mm) soil aggregate index was calculated using the following formula (Wang et al., 2015):

MWD =

∑ Y iX i

(6)

where D is the soil particle fractal dimension, di is the average diameter of two adjacent soil particles with sizes di and di + 1 (mm), dmax is the mean diameter of the soil particles with the maximum grain size (mm), Wi is the cumulative mass of the soil particles with diameters less than di (g), and W0 is the total mass of the soil sample (g).

(3)

Wnc = W–W t c

lg(Wi W0 ) lg(di d max )

Horton model: f = f c + (f 0 − fc ) e−kt

(7)

where f, f0, fc and t are the infiltration rate, initial infiltration rate, steady infiltration rate (mm min− 1) and infiltration time (min), respectively, and k is an empirical parameter.

(5)

where Yi is the proportion of the total sample weight that is retained on each sieve and Xi is the mean diameter of the ith size class (mm).

General empirical model: f = at−n + b

(8)

where f and t are the infiltration rate (mm min− 1) and infiltration time (min), respectively, and a, b, and n are empirical parameters.

2.3.2. Determination of the soil particle fractal dimension Based on the formula of the soil particle fractal dimension by Turcotte (1986) and Sperry and Hacke (2002) as well as the new model formulae derived by Tyler and Wheatcraft (1989) and Yang et al. (1993), the soil particle fractal dimension formula used in this study is as follows:

2.3.4. Determination of the soil moisture characteristic curve The soil moisture characteristic curve was determined using a highspeed refrigerated centrifuge (CR21N, HITACHI, JPN). In order to make soil moisture characteristic curve, the soil water content was measured 317

Catena 156 (2017) 315–324

J. Xia et al.

soil aeration. The greatest improvements in soil aeration were observed for T. chinensis followed by P. sepium, whereas Z. jujuba performed poorly.

under different rotation speed and time. The soil moisture characteristic curve shows the relationship between soil water potential and soil water content. The empirical equation for the soil moisture characteristic curve proposed by Gardner et al. (1970) was used for fitting: (9)

θ = AS −B

3.2. Soil aggregate composition

where θ is the soil water content (%), S is the soil suction (kPa), and A and B are parameters that denote the height of the curve (i.e., the soil water-holding capacity, where higher values of A correspond to a stronger soil water-holding capacity) and the direction of the curve (i.e., the rate of decrease of soil moisture with decreasing soil water suction).

Macroaggregates larger than 0.25 mm in the dry-sieved soil accounted for 77.1%–92.1% of the total aggregates, and compared with bare soil, the amount of macroaggregates in the T. chinensis, P. sepium and Z. jujuba shrub stand soil was reduced by 16.3%, 11.6% and 6.6%, respectively (Table 2). The 1–0.25 mm coarse sand-sized content accounted for 41.1%–55.9% of the total aggregates, which indicates that the 1–0.25 mm coarse sand-sized content was the dominant fraction of the > 0.25 mm macroaggregates. Wet-sieved soil aggregates reflect the soil stable-water capacity of the aggregates, whereas the drysieved soil aggregates do not reflect this characteristic. In the wetsieved soil, macroaggregates larger than 0.25 mm accounted for 65.7%–78.3% of the total aggregates, and compared with the bare soil, the amount of macroaggregates in the P. sepium, T. chinensis and Z. jujube shrub stand soil was increased by 6.6%, 11.6% and 16.3%, respectively (65.7%). After the soil particles were soaked with water, the gravel content decreased significantly, and the content of microaggregates, such as fine sand-sized particles and silt-clay-sized aggregates, increased significantly. Overall, the content of coarse sand-sized particles was highest, followed by the fine sand-sized particles and gravel, whereas the silt-clay-sized content was the lowest. All three shrub types were effective in reducing the contents of gravel and fine sand-sized particles and increasing the contents of coarse sand-sized and silt-clay-sized particles. The best soil aggregate structure of waterstable macroaggregates (> 0.25 mm) was found in the Z. jujube stand followed by the T. chinensis and P. sepium stands, whereas bare soil had the poorest. Table 2 shows that after dry sieving, significant differences were observed in the soil mean weight diameter (MWD) values under the different shrub stands (P < 0.05). The bare soil had the highest MWD, and the P. sepium, Z. jujube and T. chinensis shrub stands had 25.9%, 16.1% and 9.9% lower MWDs compared with the bare soil, respectively. After wet sieving, the MWD of the bare soil decreased significantly, whereas that of the soils under the T. chinensis, Z. jujube and P. sepium stands increased by 39.6%, 12.5% and 6.3% compared with that of bare soil, respectively. The soil MWD values under the different shrubbery stands obtained using the dry-sieving method were significantly higher than those obtained using the wet-sieving method because a large amount of non-water-stable aggregates dissolved in water. The MWD from the bare soil was significantly lower with the wet-sieving method than with the dry-sieving method, which indicates that the non-water-stable aggregate content of the bare soil was relatively high. Furthermore, the shrubbery growth increased the quantity of soil water-stable aggregates and decreased the content of the air-dried aggregates, and our results indicate that the T. chinensis shrub stands had the best effect on improving the soil water stability.

2.4. Data processing Comprehensive evaluations of the soil moisture storage and retention capacities of the different shrubberies were performed using the fuzzy membership function method with the following formula: (10)

X (u) = (Xi − Xmin ) (Xmax − Xmin )

where X(u) is the membership function value, Xi is the average of each shrub on a given indicator, and Xmin and Xmax are the minimum and maximum values of the given indicator in different shrubs, respectively. The soil water infiltration process and soil moisture characteristic curve simulations as well as a correlation analysis and PCA of the indicators were performed using Microsoft Excel 2013 and SPSS 16.0. The soil hydro-physical characteristics are indicated by their mean values. 3. Results and analysis 3.1. Soil bulk density and porosity The bulk density (BD) and the BD from the least limiting water range (BDLLWR) of the shrubberies were lower than that of bare soil. Compared with bare soil, the BD of the T. chinensis, P. sepium and Z. jujuba shrub stands were 6.9%, 5.6% and 2.8% lower, respectively, and the BDLLWR of the T. chinensis, P. sepium and Z. jujuba shrub stands were 10.8%, 3.8% and 2.3% lower, respectively (Table 1). The degree-ofcompactness (DC) of the shrubberies was higher than that of bare soil, and the estimated DC values are presented in Table 1. Soil total porosity is the volume percentage of the total pores in the soil. The soil porosities had the following pattern: T. chinensis > P. sepium > Z. jujuba > bare soil; the total soil porosities of the T. chinensis, P. sepium and Z. jujube shrub stands were 31.5%, 16.9% and 9.6% higher than that of bare soil, respectively. Furthermore, the shrubbery stands showed higher levels of vegetation coverage, higher amounts of litter and thicker layers of humus, and the shrubs had well-developed root systems and more residual root systems, which increased the soil porosity and thus effectively improved soil aeration. The soil porosity of the shrubbery stands was relatively high, and the soil had a loose texture and good permeability. Void ratio is the ratio of pore volume and solid soil volume in soil. The soil void ratios of the T. chinensis, P. sepium and Z. jujuba shrub stands were 1.76, 1.35 and 1.21 times that of bare soil, respectively. The shrubbery stands effectively improved the BD and porosity of the shell sand habitat, which greatly improved the

3.3. Soil particle fractal dimension The fractal dimensions (D) of the water-stable soil aggregates under

Table 1 Soil bulk density (BD), BD from the least limiting water range (BDLLWR), degree-of-compactness (DC) and soil porosity under different shrubbery stands⁎. Shrub types

BD (g cm− 3)

BDLLWR (mg m− 3)

DC (%)

Total porosity (%)

Capillary porosity (%)

Non-capillary porosity (%)

Void ratio

TC PS ZJ CK

1.34 1.36 1.40 1.44

1.62 1.75 1.78 1.82

82.5 77.7 79.8 79.3

57.6 51.2 48.0 43.8

52.6 46.2 45.2 39.8

5.1 5.0 2.8 4.0

1.37 1.05 0.94 0.78

⁎

TC: T. chinensis Lour; PS: P. sepium Bunge; ZJ: Z. jujuba var. spinosa; CK: Bare soil as control check.

318

Catena 156 (2017) 315–324

J. Xia et al.

Table 2 Soil aggregate composition and content under different shrubbery stands⁎. Sieve method

Dry sieve

Wet sieve

⁎

Shrub types

TC PS ZJ CK TC PS ZJ CK

Soil aggregate size distribution/%

Soil aggregate mean weight diameter (mm)

> 2 mm

Gravel 2–1 mm

Coarse sand 1–0.25 mm

Fine sand 0.25–0.05 mm

Silt-clay < 0.05 mm

17.10 6.73 14.32 17.19 16.30 5.54 6.20 5.92

18.92 23.38 17.68 24.15 16.14 17.20 14.77 17.59

41.11 55.90 49.45 50.78 42.45 50.11 57.36 42.21

19.08 12.46 17.50 7.51 19.04 24.08 17.30 29.10

3.80 1.54 1.06 0.38 6.08 3.08 4.38 5.19

0.73 0.60 0.68 0.81 0.67 0.51 0.54 0.48

TC: T. chinensis Lour; PS: P. sepium Bunge; ZJ: Z. jujuba var. spinosa; CK: Bare soil as control check.

Total water-storage capacity Capillary water-storage capacity Non-capillary water-storage capacity 1000 a -2

Water-storage capacity (t hm

b

e

c

800

d

f

f

g 600

400

200 g

i

g

h

0 TC

PS ZJ Shrub types

CK

Fig. 4. Soil water-storage capacity under different shrubbery stands (TC: T. chinensis Lour; PS: P. sepium Bunge; ZJ: Z. jujuba var. spinosa; CK: Bare soil as control check.) The little English letters indicate the difference in 5% by LSD test.

Fig. 3. Soil particle fractal dimension under different shrubbery stands (TC: T. chinensis Lour; PS: P. sepium Bunge; ZJ: Z. jujuba var. spinosa; CK: Bare soil as control check.) The little English by different letters indicate the difference in 5% by Least Significant Difference (LSD) test.

3.4. Soil water storage capacity Fig. 4 shows that three shrubbery stands had significantly different saturated soil water storage and capillary water storage values (P < 0.05). The mean values of the saturated soil water storage and soil capillary water storage were in the order T. chinensis > Z. jujube > P. sepium > bare soil. Compared with bare soil, the T. chinensis, Z. jujube and P. sepium shrub stands presented saturated soil water storage values that were 31.7%, 16.9% and 9.7% higher, respectively (656.46 t hm− 2), and soil capillary water storage capacities that were 32.1%, 16.1% and 13.6% higher, respectively (596.69 t hm− 2). The non-capillary water storage also differed significantly by vegetation type (P < 0.05), and the mean values were in the order T. chinensis > Z. jujube > bare soil > P. sepium.

the different shrubbery stands were significantly higher than those of the air-dried aggregates. Compared with the D values of the air-dried aggregates, the D values of the water-stable soil aggregates under the T. chinensis, P. sepium, Z. jujube shrub stands and bare soil were 1.06, 1.11, 1.18 and 1.36 times higher, respectively (Fig. 3). The D values of the soil particles under the three shrubbery stands were significantly different (P < 0.05), and the mean D values of the air-dried aggregates were in the order T. chinensis > P. sepium > Z. jujube > bare soil. Compared with bare soil, the mean D values of the shrubs were 35.2%, 15.6% and 13.3% higher, respectively. The bare soil received little interference and thus had a smaller decentralized structure, which leads to significantly lower D values of the air-dried aggregates in the bare soil. The mean D values of the water-stable aggregates were in the order T. chinensis > bare soil > Z. jujube > P. sepium. The significantly higher D value of the water-stable aggregates of the bare soil was associated with the increased clay content and finer texture after the shell sand soil was immersed in water.

3.5. Soil infiltration characteristics Both the Horton model and the general empirical model well fit the soil infiltration process of three shrubberies (Table 3 and Fig. 5). The infiltration curves had similar trends and consisted of three stages: an abrupt change in the early infiltration stage, a gradual changing stage and a stable stage. The fc values in the Horton model were similar to the measured value, and the P. sepium shrub had the lowest k value. This 319

Catena 156 (2017) 315–324

J. Xia et al.

Table 3 Models fitting of soil infiltration of under different shrubbery stands⁎. Shrub types

TC PS ZJ CK

Measured parameters(mm min− 1)

Parameters of Horton model

Parameters of current model

f0

fc

f0

fc

k

R2

a

b

n

R2

14.82 12.30 11.23 8.07

6.23 3.92 2.62 1.25

19.33 17.03 15.53 14.22

6.36 3.93 2.75 1.23

0.122 0.106 0.212 0.121

0.993 0.999 0.997 0.996

22.91 26.21 16.45 36.81

4.36 1.57 1.68 0.86

0.617 0.603 0.732 0.979

0.989 0.994 0.997 0.983

⁎ TC: T. chinensis Lour; PS: P. sepium Bunge; ZJ: Z. jujuba var. spinosa; CK: Bare soil as control check; f0: initial infiltration rate; fc: steady infiltration; k, a, b, and n are empirical parameters.

16

16 PS

ZJ

CK

TC

14

14

12

12 Soil water content (%)

-1

Infiltration rate (mm min )

TC

10 8 6 4 2

PS

ZJ

CK

10 8 6 4 2

0

0 0

10

20

30 40 Time (min)

50

60

70

0

Fig. 5. The characteristic curve of soil infiltration under different shrubbery stands (TC: T. chinensis Lour; PS: P. sepium Bunge; ZJ: Z. jujuba var. spinosa; CK: Bare soil as control check.).

2

4 6 8 2 Water suetion (10 kPa)

10

12

Fig. 6. Soil water characteristic curves under different shrubbery stands (TC: T. chinensis Lour; PS: P. sepium Bunge; ZJ: Z. jujuba var. spinosa; CK: Bare soil as control check.).

cantly with increasing soil water suction (Fig. 6). In the low soil suction range (1–10 kPa), the soil released more moisture and the soil moisture characteristic curves were all relatively steep because the capillary pores were large; thus, when a small amount of suction was applied to the soil, water in the large pores was released and could be directly utilized by plants; thus, it became available water for plant growth. The ability of the soils to release the available water in the low suction range was in the order T. chinensis > P. sepium > Z. jujube > bare soil. In the moderate suction range (300–1000 kPa), the highly compacted soil had lower porosity, and the number of large pores decreased significantly while the number of mid-size pores increased. The curve was relatively flat with increasing water suction. With an increase in soil water suction, the capacity for a gradual release of water into the soil decreased, and the soil moisture began to stabilize, which gradually decreased the difference in water contents between the soils under the three shrubberies. This trend was particularly evident for the differences observed between the soils under the Z. jujube shrub stand and the bare soil. Under the same matric potential, the soil water content under the T. chinensis shrub stand was significantly higher than that under the P. sepium stand, whereas the soil water contents under the Z. jujube shrub stand and bare soil were relatively low and the differences were insignificant (P > 0.05). The empirical equation proposed by Gardner et al. (1970) was sufficient for simulating the soil moisture curves under the three shrubbery stands on the shell ridges with a high correlation coefficient R2. The parameters of the curves are shown in Table 4. Parameter A, which is the soil water-holding capacity, was found in the order T.

finding indicated that the P. sepium shrub stand had the slowest rate of change from the initially decreasing infiltration rate to a more steady rate of infiltration and required the longest time to reach steady infiltration. The infiltration rates of the T. chinensis and bare soil were not as long as that of P. sepium, and the Z. jujube shrub stand required the shortest time to reach steady infiltration. The values of b in the general model were significantly lower than the corresponding measured steady infiltration rate. The R2 values for the measured initial infiltration rate and the steady infiltration rate showed that the Horton model had a better fit and its results were closer to the measurements than the general model, which indicates that the Horton model more accurately described the infiltration characteristics of the shell ridge soils under different shrubbery stands. The measured initial infiltration rates of the T. chinensis, P. sepium and Z. jujube shrub stands were 1.36, 1.20 and 1.09 times higher than that of bare soil, respectively, and the measured steady infiltration rates of the T. chinensis, P. sepium and Z. jujube shrubbery stands were 5.17, 3.20 and 2.24 times higher than that of bare soil, respectively. The soil permeability under the different shrubbery stands varied remarkably, with the best soil infiltration capacity observed in the T. chinensis shrub stand, followed by the P. sepium and Z. jujube stands, and the poorest soil infiltration observed in the bare soil.

3.6. Soil water retention characteristics The soil moisture of the three shrubbery stands decreased signifi320

Catena 156 (2017) 315–324

J. Xia et al.

Table 4 Parameters of soil water characteristic curve under different shrubbery stands⁎. Shrub types

Parameters A

Parameters B

R2

TC PS ZJ CK

6.6986 5.7517 4.8692 4.739

0.1791 0.1699 0.1911 0.2062

0.8798 0.8923 0.8779 0.883

Table 6 Subordinate function value of soil water storage and holding principal factor of different shrubbery stands⁎. Principal factors

Total porosity Capillary porosity Void ratio Soil mean weight diameter of dry-aggregate Soil mean weight diameter of water- stable aggregate Fractal dimension of dry-aggregate Total water-storage capacity Capillary water-storage capacity First infiltration rate Stable infiltration rate Soil water holding capacity (Parameters A) Total

⁎

TC: T. chinensis Lour; PS: P. sepium Bunge; ZJ: Z. jujuba var. spinosa; CK: Bare soil as control check.

chinensis > P. sepium > Z. jujube > bare soil, and compared with the bare soil, the soil water-holding capacity of the soils under the shrubs was 1.41, 1.21 and 1.03 times higher, respectively. The declining rates of soil moisture with decreasing soil suction (parameter B) were in the order P. sepium < T. chinensis < Z. jujube < bare soil. Shrub growth significantly improved the water-holding capacity and water availability of the shell sand soil. The best water-holding capacity and the most available water supply were observed in the T. chinensis shrub stand followed by the P. sepium and Z. jujube shrub stands. In contrast, the bare soil had the lowest water-holding capacity because of the lack of vegetation cover, low humus content, low porosity and high soil compactness.

Bulk density Total porosity Capillary porosity Non-capillary porosity Void ratio Soil mean weight diameter of dry-aggregate Soil mean weight diameter of water- stable aggregate Fractal dimension of dry-aggregate Fractal dimension of water-stable aggregate Total water-storage capacity Capillary water-storage capacity Non-capillary water-storage capacity First infiltration rate Stable infiltration rate Soil water holding capacity (Parameters A) Contribution rate (%) Cumulative contribution rate (%)

0.524 0.088 − 0.043 0.785 0.089 0.848 0.160 − 0.147 0.701 0.077 − 0.075 0.800 − 0.309 − 0.271 − 0.317 20.580 94.820

CK

1.000 1.000 1.000 0.619 1.000

0.307 0.424 0.263 0.000 0.158

0.534 0.500 0.466 0.381 0.316

0.000 0.000 0.000 1.000 0.000

1.000 1.000 1.000 1.000 1.000 1.000 10.619

0.444 0.307 0.424 0.550 0.527 0.517 3.920

0.379 0.534 0.500 0.257 0.298 0.066 4.232

0.000 0.000 0.000 0.000 0.000 0.000 1.000

4.1. Soil water physical characteristics of shrubbery stands and its interaction Mutual feedback occurs between plants and soil, and the physical and chemical properties of soil vary markedly under different vegetation types (An et al., 2010; Martinez-Murillo et al., 2013; Zhang et al., 2016; van Hall et al., 2017). Our findings indicated that the lowest DC (77.7%, P. sepium) was nearly equal to the lower limit (77%) defined by Suzuki et al. (2007) and the highest DC (82.5%, T. chinensis) was lower than the 88% upper limit or 90% upper limit, which was suggested by Reinert et al. (2008) and indicates moderate compaction under the different shrubbery stands on the shell ridges. The soil DC was consistent with the soil BD (Mentges et al., 2016). The BDLLWR decreased as the content of clay and clay plus silt increased, and the DC affects important ecological properties, such as water and air flow as well as root growth and function and plant growth (Reichert et al., 2009). The growth of shrubs on the shell sand improved the BD, DC and porosity because of the return of litter to the soil surface, the formation of a humus layer (An et al., 2010; Xia et al., 2012) and the accumulation of organic matter and biological activity (Mentges et al., 2016; van Hall et al., 2017). Furthermore, higher coverage decreased the effects of wind erosion and soil erosion as well as the effects of leaching by rainfall, and the presence of more residual roots increased the capillary

Principal component

− 0.851 0.990 0.997 0.534 0.994 − 0.177 0.983 0.989 0.619 0.985 0.991 0.532 0.937 0.955 0.884 74.239 74.239

ZJ

4. Discussion

Table 5 Contribution rate and factor loading of principal component of soil water storage and holding parameters.

Y(2)

PS

The soil water-holding capacity was significantly affected by the soil porosity, saturated water storage capacity, capillary water storage capacity, aggregate contents, D value of air-dried aggregates, permeability and water retention parameters. A single indicator could not accurately reflect the water-holding capacity and water storage capacity of the shell sand soil; thus, an integrated evaluation using multiple indicators was required. Based on the PCA, the indicators used in the integrated evaluation of the water-holding capacity and water storage capacity of the shell sand soil were the total soil porosity, capillary porosity, porosity ratio, MWD of the air-dried aggregates, MWD of the water-stable aggregates, D value of the air-dried aggregates, saturated water storage capacity, capillary water storage capacity, initial infiltration rate, steady infiltration rate and water retention parameter. A comprehensive evaluation of the soil water-holding capacity of the shrubbery stands was performed using the fuzzy membership function method, and Table 6 shows the membership function values of the main factors. The soil water-holding capacities and water storage capacities under the shrubbery stands were as follows: T. chinensis > Z. jujube > P. sepium > bare soil.

The BD, porosity, aggregate content, particle D values and soil water-holding capacity were correlated, and these factors had different effects on the soil water-holding capacity and water storage capacity under the different shrubbery stands on the shell ridges. A PCA of these factors was conducted (Table 5) to estimate the main factors that affect the soil water-holding capacity and water storage capacity of the shell ridges. The cumulative contribution of the two principle components was 94.8% and reflected most of the information on the measured indicators (Table 5). The first principal component Y (1) had accounted for 74.2% and was the largest principal component. The components with high factor loadings were the soil porosity, MWD of the waterstable aggregates, D value of the air-dried aggregates, saturated water storage capacity, capillary water storage capacity, permeability and water-holding capacity. The components of the second principle component Y (2) with high factor loadings included the MWD of the air-dried aggregates.

Y(1)

TC

⁎ TC: T. chinensis Lour; PS: P. sepium Bunge; ZJ: Z. jujuba var. spinosa; CK: Bare soil as control check.

3.7. Evaluation of the soil water-holding capacity and water storage capacity

Factors

Shrub types

321

Catena 156 (2017) 315–324

J. Xia et al.

shell sand soils with high D values had higher contents of fine sand and clay, and high soil clay contents are conducive to soil aggregate formation and improve soil aeration, permeability and capillary porosity (Xia et al., 2013). Moreover, a greater surface area per unit of soil particles corresponds to a higher soil adsorption of water molecules and a stronger soil water-holding capacity (Briar et al., 2011). With changes in the composition, size and content of the soil particles, threshold effects are observed for the D value of the soil particles, the looseness of the soil and the soil aeration properties. The D values of the soil types in China range from 1.834 to 2.904 and increase from sandy soil-loam-clay to loam-clay (Li and Zhang, 2000). The D values of the air-dried soil particles in the shell sand soil ranged from 1.6029 to 2.1664 and were significantly lower than those of loam, clay loam and clay. The D values of the shell sand soil under the shrubbery stands were in the order T. chinensis > P. sepium > Z. jujube and averaged 1.9452, which was only slightly higher than the lower limit of loamy sand (1.834) (Li and Zhang, 2000). This finding indicates that the shell sand soil contains more coarse particles and has a lower D value than loamy sand soil. The soil macroaggregates in the shell sand broke down into small aggregates after soaking in water and the soil texture became finer, whereas the D values of the water-stable aggregates increased significantly. Nevertheless, shrub growth on the shell ridges improved the D value in the same habitat, which indicates that the soils under the shrubbery stands had more heterogeneous particle size distributions. The soils also had an uneven texture, although they were finer and had higher D values in the shell sand. Consequently, these soils tended to form a good soil structure, especially under the T. chinensis shrub stand, which had the highest D value and was associated with the highest vegetation coverage of 75%. These soil types also had the thickest layer of decomposing litter, which promoted increases in the nutrient and silt-clay contents (the highest silt-clay content was 3.80%). Furthermore, the growth of tree roots influenced the soil physicochemical and biological properties and accelerated the rate of soil weathering and the formation of humus, which promoted the fixation of fine sand materials (Briar et al., 2011; Xia et al., 2013; Li et al., 2015). The canopy density and vegetation coverage of the P. sepium and Z. jujube shrub stands were low, and the shrub stands had low litter reserves, which led to the wind erosion of fine particles, increased the fractions of coarse-grained soil particles, and generated relatively uniform soil textures, low degrees of heterogeneity and lower D values. On the bare soil in the shell ridges, which was not covered by any vegetation, fine particles and nutrients were lost because of wind erosion, which promoted the development of coarse-grained soils that presented the highest contents of coarse sand, the lowest D values, the poorest water retention capabilities, and reduced plant growth.

porosity, which led to slight improvements of soil aeration (Briar et al., 2011; Abdul et al., 2014; Li et al., 2015). Exudates and organic matter, provided respectively by living and dead roots, can be used as the cementing agent of soil aggregate (Liu et al., 2009). At the same time, the extrusion and winding of fibrous roots play an important role in improving anti erosion and physical structure of soil (An et al., 2013; Liu et al., 2015). The BD of the bare soil was high, and the total porosity was low; these characteristics are mainly associated with the lack of vegetation coverage and the lack of exogenous organic substance inputs. BD formation and porosity were affected by the composition and content of the soil aggregates, whereas the quantity and size distribution of the soil aggregates were critical indicators of the rate and magnitude of physical processes, such as soil erosion, compactness, and compaction (Oades and Waters, 1991; Bertolino et al., 2010; Jiang et al., 2013); therefore, these factors were closely related to the resistance of soil to corrosion and the environmental quality of the soil (Bronick and Lal, 2005; Jordán et al., 2010). The largest fraction of the water-stable aggregates of the shell sand in the Yellow River Delta was 0.25–1.0 mm coarse sand (42%–57%) followed by fine sand (0.05–0.25 mm) and gravel, whereas the silt-clay-sized content was the lowest. The coarse sand- and fine sand-sized contents in the shell sand habitat was significantly higher than that of gravel and silt-clay; thus, the soil has the characteristics of loamy sand and can be categorized as gravelly coarse sandy loam (Li and Zhang, 2000). Macroaggregates larger than 0.25 mm generally represent the soil aggregate structure, and the water-stable aggregate structure represents the best soil structure, and its quantity is positively correlated with soil fertility (Six et al., 2000; An et al., 2010; Liu et al., 2015). The shell sand soils of the shrubbery stands were dominated by water-stable aggregates; > 0.25 mm water-stable macroaggregates accounted for 66%–78% of the total aggregates and mainly consisted of 1–0.25 mm coarse sand, which indicates that the shell sand soil under the shrubbery stands had a good soil aggregate structure. These findings indicate that the content of the > 0.25 mm water-stable aggregates is a major factor that controls the MWD of soil aggregates, which is consistent with the results of previous studies (Leutel et al., 2012; Jiang et al., 2013; Liu et al., 2015). T. chinensis performed better than Z. jujube in terms of the overall soil water stability and air-dried aggregates, whereas P. sepium performed relatively poorly, especially when compared with the soil water-stable aggregates of the T. chinensis shrub stand, which had the highest content, the best aggregation and the best stability. Compared with the bare soil, shrub growth altered the shell sand from large particles to fine particles and provided conditions that were more conducive to the formation of large water-stable aggregates and improved the stability of the soil structure, which was likely because of the higher vegetation coverage, the thick humus layer that formed from the forest litter, the penetration of plant roots and the greater plant secretions in the soil under the shrubbery stands (Liu et al., 2009; Liu et al., 2015). The soil particle fractal dimension not only depicts the soil particle sizes and pore size distribution (Tyler and Wheatcraft, 1992; Bittelli et al., 1999) but also reflects the soil hydraulic properties (Tyler and Wheatcraft, 1989), texture uniformity, aeration, and permeability (Briar et al., 2011; Tang et al., 2013). Previous studies have shown that the D value of soil particles in the shell sand soil was significantly negatively correlated with the BD, significantly positively correlated with the capillary porosity and total porosity but insignificantly correlated with the non-capillary porosity (Xia et al., 2013), and similar correlations were observed in this study. The Z. jujube shrub stand had the highest content of soil water-stable aggregates and a lower D value of water-stable aggregates, whereas the T. chinensis shrub stand had the highest soil D value. The shell sand soil had high porosity and good aeration and permeability, which led to a high D value. Moreover, as the BD increased, the D value decreased, which is inconsistent with the conditions of mountain soils under forest stands (Liu et al., 2009). The

4.2. Factors affecting the soil water-holding capacity and water storage capacity of different shrubbery stands The water storage capacity of soil is the main indicator of the potential water storage and regulation capacities (Abu and Abubakar, 2013; Shi et al., 2016; Zhang et al., 2016). The soil saturated storage reflects the function of vegetation in reducing surface runoff and preventing soil erosion (Zhang et al., 2006; Christian et al., 2011; Rosa et al., 2012). Soil water storage is closely correlated with BD and porosity, and an increase in BD suppresses increases in soil water content; the MWD of water-stable soil aggregates and capillary porosity both promote soil moisture (Liu et al., 2015). The shrubbery stands showed reduced surface runoff, and they improved the soil water storage of the shell sand and prevented soil erosion; however, the soil water storage capacities varied significantly by shrubbery type. The T. chinensis shrub stand (followed by the Z. jujube shrub stand) performed best in supplying effective water to the plants and had the best water conservation potential, and both of these parameters were conducive to the effective use of water by plant roots. The P. sepium shrub stand had 322

Catena 156 (2017) 315–324

J. Xia et al.

Table 7 Correlation coefficient between soil water storage and holding index and physical parameters. Soil water physical parameters

Bulk density Total porosity Capillary porosity Non-capillary porosity Void ratio Soil mean weight diameter of dryaggregate Soil mean weight diameter of waterstable aggregate Fractal dimension of dry-aggregate Fractal dimension of water-stable aggregate Total water-storage capacity Capillary water-storage capacity Non-capillary water-storage capacity First infiltration rate Stable infiltration rate Soil water holding capacity (Parameters B) ⁎ ⁎

Soil water storage and holding parameters Total water-storage capacity

Capillary waterstorage capacity

Non-capillary waterstorage capacity

First infiltration rate

Stable infiltration rate

Soil water holding capacity

−0.797 0.999** 0.989* 0.634 0.996** −0.185

− 0.882 0.987* 0.998** 0.505 0.985* − 0.295

− 0.030 0.628 0.515 0.999** 0.618 0.446

−0.962* 0.881 0.936 0.207 0.893 −0.345

− 0.957* 0.908 0.955* 0.259 0.917 − 0.339

−0.923 0.808 0.871 0.116 0.827 −0.252

0.968*

0.953*

0.627

0.886

0.906

0.848

0.959* 0.610

0.988* 0.520

0.401 0.791

0.976* 0.421

0.987* 0.443

0.930 0.447

1 0.988* 0.628 0.875 0.902 0.794

0.988* 1 0.498 0.933 0.953* 0.860

0.628 0.498 1 0.206 0.258 0.121

0.875 0.933 0.206 1 0.998** 0.982*

0.902 0.953* 0.258 0.998** 1 0.971*

0.794 0.860 0.121 0.982* 0.971* 1

P < 0.05;⁎⁎P < 0.01. Correlation is significant at the 0.05 level;⁎⁎Correlation is significant at the 0.01 level.

soil moisture retention capacity. This result indicated that shrub growth improved the soil structure, increased the soil porosity and MWD and decreased the BD, which led to increases in the water-holding capacity and available water content in the shell sand soil under the shrubbery stands. As a result, the water was more readily released and absorbed by plants in these soils. This study demonstrates that the shell sand soil of the T. chinensis shrub stand has the lowest BD and highest porosity values (Table 1), the highest soil water storage capacity (Fig. 4) and the highest initial and steady infiltration rates (Table 3 and Fig. 5). These factors led to the highest water retention capacity (Fig. 6 and Table 4) and the best soil hydrological regulation and storage functions (Table 6), which are consistent with the soil water storage and retention mechanisms of the grasslands of the Yellow River Delta (Liu et al., 2015). The soil water storage capacity (Fig. 4) and water-holding capacity (Table 6) of the Z. jujube shrub stand were significantly higher than those of the P. sepium shrub stand, whereas the soil porosity, porosity ratio (Table 1), permeability (Fig. 5 and Table 3) and water-holding capacity (Table 4) of the Z. jujube shrub stand were all lower than those of the P. sepium shrub stand. These results indicated that the changes in the soil water storage capacities of the Z. jujube and P. sepium shrub were not completely consistent with the changes in the BD and porosity and showed that the soil water-holding capacities of the two shrubbery stands were not only affected by the BD and porosity but were also likely associated with the soil aggregate structure, soil water content, organic matter content and root distribution (Xia et al., 2012; Alam et al., 2013; Vogelmann et al., 2013).

the poorest water storage and a low water storage capacity for plant growth and development. The correlation analysis (Table 7) showed that the saturated water content of the shell sand soil had significantly positive correlations with the total porosity and porosity ratio (P < 0.01); in addition, significantly positive correlations were observed with the MWD of the water-stable aggregates and the D values of the air-dried aggregates (P < 0.05). The capillary water storage had significantly positive correlations with the capillary porosity (P < 0.01) and the D values of the air-dried aggregates, the total porosity, the porosity ratio and the MWD of the water-stable aggregates (P < 0.05). The non-capillary water storage had significantly positive correlations with the non-capillary porosity (P < 0.01) and affected the order of non-capillary water storage. This study revealed that the initial infiltration rate was significantly positively correlated with a steady infiltration rate (P < 0.01) and the D value of the air-dried aggregates (P < 0.05) and was significantly negatively correlated with the BD (P < 0.05). The T. chinensis shrub stand had a higher impact on improving the BD and porosity, yielded the highest water storage capacity and consequently had the greatest effect in promoting the in situ infiltration and absorption of rainfall and surface run-off in soil. The correlation analysis of the indicators (Table 7) suggested that the soil water-holding capacity was closely correlated with the permeability, which in turn was closely correlated with the D value of the airdried aggregates. In addition, the D value of the air-dried aggregates was closely correlated with the capillary porosity because the saturation storage capacity was mainly controlled by the total porosity; therefore, the first category of the principle component of the waterholding capacity of the shell sand soil is represented by the soil porosity and the MWD of the water-stable aggregates. The amount of water that could be released by the soil with a low suction force was determined by the distribution of large pores and was mainly affected by the capillary force. The soil capillary porosity of the T. chinensis shrub stand was significantly higher than that of the P. sepium and Z. jujube shrub stands; therefore, under identical soil suction levels, the soil moisture under the T. chinensis shrub stand was significantly higher than that of the P. sepium and Z. jujube shrub stands. The water-holding capacity of the soil zones with moderate to high soil suction was mainly determined by the surface absorption of soil particles (Zhang et al., 2006; Abdul et al., 2014). The bare soil had the highest BD but the lowest porosity and MWD of water-stable aggregates; thus, it had the poorest

5. Conclusions The shrubbery stands on the shell ridges significantly affected the soil hydro-physical properties, infiltration characteristics and waterholding capacities and resulted in large differences that varied by stand. Compared with bare soil, the soil under the shrub stands showed reduced BD and BDLLWR and increased soil porosity ratios and DC values. These differences led to looser soil textures, significantly increased contents of silt-clay in the shell sand soil and improved stability of the aggregates. The T. chinensis shrub exhibited noteworthy improvements in the soil pore structure, which led to coordinated soil aeration and water permeability and the best overall permeability, 323

Catena 156 (2017) 315–324

J. Xia et al.

Biochem. 60, 1–9. Jordán, A., Zavala, L.M., Gil, J., 2010. Effects of mulching on soil physical properties and runoff under semi-arid conditions in southern Spain. Catena 81, 77–85. Leutel, S., Bouckaert, L., Buchan, D., Van Loo, D., Cornelis, W.M., Sanga, H.G., 2012. Manipulation of the soil pore and microbial community structure in soil mesocosm incubation studies. Soil Biol. Biochem. 45, 40–48. Li, D.C., Zhang, T.L., 2000. Fractal features of particle size distribution of soils in China. Soil Environ. Sci. 9, 263–265 (in Chinese with English abstract). Li, X.Z., Shao, M.A., Jia, X.X., Wei, X.R., He, L., 2015. Depth persistence of the spatial pattern of soil–water storage along a small transect in the Loess Plateau of China. J. Hydrol. 529, 685–695. Liu, X., Zhang, G.C., Heathman, G.C., Wang, Y.Q., Huang, C.H., 2009. Fractal features of soil particle-size distribution as affected by plant communities in the forested region of Mountain Yimeng, China. Geoderma 154, 123–130. Liu, Y.L., Li, C.L., Gao, M.X., Zhang, M., Zhao, G.X., 2015. Effect of different land-use patterns on physical characteristics of the soil in the Yellow River delta region. Acta Ecol. Sin. 35, 5183–5190 (in Chinese with English abstract). Martinez-murillo, J.F., Gabarron-galeote, M.A., Ruiz-sinoga, J.D., 2013. Soil water repellency in Mediterranean rangelands under contrasted climatic, slope and patch conditions in southern Spain. Catena 110, 196–206. Meldahl, K.H., 1995. Pleistocene shoreline ridges from tide-dominated and wavedominated coasts: northern gulf of California and western Baja California, Mexico. Mar. Geol. 123, 61–72. Mentges, M.I., Reichert, J.M., Rodrigues, M.F., Awe, G.O., Mentges, L.R., 2016. Capacity and intensity soil aeration properties affected by granulometry, moisture and structure in no-tillage soils. Geoderma 263, 47–59. Oades, J.M., Waters, A.G., 1991. Aggregate hierarchy in soils. Aust. J. of Soil Res. 29, 815–828. Reichert, J.M., Suzuki, L.E.A.S., Reinert, D.J., Horn, R., Hakansson, I., 2009. Reference bulk density and critical degree-of-compactness for no-till crop production in subtropical highly weathered soils. Soil Tillage Res. 102, 242–254. Reinert, D.J., Albuquerque, J.A., Reichert, J.M., Aita, C., Andrada, M.M.C., 2008. Limites críticos de densidade do solo para o crescimento de raízes de plantas de cobertura em argissolo vermelho. Rev. Bras. Ciênc. Solo. 32, 1805–1816. Rosa, J.D., Cooper, M., Darboux, F., Medeiros, J.C., 2012. Soil roughness evolution in different tillage systems under simulated rainfall using a semivariogram-based index. Soil Tillage Res. 124, 226–232. Saito, Y., Wei, H.L., Zhou, Y.Q., Nishimura, A., Sato, Y., Yokota, S., 2000. Delta progradation and chenier formation in the Huanghe (Yellow River) delta, China. J. Asian Earth Sci. 18, 489–497. Shi, D.M., Wang, W.L., Jiang, G.Y., Peng, X.D., Yu, Y.L., Li, Y.X., Ding, W.B., 2016. Effects of disturbed landforms on the soil water retention function during urbanization process in the three gorges reservoir region, China. Catena 144, 84–93. Six, J., Elliott, E.T., Paustian, K., 2000. Soil structure and soil organic matter II. A normalized stability index and the effect of mineralogy. Soil Sci. Soc. Am. J. 64, 1042–1049. Sperry, J.S., Hacke, U.G., 2002. Desert shrub water relations with respect to soil characteristics and plant functional type. Funct. Ecol. 16, 367–378. Suzuki, L.E.A.S., Reichert, J.M., Reinert, D.J., Lima, C.L.R., 2007. Grau de compactação, propriedades físicas e rendimento de culturas em Latossolo e Argissolo. Pesqui. Agropecu. Bras. 42, 1159–1167. Tang, Y.Q., Li, J., Zhang, X.H., Yang, P., Wang, J.X., Zhou, N.Q., 2013. Fractal characteristics and stability of soil aggregates in karst rocky desertification areas. Nat. Hazards 65, 563–579. Tian, J.Y., Xia, J.B., Sun, J.K., 2011. The Ecological Protection and Recovery Techniques of Shell Islands in Yellow River Delta (in Chinese). Chemical Industry Press, Beijing. Turcotte, D.L., 1986. Fractal fragmentation. J. Geog. Res. 91, 1921–1926. Tyler, S.W., Wheatcraft, S.W., 1989. Application of fractal mathematics to soil water retention estimation. Soil Sci. Soc. Am. J. 53, 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, 362–369. Vogelmann, E.S., Reichert, J.M., Prevedello, J., Consensa, C.O.B., Oliveira, A.É., Awe, G.O., Mataix-Solera, J., 2013. Threshold water content beyond which hydrophobic soils become hydrophilic: the role of soil texture and organic matter content. Geoderma 209, 177–187. Wang, J., Zhang, Y.H., Zhang, Z.H., 2015. Influences of intensive tillage on water-stable aggregate distribution on a steep hillslope. Soil Tillage Res. 151, 82–92. Xia, J.B., Kong, X.H., Lu, Z.H., Liu, J.T., Han, R.D., Zhu, J.F., 2012. Capacity of soil water storage and regulation under different density forests of Tamarix chinensis Lour in coastal wetland. Adv. Water Sci. 23, 563–569 (in Chinese with English abstract). Xia, J.B., Zhang, S.Y., Wang, R.R., Zhao, Y.Y., Sun, J.K., Liu, J.T., Liu, Q., 2013. Water ecology and fractal characteristics of soil particle size distribution of three typical vegetations in Shell Island. Acta Ecol. Sin. 33, 7013–7022 (in Chinese with English abstract). Xie, W.J., Zhao, Y.Y., Zhang, Z.D., Liu, Q., Xia, J.B., Sun, J.K., Tian, J.Y., Sun, T.Q., 2012. Shell sand properties and vegetative distribution on shell ridges of the southwestern coast of Bohai Bay. Environ. Earth Sci. 67, 1357–1362. Yang, P.L., Luo, Y.P., Shi, Y.C., 1993. Soil fractal character token by particle mass distribution. Chin. Sci. Bull. 38, 1896–1899 (in Chinese with English abstract). Zhang, Y.D., Liu, S.R., Ma, J.M., 2006. Water-holding capacity of ground covers and soils in alpine and sub-alpine shrubs in western Sichuan, China. Acta Ecol. Sin. 26, 2775–2781. Zhang, Y.W., Deng, L., Yan, W.M., Shangguan, Z.P., 2016. Interaction of soil water storage dynamics and long-term natural vegetation succession on the Loess Plateau, China. Catena 137, 52–60.

water storage and water-holding capacities. Shrub growth increased the amount of soil water-stable aggregates and decreased the content of the air-dried aggregates, which significantly improved the soil water storage capacity and water-holding capacity. The content of the water-stable aggregates larger than 0.25 mm was the principle factor in determining the MWD of the soil aggregates. The structure of the soil aggregates of the T. chinensis shrub stand was better than that of the Z. jujube, whereas the structure of the soil aggregates of the P. sepium shrub stand was the poorest. The content of coarse sand particles ranging from 1 to 0.25 mm in the shell sand habitat was the highest and had a relatively low D. Shrubbery growth improved the D value, which increased from 1.6029 for bare land to 1.8162–2.1664 for the shrubbery stands. The D values of the soil water-stable aggregates under the different shrubbery stands were significantly higher than that of the air-dried aggregates, whereas the D value of the soil particles under the T. chinensis shrub stand was the highest. The major indicators that reflect the shell sand soil water-holding capacity and water storage capacity could be sorted into two categories (principle components), with one category including the soil porosity and the MWD of the water-stable aggregates and the other including the MWD of the air-dried aggregates. These two categories accounted for a cumulative contribution of 94.82%. The soil water-holding capacity and water storage capacity of the shell ridges were in the order T. chinensis > Z. jujube > P. sepium > bare soil. Acknowledgements This work was supported by the National Natural Science Foundation of China [grant number 31370702]; the Key Project of Natural Science Foundation of Shandong Province [grant number ZR2015JL014] and the Science and Technology Projects of Shandong Province [grant number 2015GNC111022]. References Abdul, M.M., Saad, A.A., Boyan, K., Toby, W., 2014. Multiple on-line soil sensors and data fusion approach for delineation of water holding capacity zones for site specific irrigation. Soil Tillage Res. 143, 95–105. Abu, S.T., Abubakar, I.U., 2013. Evaluating the effects of tillage techniques on soil hydrophysical properties in Guinea Savanna of Nigeria. Soil Tillage Res. 126, 159–168. Alam, S., Sengupta, D., Kole, R.K., Bhattacharyya, A., 2013. Dissipation kinetics of tetraconazole in three types of soil and water under laboratory condition. Environ. Monit. Assess. 185, 9819–9824. An, S., Mentler, A., Mayer, H., Blum, W.E.H., 2010. Soil aggregation, aggregate stability, organic carbon and nitrogen in different soil aggregate fractions under forest and shrub vegetation on the Loess Plateau, China. Catena 81, 226–233. An, S.S., Darboux, F., Cheng, M., 2013. Revegetation as an efficient means of increasing soil aggregate stability on the Loess Plateau (China). Geoderma 209–210, 75–85. Antinoro, C., Arnone, Elisa, Noto, V., 2017. The use of soil water retention curve models in analyzing slope stability in differently structured soils. Catena 150, 133–145. Bertolino, A.V.F.A., Fernandes, N.F., Miranda, J.P.L., Souza, A.P., Lopes, M.R.S., Palmieri, F., 2010. Effects of plough pan development on surface hydrology and on soil physical properties in Southeastern Brazilian plateau. J. Hydrol. 393, 94–104. Bittelli, M., Campbell, G.S., Flury, M., 1999. Characterization of particle-size distribution in soils with a fragmentation model. Soil Sci. Soc. Am. J. 63, 782–788. Briar, S.S., Fonte, S.J., Park, I., Six, J., Scow, K., Ferris, H., 2011. The distribution of nematodes and soil microbial communities across soil aggregate fractions and farm management systems. Soil Biol. Biochem. 43, 905–914. Bronick, C.J., Lal, R., 2005. Soil structure and management: a review. Geoderma 124, 3–22. Christian, P., Jean-Claude, G., Ary, B., Ingrid, S., 2011. Mapping soil water holding capacity over large areas to predict potential production of forest stands. Geoderma 160, 355–366. Gardner, W.R., Hillel, D., Benyamini, Y., 1970. Post irrigation movement of soil water: I. Redistribution. Water Resour. Res. 6, 851–861. van Hall, R.L., Cammeraat, L.H., Keesstra, S.D., Zorn, M., 2017. Impact of secondary vegetation succession on soil quality in a humid Mediterranean landscape. Catena 149, 836–843. Institute of Soil Science, Chinese Academy of Sciences. ed, 1978. Soil Physical and Chemical Analysis (in Chinese). 511–512. Shanghai Science and Technology Presspp. 522–524. Jiang, Y.J., Sun, B., Jin, C., Wang, F., 2013. Soil aggregate stratification of nematodes and microbial communities affects the metabolic quotient in an acid soil. Soil Biol.

324