Effect of soil wetting conditions on seal formation ... - CSIRO Publishing

CSIRO PUBLISHING www.publish.csiro.au/journals/ajsr

Review Australian Journal of Soil Research, 2008, 46, 191–202

Effect of soil wetting conditions on seal formation, runoff, and soil loss in arid and semiarid soils — a review Meni Ben-HurA,C and Marcos LadoB A

Institute of Soil, Water and Environmental Sciences, the Volcani Centre, ARO, Bet-Dagan 50250, Israel. Area of Soil Science. Faculty of Sciences, University of A Coruna, A Zapateira s/n 15071, Spain. C Corresponding author. Email: [email protected] B

Contribution from the Agricultural Research Organization, the Volcani Center, no. 607/07 series

Abstract. Soil surface sealing is one of the main causes for low infiltration rate (IR) and high runoff and soil loss under raindrop impact conditions in arid and semiarid regions. Many studies have focused on the effects of soil properties on seal formation under fast wetting conditions. However, in the field, soils can be exposed to different wetting conditions, before an intense rainfall event, which can affect the role of the soil properties on seal formation. The present paper reviews the effects of different initial wetting conditions and their interactions with soil properties on seal formation, IR, runoff, and soil loss in smectitic soils. Fast wetting of soil causes aggregate slaking, which enhances seal formation, runoff, and soil loss under rainfall, mainly in soils with > 40% clay content. An increase in clay content of the soil increases aggregate strength, but at the same time increases the slaking forces. Hence, in soils with low clay content ( 40% clay content and high aggregate stability, slaking plays an important role in aggregate breakdown and seal formation. An increase of raindrop kinetic energy, from 8 to 15.9 kJ/m3, decreased the effect of the slaking forces on seal formation and runoff. It was suggested that the effects of raindrop kinetic energy and of the slaking forces on aggregate disintegration and seal formation are complementary. An increase in soil exchangeable sodium percentage (ESP), from 0.9 to 20.4%, decreased the effect of slaking forces on seal formation and runoff production under rainfall with 15.9 kJ/m3 kinetic energy. Probably, increasing the ESP increased the soil dispersivity, and therefore diminished the effect of the slaking forces on aggregate disintegration and seal formation. Aging (the time since wetting) of soil increased the stability of soil structure, decreased the seal formation, maintained high IR, and diminished soil loss amounts. These effects of soil aging depend on both the prewetting rate of the soil and soil texture.

Introduction Arid and semiarid regions are characterised by water scarcity and highly variable precipitation. In addition, the water resources in some parts of these regions are expected to decrease as a result of global warming. The soils in these regions are characterised by relatively low organic matter content, high levels of salinity and sodicity, and low vegetative cover, which makes them especially sensitive to structural degradation and, therefore, to changes in their hydraulic properties. One of these properties is the infiltration rate (IR), an important parameter in the soil phase of the hydrologic cycle since it not only determines the supply of water to the soil profile but also the amount of overland flow. Thus, IR affects both runoff and soil erosion, which can cause widespread land and water degradation problems in terms of water loss, soil degradation, and environmental pollution. Surface runoff could cause flooding downstream, and potentially could pollute good water sources. Moreover, local runoff within agricultural fields leads to poor water distribution in the fields, which could decrease the crop yield per mm of water (Letey et al. 1984; Ben-Hur et al. 1995). CSIRO 2008

Erosion of the upper soil layer and the transport of fertilisers and pesticides, which are adsorbed on the eroded sediment, from cultivated lands could decrease field fertility and increase pollution of water bodies downstream. Runoff occurs when the rainfall intensity exceeds the soil IR and the soil surface water-holding capacity. Soil erosion involves 2 major processes: (i) detachment of soil material from the soil surface; and (ii) transport of the resulting sediment. In interrill erosion, soil detachment is caused essentially by raindrop impact and soil transport by raindrop splash and runoff sheet flow (Watson and Laflen 1986). In the present review, only interrill erosion is considered. In arid and semiarid regions, the main factor that controls the soil IR under rainfall conditions is seal formation at the soil surface (Morin et al. 1981; Ben-Hur et al. 1985, 1987; Assouline and Mualem 1997). This seal is thin, only a few mm deep, and is characterised by greater density, higher strength, finer pores, and lower saturated hydraulic conductivity than the underlying soil (McIntyre 1958; Chen et al. 1980; Gal et al. 1984; Onofiok and Singer 1984; West et al. 1992; Wakindiki and Ben-Hur 2002), and its formation 10.1071/SR07168

0004-9573/08/030191

192

Australian Journal of Soil Research

M. Ben-Hur and M. Lado

decreases the IR of the soil (Morin et al. 1981; Assouline and Mualem 1997). Agassi et al. (1981) suggested that formation of a structural seal is a result of 2 complementary mechanisms: (i) a physical disintegration of surface soil aggregates, caused by the impact energy of the raindrops; and (ii) the physicochemical dispersion of soil-clays, which migrate into the soil with the infiltrating water and clog the pores immediately beneath the surface to form the ‘washed-in’ zone. The relative importance of the latter mechanism depends on the electrolyte concentration of the soil solution, and the exchangeable sodium percentage (ESP) of the soil surface. As the electrolyte concentration decreases and the ESP increases, seal formation and IR reduction during a rainstorm are amplified (Kazman et al. 1983). Aggregate stability is an index of the propensity of the aggregate to resist breakdown. The tendency of a soil to form a seal, and the amount of the resulting runoff and soil loss, depends on aggregate stability and, therefore, on soil properties such as soil texture (e.g. Ben-Hur et al. 1985; Romkens et al. 1995; Sharma et al. 1995), organic matter content (e.g. Ekwue 1991; Fullen 1991; Guerra 1994; Le Bissonnais and Arrouays 1997; Lado et al. 2004b), soil mineralogy (e.g. Stern et al. 1991; Ben-Hur et al. 1992; Mermut et al. 1995, 1997; Romkens et al. 1995; Wakindiki and Ben-Hur 2002; Norton et al. 2006), and soil salinity and sodicity (Agassi et al. 1981, 1985, 1994; Kazman et al. 1983; Ben-Hur et al. 1998). Many reviews have been published on the effects of these soil properties on aggregate stability (e.g. Kay and Angers 1999; Lado and Ben-Hur 2004) and seal formation, IR, and interrill erosion (e.g. Shainberg and Letey 1984; Sumner 1993; Lado and Ben-Hur 2004). Most of these studies were based on results obtained from air-dry soils that were initially exposed to fast wetting rates. However, under

field conditions the soils can be wetted at different rates, according to the different rain or irrigation intensities, or they may be subject to ageing (time since wetting) and have different antecedent moisture contents before being exposed to rainstorms with high intensity and high kinetic energy. These different wetting conditions could change the effects of the physical and chemical properties of the soil on seal formation, IR, runoff and soil loss. These effects are particularly important in arid and semiarid regions, where the rain intensity can be high and the dry periods between the rainstorms can be long, allowing the soil surface to dry before the next rain event. The present paper reviews and discusses the effects of different initial wetting conditions on the response of various soils to seal formation and to changes in IR, runoff, and interrill erosion under rainfall conditions. The soils discussed in this paper are smectitic soils (the dominant clay in the soil is smectite) collected from arid and semiarid regions. Their general properties are presented in Table 1. Wetting rate and antecedent moisture content When water is supplied to the soil surface by precipitation or irrigation, it typically penetrates the surface and is absorbed into successively deeper layers of the profile. Several factors can reduce IR with time during rainstorms or irrigation. The initial IR of dry soil is high because of the combined gravitational and matric potential gradients that draw water into the soil. As the water content in the soil increases and the wetting front becomes deeper, the matric potential gradient decreases, which, in turn, decreases the IR (Hillel 2004). For soils with a stable structure that does not deteriorate during wetting and exhibits a sharp wetting front, a simplistic equation that

Table 1. Particle size distribution, CaCO3 content, organic matter content (OM), cation exchange capacity (CEC), and exchangeable sodium percentage (ESP) of the studied soils Soil texture

Clay Loam Clay Clay Sand Loam Sandy Clay Clay Clay Loam Sandy Clay Clay Sand Loam Sandy Sandy Clay Clay

clay

clay

clay loam clay

Soil type

Typic Chromoxerert Calcic Haploxeralf Chromic Haploxerert Chromic Haploxerert Typic Rhodoxeralf Calcic Haploxeralf Chromic Haploxerert Typic Haploxerert Typic Haploxerert Tepic Haploxerert Calcic Haploxeralf Chromic Haploxerert Chromic Haploxerert Chromic Haploxerert Typic Haploxeralf Calcic Haploxeralf Typic Haploxeralf Chromic Haploxerert Chromic Haploxerert Chromic Haploxerert

Citation

Levy et al. 1997 Levy et al. 1997 Ben-Hur et al. 1998 Ben-Hur et al. 1998 Mamedov et al. 2001 Mamedov et al. 2001 Mamedov et al. 2001 Mamedov et al. 2001 Mamedov et al. 2001 Mamedov et al. 2001 Shainberg et al. 2003 Shainberg et al. 2003 Shainberg et al. 2003 Shainberg et al. 2003 Lado et al. 2004a Lado et al. 2004a Lado et al. 2004a Lado et al. 2004a Lado et al. 2004a Lado et al. 2004a

Particle size distribution (%) Sand Silt Clay 31 50 9 8 86 51 41 21 18 15 41 47 15 25 90 41 68 44 19 25

22 31 26 30 5 26 19 27 20 17 36 15 34 14 2 36 2 15 31 13

47 19 65 62 9 23 40 52 62 68 23 38 51 61 8 23 30 41 50 62

CaCO3

OM

(%) 15.1 18.2 6.8 6.6 1.8 16.4 13.8 16.9 10.8 1.5 18.2 9.6 20.2 5.0 0.6 18.0 0.8 10.7 15.4 4.9

2.1 1.5 3 3.9 0.4 1.1 1.0 1.2 1.3 3.5 1.2 1.1 1.8 1.7 0.6 2.1 0.9 3.4 2.2 3.8

CEC

ESP

(cmolc/kg)

(%)

38.7 17.5 – – 8.3 18.9 33.4 50.8 57.4 61.2 17.7 34.8 57.4 65.0 4.3 17.7 18.5 34.2 39.5 65.0

1.7 3.1 0.4 11.8 1.5–20.3 2.1–19.7 1.6–16.9 1.6–15.1 0.9–20.4 2.5–8.6 2.1 1.6 1.6 1.0 1.7 2.1 1.1 2.3 1.8 0.9

Soil wetting conditions and seal formation


IR ¼ IRc þ

b l

ð1Þ

where IR is infiltration rate, IRc is the steady state IR, l is the cumulative water infiltration, and b is a constant. In this case, the antecedent moisture content affects the soil IR by determining the matric potential gradient—the higher the antecedent moisture content, the smaller the matric potential gradient and the IR. It is important to emphasise here that, in the case of Eqn 1, no change in soil structure is considered during soil wetting. In contrast, for soil where surface structure is changed and a seal is developed, the change in IR with time during the rainstorm is described by Eqn 2 (Morin et al. 1981): IRt ¼ IRf þ ðIRi IRf Þ expðg p tÞ

ð2Þ

where IRt is the IR in time ti; IRf is the final IR, which is fairly constant and close to steady state IR; IRi is the initial IR; g is a soil coefficient that related to the structure stability of the soil surface; p is the rainfall intensity; and t is the time from beginning of the rainstorm. Formation of a seal at the soil surface results in a system of 2 layers with different hydraulic conductivities. In this case, the soil IR is strongly controlled by the effective hydraulic conductivity (Kc) of this 2-layer system. This Kc, under steadystate conditions, is related to the hydraulic conductivity (K) and the thickness (L) values of the individual layers according to Eqn 3: L1 þ L2 Kc ¼ L1 L2 þ K1 K2

ð3Þ

where the subscripts 1 and 2 refer to seal and the underlying layer, respectively. This equation demonstrates that the Kc of soil with a seal is not a simple average of K1 and K2. In this case, the Kc is strongly controlled by the layer with the lower K value. Thus, a thin seal layer at the soil surface with very low K can effectively reduce theIRofthesoil.Underconditionswhereasealmightbeformed,the soil antecedent moisture content can also affect the seal formation process and its properties by changing the wetting rate of the upper soil layer during the rainstorm. Interaction with clay content In addition to the impact energy of the raindrops and the physicochemical dispersion of soil-clay (Agassi et al. 1981; Shainberg and Letey 1984), another important mechanism of seal formation is aggregate disintegration caused by the slaking process (Lado et al. 2004a). This mechanism of aggregate breakdown appears when the aggregate is not strong enough to withstand the stresses produced by differential swelling, entrapped air, rapid release of heat during wetting, and the mechanical action of moving water (Emerson 1977; CollisGeorge and Green 1979; Kay and Angers 1999). These stresses are defined as slaking forces. The slaking process is controlled by the wetting rate of the soil; the faster the wetting rate, the stronger are the slaking forces. In order to determine the effects of the slaking process on seal formation, Lado et al. (2004a) exposed 3 soils with low

ESP ( 1, the higher the slaking value, the greater the role of the slaking forces on aggregate disintegration. The calculated SV for each of the 6 soils is presented in Fig. 2b as a function of the soil clay content; a significant linear increase of SV with an increase in clay content was observed. Increasing clay content in the aggregate probably increases the differential swelling, the effect of entrapped air, and the rapid release of heat during the fast wetting, which leads to stronger slaking forces. This could explain the positive relationship between SV and the clay content in the soils (Fig. 2b).

Slow wetting (a) Fast wetting (b)

2.0

y = 0.02x + 0.55 r 2 = 0.84 a

1.6 1.2 0.8

b

0.4 0.0 (b)

3.0

2.0

y = 0.44 + 0.04x r 2 = 0.85 1.0

0.0 0

where wi is the weight fraction of aggregates in the size class i with a diameter xi . The higher the MWD value, the higher was the aggregate stability. Aggregate strength as a function of clay content was determined by the slow wetting method, and is expressed by the line ‘a’ in Fig. 2a. Due to the slow wetting rate, slaking of the aggregates was minimised, and consequently, an increase in aggregate stability with the increase in clay content was observed because of the greater cementing effect of the clay within aggregates (Kemper and Koch 1966; Kay and Angers 1999). In contrast, when the dry aggregates were exposed to fast wetting, extensive slaking of the aggregates took place and there was no increase in the MWD values with an increase in clay content (line ‘b’ in Fig. 2a). To identify the role of slaking in aggregate breakdown, the slaking value (SV) for a given soil was calculated using Eqn 5: SV ¼

2.4

Mean weight diameter (mm)

decrease during wetting (Ghezzehei and Or 2000), which, in turn, should make the aggregates easier to break down. Despite this, applying rain to an air-dry soil with 62% clay led to more aggregate breakdown and a lower final IR than after prewetting the same soil at a slow rate. These results suggest that, when the clay soil is wetted at a fast rate, the slaking mechanism played an important role in seal formation, while in the soils with clay content 41%, the effect of slaking on seal formation was less important. These differences in the role of slaking on seal formation in soils with various clay contents were probably due to the concurrent effects of clay content on both the strength of the aggregates and also on the aggregate disintegration due to the slaking forces. This issue is discussed below. Aggregate stability of 6 soils with different clay contents and low ESP ( 2, 2–1, 1–0.5, 0.5–0.25, 0.25–0.1, 0.1–0.05, and < 0.05 mm, using Eqn 4:

Slaking ratio

194

10

20

30

40

50

60

Clay content in the soil (%) Fig. 2. (a) Mean weight diameter determined by fast wetting and by slow wetting methods as a function of clay content in the soil. (b) Slaking ratio as a function of the clay content in the soil. Bars indicate standard deviation (after Lado et al. 2004a).

Considering the concurrent effects of clay content on the aggregate stability and the slaking of the aggregates, the interaction between clay content and wetting rate effects on the final IR and seal properties observed in Fig. 1 can be explained as follows. In soils with 41% clay, the aggregate strength was low, and as a result, raindrop impact alone was strong enough to break down the aggregates and to form welldeveloped seals. Therefore, the differences in the final IR values between the 2 wetting rates for these soils were small (Fig. 1). Conversely, in the clay soil, aggregate strength was greater (Fig. 2a) and aggregate breakdown by raindrop impact, when slaking was prevented by wetting the soil slowly, was insufficient to completely disintegrate the aggregates and to form a well-developed seal (Fig. 1). However, slaking of the aggregates by fast wetting in this soil enhanced seal formation and significantly decreased the final IR compared with the slow wetting treatment (Fig. 1). The effect of wetting rate on soil loss was more pronounced than on final IR and total runoff. The total soil loss values from the former 3 soils subjected to the 2 former wetting rates are presented in Fig. 3. The soil loss was higher for the fast than for the slow wetting treatment for each soil texture (Fig. 3), and these differences were more pronounced than the differences in the runoff volumes (Table 2) and final IR values (Fig. 1) for the same wetting treatments and soil types. Whereas IR and runoff production depends on aggregate disintegration that enhances seal formation, for interrill erosion, aggregate disintegration that was caused by the slaking forces increased both the runoff


1200 1000

60

Fast wetting Slow wetting

Aa

50

Ba

800

Ca

600

Ab

Ab

400 200

Bb

0 23

41

195

Rain KE (kJ/m3) WR (mm/h) 8 15 2 8 64

(a)

Total runoff (mm)

Total soil loss (g/m2)

1400


40 30 20 10

62

Clay content in the soil (%) 0

Fig. 3. Total soil loss of the 3 soils exposed to fast wetting and slow wetting conditions. Values within a treatment between soils followed by the same upper case letter and values between treatments within a soil followed by the same lower case letter are not significantly different (a = 0.05) (after Lado et al. 2004a).

Wetting treatment Slow wetting Fast wetting

38

(b)

51

61

y = 0.02e0.11x + 1.11 r 2 = 0.99

25

Slaking factor

Table 2. Total runoff (mm) of 3 soils with different clay content and 2 wetting treatments (after Lado et al. 2004a) Values within a treatment between soils followed by the same lower case letter and values between treatments within a soil followed by the same upper case letter are not significantly different (a = 0.05)

22 30

20 Rain kinetic energy kJ/m3

15

8 15.9

10

y = 0.01e0.12x + 0.85 r 2 = 0.99

5

23

Clay content (%) 41

62

57.5bA 61.4abB

61.1bA 63.5bA

29.5aA 57.1aB

amount and the amount of material available for transport in the runoff. Therefore, soil loss was more sensitive than runoff production to the different wetting treatments (Fig. 2, Table 2).

Interaction with raindrop energy The first stage of seal formation is the disintegration of aggregates at the soil surface. Aggregate disintegration can occur due to the impact energy of the raindrops and also the slaking forces resulting from fast wetting of the soil surface (Lado et al. 2004a). The interaction effect between the kinetic energy of the raindrops and the wetting rate of the soil surface on seal formation, runoff production, and interrill erosion was studied by Shainberg et al. (2003). That study used a drip-type rainfall simulator and 4 soils with low ESP ( 1, the higher the slaking factor, the greater the importance of the slaking forces on seal formation. An increase in the clay content of the soils, up to ~60%, resulted in an exponential increase in the slaking factor in seal formation for the 2 kinetic energies (Fig. 4b). The results in Fig. 4b differ from those presented in Fig. 2b, which showed that an increase in the clay content in the soils, up to ~60%, resulted in a linear increase in SV (Fig. 2b). The results in Figs 2b and 4b suggest that the clay content of the soil has 2 main functions in seal formation under slaking conditions: (i) an increase in clay content increases linearly


Interaction with soil sodicity The effect of physicochemical dispersion of soil-clay in seal formation is strongly affected by the soil ESP. Agassi et al. (1981) and Kazman et al. (1983) found that under fast wetting conditions, an increase in soil ESP enhanced seal formation and decreased IR. Likewise, Ben-Hur et al. (1998) found that the final IR values of a vertisol with > 62% clay content, under fast wetting conditions, were affected by changes in soil sodicity, mainly when the ESP was < 5. However, changing the wetting rate of the soil can influence the effects of ESP on seal formation. The effects of ESP and the wetting rate of soil on seal formation and runoff production were studied by Mamedov et al. (2001) for 6 soils with clay contents ranging from 9 to 68% and ESPs ranging from 0.9 to 20.4%. In this study, each soil, with the various ESP values, was prewetted from below at 3 different wetting rates, 2, 8, and 64 mm/h, and then exposed to a rainstorm of 60 mm of deionised water, at an intensity of 36 mm/h and kinetic energy of 15.9 kJ/m3, using a drip-type

1200 Rain KE (kJ/m3) WR (mm/h) 8 15 2 8 64

(a) 1000

Total soil loss (g/m2)

the slaking forces that contribute to aggregate breakdown and to seal formation; (ii) disintegration of aggregates with higher clay contents increases the number of clay particles available to form the seal which, in turn, results in a more developed seal with lower hydraulic conductivity (Ben-Hur et al. 1985). These 2 functions, operating together, are probably the reason for the exponential increase in the effect of slaking forces on seal formation with increasing clay content of the soil (Fig. 4b). Another possible explanation for the differences between the effects of clay content on the SV (Fig. 2b) and on the slaking factor (Fig. 4b) is that the hydraulic conductivity of the soil is not linearly proportional to the pore radius (aggregates sizes). In this case, breakdown of aggregates and formation of a seal could result in an exponential increase in the slaking factor (Fig. 4b). The slaking factor was higher under a rain kinetic energy of 8 than 15.9, kJ/m3, and this difference increased with an increase in clay content (Fig. 4b). An increase in the kinetic energy of the raindrop decreased the importance of the slaking forces on seal formation, and this decrease was more pronounced in soils with higher clay content. This was probably because the effects of the kinetic energy of the raindrop and the slaking forces on aggregate disintegration and seal formation are complementary. The effects of wetting rate, raindrop kinetic energy, clay content, and their interactions on total soil loss over the entire rainstorm under the same experimental conditions (Shainberg et al. 2003) are presented in Fig. 5a. The slaking factors for soil loss (i.e. a ratio between the soil loss obtained after fast prewetting of 64 mm/h and that obtained after slow prewetting of 2 mm/h) increased exponentially with increasing clay content of the soil (Fig. 5b), as was the case for the slaking factor for seal formation and runoff production (Fig. 4b). However, in the case of the slaking factors for soil loss, no significant differences were found between the effect of raindrop kinetic energy of 8 and 15.9 kJ/m3 (Fig. 5b). Probably, a raindrop kinetic energy of 8 was sufficient to obtain the maximum effect of the slaking forces on soil loss.


800

600

400

200

0 22

38

51

61

30 (b) 25

Slaking factor

196

y = 0.006e0.13x + 1.91 r 2 = 0.79

20 15 10 5 0 0

20

40

60

Clay content (%) Fig. 5. (a) Effects of rain kinetic energy (KE) and soil wetting rate (WR) on soil loss as a function of clay content. Bar indicates a single confidence interval value at P = 0.05 (after Shainberg et al. 2003). (b) Ratio between soil loss after fast wetting and soil loss after slow wetting (slaking factor) as a function of the clay content in the soil.

rainfall simulator. The results of this study are presented in Fig. 6a. Again, from this figure, the total runoff amount generated during the entire rainstorm was used as an indicator of the processes of aggregate breakdown at the soil surface and seal formation. The soils were divided according to their sodicity into 4 groups: (i) soils with ESP < 2.5; (ii) soils with ESP 3.7–6.6; (iii) soils with ESP 7.5–10.2; (iv) soils with ESP > 15.1. Increasing the soil ESP increased the runoff amounts from most of the soils and for the different wetting rates (Fig. 6a). The increase in soil ESP enhanced soil dispersivity (van Olphen 1977), which in turn promoted aggregate disintegration at the soil surface and released clay particles. The dispersed clay particles moved into the upper layer of the soil profile, clogging the pores and forming a ‘washed-in’ zone, thus decreasing IR during the rainstorm (Agassi et al. 1981; Kazman et al. 1983). In order to study the interactions between the slaking forces, clay content in the soil, and ESP, the runoff ratio between the runoff amount obtained after fast prewetting at 64 mm/h and after slow prewetting at 2 mm/h (designated as the slaking factor for ESP) was determined for each soil, and ESP and was plotted as a function of the soil clay content (Fig. 6b).



197

(a)

Low ESP

60

Wetting rate (mm/h) 8 2

50

Medium ESP

64

40 30

Total runoff (mm)

20 10 0

High ESP

60

Very high ESP

50 40 30 20 10 0 8

22

40

52

62

68

8

22

40

52

62

68

(b) 50

Low ESP

y = 0.25e0.07x r 2 = 0.92

Moderate ESP

40

Slaking factor

High ESP

30

y = 4*10–6e0.23x + 1.09 r 2 = 0.99 20

10

0 0

20

40

60

Clay content (%) Fig. 6. (a) Effect of wetting rate on the total runoff from 6 soils as a function of clay content, for 4 exchangeable sodium percentage (ESP) levels. Bars indicate a single confidence interval value at a = 0.001 (after Mamedov et al. 2001). (b) Ratio between runoff after fast wetting and runoff after slow wetting (slaking factor) as a function of the clay content of soils with low (15) ESP.

Regarding these slaking factor values, the soils could be divided into 3 main groups [(i) soils with low ESP (15)], since no significant differences between the slaking factor values were found among the soils with ESP < 7 (Fig. 6b). These results indicated that an increase in the soil ESP decreased the effect of the slaking forces on seal formation and runoff

production, and that this decrease was more pronounced as the clay content of the soil increased. No effect of the slaking forces on seal formation was found in the soils with high ESP (>15) (Fig. 6b). Probably, increasing the ESP increased the dispersivity of the soil and, therefore, the disintegration of the aggregates by raindrop impact during the rainstorm (kinetic energy of 15.9 kJ/m3), to such a degree that seal

198



formation was more pronounced, even in soil with high clay content, thus reducing the importance of the slaking forces for seal formation.

30

Interaction with wetting rate The interaction between soil wetting rate and ageing duration effects on seal formation, IR, and soil loss was studied by Levy et al. (1997). In this study, air-dry soils, a loam soil with 19% clay and 31% silt and a clay soil with 47% clay and 23% silt, were first prewetted with 5 mm of deionised rainwater with a kinetic energy of 12.4 kJ/m3, at 3 different rates of 1, 6, or 30 mm/h, using a drip-type rainfall simulator. The soils were left to age for either 15 min or 18 h and were then subjected to a further 60 mm of rain at an intensity of 33 mm/h. The IR values of these soils as a function of cumulative rainfall for the different aging durations and wetting rates are presented in Fig. 7. In general, for the 2 soils and for both ageing durations, an increase in the prewetting rate resulted in decreased IR values. These results are consistent with the results obtained by Lado et al. (2004a) and Shainberg et al. (2003). The effects of ageing on IR curves depended on both the prewetting rate and the soil texture (Fig. 7). In the loam soil, prewetting at 1 mm/h reduced slaking of the aggregates and seal formation. In this case, no effect of ageing could be observed because the IR values exceeded the rainfall intensity (33 mm/h) for all ageing treatments. Increasing the ageing duration from 15 min to 18 h in this soil increased the IR values in the fast (30 mm/h) and medium (6 mm/h) wetting rates, and this increase was larger in the fast than the medium prewetting rate (Fig. 7). In contrast, when the clay soil was prewetted at a rate of 30 mm/h, the 2 ageing treatments resulted in similar IR

18 h ageing

25

Ageing process

Prewetting rate, mm/h 1 6 30

20 15 10

Infiltration rate (mm/h)

Another condition associated with the process of soil wetting that could affect soil structure stability is ageing (the time since wetting). The effect of ageing is partly attributed to the activity of soil microorganisms that produces soil conditioners, i.e. chemicals that enhance soil stability, during the ageing process (Martin et al. 1995). In addition, cohesion forces associated with capillary water have been hypothesised to account for soil strength in moist soils, and solid-phase bonds to account for aggregate stability (Kemper and Rosenau 1984; Kemper et al. 1987). Shainberg et al. (1996) studied the effect of ageing on a rill erodibility factor for 3 different montmorillonitic soils (clay, loam, and loamy sand soils) by means of a small hydraulic flume. Shainberg et al. (1996) found that, at water contents above air-dried conditions, soil erodibility decreased with ageing. The ageing effect was more effective in the clay soil than in the other soils. With no ageing (no soil prewetting), soil erodibilty was 3.26 103 s/m in the clay soil and 0.52 103 s/m in the loamy sand. When the ageing was increased to 4 and 24 h, the soil erodibility of the clay decreased to 0.24 103 s/m, but that of the loamy sand was not affected. Shainberg et al. (1996) related the decrease in the soil erodibilty to the development of biological and chemical cohesion forces between soil particles during ageing, which increased aggregate stability and decreased soil erodibility.

Loam soil

35

15 min ageing

5

Clay soil

35 30 25 20

15 min ageing

18 h ageing

15 10 5 0 0

10

20

30

40

50

60

Cumulative rainfall (mm) Fig. 7. Effect of wetting rate by rain (slow = 1 mm/h, medium = 6 mm/h, fast = 30 mm/h) and ageing on the infiltration rate of a loamy and a clayey soils as a function of cumulative rainfall. Bars represent 2 standard deviations (after Levy et al. 1997).

curves. When this soil was prewetted at 6 mm/h, the difference between the IR curves for the 2 ageing durations was largest, and when the prewetting rate was 1 mm/h, the differences between the IR curves were small. In this soil, slow prewetting of the soil with 1 mm/h minimised the slaking forces, and therefore, the IRs were high, even for the short ageing duration (15 min). In contrast, during fast prewetting (30 mm/h) of the clay soil, the slaking forces were stronger than the stabilising ones developed during the longer ageing time (18 h), resulting in extensive aggregate slaking, seal formation, and lowered IRs in this soil. Total soil loss resulting from interrill erosion from the clay and loam soils as a function of the prewetting rate for the 2 ageing durations are presented in Fig. 8. For both soils, interrill erosion increased linearly with an increase in prewetting rate. Likewise, increasing the ageing duration reduced the interrill erosion. The differences between soil losses for the 2 ageing durations were larger in the loam than the clay soil and increased as the wetting rate increased (Fig. 8). Interaction of ageing with raindrop impact energy and soil sodicity The interaction effect of ageing, raindrop impact energy, and soil ESP on seal formation and IR was studied by Ben-Hur et al. (1998) for a clay soil (Vertisol) with 65% clay content and ESP of 0.4 or 11.8%. The soils with different ESP values were prewetted at a fast rate to near



199

400 Clay soil

Soil loss (g/m2)

Loam soil 300

Duration of ageing 15 min 18 h

y = 9.42x + 68.81 r 2 = 0.78

y = 9.96x + 6.52 r 2 = 0.99

200 y = 1.84x + 18.68 r 2 = 0.74

100

y = 4.04x + 14.54 r 2 = 0.94

0 0

10

20

30

0

10

20

30

40

Wetting rate (mm/h) Fig. 8. Soil loss from a loamy and a clayey soils after 2 ageing treatments as a function of wetting rate of the soil (after Levy et al. 1997).

(a) No ageing

ESP 0.4 ESP 11.8

60

40 Covered soil

Infiltration rate (mm/h)

saturation with saline water (EC 5 dS/m and SAR similar to the soil ESP) from below, and then allowed to age for 0, 1, 4, 8, or 24 days. After the ageing process, the soils were exposed in a rainfall simulator to 58-mm rainstorms, with an intensity of 44 mm/h of deionised water, under 2 different conditions: (i) covered soil (zero raindrop impact energy), and (ii) uncovered soil (full raindrop impact energy of 18.1 J/mm.m2). The soils that were aged for 0 and 4 days were subjected to both kinds of rainstorm, while the soils exposed to the other ageing treatments were subjected only to zero impact energy rainfall. The IR values of the covered and uncovered soils, with both ESP values and no ageing, as a function of cumulative rainfall are presented in Fig. 9a. The IR values of the covered soils were higher than those of the uncovered soils at any cumulative rainfall depth for both ESP values. In the case of the uncovered soils, the raindrop impact broke down the aggregates at the soil surface and formed a seal that reduced the IR. In contrast, covering the soil surface prevented the mechanical breakdown of the aggregates at the soil surface, and therefore, a seal did not develop. Under such conditions, the IR is determined mainly by the saturated hydraulic conductivity of the soil layer (Hillel 1980). The IR values of the high-ESP soil were lower than those of the low-ESP soil at a given depth of cumulative rainfall for both covered and uncovered conditions. An increase in ESP increases the swelling and the dispersion of the clay in the soil (van Olphen 1977), a process that can enhance seal formation and blockage of the conducting pores in the soil matrix (Agassi et al. 1981; Kazman et al. 1983). Infiltration rates of the uncovered soils with high and low ESP as a function of cumulative rainfall after 4 days of ageing are presented in Fig. 9b. Comparison of the IRs of uncovered soils after 4 days of ageing (Fig. 9b) with those after no ageing (Fig. 9a) indicates that ageing affected the formation of the seal and its properties. For both ESP values, the IR values were higher, and the IR reduction was more gradual, after 4 days of ageing (Fig. 9b) than with no ageing (Fig. 9a). The final IR values of the uncovered soils with ESP of 0.4 and 11.8 and no ageing were 6 and 3 mm/h, respectively, compared with 14 and 12 mm/h, respectively, after 4 days of ageing.

20

Uncovered soil 0 (b) 4 days of ageing

ESP 0.4 ESP 11.8

30

Uncovered soil

20

0 0

10

20

30

40

50

60

Cumulative rainfall (mm) Fig. 9. (a) Infiltration rates of covered and uncovered soils with 2 exchangeable Na percentages (ESP) and without ageing and compaction, as a function of cumulative deionised-water rain. Bars indicate standard deviations (after Ben-Hur et al. 1998). (b) Infiltration rate of uncompacted, uncovered soil with high or low exchangeable Na percentage (ESP), after 4 days of ageing. Bars indicate standard deviations (after Ben-Hur et al. 1998).

Ben-Hur et al. (1998) also found that ageing affected the final IR of covered soils (no seal at the soil surface) that were subjected to the 58 mm rainstorm (Fig. 10). The final IR increased from 17 mm/h with no ageing to 42 mm/h after 28 days of ageing in the high-ESP soil. This effect was most pronounced during the first 4 days of ageing, after which the final IR was 90% of the value obtained after 28 days of aging. The EC of leachate collected from the soil at the end of the

200



Final infiltration rate (mm/h)

50

40

(2)

30

ESP 0.4

20

ESP 11.8

10 0

5

10

15

20

25

30

(3)

Duration of ageing (days) Fig. 10. Final infiltration rate of soils with 2 exchangeable sodium percentage (ESP) levels, as a function of duration of ageing. Bars indicate standard deviation (after Ben-Hur et al. 1998).

rainstorm for this treatment was 4.7 dS/m, which is higher than the flocculation value of the soil (van Olphen 1977). Therefore, the low final IR of the high-ESP soil with no ageing was mainly the result of clay swelling. The effect of ageing duration on the final IR of the low-ESP soil (Fig. 10) could not be determined because the final IR was higher than the rainfall intensity for all ageing treatments. Ben-Hur et al. (1998) concluded that the beneficial effect of ageing in maintaining high IR (Figs 9 and 10) could not be related to dissolution of CaCO3 and alteration of the soil pH. They suggested that under near-saturation conditions, a physico-chemical mechanism supplements the microbiological mechanism in binding soil particles together. Montmorillonite platelets are long and flexible, and they can interact at several cohesive junction points, such as at the edges or planar surfaces with high charge density (van Olphen 1977; Keren et al. 1988). Conditions of high water content, and extended ageing periods, are favourable for clay-to-clay contact, which increases the resistance of the soil aggregates to breakdown forces.

(4)

(5)

(6)

Summary and conclusions (1) An increase in soil wetting rate increases the aggregate disintegration as a result of the slaking mechanism. This mechanism plays an important role in seal formation, and runoff and soil loss generation, mainly in soils with> 41% clay content.Thesedifferencesintheroleoftheslakingmechanism in soils with various clay contents are due to the concurrent effects of clay content on both the strength of the aggregates and the aggregate disintegration due to the slaking forces. An increase in clay content enhances the cementing forces between the primary particles in an aggregate, which, in turn, increases its aggregate stability. In contrast, increasing clay content leads to stronger slaking forces. In soils with 41% clay, the aggregate strength is low, and therefore, raindrop impact alone is strong enough to break down the aggregates and to form well-developed seals. In this case, the role that the slaking mechanism plays in aggregate disintegration is less important in seal formation.

(7)

Conversely, in soils with > 41% clay content, aggregate breakdown by raindrop impact only is insufficient to completely disintegrate the aggregates and to form a welldeveloped seal. Therefore, disintegration of the aggregates and seal formation by the slaking forces in these soils is important. The effect of wetting rate on soil loss is more pronounced than on runoff. Whereas IR and runoff production depend on aggregate disintegration that enhances seal formation, for interrill erosion, aggregate disintegration increases both the runoff and the amount of material available for transport in the runoff. Therefore, soil loss is more sensitive than runoff to slaking forces. The clay content in soil could have 2 main functions in seal formation under slaking conditions: (i) an increase in clay content increases linearly the slaking forces that contribute to aggregate breakdown and to seal formation; (ii) disintegration of aggregates with high clay contents increases the number of clay particles available to form the seal, which, in turn, result in a more developed seal with lower hydraulic conductivity. These 2 functions, which operate together, lead to an exponential increase in the effect of slaking forces on seal formation with increasing clay content in the soil. An increase in kinetic energy of the raindrop (from 8 to 15.9 kJ/m3) decreases the effect of slaking forces on seal formation and runoff amount. This suggests that the effects of the kinetic energy of the raindrop and the slaking forces on aggregate disintegration and seal formation are complementary. An increase in soil ESP (from 0.9 to 20.4%) decreases the effect of slaking forces on seal formation and runoff production under high raindrop kinetic energy (15.9 kJ/m3), and this decrease is more pronounced as the clay content in the soil increases. The slaking forces have no effect on seal formation in soils with ESP > 15%. Probably, increasing the ESP increased the soil dispersivity and, therefore, diminished the effect of the slaking forces on aggregate disintegration and seal formation. Effects of ageing on seal formation, IR, and soil loss depend on both the prewetting rate of the soil and soil texture. In loam soil, increasing the ageing duration from 15 min to 18 h increased the IR values in fast (30 mm/h) and medium (6 mm/h) wetting rates of the soil. In contrast, in clay soil the slaking forces during fast prewetting (30 mm/h) were stronger than the stabilising ones developed during ageing of 18 h, resulting in extensive aggregate slaking, seal formation, and low IR. It is suggested that under nearsaturation conditions, a physico-chemical mechanism supplements the microbiological mechanism in binding soil particles together. Conditions of high water content and extended ageing periods are favourable for microorganism activity and clay-to-clay contact, which increase the aggregate stability. The effects of wetting rate on soil structure degradation, seal formation, IR reduction, and increasing of runoff and soil loss amounts could have practical agronomic impacts. The wetting rate of soil in irrigated fields could be controlled by irrigation management and methods. In soils sensitive


to aggregate disintegration by slaking forces, using drip or mini-sprinkler irrigation systems, which maintain slow wetting rates, could minimise the soil structure degradation and the increase in runoff and soil loss. References Agassi M, Bloem D, Ben-Hur M (1994) Effect of drop energy and soil and water chemistry on infiltration and erosion. Water Resources Research 30, 1187–1193. doi: 10.1029/93WR02880 Agassi M, Morin J, Shainberg I (1985) Effect of raindrop impact energy and water salinity on infiltration rates of sodic soils. Soil Science Society of America Journal 49, 186–190. Agassi M, Shainberg I, Morin J (1981) Effect of electrolyte concentration and soil sodicity on the infiltration rate and crust formation. Soil Science Society of America Journal 45, 848–851. Assouline S, Mualem Y (1997) Modeling the dynamics of seal formation and its effect on infiltration as related to soil and rainfall characteristics. Water Resources Research 33, 1527–1536. doi: 10.1029/96WR02674 Ben-Hur M, Agassi M, Keren R, Zhang J (1998) Compaction, aging, and raindrop-impact effects on hydraulic properties of saline and sodic vertisols. Soil Science Society of America Journal 62, 1377–1383. Ben-Hur M, Plaut Z, Levy GJ, Agassi M, Shainberg I (1995) Surface runoff, uniformity of water distribution, and yield of peanut irrigated with a moving sprinkler system. Agronomy Journal 87, 609–613. Ben-Hur M, Shainberg I, Bakker D, Keren R (1985) Effect of soil texture and CaCO3 content on water infiltration in crusted soils as related to water salinity. Irrigation Science 6, 281–284. doi: 10.1007/ BF00262473 Ben-Hur M, Shainberg I, Morin J (1987) Variability of infiltration in a field with surface-sealed soil. Soil Science Society of America Journal 51, 1299–1302. Ben-Hur M, Stern R, Van der Merwe AJ, Shainberg I (1992) Slope and gypsum effects on infiltration and erodibility of dispersive and nondispersive. Soil Science Society of America Journal 56, 1571–1576. Chen Y, Tarchitzky J, Brower J, Morin J, Banin A (1980) Scanning electron microscope observations on soil crusts and their formation. Soil Science 130, 49–55. Collis-George N, Green RSB (1979) The effect of aggregate size on the infiltration behaviour of a slaking soil and its relevance to ponded irrigation. Australian Journal of Soil Research 17, 65–73. doi: 10.1071/ SR9790065 Ekwue EI (1991) The effects of soil organic matter content, rainfall duration and aggregates size on soil detachment. Soil Technology 4, 197–207. doi: 10.1016/0933-3630(91)90001-4 Emerson WW (1977) Physical properties and structure. In ‘Soil factors in crop production in a semi-arid environment’. (Eds JS Russell, EL Greacen) pp. 78–104. (University of Queensland Press: Queensland) Fullen MA (1991) Soil organic matter and erosion processes on arable loamy sand soils in the West Midlands of England. Soil Technology 4, 19–31. doi: 10.1016/0933-3630(91)90037-N Gal M, Arcon L, Shainberg I, Keren R (1984) The effect of exchangeable sodium and phosphogypsum on the structure of soil crust – scanning electron-microscope observation. Soil Science Society of America Journal 48, 872–878. Ghezzehei TA, Or D (2000) Dynamics of soil aggregate coalescence governed by capillary and rheological processes. Water Resources Research 36, 367–379. doi: 10.1029/1999WR900316 Guerra A (1994) The effect of organic matter content on soil erosion in simulated rainfall experiments in W. Sussex, UK. Soil Use and Management 10, 60–64. doi: 10.1111/j.1475-2743.1994.tb00460.x Hillel D (1980) ‘Application of soil physics.’ (Academic Press: New York) Hillel D (2004) ‘Introduction to environmental soil physics.’ (Academic Press: San Diego, CA)


201

Kay BP, Angers DA (1999) Soil structure. In ‘Handbook of soil science’. (Ed. ME Sumner) pp. A-229–A-269. (CRC Press: New York) Kazman Z, Shainberg I, Gal M (1983) Effect of low levels of exchangeable Na and applied phosphogypsum on the infiltration rate of various soils. Soil Science 35, 184–192. Kemper WD, Koch EJ (1966) ‘Aggregate stability of soils from western USA and Canada.’ USDA Technical Bulletin No. 1355. (US Government Printing Office: Washington, DC) Kemper WD, Rosenau RC (1984) Soil cohesion as affected by time and water content. Soil Science Society of America Journal 48, 1001–1006. Kemper WD, Rosenau RC, Dexter AR (1987) Cohesion development in disrupted soils as affected by clay and organic matter content and temperature. Soil Science Society of America Journal 51, 860–867. Keren R, Shainberg I, Klein E (1988) Settling and flocculation value of sodium montmorillonite particles in aqueous media. Soil Science Society of America Journal 52, 76–80. Lado M, Ben-Hur M (2004) Soil mineralogy effects on seal formation, runoff and soil loss. Applied Clay Science 24, 209–224. doi: 10.1016/ j.clay.2003.03.002 Lado M, Ben-Hur M, Shainberg I (2004a) Soil wetting and texture effects on aggregate stability, seal formation and erosion. Soil Science Society of America Journal 68, 1992–1999. Lado M, Paz A, Ben-Hur M (2004b) Organic matter and aggregate size interactions in infiltration, seal formation and soil loss. Soil Science Society of America Journal 68, 935–942. Le Bissonnais Y (1996) Aggregate stability and assessment of soil crusting and erodibility: 1. Theory and methodology. European Journal of Soil Science 47, 425–437. doi: 10.1111/j.1365-2389.1996.tb01843.x Le Bissonnais Y, Arrouays D (1997) Aggregate stability and assessment of soil crustability and erodibility: II. Application to humic loamy soils with various organic carbon contents. European Journal of Soil Science 48, 39–48. doi: 10.1111/ j.1365-2389.1997.tb00183.x Letey J, Vaux HJ Jr, Feinerman E (1984) Optimum crop water application as affected by uniformity of water infiltration. Agronomy Journal 76, 435–441. Levy GJ, Levin J, Shainberg I (1997) Prewetting rate and aging effects on seal formation and interrill soil erosion. Soil Science 162, 131–139. doi: 10.1097/00010694-199702000-00006 Mamedov AI, Levy GJ, Shainberg I, Letey J (2001) Wetting rate and soil texture effect on infiltration rate and runoff. Australian Journal of Soil Research 39, 1293–1305. doi: 10.1071/SR01029 Martin JP, Martin WP, Page JB, Raney WA, De Met JD (1995) Soil aggregation. Advances in Agronomy 7, 1–37. McIntyre DS (1958) Permeability measurements of soil crust formed by raindrop impact. Soil Science 85, 158–189. Mermut AR, Luk SH, Romkens MJM, Poesen JWA (1995) Micromorphological and mineralogical components of surface sealing in loess soils from different geographic regions. Geoderma 66, 71–84. doi: 10.1016/0016-7061(94)00053-D Mermut AR, Luk SH, Romkens MJM, Poesen JWA (1997) Soil loss by splash and wash during rainfall from two loess soils. Geoderma 75, 203–214. doi: 10.1016/S0016-7061(96)00091-2 Morin J, Benyamini Y, Michaeli A (1981) The effect of raindrop impact on the dynamics of soil surface crusting and water movement in the profile. Journal of Hydrology 52, 321–335. doi: 10.1016/0022-1694(81) 90178-5 Norton LD, Mamedov AI, Huang CH, Levy GJ (2006) Soil aggregate stability as affected by long-term tillage and clay mineralogy. Advances in Geoecology 39, 422–429. Onofiok O, Singer MJ (1984) Scanning electron microscope studies of surface crusts formed by simulated rainfall. Soil Science Society of America Journal 48, 1137–1143.

202



Romkens MJM, Luk SH, Poesen JWA, Mermut AR (1995) Rainfall infiltration into loess soils from different geographic regions. CATENA 25, 21–32. doi: 10.1016/0341-8162(94)00039-H Shainberg I, Goldstein D, Levy GJ (1996) Rill erosion dependence on soil water content, aging and temperature. Soil Science Society of America Journal 60, 916–922. Shainberg I, Letey J (1984) Response of soils to sodic and saline conditions. Hilgardia 52, 1–57. Shainberg I, Mamedov AI, Levy GJ (2003) Role of wetting rate and rain energy in seal formation and interrill erosion. Soil Science 168, 54–62. doi: 10.1097/00010694-200301000-00007 Sharma PP, Gupta SC, Foster GR (1995) Raindrop-induced soil detachment and sediment transport from interrill areas. Soil Science Society of America Journal 59, 727–734. Stern R, Ben-Hur M, Shainberg I (1991) Clay mineralogy effect on rain infiltration, seal formation and soil losses. Soil Science 152, 455–462. doi: 10.1097/00010694-199112000-00008

Sumner ME (1993) Sodic soils – new perspectives. Australian Journal of Soil Research 31, 683–750. doi: 10.1071/SR9930683 van Olphen (1977) ‘An introduction to clay colloid chemistry.’ 2nd edn (Interscience Publications: New York) Wakindiki IIC, Ben-Hur M (2002) Soil mineralogy and texture effects on crust micromorphology, infiltration and erosion. Soil Science Society of America Journal 66, 897–905. Watson DA, Laflen JM (1986) Soil strength, slope and rainfall intensity effects on interrill erosion. Transactions of the American Society of Agricultural Engineers 29, 98–102. West LT, Chiang SC, Norton LD (1992) The morphology of surface crusts. In ‘Soil crusting: chemical and physical processes.’ Advances in Soil Science. (Eds ME Sumner, BA Stewart) pp. 73–92. (Lewis Publishers: Boca Raton, FL)

Manuscript received 24 October 2007, accepted 25 February 2008

http://www.publish.csiro.au/journals/ajsr