Published online 27 April 2006
Published in Vadose Zone J 5:554-563 (2006)
DOI: 10.2136/vzj2005.0083
© 2006 Soil Science Society of America
677 S. Segoe Rd., Madison, WI 53711 USA
ORIGINAL RESEARCH
Modeling the Relationship between Soil Bulk Density and the Water Retention Curve
S. Assouline*
Institute of Soil, Water and Environmental Sciences, A.R.O., the Volcani Center, P.O.B. 6, Bet Dagan 50250, Israel
* Corresponding author (vwshmuel{at}agri.gov.il)
Contribution of the Agricultural Research Organization, Institute of Soil, Water and Environmental Sciences, Bet Dagan, Israel, No. 607/05.
Received 10 July 2005.
 |
ABSTRACT
|
|---|
Increases in the soil bulk density during compaction may influence many aspects of the soil–water–plant–atmosphere system. An empirical approach is suggested that could model the effect of an increase in soil bulk density on the water retention curve (WRC). Two expressions of the WRC are considered. Relationships between their parameters and the bulk density of a compacted soil were calibrated and validated against experimental WRC data of soils at various levels of compaction. These relationships enable a relatively good prediction of the effect of bulk density on the WRC. The relationship between the pore size distribution index and the coefficient of variation of the WRC, previously found for a wide range of soil types, is appropriate also when soils at different bulk densities are considered. The silt/clay content ratio was found to be an important factor in a quantitative expression of soil strength.
Abbreviations: WRC, water retention curve
|
INTRODUCTION
|
|---|
THE SOIL BULK DENSITY, 
, is defined as the mass of an oven-dry sample of undisturbed soil per unit of bulk volume (ISSS Working Group, 1998, p. 153). It is a basic soil physical property, used as an indicator of soil porosity and compactness. An increase in the soil bulk density is a common feature of agricultural soils. Increases occur at a variety of scales and may result from various natural as well as artificial processes. The bulk density of the soil around a root increases as the soil particles are pushed away by the growing root (Dexter, 1987; Bruand et al., 1996). The compaction of soil around the root can affect many physical, chemical, and biological processes and properties (Gli
ski and Lipiec, 1990). During rainfall, the impacts of raindrops on a bare soil lead to the formation of a dense layer at the soil surface (Tackett and Pearson, 1965; Bresson et al., 2004; Assouline, 2004b). This layer affects infiltration (Morin and Benyamini, 1977; Baumhardt et al., 1990; Assouline, 2004b), evaporation (Bresler and Kemper, 1970; Jones et al., 1994; Assouline and Mualem, 2003), and seedling emergence (Weaich et al., 1992; Hadas, 2004). Grazing animals also apply pressure to the soil surface, causing the compaction of the upper 50 to 150 mm of soil under pasture (Greenwood and McKenzie, 2001). Cycles of wetting and drying after tillage may also cause the soil bulk density to increase because of natural soil reconsolidation, thereby affecting the hydraulic properties of the soil profile (Mapa et al., 1986; Rousseva et al., 1988; Or et al., 2000). Vehicular traffic involved in modern management practices induces mechanical compaction of soils (Lindstrom and Voorhees, 1995; van Dijck and van Asch, 2002), changing their basic hydraulic properties and affecting water, heat, and gas exchange (Warkentin, 1971; Willis and Raney, 1971; Stepniewski et al., 1994), root growth (Voorhees et al., 1975; Dexter, 1987; Lipiec and Hatano, 2003), and consequently crop production and environment quality (Håkansson et al., 1988; Soane and van Ouwerkerk, 1995).
From the above, it appears that an increase in influences many aspects of the soil–water–plant–atmosphere system. It is, therefore, essential to develop an ability to quantify and predict these effects to account for compaction in agricultural, hydrological, and environmental systems. Different factors, such as initial water content (Perdok et al., 2002), transient stress (Horn and Baumgartl, 1999; Or and Ghezzehei, 2002), and the number of trips (Lenhard, 1986; Startsev and McNabb, 2001), were found to affect soil structure during compaction by vehicular traffic. The bulk density change due to compaction is an integrative variable that reflects the total change in the voids volume of the soil. Subtle changes in the voids volume distribution, tortuosity, or connectivity could still occur during compaction, especially during elastic deformation, while no corresponding information may be noticeable in terms of changes in bulk density (Lenhard, 1986). However, the premise of this study is that a measurable change in bulk density will have a major effect on the WRC. Therefore, an empirical approach is suggested that quantifies and predicts the effect of an increase in bulk density on one of the most important soil properties, the water retention curve.
|
APPROACHES IN MODELING THE EFFECT OF BULK DENSITY ON THE WATER RETENTION CURVE
|
|---|
The WRC describes the relationship between two fundamental state variables of water in soil, the capillary head, 
, and the volumetric water content, 
. Furthermore, different models allow definition of the hydraulic conductivity function in terms of the WRC (Mualem, 1986; Assouline, 2001, 2004a). Therefore, the WRC is a fundamental hydraulic characteristic of a soil, indispensable for the solution of equations that describe flow processes in soils.
Approaches to model the effect of the increase in the soil bulk density on the WRC are very limited. Gupta and Larson (1979) and Rawls et al. (1983) presented regression equations to predict the effects of soil texture, bulk density, and organic matter content on the water content, , at given values of . These equations can be used to estimate changes in the WRC caused by compaction or reconsolidation, but their use as predictive tools "is awaiting future research" (Green et al., 2003, p. 7).
Rajkai et al. (1996) used a pedotransfer function to predict the WRC. The best result was obtained when fitted cumulative particle-size data, clay and silt fractions, and the bulk density were used. Pachepsky et al. (1998) found that inserting a penetration resistance parameter into pedotransfer functions improved the WRC estimates based on soil texture and bulk density. However, the approach could not account for the effects of soil structure on the pedotransfer function estimates, nor for changes in soil pore/void distribution resulting from compaction
Ahuja et al. (1998) proposed two methods to estimate the WRC of a tilled soil of known bulk density from that of the untilled soil. They estimated changes in the parameters of the expression of Brooks and Corey (1964) for the WRC, namely, the pore-size distribution index, 
, and the air entry value, a, based on the assumptions that (i) tillage does not significantly affect a and (ii) increases with tillage only at the wet end of the WRC between = a and = 10a. These assumptions are in disagreement with data showing a clear increase in a with compaction (Laliberte et al., 1966; Or et al., 2000).
Additional approaches for estimating the effects of bulk density on the WRC resulted from studies of the formation of dense seal layers at the surface of a bare soil exposed to high-energy rainfall (Baumhardt et al., 1990; Mualem and Assouline, 1989). The subscript "c" is used in the following to indicate a compacted state. The approach of Baumhardt et al. (1990) was based on the following steps: (i) estimate experimentally the seal saturated hydraulic conductivity; (ii) based on this estimate, compute the seal saturated water content, sc, using the Kozeny–Carman relationship, and the water entry value, ac, using the Poiseuille equation; (iii) determine the Brooks and Corey (1964) exponent c for the seal layer graphically assuming a proportional increase in the initial water content of the seal, ic, with bulk density. The approach of Mualem and Assouline (1989) was more conceptual by involving an explicit relationship between the soil parameters needed to express the WRC and the soil bulk density. This model was used by Assouline et al. (1997) to quantify the effect of compaction on the hydraulic properties of two Brazilian Oxisols. The model assumes a power relationship between ac and the relative soil bulk density, and a linear relationship between c and the change in soil bulk density. However, this linear relationship had to be calibrated for each specific soil; therefore, the calibrated model could quantify the effect of compaction on the WRC, but could not be used to predict this effect only from knowledge of the initial WRC of the soil before compaction. The main objective of the present study is to present an empirical approach that can both quantify and predict the effect of soil bulk density changes on the WRC. The approach is applied to the WRC expressions suggested by both Brooks and Corey (1964) and Assouline et al. (1998, 2000).
|
PROPOSED MODEL
|
|---|
The WRC model of Assouline et al. (1998) considers that the particle volume distribution in soils results from a series of sequential fragmentations. Given the assumptions that the fragmentation process is uniform and random, that the probability for a particle to be fragmented is proportional to its volume, and that particle size and pore size are related by a power function, the following expression is proposed for the WRC, Se(), which relates the effective degree of saturation Spore-size = ( – r)/(s – r) to :
 | [1] |
where 
and µ are two fitting parameters, s is the saturated volumetric water content, r is the residual water content, and L is the capillary head corresponding to a very low water content, L, that represents the limit of the domain of interest of the WRC. In our study, |L| = 15 bar and L = r. Insight into the pore size distribution characterizing the soil can be obtained from the first and second moments of the WRC expression in Eq. [1], namely, the mean, rG, and the variance, 
2:
 | [2] |
 | [3] |
where 
= µ and 
denotes the 
function. The first moment is proportional to the mean pore radius in the soil, while the second moment is proportional to the variance of the distribution of the pore radii in the soil. The ratio, 
, between the square root of the second and the first moment is the coefficient of variation of the pore size distribution:
 | [4] |
For a given homogeneous soil with an initial bulk density, , the application of pressure increases the soil bulk density to a value c that depends on the applied load and the initial soil water content, i. The saturated water content of the compacted soil, sc, is directly related to the initial saturated water content of the soil before compaction, s, and to the compacted bulk density, c:
 | [5] |
where s is the solid particle density. When s is equal to the soil porosity, n, Eq. [5] indicates that sc = nc, the porosity of the compacted soil. When s< n, because of entrapped air, for example, Eq. [5] implies that (sc/i) = (nc/n).
The residual water content, r, is considered to be mainly a function of the surface area of the soil particles. Therefore, as a first approximation, r can be considered practically unaffected by the increase in when expressed on a weight basis (Mualem and Assouline, 1989). Consequently, the volumetric residual water content of the compacted soil, rc, is given by
 | [6] |
When the soil is compacted, the two parameters c and µc representing the WRC according to Eq. [1] are assumed to be functions, f1 and f2, of the initial soil parameters and µ, and the relative bulk density (c/) as follows:
 | [7] |
 | [8] |
Consequently, the WRC of the compacted soil and the corresponding coefficient of variation, c, can be computed from Eq. [1]
through [4], in which c and µc replace and µ.
The Brooks and Corey (1964) expression for the WRC is
 | [9] |
where a is generally taken as the air entry value and is the pore-size distribution index. This expression for the WRC can also be used for the WRC of a compacted soil, provided that appropriate ac and c values replace a and . In this regard, the relationships for sc and rc given by Eq. [5] and [6] are applicable. An expression relating the air entry value of the compacted soil, ac, to the relative bulk density, as suggested by Mualem and Assouline (1989),
 | [10] |
can also be used, or adjusted according to available data. Finally, Assouline (2005) showed that a strong relationship exists between the pore-size distribution index, , and the coefficient of variation, :
 | [11] |
The hypothesis of the present study is that there is a relationship between c and c for the compacted state as well. Therefore, estimating c by means of Eq. [2], [3], [4], [7], and [8] and applying the relationships between ac and c (Eq. [10]) and between c and c should enable predictions of the WRC of a compacted soil with the Brooks and Corey (1964) expression. In principle, the suggested relationships can be used also to derive the WRC of a compacted soil based on the van Genuchten (1980) expression in those cases where the equivalences between the van Genuchten and the Brooks and Corey parameters, as suggested by van Genuchten (1980) and Morel-Seytoux et al. (1996), are applicable.
|
METHODOLOGY AND DATA
|
|---|
The methodology used for calibration and validation of the different relationships was as follows: A curve fitting procedure was applied to the relevant data from the calibration set to determine the suggested expressions in Eq. [7], [8], [10], and [11]. These fitted equations were next used to predict the properties of the compacted variants of the soils of the validation data set.
The suggested relationships were calibrated against experimental data from Laliberte et al. (1966), Reicovsky et al. (1980), and Assouline et al. (1997) (Table 1). Laliberte et al. (1966) measured the main drying WRC of three soils—Columbia (coarse-loamy, mixed, superactive, nonacid, thermic Oxyaquic Xerofluvents) sandy loam, Touchet (coarse-silty, mixed, superactive, mesic Cumulic Haploxerolls) silt loam, and an unconsolidated sand—at various bulk densities. The different bulk densities were obtained by vibrating the soil columns after filling them with air-dried sieved soil. Since Soltrol "C", a light hydrocarbon oil, was used as the wetting fluid to prevent swelling or clay dispersion, the capillary pressure values were corrected to account for the differing densities of Soltrol and water. Reicovsky et al. (1980) measured the main drying WRC of Barnes (fine-loamy, mixed, superactive, frigid Calcic Hapludolls) loam at four bulk densities ranging from 0.99 to 1.59 g cm–3; they achieved the various levels of compaction by means of a standard laboratory press with pistons that slightly compressed moist disturbed soil samples simultaneously from both ends. Assouline et al. (1997) measured the main wetting and drying curves of two compacted Oxisols from Cascavel and Palotina, Brazil; their soil was compacted by applying a pressure of 10 bars on disturbed soil samples that had been pre-equilibrated at = –0.32 bar.
The model for estimating the effects of bulk density on the WRC was validated by comparing the predicted WRCs of compacted soils with the corresponding data from another experimental data set different from the one used for calibration. The validation data set comprised measured WRCs of (i) mechanically compacted disturbed soil samples of Bet Dagan sandy loam and (ii) undisturbed samples taken from tilled and untilled soil layers of Kitasato sandy loam (Moroizumi and Horino, 2004), Mattapex (fine-silty, mixed, active, mesic Aquic Hapludults) silt loam (Hill, 1990), and Cotto clay (Or et al., 2000; Snyder et al., 2000). The Bet Dagan soil samples were compacted to various bulk densities by vibrating soil columns that had been filled with air-dried sieved soil, while their WRCs were measured with a tension table. Information on the mechanical composition of the soils and their initial and compacted bulk densities is given in Table 2. The mechanical composition of the Kitasato sandy loam was provided by Moroizumi (personal communication, July 2004). For each of the soils, the lowest bulk density was considered the initial bulk density, , with the remaining values of c representing the various levels of compaction.
The Se() expressions given by Eq. [1] and [9] were fitted to the measured data. The values for s and sc, corresponding to the initial state and to the various levels of compaction applied during the experiments, were the reported volumetric water contents at = 0. In all cases, the value of L was set equal to –15 bar, while for r and rc the reported volumetric water contents at = L were used. The corresponding values of and µ, and of c and µc were determined by best-fit. The respective and c values were computed by means of Eq. [2], [3], and [4]. The soils properties, the best-fit values of the parameters in Eq. [1] and [9], and the computed value of the coefficients of variation, and c, are presented in Table 1 for the calibration data set. The mechanical composition of the Barnes soil was reported by Young (1984).
The applied curve fitting procedure was an iterative nonlinear regression using the Levenberg–Marquardt method to minimize the sum of the squares of the differences between observed and computed values of the dependent variable (Glantz and Slinker, 1990). For each data set, the fitted curve (shown by a solid line) and the 95% confidence limits resulting from the regression (dashed lines) along with the calibration data (solid circles) are presented. The data from the validation data set will be plotted on the same figure (open circles) to check the applicability of the fitted relationships within the 95% confidence limits.
|
RESULTS
|
|---|
Calibration of the Model
The required relationships (Eq. [7], [8], [10], and [11]) were derived from an analysis of the best-fit results on the calibration data set (Table 1). The ratio (c/) was related to the relative bulk density (c/) according to
 | [12] |
This expression is depicted in Fig. 1
. A similar power function was assumed to relate the ratio of (µc/µ) to (c/):
 | [13] |
However, in contrast to the expression for , Eq. [13] was found to be soil-type dependent, leading to a different value of 
for each soil. The values of were found to be inversely related to the ratio between the silt content, SC, and the clay content, CC, according to the expression:
 | [14] |
This equation (Fig. 2
) was adopted to complete the relationship given by Eq. [13].
View larger version (14K):
[in this window]
[in a new window]
|
Fig. 2. Plot of the power in Eq. [13] for each soil of the calibration data set vs. the corresponding silt/clay content ratio (solid circles) and the fitted expression (Eq. [14]) (solid line). The dashed lines represent the 95% confidence limits of the fitted curve; the open circles represent the validation data.
|
|
In terms of the parameters of the Brooks and Corey (1964) model (Eq. [9]), our analysis revealed the following relationship between (ac/a) and the relative bulk density (c/) (Fig. 3
):
 | [15] |
which closely resembles Eq. [10].
The coefficients of variation, c, for the various compaction levels of each soil were computed from the respective best-fit values of c and µc (Eq. [2], [3], and [4]). The () relationship based on the data of all 34 soils used in the study of Assouline (2005) ("+" symbols in Fig. 4
) is:
 | [16] |
This equation is plotted in Fig. 4, along with the (c,c) points corresponding to the calibration (full circles) and validation (open circles) data sets. It appears that the () expression in Eq. [16] remains valid for the compacted soils. This strengthens previous finding on the relationship between the pore size distribution index, , and the WRC characteristic, (Assouline, 2005), and that is not merely an empirical parameter but rather strongly depends on the statistical characteristics of the pore size distribution. Equation [16] was adopted in the following to represent the c(c) relationship. Since is mainly related to µ (Eq. [2], [3] and [4]), a strong relationship between and µ can be expected, as was shown by Assouline (2005). This relationship hence could provide an alternative estimate for c. However, the c(c) expression given by Eq. [16] is preferred as it relates c to a statistical characteristic of the WRC, rather than to an empirical parameter.
Figure 5
depicts the reproduced WRCs for the various compaction states of Columbia sandy loam and Touchet silt loam soils, computed according to Eq. [1] (top graphs) and [9] (bottom graphs), while applying the relationships given by Eq. [5], [6], and [12] through [16]. The overall agreement between the curves and the measured data is good, indicating that the model successfully reproduced the effect of the increase in bulk density on the WRC. The performance of the model was essentially the same for the two respective WRC expressions, although it was better for Eq. [9] in the case of Touchet silt loam, which has a relatively high bulk density. This indicates that the relationship between (ac/a) and (c/) and the suggested method to estimate c based on c accurately reproduce the effect of increases in bulk density on the WRC when expressed in terms of the Brooks and Corey (1964) model.
View larger version (27K):
[in this window]
[in a new window]
|
Fig. 5. Reproduced WRCs (solid line) for different compaction levels of Columbia sandy loam and Touchet silt loam soils of the calibration data set, using Eq. [1] (top) and Eq. [9] (bottom), and the corresponding measured data (solid circles). The open circles and dashed line represent the soil at its initial bulk density, .
|
|
Validation of the Model
Using the WRC parameters of each soil at its initial bulk density, , as they were determined by the best-fit procedure, the various parameters corresponding to the compacted state, c, were estimated by means of the relationships presented above: Eq. [5] and [6] for sc and rc; Eq. [12], [13], and [14] for c and µc; and Eq. [15] and [16] for ac and c. The fitted WRCs at and the predicted WRCs at c for the mechanically compacted sandy loam soil are depicted in Fig. 6a
and 6b, along with the respective measured data. For the expressions of both Assouline et al. (1998) (Eq. [1], Fig. 6a) and Brooks and Corey (1964) (Eq. [9], Fig. 6b), the model has provided good predictions of the WRCs of the compacted soil, although the overall performance of Eq. [1] was somewhat better than that of Eq. [9]. This indicates that the effect of compaction can be satisfactorily predicted with the suggested relationships.
View larger version (16K):
[in this window]
[in a new window]
|
Fig. 6. Predicted WRCs (solid line) of the mechanically compacted sandy loam soil of the validation data set using (a) Eq. [1] and (b) Eq. [9] compared with the corresponding measured data (solid circles). The open circles and dashed line represent the soils at their initial bulk density .
|
|
The predictive capability of the model is further illustrated for the case where changes in soil bulk density resulted from tillage practices. The measured data for the tilled and untilled soil samples, the corresponding fitted WRCs for , and the predicted WRCs for c by means of Eq. [1] and [9] are depicted in Fig. 7a
and 7b. The performance is good for the clay and sandy loam soils, and acceptable for the silt loam soils. In this case too, the overall performance of Eq. [1] was somewhat better than that of Eq. [9].
The suggested expressions were calibrated against data representing the effect of mechanical compaction on soil structure. The changes in soil structure resulting from natural soil reconsolidation after tillage to a given bulk density c are not necessarily similar to those caused by mechanical compaction to the same bulk density. Therefore, it is conceivable that the calibrated model may not perform as well for reconsolidation as compared with mechanical compaction. However, the results in Fig. 7 show that the main trends are predicted, thus indicating that the approach proposed in this study may be applied, at least as a first approximation, to reconsolidation problems as well.
|
DISCUSSION
|
|---|
An increase in soil bulk density due to compaction or reconsolidation reshapes the pore size distribution and consequently the water retention curve. During the compaction process, soil aggregates can move, deform, or collapse as a result of the applied stress. The movement of the aggregates relative to each other may be the main cause for a rapid increase in bulk density at low applied stresses, as generally observed and modeled (Bailey et al., 1986; Fritton, 2001; Assouline, 2002). It may also explain that the main impact of vehicular trafficking on soil bulk density occurs during the first few trips (Lenhard, 1986). This process is likely to affect mostly the larger pores. Consequently, most of the changes in the pore size distribution occur at the expense of larger pores, with concomitant reductions in water retention at relatively high capillary heads near saturation (Bruand and Cousin, 1995). Equations. [12] and [15] provide a quantitative estimate of this trend. The parameters a and , which are strongly related (Assouline, 2005), are good representatives of the larger pores. The power functions describing the response of these two parameters to the increase in soil bulk density illustrate the strong reduction of the larger pores. As a direct consequence of this effect, but also due to changes at the smaller pores level, the relative fraction of smaller pores in the distribution becomes more important following an increase of . This deformation, and especially the fragmentation of aggregates as it produces smaller units that create additional micropore spaces, increase the relative proportion of smaller pores, and consequently, the water retention at low capillary heads (Hill and Sumner, 1967). This trend is expressed by the dependency of the parameters and µ, which are also strongly interrelated (Assouline, 2005), on . These dependencies are a function of soil type, and more precisely, of the silt/clay ratio.
The rearrangement of aggregates also directly affects soil strength characteristics. Soil strength is determined by a large set of factors such as initial water content, organic matter content, pH, the presence of oxides, and clay type and particle size (Kemper and Roseneau, 1984; McBride, 1989; O'Sullivan, 1992; Tessier, 1991). The impact of particle type and size on soil strength can be expressed through the effect of the silt/clay ratio (Eq. [14]) on the suggested dependencies of µ and , on . Soil strength is traditionally explained in terms of the absolute contents of the soil mechanical components rather than their ratios. Ibanga et al. (1980) examined the effect of particle size on soil hardness of 17 artificial soil-texture mixtures that comprised various combinations of dried sand, silt, and clay fractions, and calculated the modulus of rupture, s, of these mixtures according to Reeve (1965). They concluded that soil strength was directly related to the percentage of clay, less strongly so to the percentage of sand, and inversely related to the percentage of silt. Reanalysis of the data of Ibanga et al. (1980) (Fig. 8
) showed that s was strongly related to the silt/clay ratio. An expression similar to Eq. [14] could be fitted to the data:
 | [17] |
This result confirms the empirical relationships given by Eq. [14] and indicates that the silt/clay ratio may play an important role in determining the resistance of soils to deformation or fragmentation.
View larger version (15K):
[in this window]
[in a new window]
|
Fig. 8. Plot of the modulus of rupture, s, as a function of the silt/clay content ratio for the data set of Ibanga et al. (1980) (solid circles) and the fitted expression (Eq. [17]) (solid line). The dashed line represents the 95% confidence limits of the fitted curve.
|
|
The suggested expressions for the effect of soil bulk density on the WRC pertain to two different models of the WRC, Eq. [1] and [9]. Adopting the capillary model, the radius of a pore, r, is inversely proportional to the capillary head, . Consequently, the pore size distribution, PSD, corresponding to a given WRC can be derived and expressed in terms of dSe(r)/dr. The PSDs corresponding to the uncompacted and compacted sandy loam soils (Fig. 6), and resulting from the application of Eq. [1] and [9], are depicted in Fig. 9a
. Using Eq. [9] for the WRC assumes an exponential-like PSD, but truncated at the values of r corresponding to a and ac. Using Eq. [1] for the WRC leads to a continuous Weibull-type PSD (Assouline et al., 1998). The impact of compaction on the PSD, as applied to the two WRC models, is illustrated in Fig. 9b, where the differences between the PSD of the compacted state and that of the uncompacted one are shown for Eq. [1] and [9]. The negative parts of the curves reflect the decreases in the larger pores, while the positive parts represent increases in the smaller pores. The assumed changes in the PSD resulting from an increase in are highly dependent on the selected WRC model. Equation [9] assumes a practically constant reduction in large pores from 0.100 to 0.065 mm, balanced by an essentially constant increase in the smaller pores range. On the contrary, Eq. [1] suggests a gradually increasing reduction in larger pores, from 0.150 to 0.027 mm, and a gradually increasing augmentation of the smaller pores. These differences could have significant consequences on the evaluation of the hydraulic conductivity functions of compacted soils.
View larger version (22K):
[in this window]
[in a new window]
|
Fig. 9. (a) Pore size distributions for the uncompacted (dashed lines) and compacted (solid lines) sandy loam soil, derived from the WRC expressions in Eq. [1] and Eq. [9] and shown in Fig. (6). (b) Differences between the pore size distributions for the compacted and uncompacted sandy loam soil, presented in (a), for the WRC expressions in Eq. [1] (thicker line) and Eq. [9] (thinner line).
|
|
|
CONCLUSIONS
|
|---|
An increase in the soil bulk density affects many aspects of the soil–water–plant–atmosphere system. Approaches to modeling and predicting the relationship between soil bulk density and WRC are very limited. This study suggests relationships between the parameters of two WRC models and the soil bulk density. These relationships quantify and predict the effect of increases in the bulk density on the soil WRC, given that the WRC data for an initial or reference bulk density are available.
The suggested relationships were calibrated using data corresponding to mechanical compaction of disturbed soil samples and were validated against data for mechanical compaction a well as for natural soil reconsolidation after tillage. The relationships predicted the WRCs of the compacted soils of the validation data set satisfactorily and also appear to account for the main effects of a bulk density increase on the WRC.
The suggested relationships simulate a decrease in the fraction of larger pores and a resulting decrease in water retention at high capillary heads, as well as an increase in smaller pores and the related increase in water retention at relatively low capillary heads. However, the simulated extent of these changes depends on the involved WRC model. The relationships for the parameters µ and were found to depend on soil type, and, more precisely, on the silt/clay ratio, which is one of the factors determining soil strength.
Compaction is a complex process affected by soil properties, initial conditions, and the manner in which stress is transmitted to the soil. Changes in soil structure may occur during compaction, especially during elastic deformations, without necessarily being reflected by measurable corresponding changes in terms of bulk density. However, the premise of this study was that a measurable change in bulk density still will account for the major effect on the WRC. The results presented here tend to support this premise, at least at a reasonable level of accuracy, and suggest our proposal is suitable for a large number of agricultural, hydrological, and environmental applications. Still, the performance of our approach could be improved in the future when more experimental data and a better quantitative expression of the role of the different factors in compaction become available.
|
ACKNOWLEDGMENTS
|
|---|
Thanks to L. Abramovitz for his technical assistance, and to A. Shaviv for his help in measuring the WRCs of the Bet Dagan sandy loam. The author also thanks Robert Lenhard and two anonymous reviewers for their constructive comments, and Rien van Genuchten for improving the writing style. This research was supported by Research Grant no. US-3393-03 R from BARD, the United States-Israel Binational Agricultural Research and Development Fund; this support is gratefully acknowledged.
|
REFERENCES
|
|---|
- Ahuja, L.R., F. Fiedler, G.H. Dunn, J.G. Benjamin, and A. Garrison. 1998. Changes in soil water retention curves due to tillage and natural reconsolidation. Soil Sci. Soc. Am. J. 62:1228–1233.[Abstract/Free Full Text]
- Assouline, S. 2001. A model of soil relative hydraulic conductivity based on water retention curve characteristics. Water Resour. Res. 37:265–271.[CrossRef]
- Assouline, S. 2002. Modeling soil compaction under uniaxial compression. Soil Sci. Soc. Am. J. 66:1784–1787.[Abstract/Free Full Text]
- Assouline, S. 2004a. Correction to "A model of soil relative hydraulic conductivity based on water retention curve characteristics". Water Resour. Res. 40:W02901. doi:10.1029/2004WR003025.[CrossRef]
- Assouline, S. 2004b. Rainfall-induced soil surface sealing: A critical review of observations, conceptual models, and solutions. Available at www.vadosezonejournal.org. Vadose Zone J. 3:570–591.[Abstract/Free Full Text]
- Assouline, S. 2005. On the relationship between the pore size distribution index and characteristics of the soil hydraulic functions. Water Resour. Res. 41:W07019. doi:10.1029/2004WR003511.[CrossRef]
- Assouline, S., and Y. Mualem. 2003. Effect of rainfall induced soil seals on the soil water regime: Drying interval and subsequent wetting. Transp. Porous Media 53:75–94.[CrossRef]
- Assouline, S., J. Tavares-Filho, and D. Tessier. 1997. Effect of compaction on soil physical and hydraulic properties: Experimental results and modeling. Soil Sci. Soc. Am. J. 61:390–398.[Abstract/Free Full Text]
- Assouline, S., D. Tessier, and A. Bruand. 1998. A conceptual model of the soil water retention curve. Water Resour. Res. 34:223–231.
- Assouline, S., D. Tessier, and A. Bruand. 2000. Correction to "A conceptual model of the soil water retention curve". Water Resour. Res. 36:3769.[CrossRef]
- Bailey, A.C., C.E. Johnson, and R.L. Schafer. 1986. A model for agricultural soil compaction. J. Agric. Eng. Res. 33:257–262.[CrossRef]
- Baumhardt, R.L., M.J.M. Römkens, F.D. Whisler, and J.Y. Parlange. 1990. Modeling infiltration into a sealing soil. Water Resour. Res. 26:2497–2505.[CrossRef]
- Bresler, E., and W.D. Kemper. 1970. Soil water evaporation as affected by wetting methods and crust formation. Soil Sci. Soc. Am. J. 34:3–8.
- Bresson, L.M., C.J. Moran, and S. Assouline. 2004. Use of bulk density profiles from X-radiography to examine structural crusts models. Soil Sci. Soc. Am. J. 68:1169–1176.[Abstract/Free Full Text]
- Brooks, R.H., and A.T. Corey. 1964. Hydraulic properties of porous media. Hydrol. Pap. 3. Colorado State Univ., Fort Collins.
- Bruand, A., and I. Cousin. 1995. Variation of textural porosity of a clay-loam soil during compaction. Eur. J. Soil Sci. 46:377–385.
- Bruand, A., I. Cousin, B. Nicoullaud, O. Duval, and J.C. Begon. 1996. Backscattered electron scanning images of soil porosity for analyzing soil compaction around roots. Soil Sci. Soc. Am. J. 60:895–901.[Abstract/Free Full Text]
- Dexter, A.R. 1987. Compression of soil around roots. Plant Soil 97:401–406.[CrossRef]
- Fritton, D.D. 2001. An improved empirical equation for uniaxial soil compression for a wide range of applied stresses. Soil Sci. Soc. Am. Proc. 65:678–684.
- Glantz, S.A., and B.K. Slinker. 1990. primer of applied regression and analysis of variance. McGraw-Hill, New York.
- Gliski, J., and J. Lipiec. 1990. Soil physical conditions and plant roots. CRC Press, Boca Raton, FL.
- Green, T.R., L.R. Ahuja, and J.G. Benjamin. 2003. Advances and challenges in predicting agricultural management effects on soil hydraulic properties. Geoderma 116:3–27.[CrossRef][ISI]
- Greenwood, K.L., and B.M. McKenzie. 2001. Grazing effects on soil physical properties and the consequences for pasture: A review. Aust. J. Exp. Agric. 41:1231–1250.[CrossRef]
- Gupta, S.C., and W.E. Larson. 1979. Estimating soil water retention characteristics from particle size distribution, organic matter percent and bulk density. Water Resour. Res. 15:1633–1635.[CrossRef]
- Hadas, A. 2004. Seedbed preparation—The soil physical environment of germination seeds. p. 3–49. In R.L. Benech-Arnold and R.A. Sanchez (ed.) Handbook of seed physiology, applications to agriculture. Haworth Press, New York.
- Håkansson, I., W.B. Voorhees, and H. Riley. 1988. Vehicle and wheel factors influencing soil compaction and crop response in different traffic regimes. Soil Tillage Res. 11:239–282.[CrossRef]
- Hill, R.L. 1990. Long-term conventional and no-tillage effects on selected soil physical properties. Soil Sci. Soc. Am. J. 54:161–166.[Abstract/Free Full Text]
- Hill, J.N.S., and M.E. Sumner. 1967. Effect of bulk density on moisture characteristics of soils. Soil Sci. 103:324–328.
- Horn, R., and T. Baumgartl. 1999. Dynamic properties of soils. p. A19–A51. In M.E. Sumner (ed.) Handbook of soil science. CRC Press, Boca Raton, FL.
- Ibanga, I.J., O.W. Bidwell, W.L. Powers, A.M. Feyerherm, and W.W. Williams. 1980. Soil consistence: Effect of particle size. Soil Sci. Soc. Am. J. 44:1124–1126.[Abstract/Free Full Text]
- ISSS Working Group. 1998. World reference base of soil resources: Introduction. 1st ed. International Society of Soil Science, International Soil Reference and Information Center, and FAO, United Nations, Acco, Leuven, the Netherlands.
- Jones, O.R., V.L. Hauser, and T.W. Popham. 1994. No-tillage effects on infiltration, runoff, and water conservation on dryland. Trans. ASAE 37:473–479.
- Kemper, W.D., and R.C. Roseneau. 1984. Soil cohesion as affected by time and water content. Soil Sci. Soc. Am. J. 48:462–464.
- Laliberte, G.E., A.T. Corey, and R.H. Brooks. 1966. Properties of unsaturated porous media. Hydrology paper. 17. Colorado State Univ., Fort Collins.
- Lenhard, R.J. 1986. Changes in void distribution and volume during compaction of a forest soil. Soil Sci. Soc. Am. J. 50:1001–1006.
- Lindstrom, M.J., and W.B. Voorhees. 1995. Soil properties across a landscape continuum as affected by planting wheel traffic. p. 351–363. In Site-specific management for agricultural systems. ASA, CSSA, and SSSA, Madison, WI.
- Lipiec, J., and R. Hatano. 2003. Quantification of compaction effects on soil physical properties and crop growth. Geoderma 116:107–136.[CrossRef][ISI]
- Mapa, R.B., R.E. Green, and L. Santo. 1986. Temporal variability of soil hydraulic properties with wetting and drying subsequent to tillage. Soil Sci. Soc. Am. J. 50:1133–1138.[Abstract/Free Full Text]
- McBride, M.B. 1989. Surface chemistry of soil minerals. p. 35–88. In J.B. Dixon and S.B. Weed (ed.) Minerals in soil environment. 2nd ed. SSSA Book Ser. 1. SSSA, Madison, WI.
- Morel-Seytoux, H.J., P.D. Meyer, M. Nachabe, J. Touma, M.Th. van Genuchten, and R.J. Lenhard. 1996. Parameter equivalence for the Brooks-Corey and van Genuchten soil characteristics: Preserving the effective capillary drive. Water Resour. Res. 32:1251–1258.[CrossRef]
- Morin, J., and Y. Benyamini. 1977. Rainfall infiltration into bare soils. Water Resour. Res. 13:813–817.
- Moroizumi, T., and H. Horino. 2004. Tillage effects on subsurface drainage. Soil Sci. Soc. Am. J. 68:1138–1144.[Abstract/Free Full Text]
- Mualem, Y. 1986. Hydraulic conductivity of unsaturated soils: Predictions and formulas. p. 799–823. In A. Klute (ed.) Methods of soil analysis. Agron. Monogr. 9. ASA and SSSA, Madison, WI.
- Mualem, Y., and S. Assouline. 1989. Modeling soil seal as a nonuniform layer. Water Resour. Res. 25:2101–2108.[CrossRef]
- Or, D., and T.A. Ghezzehei. 2002. Modeling post-tillage soil structure dynamics: A review. Soil Tillage Res. 64:41–59.[CrossRef]
- Or, D., F.J. Leij, V. Snyder, and T.A. Ghezzehei. 2000. Stochastic model of posttillage soil pore space evolution. Water Resour. Res. 36:1641–1652.[CrossRef]
- O'Sullivan, M.F. 1992. Uniaxial compaction effects on soil physical properties in relation to soil type and cultivation. Soil Tillage Res. 24:257–269.[CrossRef]
- Pachepsky, Ya.A., W.J. Rawls, D. Gimenez, and J.P.C. Watt. 1998. Use of soil penetration resistance and group method of data handling to improve soil water retention estimates. Soil Tillage Res. 49:117–126.
- Perdok, U.D., B. Kroesbergen, and W.B. Hoogmoed. 2002. Possibilities for modeling the effect of compression on mechanical and physical properties of various Dutch soil types. Soil Tillage Res. 65:61–75.[CrossRef]
- Rajkai, K., S. Kabos, M.Th. van Genuchten, and P.E. Jansson. 1996. Estimation of water retention characteristics from bulk density and particle distribution of Swedish soils. Soil Sci. 161:832–845.[CrossRef]
- Rawls, W.J., D.L. Brakensiek, and B. Soni. 1983. Agricultural management effects on soil water processes. Part I: Soil water retention and Green and Ampt infiltration parameters. Trans. ASAE 26:1747–1752.
- Reeve, R.C. 1965. Modulus of rupture. p. 466–471. In C.A. Black (ed.) Methods of soil analysis. Part 1. Agron. Monogr. 9. ASA, Madison, WI.
- Reicovsky, D.C., W.B. Voorhees, and J.K. Radke. 1980. Unsaturated water flow through a simulated wheel track. Soil Sci. Soc. Am. J. 45:3–8.
- Rousseva, S.S., L.R. Ahuja, and G.C. Heathman. 1988. Use of a surface gamma-neutron gauge for in-situ measurements of changes in bulk density of the tilled zone. Soil Tillage Res. 12:235–251.
- Soane, B.D., and C. van Ouwerkerk. 1995. Implications of soil compaction in crop production for the quality of the environment. Soil Tillage Res. 35:5–22.
- Snyder, V.A., J. Rivadeneira, and H.M. Lugo. 2000. Temporal changes in soil structure and hydraulic properties in the plow layer of an oxisol (orthic ferralsol) following tillage. p. 314–324. In R. Horn et al. (ed.) Subsoil compaction. Adv. in Geoecology 32. Catena Verlag, Reiskirchen, Germany.
- Startsev, A.D., and D.H. McNabb. 2001. Skidder traffic effect on water retention, pore-size distribution and van Genuchten parameters of boreal soils. Soil Sci. Soc. Am. J. 65:224–231.[Abstract/Free Full Text]
- Stepniewski, W., J. Gliski, and B.C. Ball. 1994. Effects of compaction on soil aeration properties. p. 167–190. In B.D. Soane and C. van Ouwerkerk (ed.) Soil compaction in crop production. Elsevier, Amsterdam.
- Tackett, J.L., and R.W. Pearson. 1965. Some characteristics of soil crusts formed by simulated rainfall. Soil Sci. 99:407–413.
- Tessier, D. 1991. Microstructure and behavior of clay minerals. p. 387–415. In M.F. De Boodt (ed.) Soil colloids and their associations in aggregates. NATO Advanced Series. Plenum Press, New York.
- van Dijck, S.J.E., and Th.W.J. van Asch. 2002. Compaction of loamy soils due to tractor traffic in vineyards and orchards and its effect on infiltration in southern France. Soil Tillage Res. 63:141–153.[CrossRef]
- van Genuchten, M.Th. 1980. A closed-form equation for predicting the hydraulic conductivity of unsaturated soils. Soil Sci. Soc. Am. J. 44:892–898.[Abstract/Free Full Text]
- Voorhees, W.B., D.A. Farrell, and W.E. Larson. 1975. Soil strength and aeration effects on root elongation. Soil Sci. Soc. Am. Proc. 39:948–953.
- Warkentin, B. p. 1971. Effects of compaction on content and transmission of water in soils. p. 126–153. In K.K. Barnes et al. (ed.) Compaction of agricultural soils. ASAE, St. Joseph, MI.
- Weaich, K., A. Cass, and K.L. Bristow. 1992. Use of a penetration resistance characteristic to predict soil strength development during drying. Soil Tillage Res. 25:149–166.[CrossRef]
- Willis, W.O., and W.A. Raney. 1971. Effects of compaction on content and transmission of heat in soils. p. 165–177. In K.K. Barnes et al. (ed.) Compaction of agricultural soils. ASAE, St. Joseph, MI.
TOP
ABSTRACT
INTRODUCTION
