| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
a Universidad Nacional del Litoral, Facultad de Ciencias Agrarias, 86– Kreder 2805– S 3080 HOF, Esperanza (Sta Fe), Argentina
b Universidade de São Paulo, Escola Superior de Agricultura Luiz de Queiroz, Depto. De Solos e Nutricão de Plantas, Av. Padua Dias 11, Caixa Postal 09, CEP: 13418-900, Piracicaba (SP), Brazil
c Dep. of Land Resource Science, Univ. of Guelph, Guelph, ON N1G 2W1, Canada
* Corresponding author (apisilva{at}esalq.usp.br).
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONSUMMARY AND CONCLUSIONSREFERENCES |
|---|
Abbreviations: PTF, pedotransfer function • VIF, variance inflation factor
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONSUMMARY AND CONCLUSIONSREFERENCES |
|---|
The soil resistance to compaction depends on the intrinsic soil attributes, with texture being one of the most relevant (Larson et al., 1980; Horn, 1988; McBride, 1989). Several researches have demonstrated that clay content and mineralogy affect the process of soil compression (Larson et al., 1980; Horn, 1988, McBride and Watson, 1990; Smith et al., 1997). Coarse-textured soils are less susceptible to compaction than those with a fine texture (Horn, 1988; McNabb and Boersma, 1993; Horn and Lebert, 1994; McBride and Joosse, 1996).
Soil response to compaction is also influenced by organic matter content (Larson et al., 1980; McBride, 1989; McBride and Watson, 1990; Soane, 1990). The susceptibility to compaction decreases as soil organic C content increases (O'Sullivan, 1992; Zhang et al., 1997). However, the effect of organic matter on the reduction of soil compressibility seems to be dependent on soil moisture at the time of load application (Soane, 1990).
Soil moisture has been widely recognized as a determinant of soil compressibility (Larson and Gupta, 1980; Soane, 1986; McBride, 1989; Soane, 1990; O'Sullivan, 1992; McNabb and Boersma, 1996; Sánchez-Girón et al., 1998). However, uncertainty exists regarding to the effect of soil moisture on the susceptibility to compaction. Larson et al. (1980) and O'Sullivan (1992) indicated that susceptibility to compaction does not depend on the soil water content, which is in direct contrast to the findings of Sánchez-Girón et al. (1998), Silva et al. (2000), and Sánchez-Girón et al. (2001). The differences in the soil susceptibility to compaction seems to be related to the mechanism whereby a decrease in soil water content increases the number of contacts between particles, which is directly dependent on soil texture (McNabb and Boersma, 1996; Harte, 2000). Crystalline clay minerals may lose more water from the external surface than noncrystalline minerals. As a result, the number of contacts between particles increases at a faster rate in the former than in the later as the soils dried (McNabb and Boersma, 1996).
The soil compression behavior is also affected by soil bulk density. Soil deformation occurs when some individualized (crystals) or grouped (domains) particles are able to separate and move in relation to each other. This movement is restricted by friction forces and by the bonds existing between particles. The denser the soil and the more intricate the particle arrangement, the smaller the pore space available for particle movement is and the higher the friction forces between them are. Thus, displacement and rearrangement of solid particles to closer positions (deformation) becomes more difficult as bulk density increase (Paz and Guérif, 2000).
Soils classified as Hapludox may have a wide range of texture and organic matter content leading to variations in other soil properties such as bulk density and water retention capacity, which in turn influence the soils compressibility (Hakansson and Voorhees, 1998).
Understanding the compressive behavior of soils is essential to predict the alterations that might occur in soil structure when submitted to stress caused by agricultural implements (McBride, 1989). Soil compression curve has been used to understand the process of compression (Larson and Gupta, 1980; Larson et al., 1980; Bailey et al., 1986; O'Sullivan et al., 1992; McNabb and Boersma, 1993). The curve, which describes the relationship between the logarithm of the load applied and void ratio (or bulk density), generally consists of two sections: a linear part, named virgin compression line whose slope is called compression index, Cc, (Terzaghi and Peck, 1967), and a nonlinear portion which results from the recompression of a previously decompacted soil. The intersect of the two sections of the curve is known as preconsolidation pressure, 
p (Casagrande, 1936). The preconsolidation pressure has been widely accepted as an indicator of the history of stress to which the soil was submitted in the past and of its load support capacity (Veenhof and McBride, 1996), while the compression index is used as an indicator of soil susceptibility to compaction (Larson et al., 1980).
Soil compressibility, defined as the resistance against volume decrease when soil is subjected to a mechanical load, has been described by the shape of the soil compression curve (Horn and Lebert, 1994). Several nonlinear models that describe the entire compression curve and account for the influence of soil properties, such as differences in initial bulk density or in soil water content, on the soil compressibility has been proposed (Bailey et al., 1986; McNabb and Boersma, 1993; McNabb and Boersma, 1996; Assouline et al., 1997; Assouline, 2002).
Pedotransfer functions, which quantitatively describe the relationship between soil attributes that are generally difficult to measure, such as indices obtained from the compression curve, and others more easily measurable such as texture, have been developed for temperate regions to establish indicators of soil quality (Larson et al., 1980; McBride, 1989; O'Sullivan, 1992; McBride and Joosse, 1996). McBride and Joosse (1996) reported the usefulness of PTFs for the characterization of the compaction degree of Canadian soils and suggested that they could be used as indicators of the soil physical quality at a regional level. Despite the usefulness of these functions, little information exists for tropical soils (Assouline et al., 1997).
In the present study, we tested the hypothesis of whether the compressive behavior of Hapludox could be estimated using PTFs. The specifics objectives were to: (i) quantify the compression index, the preconsolidation pressure, and the soil compression curve, (ii) determine the effect of water content, organic matter, texture, and bulk density on these indicators, and (iii) develop PTFs that allow to estimate the susceptibility to compaction, load support capacity, and soil bulk density after load removal of Hapludox based on easily measurable soil attributes.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONSUMMARY AND CONCLUSIONSREFERENCES |
|---|
In this area, 25 sampling points spaced about 50 m apart, were established between the sugarcane rows following a transect of about 1400 m. At each point, two undisturbed samples were collected from the surface (0–0.15 m) using cores with an inside diameter of 7 cm and a height of 2.5 cm, providing a total of 50 samples.
The samples were saturated in water for 24 h. The cores were then split into two groups of 25 samples, with each group being submitted to one of two matric potentials; that is –10 and –100 kPa, using pressure chambers according to the method of Klute (1986) to create variations in soil water content.
After reaching equilibrium, the samples were submitted to a uniaxial compression test using the automatized Satron MCT-2000 consolidation system (MIRAE Engineering, Inc., Buscan, Korea). Displacement occurring at each pressure applied was recorded with a sensor connected to a data acquisition system. The samples were subjected to the following pressures: 25, 50, 100, 200, 400, 600, 800, 1000, 1300, and 1600 kPa. The load was applied during a period of 15 min. Tests (Silva et al., 2000) have shown that this time was enough to reach 90% of the maximum soil deformation for Hapludox. Assouline et al. (1997) also mentioned that a few minutes were necessary to reach equilibrium for Oxisols.
After withdrawal of the load, the samples were oven dried at 105°C for 24 h and the dry mass (g) was determined. Bulk density, 
, was determined based on dry mass and the volume calculated for each pressure applied (Blake and Hartge, 1986a). The bulk density of the soil before application of the selected pressures was defined as the initial bulk density, i. Gravimetric water content, w, was determined based on the weight of the soil samples at the beginning of compression tests and their dry mass.
The samples were then ground and sieved through a 2-mm sieve and particle-size distribution was determined by the densimeter method (Gee and Bauder, 1986). The particle density was determined using the pycnometer method (Blake and Hartge, 1986b) and soil organic C content was ascertained through oxidation with potassium dichromate. The void ratio, e, was calculated for each sample based on soil bulk density, , and particle density, p, {e = [(p/) – 1]} (McBride and Joosse, 1996).
Based on these values, a soil compression curve was constructed for each sample. This curve graphically represents the relationship between the logarithm of the applied pressure, , (x axis) and void ratio, e, (y axis).
A program was developed to calculate the compression index, Cc, and preconsolidation pressure, p, using the Mathcad (Mathsoft Inc., 2000) software. In one of the first steps, the program determines the point of maximum curvature of the compression curve by calculating the second derivative along the compression line (Fig. 1
, Point a). The point of maximum curvature is the point at which the second derivative reaches its minimum value. In successive steps, the program calculates the tangent to the maximum curvature point (Fig. 1, Line b), the horizontal line that passes through that point (Fig. 1, Line c), the bisector line of the angle formed between the tangent and the horizontal line at the point of maximum curvature (Fig. 1, Line d), and the virgin compression line (the line calculated using the last three points of the compression curve) (Fig. 1, Line e). Finally, the intersect between the bisector line and the virgin compression line, called preconsolidation pressure, p, is calculated (Fig. 1, Point f). Thus, p is calculated as proposed by Casagrande (1936) but without the subjectivity of the manual method. The preconsolidation pressure is an indicator of the soil's load support capacity. The slope of the virgin compression line, called compression index, was determined as follows: Cc = de/dlog. The Cc represents an indicator of soil susceptibility to compaction.
|
i, on the estimation of the final bulk density, , as follows:
 | [1] |
, natural logarithm of the final bulk density, (Mg m–3); 0, bulk density for the stress value equals 0 MPa (Mg m–3); i, initial bulk density (Mg m–3); mean, mean bulk density (Mg m–3); 
i = i/mean; equals applied pressure (MPa); c = (i – 1) x 0 (Mg m–3); a, b, c, and d are parameters of the model that describe the shape of the curve.
The parameter i standardizes 0 to account for the differences in the initial bulk density values, while c fits the complete compression curve for differences in the initial bulk density values of each soil sample.
The influence of water and organic matter content, texture, and soil bulk density on the Cc, p, and soil compressibility was quantified by linear and nonlinear multivariate regression analysis using the SAS software (SAS Institute, Inc., 1989). The effects of multicollinearity were tested using the variance inflation factor (VIF). The VIF is an indicator of the magnitude at which the variance of the parameters of the model increases when independent variables are closely correlated, compared with the situation when the variables are not correlated. Variance inflation factor values lower than 10 indicate that the multicollinearity effect is not influencing the regression result (Neter et al., 1989). In the present study, only variables with a VIF value lower than 10 were included in the developed models.
|
RESULTS AND DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONSUMMARY AND CONCLUSIONSREFERENCES |
|---|
|
p = 138 kPa), an indicator of the pressure exerted on the soil in the past by agricultural machinery, was similar to that reported by Kanali et al. (1997) (p = 122 kPa) as a critical value for soils under sugarcane cultivation. Agricultural machinery can exert pressures ranging from 70 to 350 kPa, while transport vehicles might exert pressures of up to 600 kPa (Soane, 1986; Kanali et al., 1997; Tijink and Van der Linden, 2000). The variation in preconsolidation pressure from 22 to 305 kPa (Table 1) suggests that the soils studied has not been subject to excessively high loads, in agreement with the fact that in the studied area conventional agricultural machinery is used for soil tillage and sugarcane is harvested manually. Sugarcane transport vehicles (trailers) with high weight per axis and high air tire inflation pressure only travel in designated areas, thus preventing the occurrence of excessive soil compaction.
Effect of Physical Attributes on Soil Susceptibility to Compaction
The Cc increases linearly with clay content up to a value of 30%, remaining relatively constant thereafter (Fig. 2)
. Larson et al. (1980) observed a similar behavior of the Cc. These authors, studying different soil classes (Alfisol, Ultisol, and Oxisol), found a linear and positive correlation between clay content and Cc up to a clay content of 33%, with the Cc remaining constant thereafter. According to these authors, soils containing more than 33% clay basically consist on a clayey matrix with the coarse material (sand grains) being dispersed within this matrix. This is the reason why the Cc remains constant from this value on. Similar results have been reported by Lebert and Horn (1991), Veenhof and McBride (1996), and Smith et al. (1997).
|
 | [2] |
 | [3] |
equals bulk density (Mg m–3).
|
The difference between the Cc at the plateau determined in this study (Cc = 0.24) and the value reported by Larson et al. (1980) (Cc = 0.46) might be attributable to the fact that these authors included different soil orders (different clay mineralogy) in their study and used disturbed samples. Similar Cc values to those obtained in the present study were reported by Kondo and Dias Junior (1999b) and Silva et al. (2000) for Hapludox using undisturbed samples.
Bulk density had a negative influence on the Cc. Therefore, the higher the bulk density the lower the soil deformation and the soil susceptibility to compaction, as suggested by Paz and Guérif (2000). Similar results have been reported by Culley and Larson (1987), Lebert and Horn (1991), Veenhof and McBride (1996), Silva et al. (2000), and Silva et al. (2002a).
According to ours results, Hapludox with a coarse texture (predominance of the sandy fraction) are less susceptible to compaction than those with a predominantly clay fraction in their granulometric composition, as has been also demonstrated for other soils (Horn, 1988; McNabb and Boersma, 1993; Horn and Lebert, 1994; McBride and Joosse, 1996). In addition, for a given clay content, the soil becomes more susceptible to compaction as its bulk density decreases, as proposed by Horn (1988). Thus, soils with a potentially better structural quality associated with a lower bulk density and/or higher clay content will be more susceptible to degradation.
Influence of Physical Attributes on the Soil Load Support Capacity
The effect of soil water content, organic matter, texture, and bulk density on the preconsolidation pressure, p, was determined by multiple regression analysis and the result is shown in Table 3. The model explained 70% of the variability of the data. The residual values showed a normal distribution according to the Shapiro–Wilk test (W = 0.98, P < W = 0.46). The preconsolidation pressure, p, was significantly and positively related with soil bulk density and clay content, and negatively correlated with soil water content. Organic matter content did not have any significant effect on the p.
|
p and soil bulk density has been reported by Lebert and Horn (1991), Alexandrou and Earl (1998), Canarache et al. (2000), Silva et al. (2000), and Silva et al. (2002a). Although Lebert and Horn (1991), Alexandrou and Earl (1998), or Silva et al. (2000) did not incorporate the clay content variable in their model, they suggested that the relationship between preconsolidation pressure and soil bulk density was dependent on soil texture. Lebert and Horn (1991) emphasized that the influence of soil bulk density as an independent variable decreases with increasing clay content.
In general, authors agree on the assumption that the effect of soil bulk density on p is due to an increase in the friction forces between particles, while the influence of clay content is attributable to cohesion forces between particles. These cohesion forces impede the separation and displacement of the soil particles increasing its load support capacity. Table 3 shows that the higher the clay content and soil initial bulk density at a given moisture, the higher the load support capacity of the soil.
Although an increase in clay content has a positive effect on the load support capacity of soils, it should be noted that soils with a higher clay content are more susceptible to compaction (higher Cc, Eq. [2] and [3]) if the value of p is exceeded. Therefore, additional care should be taken not to exceed the p value as the clay content increases.
The p decreased linearly with increasing gravimetric soil water content. The soil water increase induces a decline in the binding forces between particles, and a consequent reduction in the soil load support capacity. This behavior has been observed by several authors (Horn and Lebert, 1994; Veenhof and McBride, 1996; Alexandrou and Earl, 1998; Kondo and Dias Junior, 1999a; Silva et al. 2002b). The positive influence of soil bulk density and the negative influence of the degree of saturation on p have also been reported by Silva et al. (2000) for a dark Hapludox.
Figure 3
illustrates the effect of clay content and bulk density on p for three soil gravimetric water content values, w. The w values adopted were: the w at field capacity (
= –10kPa) (wfc = 0.20 kg kg–1), one standard deviation above wfc (w = 0.26 kg kg–1), and one standard deviation below wfc (w = 0.14 kg kg–1). In all cases, the p decreased with increasing w, in agreement with the results obtained by Horn and Lebert (1994), Veenhof and McBride (1996), Alexandrou and Earl (1998), Kondo and Dias Junior (1999a), and Silva et al. (2002b).
|
p was lower than 120 kPa within almost the whole range of clay content and bulk density. At wfc, p was lower than 120 kPa for soil bulk densities below 1.6 Mg m–3 and clay contents lower than 45%. Thus, soils with a coarse texture that had been decompacted as a result of tillage (lower soil bulk density) easily would suffer additional compaction. Considering that agricultural machinery employed in sugarcane cultivation exerts, on average, pressures of 120 kPa (Kanali et al., 1997; Tijink and Van der Linden, 2000), care should be taken in terms of selecting tillage timing; that is, adequate soil moisture, to prevent excessive compaction. As the soil dries, its load support capacity increases. At w = 0.14 kg kg–1, the soils may resist the pressures exerted by agricultural machineries without suffering further compaction. However, they may not resist the traffic of trucks or trailers that, due to their characteristic of weight per axis and air tire inflation pressure, exert high pressures on the soil. Vehicles with a weight per axis of 60 kN and an air inflation pressure of 180 kPa are widely employed in sugarcane transport (Kanali et al., 1997).
Droogers et al. (1996) reported that an optimum soil water content for tillage and agricultural machinery traffic exists for each soil. This optimum value is defined as the soil water content at which tillage or traffic can be performed without causing soil structure deterioration. In the case of drier soils, tillage can be performed without causing additional compaction, although the required energy is higher.
In the present study, clayey soils showed a higher load support capacity than those with a predominantly sandy fraction for the water content range studied (Fig. 3). Similar results have been reported by Kondo and Dias Junior (1999a) and Silva et al. (2002b) for Hapludox. The results suggest that in toposequences with a wide variation in clay content, tillage should be performed taking into account the optimum soil water content of the soils with lower clay content, which have a lower load support capacity, to prevent irreversible deformation and, consequently, degradation of the structural quality of the soil. It has been mentioned that in soil compaction not only static load pressure but also dynamic stresses are important. Static load pressure (normal stress) is related to vertical wheel load while shear stress (horizontal component) is related to wheel slip. Hence, parameters related to the shear strength of the soil, such as the angle of internal friction and cohesion must be determined to better understand the soil behavior, and to improve the PTFs developed, as mentioned by Lebert and Horn (1991).
Influence of Physical Attributes on Soil Compressibility
Equation 1 was used to fit the data of the compression curves to assess the compressibility of the studied soils. The model parameters obtained by nonlinear regression analysis are shown in Table 4. The model explained 90% of the variability of the bulk density data, with all parameters being significant (P < 0.05) (Glantz and Slinker, 1990). The Shapiro–Wilk test showed normal distribution of the residual values (observed value minus estimated value at the pressure applied) (P < W = 0.10), indicating adequate fit of the data (Neter et al., 1989). The approach applied by da Silva and Kay (1997) was used to describe the influence of soil texture, soil organic matter, and soil moisture on the estimated bulk density. The parameters of the model were not related to soil texture, soil organic matter, and soil water content. The initial bulk density seems to be one of the most important soil properties influencing the value of soil bulk density after load removal, as suggested by McNabb and Boersma (1993).
|
) that will be reached when a given pressure be applied under similar soil moisture to those of the present study.
|
Several studies have demonstrated that agricultural machinery and transport equipment can range in weight from 10 to 150 kN and use air tire inflation pressures ranging from 50 to 800 kPa. As a consequence, this equipment can exert pressures on the soil ranging from 50 to 850 kPa (Soane et al., 1981a,b; Koolen et al., 1992; Vermeulen and Perdok, 1994; Kanali et al., 1997). Figure 5
shows the values estimated by the proposed model based on a wide variation in the initial and in the pressure exerted by the agricultural machinery.
|
Studies have demonstrated that crop productivity is reduced when the critical soil bulk density (defined as the at which plant development is limited) is exceeded (da Silva and Kay, 1997; Hakansson and Lipiec, 2000). In this respect, the presented model might be a useful tool since the final bulk density can be estimated based on the determination of the initial bulk density and on the knowledge of the pressures exerted on the soil by the used agricultural machinery. The model permits to predict whether the critical soil bulk density will be exceeded. Additionally, the model can be used as a basis for the decision as to which type of agricultural machinery (load capacity, air tire inflation pressure) should be used in each situation.
|
SUMMARY AND CONCLUSIONS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONSUMMARY AND CONCLUSIONSREFERENCES |
|---|
|
ACKNOWLEDGMENTS |
|---|
Received for publication September 23, 2002.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONSUMMARY AND CONCLUSIONSREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  G. Yoo and M. M. Wander Tillage Effects on Aggregate Turnover and Sequestration of Particulate and Humified Soil Organic Carbon Soil Sci. Soc. Am. J., May 1, 2008; 72(3): 670 - 676. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 G. Yoo, T. M. Nissen, and M. M. Wander Use of physical properties to predict the effects of tillage practices on organic matter dynamics in three illinois soils. J. Environ. Qual., July 1, 2006; 35(4): 1576 - 1583. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
D. D. Fritton Fitting Uniaxial Soil Compression Using Initial Bulk Density, Water Content, and Matric Potential Soil Sci. Soc. Am. J., June 21, 2006; 70(4): 1262 - 1271. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| The SCI Journals | Agronomy Journal | Crop Science | |||
| Vadose Zone Journal | Journal of Plant Registrations | ||||
| Journal of Natural Resources and Life Sciences Education |
Journal of Environmental Quality |
||||
,
