| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
a USDA-ARS, National Soil Erosion Research Lab., 275 S. Russell St., Purdue Univ., West Lafayette, IN 47907-2077
b current address: Wind Erosion Research Unit, USDA-ARS-NPA, Manhattan, KS, 66502
c Univ. of Regensburg, Dep. of Landscape Ecology and Soil Science, Regensburg, Germany
d Institute of Soil, Water and Environmental Sciences, ARO, The Volcani Center, P.O. Box 6, Bet Dagan 50250, Israel
* Corresponding author (amirakh.mamedov{at}ars.usda.gov).
 |
ABSTRACT |
|---|
 TOPNOTES ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
Abbreviations: DW, deionized water • GS, gypsum solution • MC, moisture content • MW, molecular weight • PAM, polyacrylamide • SR, stability ratio • VDP, volume of drainable pores
|
INTRODUCTION |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
The effects of PAM as a soil conditioner could be influenced by the concentration and composition of the soil solution. The presence of salts in the soil solution enhanced the favorable effect of PAM on infiltration (Shainberg et al., 1990) and hydraulic conductivity (El-Morsy et al., 1991). In the presence of electrolytes, the negative charge and the thickness of the diffuse double layer at the clay and polymer surfaces was suppressed, resulting in decreased repulsion forces and greater adsorption of soil particles to an anionic polymer (Letey, 1994; Shainberg and Levy, 1994). Thus, treatment of a soil with PAM together with a source of electrolytes, such as gypsum, was found to be very effective in controlling seal formation and runoff, because it slows both the physical disintegration of surface aggregates and the chemical dispersion of the soil clays (Levin et al., 1991; Shainberg et al., 1990; Tang et al., 2006).
When the soil is leached with a PAM solution, the presence of PAM increases the viscosity of the solution and leads to a reduction in soil hydraulic conductivity (Malik and Letey, 1992). In the presence of electrolytes, the adverse effects of PAM are alleviated, especially when the electrolytes contain Ca salts (Ajwa and Trout, 2006). It has been further suggested that in the presence of electrolytes, the PAM molecules coil and form short chains that are less effective in clogging the soil pores and reducing its permeability (Laird, 1997; Lentz, 2003; Yu et al., 2003; Ajwa and Trout, 2006). On the other hand, coiling of the PAM molecules in the presence of electrolytes in the soil solution may have an adverse effect on the soil stabilization efficiency of the PAM, since the "grappling distance" of the coiled molecules is substantially shorter compared with conditions where the molecules assume a stretched configuration.
Some uncertainty exists regarding the issue of whether PAM penetrates into aggregates or whether it adsorbs on the aggregates' exterior surfaces, and thus stabilizes only these exterior surfaces. Malik and Letey (1991) studied adsorption of high-MW anionic PAM (107 Da) by soil, soil clay, and washed quartz. Adsorption of PAM on soils and washed quartz was similar and three orders of magnitude lower than that on soil clay. They concluded that PAM adsorption by soils is mostly on exterior surfaces, and that PAM did not penetrate into aggregates. Conversely, the studies of Shaviv et al. (1987) for low-MW PAM and those of Miller et al. (1998) and Levy and Miller (1999) for high-MW PAM have shown that the PAM penetrated into pores within aggregates. Lu and Wu (2003) suggested that the depth of PAM penetration depends on PAM properties, the method of application, and the properties of the soil and water used. Levy and Miller (1999) emphasized the aspect of scale: when high-MW PAM is used, pores in small-size aggregates (<1 mm) are narrow and may not allow the large molecules of PAM to penetrate into the aggregates, while the opposite is true for large aggregates having more macroporosity and interaggregate porosity.
The effects of PAM MW on its performance as a soil amendment also deserve some attention. The MW of commercially available PAM ranges from a few thousand daltons to 20 x 106 Da (Barvenik, 1994). A PAM with high MW has longer molecules than a low-MW PAM, and is therefore considered a more efficient flocculent of colloidal systems (Theng, 1982). This efficiency of the high-MW PAM is related to the large "grappling distance" of its long molecules, which can facilitate the formation of interparticle bridging (Theng, 1982). Similarly, Heller and Keren (2002), who studied the rheological behavior of a Na-montmorillonite suspension, reported that the higher the MW of PAM, the more effective its ability to stabilize flocs of clay in a free electrolyte clay suspension. It can be envisaged, however, that PAM of low to moderate MW will be more effective than PAM of high MW in penetrating into aggregates and contributing to their stabilization.
Contrary to the studies on clay materials, investigations of the response of soils to application of PAM with different MWs has yielded inconsistent results. Levy and Agassi (1995) studied the effects of 2 x 105 and 2 x 107 Da anionic PAMs on infiltration rate and erosion in three soils of different textures. In coarse- and medium-textured soils, high-MW PAM was more effective than low-MW PAM in maintaining a high infiltration rate. In the fine-textured soil, the effect of both polymers on infiltration was comparable. Furthermore, both polymers had similar effects on reducing erosion in the three soils (Levy and Agassi, 1995). Green et al. (2000) also noted a considerable interaction between soil type and the MW of PAM regarding its effect on the soil infiltration rate. They showed that MW played a significant role in determining the effectiveness of PAM in the stabilization of coarse-textured soils, but not of fine-textured ones (Green et al., 2000). In a later study, Green et al. (2004) observed that PAM formulas with MWs in the range of 6 to 18 x 106 Da had comparably favorable effects on enhancing the aggregate stability of soils with different textures. Conversely, for the aggregate slaking test, the effects of PAM were soil dependent, with only the low-MW PAM being effective in reducing aggregate slaking in an unstable silt loam soil (Green et al., 2004).
We hypothesized that (i) the effects of PAM on soil structural stability may depend on the ability of PAM to penetrate into aggregates and thus stabilize both exterior and interior aggregate surfaces or interaggregate porosity (i.e., macroporosity), and (ii) penetration of PAM into aggregates depends on PAM MW and solution ionic strength. Our objective was, therefore, to examine the impact of two PAM polymers having different MWs on the stability of two size classes of aggregates from four smectitic soils varying in clay content at solutions of different ionic strengths.
|
MATERIALS AND METHODS |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
|
|
|
0.9 dS m–1, sodium adsorption ratio of 2 (mmolc L–1)0.5, and pH of 6.7] under constant stirring and slow addition of polymer granules over 4 h. Polymer solutions were prepared with tap water rather than deionized water to improve the dissolution of the polymers and minimize the impact of the dissolved polymer on the viscosity of the solution (Barvenik, 1994). We expected that the use of a less viscous solution would reduce the undesired effect of solution viscosity on the degree of PAM penetration into intraaggregate porosity.
Aggregate Preparation
The soil samples taken from the field were air dried, crushed, and sieved to 1 to 2 and 0.5 to 1 mm. We developed a procedure that ensured that (i) aggregates would not slake during wetting with the polymer solution, (ii) each individual aggregate would come in contact with the polymer solution, and (iii) the polymer in the solution would have the opportunity to penetrate into the aggregates. The following procedure was used. Plastic boxes (25 by 30 cm) were filled with a very coarse sand to form a 15-mm-thick layer that was then covered with a high-porosity (40–60-µm) filter paper. Aggregates from a given soil and size were gently spread on the filter paper to form a monolayer of aggregates. The aggregates were saturated at a rate of 2 mm h–1 from below with tap water or PAM solution, using a peristaltic pump, and were then kept in their respective solution for 24 h to reach equilibrium. The boxes were covered with plastic lids to eliminate possible evaporation. After 24 h, the solutions from the boxes were drained and the aggregates were left to dry in an oven at 40°C for 24 h. The small aggregates (0.5–1 mm) were sieved after drying to eliminate the broken ones. The bigger aggregates (1–2 mm) were sieved and crushed to a size of 0.5 to 1 mm. Finally, the polymer concentration in the solution before and after saturating the aggregates was checked. The analysis by a total C analyzer showed no considerable decrease in the polymer concentration, indicating no shortage of polymer for adsorption by the aggregates.
Aggregate Stability Determination
Theory
Soil structural stability was characterized by determining aggregate stability with the modified version of Levy and Mamedov (2002) to the High Energy Moisture Characteristics (HEMC) method. The HEMC method has been found to be a sensitive and useful method for determining the aggregate stability of arid- and humid-zone soils having a wide range of stability levels (Pierson and Mulla, 1989; Crescimano and Provenzano, 1999; Levy and Miller, 1997; Levy and Mamedov, 2002; Levy et al., 2003; Norton et al., 2006).
In the HEMC method, aggregates are wetted either slowly or rapidly in a controlled manner, and a moisture content (MC) curve at high energies (i.e., energies up to 500 mm H2O tension) is performed. An index of aggregate stability is obtained by quantifying differences in MC curves for fast and slow wetting (Fig. 1a
). For a given wetting rate, a structural index is defined as the ratio of the volume of drainable pores (VDP) to modal suction (Collis-George and Figueroa, 1984). Modal suction corresponds to matric potential (
, J kg–1) at the peak of the specific water capacity curve (d
/d), where is the water content (kg kg–1) (Fig. 1b). The VDP is defined as the integral of the area under the specific water capacity curve and above its baseline (Fig. 1b) and thus is expressed in units of water content (kg kg–1). The ratio of the structural index obtained from fast wetting to the structural index obtained from slow wetting is termed the stability ratio (SR). In general, the SR is used to compare the stability of aggregates on a relative scale of zero to one (0 < SR < 1). Unity SR indicates stable aggregates that resisted slaking by fast wetting. A value of zero SR indicates that fast wetting destroyed the aggregates to the extent that all pores that drain at the matric range applied no longer exist. The use of other indices, however, to describe the stability of aggregates when using the HEMC method is also possible (Crescimano and Provenzano, 1999).
|
A MC curve, at a matric potential range of 0 to –5.0 J kg–1, was obtained using a hanging water column, whereby the height of the meniscus in the pipette was decreased in increments of 0.1 to 0.2 J kg–1, thereby increasing the suction applied. The volume of water that drained from the aggregates at each matric potential was recorded after a 2-min equilibrium period, and the corresponding water content of the aggregates was calculated. Preliminary studies showed that, under our experimental conditions, no additional change in the volume of drainage was noted at equilibrium times >2 min. Each treatment was duplicated. The coefficient of variation between replicates of water content (, kg kg–1) was <6%.
Data Analysis
To accurately calculate the VDP and modal suction, modeling of MC curves was performed with the following seven-parameter modified van Genuchten model (Pierson and Mulla, 1989):
 | [1] |
s and r are pseudo saturated and residual gravimetric water contents, respectively; 
and n control the location and steepness, respectively, of the S-shaped inflection of the MC curve; and A, B, and C are the quadratic terms added by Pierson and Mulla (1989) to improve fitting of the model to the MC curve. The specific water capacity curve (d/d), needed for obtaining the value of modal suction, was computed by differentiating Eq. [1] with respect to matric potential, and had the explicit form
 | [2] |
The VDP was calculated by subtracting the terms for pore shrinkage (2A + B) from Eq. [2] and analytically integrating the remainder of that equation.
We deviated from the usual procedure used to determine the SR (e.g., Pierson and Mulla, 1989; Levy and Miller, 1997; Levy and Mamedov, 2002), and calculated the SR as the ratio of the structural index for any given treatment obtained by fast wetting with DW or GS to the structural index obtained with ethanol (slow wetting) for the same soil. Other researchers have also used wetting by ethanol in their aggregate stability studies for the purpose of avoiding the breakdown of the aggregates during the wetting process (e.g., Le Bissonnais, 1996; Lado et al., 2004; Annabi et al., 2007). Furthermore, we observed no notable differences between fast and slow wetting on the MC curves of the samples for wetting with ethanol. Thus, use of the structural index of ethanol as a reference allowed us to compare the effects of treatments on slaking (or pore size distribution) and aggregate stability of all soils in the range of 0 to 1.
Statistical Analysis
The aggregate stability study tested four main treatments, four soil types, two levels of solution ionic strength, two aggregate sizes, and three PAM levels [control, PAM(H), and PAM(M)]. A full factorial design was used for this study of 48 treatments (4 x 2 x 2 x 3), each in two replicates. A multifactor ANOVA procedure (SAS Institute, 1995) was used to compare the effects of the four main treatments and their interactions on aggregate stability. When significant interactions (P < 0.05) were found, treatment mean comparisons were made using the Tukey–Kramer honestly significant difference test at a significance level of 0.05 (SAS Institute, 1995).
|
RESULTS AND DISCUSSION |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
/d] curves) for the loam and clay-Y soils wetted with DW or ethanol (fast and slow) are presented in Fig. 1; for ethanol, only one curve is presented as no differences in the curves for slow and fast wetting were observed. When DW was used, the differences between the fast- and slow-wetted curves was ascribed to aggregate slaking that itself was attributed to entrapped air, hydration of the exchangeable cations and clay surfaces of the soil particles, and differential swelling (Le Bissonnais, 1996). The hydration of exchangeable cations and clay surfaces, measured in kilojoules per gram of soil (Keren and Shainberg, 1975), weakens the bonds between soil colloids. In the case of ethanol, because of its relatively low dielectric constant (
= 25.3) compared with water ( = 80.1), no hydration was expected. The similarity in the fast and slow wetting curves for ethanol (Fig. 1) indicated that, in the absence of hydration, also no slaking took place. We postulate that, under conditions of no hydration, the soil aggregates remain strong. Thus, the aggregates can overcome the pressure built by entrapped air during wetting, and air can escape easily from the soil pores during wetting as the size of the pores is no longer reduced by the hydrated exchangeable cations and clay surface. Although soil texture differed among the soils tested, they all exhibited similar water retention curves when wetted with ethanol (see example in Fig. 1 for data from the loam and clay-Y soils). For all the soils, the mean VDP was 0.412 ± 0.011 kg kg–1 and the mean modal suction was 0.7 ± 0.015 J kg–1. We used low levels of matric potential (0–50 cm) to drain the aggregates. Under these conditions, it was mostly the interaggregate porosity that was being drained. Ethanol, being a nonpolar liquid, did not hydrate the soil and therefore filled mainly the interaggregate porosity. Consequently, when no hydration was involved, the fact that the aggregates from the different soils all had the same size (0.5–1.0 mm), and thus similar interaggregate porosity, led to comparable characteristics of the water retention curves for the soils studied.
The water retention curves of some selected treatments for the loam and clay-Y soils are presented in Fig. 2 . The use of PAM, GS, or both had considerable effects on the shape of the water retention curves. These observations suggested that the treatments studied impacted the sensitivity of the aggregates to slaking due to entrapped air or osmotic stress (low electrolyte concentration). To quantify the effects of the treatments, we opted to use the data obtained from the wetting with ethanol as a reference, and therefore calculated the SR of the aggregates as the ratio of the stability index for a given treatment to the stability index obtained from the wetting by ethanol.
|
|
Aggregate slaking generally results in the formation of a larger number of particles of smaller size than the original aggregates. This, in turn, causes the interparticle pore size distribution to shift toward a larger number of smaller pores and thus to a decrease in the VDP. Certain conditions enhance aggregate resistance to slaking, such as the presence of cementing or stabilizing agents, e.g., clay (Kemper and Koch, 1966) or PAM (Yu et al., 2003), and the suppression of differential swelling via the addition of electrolytes to the soil solution (e.g., GS treatment). Under these conditions, major changes in the original pore size distribution among the tested aggregates occurred, and hence high VDP levels were maintained.
Stability Ratio
The SR data (Fig. 4
) represent aggregate stability in relative terms, i.e., the sensitivity to slaking of aggregates from a given treatment relative to their sensitivity to slaking when wetted by ethanol. Similar to the results for the VDP data, results of a multifactor ANOVA test for the SR data showed that soil treatment accounted for >70% of the variance (Table 2). Thus, we repeated the ANOVA test for each soil separately and noted that, generally, the main treatments and their interactions had significant effects on the SR data (Table 4
).
|
|
Detailed analysis of the combined impact of PAM MW, aggregate size, and water quality led to the division of soils into two groups with respect to their response to these treatments. The first group included the sandy clay and the two clay soils and the other included the loam.
Sandy Clay and Clay Soils
For both types of PAMs, the SR of the initially small aggregates was, in general, greater than that of the initially large aggregates (Fig. 4). The lesser SR of the initially large aggregates was explained as follows. Some of the interior surfaces of the large aggregates have become external surfaces following the process of sieving and gentle crushing to a size of 0.5 to 1 mm. We suggest that, during the stage where the aggregates were saturated by a PAM solution, part of the PAM molecules penetrated to a certain depth into the pores within the aggregates (Lu and Wu, 2003), thus stabilizing interior pore surfaces of the aggregates only to a certain degree. Consequently, after sieving and crushing, the exterior surfaces of the aggregates contained a mixture of surfaces. One part consisted of surfaces that were exposed to PAM and were thus stabilized; the other part consisted of formerly interior surfaces that mainly were not stabilized by the PAM and thus exhibited lower resistance to slaking. Therefore, the overall stability of the initially large aggregates was less than that of the initially small aggregates, where all the external surfaces were stabilized by the PAM.
Based on the difference in the SR between the two groups, we conclude that (i) to enhance aggregates' resistance to slaking, it is enough to stabilize the exterior surfaces of the aggregates with PAM, and (ii) most of the PAM added to the aggregates was adsorbed on the exterior surfaces of the aggregates and only a small fraction of the PAM added, if any, entered into pores within the aggregates. The latter conclusion was consistent with the findings of Malik and Letey (1991), who also studied small aggregates. Evidently, pores in small-size aggregates are narrow and may not allow significant penetration of the large molecules of PAM into the aggregates; conversely, it has been suggested that, in large aggregates, the large molecules of PAM may penetrate into the aggregates (Levy and Miller, 1999).
In the PAM(H)-treated aggregates, water quality had no effect on the SR of the initially small aggregates; conversely, the initially large aggregates' SR was greater when wetted with GS than with DW (Fig. 4). The inconsistent impact of water quality in the two groups of aggregates was related to the aforementioned limited ability of the PAM molecules to penetrate into the soil aggregates and stabilize interior surfaces. In the case of the initially small aggregates, treating the exterior surfaces of the aggregates with PAM was sufficient to maintain the stability of the aggregates and the use of a saline solution had no advantage over the use of DW with respect to aggregate stability. In the initially large aggregates, however, where PAM-treated and fresh untreated surfaces were exposed to the wetting solution, the presence of electrolytes in the GS that suppress differential swelling contributed to the resistance of the aggregates to slaking and thus yielded greater SR values in comparison to aggregates wetted by DW.
In the PAM(M)-treated aggregates, the effect of water quality depended on the initial size of the aggregates. In the initially large aggregates, as noted for the PAM(H) treatment and for similar reasons, the use of GS resulted in a greater SR than the use of DW. For the initially small aggregates, the opposite was noted (Fig. 4), which was ascribed to the difference in the configuration of the PAM molecules in the two water qualities (DW and GS). In the presence of electrolytes (GS), the PAM molecules apparently tend to coil (Laird, 1997; Ajwa and Trout, 2006), thus the "grappling distance" of the coiled molecules is shorter than that of uncoiled molecules. Hence, although the use of GS reduced differential swelling, it did not enable the PAM(M), because of the coiling of its already relatively short molecules, to exhibit its full potential as a conditioner. Consequently, the SR of the aggregates in DW, where the PAM molecules could assume a stretched configuration, was greater.
The two PAMs studied had, in general, comparable SRs (for a given initial aggregate size) when GS was used (Fig. 4). As mentioned above, in the presence of GS the PAM molecules tend to coil. We postulate that when coiled, the differences in the length of the molecules arising from the differences in the MW between the two PAM polymers was reduced to such a degree that both PAM polymers had a similar impact on the stability of similar-size aggregates. When DW was used, the effect of PAM MW depended on the initial aggregate size. For the initially small aggregates, PAM(M) had a more favorable effect on the SR than PAM(H) (Fig. 4). This observation was in agreement with the findings of Green et al. (2004) for a silt loam. Furthermore, we suggest that, in comparison to the PAM(H), probably more of the PAM(M) was able to penetrate into the aggregates, and thus improve their overall stability. In the initially large aggregates, the effect of PAM MW differed among the three soils (Fig. 4). In general, however, the PAM(H) was more effective in the soil with relatively low clay content (sandy clay), while the PAM(M) was more effective in stabilizing the soil with the highest clay content (Clay-E). Green et al. (2000) also noted that different soils responded in dissimilar ways to the MW and charge density of PAM and therefore no specific PAM could have been singled out as preferable among the PAMs tested.
Loam
Unlike the sandy clay and the two clay soils, where PAM MW had a significant effect on the SR, in the loam, the two PAM treatments studied had, in general, comparable SR values for a given aggregate size (Fig. 4). This observation further illustrates the intricate relation between PAM MW and soil type with respect to the efficiency of PAM as a stabilizing agent.
The effects of water quality on the SR of the loam aggregates were opposite to those noted in the other three soils. In the initially small aggregates, the use of DW in the loam resulted in a significantly greater SR than the use of GS, irrespective of PAM MW; in the other three soils, this observation was noted only with the PAM(M) (Fig. 4). Evidently the importance of the PAM molecules assuming a stretched configuration (as explained above) was essential in stabilizing the aggregates of the loam soil, not only in the PAM with medium MW, but also in the PAM with the high MW. In the initially large aggregates, water quality had no effect on the SR of the loam, whereas in the other three soils the GS yielded greater SR values than DW (Fig. 4). We postulate that the contribution of differential swelling to aggregate slaking in PAM-treated aggregates was more important in the fine-textured soils than in the loam. Therefore, in the loam, the use of the GS had no advantage over the use of DW in determining aggregate stability.
|
CONCLUSIONS |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
Our results support the use of PAM as an amendment for stabilizing soil aggregates, which, in turn, could assist in the control of seal formation, runoff, and erosion. The choice of PAM to be used as a soil amendment, however, needs to be made in accordance with the soil to be treated and the specific conditions prevailing in the field. Further studies are needed to evaluate the role of PAM MW in stabilizing soil aggregates under prolonged aging duration and cycles of wetting and drying.
|
NOTES |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
Received for publication March 8, 2007.
|
REFERENCES |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  A. K. Bhardwaj, R. A. McLaughlin, I. Shainberg, and G. J. Levy Hydraulic Characteristics of Depositional Seals as Affected by Exchangeable Cations, Clay Mineralogy, and Polyacrylamide Soil Sci. Soc. Am. J., April 21, 2009; 73(3): 910 - 918. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| The SCI Journals | Agronomy Journal | Crop Science | |||
| Journal of Natural Resources and Life Sciences Education |
Vadose Zone Journal | ||||
| Journal of Plant Registrations | Journal of Environmental Quality |
The Plant Genome | |||
, polyacrylamide [PAM] type
, and aggregate size) effects on the stability ratio for each soil.