| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
a Usda-Ars, Weslaco, TX 78596 USA
b Dep. of Soil and Atmospheric Sciences, Univ. of Missouri-Columbia, Columbia, MO 65211 USA
bradford{at}pop.tamu.edu
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMaterials and methodsResults and discussionREFERENCES |
|---|
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMaterials and methodsResults and discussionREFERENCES |
|---|
Differences in interparticle forces of attraction in a clay–water–ion system contribute to differences in soil resistance to detachment forces of concentrated flow. Clay mineralogy, type of exchangeable cation, and chemical characteristics of the pore and eroding fluids affect the ability of clay particles to bind soil particles and increase soil strength or resistance to soil detachment. Grissinger (1966) found that increased concentrations of clay minerals generally induced greater soil stability in a small flume and that Na–montmorillonite was 20 times as effective as Ca–montmorillonite. Grissinger concluded that Na–montmorillonite increased the strength of the soil bed by increasing the number of contacts between soil particles. Calcium–montmorillonite forms packets of clay platelets condensed together with low external surface area (Shainberg and Letey, 1984). Similar observations were made by Shaikh et al. (1987) who found that the erodibility of Ca–montmorillonite in a flume was about two orders of magnitude higher than that of Na–montmorillonite and one order of magnitude higher than that of kaolinite.
Zhao et al. (1991), by determining the rheological properties of clay suspensions, studied the effects of pH and electrolyte concentration on particle interaction for three homoionic sodium soil clay suspensions. The extrapolated shear stress or Bingham yield value (
), the stress value at which flow is initiated, is a function of both the number of particle-particle bonds and the energy required to break these bonds. In the Zhao et al. (1991) study, increased as the clay concentration increased from 2.5 to 8%, with the magnitude of increasing as kaolinitic soil clay > smectitic soil clay > illitic soil clay. A reasonable next step in extending the results from soil clay suspensions (Zhao et al., 1991) is the determination of cohesive or detachment properties of artificially prepared clay and sand mixtures under concentrated flow conditions. The results of erosion studies on pure clays may not represent the erosional behavior of more complex cohesive soils; however, determination of factors controlling erosional behavior of soils may begin with the simplest possible conditions.
Shainberg et al. (1994, 1996) determined soil detachment under rill flow using the mini-flume procedures. He isolated the effects of soil sodicity, water quality, soil water content, aging, and temperature on soil detachment. Following Kemper et al. (1987), it is assumed that the cohesive forces between soil particles (and the soil's resistance to detachment by flow) increase with number and strength of particle to particle contacts and decrease with those soil properties that weaken the attraction forces between soil particles. The objective of this study was to evaluate the effects of clay mineralogy and water quality on rill erodibility of clay–sand mixtures as determined in small laboratory mini-flumes. We hypothesized that the cohesive forces between particles decrease and rill erodibility increases as the dominant clay mineral changes from montmorillonite to illite to kaolinite, as total salt in the eroding fluid decreases, and as the ratio of Ca to Na in the system decreases.
|
Materials and methods |
|---|
|
TOPABSTRACTINTRODUCTIONMaterials and methodsResults and discussionREFERENCES |
|---|
Clays used in this experiment were obtained from the following sources: (i) Illite (IMt-1), referred to as Silver Hill illite, was from the Source Clays Repository, Dep. of Geology, Univ. of Missouri, Columbia, MO. The characterization of this clay is presented in the paper by Hower and Mowatt (1966). (ii) The kaolinite sample was supplied by the J.M. Huber Corp., Macon, GA. (iii) The sample referred to as montmorillonite was a Upton Wyoming bentonite (Volclay Premium Gel) in the sodium form from the American Colloidal Company, Aberdeen, MS.
Clay samples were ground to pass a sieve with 0.15-mm openings. Three kilograms of clay were placed in 15-L plastic pails, and 10 L of either 1 mol L-1 NaCl or 1 mol L-1 CaCl2 were added. The pH of the solution was measured and adjusted to pH 7 by adding either HCl or NaOH. Suspensions were stirred and allowed to settle for 24 h. The solution above the clay surface was siphoned and the original volume restored by adding either 1 mol L-1 NaCl or 1 mol L-1 CaCl2. The suspensions were again adjusted to pH 7, stirred, and allowed to settle overnight. This process was repeated three times.
Salts were leached from the clays by a combination of settling–decanting and vacuum filtration. Prior to filtration, Ca-saturated clays were allowed to settle 16 h, the solution siphoned off, and its specific conductance measured. Deionized water was added and the suspension stirred and allowed to settle 16 h. This dilution process was repeated until the suspensions failed to completely clear. The specific conductance of the solution where the suspension failed to clear was between 10 and 20 mS m-1. The suspensions were then transferred to the filtration units. The Na-saturated clays in 1 mol L-1 NaCl were transferred to vacuum filtration units without decanting.
Six filtration units were built by the University of Missouri Science Instrument Shop. They were formed from 2.5 cm thick by 27-cm-diam. plexiglas plates. Circular groves were formed on the surface of the plate with a 3-mm hole connecting them to a 6-mm-diam. hole drilled from the perimeter to the center of the unit. Filter paper (27 cm) was supported by a 26-cm-diam stainless steel screen with 1-mm openings placed over the circular groves. The units were sealed into 15-L plastic pails. The units were connected to a vacuum system that provided filtrate collection bottles for each unit. Whatman #3 filter paper (27 cm diam.) was wetted with deionized water and then placed in the filtration unit. The vacuum was turned on and the clay suspensions added.
Final increments of filtrate were analyzed for Ca by atomic adsorption, Na by flame emission, Cl by ion specific electrode, and total salt by electrical conductivity.
Clay suspensions were filtered one at a time. Each suspension was divided into six equal parts and placed in the filtration units resulting in a loading density of 0.9 g clay per cm2 of filtering surface. In all cases during filtration, the first 100 to 500 mL of filtrate contained suspended clay. This suspension was returned to the filtration unit and filtered again. When the solution above the clay was reduced so that the clay surface was exposed, a 10 part water and 90 part methanol mixture was added and the filtration continued. The process was repeated until the specific conductance of the filtrate was <5 mS m-1.
With this loading density, Ca-saturated illite and kaolinite filtrations were complete in 3 to 5 h. The Ca-saturated montmorillonite and Na-saturated illite and kaolinite suspensions were filtered within 24 h. Sodium-saturated montmorillonite, however, could not be filtered at a loading density of 0.9 g cm-2. The loading density was reduced to 0.2 g cm-2 and the suspension filtered for 72 h. As the salt concentration approached 0.005 mole L-1 , the Na-saturated montmorillonite formed a gel containing about 6 parts of water to 1 part of clay and the filtrate contained suspended material. These filtrates were filtered again on a new set of filter papers. The filtrate was repeatedly returned to the surface until most of the suspended clay was retained on the surface. This part of the process was complete within 48 h.
Clays were removed from the filter paper, placed in aluminum cake pans, and dried in a forced air oven at 90°C for 48 h. Dried clay was broken into small pieces with a mortar and pestle, placed in a ball mill, and ground for 24 h. Illite clays grounded to a fine powder in the ball mill. Kaolinite clays did not grind well in the ball mill but formed a compact mass on the perimeter as the balls rolled in the middle. Kaolinite, however, was easily reduced to a powder with a mortar and pestle. Montmorillonitic clays were extremely hard to grind and were first crushed with a power mortar and pestle and then ground in the ball mill. After grinding, the materials were sieved to pass through 0.15-mm openings.
Cation exchange capacity and exchangeable bases were analyzed by the ammonium acetate method at pH 7 (Soil Survey Staff, 1975). X-ray diffraction analysis of the clays was done as described by Whittig and Allardice (1986). A Syntax-Padeo V diffractometer (Sintax Incorporated, Sunny Valley, CA) with a Cu target set at 40 kV, 30 mA was used.
Clay–sand mixtures varying in composition from 2 to 20% (w/w) were prepared. Portions of oven-dried clay, which had equilibrated with air, were added to 2000 g of Chelsea fine sand in a 7-L pail and thoroughly mixed with a spatula. Samples were wetted to an estimated pressure of -5-cm water head by adding distilled water for the 0.001 mol L-1, 0.009 mol L-1 salt for the 0.01 mol L-1, and 0.099 mol L-1 salt for the 0.1 mol L-1 eroding salt treatments (salts present in the clays resulted in a solution salt concentration of about 0.001 mol L-1). Solution was added and the mixture stirred with a spatula until it glistened and released some water when strongly pressed with the spatula. A 10-cm high wall was formed along the sides of the pail and the appropriate amount of water judged as the point where the lower layer barely glistened with free water. The estimated pressure was verified by placing about 75 g of this mixture on a filter paper on top of a sand tension plate at -5 cm (-490 Pa) water pressure, equilibrating overnight, and comparing the water percentage to that of the mixture in the pail.
The sample was stored for 7 d and allowed to consolidate. The soil water content was adjusted so that the phreatic surface was at the soil surface. From this condition, water content was measured gravimetrically for each clay–sand mixture. The mixture was then packed into the mini-flume and bulk density calculated from the water content, mass, and volume.
The mini-flume was 50 cm long, 4.6 cm wide, and 12 cm deep. A 90° V-shaped channel, 20 cm long, was attached to each end of the central flume area. The inlet channel was necessary to minimize flow turbulence to the central test area. Sediment was collected at the outlet channel at eight water flow rates. All tests were conducted at a flume-bed slope of 0.20 m m-1. Flow rates from 0.01 to 1.0 L min-1, depending on the erodibility of the soil material, were applied and controlled with a peristaltic pump. Packing of soil into the flume was accomplished by placing a V-shaped, 90°, plexiglas plunger over the upper flume surface, inverting the flume so that top rested on a table, and filling the flume from the bottom with the clay–sand mixture. After filling, a plexiglas base was attached with tape and the flume inverted again to an upright position. The V-shaped plunger was removed from the top by sliding across the surface, and the mixture was allowed to sit for 1 h before beginning the test.
The reason for not packing air-dried material in the flume and then slowly wetting, or testing in an air-dried condition as done by Shainberg et al. (1994), was that preliminary study showed rapid consolidation or rapid swelling of some clay–sand mixtures upon wetting. If the tests were conducted on dry materials, the montmorillonite clay–sand mixtures wet so slowly that dry materials were eroded during the mini-flume test. If the air-dried materials were placed in the flumes and then slowly wetted, the montmorillonite clay–sand mixtures swelled nonuniformly and cracks appeared along the soil surface, or swelled to an extent that the soil surface protruded greatly above the flume intake and outlet surfaces. Prewetting of kaolinite clay–sand mixtures resulted in nonuniform consolidation of the soil surface.
Mini-flume tests were conducted at three salt concentrations. Pore water and eroding fluid solutions were 0.001, 0.01, and 0.1 mol L-1 NaCl or CaCl2. The measurement began as the eroding solution at a rate of 0.01 L min-1 was applied with a peristaltic pump, and both discharge rate (volume of flow) and sediment concentration in runoff were determined. Following the collection of three runoff volumes and sediment concentrations, the flow rate was increased and three more samples were taken. Discharge was collected for a total of eight flow rates. Rate of increase of applied flow depended upon the erodibility of the clay–sand mixture. Each test was replicated at least two times.
Soil detachment rate (Dc) was plotted as a linear function of shear stress (
) assuming
 | (1) |
c is the critical shear stress (Pa), and is the shear stress of flowing water (Pa), given by
 | (2) |
is the specific weight of water (N m-3), S is the rill slope, and rh is hydraulic radius (defined as the ratio of flow cross-sectional area to the wetted perimeter of flow). |
Results and discussion |
|---|
|
TOPABSTRACTINTRODUCTIONMaterials and methodsResults and discussionREFERENCES |
|---|
|
Weaver and Pollard (1973) have collected data on the composition of various clay minerals. The ideal kaolinite is: Al4(Si4O10)(OH)8, 46.54% SiO2, 39.5% Al2O3, 13.96% HOH. In nature, this composition is seldom, if ever, found, and impurities of Fe, Ti, Mg, Ca, K, and Na are usually present. Our sample has a CEC of 5.05 cmolc kg-1 in the Ca–kaolinite form (Table 2) , which is at the small end of the range of CEC, 4 to 8 cmolc kg-1, reported by Weaver and Pollard (1973). Some researchers believe that the CEC of kaolinite may be due to impurities, and pure kaolinite has CEC values in the 1 to 4 cmolc kg-1 range. X-ray analysis of our kaolinite sample indicates sharp peaks at 0.702 and 0.356 nm which disappear when the sample is heated to 550°C (Fig. 1) , which is the criteria for the identification of kaolinite (Whittig and Allardice, 1986). No other crystalline materials were detected on the x-ray diffractogram.
|
|
 | (3) |
They report this as one charge per 67 x 10-8 µm2 and indicate that illite and montmorillonite have values from (75 to 100) x 10-8 µm2.
The CEC of Na–illite was 11.59 cmolc kg-1 and Ca–illite was 18.24 cmolc kg-1. Silver Hill illite CEC values of 12 cmolc kg-1 (Na/Mg exchange) and 15 cmolc kg-1 (mixed cations displaced by Sr), were reported by Thellier and Sposito (1988). Cation exchange capacity values greater than 15 cmolc kg-1 in illite samples have been attributed to the presence of expanding layers. Ormsby and Sand (1954) found that illite with all layers contracted would have a CEC of 15 cmolc kg-1.
Our x-ray analysis showed peaks at 0.981 and 0.490 nm (illite) and small peaks at 0.443 nm (CaSO46H2O) and 0.419 and 0.330 nm (quartz) (Fig. 1). Thellier and Sposito (1988) presented x-ray evidence to indicate that the Silver Hill illite was a sample with characteristics of ideal muscovite. They found traces of expanding layer clays and kaolinite in the sample. Our sample does not contain expanding layers or kaolinite and appears to be a representative illite with muscovite characteristics, but with quartz impurities as well as trace impurities of CaSO46H2O.
Using the CEC of Ca–illite and the surface charge density of 1 charge per 87.5 x 10-8 µm2 suggested by Van der Marel (1958), the estimated surface area of this illite sample is:
 | (4) |
The CEC of our Ca–montmorillonite sample was 64.45 cmolc kg-1 (Table 2), which is smaller than the range of values, 70 to 130 cmolc kg-1, reported for montmorillonite clays by Weaver and Pollard (1973). Shainberg et al. (1987) measured CEC of several montmorillonites and found values from 80 to 122 cmolc kg-1. They reported surface areas of 55.2 to 80.0 x 104 m2 kg-1 and charge densities of 1 charge per 80 to 115 x 10-8 µm2. If we use the charge density value of Van der Marel (1958), 87.5 x 10-8 µm2, our estimated surface area is 34 x 104 m2 kg-1. The estimated surface area and the CEC of the montmorillonite used in this study is less than most values reported for this clay.
X-ray diffraction data for the montmorillonite used in this study indicated the sample was well crystallized and contained a quartz impurity (Fig. 1). Spacings characteristic of montmorillonite are the 1.73-nm peak when Mg saturated and glycolated, collapsing to 1.47 nm when only Mg saturated, further decreasing to 1.15 nm when K saturated, and finally decreasing to 0.98 nm when K saturated and heated. Spacing at 0.332 and 0.42 nm indicate the presence of quartz.
Bulk Density and Water Contents of Clay–Sand Mixtures
Water contents of the clay–sand mixtures used in the flume experiments were adjusted to give matrix water pressure of -5 cm, determined by the criteria described for in-pail estimation. Water contents adjusted in pail were compared with those of samples measured at a -5 cm (-490 kPa) water head with a sand tension plate and the values were in close agreement (Fig. 2)
. Because the air-entry value was not exceeded, water contents at saturation, -5 cm (-490 kPa) and -10 cm (-980 kPa) of water head, should be equal. With air-filled porosity near zero, the bulk density of clay–sand mixtures saturated with water was a predictable function of water content and decreased as water content increased.
|
|
c and Kr values was similar to that for soil detachment at = 3 Pa. The detachment rate of 10% Na–illite and Na–kaolinite sand mixtures was about 2000 times > Na–montmorillonite. The erosion rate of Ca–illite was about 100 times > Ca–montmorillonite, and Ca–kaolinite was about 4 times > Ca–montmorillonite. These differences in erodibility may be related to differences in the effective surface areas of the clays.
Water contents of clay at a given matrix pressure are assumed to be related to effective surface area of clay. Water contents at saturation varied from 0.23 to 6.86 g g-1 (Table 1) and detachment rates at = 3 Pa from 0.01 to 21.39 x 10-3 kg m-2 s-1. It is assumed that water content (W) and detachment rate (Dc) are similar functions of clay surface area. Then changes in Dc with W, d(Dc)/d(W), due to varied clay surface areas are proportional to the ratio Dc/W:
 | (5) |
 | (6) |
 | (7) |
A plot of measured ln(Dc) vs. ln(W) are shown in Fig. 4 .
|
Soil detachment of 10% Ca–montmorillonite was 7 times greater than that of a 10% Na–montmorillonite mixture at = 3 Pa, and soil detachment at = 3 Pa of Ca– and Na–montmorillonite was a small fraction of that for Na–illite and Na–kaolinite (Table 3)
. Shaikh et al. (1988a) determined that the erosion rate of Ca–montmorillonite clay was two orders of magnitude greater than that of Na–montmorillonite. In the Shaikh et al. (1988a) experiment, both clays were unsaturated and compacted at the optimum water content and the erosion tests were conducted on 100% clay materials. In our study and comparison, the test material was 10% clay and 90% sand and the test was run on a saturated mixture. Shaikh et al. (1988b) attributed the greater erosion rate for Ca–montmorillonite to slaking, the breakdown of soil aggregates upon immersion in water, and not dispersion. They reported that the Ca–montmorillonite slaked when immersed in water, but the Na–montmorillonite samples did not slake. In our study, the clay and sand mixtures were completely saturated before the erosion run; therefore, slaking was not the cause for differences in erosion between the Na– and Ca–montmorillonite clays. Ca–montmorillonite has less strength than Na–montmorillonite when compared at the same void ratio (Warkentin and Yong, 1962) . This results because less surface area and repulsion forces enhanced tactoid formation and reduced interdomainal contact. Dowdy and Larson (1971) found that the tensile strength of Ca–montmorillonite, determined by direct tensile stressing of oriented clay films, was less than that of Na–montmorillonite. Results were explained on the basis of larger domains in the Ca systems with a reduced area of interdomainal contact within a given cross section, thus reducing the strength. Na–smectites are single platelets with surface area of 800 m2 g-1. Ca–smectites form tactoids with 5 to 10 platelets in each tactoid, and the active surface area is only 100 m2 g-1. Thus we conclude that the smaller erosion rates found for Na–montmorillonite than for Ca–montmorillonite are due to smaller domains and a larger number of strength-bearing contact areas per unit cross section.
|
Soil detachment rate at = 3 Pa is about 2000 times greater for 10% (w/w) clay–sand mixtures of Na–illite and Na–kaolinite that for Na–montmorillonite (Table 3). Water holding capacity of Na–montmorillonite was also about 30 times > Na–illite and 18 times > Na–kaolinite (Table 2). Differences in detachment and water holding between montmorillonite and illite or kaolinite are consistent with surface areas which range from 700 to 800 m2 g-1 for Na–montmorillonite, 10 to 100 m2 g-1 for Na–kaolinite and Na–illite (Weaver and Pollard, 1973).
The water holding capacity of Na–kaolinite was nearly 2 times that of Na–illite while the soil detachment rate at = 3 Pa was only slightly greater for Na–illite than Na–kaolinite. Ionic bonding between faces and edges with respect to hydrogen bonding between faces may play a more important role in determining properties of kaolinite than illite. Zhao et al. (1991) determined extrapolated shear stress values for Na-saturated kaolinite, illite, and montmorillonite soil clays. Smaller extrapolated shear, at lesser salt concentrations, for Na–illite than Na–kaolinite indicated Na–illite had a smaller number of interparticle bonds resulting from a denser, face-to-face packing of individual particles into tactoids (Zhao et al., 1991).
The dominant factor responsible for increased water retention and decreased erodibililty with Na–montmorillonite in respect to Ca–montmorillonite may be the exposure of interlayers to interparticle interaction. Na–montmorillonite may swell to infinite thickness, but Ca–montmorillonite is limited to about three water layers (Shainberg et al., 1987).
There is a three-fold increase in soil detachment at = 3 Pa when Na replaces Ca as the saturating cation of illite (Table 3). Shainberg and Kemper (1966) found that at relative humidities near 100% exchangeable cations are hydrated. Water molecules associated with adsorbed cations are oriented with their positive dipole outward and provide a mechanism for hydrogen bonding in illite because the oxygen of silica tetrahedrons are oriented with their negative dipole outward. We believe that replacing Ca with more highly hydrated Na as the saturating cation of illite would tend to increase hydrogen bonding to silica tetrahedron, and increase effective particle size with a decrease in CEC and water holding.
A decrease in water holding at -5 cm (-490 Pa) pressure and a 63 fold increase in soil detachment at = 3 Pa was found when Na replaced Ca as the saturating cation of kaolinite (Table 3). Water molecules provide a mechanism for hydrogen bonding between kaolinite faces since the oxygen of silica tetrahedrons are oriented with their negative dipole outward and those bonded to hydroxyl groups of aluminum octahedral layers positive dipole outward. Edges of kaolinite particles are positive dipoles with hydroxyl groups associated with aluminum octahedral and silica tetrahedral structures. Replacing Ca with more highly hydrated Na as the saturating cation in the silica tetrahedral layer would tend to decrease the hydrogen bonding between tetrahedral and octahedral structures and is inconsistent with the observed decrease in water holding and CEC and increase in detachment.
Replacing Ca with Na on the tetrahedral face kaolinite may result in more edge to face ionic bonding between kaolinite particles. The CEC of Ca–kaolinite was 5.05 and Na–kaolinite was 3.78 cmolc kg-1. Increased interparticle ionic bonding and increased particle size of kaolinite when Ca is replaced with Na can be offered as an explanation for decreased CEC, decreased water holding, and increased detachment.
Erosion rates for Na-saturated illite and kaolinite are significantly greater than those of Ca-saturated illite and kaolinite because of greater interparticle bonding between the Na-saturated clays. Decreased erosion of Na- over Ca-saturated montmorillonite, is attributed to an increased number of particle–particle bonds because of interlayer exposure on Na saturation.
Effect of Water Quality
Effect of salt concentration on c, Kr, and soil detachment rate at = 3 Pa (3) was evaluated for only one clay content for each clay (Table 4)
. This clay content was the percentage clay required to cause similar erosion rates among all clays at a hydraulic shear of 3 Pa. Rill erodibility (Kr) is a linear function of soil detachment (Dc) as shown in Eq. [8]:
 | (8) |
|
3) for both Ca–illite and Na–montmorillonite. Detachment rates for salt concentrations of 0.01 and 0.1 M CaCl2 or NaCl were not significantly different (P < 0.05), except for Ca–montmorillonite. Salt concentrations did not significantly affect 3 for Ca–kaolinite. Salt concentration did not affect rill erodibility for Ca–kaolinite and Na–montmorillonite, but decreasing the salt concentration to 0.001 M increased Kr for both Ca–illite and Ca–montmorillonite.
Critical shear stress (c) was significantly increased as salt concentration increased from 0.001 to 0.01 M. For Na–montmorillonite, each increase in salt concentration increased c; for the other clays, there were no differences between c for salt concentrations of 0.1 and 0.01 M. The effects of exchangeable sodium percentage (ESP) and water quality on rill erosion was discussed by Shainberg et al. (1994). For Ca-saturated soils (ESP = 0), water quality (deionized water and 0.005 M CaCl2 solution) had no significant effect on rill erodibility. For soils with ESP of 5 and 10, erodibility was greater with deionized water as the eroding fluid compared to a 0.0005 M CaCl2 solution (Shainberg et al., 1994).
The critical shear stress for Na–montmorillonite was near zero. This can be explained by the fact that clay detachment from the matrix can be either spontaneous or results from the application of a mechanical force equal to or greater than the shear force. Spontaneous dispersion will take place under two conditions: very low electrolyte concentrations (0.001 M) and low cohesion forces between the clay and the matrix. Thus with a 0.001 M solution, spontaneous dispersion occurred for Na–montmorillonite, but did not occur for the other clays. This agrees with conclusions of Arora and Coleman (1979) that smectite and illite clays are more dispersive than kaolinite clays. Negative erodibility for Na–montmorillonite is also possible due to spontaneous dispersion. At a large rate of flow, which is equivalent to a great shear stress, there is not sufficient time for the clay to disperse; thus the concentration of Na clay in the solution is less, the volume of water is large, and a negative erodibility is calculated.
|
ACKNOWLEDGMENTS |
|---|
Received for publication December 24, 1997.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMaterials and methodsResults and discussionREFERENCES |
|---|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| 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 | |||
