SSSAJ Grow Your Career with SSSA
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF) Free
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via ISI Web of Science (1)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Selim, H. M.
Right arrow Articles by Xue, S. K.
Right arrow Search for Related Content
PubMed
Right arrow Articles by Selim, H. M.
Right arrow Articles by Xue, S. K.
GeoRef
Right arrow GeoRef Citation
Agricola
Right arrow Articles by Selim, H. M.
Right arrow Articles by Xue, S. K.
Related Collections
Right arrow Flow
Right arrow Other Environmental Contamination
Right arrow Water Pollution
Soil Science Society of America Journal 66:1773-1783 (2002)
© 2002 Soil Science Society of America

DIVISION S-1—SOIL PHYSICS

Alachlor Transport during Transient Water Flow in Unsaturated Soils

H. M. Selim*, L. Zhou and S. K. Xue

Agronomy Dep., LAES, Louisiana State Univ., Baton Rouge, LA 70803

* Corresponding author (mselim{at}agctr.lsu.edu)


   ABSTRACT
TOP
 NOTES
 ABSTRACT
INTRODUCTION
 THEORY
 MATERIAL AND METHODS
 RESULTS AND DISCUSSION
 SUMMARY AND CONCLUSIONS
 REFERENCES
 
In most herbicide mobility investigations, laboratory experiments are often restricted to the traditional miscible displacement methods where water saturated conditions and constant flux are maintained. In this study, we examine the mobility of alachlor herbicide under soil-water unsaturated and transient conditions. Gigger (a Typic Fragiudalf) surface soil samples were collected from long-term soybean management system under no-till with wheat (NTW, Triticum aestivum L.), conventional tillage with wheat (CTW), and conventional tillage with hairy vetch (CTV, Vicia villosa Roth). Air-dried soil was packed in columns which were horizontally maintained. The inflow end was connected to a water supply which was maintained at zero head. The input pulse was made up of tritium and 14C-alachlor solution. During horizontal infiltration, the advance of the wetting front and volume infiltrated into the soil was monitored over time. When the wetting front reached 25 cm, the column was dismantled and the soil moisture and alachlor concentrations in each 1-cm section quantified. All alachlor distribution profiles appeared retarded and did not show sharp solute fronts. Rather the concentration distributions were of moderately shaped fronts reaching some 10 cm from the source and followed by a gradual decrease up to 20 cm. Alachlor movement was modeled based on transient flow in unsaturated soils. Such results were compared with equilibrium type models of the linear and nonlinear type under steady and transient flow conditions. Parameters needed were soil water-diffusivity, solute-dispersion coefficients and retention parameters for alachlor for individual soils. Independently derived parameters provided a good description of the BTC under unsaturated conditions.


   INTRODUCTION
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
THEORY
 MATERIAL AND METHODS
 RESULTS AND DISCUSSION
 SUMMARY AND CONCLUSIONS
 REFERENCES
 
CONTAMINATION of surface and groundwater from applied agricultural chemicals is a national concern. A primary purpose of the USEPA Clean Water Act (Section 319) and the Coastal Zone Management Act (Section 6217) is to evaluate, demonstrate, and implement best management strategies to improve water quality in watersheds nationwide. Assessment of herbicide sorption–desorption during transport is a prerequisite in minimizing their mobility in the soil profile and potential contamination of the ground water. Alachlor [2-chloro-N-(2,6-diethylphenyl)-N-(methoxymethyl) acetamide] is a commonly used preemergence herbicide for the control of annual grasses and certain broad leaf weeds in corn (Zea mays L.) and soybean [Glycine max (L.) Merr.] fields, among others (Chesters et al., 1989). Alachlor has a water solubility of 240 mg L-1 and vapor pressure of 2.2 x 10-5 mM Hg at 25°C. Alachlor has been classified as Group B2, a probable human carcinogen. USEPA proposed 2 µg mL-1 of alachlor a Maximum Contaminant Level for drinking water (USEPA, 1993). Nevertheless, alachlor has been detected in groundwater in several states including Iowa, Pennsylvania, Maryland, and Minnesota (Cohen, 1992).

Over the last three decades, modeling efforts focused on describing retention mechanisms of herbicides in the soil environment are based on results from batch adsorption methods. Batch data based on 24-h adsorption are commonly regarded as a measure of equilibrium-type retention. Kinetic batch data, which are infrequently measured, have the advantage of taking into account possible time-dependent reactions for adsorption, release, or desorption. Nonequilibrium conditions may be due to heterogeneity of sorption sites and slow diffusion to sites within the soil matrix, i.e., slowly accessible sites. The Freundlich (equilibrium) approach was utilized, for example, to describe alachlor retention in different soils by Bosetto et al. (1993), Clay and Koskinen (1990), and Senesi et al. (1994). Although the Freundlich approach well described alachlor retention, the goodness of fit alone does not provide definite information on the retention mechanisms. Numerous examples which illustrated the use of the equilibrium approach are available for other herbicides including atrazine (2-chloro-4-ethylamine-6-isopropylamino-S-triazine), metribuzin (4-amino-6-tert-butyl-4,5-dihydro-3-methyltio-1,2,4-triazin-5-one), and metolachlor [2-chloro-6'-ethyl-N-(2-methoxy-1-methylethyl)acet-o-toluidide] (see Wauchope and Meyers 1985; Johnson 2001). Examples of kinetic or nonequilibrium models include Kookana et al. (1992) who assessed several kinetic models for describing batch data for simazine (6-chloro-N2,N4-diethyl-1,3,5-triazine-2,4-diamine) and linuron [3-(3,4-dichlorophenyl)-1-methoxy-1-methylurea] in soils. Furthermore, Locke (1992) utilized a three-site model to describe alachlor adsorption versus time for soils under no-till and conventional tillage. Similarly, Xue and Selim (1995) used a multireaction kinetic approach that accounts for reversible and irreversible processes to describe hysteretic adsorption–desorption of alachlor behavior under no-till and conventional tillage.

A number of transport models incorporating various types of sorption mechanisms of herbicides in soils are reported in the literature for describing the mobility of herbicides in soils. Two-site and multisite models with instantaneous reaction type sites and kinetic reaction type sites were developed to describe breakthrough curves (BTCs) from several herbicides in soils. Other transport models that account for equilibrium and kinetic retention during transport were utilized by Gamerdinger et al. (1990)( 1991) for s-triazines in soils and Gaber et al. (1995) for atrazine BTCs in soils. Other classes of transport models are those which incorporate multireactions of the reversible and irreversible type as well as those of the multiregion (mobile–immobile) type models. Examples of such models include the second-order approach that was implemented to describe atrazine BTCs (Ma and Selim 1994). Another model is that of Wilson et al. (1998) who utilized a multireaction approach to describe fluometruron BTCs from soils under different tillage treatments.

Discrepancies between model simulated and observed BTCs are commonly reported from miscible displacement experiments for herbicides in soil columns. Such differences have been attributed to experimental inaccuracies or inconsistencies in different experiments performed. Other factors include constraints related to parameter estimation as well as inaccuracies of inherent assumptions in model formulations (e.g., Xue et al., 1997) and failure of models to adequately describe transport–retention processes. Furthermore, herbicide retention studies are commonly carried out in aqueous soil suspensions by means of batch type reactor or stirred slurry techniques. Other investigations utilize miscible displacement where soil columns are maintained under steady flow and the soil is water saturated. Little information has been reported for herbicide retention in unsaturated soils especially for alachlor sorption in relatively dry soils when a rainfall begins. Accurate description of herbicide concentration distribution resulting from the first infiltration event following alachlor application is paramount in the ability to predict the fate of alachlor during redistribution and subsequent infiltration events.

Because of complexity of simultaneously solving water flow and solute transport equations, several studies attempted to use equivalent steady-state solution to approximate solute transport under transient flow. Wierenga (1977) conducted a study to compare two numerical models for simultaneous movement of water and salts in soil profiles, a steady-state model and a transient model. His results indicated that both techniques yield comparable results when concentrations are plotted versus cumulative drainage. With similar approaches, Russo et al. (1989a) showed that under transient nonmonotonic water flow, calculations of solute transport based on the assumption of an "equivalent" steady-state water flow may considerably overestimate solute displacement. If the immobile water is taken into consideration, Russo et al. (1989b) found that the steady-state profiles of total resident solute concentration at early times are characterized by higher concentration peaks and smaller travel distances than the corresponding transient profiles. At large time, however, they found that the steady-state profiles with relatively large mass transfer coefficient between mobile and immobile phases are characterized by higher concentration peaks but larger travel distances than the transient profiles. On the basis of theoretical derivation, Bond and Wierenga (1990) showed that simple solutions of the convection–dispersion equation for transient flow conditions have the same mathematical form as those for steady flow. However, a different coordinate system and a time-averaged, water-content-weighted dispersion coefficient must be used in the solutions for unsteady-flow conditions. Later, Porro and Wierenga (1993) showed that transport parameters determined under steady-state conditions can be useful in describing transport under transient conditions.

The objectives of this study were to: (i) determine experimentally the extent of alachlor retention during infiltration into air-dry soils; (ii) quantify alachlor mobility, water-extractable, and residual alachlor distribution during infiltration in soils from fields with different tillage treatments; (iii) assess the capability of predicting alachlor mobility on the basis of the convection–dispersion equation (CDE) assuming steady flow vs. models that account for transient water flow in water-unsaturated soils; and (iv) assess the capability of independently derived parameters of the equilibrium as well as the kinetic type in describing alachlor transport during infiltration into air-dry soils.

To achieve these objectives, the mobility of alachlor in unsaturated soil columns subject to horizontal infiltration was carried out for soils from three different tillage treatments. Results of alachlor concentration profiles in each soil were predicted by means of independently derived model parameters from retention kinetic (batch) adsorption experiments. In addition, we predicted alachlor concentrations on the basis of transient flow models as well as on simplified steady water flow models.


   THEORY
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
 THEORY
MATERIAL AND METHODS
 RESULTS AND DISCUSSION
 SUMMARY AND CONCLUSIONS
 REFERENCES
 
Transient Water Flow
A prerequisite for describing reactive solutes under transient flow conditions is knowledge of water content and flux density during transport in soil. For one-dimensional transient water flow in a uniform horizontal soil, the governing equation is,

[1]
where {theta} is the volumetric water content (cm3 cm-3), t is time (s), and x is the horizontal distance (cm). D({theta}) (cm2 s-1) is the soil water diffusivity function and is assumed to be of the exponential form,

[2]
where a and b are constant. The soil water diffusivity function was obtained by Boltzmann transformation {chi} = {chi}({theta}) = x t-1/2, given by

[3]

The use of Boltzmann transformation assumes that {theta} is a single-valued function of {chi}, where {theta}i is the initial water content. Because the measured diffusivity function cannot be described very well by a single exponential formula, it was expressed in piecewise exponential format. The initial and boundary conditions for the transient water flow are

[4]

[5]

[6]
where {theta}i and {theta}0 are initial water content and saturated water content, respectively, and L (cm) is the length of the horizontal soil column.

Transient Solute Transport
The equation describing the transport of reactive solutes during transient soil-water unsaturated flow conditions is given by

[7]
where Jw [= Jw(x, t)] is the transient water flux density (cm s-1) and Ds is solute dispersion coefficient (cm2 s-1). Ds was assumed to follow

[8]
where Do is the molecular diffusion coefficient (cm2 s-1) of a solute in a free solution, a and b are dimensionless constant, and {lambda} is the dispersivity (cm). The initial and boundary conditions are

[9]

[10]

[11]
where Cs is the concentration (µg mL-1) of the applied pulse. The adsorption phase S can be described by a linear equilibrium retention model (Eq. [16]) or by a multireaction model proposed by Xue and Selim (1995),

[12]

[13]

[14]

[15]
where Se is the amount retained by equilibrium type sites, ke is the associated distribution coefficient (cm3 g-1), and n is a nonlinear Freundlich parameter (dimensionless). In addition, S1 is the amount retained by kinetic type sites, S2 represents the amount of alachlor retained irreversibly and S is the total amount of solute (herbicide) retained by the soil matrix (µg g-1 soil). The parameters k1, k2, and k3 are the associated rate coefficient (s-1). The coupled transient water flow and herbicide transport Eq. [1] and [7] subject to initial and boundary conditions Eq. [4] through [6] and Eq. [9] through [11], respectively, were solved by finite difference method to obtain water content and herbicide distributions at different transport time.

Steady-State Flow Model
Because analytical solutions for solute transport in unsaturated soil are quite complex, an alternative is the use of analytical solution of solute transport under steady-state flow conditions. Furthermore, if one assumes that adsorption of a herbicide can be described by a linear equilibrium model,

[16]
where kd is the distribution constant (cm3 g-1), then, Eq. [7] can be rewritten as

[17]

The parameter R (= 1 + {rho}kd/{theta}) is the retardation factor, V (= q/{theta}) is the pore water velocity, and q is Darcian (steady) water flux (cm s-1). The associated initial and boundary conditions are

[18]

[19]

[20]

Here Cs is the concentration (µg mL-1) of the applied solution. Equation [17] subject to Eq. [18] through [20] can be solved through Laplace transform (Lindstrom et al., 1967; van Genuchten and Alves, 1982), and the analytical solution is

[21]
where erfc(z) is the complementary error function at z. Equation [21] was used to describe alachlor distribution profiles in unsaturated soil columns. Here averages for Darcian water flux and average moisture content during infiltration into each soil column were assumed. Our average soil moisture content was calculated by

[22]
where I is the accumulative infiltration solution in cm, Lf is the advance of wetting front in cm, and {theta}i the initial soil moisture content. The average Darcian water flux was based on

[23]
where T is the total time for infiltration (s). The average pore water velocity is thus given by

[24]


   MATERIAL AND METHODS
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
 THEORY
 MATERIAL AND METHODS
RESULTS AND DISCUSSION
 SUMMARY AND CONCLUSIONS
 REFERENCES
 
Soils
Surface soils (0–15 cm) were sampled from long-term, no-till and conventional tillage cotton (Gossypium hirsutum L.) plots and under different winter cover treatments, from Macon Ridge Research Station at Winnsboro, LA. The soil was a Gigger soil (Silt mixed thermic Typic Fragiudalf) and contained 22% (w/w) sand, 66% (w/w) silt, and 12% (w/w) clay. The pH (1:1 soil:water ratio) was 5.37 in the no-till and 5.23 in the conventionally tilled soil. Soils from three different tillage treatments were used; NTW: no till with wheat as winter cover crop; CTW: conventional tillage with wheat as winter cover, and CTV: conventional tillage with hairy vetch as winter cover. Soil organic matter content was 1.45% (w/w) for NTW, 0.81%(w/w) for CTW, and 0.98%(w/w) for CTV treatments.

Horizontal Infiltration Experiments
The experimental setup for horizontal infiltration into a uniformly packed air-dry soil is shown in Fig. 1 . Each soil column was made of 1-cm sectioned acrylic plastic tubing with 3.15-cm-i.d. At the inflow end a high-conductance filter paper was installed to distribute the flow uniformly. The water supply was a Marriotte burette with zero pressure head at the midpoint of the horizontal soil column in order to maintain the boundary x = 0 at saturation. In the outlet end of the column several holes were installed in order to allow for air outflow during water infiltration. The input pulse solution was made of 0.9711 µg mL-1 of alachlor in 0.005 M CaCl2. This is equivalent to application rate of 1 kg active ingredient alachlor per hectare. Carbon-14 labeled alachlor was utilized as a reactive tracer solute in this study.



View larger version (21K):
[in this window]
[in a new window]
  		Fig. 1. Experimental setup for horizontal infiltration of alachlor and tritium solution into a column of dry soil.

 
Infiltration was initiated by connecting the inflow end of the column to the solution reservoir. The volume of infiltrated solution was monitored with time. In addition, the advance of the wetting front into the air-dry soil was recorded visually. When the advancing wetting front reached 25 cm from the source, infiltration was terminated. Column sections were quickly dismantled and the soil in each 1-cm section placed in previously weighed beakers. The amount of alachlor in individual soil sections was quantified by the extraction procedure outlined below and the soil moisture content versus horizontal distance was determined gravimetrically. Extraction on individual sections commenced immediately following infiltration to minimize the redistribution of alachlor between the sorbed and solution phase. Further assessment of solute mobility in the soil was achieved by the addition of tritium, as a reference or nonreactive tracer solute, in the input pulse solution. This was carried out for the NTW column only. Additional information on the experimental conditions of infiltration is given in Table 1.


View this table:
[in this window]
[in a new window]
  		Table 1. Experimental conditions of the horizontal soil columns and alachlor mass balance for Gigger soil from three different tillage treatments.

 
Alachlor Extraction
Water extractable alachlor was quantified by mixing 20 mL of 0.005 M CaCl2 to the soil from each 1-cm section. The soil solution was then equilibrated for 24 h, centrifuged at 3300 g, and the supernatant collected in 250-mL flasks. Repeated additions of CaCl2 solution, mixing for 5 min, centrifuging, and decanting of supernatant were carried out three times. The amount of water extractable alachlor was thus based on total amount extracted in these steps. Methanol residual or strongly retained alachlor was based on subsequent extraction of the soil with methanol. Here 20 mL of methanol was added and mixed with the soil; the mixture was shaken for 12 h, and then centrifuged. The supernatant was decanted and the tube with soil was weighed for determination of the residual or methanol extractable alachlor from the soil. Alachlor as well as tritium were analyzed by liquid scintillation counting (LSC).

Batch Kinetics
Retention (adsorption–desorption) of alachlor in the three soils was carried out earlier by Xue et al. (1997). Here we focus on the adsorption experiments which were conducted (in duplicates) by mixing 10 g of air-dry soil and 20 mL of alachlor solutions of varying concentrations in 40-mL polytetrafluoroethylene tubes. Alachlor initial concentrations (Co) used were 0.5, 1, 2, 5, 10, 20, 30, and 50 mg L-1 in 0.005 M CaCl2 background solution with four reaction times (24, 48, 120, and 528 h). Samples were shaken until each reaction time was reached, and then were centrifuged, and the supernatants were withdrawn for alachlor analysis.


   RESULTS AND DISCUSSION
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
 THEORY
 MATERIAL AND METHODS
 RESULTS AND DISCUSSION
SUMMARY AND CONCLUSIONS
 REFERENCES
 
Soil Moisture Profiles
Measured and calculated soil moisture profiles in the horizontal soil columns are shown in Fig. 2 . These profiles, which represent the moisture distribution at the termination of horizontal infiltration, clearly show sharp wetting fronts. Such sharp wetting fronts are not unexpected in these air-dry soils. The predicted soil moisture profiles are illustrated by the solid curves and well describe the entire moisture distributions. These predictions were obtained on the basis of the water flow Eq. [1] and the parameters used for each soil are given in Table 2. The soil-moisture diffusivity vs. soil moisture content for all three columns and was obtained from Eq. [3] and are shown in Fig. 3 . The measured D({theta}) vs. {theta} was fitted by means of Eq. [2] and best-fit values for the parameters a and b obtained. The fitted diffusivity functions were subsequently utilized to model horizontal water infiltration into our air-dry soils. We found that it was necessary to adjust the parameters a and b to achieve the moisture predictions shown in Fig. 2. The adjustments were necessary for the NTW case only and was achieved by multiplying parameters a and b by a factor of 1.05 to 1.03. With the original a and b parameters, the calculated moisture front lagged some 2 to 3 cm behind that measured, and reasons for such adjustments were not clearly understood.



View larger version (21K):
[in this window]
[in a new window]
  		Fig. 2. Measured and calculated soil moisture content versus horizontal distance in the NTW, CTW, and CTV soil columns.

 

View this table:
[in this window]
[in a new window]
  		Table 2. Soil-water diffusivity parameters of Eq. [9] used in modeling horizontal water infiltration in unsaturated Gigger soil columns.

 


View larger version (19K):
[in this window]
[in a new window]
  		Fig. 3. Measured soil water diffusivity versus soil moisture content for the NTW, CTW, and CTV soil columns.

 
The applicability of the flow equation in describing water flow in these soils is further depicted by the linear relation of cumulative infiltration vs. square root of time for all three soils (figure not shown). For the NTW and CTW treatments, similar rates of infiltration were observed. In contrast, enhanced rates of infiltration were observed for the CTV where the hairy vetch was the cover crop during the winter months. Such differences were perhaps due to differences in bulk densities and saturated water contents among the three treatments (see Table 1).

Further assessment of solute mobility in the soil was achieved by the addition of tritium, as a reference or nonreactive tracer solute, in the input pulse solution. This was carried out for the NTW column only and is illustrated in Fig. 4 . Here relative concentration for tritium coincides well with the soil moisture profile. Thus both the water and tritium did not separate. Furthermore, the tritium front did not lag behind the water front, which clearly illustrates that no retardation of the applied tritium was observed. Such observation is expected for the transport of nonreactive chemicals during infiltration in an initially dry soil.



View larger version (17K):
[in this window]
[in a new window]
  		Fig. 4. Measured soil moisture content and tritium concentration versus horizontal distance in the NTW soil column.

 
Alachlor Distribution Profiles
Alachlor concentration distributions are shown in Fig. 5 for the three columns. The results are presented for water and methanol extractable fractions and are expressed on soil mass basis. For all three soil treatments, the water extractable fraction exceeded the sorbed (methanol) fraction. Such a finding is inconsistent with data on alachlor retention kinetics as well as transport in various soils (Chesters et al., 1989). Because of the soil extraction procedure, it is believed that the water-extractable phase was overestimated at the expense of the sorbed phase. This is because during extraction a fraction of that sorbed on the soil matrix was likely released into the solution phase. Water extraction preceded extraction with methanol.



View larger version (28K):
[in this window]
[in a new window]
  		Fig. 5. Measured soil-water, methanol, and total extractable alachlor concentration versus horizontal distance in the NTW, CTW, and CTV soil columns.

 
Both water and methanol extractable phases followed each other with the highest concentrations close to the source (x = 0) and gradually decreasing with distance (see Fig. 5). Generally, alachlor concentration profiles were retarded and did not show sharp fronts. Rather the profiles were of moderately shaped fronts reaching some 10 cm from the (input) source and were followed by a gradual decrease in concentrations for distances of 10 to 20 cm from the source. On the basis of these concentration profiles, a mass balance was calculated. The extent of recovery of alachlor averaged 95% of that applied to individual columns (see Table 1). These high recoveries are perhaps due to time of reaction of alachlor with the soil matrix in our horizontal column experiments. This was a direct consequence of our primary objective of quantifying herbicide mobility in the soil at the termination of water infiltration. We recognize that the amount of nondesorbable herbicide almost always increases with time (Wauchope and Myers, 1985) and lack of complete alachlor recovery from soils has been reported. For example, Helling et al. (1988) found that alachlor recovery (1 d after application) was 57% of that surface broadcast as an emulsifiable solution. Bosetto et al. (1993) reported only 50% recovery was obtained after several washings in distilled water.

Transient Flow Model
A prerequisite to modeling solute mobility under transient flow conditions is that water flow conditions are well described as was achieved above. As a result, alachlor mobility was modeled by the solute transport Eq. [7] for transient flow conditions in water-unsaturated soils. A comparison of measured and model predicted distributions of the total amount of alachlor (sum of water and methanol extractable phases) are given in Fig. 6 . Two modeling efforts are illustrated in Fig. 6. The solid line represents predictions based on the multireaction model (Eq. [12]–[15]) where the model parameters (ke, n, k1, k2, and k3) were based on independent estimates from the kinetic batch experiments of Xue et al. (1997). These parameters are given in Table 3 and provide best-fit descriptions of alachlor retention versus time for a wide range of initial (input) concentrations, as illustrated in Fig. 7 . As illustrated by the calculated alachlor profile distributions, poor model predictions were obtained. Most disappointing was the failure of the model to describe the extent of the advance of the solute front and excessive tailing of the alachlor distributions for all three-soil treatments.



View larger version (25K):
[in this window]
[in a new window]
  		Fig. 6. Total concentration of alachlor versus horizontal distance in the NTW, CTW, and CTV soil columns. Solid curves are based on the kinetic multireaction model where model parameters were independently derived from batch kinetic and miscible displacement (transport) experiments given in Table 3 and 4, respectively.

 

View this table:
[in this window]
[in a new window]
  		Table 3. Equilibrium and kinetic model parameters based on batch data for alachlor retention for the three tillage treatments of Gigger soil.

 

View this table:
[in this window]
[in a new window]
  		Table 4. Parameters based on model calibration using alachlor BTCs from miscible displacement in Gigger soil{dagger}.

 


View larger version (23K):
[in this window]
[in a new window]
  		Fig. 8. Measured sorption isotherms for Gigger soil after 1 d of reaction for the NTW, CTW, and CTV treatments. Solid curves are fitted based on nonlinear (Freundlich) adsorption.

 


View larger version (21K):
[in this window]
[in a new window]
  		Fig. 7. Alachlor concentration in soil solution versus time from kinetic batch experiment for Gigger soil (CTW treatment) for a wide range of initial concentrations (Ci's). Solid curves are based on the kinetic mutlireaction model.

 
Failure of batch-derived model parameters to describe alachlor transport was also reported by Xue et al. (1997). Specifically, attempts to describe BTCs from several miscible displacement columns were made. In fact, their predicted BTCs were always less retarded than measured BTCs from six different soil (column) tillage treatments. In their experiment, each soil column was fully saturated and steady water flow conditions were maintained. Because of these failures, Xue et al. (1997) calibrated the multireaction model with the BTC data using nonlinear parameter optimization techniques. Best-fit (calibrated) model parameters (ke, n, k1, k2, and k3) are given in Table 4. These parameters were used in our modeling efforts for our unsaturated transient flow conditions (see dashed curves in Fig. 6). For all three soil columns, improved description of the concentration profiles were achieved when calibrated model parameters were used.

Steady-State Model
In an effort to test whether simplified modeling approaches are capable of predicting alachlor distributions into our air-dry soils, we decided to utilize the simple concept of the distribution coefficient (Kd) as a measure of retention during infiltration. To test this concept, we fitted the analytical solution for steady water state flow, i.e., Eq. [21], using CXTFIT 2.0 (Toride et al., 1995). Here average water flux and average soil moisture contents were based on our experimental column measurements (Eq. [22]–[24]). The Kd for the three soil treatments were derived from the batch results of Xue et al. (1997). The Kd values were obtained from best fit of alachlor isotherms shown in Fig. 8 for a 24-h adsorption. The best-fit values for Kd are given in Table 3. Therefore, the only parameter we optimized using CXTFIT was the hydrodynamic dispersion coefficient Ds for each soil column. On the basis of the fitted dispersion coefficients Ds, a dispersivity {lambda} (= Ds/V) was computed. The obtained dispersivity values are 6.94, 0.687, and 1.165 cm for NTW, CTW, and CTV, respectively. These values for {lambda} were used in all simulations.

Alachlor predictions based on the steady-state model were surprisingly good for all soil columns (see Fig. 9) . In fact, these predictions were better than those based on transient and unsaturated flow using the multireaction approach (see Fig. 6). It is conceivable that Kd adequately represents retention over short infiltration events. Duration of infiltration events is relatively short when compared with redistribution events subsequent to infiltration where Kd is perhaps not valid. When we utilized this simplified approach and incorporated Kd with our transient flow model, the predictions were similar to those based on the steady-state model. These predictions are shown in Fig. 10 (dashed curves) and illustrate the useful nature of this simplified approach where average values for soil moisture and water flux were used.



View larger version (23K):
[in this window]
[in a new window]
  		Fig. 9. Total concentration of alachlor versus horizontal distance in the NTW, CTW, and CTV soil columns. Solid curves are based on steady-state flow and linear equilibrium (CXTFIT model).

 


View larger version (24K):
[in this window]
[in a new window]
  		Fig. 10. Total concentration of alachlor versus horizontal distance in the NTW, CTW, and CTV soil columns. Solid and dashed curves are based on alachlor transport in unsaturated and transient flow conditions. Retention was characterized by linear (Kd) and Freundlich (Kf and N) equilibrium models.

 
Because of the observed nonlinear behavior of alachlor isotherms shown in Fig. 8, we also tested the capability of our transient flow model in describing alachlor concentration distributions when retention based on nonlinear retention of the equilibrium type is assumed. Specifically the nonlinear (Freundlich) model,

[25]
was used to describe the isotherms of Fig 8 and estimates for the associated coefficients for the Freundlich parameter Kf and the dimensionless parameter N were obtained by means of nonlinear least-square optimization. These values are given in Table 3 and were used in the predictions shown by the solid curves in Fig 10. For all cases, the predictions based on the nonlinear approach were inferior to those obtained by a simple linear adsorption and were perhaps due to the low input concentration used in our transport experiments. The only exception is that for CTV where both predictions based on linear and nonlinear retention were similar. This is most likely due to the fact that for CTV the nonlinear parameter N (= 0.97) was close to 1.

The results based on our efforts where Kd was the only retention mechanism illustrate the fact that the use of the multireaction model did not yield adequate predictions. This implies that multireaction model failed to describe this data set. In addition, none of our modeling efforts were capable of describing the entire concentration distributions versus distance adequately. Therefore, new models must be developed to describe the concentration peaks or front along with the extended or gradual concentration versus distance. Results from similar horizontal as well as vertical infiltration experiments did not exhibit such concentration distributions. In fact, results for phosphors, chloride, tritium, and nitrates showed clearly sharp peaks with no extended or gradual concentration decrease ahead of the solute fronts (Bond et al., 1982; Tan et al., 1992; Chen et al., 1996; Kato et al., 1996). We are not aware of such behavior for herbicide transport in unsaturated soil columns, however. One possible explanation could be that, in our soil columns, we are dealing with a multiregion or dual-porosity system. It is conceivable that a bimodal distribution of the flow region is responsible for our measured concentration distributions. A fraction of soil region was perhaps more conducting, thus, resulting in an unequal rates of flow in the soil, i.e., preferential flow. Field and laboratory observations that support the two flow domain concept are abundant in the literature. For example, bimodal peaks for nonreactive solutes were reported for conditions where steady soil water flow was dominant (Hornberger et al., 1990; Leij and Dane, 1989; Hamlen and Kachanoski, 1992). Hornberger et al. (1990) observed dual peaks in a transport study in an uphill forest soil where an impulse bromide was applied.

Tillage Effects
If preferential flow is greater under no-till conditions, then BTCs for tracer solutes should occur sooner than for conventional-tillage columns. Wilson et al. (1998) found that the hydrodynamic dispersion coefficient Ds was consistently higher for soil columns from no till than from conventional till. They also found that the response of hairy vetch cover systems was inconsistent, however. Specifically, Wilson et al. (1998) found that Ds for hairy vetch treatments with till and no till showed no numerical differences. Our results for hydrodynamic dispersion, (Ds of Eq. [17]) determined on the basis of CXTFIT, were much higher for no-till wheat or NTW than conventional-till wheat or CTW (Ds of 14.49 versus 1.41 cm2 h-1). In a long-term corn study in Iowa, Singh and Kanwar (1991) found that values for hydrodynamic dispersion coefficient for chloride from no-till columns were 2.5 times higher than average values for conventional tillage. Our results are also consistent with results of Xue et al. (1997), which are based on tritium BTCs in undisturbed columns under steady and water-saturated flow conditions. Xue et al. (1997) concluded that high Ds values for no-till treatments was due to the fact that wheat roots have stronger penetration than other crop roots and may thus account for the larger dispersion. High dispersion is indicative of greater variation in pore velocities, which one would anticipate to occur under no-till conditions. The Ds value for the hairy vetch column (CTV) was 4.03 cm2 h-1, which is higher than that for CTW. This higher dispersion is most likely due to increases in organic matter for the vetch cover system (CTV). Organic matter increases due to vetch cover promote greater aggregation and stability of aggregates.


   SUMMARY AND CONCLUSIONS
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
 THEORY
 MATERIAL AND METHODS
 RESULTS AND DISCUSSION
 SUMMARY AND CONCLUSIONS
REFERENCES
 
We examined the mobility of alachlor in unsaturated soil columns subjected to horizontal infiltration. An input pulse of 14C-alachlor solution having low concentration was introduced to the column and the advance of the volume infiltrated was monitored. When the wetting front reached 25 cm, each column was dismantled and soil moisture and alachlor concentrations in 1-cm sections were quantified. Alachlor concentrations versus distance were retarded behind the wetting front and did not show sharp fronts. Alachlor distributions were of moderately shaped fronts reaching some 10 cm from the (input) source and were followed by a gradual decrease in concentrations with distance. Attempts were made to describe alachlor movement by solute transport equation for transient flow in water-unsaturated soils. Results were compared with equilibrium type models of the linear and nonlinear type under steady as well as transient flow. Our major findings are as follows.

  1. Use of kinetic approaches for alachlor retention along with transient and water-unsaturated flow conditions yielded inadequate predictions of the alachlor fronts for all soil columns.
  2. Best predictions were obtained when steady water flow conditions and linear adsorption were assumed. This is perhaps due to time of reaction of alachlor infiltration into the dry soil.
  3. The value for hydrodynamic dispersion Ds was substantially higher for columns from the no-till than from the conventional-till treatment, which was indicative of improved microstructure under no-till conditions. Moreover, Ds for the hairy vetch column (CTV) was higher than that for conventional till with wheat (CTW). This is due to increases in organic matter in the vetch cover treatment which promotes greater aggregation and stability of aggregates.
  4. None of the models investigated were capable of describing the gradual decrease in alachlor concentrations beyond the fronts.


   NOTES
TOP
 NOTES
ABSTRACT
 INTRODUCTION
 THEORY
 MATERIAL AND METHODS
 RESULTS AND DISCUSSION
 SUMMARY AND CONCLUSIONS
 REFERENCES
 
Louisiana Agric. Exp. Stn. Manuscript no. 02-09-0107.

Received for publication November 19, 2001.


   REFERENCES
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
 THEORY
 MATERIAL AND METHODS
 RESULTS AND DISCUSSION
 SUMMARY AND CONCLUSIONS
 REFERENCES
 




This article has been cited by other articles:


Home page
 Soil Sci.Home page
F. Sopena, C. Maqueda, and E. Morillo
Influence of Soil Characteristics and Formulation on Alachlor Dissipation in Soil
Soil Sci. Soc. Am. J., May 1, 2008; 72(3): 767 - 774.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF) Free
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via ISI Web of Science (1)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Selim, H. M.
Right arrow Articles by Xue, S. K.
Right arrow Search for Related Content
PubMed
Right arrow Articles by Selim, H. M.
Right arrow Articles by Xue, S. K.
GeoRef
Right arrow GeoRef Citation
Agricola
Right arrow Articles by Selim, H. M.
Right arrow Articles by Xue, S. K.
Related Collections
Right arrow Flow
Right arrow Other Environmental Contamination
Right arrow Water Pollution

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE <