JEQ Journal of Natural Resources and Life Sciences Education
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 Related articles in JEQ
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Citing Articles
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 Zhu, H.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Selim, H. M.
Right arrow Articles by Zhu, H.
Agricola
Right arrow Articles by Selim, H. M.
Right arrow Articles by Zhu, H.
Related Collections
Right arrow Organic Compounds
Right arrow Agricultural Pesticides
Published in J. Environ. Qual. 34:325-335 (2005).
© ASA, CSSA, SSSA
677 S. Segoe Rd., Madison, WI 53711 USA

TECHNICAL REPORTS

Organic Compounds in the Environment

Atrazine Sorption–Desorption Hysteresis by Sugarcane Mulch Residue

H. M. Selim* and H. Zhu

Sturgis Hall, Department of Agronomy and Environmental Management, Louisiana State University Agricultural Center, Baton Rouge, LA 70803

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

Received for publication November 12, 2003.

   ABSTRACT
TOP
 NOTES
 ABSTRACT
INTRODUCTION
 MODEL FORMULATION
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 CONCLUSIONS
 REFERENCES
 
Sorption and desorption kinetics are essential components for modeling the movement and retention of applied agricultural chemicals in soils and the fraction of chemicals susceptible to runoff. In this study, we investigated the retention characteristics of sugarcane (Saccharum spp. hybrid) mulch residue for atrazine (2-chloro-4-ethylamino-6-isopropylamino-1,3,5-triazine) based on studies of sorption–desorption kinetics. A sorption kinetic batch method was used to quantify retention of the mulch residue for a wide range of atrazine concentrations and reaction times. Desorption was performed following 504 h of sorption using successive dilutions, followed by methanol extraction. Atrazine retention by the mulch residue was well described using a linear model where the partitioning coefficient (Kd) increased with reaction time from 10.40 to 23.4 cm3 g–1 after 2 and 504 h, respectively. Values for mulch residue Kd were an order of magnitude higher than those found for Commerce silt loam (fine-silty, mixed, superactive, nonacid, thermic Fluvaquentic Endoaquepts) where the sugarcane crop was grown. A kinetic multireaction model was successful in describing sorption behavior with reaction time. The model was equally successful in describing observed hysteretic atrazine behavior during desorption for all input concentrations. The model was concentration independent where one set of model parameters, which was derived from all batch results, was valid for the entire atrazine concentration range. Average atrazine recovery following six successive desorption steps were 63.67 ± 4.38% of the amount adsorbed. Moreover, a hysteresis coefficient based on the difference in the area between sorption and desorption isotherms was capable of quantifying hysteresis of desorption isotherms.

Abbreviations: MRM, multireaction model


   INTRODUCTION
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
MODEL FORMULATION
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 CONCLUSIONS
 REFERENCES
 
A RECENT SHIFT of sugarcane harvesting technology from burning leaves to leaving leaves on the soil surface is raising several economic and environmental concerns. The traditional whole-stalk harvest system is where the whole stalks of sugarcane plants are piled, cut, burned, picked up, and transported to the mill. The new system involves the use of a combine harvester that cuts the cane stalks into billets, which are directly loaded into wagons for transport to the mill. Extractor fans in the combine harvester separate leaf material from billets. The plant residue falls to the soil surface. It has been reported that mulch produced from the leaf material and plant residue can promote disease and lower yields in the subsequent ratoon crop (Richard, 1999). As a result, burning the leaves off the whole stalks before harvest or burning of the crop residue on the soil surface following harvest are common practices to reduce their effect on disease and/or possible yield reduction.

The presence of mulch residue on the soil surface protects it from water and wind erosion and conserves soil water for crop production (Unger and Wiese, 1979). Recycled crop residues can be a temporary storage medium for herbicides, altering patterns of chemical dispersion in conservation tillage when compared with conventional practices (Dao, 1991). The effect of surface crop residues on interception, subsequent wash-off, and movement of herbicides through soil are major concerns associated with no-tillage practices.

A number of studies quantified pesticide concentrations retained by crop residues. It is often assumed that applied herbicides eventually would leach and be recovered by the soil. However, a previous year's crop residues may intercept chemical spray and thus reduce the efficacy of soil-applied herbicide (Banks and Robinson, 1982; Ghadiri et al., 1984; Crutchfield et al., 1985). In laboratory and field studies, Dao (1991) reported that wheat straw exhibited a strong affinity for metribuzin and its S-ethyl analog. The retention capacity was associated with the lignin fraction of the wheat residues. Dao (1995) reported that the addition of wheat straw to the soil elevated organic C concentration in the near-surface zone of no-till soils, which resulted in two- to fivefold increase in metribuzin retention. In a Michigan study, corn residues showed a sorption capacity for benzene, ethylbenzene, and 1,2,3-trichlorobenzene that was 35 to 60 times greater than that of the surface soil (Boyd et al., 1990). The sorption capabilities of corn residues and soil organic matter for nonionic organic compounds were nearly identical as indicated by the similarity of their Koc values (Koc = Kd/OC). Here, Kd represents the distribution coefficient for retention and OC the organic carbon fraction in the soil. They concluded that it maybe unnecessary to distinguish the organic carbon in crop residue from humus carbon in soil when predicting nonionic organic compounds sorption coefficients. In all these investigations, the time-dependent nature of herbicide retention by mulch residues on Kd during sorption and desorption was not addressed.

Atrazine is a triazine family herbicide used worldwide since 1952 to control annual weeds in corn and sorghum, among other crops. Atrazine is a major herbicide that is used extensively in sugarcane production in Louisiana. Chemical weed control programs for sugarcane usually require two herbicide applications, one in early spring and another before the crop canopy closes. Sugarcane producers in southern Louisiana refer to the latter application as the layby treatment, which follows the last cultivation in the sugarcane growing season. Moreover, atrazine is recommended as a postemergence treatment to control winter or early-spring weeds or as a postemergence treatment for layby or fallow fields.

Numerous studies characterizing sorption of atrazine on humic acids, clays, and oxyhydroxides have been published (Kalouskova, 1989; Piccolo et al., 1992; Senesi et al., 1995; Laird et al., 1994). In contrast, data on atrazine sorption and desorption on crop residues are at best sparse. Moreover, a literature search reveals that no studies have been performed on the sorption–desorption of atrazine by sugarcane mulch residue. In this study, the objectives were to (i) determine experimentally the sorption and desorption characteristics of atrazine by sugarcane mulch residue, (ii) assess the extent of hysteretic desorption behavior for atrazine by mulch residue, (iii) quantify the sorption capacity for atrazine by sugarcane residue, and (iv) examine the ability of a multireaction model (MRM) in describing atrazine retention kinetics by the sugarcane residue. Such information is essential for predicting the fate of applied herbicides and the implementation of corrective actions needed to reduce their off-target movement from sugarcane fields.


   MODEL FORMULATION
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
 MODEL FORMULATION
MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 CONCLUSIONS
 REFERENCES
 
Freundlich Model
The Freundlich equation is empirical in nature and relates the amounts sorbed and solution-phase concentrations at equilibrium. It can be expressed as:

[1]
where S is the amount of solute retained by the soil or mulch residue (mg kg–1), C is the solution concentration (mg L–1), Kf is the Freundlich distribution or partitioning coefficient (mg kg–1)/(mg L–1)N, and N is a dimensionless parameter commonly less than unity. The parameter N is a measure of the extent of the heterogeneity of sorption sites having different affinities for solute retention by matrix surfaces, where sorption by the highest energy sites takes place preferentially at the lowest solution concentrations. This approach was used frequently in describing herbicide retention including atrazine sorption by soils (Ma et al., 1993). When N = 1, the Freundlich model approaches to that of a linear model (S = KdC), and the Kd (cm3 mL–1) is used instead of Kf.

Multireaction Model
The two-site equilibrium-kinetic model of Selim et al. (1976) is perhaps one of the earliest multisite or multireaction approaches for describing retention and transport behavior of reactive solutes in porous media. Basic to the multisite approach is that the soil solid phase is made up of different constituents (soil minerals, organic matter, iron and aluminum oxides), and that a solute species is likely to react with various constituents (sites) by different mechanisms. The model is multipurpose in nature, which accounts for linear as well as nonlinear reaction processes of the equilibrium and/or kinetic (reversible and irreversible) type. The capability of the model is not limited to describing commonly measured batch-type sorption data (following a specific reaction time, e.g., 1 d) but also in describing changes in concentration with time of reaction during sorption as well as desorption. The uniqueness of this model is that its aim is to describe the reactivity of solutes with natural systems versus time during sorption or desorption. In contrast, for most models (e.g., simple linear, Freundlich, Langmuir, dual domain reactivity models [DDRM], and treble domain reactivity models [TDRM]), two distinct sets of parameters are obtained, one for sorption and one for desorption (Huang et al., 1997; Huang and Weber, 1998; Lesan and Bhandari, 2003). Moreover, the use of such models yields a set of parameters that is only applicable for a specific reaction time. On the other hand, the MRM provides a comprehensive accounting of the sorption–desorption processes, where a single set of parameters is sought that is applicable for an entire data set and for a wide range of initial (or input) concentrations.

Consistent with the multiple reactions considered in our model, Clay and Koskinen (1990) and Locke (1992), among others, argued that sorption for several herbicides may be assumed to react at different rates with different sites on matrix surfaces. Along the same lines, a multiple or distributive reactivity approach was proposed by Weber et al. (1992). In fact, Huang et al. (1997) tested linear and Freundlich models as well as DDRM in describing sorption isotherms for a wide range of soils and sediments. They concluded that the success of multiple domain reactivity models (dual or treble), comprised of linear and one (or two) Langmuir-type sorption mechanisms, implies that simple linear-phase partitioning is not valid. More recently, based on Freundlich fittings of sorption for different reaction times, Huang and Weber (1998) showed that Koc parameters are highly time and concentration dependent. Such a finding is consistent with numerous studies, which demonstrated that sorption by natural solids requires times ranging from a few days to months for apparent equilibrium to be attained. Pignatello and Xing (1996) summarized the results of several slow sorption studies and suggested that times for sorption by soils and sediments to attain equilibrium may vary from days to years. To explain the slow sorption (and desorption) observations, they proposed several diffusion-type mechanisms within soil organic matter and in intraparticle pores of different scales (macro, meso, micro, and nano).

In a similar formulation to the hypothetical model of Gamble and Khan (1990), we assume that atrazine in the soil (or crop residue) is present in four phases (Fig. 1) . Here, C is the solute concentration in solution (mg L–1), Se is assumed to represent the amount retained on the equilibrium sites (mg kg–1) and has a low binding energy, Sk is the amount retained on the kinetic sites (mg kg–1) through strong interactions with the mulch residue, Si represents the amount retained irreversibly (mg kg–1), Ke is an equilibrium constant (dimensionless), and k1 and k2 (h–1) are the forward and backward reaction rate coefficients associated with the kinetic sites, respectively. The parameter k3 (h–1) is the irreversible rate coefficient associated with the kinetic sites. The multireaction model presented here is a version of the more comprehensive MRM as recently described by Ma and Selim (1997), and is based on the assumption that the number of sorption sites is not limited. The retention reactions associated with the MRM are expressed as:

[2]

[3]

[4]
and the total amount retained S is now defined as:

[5]
where the parameters n and m are the reaction orders associated with Se and Sk, respectively, {theta} is the soil water content (cm3 cm–3), {rho} is the soil bulk density (g cm–3), and t is time (h). The set of nonlinear Eq. [2], [3], and [4], subject to the appropriate initial conditions, was solved numerically based on finite-difference approximation. The numerical scheme is documented in FORTRAN code MRM, which is available from the authors.



View larger version (6K):
[in this window]
[in a new window]
  		Fig. 1. A schematic of the proposed multireaction model (MRM) with equilibrium and kinetic adsorption sites and one first-order degradation (irreversible) reaction. The term C is concentration in soil solution, Se represents the amount sorbed on the equilibrium sites, Sk is that sorbed kinetic sites, and Si the amount retained irreversibly, where Ke, k1, k2, and k3 are the respective rates of reactions.

 

   MATERIALS AND METHODS
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
 MODEL FORMULATION
 MATERIALS AND METHODS
RESULTS AND DISCUSSION
 CONCLUSIONS
 REFERENCES
 
Sorption–Desorption
A bulk sample of sugarcane residue from first stubble (sugarcane variety LCP85-384) was collected from a private farm located about 8 km south of Baton Rouge (on 16 Apr. 1999) before spring application of herbicides. The soil was a Commerce loamy soil with the following properties: pH = 5.93, organic carbon = 0.65%, sand = 30%, silt = 54%, clay = 16%, and cation exchange capacity = 165 cmolc kg–1. The amount of mulch residue cover in the field was quantified based on eight replications as 4840 ± 423 kg ha–1 and 5 cm in thickness. This site was chosen to evaluate several best management practices (BMPs) including mulch management and quantify the effect of mulch on herbicide retention and runoff losses (for details see Selim et al., 2003). The mulch residue used in our atrazine retention study was dried at 55°C for 24 h and then cut into 1-cm-long sections and stored at 5°C.

Atrazine sorption by the sugarcane mulch residue was performed using a batch equilibration technique (Zhu and Selim, 2000). Radioactive atrazine was used as a tracer to monitor the extent of retention. Six 14C-atrazine spiked solutions having initial concentrations (Ci) of 3.37, 6.36, 12.34, 18.22, 24.30, and 30.16 mg L–1 in 0.005 M CaCl2 background solution were used. Three replicates were used for each initial concentration. Sorption was initiated by mixing 1 g of dried and cut sugarcane residue with 30 mL of the various atrazine concentration solutions in a 40-mL Teflon centrifuge tube. The mixtures were shaken and centrifuged at 500 x g for 10 min for each specific reaction time before sampling. A 0.5-mL aliquot was sampled from the supernatant at reaction times of 2, 8, 24, 48, 96, 192, 288, and 504 h. The mixtures were returned to the shaker after each sampling. The collected samples were analyzed using a liquid scintillation analyzer (TRI-CARB 2100TR; Packard Instrument Company, Meriden, CT) by mixing the 0.5-mL aliquot with 5 mL of Ultima-Gold-LS cocktail. The amount of atrazine adsorbed by the sugarcane residue was calculated from the difference between concentrations of the supernatant and initial solutions.

Atrazine desorption commenced immediately after the last sorption time step (504 h). Each desorption step was conducted by replacing the supernatant with atrazine-free 0.005 M CaCl2 background solution and shaking for 24 h. Six desorption steps were performed with a total desorption time of 6 d. Following the sixth step, one further extraction using pure methanol was performed. Atrazine in the supernatant solution during desorption was analyzed using liquid scintillation and the amount of atrazine desorbed from the mulch residue was calculated based on the change of atrazine concentration in solution (before and after desorption).

In a separate experiment, atrazine retention by the Commerce surface soil where the sugarcane was grown was measured in a similar fashion as that for the mulch residue. The only exception was that in the batch study the soil-to-atrazine solution used was 10 g of air-dried soil (in triplicates) mixed with 30 mL of atrazine solution. Six input or initial atrazine concentrations Ci values (1.80, 2.5, 5.4, 10.3, 20.2, and 30.0 mg L–1 in 0.005 M CaCl2 solution) were used. Moreover, the reactions times were 2, 6, 12, 24, 48, 96, 192, and 384 h. Details on atrazine desorption following sorption by the Commerce soil are given in Selim (2003).

Data Analysis
The isotherms of atrazine sorption and desorption by sugarcane mulch residue were used to estimate Freundlich parameters (i.e., distribution coefficient Kf and nonlinear dimensionless parameter N). Proc NLIN in SAS Version 8 (SAS Institute, 1999) was used to carry out the nonlinear regression of the Freundlich equation. Means and 95% confidence intervals were computed for regression coefficients derived from the Freundlich equation. Desorption isotherm coefficients were also estimated using nonlinear regression. The kinetic retention of atrazine was described using the kinetic multireaction model where the system of Eq. [2], [3], and [4] was solved using a finite-difference (explicit-implicit) iteration method and model parameters were estimated based on nonlinear least-square optimization.


   RESULTS AND DISCUSSION
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
 MODEL FORMULATION
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
CONCLUSIONS
 REFERENCES
 
Sorption
The amount of atrazine in soil solution versus that retained by the mulch residue is presented in Fig. 2 . These results are for the various reaction times used and are often referred to as sorption isotherms. Retention results for atrazine by the mulch residue were well described using the linear model (S = KdC). This is clearly depicted by the shape of the isotherms shown in Fig. 2. Estimated values for the distribution coefficient (Kd) are given in Table 1 and exhibit an increase with reaction time (from 10.4 to 23.40 L kg–1 after 2 and 504 h, respectively). The change of Kd values for the mulch residue versus time is shown in Fig. 3 . We should emphasize also that the Freundlich model (Eq. [1]) provided equally good description of atrazine sorption where the nonlinear parameter N varied between 0.91 and 0.97 with an average value of 0.943 (Table 1). Recent studies by Lesan and Bhandari (2003) indicated that N did not exhibit a consistent trend in a 12-wk sorption experiment for two soils. They reported a range of 0.90 to 0.94, which is in line with our results. Based on sorption results given in Fig. 1, one may regard atrazine sorption by the sugarcane residue as essentially linear for all reaction times. The lack of nonlinear or concentration-dependent behavior of the sorption patterns is indicative of the lack of heterogeneity of sorption-site energies. Rather, sorption-site energies for the linear sorption process of atrazine by mulch residue are relatively homogenous. Since N did not differ greatly from unity, the Kd parameter is more meaningful for comparison at different reaction times as well as with soils or other residues. As expected, Kf exhibited significant dependence on the time of reaction in a similar fashion to Kd. Lesan and Bhandari (2003) results showed strong kinetic dependence of atrazine sorption with at least a doubling of Kf values over a 12-wk sorption for two surface soils.



View larger version (21K):
[in this window]
[in a new window]
  		Fig. 2. Adsorption isotherms of atrazine by sugarcane mulch residue at different reaction times. The solid lines are predictions using a linear model.

 

View this table:
[in this window]
[in a new window]
  		Table 1. Estimated parameters with one standard deviation for atrazine adsorption by sugarcane mulch residue at different reaction times.

 


View larger version (17K):
[in this window]
[in a new window]
  		Fig. 3. Measured atrazine distribution coefficient (Kd) versus adsorption reaction time for sugarcane mulch residue and Commerce soil. Error bars represent one standard deviation.

 
Values for mulch residue Kd given in Table 1 were an order of magnitude higher than those found for the soil matrix of Commerce silt loam soil where our sugarcane crop was grown. This was expected since organic matter is the principal soil component affecting the sorption of many herbicides in the soil environment. The batch kinetic procedure was used to generate the isotherms for Commerce soil within the atrazine concentration range in a similar fashion as that used for the mulch residue. The only exception was that the soil to solution ratio, for the different Ci values, was 1 to 3. The sorption results are clearly illustrated when we compare our sorption isotherms for the soil matrix given in Fig. 4 with those for the mulch residue of Fig. 2. The Kd values for Commerce soil matrix ranged from 2.10 to 2.35 cm3 g–1 after 24 and 384 h of reaction time, respectively. Based on an organic carbon content of 0.65% for our Commerce soil, these Kd values for atrazine transform to Koc values of 320 to 258 cm3 g–1 after 24 and 384 h of reaction time, respectively. Based on literature values, Wauchope et al. (1992) reported an average soil Koc of 100 cm3 g–1 for atrazine.



View larger version (20K):
[in this window]
[in a new window]
  		Fig. 4. Adsorption isotherms for atrazine by Commerce soil for 2-, 24-, and 384-h reaction times. Solid lines are predictions using a linear model.

 
As illustrated in Fig. 4, the Commerce soil exhibited limited kinetic sorption behavior, which is in contrast with the extensive kinetics for atrazine retention by the mulch residue. A Kd for mulch residue (23.40 cm3 g–1) was an order of magnitude higher than for the Commerce soil (2.35 cm3 g–1). Other authors measured strong sorption of atrazine by the mulch residue from other crops. For example, Abdelhafid et al. (2000) measured high Kd values for wheat straw compared with soil (15.01 versus 0.77 cm3 g–1, respectively). Their results were based on a 24-h equilibration. We are not aware of earlier studies where atrazine Kd by sugarcane mulch was measured. Higher atrazine retention by the mulch residue in comparison with mineral soils was recently observed in field experiments. For example, Selim et al. (2003) found that the amount of extractable atrazine was 10 to 20 times higher for the mulch residue compared with that retained by the underlying (Commerce silt loam) surface soil layer. Therefore, we conclude that results from our laboratory study of the retention kinetics of the mulch residue were consistent with field measurements.

The strong atrazine retention by the mulch residue as indicated by the Kd values of Table 1 is highly beneficial in minimizing potential runoff and downward movement in the soil. In fact, Selim et al. (2003) found that when the sugarcane residue was not removed from the soil surface, a reduction in runoff-effluent concentrations, as much as 50%, for atrazine and pendimethalin was realized. They also reported that field investigations during the 1999 and 2000 growing seasons showed that the presence of mulch residue resulted in consistently lower estimates for rates of decay or disappearance of atrazine and pendimethalin in the surface soil.

Kinetics
The Kd as well as Kf parameter estimates given in Table 1 indicate a time-dependent behavior of atrazine sorption by the mulch residue. Therefore, the use of the MRM to describe such time-dependent behavior for the mulch is justified. Based on model simulations shown in Fig. 5 and 6 , the time-dependent behavior of atrazine by the mulch was well described by our MRM. The sorption patterns indicate an initial fast sorption that occurred within the first 24 h. This was followed by slow reactions that appear to be the dominant processes. This assessment is consistent with increased Kd during sorption.



View larger version (26K):
[in this window]
[in a new window]
  		Fig. 5. Atrazine concentration in solution during adsorption and desorption versus time. Results are from the batch kinetic experiment having a mulch to solution ratio of 1:30 and for several initial concentrations (Ci). The curves are multireaction model (MRM) predictions using the overall set of model parameters given in Table 2.

 


View larger version (23K):
[in this window]
[in a new window]
  		Fig. 6. Amount of atrazine sorbed by sugarcane mulch residue. Results are from the batch kinetic experiment having a mulch to solution ratio of 1:30 and for several initial concentrations (Ci). The curves are multireaction model (MRM) predictions using the overall set of model parameters given in Table 2.

 

View this table:
[in this window]
[in a new window]
  		Table 2. Model parameter estimates, standard error (SE), root mean square errors (RMSE), and coefficient of correlation (r2) for atrazine retention in sugarcane mulch residue for different initial (input) atrazine concentrations (Ci). Model parameters (Ke, k1, k2, and k3) are given for adsorption and adsorption–desorption data separately.

 
The goodness-of-fit of the MRM was tested using our sorption–desorption data set for all initial concentrations. In all MRM model calculations, we assumed the empirical coefficients n and m to be n = m (see Eq. [2] and [3]). This assumption was used since there is no known method for estimating n or m independently [for details see Ma and Selim (1997) and Selim (2003)]. The parameter value used was based on the average of the Freundlich N (i.e., n = m = 0.94). The MRM parameter estimates given in Table 2 were obtained using nonlinear least square optimization for each initial concentration Ci (3.37, 6.36, 12.34, 18.22, 24.30 and 30.16 mg L–1). Moreover, we obtained one set of parameters for the MRM where the entire data set for all Ci values was used in the nonlinear least-square parameter estimation procedure. As a result, a set of model parameter estimates, hereafter referred to as the overall set of parameters, and parameters corresponding to each Ci, were obtained (Table 2). The kinetics of atrazine sorption and desorption for the range of concentrations and time were well described by the MRM. This is clearly shown by the solid curves, which represent MRM predictions based on the overall set of model parameters (see Fig. 5 and 6).

To further test the capability of the MRM to describe atrazine retention, we used the MRM in a descriptive or simulation mode where the necessary parameters were based on both sorption and desorption results (i.e., the entire batch results as one data set). The resulting parameter estimates for different Ci values are given in Table 2. Based on the goodness of fit and their statistics, sorption as well as desorption results were well described by the MRM as illustrated by the low parameter standard errors and high r2 values. Moreover, parameter estimates listed in Table 2 were not significantly different when data sets from individual concentrations (Ci) were used versus the overall data (i.e., all applied concentrations; p > 0.05). Moreover, parameter estimates based on sorption data alone were not significantly different from those based on our sorption–desorption results (all p > 0.05). Thus, we conclude that one set of model parameters is capable of describing the entire data set. Thus, the model is independent of concentration and is applicable in describing not only sorption but also desorption reactions by the mulch residue (Table 2).

To test whether a simpler version of the model adequately describes our batch results, we used the model version where only reversible reactions were included. Therefore, we did not invoke irreversible sorption (i.e., k3 was set to zero) and new parameter estimates were obtained (not shown). Much higher root mean square error (RMSE = 0.3838) was obtained in the absence of an irreversible reaction compared with RMSE of 0.1114 when k3 was included in the model. Moreover, the extra sum of squares principal was used to determine if there was statistically significant improvement in the fit of the model to the data by adding more parameters to the original model (see Ma and Selim, 1997). The F test indicated that, at the 0.05 confidence level, the model with irreversible sorption provided superior fitting in describing our batch results compared with the simplified model version. As a result, we conclude that the use of fully reversible reactions (equilibrium plus kinetic) alone provides less than adequate description of the data. Therefore, irreversible sorption can account in part for the observed hysteresis as discussed in the section below. Ma and Selim (1997) arrived at a similar conclusion when various MRM versions having different numbers of parameters were used in describing atrazine retention and mobility in an aggregated soil where chemical as well as physical equilibrium models were examined.

Although a comparison of parameter estimates based on MRM, given in Table 2, and those based on equilibrium (linear and Freundlich) models, given in Table 1, is not explicitly applicable, several consistent trends need to be emphasized. Retention is initially fast as indicated by the reactions associated with the equilibrium sites (Se) and represented by Ke, whereas for kinetic reactions associated with Sk and Si, the magnitude of the rate coefficients (k1 and k2) compared with k3 depicts kinetic or slow reactions that continued following the initially fast sorption, which occurred within the first few hours. For large times, (t -> {infty}) values for Kf or Kd can be approximated based on the contributions from the fast reactions approximated by ({theta}/{rho}) Ke and those of the kinetic-type reactions ({theta}/{rho}) k1/k2. Based on our overall set of parameters of Table 2, our approximation yields a value of 20 (dimensionless), which is within the range of Kd given in Table 1. Moreover, based on the overall set of parameters, 50% of the sites of the mulch residue are of the equilibrium type whereas the remaining are kinetically dominant. This finding is contrary to that for atrazine retention by the Commerce soil, where nearly all the reactions were of the equilibrium type and where kinetics was lacking (Selim, 2003).

Hysteresis and Atrazine Recovery
The amount of atrazine desorbed at the end of six successive desorption steps was quantified. In addition, we calculated the amount of atrazine released subsequently using methanol as the final extraction step. These results, along with the amount retained after 504 h of sorption, are given in Table 3. Based on six successive desorption steps, the total amount of atrazine desorbed, as a percentage of total adsorbed after 504 h retention, was 58.18 ± 4.36% over the entire range of atrazine initial input concentrations. The recovery of atrazine for the lowest Ci (3.37 mg L–1) was significantly different from the recoveries for all other Ci values with a p value of 0.005. No significant difference in atrazine recovery was observed for all other Ci values when the lowest Ci was excluded (p = 0.239). The amount of atrazine desorbed represents the sum of water-soluble or readily desorbable atrazine fractions. As the initial input Ci increased from 3.37 to 30.16 mg L–1, the percentage of desorbed atrazine based on the total amount adsorbed increased from 51.29 to 60.91%.


View this table:
[in this window]
[in a new window]
  		Table 3. Amount of atrazine adsorbed and desorbed by 0.005 M CaCl2 or pure methanol and atrazine recovery.

 
Seybold and Mersie (1996) reported an average recovery, following four 24-h desorption steps, of 43 ± 4.3% for atrazine from an Emporia loam (fine-loamy, siliceous, subactive, thermic Typic Hapludults) and Cullen clay (very-fine, kaolinitic, thermic Typic Hapludults). The measured Kd values following 24 h sorption were 1.46 and 1.82 cm3 g–1, and the respective Koc values were 234 and 140 cm3 g–1, for Emporia and Cullen soils, respectively. The organic C content was reported as 0.63 and 1.3% for the Emporia and Cullen soils, respectively. Seybold and Mersie (1996) concluded that no significant differences of atrazine recoveries were observed between their two soils. In their study, the background solution (CaCl2) used was similar to that used in this study. However, unlike our study, Seybold and Mersie (1996) limited the time for sorption to only 24 h. We postulate that the higher release of atrazine from the mulch compared with the Commerce soil is probably due to the higher amount of atrazine sorbed on kinetic rather than instantaneous equilibrium sites as described by the MRM.

Our desorption results in Fig. 6 exhibit several features. During the first two desorption steps, a major amount of atrazine was desorbed from the mulch residue, and less atrazine was released during subsequent desorption as the desorption steps increased (see Fig. 6). Desorbed atrazine during the first few hours is likely to come from the most accessible sites and/or from the less-energy-release sorption mechanisms, whereas atrazine sorbed on less accessible sites and/or more strongly adsorbed (high-energy) sites is not susceptible to desorption initially and is subsequently subject to slow release over time. The availability of pesticide residues in soils can be evaluated in relation to their extractability (Moreau and Mouvet, 1997, 1998). For instance, all solvent-extractable residues are potentially available and water-extractable residues are the most readily available phase. Atrazine extractable with methanol is probably less readily available and could possibly be extractable with water if the extraction procedure was repeated enough times. Our results indicate that the final extraction step using pure methanol resulted in an average atrazine recovery of only 5.49 ± 0.69% of applied atrazine (Table 3). The amount of nonextractable (i.e., strongly held) atrazine was related to the concentration initially bound to the solid and the energy level of the sorption mechanisms (Moreau and Mouvet, 1998). The higher atrazine recovery from our sugarcane mulch residue than from soils is indicative of the higher availability of atrazine retained by the mulch residue.

Sorption–desorption results of atrazine by the mulch residue are presented as isotherms in the traditional manner in Fig. 7 . These isotherms, or more specifically C versus S at several times, clearly indicate considerable hysteresis. This hysteretic behavior resulting from discrepancy between sorption and desorption isotherms was not surprising in view of the kinetic retention behavior of atrazine by the mulch residue. Selim et al. (1976) showed that observed hysteresis in batch experiments may be explained by using a two-site equilibrium-kinetic model. They showed that lack of equilibrium conditions may be responsible for observed desorption hysteresis. The likely time-dependent nature of hysteresis of atrazine sorption–desorption in soils was reported by Ma et al. (1993) and Moreau and Mouvet (1998). Such kinetic behavior may partially explain the discrepancy between measured data and predictions based on the Freundlich model shown by the solid curves in Fig. 7. Based on one-step (24-h) desorption, Lesan and Bhandari (2003) showed that the extent of hysteresis was a function of the reaction time for sorption. Such results suggest strong non-equilibrium retention where for short contact times, desorption isotherms exhibited little deviation from those for sorption. Lesan and Bhandari (2003) did not carry out additional desorption steps beyond 24 h.



View larger version (30K):
[in this window]
[in a new window]
  		Fig. 7. Traditional desorption isotherms of atrazine by sugarcane mulch residue. The solid line is the adsorption isotherm for the 504-h reaction. The dashed curves are predictions using the (a) multireaction model (MRM) and (b) Freundlich model.

 
In addition to the traditional isotherms (see Fig. 7), we also present desorption results based on the time of reaction as described by Zhu and Selim (2000). Specifically, atrazine desorption results for the mulch residue are now presented for each time of reaction (sorption and desorption) and are shown in Fig. 8 . As a result, we have a family of desorption isotherms that are presented in a similar manner to those for sorption. In this way, we maintained the overall isotherm definition of C versus S for a given reaction time of sorption or sorption and desorption. Therefore, this family of desorption isotherms is shown here for each desorption time (or step), and can be referred to as time-dependent desorption isotherms. If hysteresis is absent, sorption and desorption results should coalesce, and the dashed and solid curves should be indistinguishable (Zhu and Selim, 2000). One can also regard such a family of curves as the kinetic type since each curve represents desorption (release) at a given desorption time following sorption.



View larger version (26K):
[in this window]
[in a new window]
  		Fig. 8. Time-dependent desorption isotherms of atrazine by sugarcane mulch residue. The solid line is the adsorption isotherm for the 504-h reaction. The dashed curves are predictions using the (a) multireaction model (MRM) and (b) Freundlich model.

 
The Freundlich model (Eq. [1]) was also used to describe our atrazine desorption data based on the traditional isotherm approach (Fig. 7) and those based on our time-dependent isotherms (Fig. 8). As a result, two sets of values for best estimates of Kf and N were obtained and are given in Table 4. For the traditional desorption isotherms (see Fig. 7), Kf values were consistently higher than those associated with sorption isotherms (see Table 1). The opposite trend was observed for the nonlinear parameter N. Moreover, Kf for desorption was significantly dependent on initial atrazine concentration Ci. Similar findings for atrazine in soils were previously reported (Clay and Koskinen, 1990; Ma et al., 1993). Moreover, this comparison is further complicated since exceedingly small N values for desorption were estimated (0.197–0.267). For the time-dependent isotherms (see Fig. 8), the estimated N values were similar to those for sorption (Tables 1 and 4), which is consistent with results for metolachlor sorption–desorption by a Sharkey clay (very-fine, smectitic, thermic Chromic Epiaquerts) (Zhu and Selim, 2000). Although Kd values based on 24-h equilibration are widely reported, our retention results indicate that parameter estimates based on time-dependent sorption–desorption isotherms are more appropriate to predict herbicide residue in the field, and thus more meaningful in environmental assessment.


View this table:
[in this window]
[in a new window]
  		Table 4. Freundlich parameters for atrazine desorption based on the traditional and time-dependent isotherms for sugarcane mulch residue and the values of hysteresis coefficients.

 
The extent of hysteresis was quantified using the hysteresis coefficients {omega} (Ma et al., 1993) and H (Cox et al., 1997). These coefficients were defined based on the discrepancy between the sorption and desorption isotherms, and calculated using Freundlich parameters estimated from sorption and desorption isotherms separately. They can be expressed as:

[6]

[7]
where Na and Nd are Freundlich N for sorption and desorption, respectively. The coefficient H is a simple one and easy to use, whereas {omega} is only applicable for the traditional type isotherms of successive desorption (shown in Fig. 6). Recently, Zhu and Selim (2000) proposed an alternative hysteresis coefficient {lambda} based on the difference in the areas between sorption and desorption isotherms. Zhu and Selim (2000) derived the following expression for the parameter {lambda} for the traditional isotherms (Fig. 7) as:

[8]

For time-dependent desorption isotherms (see Fig. 8), the expression for {lambda} was given as:

[9]
where Na and Nd for Eq. [8] are the same as in Eq. [6] and [7], whereas (Kf)d and (Kf)a are the Freundlich parameters for the time-dependent sorption–desorption isotherms. Based on the above formulations, we obtained values for {lambda}, {omega}, and H, which are presented in Table 4. For the traditional isotherms, values for {lambda} and {omega} decreased as Ci increased, whereas the opposite was observed for H value. This is similar to metolachlor sorption–desorption in Sharkey clay soil as reported by Zhu and Selim (2000). Ma et al. (1993) calculated {omega} for atrazine on Sharkey clay soil and indicated that {omega} increased linearly with incubation time, which is the time interval between the end of sorption and the beginning of the desorption process. However, they did not observe an effect of Ci on {omega}. Seybold and Mersie (1996) calculated {omega} for metolachlor in two soils and found that {omega} is Ci dependent for Cullen soil, which contained 31% clay and 1.3% organic carbon, but this phenomena was not apparent in Emporia soil, which contains less clay and less organic carbon. Thus, the dependency of {omega} on Ci in our study is consistent with Seybold and Mersie (1996) results. For the time-dependent desorption isotherms shown in Fig. 8, values for {lambda} increased with desorption time, which is indicative of dependency on the desorption history. Such behavior might be explained by the existence of irreversible reactions, which cause a decrease in desorbed herbicide amounts as desorption time increased.


   CONCLUSIONS
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
 MODEL FORMULATION
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 CONCLUSIONS
REFERENCES
 
We investigated the characteristics of sugarcane mulch residue for atrazine sorption–desorption. A sorption kinetic-batch method was used to quantify retention of mulch for a wide range of atrazine concentrations and reaction times. Sorption as well as desorption of atrazine by the mulch residue exhibited extensive kinetic behavior. Sorption isotherms appeared linear for all retention times with Kd values increasing from 10.40 to 23.4 cm3 g–1 after 2 and 504 h, respectively. Values for mulch residue Kd were an order of magnitude higher than those found for Commerce silt loam soil where the sugarcane crop was grown. Such strong atrazine retention by the mulch residue is highly beneficial in minimizing potential runoff and downward movement in the soil. Use of an equilibrium-kinetic multireaction model was successful in describing sorption results for the entire range of concentrations. Desorption exhibited strong hysteresis for all input concentrations and was equally well-described by the MRM. Moreover, one set of model parameters estimated from the entire data set including both sorption and desorption results for the entire range of initial input concentrations was adequate in describing the batch results. Recovery of applied atrazine based on six successive desorptions and subsequent extraction with methanol accounted for 63.67 ± 4.38% of that adsorbed over the entire input concentration range. Average atrazine recovery for the different initial atrazine input concentrations was found to be concentration independent except for the lowest Ci (3.37 mg L–1). Moreover, a hysteresis coefficient based on the difference in the area between sorption and desorption isotherms was capable of quantifying hysteresis of desorption isotherms.


   ACKNOWLEDGMENTS
 
This study was funded in part by the Nonpoint Source Program (Section 319), Louisiana Department of Environmental Quality (Contract no. 24400-93-32), Jan Boydstun, Program Director. Special thanks to Bayer Corporation for providing radio-labeled atrazine isotope.


   NOTES
TOP
 NOTES
ABSTRACT
 INTRODUCTION
 MODEL FORMULATION
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 CONCLUSIONS
 REFERENCES
 
Approved by the director of the Louisiana Agricultural Experiment Station as Manuscript no. 03-14-1562.


   REFERENCES
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
 MODEL FORMULATION
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 CONCLUSIONS
 REFERENCES
 


Related articles in JEQ:

This Issue in Journal of Environmental Quality

JEQ 2005 34: 1-6. [Full Text]  




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 Related articles in JEQ
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Citing Articles
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 Zhu, H.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Selim, H. M.
Right arrow Articles by Zhu, H.
Agricola
Right arrow Articles by Selim, H. M.
Right arrow Articles by Zhu, H.
Related Collections
Right arrow Organic Compounds
Right arrow Agricultural Pesticides

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
The SCI Journals