Modeling Physical and Chemical Nonequilibrium Transport of

Louisiana State UniversityLSU Digital Commons

LSU Historical Dissertations and Theses Graduate School

1996

Modeling Physical and Chemical NonequilibriumTransport of Herbicide in Soils From DifferentTillage Systems.Shikui XueLouisiana State University and Agricultural & Mechanical College

Follow this and additional works at: https://digitalcommons.lsu.edu/gradschool_disstheses

This Dissertation is brought to you for free and open access by the Graduate School at LSU Digital Commons. It has been accepted for inclusion inLSU Historical Dissertations and Theses by an authorized administrator of LSU Digital Commons. For more information, please [email protected].

Recommended CitationXue, Shikui, "Modeling Physical and Chemical Nonequilibrium Transport of Herbicide in Soils From Different Tillage Systems."(1996). LSU Historical Dissertations and Theses. 6225.https://digitalcommons.lsu.edu/gradschool_disstheses/6225

https://digitalcommons.lsu.edu?utm_source=digitalcommons.lsu.edu%2Fgradschool_disstheses%2F6225&utm_medium=PDF&utm_campaign=PDFCoverPages

https://digitalcommons.lsu.edu/gradschool_disstheses?utm_source=digitalcommons.lsu.edu%2Fgradschool_disstheses%2F6225&utm_medium=PDF&utm_campaign=PDFCoverPages

https://digitalcommons.lsu.edu/gradschool?utm_source=digitalcommons.lsu.edu%2Fgradschool_disstheses%2F6225&utm_medium=PDF&utm_campaign=PDFCoverPages

https://digitalcommons.lsu.edu/gradschool_disstheses?utm_source=digitalcommons.lsu.edu%2Fgradschool_disstheses%2F6225&utm_medium=PDF&utm_campaign=PDFCoverPages

https://digitalcommons.lsu.edu/gradschool_disstheses/6225?utm_source=digitalcommons.lsu.edu%2Fgradschool_disstheses%2F6225&utm_medium=PDF&utm_campaign=PDFCoverPages

mailto:[email protected]

MODELING PHYSICAL AND CHEMICAL NONEQUILIBRIUM TRANSPORT OF

HERBICIDE IN SOILS FROM DIFFERENT TILLAGE SYSTEMS

A Dissertation

Submitted to the Graduate School of Louisiana State University and

Agricultural and Mechanical College in partial fulfillment of the

requirements for the degree of Doctor of Philosophy

in

The Department of Agronomy

ByShikui Xue

B.S. Huazhong Agricultural University, P.R. China 1984 M.S. Chinese Academy of Sciences, 1987

May, 1996

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

ACKNOWLEDGEMENTS

I would like to take this opportunity to thank the many people who gave advice

and assistance while I was working on my Ph.D program. First, I would like to express

my gratitude to my major professor, Dr. H. M. Selim for generously providing

guidance by sharing his time, ideas, and experience with me throughout my graduate

program.

In addition, I am grateful to Dr. Liwang Ma for his suggestions throughout my

research, and to Mr. D. C. Johnson and Steven L. McGowen for their assistance in the

field and laboratory.

I would also like to acknowledge Dr. Glenn V. Wilson from the University of

Tennessee for his many invaluable suggestions, Dr. L.M . Southwick for providing

standard compounds used in this study, Dr. Robert L. Hutchinson for providing the

tillage experiment plots used in this study, and Dr. C. W. Lindau for providing the

materials used for quick freezing o f soil columns.

Finally, I wish to thank the members of my graduate committee: Dr. R. P.

Gambrell, Dr. Richard M. Johnson, Dr. K. T. Valsaraj and Dr. David J. Longstreth

for their advice and suggestions during my graduate program.

Last, but not the least, I appreciate the convenience provided by all faculty and

staff members o f the Department of Agronomy throughout my program.

Funding for this research was supported in part by a USDA-NRI grant under the

supervision o f Professor Selim.

ii


TABLE OF CONTENTS

ACKNOWLEDGEMENTS...........................................................................................ii

ABSTRACT..................................................................................................................vi

CHAPTER1 A REVIEW OF HERBICIDE BEHAVIOR IN SOILS

OF DIFFERENT TILLAGE SYSTEMS.................................................... 11.1 Introduction.................................................................................. 11.2 Herbicide Transport in S o ils .................................................... 4

1.2.1 Retention M echanism ......................................................... 51.2.2 Equilibrium Retention Models ........................................ 61.2.3 Kinetic Retention Models ................................................ 71.2.4 Factors Affecting Retention R eactions........................... 91.2.3 Hydrodynamic D isp e rs io n .................................................101.2.6 Preferential Transport .........................................................131.2.7 Biological D ecom position ...................................................14

1.3 Modeling the Water and Herbicide Transport.........................161.3.1 Modeling Water Flow .........................................................161.3.2 Modeling Herbicide T ra n s p o r t ..........................................18

1.4 Alachlor Behavior in S o i l ........................................................... 231.5 Current Experimental Methods on the Fate

and Transport of H erbicide................................................... 251.5.1 Retention and Mobility Studies o f H e rb ic id e s 251.5.2 Tracer Studies for Characterizing Preferential Flow . 261.5.3 Herbicide Transport under Unsaturated F l o w 29

1.6 O bjectives...................................................................................... 31

2 MODELING ADSORPTION-DESORPTION KINETICS OF ALACHLOR IN SOILS OF DIFFERENTTILLAGE SYSTEMS ................................................................................... 33

2.1 Introduction................................................................................... 332.2 M ethods.........................................................................................36

2.2.1 S o il ............................................................................................362.2.2 Batch E x p erim en t................................................................. 362.2.3 Multiple Reaction Modeling

and Statistical A n a ly s is ..................................................... 372.3 Results and Discussions .............................................................. 38

2.3.1 Adsorption-Desorption K inetics......................................... 382.3.2 Alachlor H y ste res is .............................................................. 522.3.3 V alidation................................................................................55


3 PREFERENTIAL FLOW IN SOILS FROM NO-TILL ANDCONVENTIONAL TILLAGE SYSTEMS................................................... 59

3.1 Introduction................................................................................... 593.2 Materials and M ethods.................................................................63

3.2.1 Transport E xperim en ts........................................................ 633.2.2 Breakthrough Curve A n a ly s is ............................................ 653.2.3 Superposition o f Short Pulse ............................................ 66

3.3 Results and Discussions .............................................................. 673.3.1 The Flow Paths in Undisturbed C o lu m n s....................... 683.3.2 BTC Characteristics o f T ritiu m ......................................... 803.3.3 Modeling BTCs of Tritium ............................................... 84

3.4 Conclusions....................................................................................85

4 MODELING ALACHLOR TRANSPORT IN SATURATED SOILS FROM NO-TILL ANDCONVENTIONAL TILLAGE SYSTEMS................................................... 87

4.1 Introduction....................................................................................874.2 Materials and M ethods.................................................................90

4.2.1 Batch Experiments .............................................................. 904.2.2 Miscible Displacement E xperim en ts................................ 904.2.3 Equilibrium Retention M o d e lin g ...................................... 914.2.4 Multireaction Modeling ..................................................... 934.2.5 Multireaction Transport M o d e lin g ................................... 93

4.3 Results and Discussions ..............................................................954.3.1 Soil P ro p e rtie s .......................................................................954.3.2 Modeling Alachlor A d so rp tio n ......................................... 954.3.3 The tillage Effect on Alachlor A d so rp tio n .................... 994.3.4 Hydrodynamic Dispersion C o effic ien ts .........................1054.3.5 Alachlor BTCs in Soils

from Different Tillage S y s te m s .....................................1074.3.6 Transport M odeling ............................................................ 1084.3.7 Retention Parameters o f Alachlor During

Transport in Undisturbed Soil C o lu m n s .................. 1124.3.8 Alachlor Recovery, Degradation

and Irreversible R eac tio n ................................................ 1224.3 .9 Tillage Effect on Alachlor Transport ............................123

4.4 Conclusions..................................................................................124

5 SUMMARY AND CONCLUSIONS............................................................ 126

REFERENCES . . .................................................................. 132

iv


APPENDIX: LETTER OF PERMISSION . . . . . . . 154

VITA . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 155

v


ABSTRACT

The physical and chemical nonequilibrium transport o f alachlor were studied in

a surface Gigger soil from different tillages through tracer studies, and batch and

miscible displacement experiments. Batch experiments indicated initially fast reaction

followed by slow adsorption. Adsorption and desorption results indicated time

dependent hysteretic behavior and was best described by a multireaction model

incorporating nonlinear equilibrium reaction, a reversible kinetic mechanism, and a

consecutive irreversible mechanism. The model predicted alachlor hysteresis and

adsorption-desorption kinetics satisfactorily based on parameters obtained from

adsorption experiments.

Tracer (Eosin Y and Blue FCF dyes) studies showed non-uniformly stained areas

in undisturbed soil cores (6.4 cm i.d, 15 cm length) and indicated more pronounced

preferential flow and physical nonequilibrium solute transport in no-till than in

conventional tillage. Tritium breakthrough curves (BTCs) indicated earlier breakthrough

associated with bimodal peaks in short pulses for no-till. The shape o f BTCs were also

dependent on flow direction. The superimposed experimental data from short pulses

well predicted the data of long pulses. The classical convective-dispersive equation was

inadequate and there was no improvement in describing tritium BTCs using physical

nonequilibrium models (mobile-immobile and stochastic models) for soils from no-till.

Miscible displacement results indicated that alachlor BTCs in soils of no-till

were more asymmetrical, with earlier breakthrough and longer tailing than soils from

conventional tillage. A multireaction transport model (MRTM) was not satisfactory for

vi


alachlor prediction using independently measured parameters from batch experiments.

However, MRTM successfully described alachlor BTCs in a calibration mode where

physical and chemical nonequilibrium were dominant. Best-fit parameters indicated the

dominance of kinetic reactions compared with parameters from batch experiments and

may be attributed to soil heterogeneity. Although no-till increased alachlor retention in

batch experiments, an overall estimation based on the sum o f kinetic and equilibrium

retention showed no significant influence on retention by tillage. High pressure liquid

chromatography (HPLC) chromatograms, fitted transport parameters, flow interruptions

and percent recoveries indicated a significant consecutive irreversible reaction in soils

o f conventional tillage. Moreover, no-till increased alachlor transport based on

breakthrough time compared with conventional tillage.

vii


CHAPTER 1

A REVIEW OF HERBICIDE BEHAVIOR IN SOILS FROM DIFFERENTTILLAGE SYSTEMS

1.1 Introduction

Conservation tillage practices are widely adopted in the United States. It was

estimated that by the year 2010, 95% of all cropland in the United States will be

farmed with conservation tillage practices (McWhorter, 1984). No-till is one of the

most common conservation tillages. With no-till, there is effective control o f erosion

and surface water runoff (Shelton et al., 1983), however, it increases amount of water

entering the soil and increases usage o f pesticides to control enhanced weed and insect

problems. In addition, no-till generally depends more on herbicides and less on

cultivation to control weeds than conventional tillage. More herbicide may be needed

also because surface residues intercept the spray (Crosson et al., 1986).

The effect of no-till on macropore development and the flow o f water and

chemicals through macropores has been shown to be significant (Tyler and Thomas,

1977; Edwards et al., 1988; Andreini and Steenhuis, 1990). There are various

viewpoints regarding the fate and transport of applied agricultural chemicals in different

tillage system. Because soil is left practically undisturbed under no-till, macropores

(cracks between aggregates and biological continuous macropores towards the

groundwater) increase the hydraulic conductivity (Mahboubi, et a l., 1993) thus increase

infiltration of water with decreased attenuation of dissolved chemicals due to low pore-

1


2

surface area. Under undisturbed flow conditions, there is a higher leaching under no-till

than conventional tillage (Gish, 1991; Dick et al., 1986; Hall et al., 1989; Isensee et

al., 1990). However, macropores may not be hydrologically active and transport of

herbicides through smaller pores may result in greater reactivity in no-till than

conventional tillage. Typically, in situ storm events are not of sufficient size or intensity

to cause saturated flow in the vadose zone, unless perched water table or downslope

convergent processes are prevalent (Wilson, et al., 1991). As a result, the no-till

practice may not enhance the leaching of herbicides. Fermanich and Daniel (1991)

reported greater carbofuran transport in tilled than in no-till in laboratory soil columns.

The significance o f unsaturated versus saturated flow conditions to preferential transport

of herbicides is not well documented.

Conservation tillage practices may alter the chemical and biological properties

of soils. Important soil changes take place with continuous no-till. Soil organic matter

content increases and its distribution in the soil profile is changed. Plow-tillage buries

organic debris, whereas that which accumulates on the surface of untilled soil may

reduce herbicide effectiveness. Increased soil organic matter content causes increased

adsorption of most herbicides, requiring increased rates of application. (Triplett and

Worsham, 1985). No-till also increases the cation exchange capacity (CEC) of the

surface soil compared to conventional tillage (Mahboubi, et al., 1993; Stearman et al.,

1989; Phillips and Phillips, 1984). Although the formation o f macropores may increase

the potential for herbicide leaching through the vadose zone, changes in the soil

chemical properties may enhance herbicide retention in the surface horizons of no-till


3

soils. This was demonstrated by Levanon et al., (1993) who reported higher

concentration of the herbicides occurred in the leachates of conventional column than

the no-till columns. Shallow soil samples have revealed that pH decreases up to 1 unit

after conservation tillage is practiced, due primarily to nitrogen fertilization (Blevin et

al., 1982). Moreover, increased organic matter content in no-till results in reduced

herbicide activity (i.e., concentration levels) and may result in greater herbicide

retention. However, dissolved soil organic carbon may compete with soil organic matter

for herbicide retention. This would result in increased herbicide solubility and mobility,

and decreased retention o f herbicide (Chiou and Manes, 1986; Chiou et a l., 1987) and

therefore alter the fate and transport of herbicides (Madhun et al., 1986; Lee and

Farmer, 1989; McCarthy and Zachara, 1989; Abdul et al. 1990; Chin et al., 1990; Lee

et al., 1990). Because dissolved organic matter is mobile in the soil environment

(Jardine and Jacobs, 1990), herbicides are transported to greater soil depths. How these

changes in chemical and biological properties due to tillage practices influence the

retention of herbicides is not well verified.

Research plots at the West Tennessee Experiment Station (Jackson, TN), were

established in 1981 under continuous no-tillage and conventional tillage cotton with

wheat and hairy vetch cover crops. A similar long-term study was initiated at LSU

Agricultural Center, Macron Ridge Research Station, Winnsboro, LA, in the Fall of

1986 (Hutchinson, et al., 1993) to evaluate the agronomic and economic feasibility of

alternative tillage systems and winter cover crops for cotton on a highly erodible loess

soil. Other goals o f this study were to identify soil, environmental, and biotic factors


4

that influence cotton response to alternative tillage systems and winter cover crops.

These plots include no-till and conventional tillage of cotton with wheat, crimson

clover, and hairy vetch and no winter cover crops. The preemergence herbicides were

used for controlling the weeds. These plots are ideal for studying long-term tillage

effects on herbicide transport through highly erosive soils common to the southern

region.

1.2 Herbicide Transport in Soils

The classical convection-dispersion equation that is generally acceptable for the

description o f dissolved chemicals in the soil solution is (Selim, 1992):

= ° - 1) St dt Sz r

where C is solute concentration in solution (pg m l'1), 6 is the soil water content (cm3

cm'3), p is the soil bulk density (g cm'3), D is the hydrodynamic dispersion coefficient

(cm2 h '1), and v is Darcy’s water flux density (cm h '1). In addition, S is the solute

concentration associated with the solid phase of the soil (pg g' 1 soil) and \{/l are the rates

of solute removal (or supply) from soil solution (pg cm '3 h '1) and are not included in

S. The (dS/dt) term is a fully reversible process between the solution and the solid

phases, and \J/t are irreversible (sinks or sources) rates of reactions, i.e ., transformation

reactions. Processes governing the interactions of individual solute species must be

identified if prediction of the fate o f contaminants in the soil using the convection-

dispersion equation (1.1) is sought. The transport equation indicates that chemical


5

retention, hydrodynamic dispersion, water flux and biological degradation may

influence the fate o f herbicides in soils of different tillage systems.

1.2.1 Retention Mechanism

Retention (i.e., sorption and desorption) reactions play a significant role in the

transport o f herbicides in different tillage systems. Sorption may be regarded as a

partitioning of the herbicide between the bulk solution phase and the solid phase (soil

matrix). (Greenland et al., 1965; Biggar and Cheung, 1973). The partition processes

is characterized by a distribution coefficient (K<,) given by

(1.2)a e y ec e

where as = activity of the adsorbed solute, a,, = activity of the solute in the

equilibrium solution, Cs = micrograms of solute adsorbed per milliliter of solvent in

contact with the adsorbent surface, Ce = micrograms of solute per milliliter of solvent

in equilibrium solution, 7 S= activity coefficient o f adsorbed solute, and = activity

coefficient of the solute in the equilibrium solution. It is assumed that K0 is related to

the standard differential Gibbs energy of adsorption (AG00) in the limit at infinitely

dilute adsorbate solution. Thus,

(1 3 )

c . - ° C , tte 0

when the 7 terms approach unity, we have AG00 =-RTlnK0. The value o f K0 is

obtained by plotting h^Cg/Cg) vs. Cs, and then extrapolating to zero Cs. The intercept


vary with time. Linear, Freundlich, and one-, and two-site Langmuir equations are

perhaps most commonly used to describe equilibrium reactions. Since a sorption

maximum is rarely observed, the Freundlich equation seems to be more applicable than

the Langmuir. At low herbicide concentrations, the sorption can be described by a

linear sorption isotherm that passes through the origin (Johnson and Farmer, 1993).

Models assuming an equilibrium partitioning are less complex, require fewer parameters

to be estimated, and their solution are more readily obtained. In certain cases, however,

the equilibrium model may sacrifice accuracy relative to more appropriate

nonequilibrium model. As a result, the first-order reversible kinetic reaction has been

extended to include the nonlinear kinetic type (Mansell et al., 1977).

1.2.3 Kinetic Retention Models

For some solutes, retention reactions in the soil solution have been observed to

be strongly time dependent (Selim, 1992). A number of empirical models have been

proposed to describe kinetic retention reactions of solutes in the solution phase. The

earliest model is the first-order kinetic reaction, which was first introduced into the

convective-dispersive transport equation (1.1) by Lapidus and Amundson (1952).

Multiple Reaction Models

Multisite or multireaction models deal with the multiple interactions of one

species in the soil environment. Such models are based on the assumption that a

fraction of the total sites are highly kinetic whereas the remaining fraction of sites

interact slowly or instantaneously with those in the soil solution. The earliest

multireaction models is the two-site model proposed by Selim et al. (1976). Such a two-


site approach proved successful in describing observed extensive tailing o f breakthrough

results o f herbicides.

A general multireaction model is presented by Selim (1992). In this model, the

solute was considered in the solution phase (C) and in five phases representing solute

retained by the soil matrix as Se, S 1} S2, S3, and Sirr, where Se, and S2 are in direct

contact with the solution phase and are governed by concurrent-type reactions. Se as the

amount o f solute that is sorbed reversibly and is in equilibrium with C at all times. The

governing equilibrium retention/release mechanism was that o f the nonlinear Freundlich

type. The retention/release reaction associated with Sj and S2 were considered to be

direct contact with C as the two site reaction. The multireaction model also considers

irreversible solute removal via a retention sink term Q in order to account for

irreversible reactions such as precipitation/dissolution, mineralization and

immobilization. The multireaction model also includes an additional retention phase (S3)

that is governed by a consecutive reaction with S2. This phase represents the amount

of solute strongly retained by the soil that reacts slowly and reversibly with S2 and may

be a result o f further rearrangements of the solute retained on matrix surfaces. The

reaction between S2 and S3 was considered to be of the kinetic first-order type.

Second-Order Two-Site Models

Considering the sites on the soil matrix that are accessible for retention of the

reactive solutes in solution, a second-order two-site model was proposed (Selim, 1992).

In this model, a fraction factor F was introduced to denote type 1 and type 2 sites to

the total amount o f sites S j, and </> as the amount of vacant sites in the soil. The


9

forward processes is controlled by the product of concentration (C) and vacant sites (<£),

by the reversible processes.

1.2.4 Factors Affecting Retention Reactions

Organic Matter: Laboratory studies have, in general, indicated that organic matter

plays a major role in the performance of soil applied herbicides. Such studies often

involved multiple correlation analyses for herbicide adsorption by a series o f soils with

widely different properties, including organic matter content, clay content, clay mineral

type pH and CEC (Stevenson, 1982). Huang (1984) indicated that besides organic

matter, the noncrystalline to poorly crystalline Al and Fe components (extracted by the

sodium dithionite-citrate bicarbonate method) and other inorganic constituents present

in a series o f particle size fractions of the soils, especially < 2 0 pm fractions, provide

adsorption sites for atrazine. Several other researchers (Wood et al., 1987; Peter and

Weber, 1985) have found high correlations between metolachlor and metribuzin

sorption and organic matter content. Brouwer et al. (1990) reported a linear increase

of distribution coefficient Kd with organic matter content. Thus, Koc (Kd divided by

percent organic carbon, %OC) was proposed as a characteristic property o f the

herbicide and was assumed to be independent of soil properties (Yaron et al. 1985).

The use of Koc showed less variability between soils. Groundwater contamination has

been frequently observed in regions where herbicides are applied to coarse-textured

soils with low organic carbon (OC) contents (Cohen et al., 1984). Because soil organic

matter is the primary sorbent for herbicides (Chiou, 1989), soils low in OC have a

lower capacity for retarding herbicide mobility.


10

Dissolved Organic Matter: Another important factor influencing the retention of

herbicides in soils is dissolved organic carbon. As it influences herbicide retention in

soils, the solubility o f a compound is enhanced by a fraction o f organic matter dispersed

into water. The presence o f high-molecular-weight humic material in water even at

trace quantities can significantly enhance the apparent water solubility of some

otherwise extremely insoluble organic compounds by partition-like interactions with the

microscopic organic environment of the dissolved humic material (Chiou and Manes,

1986; Gschwend and Wu, 1985). Complexation o f organic compounds such as

herbicides with suspended and dissolved organic matter has been demonstrated in the

laboratory (Stevenson, 1972). Several studies have shown that colloidal organic matter

and soluble macromolecules may be mobile through aquifer material and soils (Jardine

et al., 1992).

1.2.5 Hydrodynamic Dispersion

The various mechanisms that may contribute to dispersion have been under

investigation in several fields, including chemical engineering, chromatography, soil

science and hydrology. These mechanisms include axial diffusion, hydrodynamic

dispersion, boundary layer or film diffusion, and intraparticle diffusion. There are two

general ways in which the effects of multiple sources of dispersion can be modeled. The

first involves the use of an explicit terms and separate governing equations for each

mechanism contributing to dispersion. The effect of intraparticle diffusion on herbicide

transport, for example, is simulated with an appropriate equation describing flux in and

out of porous particles. The hydrodynamic dispersion phenomena is due to the


11

nonuniform flow velocity distribution during fluid flow in the porous soil medium. This

is responsible for the spreading of the herbicide pattern. When this spreading is

considered to be a random processes, the equivalent flux can be taken to be

proportional to the concentration gradient, with the dispersion coefficient D being the

proportionality constant.

The research o f hydrodynamic dispersion into the nature and extent of its effect

in agricultural soils has been reviewed by Biggar and Nielson (1967). Tritiated water

was used in these investigations. The phenomenon is connected with differences in flow

rate between various parts o f the water phase. These differences exist both within a

pore section and between different pores and pore sections. Values for the dispersion

coefficient can be obtained from carefully controlled soil-column experiments. Amount

o f spreading in such an experiment is shown by concentration distribution with depth

in the column or with times in the effluent. Average water-flow velocity in pores exerts

an influence, as does degree of water saturation. Diffusion in the water phase has a

two-fold effect: on the one hand there is a contribution to spreading in the flow

direction, on the other hand the diffusion component perpendicular to the flow direction

tends to eliminate concentration differences caused by hydrodynamic dispersion.

Generally, the hydrodynamic dispersion can be expressed as (Brusseau, 1993)

D=— *av (1-4)r

where D0 is the molecular diffusion coefficient in water (L2/T), r is tortuosity, a is

dispersivity (cm) and v is pore water velocity. The molecular diffusion coefficient of


ionic species in free solution can be estimated by the Nemst-Einstein equation (Ibl,

1983)

D - mRT (1 51D° JZJF (L 5 )

where m = the mobility (i.e. the steady state velocity attained by the ion under action

of unit force), R=universal gas constant, T is temperature, Z =charge of the ion, F is

Faraday constant.

For liquid organic compounds, the Stock-Einstein equation has been shown to

be adequate for describing the molecular diffusion coefficients (Thibodeaux, 1979). An

approximate analytical relation of diffusion coefficient in cm2 s' 1 for solute in water:

D o= 7 .4 x l0 ~ 8(^ X l8)1 /2r (1.6)Mv°-6

where V is the molar volume of the solute in cm3 mol' 1 as liquid at its normal boiling

point, n is the viscosity of the solution in centipoise (i.e. 0.01 g cm"1 s '1, /* are 0.01787

and 0.01022 poise at 0°C and 20°C respectively), Sk is the "association parameter" for

water, and T is the absolute temperature in K. The recommended value o f ¥ for water

is 2.6. This equation is good only for dilute solutions o f non-dissociating solute. It is

noted that the molecular diffusion coefficients are generally in the order o f 8-15x10^

cm2 s' 1 (Acar, et al., 1990).

Temperature corrections o f molecular diffusion coefficients for liquids can be

inferred from the above equation. Since viscosity is strongly temperature dependent, the

following proportionality should be used:


where the subscripts 1 and 2 refer to different temperatures respectively.

The amount o f herbicide diffusing through a unit area of soil at any point is the

product o f the gradient of concentration, and the apparent diffusion coefficient whose

value is determined by the geometric factors and partition o f a herbicide over the soil

phases. The molecular diffusion mechanism is due to the random thermal motion of

molecules in solution and is an active process regardless o f whether there is net water

flow in the soil. Herbicides diffuse more slowly in soil than in air or free solution

because the pathway through the pores is restricted and tortuous and because part of the

chemical may be retarded by sorption on the solids or because the chemical material

may influence the solubility. Graham-Bryce (1969) reported that diffusion coefficients

varied little with concentration for disulfoton and dimethoate in a silt loam soil, but

increased rapidly with increase in moisture content for dimethoate from 3.31x10'

8cm 2 s' 1 at 10% volumetric moisture content to l^ lx lO ^ c m 2 s' 1 at 43% moisture

content. In contrast, for disulfoton which is more volatile, less soluble and more

strongly sorbed than dimethoate, diffusion coefficients were smaller (2.83xlC 8cm2 s' 1

at 41 % moisture content) but did not change much as the soil became drier (2.74xl0 ' 8

cm2 s '1, at 8 % moisture content).

1.2.6 Preferential Transport

There are two different processes which are involved in herbicide transport, one

is the movement through the soil matrix and one is rapid downward movement through


14

macropores (Starr and Glotfelty. 1990). High water fluxes and flow through large pores

between aggregates increases nonequilibrium between liquid and solid phase.

Preferential transport has been cited as being responsible for the rapid movement of

agricultural chemicals to lower soil depths (Czapar et al., 1992). It increases the

dispersion coefficient D and water flux v and thus increases the transport. Preferential

solute transport and water flow in tilled and untilled soils has been reported by Andreini

and Steenhuis (1990). They used grid lysimeters to separate the water flowing through

undisturbed soil columns from conventional and conservation tillage systems both

temporally and spatially. They sprinkled two soil columns continuous with water (2 cm

d"1) for 10 days, applied a spike of bromide and blue dye for 24 h in the sprinkler

water, and then watered the columns for an additional 30 days. At no time during the

experiment did water accumulate on the soil surface. Following water application the

columns were excavated to determine water flow paths. In both tillage systems, dye

flow pathways led to areas where water and solutes exited the columns. In the no-till

column, nearly the entire depth of the profile was short-circuited by preferential flow,

but in the tilled column solute passed through the mixed, unstructured plow layer. By

use of a soluble dye to denote flow pathways in soil, Trundgill et al. (1983) found only

a small fraction o f soil stained. Less than 3% of total soil volume was involved in the

transport processes. Watson and Luxmoore (1988) found < 1 % of the soil volume was

used in transport o f 96% o f the water flux in a forest soil.


15

1.2.7 Biological Decomposition

Decomposition is another main phenomenon influencing the fate of herbicide

transport. Alachlor is primarily degraded microbially in soil. Degradation is a slow

biological processes for alachlor and other acetanilide herbicides (Sharp, 1988), and

only small amounts were mineralized in surface soils (Novick et al. 1986). It has been

found that microbial degradation was of greater importance than volatilization and

leaching for acetanilide herbicides (Zimdahl and Clark 1982). The half-lives of alachlor

at 20°C and 80% saturated soil moisture, was 18 days in a clay loam soil (Zimdahl and

Clark, 1982; Ivany et al., 1983; David, 1990). Microbial degradation of herbicides

increases with increased soil moisture and soil temperature. A quantity is needed to

express the decomposition-rate equation. Usually these rates are considerably higher in

soil than in water at the same pH. So far the first-order rate equation has been used

most frequently. It is assumed that the decomposition rate is proportional to the amount

of herbicide remaining at time t

where K, is the rate constant in day'1. Equations of the hyperbolic type represent

breakdown patterns with a comparatively high rate for lower amounts. For C = C 0/2,

the half-life (t1/2) can be calculated as

(1.8)

, 0.693M/2 ~ (1.9)


1.3 M odeling the W ater an d Herbicide T ransport

1.3.1 Modeling Water Flow

The flow through porous media depends primarily on the dimensions of the void

fraction. Since the size of the pores is difficult to characterize, a dimension of the solid

phase in the case o f discrete particles and the surface area o f the solid phase in the case

o f consolidate media are usually used for characterization. In one method, the packed

theory is then developed by applying single straight tubes to the collection of crooked

tubes. It is assumed that the medium is uniform and that there is no "channeling".

According to Blake-Kozeny equation and Darcy’s Law, the permeability of the column

is (Bird et al. 1960)

* = 2 5 ° < I ^ £ (1.10)d 2 e3

where d is equivalent diameter of aggregate and e is porosity o f the soil column.

Another method o f modeling water flow is to describe the characteristics of

channel flow. Bouma and Dekker (1978) used a dilute solution o f methylene blue in

water (0.03% by weight) and stained walls of pores in the soil were observed and

counted after subsequent excavation. The stain pattern in the entire soil mass on vertical

ped faces and in channels had been observed for 10 cm depth intervals in terms o f type

(bands on ped faces or in tubular pores), sizes and quantities. Presence of stain on the

wall o f a soil pore is a clear indication of water movement. But lack of stain doesliot

necessarily imply complete lack of movement. Stains are produced by complete or

partial penetration of dye water from large pores into dry, fine porous peds. Leaving


17

a stain on the outside as de-colored water proceeds to penetrate the ped and dye water

move further downward along the ped face. The number o f the channel bands > 7.5 mm

wide, and bands 2.5-7.5 mm wide were counted for every 10 cm interval in the profile.

They reported that the number o f pathways along large vertical pores that were

involved in infiltration (as indicated by stains) was determined by the flow region. The

number o f colored bands increases with rain intensity. The morphological observations

presented were transformed into diagrams showing the entire stained contact area for

any given depth interval. The observed infiltration pattern resulted in a small vertical

contact area between soil and infiltrating water through the large pores. Only a small

fraction (2 % or less) o f potentially available contact area in a dry soil was used for

vertical flow through the large pores). Bouma, et al., (1979) used Methylene blue to

indicate flow patterns. A statistical planar and tubular void-interaction model was used

to estimate sizes o f "necks" in the flow systems, which were assumed to determine the

saturated hydraulic conductivity. Two types of data, obtained from horizontal thin

sections, (i) measured width distribution of stained planar voids (dj, di+1, di+2...), and

(ii) the number of measurements in each width class dj (Nj, Ni+1, N i+2...). The data

were used to calculated hydraulic conductivities of the preferential flow based on

observed geometry of voids. For a plane slit of unit length (perpendicular to the

direction of the flow) it follows:

Q - P g * r \ H (1.11)t 817

where Q/t = volume of liquid conducted per unit time and length (cm3 sec'1); 17 =


18

viscosity (g cm ' 1 sec_1); p = liquid density (g cm'3); g = gravitational constant(m s'2);

VH = hydraulic gradient (m m"1); and d = width o f plane slit (cm). For a tubular void

with radius r (cm) it follows:

Q . P g r r \ F (1.12)t 817

1.3.2 Modeling Herbicide Transport

The traditional approach to modelling and predicting herbicide transport is to

solve the convection-dispersion equation ( 1. 1) analytically or numerically, depending

on the problem addressed (van Genuchten and Alves, 1982). This equation is based on

a differential mass balance of a control volume. Even though the solution to this

equation will result in the knowledge of the herbicide distribution in space and time,

in most cases only the concentration is monitored with time at a particular point. This

is especially true for laboratory studies performed for the determination of solute

behavior during transport in soil columns. The effluent versus time is referred to as the

breakthrough curve (BTC). The BTCs are usually modelled using the CDE, e.g. the

outflow is predicted by the determination of the solute distribution inside the control

volume. When the soil is nonhomogeneous, physical nonequilibrium models are often

proposed (van Genuchten, 1981).

Mobile-Immobile Transport Models

A mathematical model was proposed when apparent non-equilibrium condition

in the system are attributed to large heterogeneities in microscopic pore-water velocities

(Parker and van Genuchten, 1984). This approach assumes that the liquid phase can be


partitioned into mobile (dynamic or macro-porosity) and immobile (stagnant or micro

porosity) regions. Convective and dispersive transport is restricted to the mobile water

phase, while transfer o f solutes into and out o f the immobile (non-moving) liquid phase

is assumed to be diffusion-limited. The governing equation for this two-region model

mobile -immobile model are

where cm and cim are the resident concentrations of the mobile and immobile liquid

phases respectively; 0m and 0^ are the mobile and immobile volumetric water contents

such that 0 = 0m+ 0im> Dm is the hydrodynamic dispersion coefficient in the mobile

water region, q is Darcy water flux and fm represents the fraction o f the sorption sites

that equilibrated with the mobile liquid phase, and a is a mass transfer coefficient that

governs the rate of solute exchange between the mobile and immobile regions.

Applying reduced variable vm = q/0m, equation (1.13) and (1.14) become

(1.13)

(1.14)

(1.15)


The model consists of equation (1.15) and (1.16) subject to the conditions

Cm(X,T=0)*Cim(X,T-0)=0 (1.17)

7 > C im(X -« .,7 )= 0 (1.18)

(C4 l f ) lx - oS,'4ii(7) (1-19)

where 5(T) is the Dirac Delta function and A =M /(0L), where M is the mass input per

unit column cross-sectional area. The solution for the volume-averaged concentration

(cr) in terms of the reduced variables is

Cr(X ,7> C l+(C0-C,.M(X,7) 0 < T < T 0 (1.20)

Cr(X,7)=CI+(C0-C iM (X ,7 )-C ^ (X ,r-T 0) T> T0 (1.21)

where

TA(X,T)=^g(X,7)J(a,b)dT O-22)

(1.23)


21

a

J(a,b)=\-e~b f e-xI0[2sfbk]dh (1.24)

where

(1.25)

and

i . - M (r-r)(1-/3)/?

(1.26)

The function J(a, b) is refereed as Goldstei’s J-function (Goldstein, 1953); I0 in

this function represents a zero-order Bessel function.

Two - Site Mobile-Immobile Models

When considering the reaction in heterogenous systems, a Second-Order Mobile-

Immobile Model was proposed (Selim, 1992). In the model, the chemically controlled

heterogeneous reaction is governed according to the Two-Site approach, in the

meantime, the physically controlled reaction is chosen to be described by diffusion or

mass transfer of the mobile -immobile concept (Coats and Smith, 1964; van Genuchten

and Wierenga, 1976). Leaching of herbicide through macropores occurs as discrete

pulses with each large storm while transport through smaller pores is assumed

immobile. For such conditions, mobile- immobile transport models have been

successfully used to predict transport of several ions in soil columns (Selim et al., 1987;

Gaston and Selim 1990). In such models, equilibrium ion exchange reactions among

competing ions are considered in the mobile and immobile water zones and transfer


22

between zones is by diffusion. It is further assumed that the inter-aggregate surfaces are

those which are in contact with the immobile water; whereas, intra-aggregate or

macropore surface are indirect contact with the mobile water phase. This model was

extended to incorporate retention reaction of the equilibrium type as well as degradation

reactions (van Genuchten and Wagenet, 1989). In addition, a kinetic second-order

retention approach has been incorporated into the mobile-immobile concept by Selim

and Amacher (1988).

Stochastic Modeling of Herbicide Transport

Early stochastic modeling of water flow in heterogenous porous media was

proposed by Dagan (1982). It has become quite common to regard the medium as well

as the flow variable as a random variable characterized by probability distributions

rather than by well-defined deterministic values. The fractured medium were treated as

an equivalent porous medium (Neuman et al., 1985). This is appropriate when the

fractured media contains many inter-connecting fractures. In fact, there is much field

evidence that the dispersivity is not a constant due to medium heterogeneities but

depends on the travel distance and/or scale of the system. There have been many

studies to incorporate the scale dependent dispersion in modeling studies (Gelhar et a l.,

1979; Hatton and Lightfoot, 1984).

Another approach was that fracture medium was assumed by means o f a limited

number o f tortuous and intersecting channels (Tsang and Tsang, 1987). These channels

have variable apertures along their lengths. The parameters that characterize the

channels are (i) the aperture density distribution, which gives the relative probability


23

of the occurrence of a given aperture value, (ii) the effective channel length and width,

and (iii) the aperture spatial correlation length, which gives the spatial range within

which the aperture values are correlated.

Toride et a l., (1995) used three stochastic variables including pore water velocity

v, dispersion coefficient D and distribution coefficient Kd for linear adsorption or the

first-order rate coefficient for nonequilibrium adsorption a . The three different pairs

of random parameters are described with a bivariate lognormal joint probability density

function for the one dimensional solute transport.

1.4 Alachlor Behavior in Soil

Alachlor (2-chloro-N-(2,6-diethylphenyl)-N-(methoxymethyl)acetamide, a pre

emergence herbicide, is one of the most commonly used soil-active herbicides. It is

extensively used to control annual grasses and certain broad leaf weeds in com, cotton,

soybean and sugarcane etc. About 37 million kilograms o f alachlor is produced annually

in the USA (Chesters et al., 1989). The molecular structure is shown in Fig. 1.1. Its

molecular formula is C ^ f^ C lN C ^ and it has a molecular weight of 269.77. Its

solubility in water is 148-242 ppm at 25°C and with field half life of 14-49 days

(Wauchope et al. 1992). It is photolytically stable and is chemically stable up to 105°C.

The distribution coefficient varied with soils. A Kd of 1.1 and K^, of 157 was reported

by Kladivko, et al. (1991). Due to its moderate retention by soil, some studies (Wu,

1980) reported that alachlor did not leach below 8 cm in field soil and not below 4 cm

in a moist soil. However, residues of these herbicides are frequently found in

environmental samples o f soil and water due to their persistence and relatively high


24

H2

GH3

CH2CH3

O

IIN C

H2

01

CH3

Fig. 1.1 Alachlor Molecular Structure

CH2CI


25

water solubility and weak soil sorption (Cohen, 1984). Due to possible groundwater

contamination and health effects, the U .S. EPA proposed Maximum Contaminant level

Goals (MCLG) of zero ppb for alachlor (U.S. Environmental Protection Agency,

1989a). But so far, alachlor has been detected in ground water in many states including

Iowa, Pennsylvania, Maryland, Nebraska, and Minnesota etc. (Roux et al., 1991;

Cohen, 1992; Holden and Graham, 1990).

1.5 C urren t Experim ental M ethods on the Fate and T ransport of H erbicide

1.5.1 Retention and Mobility Studies of Herbicides

Batch methods are widely used method for characterizing the retention of

herbicides (Sparks, 1988) and for supporting product registration (US Environmental

Protection Agency, 1982). Thin-layer Chromatography is another method used for

degradation and mobility study of herbicide (Sharon and Koskinen, 1990). The study

via TLC conducted by Sharon and Koskinen (1990) indicated that greater than 88% of

the 14C in solution chromotographed with parent alachlor, the remaining 12% of the

14C in solution was distributed equally along the plate which may result from

degradation.

Several researchers conducted transport experiments for dissolved chemicals

using saturated soil columns (Selim and Ma, 1995; Czapar et al. 1992). It is a

commonly used method to study herbicide mobility, which is required by the USEPA

to support product registration (U.S. Environmental Protection Agency, 1982).


26

1.5.2 Tracer Studies for Characterizing Preferential Flow

Techniques to visualize macropores and cracks, and to measured their capacity

o f transporting water and solutes have been introduced by Bouma and Dekker (1978),

and Bouma et al. (1978, 1979). These methods often involve the application of an

amount o f dye solution to the soil surface. The presence o f the dye, which adsorbed at

the walls o f the cracks and macropores, can be studied. It can provide direct evidence

for the presence of preferential flow channels in the soil and gives information on the

nature and the extent of the flow channels involved.

Allen et al. (1989) proposed the mechanisms of dye adsorption in soils. The

adsorption o f a dye onto adsorbents follows three consecutive stages. First, dye

migrates through the solution to the exterior surface o f the adsorbent particles.

Secondly, the dye moves within the pores o f the particles. Then, thirdly, the dye is

adsorbed at sites on the interior surface of the adsorbent particles. They proposed that

the main resistance to the mass transfer occurs solely in the second stage, i.e., during

the movement or diffusion of the dye in the pore structure o f the adsorbent. Dubinin

(1967) suggested that the pore structure of adsorbent particle consists of macropores,

transitional pores, and micropores. The pore size distribution can be shown to be more

important than the surface area.

Dyes Used in Tracer Studies

Many non-fluorescent dyes have been used to illustrate flow paths in soils. The

choice of a dye is greatly restricted because soils often interact with the dye molecules

(Corey, 1968). The study of Corey (1968) indicated Acid Red 1 had the highest


27

mobility among anionic dyes tested. Ghodrati and Jury (1990) used the dispersed

Orange 3 dye to trace the preferential flow. However, it did not flow through the soil

profile and was filtered out o f the solution right at the soil surface, producing a crust

because of its high concentration (80 g kg '1). It penetrated only to a minimal depth of

a few centimeters, van Ommen et al., (1988) proposed a method based on the formation

of an intensely colored complex of iodine with starch. After application, horizontal

cross-sections o f the soil profile are obtained by removing slices of the soil. A whole

soil profile can be excavated in order to obtain a series of two dimensional pictures of

the phenomena. However, it is necessary that the starch and bleaching agent are

uniformly applied onto the soil’s cross section. Therefore, this approach is limited to

non-structured soils (van Ommen et al., 1988).

Although methylene blue is readily visible, it is strongly adsorbed along the

walls o f conducting pores (Smettem and Collis-George, 1985; Steenhuis et al., 1990).

Particularly in smaller pores, methylene blue is not a good indicator for water

movement.

The Brilliant Blue FCF was considered to be the best compromise between

mobility, visibility, and toxicity (Flury and Fliihler, 1994; 1995). Brilliant Blue FCF

has been used as tracer by Andreini and Steenhuis (1990) and Boll et al. (1992) in

laboratory experiments, and by Pickering et al. (1988) and Steenhuis et al. (1990) in

the field to stain flow pathways. The relative retardation was 1.2 for Brilliant Blue FCF

compared with iodine. The Brilliant Blue FCF move a shorter distance compared with

bromide.


28

Koch and Fliihler (1994) used Brilliant Blue FCF as a point source in the center

o f the column or as a pulse over the entire surface. Lateral mixing of the point source

was investigated visually by cutting the columns vertically into two halves immediately

after the dye was detected in the outflow. The dye patterns along vertical profiles

through the centers o f columns were drawn. In the same paper, Koch and Fliihler

(1994) characterized the dye patterns for a layered column with three distinct layers.

Steenhuis et al. (1990) use a 0.6% solution of Rhodamine WT dye (moderately

adsorbed), 1.0% FD&C blue #1 food coloring dye (weakly adsorbed), and 0.1%

methylene blue dye (strongly adsorbed). The dyes were used to identify the primary

solute-transport paths and to compare the flow mechanisms under the two tillage

treatments. The staining pattern in the soil by the dye solutions indicated earthworm

burrows are the most active flow conduits in the upper profile.

Ghodrati and Jury (1990) using an anionic water-soluble dye, Acid-Red 1 to

characterize preferential flow. The observed vertical and horizontal distribution patterns

o f the water-soluble dye clearly indicated preferential flow patterns for each

experimental condition. These include vertical fingering of dye tracers 5 to 20 cm wide,

which extended more than twice as deep as the mean displacement of the dye, as well

as isolated paths o f dye indicating lateral flow. In addition to photographing, accurate

traces of the stained dye patterns on the trench face were made at each site on clear

plastic sheets. This processes was repeated by cutting the trench face successfully at 10

cm interval both vertically and horizontally. The representative vertical and horizontal

flow patterns of the acid-Red 1 dye were produced directly from the photograph by re-


29

staining all colored regions with black ink. The relationship between the fraction of

vertical and horizonal surfaces covered by the dye and the soil depth for these plots

were obtained. The flow pattern are obviously three-dimensional, showing both lateral

and vertical movement along tortuous paths. Characterization o f preferential flow in

soils through water tracer studies are, in most cases, inconclusive because of the

inability of most solute sampling devices to detect the spatial pattern o f preferential flow

pathways.

1.5.3 Herbicide Transport under Unsaturated Flow

Under field conditions, macropores will empty as the soil water content

decreases, While mesopores remain water-filled and hydrological active. Mesopores,

which are capable of rapid infiltration (Wilson and Luxmoore, 1988), promote greater

attenuation o f chemicals due to increased pore-surface area (Luxmoore et a l., 1990) and

may be the key pore class to field-scale transport. To study the water and herbicide

transport through pores, an unsaturated column method was used by Jardine and Jacobs

(1990). They showed that while macropore transport is significant, solutes in smaller

pores within aggregates are also highly mobile. Jardine and Jacob (1990) demonstrated,

with undisturbed soil columns under steady-state flow conditions, significant differences

in reactive solute transport as conditions were changed from saturated to just -1.0 Kpa

pressure potential. For studying the transport of reactive Sr in undisturbed columns,

Jardine et al. (1993) conducted miscible displacement experiments under unsaturated

condition. Prior to the displacement experiments, soil column were slowly saturated

with 0.05 M CaCl2 from the bottom. The columns were then allowed to drain under


30

a desired pressure head in order to establish steady flow. The inlet pressure head was

maintained with a Mariotte device and the outlet pressure head was maintained with a

constant-vacuum source. A unit hydraulic head gradient was established within the

column by placing the inlet and outlet the same pressure head. This resulted in a

uniform matric potential and water content within the soil column. For most

displacement experiments, a pressure head o f -10 cm was selected because it allowed

macro-pores (diameter< 1 mm) in the soil columns to remain empty. The results

indicated that the application o f -10 and -15 cm pressure heads resulted in five- and 40-

fold decreases in the mean pore water flux, respectively, with relatively little change

in soil water content relative to saturated conditions. They concluded that most of the

water flux may be channeled through pores that hold water with tension < 10 cm

(macropores). Selim et al. (1977) simulated solute transport through unsaturated

multilayered soil profiles in which a steady, vertically downward water flow was

considered. BTCs for the reactive solute show lower retardation factors for the soil

profiles having a water table at z = 100 cm than at z-*oo. At steady state, a uniform soil

water content can be used to represent each soil layer in order to simplify the solute

transport problem. But for the transient water flow conditions o f unsaturated

multilayered soils, such a simplifying approach was not applicable (Selim, 1978).

Elrick et al.(1966) and Krupp and Elrick (1968) regulated the water content by

placing a perforated sample holder in a pressure chamber, while the desired flow rate

was obtained. Both ends o f the sample were in close contact with a cellulose acetate

membrane filter, supported by a porous screen and a stainless steel end cap containing


31

three plastic nipples. The sample and pressure chamber where placed on a balance to

allow continuous monitoring of the sample weight and hence water content. A driving

head, equal to the sample length, was established by increasing the pressure in the

chamber (unit gradient). A constant water content was obtained throughout the sample.

This meant that only one flow rate was possible for each water content.

The transport o f chemicals in soil columns during a constant-flux, unsteady,

unsaturated flow have been conducted by Mansell et al. (1993). Bond and Phillips

(1990) gave an analytical solution of cation transport during unsteady, unsaturated soil

water flow. Tan et al. (1992) conducted experiments using bacteria and calculated the

position of bacteria front in flux-dependent unsteady state flow condition which is

realistic in the field condition, but encountered difficulties in predicting the reactive

bacteria change with space and time.

1.6 Objectives

In herbicide investigations, the increased demand for quantitative data is a

stimulant to the use of computational models. There is a need to provide behavioral

patterns o f herbicide fate a toxicological value and to provide a discussion base for

agricultural and environmental policy. A minimum requirement is that it must be

possible to estimate and predict the concentration o f herbicide as a function o f time and

position in the soil profile. The specific objectives of this study were:

(i) Examine the influence of chemical properties (organic matter) on the sorption

and desorption kinetics of alachlor in soil of no-till and conventional tillage

systems and characterize the retention behavior o f alachlor;


32

(ii) Characterize the preferential flow and physical nonequilibrium transport of

solute in undisturbed and saturated soils from no-till and conventional tillage

systems;

(iii) Quantify physical and chemical nonequilibrium transport parameters o f alachlor

in undisturbed and saturated soils from no-till and conventional tillage systems;

and

(iv) Model physical and chemical nonequilibrium transport o f alachlor in undisturbed

and saturated soils from no-till and conventional tillage systems.


CHAPTER 2

MODELING ADSORPTION-DESORPTION KINETICS OF ALACHLOR IN SOILS OF DIFFERENT TILLAGE SYSTEMS

2.1 Introduction

Alachlor [2-chloro-N-(2,6-diethylphenyl)-N-(methoxymethyl)acetamide],apre-

emergence herbicide has been widely used for weed control for several agronomic crops

(Lawruk et al., 1992). Alachlor has been classified by EPA as Group B2, a probable

human carcinogen with a maximum contaminant level goal of zero in drinking water

(U.S. EPA, 1989).

Adsorption and desorption (or retention) reactions of pesticides such as alachlor

in soils play a significant role in their behavior and has been the subject o f numerous

investigations over the last three decades. Alachlor can be adsorbed by a coordination

bond, through a water bridge, between the C = 0 groups and exchangeable cations on

clay and organic matter (Bosetto et al., 1993). The coordination strength is directly

correlated with the polarizing power of the exchangeable cation. X-ray diffraction

analyses showed that alachlor penetrated the inter-layer space of montmorillonite

(Bosetto et al., 1993). Alachlor may be characterized as a moderately nonpolar

compound (Green and Karickhoff, 1990). Alachlor sorption has been associated with

both hydrophilic and hydrophobic soil components. Consequently, adsorption of

alachlor was found to be positively correlated with soil organic matter and clay content

(Locke, 1992).

33


34

The dependence o f pesticide retention on the concentration o f pesticides in the

soil solution has been observed by several investigators. Locke (1992) reported that a

high percentage o f alachlor applied to soils from no-till and conventional tillage

experiments was adsorbed at low initial concentration. At low concentrations, Weber

and Peter (1982) found that alachlor was adsorbed on Ca-organic matter and Ca-

montmorillonite systems in similar amounts. However, at high concentrations, higher

amounts of alachlor were adsorbed by Ca-montmorillonite than Ca-organic matter

system. Senesi et al.(1994), through infrared analysis, indicated that the alachlor

adsorption mechanism is concentration dependent, with H-binding and charge-transfer

processes being the major mechanisms occurring at low concentrations. A hydrophobic

binding of alachlor to the aliphatic parts of humic acid (HA) is dominant at high

concentration. As a result o f concentration dependant retention mechanism, a nonlinear

isotherms or Freundlich isotherms was found for a wide range o f concentration (Senesi

et al. 1994).

The fate o f pesticides and their potential mobility in soils are directly influenced

by their retention mechanisms. In a kinetic study, Clay and Koskinen (1990) reported

that alachlor retention was rapid and appeared complete within 24 h, with the first 2 h

accounting for 88% of total retention. As a result, the validity o f the local equilibrium

assumption (LEA) is often invoked. For example, the Freundlich (equilibrium)

approach was utilized to describe alachlor retention 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


35

information on the retention mechanisms. In addition, the rate at which "apparent"

equilibrium is achieved, is influenced by limitations to diffusion through the sorbent as

well as by specific sorbate-sorbent interactions. Nonequilibrium conditions may also

be due to heterogeneity o f sorption sites and slow diffusion to sites within the soil

matrix, i.e. slowly accessible sites.

In a review of pesticide adsorption and desorption behavior, Calvet (1980)

indicated that desorption is generally slower than adsorption. Based on adsorption-

desorption kinetic results for alachlor over a 24 h period, Boesten and Van der Pas

(1988) concluded that the evidence supporting the existence o f hysteresis of alachlor is

not convincing. In contrast, obvious hysteretic patterns were observed by Bosetto et al.

(1993) and Helling et al. (1988). Hysteretic behavior along with increased irreversible

alachlor retention with time were also reported by Clay and Koskinen, 1990) and Locke

(1992). Incomplete recovery of applied alachlor and commonly observed hysteretic

desorption phenomena were attributed to slower secondary sorption reactions. In fact,

Locke (1992) applied with success a three-site multireaction model to describe alachlor

adsorption versus time for soils under no-till and conventional tillage. No attempts

were made to describe alachlor retention behavior during desorption, however.

The objectives o f this study were: (i) to determine the kinetics of alachlor

retention in two surface soil samples for a wide range o f input concentrations, and (ii)

to determine the hysteretic characteristics of alachlor adsorption-desorption in soils.

The soils were Gigger silt loam collected from long term no-till and conventional tillage

cotton plots. A third objective is to assess the capability of a nonlinear multireaction


36

model for the prediction of alachlor adsorption as well as desorption in soils based on

independently derived model parameters.

2.2 Methods

2.2.1 Soil

Surface soils (0-15 cm) were sampled from long term no-till and conventional

tillage cotton plots under wheat as winter cover, from the Macon Ridge Research

Station, Winnsboro, LA. The soil used in this study was a Gigger soil and contained

22% sand, 66% silt and 12 % clay. Organic matter content was 1.45% for the soil of

no-till and 0.81 % for the soil of conventional tillage. The pH (1:1 soil: water ratio) was

5.37 in the no-till and 5.23 in the conventionally tilled soil.

2.2 .2 Batch Experiment

Alachlor retention in the two surface soils was carried out using batch methods

as outlined by Ma and Selim (1994). The adsorption experiments were conducted by

mixing 10 g o f air-dry soil and 20 ml o f alachlor solutions o f varying concentrations

in 40 ml Teflon tubes. Two batch experiments were conducted. In the first (data set I),

three initial alachlor concentrations (CQ= 1, 5, and 10 mg L"1 in 0.01 N CaCl2 solution)

and 10 reaction times (1, 2, 4, 8, 16, 32, 64, 128, 256 and 512 h) were used in

duplicate. In the second (data set II), eight alachlor initial concentrations (Co= 0 .5 , 1,

2, 5, 10, 20, 30 and 50 mg L '1 in 0.01 N CaCl2 background solution) with 4 reaction

times (24, 48, 120 and 528 hours) were carried out in duplicate. Samples were shaken

for each reaction time, then were centrifuged, and the supernatants were withdrawn,

and filtered for alachlor analysis by high performance liquid chromatography (HPLC).


37

Desorption of alachlor from the soil was carried out following sorption using

successive dilution. Desorption was carried out in the first batch experiment (data set

I) only. Desorption steps commenced following centrifugation and decanting of

supernatant for HPLC analysis and weighing o f sample tubes. The amount decanted was

replaced by 20 ml of the background solution (0.01 N CaClj). The samples were then

reweighed, vortex mixed, and returned to the shaker for the second desorption step (one

day reaction time) for a total o f six desorption steps in each desorption.

2.2.3 Multiple Reaction Modeling and Statistical Analysis

The two-site equilibrium-kinetic model of Selim et al. (1976) is perhaps one of

the earliest multi-site or multireaction approaches for describing retention and transport

behavior of reactive solutes in porous media. Basic to the multi-site approach is that the

soil solid phase is made up o f 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. This multiple or distributive reactivity

approach was recently used by Weber et. al. (1992). As reported by Clay and

Koskinen (1990) and Locke (1992), alachlor is assumed to react at different rates with

different sites on matrix surfaces. Therefore, a multireaction kinetic approach may be

considered to describe alachlor retention kinetics in soils.

The multireaction model used here considers several interactions o f one reactive

solute species (alachlor) within the soil environment. Specifically, the model assumes

that a fraction of the total sites is highly kinetic whereas the remaining fraction interacts

slowly or instantaneously with solute in the soil solution. As illustrated in Fig. 2.1, the


38

model accounts for reversible as well as irreversible (concurrent and consecutive) type

reactions;

cm '3), 8 is soil water content (cm3 cm'3), and t is reaction time (h). The parameter ke

reaction rate coefficients (h '1). The term Se is the amount retained (reversibly) by

type sites (jig g '1 soil), and n and m are empirical parameters (dimensionless). In

addition, S2 is the amount irreversibly transferred from Sl5 Sirr represents irreversible

sites (pg/g soil). Transformation of Sj to S2 (eq. 2.3) as well as decomposition of

alachlor in the solution phase (eq. 2.4) were assumed to follow a first-order reaction.

2.3 Results and Discussions

2.3.1 Adsorption-Desorption Kinetics

Different versions of the multireaction model of Fig. 2.1 represent different

reactions from which one can deduce alachlor retention mechanisms.

(2 . 1)

(2 .2)

(2.3)

(2.4)

where C is solute concentration in soil solution (mg L '1), p is soil bulk density (g

is Freundlich distribution coefficient (cm3 g '1) and k1? k2, k3 and lq„ are associated

equilibrium type sites (pg g '1 soil), Sj is the amount retained (reversibly) by kinetic


39

Fig. 2.1. Schematic diagram of the nonlinear multireaction kinetic model.


40

Four variations were examined; (1) a two parameter model with kg and kjrr, (2)

a three parameter model with kg, kj and k2, (3) a four parameter model with kg, k 1#

k2 and lq^, and (4) another four parameter model with kg, k 1# k2 and k3. Each model

variation was fitted to the experimental data using a nonlinear, least-squares, parameter

optimization scheme (van Genuchten, 1981). Criteria used for estimating the goodness-

of-fit of the model to the data were the r2 and the root mean square (rms) statistics

(Kinniburgh, 1986),

rms = [_^£_]1/2 (2.5)lN-P

where rss is the residual sum of squares, N is the number of data points and P is the

number of parameters.

In all subsequent model calculations, we assumed empirical coefficients n and

m to be non-unity and n = m (eq. 2.1-2.2). This assumption was used since there is

no known method for estimating n or m independently. Estimated values were obtained

from fitting isotherm (adsorption) results after 528 h to the Freundlich model, ST= k fCn

where ST is the total amount retained. Estimates for n were 0.608+0.028 for the soil

from no-till and 0.660±0.013 for the soil from conventional tillage. Based on the extra

sum of squares principle (Kinniburgh, 1986), there was a significant (p<0.05)

improvement in the fitting of the Freundlich model as opposed to a linear model

(ST= k dC). Therefore, these n values were used in all model calculations.

The goodness-of-fit of these model variations was tested using alachlor

adsorption-desorption (data set I) for initial concentration in solution (C0) of 1, 5, and


41

10 mg L '1. The results are listed in Table 2.1 for the no-till and conventional-till soil.

Experimental data points and model predictions for the model variations are shown in

Fig. 2.2 for the no-till soil with C0 o f 5 mg L '1. The time-dependent behavior was

well described by the various model variations indicating initial fast adsorption (closed

circles) followed by slow adsorption where kinetic reactions appear dominant.

Desorption results following 16, 128, 256, and 512 h of adsorption are shown by the

open circles and were also well described by our model.

As indicated from estimates of the rate coefficients for the no-till and

conventional-till soil (Table 2.1), model variation 4 with the consecutive irreversible

reaction (kg, k j, k2 and k3) consistently provided lowest rms and highest r2, and thus

the best overall data fit. Alachlor concentrations versus time are given in Figs. 2.3 and

2.4, where the solid curves are fitted using model variation 4. Model variation 3 with

the concurrent irreversible reaction (ke, k j, k2 and k ^ ) did not provide as good a fit

as model variation 4 with the consecutive irreversible reaction. Therefore, one may

assume that alachlor adsorption processes include a consecutive irreversible rather than

a concurrent irreversible reaction. Consecutive irreversible type sites may represent

"restricted" sites as reported by Locke (1992). Sorption may be reversible at some sites,

but with time, alachlor may diffuse into clay or organic matrices where binding may

be less reversible. Stronger retention by some clay and organic matter sites, together

with slow diffusion from the clay surfaces, may result in the so-called "restricted"

sorption phase. In addition, alachlor is known to be degradable under aerobic and

anaerobic conditions and degradation was described by first-order kinetics (Pothuluri


Reproduced with

permission

of the copyright owner.

Further reproduction prohibited

without permission

Table 2.1. Comparison o f goodness o f fit o f several variations o f the nonlinear multireaction model for describing alachlor adsorption-desorption kinetics (data set I) for Gigger soil with no-till and conventional till. Overall fit is based on model variation 4 with all Cc’s.

Co

ModelVan- Model ation r2 rms ke K r kl k2 k3

mg/L

1.0 1 0.969 0.0262

cm^/g --------------------

1.4506±0.1144 0.0150±0.0023

If1No-Till

2 0.978 0.0224 1.4600±0.1006 0.0107 +0.0018 0.0027 +0.0012 -3 0.978 0.0225 1.4604+0.1010 0.0000+0.0028 0.0107+0.0018 0.0027+0.0012 -4 0.979 0.0221 1.4404±0.1052 0.0083 +0.0028 0.0067 +0.0069 0.0034+0.0026

5.0 1 0.985 0.1377 1.9698±0.1498 0.0024±0.0004 _2 0.991 0.1113 1.8420±0.1338 0.0076 +0.0020 0.0043 +0.0018 -3 0.991 0.1120 . 1.8422±0.1350 0.0000±0.0002 0.0075 +0.0020 0.0043 +0.0018 -4 0.992 0.1098 1.8038 +0.1462 0.0079 +0.0037 0.0109 +0.0106 0.0017+0.0012

10.0 1 0.991 0.2356 2.1438 +0.1472 0.0011+0.0002 _ _2 0.996 0.1643 1.8880+0.1214 0.0102+0.0028 0.0095+0.0031 -3 0.998 0.1546 1.7282+0.1424 0.0007+0.0002 0.0215+0.0095 0.0372+0.0183 -4 0.998 0.1478 1.7252+0.1386 0.0230+0.0101 0.0406+0.0192 0.0014+0.0004

Overall 4 0.995 0.1252 1.8370+0.0966 0.0179+0.0056 0.0313+0.0114 0.0017+0.0008

(Table 2.1 continued)

Reproduced with

permission



without permission.

Conventional -T ill1.0 1 0.972 0.0313 1.1418±0.1036 0.0067±0.0011 - - -

2 0.978 0.0281 1.1368±0.0962 0.0057±0.0012 0.0018±0.0011 -3 0.978 0.0283 1.1382±0.0980 0.0000±0.0009 0.0057 ±0.0013 0.0020±0.0012 -4 0.980 0.0276 1.1054±0.1054 0.0044±0.0021 0.0107 ±0.0139 0.0028 ±0.0014

5.0 1 0.980 0.1614 1.7464±0.1620 0.0019±0.0004 _ - -

2 0.987 0.1351 1.5858±0.1530 0.0079 ±0.0027 0.0058±0.0027 -

3 0.989 0.1351 1.5066±0.1734 0.0010±0.0007 0.010±0.0059 0.0165±0.0146 -4 0.991 0.1287 1.2120±0.2928 0.0589±0.0579 0.1217±0.1028 © H- o

10.0 1 0.993 0.2127 1.7266±0.1198 0.0011 ±0.0002 _ - -

2 0.997 0.1407 1.5392±0.0914

00is©■H001o

0 1 H- © M -

3 0.997 0.1411 1.5248 ±0.0978 0.0080±0.0021 0.0094±0.0059 0.0002±0.0005 -4 0.998 0.1333 1.5222±0.0960 O 8 00 ON H- o 8

00008o+1ood

0.0010±0.0005

Overall 4 0.996 0.1117 1.5078 ±0.0696 0.0094±0.0025 0.0170±0.0067

o+1

8o

CG is alachlor concentration in soil solution; r2 is correlation coefficient; rms is the root mean square error, is Freundlich distribution coefficient; is reaction rate coefficient from solution to irreversible sites Sirr. lq and k2 are forward and backward reaction rate coefficients for reversible kinetic site Sj and k3 is rate coefficient irreversibly transformed from sites Sj to S2.

44

No-Till c0=5 m g /L_ i

\D>E

Model v a r ia t io n 1

4 _• Adsorption o Desorption

zoI—<cn

zLl Iazoa

o 100 300200 400 500 600 700

TIME (h)

Fig. 2.2 Measured alachlor concentration versus time during adsorption (closed circles) and desorption (open circles) for Gigger soil with no-till with initial concentration(C0) of 5 mg L"1. Desorption results followed 16, 128, 256, and 512 h of adsorption. Solid and dashed curves are fitted results using four variations of the nonlinear multireaction model.


45

No-TillC#- 1 m g /L

0.8F lttsd (o v e ra ll) -

F ittad (Indiv idual) • A dsorption O O ssarp tlon

0.6

H 0 .4

0.2

0.0

100 200 300 400 500 600 700

T

C0« 3 m g /L

Zo

zUJozou

100 200 300 400 500 600 700

-JN9Ew

C#» 1 0 m g /L *

h-ZUJozoa

100 2 0 0 3 0 0 4 0 0 5 0 0

TIME (h )

600 70 0

Fig. 2.3. Experimental results o f alachlor adsorption (closed circles) and desorption (open circles) in Gigger soil with no-till. Initial concentrations (C,,) were 1, 5 and 10 mg L '1. The solid curves are based on individual fitting for each CQ (model variation 4) and dashed curves are based on the overall set o f model parameters of Table 2.1.


46

Conventional Tillage

C .-1 m g /L- F itted (Ind iv idual) F itted (o v e ra ll) _

0.8

0 .6• A dsorp tion ? D esorp tion

0 .4

• \

00

100 200 300 400 500 600 7005

C#« 5 m g /L

\ * cnE

W 3Zo

2

Zu 1 o 1 z oa 0

0 100 200 3 0 0 4 00 500 600 70010

Co« 1 0 m g /L—I 6o»

6

4

liJ 2

0

3 0 0 600 700100 200 400 5000TIME (h )

Fig. 2.4. Experimental results of alachlor adsorption (closed circles) and desorption (open circles) in Gigger soil with conventional tillage. Initial concentrations (C ^ were 1, 5 and 10 mg L '1. The solid curves are based on individual fitting for each C0 (model variation 4) and dashed curves are based on the overall set of model parameters o f Table 2.1.


47

et al. 1990). Although model variation 4 provided the best fit, model parameters were

significantly different for different CD’s. As indicated in Table 2.1, estimated ke at C0=

1 mg L"1 was significantly lower than that for higher C0’s. However, there was no

significant difference in 1^ values for C0 of 5 and 10 mg L '1 for no-till and

conventional-till soils. In addition, for both soils, there was some increase in k3 as C0

decreased, which is indicative of a high fractions o f irreversible sites at low

concentrations. It may be assumed that irreversible sites are surface limited. At low

concentrations, alachlor was highly accessible to the irreversible sites and vice-versa.

Conversely, the kinetic backward rate constants k2 was higher at high C0’s. The overall

shape of experimental and model fitted curves are flatter at the higher initial

concentrations compared with lower concentrations (Figs.2.3 and 2.4).

Based on parameter estimates given in Table 2.1 for data set I, ke values for no

till soil were significantly higher (p=0.05) than those o f conventional-till soil,

regardless o f initial concentrations. This finding is also consistent with the results

from data set II (Table 2.2). No significant differences were found for k1? k2 and k3

between the two soils. This result is consistent with the report of Locke (1992) where

the Freundlich parameter Kf was found to be higher in no-till than in conventional till

soils due to increased organic matter content. Locke (1992) also reported that their

instantaneous adsorption coefficient ( k j for no-till was greater than that o f conventional

till.

To test whether adsorption-desorption mechanisms based on model descriptions

were concentration dependent, the entire data set for all CQ’s was used in the nonlinear


48

Table 2.2. Comparison o f goodness o f fit o f the nonlinear multireaction model (variation 4) for describing alachlor adsorption kinetics for C„ o f 0.5 to 50 mg/L (data set II) for Gigger soil.

Parameter No-Till Conventional-Till

n 0.608±0.028 0.660±0.013

ke (cm3/g) 2.330±0.208 0.926±0.278

ki (h-1) 0.0043 ±0.0030 0.0130±0.0090

k2 (h-1) 0.0059 ±0.0048 0.0137±0.0107

k3 (h-1) 0.0000±0.0004 0.0001 ±0.0001

r2 0.9999 0.9999

rms 0.1415 0.0488

C0 is alachlor concentration in soil solution; n is Freundlich nonlinear empirical parameter; ke is Freundlich distribution coefficient; kj and k2 are forward and backward reaction rate coefficients for reversible kinetic sites k3 is rate coefficient of irreversibly transformed from sites S t to S2; k ^ is reaction rate coefficient from solution to irreversible sites Sirr; rms is the root mean square error and r2 is correlation coefficient.


49

least-square parameter estimation procedure. As a result, a set of model parameter

estimates, hereafter referred to as the overall set of parameters, were also obtained (see

Table 2.1). The dashed curves of the time dependent alachlor adsorption-desorption

shown in Figs. 2.3 and 2.4, were obtained based on these overall set of parameters.

There were considerable discrepancies in predicting alachlor concentrations during

adsorption, especially at the low C0 in both soils used in this study. The deviation was

less pronounced at high CD. At high CQ the overall set of parameters described both

adsorption and desorption versus time behavior equally well and is considered an added

credence to the capability o f our multireaction model. Moreover, this concentration

dependent alachlor observation is consistent with that reported by Locke (1992). Senesi

et al. (1994) also indicated that alachlor retention may be governed by concentration

dependent adsorption mechanisms. As a result, the extent o f kinetic reactions became

concentration dependent. H-binding and charge-transfer processes may be irreversible

and time dependent. A similar analogy may be advanced for higher concentrations.

Results of alachlor retention behavior during adsorption from data set II are

shown in Figs. 2.5 and 2.6 for no-till and conventional-till soils, respectively. The

solid curves shown in the Figures were obtained based on the overall set of parameters

(model variation 4) and are given for both soils in Table 2.2.

Based on the goodness of fit and their statistics, adsorption results of data set

n were well described by the model as illustrated by the low rms and high i^s.

Moreover, the use of an overall set of parameters was successful except for data from


50

35Co= 5 0 m g / L NO—Till

30

25

So>E 20

30zo

20uiozoo

0.5.1 0 0 2 0 0 3 0 0 4 0 0 5 0 00

TIME (h)

Fig. 2.5. Experimental results of alachlor in Gigger soil with no-till and for awide range o f initial concentration (C0’s). The solid curves are based on overall set o f model parameters in Table 2.2.


51

45Convent ional t i l lage

40

Co= 5 0 m g / L35

«"■\ \ 30a>E

30HZUioz 20oo

0 .50 1 0 0 2 0 0 3 0 0 4 0 0 5 0 0

TIME (h)

Fig. 2.6. Experimental results of alachlor in Gigger soil with conventional tillage and for a wide range of initial concentration (CD’s). The solid curves are based on overall set of model parameters in Table 2.2.


52

low C0 ( < 1). This finding illustrates the dependence on C0 and is consistent with our

previous finding based on adsorption-desorption of data set I.

2.3.2 Alachlor Hysteresis

Adsorption-desorption results (data set I) are presented as isotherms in the

traditional manner in Figs. 2 .7 and 2.8. These isotherms clearly indicate considerable

hysteresis in both soils. This hysteretic behavior resulting from discrepancy between

adsorption and desorption isotherms was not surprising in view of the kinetic retention

behavior o f alachlor in our soils. 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. It has been shown that desorption protocol (artifact) also may

be a cause for such a discrepancy (Clay and Koskinen, 1990; Locke, 1992). Bowman

and Sans (1985) reported that a substantial amount of observed hysteresis is caused by

centrifugation o f the suspension. This artifact may partially explain the discrepancy

between model prediction and measured data as discussed in the next section.

Adsorption-desorption isotherms (Figs. 2.7 and 2.8) indicate that the amount of

irreversible or nondesorbable phase increased with time of reaction. Alachlor may be

retained by heterogeneous type sites having a wide range o f binding energies. At trace

alachlor concentrations, binding may be irreversible. The amount o f nondesorbable

herbicide almost always increased with time (Wauchope and Myers, 1985). In terms

of its energy status in soils, methanol-extractable alachlor exhibited stronger interaction

(H-bounding and charge transferring) with soil organic matter and clay than did water-


53

aUimx001a:o- iXo3<

No-Till A d s o r p t i o n t i m s - 2 5 6 h

■ Ads or pt i on D e s or p t i o n for

• C = 1 0 m g / L A C“= 5 m g / L O C = 1 m g / L

Fi t ted■ — P r e d i c t e d

‘No-TillAdsorption t lm e -5 1 2

sEwoUJcoxo(0Xo

u3<

■ Adsor pt i on De so rp t ion for

• C = 1 0 m g / L A C° = 5 m g / L O C = 1 m g / L

Fi t ted- — Pr e d i c t e d

1 2 3 4 5

ALACHLOR CONCENTRATION (mg/L)

Fig. 2.7. Experimental and predicted alachlor adsorption-desorption isotherms for Gigger soil with no-till. The solid curves are based on overall set of model parameters (Table 2.1) and dashed curves are predictions using parameters from data set II (Table 2.2).


54

I 1 I IConventiona l t i llage A d s o r p t i o n t i m e = 2 5 6 ho«

' ■ A d s o r p t i o n D e s o r p t i o n f o r • C = 1 0 m g / L A C*« 5 m g / L O C»= 1 m g / L

oUJmctoin

xo3<

— Overa l l F i t t ed - - P r e d i c t e d

0 2 3 4 5 6

C o n v e n t i o n a l t i l l a g eAdsorption t im e -5 1 2 h

■ Ads or pt ion

De so rp t i on for• C = 1 0 m g / L A C = 5 m g / L O C ° = 1 m g / L

— Fi t ted - - P r e d i c t e d

ALACHLOR CONCENTRATION (m g /L )

Fig. 2.8. Experimental and predicted alachlor adsorption-desorption isotherms for Gigger soil with conventional tillage. The solid curves are based on overall set o f model parameters (Table 2.1) and dashed curves are predictions using parameters from data set II (Table 2.2).


55

extractable fractions. Methanol-nonextractable alachlor forms the strongest bonds in soil

and is referred to as soil bound residues. Lack of complete recovery has been reported

by others. For example, Helling et al., (1988) found that alachlor recovery (1 d after

application) was 57% of that surface broadcasted as an emulsifiable concentrate. Bosetto

et al. (1993) reported only 50% recovery was obtained after several washings in

distilled water.

Decomposition may also contribute to the observed alachlor hysteresis shown

in Figs. 2.7 and 2.8. Alachlor is primarily degraded microbially in soil. Degradation

of alachlor and other acetanilide herbicides is a slow biological processes, and only

small amounts were mineralized in surface soils (Novick et al. 1986). Microbial

degradation was reported to be greater than volatilization or leaching for acetanilide

herbicides (Zimdahl and Clark, 1982). Alachlor was 50 times more persistent in a

sterile than in non-sterile soil (Beestman and Deming, 1974). The half-life of alachlor

in a clay loam soil was reported as 18 d (Zimdahl and Clark, 1982).

2.3.3 Validation

A prerequisite for the validation of a model, such as the multireaction model

proposed here, is that necessary model parameters must be estimated independently.

Such a validation should be carried out prior to model adoption for prediction of

retention and mobility of pesticides in the soils. This validation is also necessary for the

use o f a model for different soils and for a wide range of conditions. This requirement

is not always achieved because independent parameters are not often available,

however. As a result, evaluation of a model is sometimes restricted to goodness of fit


56

of the model results to experimental measurements such as the solid curves of

adsorption-desorption isotherm results o f Figs. 2.7 and 2.8. In addition, efforts to

describe adsorption-desorption often result in two significantly different sets of

parameters; one for adsorption and another for desorption (e.g. Ma et al, 1993; Locke,

1992). In contrast, for all calculations presented in this study, only one set o f model

parameters was used to describe both adsorption and desorption. We are not aware of

efforts to describe alachlor adsorption-desorption kinetics based on only one set of

model parameters as was carried out in this study.

Validation of our proposed kinetic model is illustrated by the dashed curves

shown by the isotherms of Figs 2.7 and 2.8 and the solid curves of alachlor

concentration versus time for selected C0’s shown in Fig. 2.9. Here all model

parameters (n, kg, k l5 k2 and k3) were obtained independently from experimental

measurements (i.e., from data set II). With the exception of p and 0, initial conditions

for this initial-value problem were the only input required. Based on these predictions,

we can conclude that the model predicted the overall alachlor adsorption and desorption

(hysteresis) behavior satisfactorily. However, predictions of desorption isotherms were

not considered adequate at longer time periods. In addition, the model underpredicted

experimental adsorption isotherms which directly influences subsequent predictions for

the desorption isotherms. Discrepancies between experimental and predicted are

expected if the amounts of alachlor in the various phases (C, Se, S 1# and S2) at each

desorption step were significantly different. These underpredictions also may be due to

the inherent assumptions o f the model. Specifically, the model may not account for all


57

• A d s o r p t i o n o D e s o r p t i o n — P r e d i c t e d A d s o r p t i o n ■ - ■ - P r e d i c t e d D e s o r p t i o n

C#« 1 0 m g/L

P r e d i c t e d ;

A *O o o '0 -0 -0 ° o 'o -o

600 700500300 400200too06 e A d s o r p t i o n

o D e s o r p t i o n — P r e d i c t e d A d s o r p t i o n . - ■ • P r e d i c t e d D e s o r p t i o n

No-Til lC « 5 m g /L

5

4 P r e d i c t e d

3

2

1

o O'--. 'O O'O'O ° O O'O O O C JO0

0 100 20 0 30 0 400 500 600 700

TIME (h)

Fig. 2.9. Experimental and predicted alachlor concentration versus time during adsorption (closed circles) and desorption (open circles) for initial alachlor concentration (C ^ of 5 and 10 mg L '1 for Gigger soil with notill. The solid and dashed curves are predictions using parameters from data set n (Table 2.2).


58

retention or provide incomplete description of alachlor reactions in soils. In addition,

we recognize that our parameter estimates were based on the nonlinear least-square

procedure where high concentrations are emphasized. W e also assumed that the

reversible reactions had the same nonlinearity (n=m ). The model allows for n & m but

we are not aware of experimental method for determining n and m independently.

Finally, based on literature review, most retention experiments were designed

for adsorption measurements where desorption data were not always sought. Therefore,

kinetic retention models, such as the one proposed in this study, which are capable of

predicting desorption behavior of pesticides in soils based solely on adsorption

parameters are o f practical importance. Based on our results, the overall goodness of

our model predictions, except for large desorption times, are considered adequate and

provides added credence to the applicability of our proposed model approach.


CHAPTER 3

PREFERENTIAL FLOW IN SOILS FROM NO-TILL AND CONVENTIONALTILLAGE SYSTEMS

3.1 Introduction

Preferential transport has been cited as being responsible for the rapid movement

of agricultural chemicals in structured soils (Czapar et al., 1992). Because soils remain

essentially undisturbed under no-till, macropores are developed and thus enhance

infiltration o f water (Ehlers, 1975). As a result, under undisturbed flow conditions there

is a higher leaching of pesticides under no-till (Isensee et al., 1990). Dyes are widely

used to stain the travel paths of water and solutes to provide direct evidence for the

presence of preferential flow channels in soils.

Techniques to visualize macropores and cracks, and their capacity of

transporting water and solutes have been introduced by Bouma and Dekker (1978), and

Bouma et al. (1978, 1979). These often involve the application of an amount of dye

solution to the soil surface. The presence o f a dye, which is adsorbed on the walls of

cracks and macropores, can be thus investigated. Preferential solute transport and water

flow in tilled and untilled soils has been reported by Andreini and Steenhuis (1990). A

blue dye was used to trace the water flow path. In no-till columns, nearly the entire

depth of the profile was short-circuited by preferential flow, but in the tilled column

solute passed through the mixed, unstructured plow layer.

59


60

Steenhuis et al. (1990) used dyes to identify the primary solute-transport paths

and to compare the flow mechanisms under two tillage treatments. The staining pattern

in the soil by the dye solutions indicated earthworm burrows were the most active flow

conduits in the upper profile. Ghodrati and Jury (1990) used an anionic water-soluble

dye to characterize preferential flow in field soils. The observed vertical and horizontal

distribution patterns of the water-soluble dye clearly indicated preferential flow patterns

under either ponding or daily sprinkler irrigation. These include vertical fingering of

dye tracers 5 to 20 cm wide, extending more than twice as deep as the mean

displacement, as well as isolated paths of the dye which is indicative of lateral flow.

In most cases, characterization of preferential flow in soils based on tracer studies are

inconclusive. This is mainly due to the inability o f most solute sampling devices to

detect the spatial pattern o f preferential flow pathways in the soil over time.

Transport studies in soil columns have commonly been conducted with sieved,

dried and repacked soil materials. The soil structure is destroyed through such a

pretreatment. This shortcoming can be overcome by use of undisturbed soil columns.

Solute transport studies with undisturbed soil cores have been proven to be useful in

characterizing macropore flow (White et. al., 1984). Several theoretical models,

including the convection-dispersion (CDE) equation (Parker and van Genuchten, 1984),

mobile-immobile model (van Genuchten and Wierenga, 1976), two-region transport

models (Parker and van Genuchten, 1984; Selim and Ma, 1995) and stochastic models

(Bresler and Dagan, 1982) have been developed to describe one dimensional solute

transport in laboratory and field conditions. Several authors found that their


61

experimental data could be explained equally well by a two-region model as well as by

the simple CDE model (Schulin et al., 1987), and they could not decide whether a

distinction into macropore region and micropore region was justified. Steenhuis, et al.,

(1994) proposed that solute transport consisted o f preferential flow and matrix flow with

little modification through the soil channels (preferential path) to the deep soil. The

solute concentration in matrix flow obey the convective-dispersive equation.

Coats and Smith (1964) introduced a transport model to account for diffusion-

limited mass transfer into immobile soil regions and to explain the observed large

dispersion coefficients found for structured soils. By using flow interruption, the large

impact of micropore diffusion was observed and the contributions o f macro- and

micropore transport mechanisms can be separated (Oren et al., 1996).

The soil can also have a distinct bimodal pore size distribution, water-filled

macropores are considered to contribute to the convective transport, whereas transport

into and out of the micropores is by diffusion only. During interruption, diffusion in

the column continues and may have induced a change of concentration in the effluent

when flow resumes. Solute transport in structural soil is therefore characterized by an

early arrival of solutes displaced through the macropores and a slow approach to the

final concentration caused by the slow diffusion within soil matrix. When modeling

flow interruption, convection is ignored and the dispersion coefficient D reduces to the

macropore diffusion coefficient, which is equal to molecular diffusion coefficient

divided by tortuosity. The experimental data needed to validate existing solute transport


62

models are limited and development of more refined approaches for modeling solute

preferential flow is also hindered due to the lack of experimental data (Butters and Jury,

1989).

The tillage effect on tracer (such as tritium) transport properties are fundamental

in understanding non-reactive solute behavior such as chloride, nitrate and other ions

which are not adsorbed onto soil organic matter and clays. Different solute application

methods have been investigated and the breakthrough curves (BTCs) were reported to

be dependent on the pulse concentration and pulse duration (Kluitenberg and Horton,

1990). However, little work has been done to examine the effect of flow direction on

preferential flow of solute in addressing the structuralization influence on solute

transport. Moreover, few attempts have been made on relating the flow paths to solute

BTC characterization and physical nonequilibrium solute transport o f the tillage effect

in undisturbed soil columns. This information is pertinent to tillage management

strategies as well as to the formulation and validation o f transport models describing

preferential flow.

The purpose of this study was (i) to examine flow paths and preferential

characteristics to determine the extent of heterogeneity of solute transport in undisturbed

columns under different tillage systems; (ii) to investigate the influence of pulse

duration and flow direction on tritium breakthrough in undisturbed soil columns of no

till and conventional tillage; (iii) to examine transport models for describing nonreactive

solute in undisturbed soil columns.


63

3.2 M aterials and M ethods

3.2.1 Transport Experiments

A hydraulic coring device mounted on a tractor was used to obtain 15 cm acrylic

cylinders (6.4 cm i.d.) o f the surface undisturbed soil in two tillage systems (no-till vs.

conventional tillage) of wheat (Triticum aestivum L .), hairy vetch (Vicia villosa Roth)

and no winter cover, in Gigger silty loam (silt mixed thermic Typic Fragiudalf). The

soil consists o f 22% sand, 66% silt and 12% clay. Six treatments including no-till with

wheat as a winter cover crop (NTW), no-till with hairy vetch as a winter cover crop

(NTV), no-till with no winter cover crop (NNC); conventional tillage with wheat as a

winter cover crop (CTW), conventional tillage with hairy vetch as a winter cover crop

(CTV) and conventional tillage with no winter cover crops (CNC) were studied in

duplicate. The field site was located at the LSU Agricultural Center, Macron Ridge

Research Station, Winnsboro, LA, and details were described in Chapter 2.

To eliminate the wall effect, the pedestal of soil was pushed out the column, and

then carved around to 3-4 mm, slightly smaller than column diameter. A large portion

o f pararifin wax was placed on an electric hot plate until it was fully melted (56°C),

and then filled the column annulus continuously. This wax coating was only performed

on 2 of 12 columns, namely NTV2 and NNC2. The undisturbed columns were

assembled with end plates and were equilibrated with C 0 2 for 30 minutes to improve

water saturation. Each soil column was saturated with a 0.005 M CaCl2 solution using

a stainless steel pump (piston and liner of variable speed pump, QG6 RH1, Fluid

Metering Inc., Oyster Bay, NY) which was adjusted for the desired velocity. After


64

steady flow was reached, the weights of saturated soil columns were recorded and a

small tritium pulse (duration o f 10 minutes) and high concentration (25.0 M BqL"1, i.

e ., 2 /tCi/mL) was applied from the bottom of the column (upward). After completing

the elution of the tritium pulse with 0.005 M CaCl2, the inflow tubing was switched to

allow the flow direction to be reversed. To carry out this switch in flow direction from

upward to downward, flow was interrupted by stopping the flow pump for a few

seconds. A second small tritium pulse with the same concentration and duration was

then applied, from the top to the bottom of the column (i.e., downward). After elution

o f the downward pulse, a long tritium pulse of about 1 pore volume with lower

concentration (0.35 M BqL'1, i.e ., 0.028 /uCi/mL) was subsequently applied in the

downward direction. This large pulse was thereafter eluted with 0.005 M CaCl2

solution. A 0.5 mL aliquot o f effluent solution was mixed with 5 mL of cocktail in a

6 mL plastic vial, were counted using Beckman LS 7500 Liquid Scintillation Counter

for counting 0 radiation o f tritium.

After completing the leaching of tritium, one pore volume pulse (initial

concentration o f C0 = 2 g L"1 prepared in 0.005 M CaCl2) of weakly reactive Eosin

Yellowish dye (Eosin Y; 2 ’,4’,5’,7’,-Tetrabromofluorescein Disodium salt:

C20H6Br4Na2O5, Mr691.88, color index 45380, solubility 100 g L '1, dark red) was

applied in the upward direction. This dye was used by Selsted and Becker (1986) for

reversible staining of peptides and proteins. Immediately after the termination of the

upward pulse, a downward pulse ( ® 1.0 V J of Brilliant Blue FCF, ( C j ^ g ^ N a O ^ ,

Mr787.91, color index 42090, solubility of 200 g L"1. k j of 0.19, --5.78 cm2 g '1) with


65

a concentration of 2 g L"1 was applied to the soil column in an attempt to stain the flow

path. After termination o f the displacement experiments, the columns were weighed and

put in a container for quick freezing with liquid N2. Columns was thereafter sectioned

with an electric saw at 2 cm intervals. The stained areas in each section were

photographed and drawn and the stained pattern were obtained. The dye concentration

in the effluent solution were also measured for 4 soil columns (NTV2, NNC2, CTV2,

and CNC2) with a spectrophotometer (HITACHI Model 100-20) at a wavelength of 630

nm for Blue FCF and 430 nm for Eosin Yellowish dye.

3.2.2 Breakthrough Curve Analysis

For non-reactive solutes, their transport through a homogeneous porous medium

can be written as

!>£=d ! ^ - v— (3.1)dt dx2 dx

where C is concentration (jig mL'3), x is depth below the soil surface (cm), t is time

(h), v is the average pore-water velocity (cm h '1), and D is the hydrodynamic

dispersion coefficient (cm2 h '1) which can be expressed as (Brusseau, 1993)

(3-2)r

where D0 is the molecular diffusion of solute in water (cm2 h"1), r is tortuosity, a is

dispersivity (cm) and v is pore water velocity (cm h '1).

The above equation (3.1) was subject to the initial condition

C(x,0)= q 0 < x < L (3.3a)


66

and the boundary conditions

v C -0 D ^ = v C o; x=0, t< tp (3.3b)

vC -0D — = 0, x=0, t> tD (3.3c)dx y

*£{L ,t)=0 (3.3d)dx

where C} is the initial concentration of solute in soil (mg kg"1 soil), L is the column

length (cm), CD is the pulse concentration (mg m L '1), tp is pulse duration (h) and 6 is

soil moisture content (cm3 cm"3). Equation (3.1) with boundary conditions (3.3a) to

(3.3d) were solved numerically. The relative pore volume (T =V /V 0) and relative

concentration C/C0 were used to express the results of BTCs. Equation (3.1) was fitted

to the breakthrough data to obtain the hydrodynamic dispersion coefficients by use of

non-linear least inversion method of Parker and van Genuchten (1984).

3.2.3 Superposition of Short Pulses

For input concentration of a short pulse (with duration 5 pore volume), one can

express the pulse concentration as

Ij(t)—C0j, 0 < t < 5

= 0 , tS:5 (3.4)

where Ij is the pulse with C0j as a constant concentration, and resulting relative

concentration as Cj/C0j. For the input having n tritium pulses at x = 0 , each o f duration


67

5, then the relative concentration at the effluent (x=L ) becomes

ac^tcfi-v-mic0j <3-5>

Since the relative effluent concentration at any time t is the result of the total pulses

occurring up to that time but not subsequent to it, equation (3.S) must be modified as

C lC ^ C f i - d - m /C ^ (3-6)

the time t can be taken as a discrete variable taking value at 8 time interval. The

experimental data of short pulse with duration 8 ( * 0.015 VQ) were superimposed into

duration t ( » 1.0 VQ) to calculate the BTCs of long pulses. The principle of

superposition can thus be used to characterize the behavior of soil column by unit (pulse

duration 5) BTC and to predict the response BTC of an arbitrary input (pulse duration

t). BTCs of long pulse were predicted with small pulse at the same experimental

condition using the principle of superposition.


The experimental conditions for the transport experiments are given in Table

3.1. The pore-water velocity v, was taken as the flux (cm/h) divided by saturated

moisture content 0 (cm3/cm3). The saturated moisture content were higher in soil

columns of no-till ( * 0 .5 cm3/cm3) compared with that o f conventional tillage (* 0 .4 6

cm3/cm3).


68

3.3.1 The Flow Paths o f Undisturbed Columns

The advantage o f using dyes is that one can relate observed preferential flow to

structural features or macropores in soil. The stained areas and flow paths of three

cross sections at 4, 8 and 12 cm below the soil surface o f each column are shown in

Fig 3.1. and Fig. 3.2. They indicate that there were considerable unstained areas for

the no-till soils whereas the soils were somewhat uniformly stained in soils of

conventional tillage. The only exception is the conventional tillage with no cover crop

(replicate 2, i.e. CNC2) where significant wall effects were observed. These results

revealed that water and solute was transported only through a portion of the pore space.

Li and Ghodrati (1995) reported that only a small fraction o f dye penetrated the soil

matrix while the major fraction of dye entered through earthworm holes at the surface.

In contrast, conventional tillage homogenizes and destroys part o f the structure of the

top soil and continuous macropores are disrupted. To characterize preferential flow in

undisturbed soil columns, possible wall effects should be eliminated. Attempts were

made for two columns (NTV2 and NNC2) to eliminate the wall effect. As shown in

Fig. 3.1, there was no dye flow through the wall in the columns sealed by wax. It

indicates the wall effect was partially eliminated therefore preferential flow paths can

be more clearly demonstrated with the stained areas o f the two dyes. The paths were

irregular in soil of no-till. In the cross sections near the soil surface, the entire area was

stained which indicate that the dye diffused and connected the flow paths; whereas for

cross sections at lower depths, a discontinuous flow paths were observed which is

indicative o f preferential flow. Preferential paths can develop as a consequence of


Reproduced with

permission



without permission.

Table 3.1 Experimental condition in tritium and dye transport study for soils from different tillage treatments

--------------------------------- Tritium®------------------------------------- Dye§-------Short pulse Duration Long Pulse Duration Eosin Y Blue FCF

Column"1"Number

9 0 v Upwardv/vc$

DownwardV/V0S

Upwardv/vc tp(h)

Downward V/VG tp(h)

R $ $ R$$

CTW1 1.47 0.468 1.33 0.0148 0.0148 - - 1.108 12.5 - -

CTW2 1.39 0.475 0.98 0.0109 0.0109 1.489 21.9 1.471 21.5 - -

CTV1 1.44 0.462 1.35 0.0150 0.0150 - - 1.125 12.5 - -CTV2 1.47 0.496 0.94 0.0105 0.0105 1.364 21.4 1.331 21.1 1.43+0.02 1.33+0.03CNC1 1.45 0.457 1.36 0.0151 0.0151 - - 1.133 12.5 - -CNC2 1.46 0.464 1.00 0.0110 0.0110 1.339 20.7 1.348 21.3 2.05+0.31 1.72+0.08NTW1 1.40 0.509 1.21 0.0134 0.0134 - - 1.008 12.5 - -NTW2 1.43 0.495 0.94 0.0100 0.0100 1.654 25.1 1.646 25.1 - -NTV1 1.41 0.507 1.22 0.0136 0.0136 - - 1.017 12.5 - -NTV2# 1.43 0.503 1.08 0.0133 0.0135 - - 0.699 12.6 1.69+0.05 1.96+0.06

0.301 5.40 - -NNC1 1.42 0.501 1.23 0.0137 0.0137 - - 1.025 12.5 - -

NNC2 1.40 0.510 1.22 0.0137 0.0140 - - 1.434 18.0 1.47+0.05 1.62+0.13

+CNC, conventional tillage-no cover; CTV, conventional tillage-vetch; CTW, conventional tillage-wheat; NNC, no-till-no cover; NTV, no-till-vetch; NTW, no-till-wheat.p is bulk density o f soil (g cm'3); 0 is soil moisture content (cm3 cm'3); v is pore water velocity (cm h*1); V/VQ is pore volumes o f pulse duration; Ip is the time o f pulse duration. NTV2# is the column with flow interruption.® Retardation factor was set to 1.0 in fitting tritium BTCs.5 10 min duration o f tritium input; § 15 hours duration for dye input.** R = fitted retardation factor (see solid and dashed lines o f Fig. 3.3)

70

Column N7W2 NTV2

Z = 4 cm

8 cm

12 cm

NNC2

Blue Red Unstained

Fig. 3.1 The stained areas by red and blue dyes for undisturbed soils from no-till with wheat (NTW2), vetch (NTV2) and no winter cover crop (NNC2) at 4, 8 and 12 cm depth below the soil surface.


Fig. 3.2

Blue Red UnstainedThe stained areas by red and blue dyes for undisturbed soils from conventional tillage with wheat (CTW2), vetch (CTV2) and no winter cover crop (CNC2) at 4, 8 and 12 cm depth below the soil surface.


72

oo

oo

1

0.8

0.6

0.4

0.2

0

1

0.8

0.6

0.4

0.2

0

a) Column NTV2

Red ° Measured— Fitted

Blue * Measured- - Fitted

° ° ° O o oc) Column CTV2 °0oo

’ Red ° Measured— Fitted

. Blue * Measured- - Fitted

0.5 1 1.5■" 'P W U U f l o

2.5

1

0.8

0.6

0.4

0.2

0

b) Column NNC2 ° c Red 0 Measured

— Fitted Blue • Measured

- - Fitted

• / • /

/

0.5 1 1.5

j . /0 0.5 1 1.5

V/Vn2.5

2.5

d) Column CNC2 ,0°,

° Measured— Fitted

• Measured- - Fittedo

oo

0.4

0.2

0.5 2.5V/V0

Fig. 3.3 The BTCs o f dye transport in undisturbed columns of (a) no-till with vetch as a winter cover crop (NTV2); (b) no-till with no winter cover crops (NNC2); (c) conventional tillage with vetch as a winter cover crop (CTV2) and (d) conventional tillage with no winter cover crop (CNC2).


73

0.04

0.03

a) NTW1 • Short PulseUpward 0 Measured

- - - CDE Fitted Downward . Measured

— CDE Fitted

0.01

0.04b) NTW2 - Short Pulse

Upward 0 Measured0.03 . - - - CDE Fitted

F Downward • Measured1 — CDE Fitted

0.02 11°O

0.01i V

iL

0 5 ...........

0.04

0.03

o0.02

O

0.01

c) CTW1 • Short Pulse0.04

d) CTW2 • Short PulseUpward ° Measured Upward 0 Measured

- - - CDE Fitted 0.03 - - - CDE FittedDownward • Measured Downward • Measured

& — CDE Fitted•Jh — CDE Fitted

fef 0.02 .f \ OHfb

' I \ 0.01g k

t V n .0 1 2 3 4 5 5 1 2 3 4

< < O v/v0

Fig. 3.4 Short pulse BTCs of tritium in undisturbed soil columns of (a) and (b) no-till; (c) and (d) conventional tillage with wheat as a winter cover crop.


74

0.04

0.03

OO 0.02

0.01

0.04

a) NTV1 - Short pulse

Upward 0 Measured - - - CDE Fitted

Downward • Measured — CDE Fitted

oo

c) CTV1 • Short Pulse

0.03

g 0.02

Upward o Measured - - - CDE Fitted


0.01

OD

0.04

0.03

0.02

0.01

0

0.04

0.03

0.02

0.01

b) NTV2 - Short Pulse

Upward Measured- - - CDE Fitted

Downward . Measured

.CDE Fitted

■ AAJ . __3 1 2 3 4 5

d) CTV2 - Short PulseUpward ° Measured

— CDE FittedDownward > Measured

.CDE Fitted

/ \ w _V/Vn

Fig. 3.5 Short pulse BTCs of tritium in soil columns o f (a) and (b) no-till, (c) and (d) conventional tillage with vetch as a winter cover crop.


75

0.04

0.03 •

0.01

a) NNC1 - Short Pulse

Upward ° Measured - - - CDE Fitted


0.04

0.03

0.02

0.01

b) NNC2 • Short Pulse

Upward 0 Measured CDE Fitted


0.04

0.03

OO 0.02

0.01

c) CNC1 - Short Pulse0.04

d) CNC2 - Short PulseUpward ° Measured

CDE Fitted0.03

Upward ° Measured

Downward . Measured CDE Fitted

Downward .CDE Fitted Measured

_ CDE Fitted

e #o °a

0.02 •

6 o t>

0.01 ' A .I s \ £ V—■ i 0

V/Vn V/Vn

Fig. 3.6 Short pulse BTCs of tritium in soil columns o f (a) and (b) no-till, (c) and (d) conventional tillage with no winter cover crop.


76

a) NTW1 - Long Pulse

• Downward — CDE

Superposition

* c) CTW1 • Long Pulse

Downward — CDE

Superposition

oo

1

0.8

0.6

0.4

0.2

0

OU

b) NTW2 ■ Long Pulse

U ° Measured/ J CDE

/ 'i Superposition

■ 1 I t \ \ \ D * Measured

■'tA -CDE

J o V 1°‘ «*/ l a \\\ ’ ’

Superposition

* /« (ff 1° 1\

■iti t \ s

'

0 1 2 3 4

d) CTW2 • Long pulse U ° Measured

— CDE - - - Superposition

D • Measured - - CDE

Superposition

Fig. 3.7 Long pulse BTCs of tritium in soil columns o f (a) and (b) no-till, (c) and (d) conventional tillage with wheat as a winter cover crop.


77

oo

a) NTV1 - Long Pulse Downward

— CDESuperposition

b) NTV2 - Long Pulse o Downward

CDE• - - - Superposition * • Flow Interruption

OD

1

0.8

0.6

0.4

0.2

0

OO

4 d FIOW nterrupti

. 2 d Flow Sj Interruption

c) CTV1 • Long Pulse1

d) CTV2-Long Pulser ' \ U ° Measured

a • Downward 0.8 k \ — CDEf i t — CDE f j i j - - - Superposition

/ *jt. - - - Superposition 0.6 u « D * Measured

/ Vo

O f V — CDE

/ V u f r ‘ ' Superposition1 U* 0.4

/ \

/ A0.2

■ / V

/ \_r! . . 0 m l- ------------------------------------------------------------------------

V/Vn v/vn

Fig. 3.8 Long pulse BTCs of tritium in soil columns o f (a) and (b) no-till, (c) and (d) conventional tillage with vetch as a winter cover crop.


78

oo

oo

1

0.8

0.6

0.4

0.2

a) NNC1 Long Pulse

• Downward — CDE - - - Superposition

Oo

c) CNC1 • Long Pulse

• Downward — CDE - - - Superposition

OD

(b) NNC2 - Long Pulse

• Downward — CDE

Superposition

d) CNC2 - Long Pulse U ° Measured

CDE - - Superposition

D • Measured - - CDE

Superposition

Fig. 3.9 Long pulse BTCs curves of tritium in soil column of (a) and (b) no-till, (c) and (d) conventional tillage column with no winter cover crop.


79

a) NTW1 ■ Long Pulse

0.6 • Measured- CDE ■ MIM— SM

0.6

0.4

0.2

1b) NTV1 • Long Pulse

0.8

• Measured— CDE - - MIM— SM

0.6OoD

0.4

0.2

c) NNC1 • Long Pulse

• Measured -• CDE• MIM — SM

0.8

0.6OoD

0.2

V/V0

Fig. 3.10 Modeling long pulse BTCs of tritium in soil column of (a) no-till with wheat as a winter cover crop, (b) no-till with vetch as a winter cover crop and (c) no-till with no cover crop with three models (CDE, MIM and SM).


80

abandoned root channels, burrows and fissures. On a very local scale, pore spaces

between soil structure peds may also be preferential flow paths.

The breakthrough curves of dyes are shown in Fig. 3.3. There was a higher

concentration o f dyes in effluent of no-till than that of conventional tillage, which can

be explained by the flow paths stained in Fig. 3.1 and Fig. 3.2. The fitted parameters

R is given in Table 3.1 and it ranged from 1.3 to 2.35 for both dyes in 4 columns

where the BTCs of dyes were measured. This indicated the dyes were retarded by the

soils. Flury and Fliihler (1995) reported the retardation was 1.2 for Brilliant Blue FCF

compared with Iodide. Andreini and Steenhuis (1990) found retardation factor of

Brilliant Blue FCF ranged from 1.5 to 7 in soil columns of a fine sandy loam.

3.3.2 BTC Characteristics o f Tritium

As shown from Fig. 3.4a to Fig. 3.9d, the breakthrough times were earlier in

soil columns of no-till than that of conventional tillage in all cover crops. One exception

was found in soil of conventional tillage with no cover (CNC2), where significant wall

effect was observed (Fig. 3.2) and an early (0.2 V0) tritium breakthrough was detected

in effluent. The tritium was detected after only 0.2 V0 following application in soils of

no-till. In contrast, the tritium was detected after 0.4 V0 for the conventional tillage.

All results consistently indicated no-till reduced breakthrough time. Singh and Kanwar

(1991) proposed the existence of rapid flow through macropores if the pore volume

reaches 0.5 relative concentration (C/C0) in the effluent was well before one pore

volume. Based on this criterion, preferential flow were dominant in all soil columns of

no-till. In addition to an early breakthrough, a sudden jump to peak concentrations was


81

followed by a precipitous drop to low concentration in all soil columns o f no-till. These

characteristics are similar to that of soils with earth worm holes as reported by Li and

Ghodrati (1995).

Some BTCs in short pulse from no-till columns showed double or bimodal peaks

in both upward and downward flow (Figs 3.4a, 3.5a, 3.6a and 3.6b). These bimodal

peaks were independent o f the direction o f flow. This is an indication o f the intrinsic

preferential flow paths in the undisturbed soil columns. The first peak is perhaps caused

by macropores and the second by the less permeable matrix-pore system. BTCs under

preferential flow conditions show consistently early breakthrough of solutes with

bimodal concentration peaks in other studies (Butters and Jury, 1989; Homberger et

al., 1990; Jury et al., 1986). Several investigators attributed preferential flow and

bimodal peaks to flow heterogeneities in soils (Ma and Selim, 1994; Skopp et al., 1981;

Homberger et al., 1990). Observed double peaks may be attributed to the presence of

multiple flow domains in some soils (Skopp et. al., 1981), although chemical

heterogeneities with reaction sites of different mass transfer rates and thus various

breakthrough time can also lead to double peaks (Villermanx, 1974). This is not the

case in this study since tritium is considered a non-reactive tracer. In soil with two

distinct flow domains, there may be two separate or distinct BTCs. One has shorter

travel time (preferential flow path), and another has a longer travel time. Ward et al.,

(1994) reported bimodal BTCs at three different values of fluxes. As flux decreased,

the relative amount of solute associated with the slow peak o f the BTC increased, while

the time to first arrival o f solute increased.


82

All BTCs in column of no-till showed some degree o f tailing o f the desorption

(elution) side (Figs. 3.4 - 3.8). Such tailing or asymmetry has been postulated by van

Genuchten and Wierenga (1976) to be due to the presence o f immobile water.

Convective solute transport in these models was assumed to occur only in the mobile

soils water region, while adsorption in a stagnant region of the soil which is controlled

by diffusion through the immobile (non-moving) fraction of the soil-water region.

Extreme tailing under saturated condition, is an indication o f cracks, or macropores in

soil. Solute transport between mobile (flowing) and immobile (stagnant) soil-water

regions is diffusion controlled. Thus excessive tailing and asymmetry indicate possible

existence o f nonequilibrium mass exchange between regions o f different mobilities.

Unlike the BTCs of soils from no-till, the BTCs of soils from conventional tillage

(Figs. 3.4 - 3.9) were relatively symmetric, indicating little or no preferential flow was

taking place in these columns. For the no-till with vetch as a winter cover crop in

replicate 2 (NTV2) shown in Fig. 3.8b, two flow interruptions were carried out during

tritium transport experiments. One is during the tritium pulse input (0.7 VQ), and

another was during leaching or elution at 2.7 V0. There was no concentration change

due to flow interruption, which is consistent with the results o f Koch and Fliiler (1993)

in packed porous beads. These observations contrasts to that of Oren et al. (1996)

where a significant change of outflow concentration was observed when flow was

interrupted in undisturbed soil columns. Both diffusion in macropores and diffusion into

micropores may explain the change, but the two processes may compensate each other

if the transfer in each direction is equal each other. The influence of the two diffusion


83

processes depend on the concentration gradients and the mass transfer coefficient

between macro and micropores.

To further examine the flow characteristics in the undisturbed columns of no-till

and conventional tillage, BTCs of the short pulse were used to predict the BTCs of long

pulse at the same experimental condition using principle o f superposition (Ma and

Selim, 1994). The predicted results are shown by dash curves in Fig. 3.7 - 3.9. These

BTCs were calculated using equation (3.7) by a repeated summation of several

successive short pulse BTCs until the desired pulse duration was achieved. The

superimposed BTCs were consistent with observed BTCs o f the long pulse (Figs. 3.7 -

3.8). Such predictions o f measured results are confirmation of the intrinsic nature of

the flow paths in the soil columns. Poor predictions for the two replications from no-till

with wheat cover crop (NTW1 and NTW2 of Fig 3.7) may be attributed to possible

changes in the flow paths during transport. Possible changes in the flow paths during

transport may have occurred in the column with cracks and channels as a result of

structural changes. Water-unstable aggregates may cause a change in flow paths due to

their collapse during water saturation. Additionally, macropore geometry may change

with time.

The direction of flow (i.e ., upward or downward) influenced the shape o f tritium

breakthrough results as shown in Figs. 3.4 -3.6. The BTCs in soils o f no-till were

distinctly different for each flow direction, which may be due to the structuralization

of soil of no-till. The direction o f flow had little effect on tritium breakthrough in the

soils of the conventional tillage. The only exception is that o f conventional tillage with


84

no winter cover crops (Fig. 3.6c). Conventional tillage homogenizes and destroys part

o f the structure o f the topsoil and continuous macropores are disrupted. Furthermore,

we are not aware o f previous studies where upward versus downward flow was

investigated for various soils from no-till and conventional tillage.

3.3.3 Modeling BTCs of Tritium

As shown from Fig. 3.4 - Fig. 3.9, the CDE (equation 3.1) adequately described

the tritium BTCs in soils under conventional tillage. The CDE did not adequately

describe the experimental data of soils from no-till, especially for the short pulse (Figs

3.4a, 3.5a 3.6a and 3.6b). This inadequacy is expected because the CDE (equation 3.1)

is based on assumption o f homogeneous pore distribution. However, the model

description was improved for the long pulse (Figs 3.7a, 3.8a and 3.9a). As indicated

by dye staining, the soil columns from no-till was heterogenous.

It is perhaps conceivable to describe such BTCs based on two region or two

domain flow approaches (Skopp et al., 1981), mobile and immobile model (MIM) as

proposed by van Genuchten and Wierenga (1976) and stochastic model (SM) as

proposed by Bresler and Dagan (1982). To test the physical nonequilibrium models in

describing tritium BTCs in undisturbed soil columns, the MIM and SM models of

Parker and van Genuchten (1984) were used to describe the long pulse BTCs of soils

from no-till (Fig 3.10). There was no improvement in data description. The same result

was reported by Hiroshi et al., (1994) where the BTCs of bromide were not adequated

described by either CDE or MIM models. Increasing the number o f model parameters

created more uncertainty o f parameter estimation. Li and Ghodrati (1995) did not obtain


a satisfactory fitting when using CDE, MIM, and SM for N 0 3‘ transport in soil

columns with earthworm macropores. They concluded that MIM did not seem to apply

to solute transport in soil columns containing earthworm macropores and the partition

of soil water into mobile and immobile is not sufficient to account for the range of pore

water velocities encountered. In addition, the CDE, MIM and SM could not describe

the double peaks of BTCs. Skopp et al. (1981) proposed a two-domain transport model

to explain early breakthrough and bimodal peaks in solute transport. Another alternative

is use a stochastic model (or transfer function model) to predict the bimodal BTCs by

superposition of two probability density function. Utermann et al. (1990) successfully

predicted double peaks using a bimodal probability density function.

3.4 Conclusions

The dye study indicated that the heterogeneous solute transport which

characterized the preferential flow in undisturbed soil columns from no-till with earlier

breakthrough time, longer tailing, and double peaks, compared with the BTCs from

undisturbed soil columns of conventional tillage. Such observations implies that

preferential flow was pronounced and thus enhanced the transport of non - reactive

solute in soils of no-till compared to that of conventional tillage. The flow direction

altered shape of tritium BTCs which is indicative to structural formation for the no-till.

Double peaks were observed for short pulses whereas excessive tailing was observed

for the long pulse in soil o f no-till, suggesting that the double peak effect was integrated

with pulse duration. The direction and the pulse duration had little influence on tritium

BTCs in soil of conventional tillage. The consistency between the experimental data and


86

the prediction o f BTCs of long pulse using superposition from short pulse experimental

data can be used as a confirmation in the flow paths o f soil columns. The preferential

flow in undisturbed soil columns may be attributed to macropores and cracks between

aggregates. The CDE for BTCs of soil columns from no-till where preferential flow is

pronounced, may not be strictly applicable, especially for the pulse o f high

concentration and short duration. Mobile-immobile transport model (MIM) and

stochastic model (SM) did not improve the fitting. Thus at present we are not aware of

modeling approaches that can predict solute transport in the no-till soils which displayed

preferential characteristics.


CHAPTER 4

MODELING ALACHLOR TRANSPORT IN SATURATED SOIL FROM NOTILL AND CONVENTIONAL TILLAGE SYSTEMS

4.1 Introduction

Alachlor, a pre-emergence herbicide, is one of most commonly used soil-active

herbicides. It has been extensively used to control annual grasses and certain broad leaf

weeds in com, cotton, soybean and sugarcane fields. About 37 million kilograms of

alachlor are produced annually in the USA (Chesters et al., 1989). According to the

criteria o f Green and Karickhoff (1990), alachlor is a nonionizable, moderately

nonpolar compound. It has water solubility of 148-242 mg L '1 at 25°C, and field half

life of 14-49 days (Wauchope et al. 1992). It has been considered a human carcinogen

(U.S. EPA, 1989). Because o f groundwater contamination and possible health effects,

EPA proposed zero fig L"1 for alachlor as a Maximum Contaminant Level Goal (U.S.

EPA, 1989). But so far, alachlor have been detected in ground water in many states

including Iowa, Pennsylvania, Maryland, Nebraska, and Minnesota etc. (Cohen, 1992).

Preferential transport has been cited as being responsible for the rapid movement

o f agricultural chemicals (Czapar et al., 1992). Preferential solute transport and water

flow in tilled and untilled soils has been reported by Andreini and Steenhuis (1990).

The authors used grid lysimeters to separate the water flowing through undisturbed soil

columns from conventional and conservation tillage systems both temporally and

spatially. They found that dye flow pathways led to areas where water and solutes

exited the columns. In the no-till column, nearly the entire depth of the profile was

87


88

short-circuited by preferential flow; but in the tilled column, dye passed uniformly

through the mixed, unstructured plow layer. Watson and Luxmoore (1986) found < 1 %

o f the soil volume was used in the transport of 96% of water flux in a forest soil.

Because of preferential flow, Isensee et al. (1990) detected alachlor in shallow

groundwater shortly after application in soil of no-till.

Although the formation of macropores and preferential flow may increase the

potential for herbicide leaching through the vadose zone, changes in the soil chemical

properties may enhance herbicide retention in the surface horizons o f no-till soils, thus,

higher concentration of the herbicides occurred in the leachates of conventional than no

till soils (Levanon et al., 1993). Adsorption reactions play a large role in the fate of

alachlor transport under different tillage systems. Increased sorption to soil organic

carbon may reduce leaching. Sorption provides a mean to quantify and characterize the

potential for alachlor binding and mobility in soil. No-till was reported to increase

organic carbon content and thus increase cation exchange capacity (Lai, et al., 1994).

Cover crop was reported to influence the soil organic carbon content and clover cover

caused significantly higher soil C levels compared with fallow, even after 1 year

(Kumwenda, et al., 1993). However, the greater amount of crop residue in the

conservation tillage systems did not appear to affect the amount o f alachlor retained on

the soil or its dissipation rate (Gaynor et al., 1987). Soil organic matter content is a

major factor for alachlor adsorption, and appears to occur primarily on the humic acid

and fulvic acid and to a lesser extent on clay colloids (Chesters et al., 1989). By

studying infrared spectra o f free alachlor and herbicide-clay complexes, Bosetto et al.


89

(1993) concluded that alachlor is adsorbed on homoionic montmorillonite by

coordination bond, through a water bridge, between the C = 0 group and the

exchangeable cation of the clay. Senesi et a l., (1994) studied the infrared spectra of

pure alachlor and adsorbed alachlor and indicated the interaction occurred between

C = 0 and C-N group by H-bound formation. Generally, the retention o f alachlor in soil

has been described by either (instantaneous) equilibrium (Locke, 1992) or kinetic

reactions whereby concentrations in solution and sorbed phases vary with time (Xue and

Selim, 1995).

There are limited data and contradictory conclusions regarding the fate of

agrochemicals and their transport in soils under different tillage systems (Levanon,

1993). The influence of tillage and cover crops on alachlor mobility are not well

understood. There is also a lack of knowledge of the mechanisms involved in rapid

transport processes of adsorbing chemicals. In alachlor investigations, the increased

demand for quantitative data has acted as a stimulant to the use of computational

models. These are needed to give the behavioral pattern o f a toxicological value. A

minimum requirement o f the model is that it must predict the concentration of alachlor

as a function of time and depth in the soil profile. The objectives of this study were (i)

to investigate alachlor mobility in soils of different tillage and cover crops based on

adsorption kinetics and (ii) to determine the influence o f tillage and winter cover crops

on alachlor mobility based on transport characteristics and parameters o f breakthrough

results.


90

4.2 Materials and Methods

A long-term study was initiated at LSU Agricultural Center, Macron Ridge

Station, Winnsboro, LA, in the Fall o f 1986 to study the effects o f several tillage

systems and winter cover crops on cotton (Gossypium hirsutum L .) cultivation. The soil

was a Gigger silt loam (fine silty, mixed, thermic Typic Fragiudalf). Among these, six

treatments were chosen for this study. We focused our investigation on conventional

and no-till tillage systems under three winter cover crop treatments namely; wheat

(Triticum aestivum L .), hairy vetch (Vida villosa Roth) and no cover crop.

4.2.1 Batch Experiments

Surface soil samples from six treatments were taken and passed through a 2 mm

sieve. Adsorption of alachlor in all surface soils were carried out using kinetic batch

methods. Alachlor adsorption experiments were commenced by mixing 10 g of soil and

20 mL o f different alachlor solutions in Teflon centrifuge tubes. Eight initial alachlor

concentrations (0.5, 1, 2, 5, 10, 20, 30 and 50 /xg mL"1) were used and four reaction

times (1, 2, 5, and 22 day) were studied in duplicates. At the end of each reaction time,

the samples were centrifuged for 10 minutes at 5000 ipm, the supernatants were

withdrawn, and filtered for analysis.

4.2.2 Miscible Displacement experiments

A Giddings hydraulic coring device mounted on a tractor was used to obtain

undisturbed soil column of 15 cm acrylic cylinders (6.4 cm i.d .) from the six

treatments. The acrylic cylinders were pushed into soil and then were carefully lifted

from the ground and the excess soil in the bottom of the cylinder was shaved off. The


91

columns were used for saturated flow experiments and direct parameter estimation. The

undisturbed columns were carefully assembled with glass beads on the soil surfaces to

ensure contact with both ends of column. The flow parts were connected with Teflon

tubes. The soil columns were firstly equilibrated with C 0 2 for 30 minute to replace the

air trapped in soils; then the columns were saturated using a 0.005 M CaCl2 solution

by adjusting the stainless pump (piston and liner of variable speed pump, QG6 RH1,

Fluid Metering Inc., Oyster Bay, NY) to a desired velocity until a steady flow velocity

was reached. One pore volume (V0) tritium pulse of 1.0 M Bq L '1 (0.028 /iCi mL'1),

followed by about 6 pore volume alachlor (10 mg L '1 in 0.005 M C a C y were applied

and then leached with 0.005 M CaCl2 solution for about 20 V0s. A 0.5 mL aliquot of

the supernatant was mixed with 5 mL of "ULTIMA GOLD" cocktail was mixed in a

6 mL plastic vial. The /? radiation o f tritium was measured by liquid scintillation

counter (Beckman LS 7500 Liquid Scintillation Counter). Alachlor concentration was

determined by high performance liquid chromatograph (HPLC) using a UV detector at

a wavelength of 220 nm. A Supeclco LC-18 reversed column (25 cm x 4 mm i.d.) was

used with 70:30 acetonitrile:water as the mobile phase at a flow rate of 1 mL m in'1.

The lower detection limit for alachlor was 5 fig L '1. The breakthrough curve (BTC)

were expressed as relative concentration C/CD vs. number o f pore volumes (T=V /V 0).

4.2.3 Equilibrium Retention Modelling

Linear and Freundlich models are the most widely used model for describing

herbicide adsorption. In the linear model


where S is adsorption (jig g"1 soil), kd is the concentration-independent distribution

constant (cm3 g '1) and C is the concentration at equilibrium (mg L’1). The partitioning

coefficient of organic matter was calculated as

1 ^ = 1 0 0 1 ^ /0 0 % (4.2)

where OC% is the soil organic carbon content expressed as a percent of the soil mass.

In the Freundlich model

S - k f N (4-3)

where N is nonlinear equilibrium parameter. kf is Freundlich distribution constant (cm3

g '1), C and S are as defined in linear model.

The adjusted r2 (radj2) were used to test the goodness-of-fit of models and was

calculated as

p. t _ (*-!)(!-r2) (4.4)J (n-p)

where n is the sample size and p is the number o f parameters in the model, r is

correlation coefficient.

The adsorption data for one day and 22 day were fitted using the NONLIN

procedure o f SAS (1985) to obtain equilibrium parameters o f linear and Freundlich

models.


93

4.2.4 Multireaction Modeling

The following multireaction model was presented by Selim (1992),

Se=keC N (4-5)

(4-6)dt l p 2 1 3 1

(4-7)dt 3 1

Sr =Se+S1+S2 (4.8)

where C is the solute in soil solution (mg L '1), p is the soil bulk density (g cm"3), 6 is

the water content (cm3 cm '3), t is the reaction time in hour, kg is distribution constant

(g cm '3), k ls k2 and k3 are associated rate coefficient (hr'1), Se is the amount retained

by equilibrium sites, is the amount retained by kinetic sites, S2 represented amount

retained by irreversible sites and ST is the total amount of solute retained by the soil

matrix (jxg g '1 soil). The parameter N was obtained from fitting adsorption data after

22 days of reaction using the Freundlich equilibrium equation (4.5). The above

reactions (4.5-4.8) represent initial-value problems that were solved numerically using

finite-difference approximations.

4.2.5 Multireaction Transport Modeling

The BTCs for reactive solutes are usually described using the convective-

dispersive transport equation (Selim, 1992)


where D (cm2 h*1) is hydrodynamic dispersion coefficient as expressed by Brusseau

(1993)

D=— +av (4.10)r

where D0 is the diffusion in water (cm2 h '1), r is tortuosity (dimensionless), a is

dispersivity (cm) and v is pore water velocity (cm h '1), p is bulk density (g cm'3), 8 is

volumetric saturated moisture content (cm3 cm'3), S is concentration of the solute in the

adsorbed phase (pg g '1 soil), C is the solute concentration in the liquid phase Qxg mL'

J), and z is depth (cm).

The initial condition used was

C(x,0) = Cj 0 < x < L (4.11a)

and the boundary conditions were

v C -8 D ^ = v C 0; x=0, t< tp (4.11b)

vC -8D ^£= 0, *=0, t> tD (4.11c)dx y

i£ (L ,/)= 0 (4. l id )dx

where Cj is the initial concentration of solute in soil (mg kg '1 soil), L is the column

length (cm), C0 is the pulse concentration (mg L"1), tp is pulse duration (hrs).


95

Dispersion coefficients was estimated based on tritium BTC data using nonlinear least-

squares inversion methods (Parker and van Genuchten, 1984). The above equations

were solved using explicit-implicit finite difference approximations.


4.3.1 Soil Properties

Selected soil properties for the surface soil from different tillage systems are

showed in Table 4.1. The average organic matter content was 1.37±0.439 in soils of

no-till and 0 .96±0.348 in that of conventional tillage. There was a significant (p=0.05)

difference using t-test, which indicated no-till increased organic matter content. It is

consistent with the report o f Lai et. al. (1994). However, the soil pH was not affected

by tillage which is consistent with report of Lai et. al. (1994).

4.3.2 Modeling Alachlor Adsorption

The goodness-of-fit for the linear and Freundlich equilibrium model after one

day and 22 day adsorption are listed in Table 4.2. In all tillages and cover crops, the

râdj were higher for Freundlich model compared with the linear model especially for

the 22 day adsorption. For all soils studied, the nonlinear parameter N of Freundlich

model varied from 0.6532 to 0.9353. That N was less than unity indicated availability

of sorption sites decreased as initial concentration increased. The Freundlich model

accounts in a certain way for surface heterogeneity. The alachlor adsorption isotherm

under no-till and conventional tillage under wheat as a cover crop is shown in Fig. 4.1.

It is obvious that Freundlich model improved the fitting of the data compared

with the linear model.


96

Table 4.1 Selected soil properties for the surface soils from different tillage and winter cover crops

Sample§ Organic matter content

%

PH

CTW 0.81 4.7

CTV 0.98 4.9

CNC 1.09 4.2

Conventional Tillage 0.96* 4.6

NTW 1.45 5.7

NTV 1.48 4.4

NNC 1.16 5.5

No-Till 1.36* 5.2

t 3.093* 1.32

t (p=0.05,n==4) 2.776 2.776

§ CTW, conventional tillage-wheat; CTV, conventional tillage-vetch; CNC, conventional tillage-no cover; NTW, no-till-wheat; NTV, no-till-vetch; NNC, no-till-no cover.* These values are significantly different at the 0.05 level using t-test.


Reproduced with

permission



without permission.

Table 4.2. The goodness-of-fit and parameters o f equilibrium and multireaction model in fitting alachlor adsorption in soils o f no-till and conventional tillage under different cover crops.

-------------- ------------ 1 day— — 22 day- --------------- ------- Kinetic-----------

(Linear) (Freundlich) (Linear) (Freundlich) (MRM)

Sample§ âdj N r2adj âdj fcf N âdj âdj

CTW 0.959 1.04 0.763 0.985 0.936 2.81 0.653 0.999 0.9997

CTV 0.997 0.50 0.935 0.999 0.951 1.80 0.704 0.989 0.9994

CNC 0.975 0.99 0.834 0.987 0.935 2.00 0.690 0.993 0.9994

NTW 0.980 1.64 0.796 0.999 0.950 4.10 0.689 0.995 0.9988

NTV 0.988 1.74 0.853 0.998 0.986 2.28 0.824 0.999 0.9998

NNC 0.966 1.87 0.794 0.986 0.929 3.61 0.665 0.996 0.9986

§ CTW, conventional tillage-wheat; CTV, conventional tillage-vetch; CNC, conventional tillage-no cover; NTW, no-till- wheat; NTV, no-till-vetch; NNC, no-till-no cover.

vo<1

98

Fig.4.1

a) Isotherm-1 d • CTW Data ° NTW Data

Freundlich Linear

P 30

30

crQx 20 o5

Q 10LUCDDC

8 0

b) Isotherm-22 d • CTW Data° NTW Data— Freundlich-- Linear

20C (mg/L)

Alachlor adsorption isotherms for (a) one day adsorption and (b) 22 day adsorption o f soils from no-till (NTW) and conventional tillage (CTW).


99

The fit o f the multireaction model (MRM) to the batch results of a wide range

o f initial concentrations (C0 = 0.5 - 50 mg L '1) was achieved using the four parameter

model (kg, k 1? k2 and k3) consisting of one equilibrium reaction, one reversible reaction

and one irreversible reaction. The goodness-of-fit for MRM are listed Table 4.2. For

all soils, the r2^ were higher than respective equilibrium models in soils of all tillages

and cover crops, which indicated that multireaction model improved the fitting

compared equilibrium models, especially for long adsorption times. Model parameters

for MRM are listed in Table 4.3. The nonlinear parameter N in equation was based on

the equilibrium Freundlich parameter N of 22 days (Table 4.2). Concentration as a

function of time for no-till and conventional tillage under different cover crops are

presented from Fig.4. 2 to Fig. 4.4 for various C0s. Retention o f alachlor by soil was

rapid during initial stages o f reaction and was then followed by slow and continuous

alachlor retention. The solid lines are model fitting using MRM. The reversible kinetic

reaction constants kj and k2 were significantly different from zero except in soil of

conventional tillage of vetch as winter cover crop. Slow diffusion to sites within the

soil/organic matrix probably contributed to the apparent nonequilibrium behavior.

4.3.3 The Tillage Effect on Alachlor Adsorption

The overall contribution from both organic and inorganic sorbents is often

described with a distribution coefficient of herbicide partitioning between solution and

solid phase (kj). As showed in Table 4.3, the equilibrium constant kd, in all soils of no

till were significantly (p=0.05) higher than that of soils from respective conventional


c (m

g/L)

100

b) CTWa) NTW

Co - 60 mg/L

Co * 50 mg/L

'600400200'600400200TIME (h)TIME (h)

Fig.4.2 Alachlor concentration as a function o f time for (a), no-till and (b) conventional tillage of wheat as a winter cover. All symbols represented measured concentration and all lines represented fitted concentration.


C (m

g/L)

101

46a) NTV

45b) CTV

40 40

35 35 Co - 50 mg/L

30 ■ ■Co - 50 mg/L 30 ■

25 25

20 20---------------------------- 1 30

15i 30

15 ------------------------20

10V

10• 10

5 ' 5■a u - ■ 5

n ~*~i a ■ a s s 2 1 n2

° 0 200 400 '600 ° 0 200 400 '600TIME (h) TIME (h)

F ig .4 ,3 Alachlor concentration as a function o f time for (a), no-till and(b) conventional tillage of vetch as a winter cover. All symbols represented measured values and all lines represented fitted values respectively.


C (m

g/L)

102

45

40

35

30

25

20 Nv15

a) NNC

v .10

5

0200

— . J

Co ■ 50 mg/L

" • 10

400

45

40

35

30

25

20

15

10

5

“600 ° 0

30

20

b) CNC

j —

•

Co • 50 mg/L

v .

V

* 30

“ V *u

• 1ft

--------------------------L* ,

200 400 '600TIME (h) TIME (h)

Fig.4.4 Alachlor concentration as a function o f time for (a), no-till and (b) conventional tillage o f no winter cover. All symbols represented measured concentration and all lines represented fitted concentrations.


103

tillage. There were no significant differences within the same tillage, under different

cover crops. As shown in Table 4.3, the equilibrium adsorption constants ke of the

multireaction model in all no-till treatments were significantly (p=0.05) higher than that

o f conventional tillages. There was no clear trend on kg by winter cover crops. The

influence of tillage and winter cover crop on kinetic parameters kj and k2 were not

consistent. The irreversible reaction constant k3 were relatively small and there was no

consistent trend on tillage and winter cover crops. The no-till increased alachlor

adsorption can be clearly illustrated by C vs. time in Fig. 4 .2 , Fig.4.3 and Fig.4.4,

where alachlor concentration were lower in soils of no-till than that o f the respective

conventional tillage in all cover crops and different concentrations.

Alachlor sorption has been associated with both hydrophilic and lipophilic

(hydrophobic) soil components and thus can be characterized by sorption per unit of

organic carbon (k ^ , Locke, 1992). As shown in Table 4.3, the k ^ was relatively stable

compared with k j in soils o f both no-till and conventional tillages. No-till increased the

k ^ in vetch and no cover crops compared with conventional tillage. There were no

significant difference in k ^ between no-till and conventional tillage with wheat as the

cover crop. Therefore no consistent influence on k ^ was resulted from tillage

treatments. This indicated that increases in organic matter content and/or changes in its

composition contributed to increasing alachlor adsorption. As shown in Table 4.3, k ^

was variable and ranging from 90.6 to 143.4 for the soils studied.

The k^. characterizes partitioning between aqueous solution and organic phases

in soil, but does not take into account composition characteristics of the organic phase,


Reproduced with

permission



without permission.

Table 4.3 Equilibrium and kinetic parameters for different tillages and winter cover crops in batch experiments

Sample® koc— (cm3 g '1)—

K *1 hh '1

k3

CTW 0.47+0.06 99 .2+ 9 .2 0.926+0.278 0.0130±0.0090 0.0137±0.0107 0.0001 ±0.0001CTV 0.55+0.02 95.3+3.1 0.896+0.312 0.0022±0.0088 0.0160±0.0404 0.0002 ±0.0001CNC 0.57±0.06 90 .6+ 9 .2 0.638+0.602 0.0318+0.0171 0.0693+0.0235 0.0001 ±0.0001NTW 0.92+0.08 109.6±9.9 2.330±0.208 0.0043 +0.0030 0.0059+0.0048 0.0000±0.0004NTV 1.10+0.07 129.7±8.5 1.784 ±0.146 0.0034+0.0027 0.0162+0.0175 0.0004 ±0.0002NNC 0.96+0.12 143.4+7.1 2.480+0.582 0.0087+0.0071 0.0448+0.0290 0.0001 ±0.0001

wheat; NTV, no-till-vetch; NNC, no-till-no cover; is Freundlich parameter for the equilibrium model; k^. is obtained from k j values divided by organic carbon content; kg is distribution constant, kl9 k2 and forward and backward rate coefficients for Sj, k3 is rate coefficient for consecutive irreversible reaction from Sj to S2.

2

105

which can vary widely among soils (Locke, 1992). Thus, organic composition could be

responsible for the widely varied values reported in the literature, e.g, 170 by

Wauchope et al. (1992) and 342 as by Locke (1992). As result o f its hydrophobic and

hydrophilic nature, alachlor sorption is significantly correlated to organic matter and

clay (Weber and Peter, 1982). Soil organic-matter content was one o f the soil

characteristics most highly correlated with adsorption o f alachlor. Soil organic matter

content increases and its distribution in the soil profile is altered in no-till. Plow-tillage

buries organic debris, whereas that which accumulated on the surface o f no-till soil may

reduce alachlor mobility.

4 .3 .4 Hydrodynamic Dispersion Coefficients

The BTCs of tritium and alachlor transport in soil columns of different tillage

systems are showed from Fig. 4.5 to Fig 4.10, with their corresponding experimental

conditions given in Table 4.4. Two replicates were carried out for each treatment.

Tritiated water was employed as a conservative tracer to investigate the hydrodynamic

dispersion in each undisturbed column. The details o f tritium transport study was

reported in Chapter 3, and the hydrodynamic dispersion coefficients obtained from long

pulse (duration about 1.0 V J were used in describing alachlor transport as described

in later sections. W e found good agreement among replicated columns for tritium BTCs

except CNC2 where significant wall effect was observed in dye experiment. The

hydrodynamic dispersions D for each different column are given in Table 4.4. There

was a tendency of a larger D values in all soils o f the no-till than that o f respective


106

Table 4.4 Experimental condition and parameters in transport study for different tillages and winter cover crops.

Sample§ P e V v/v0 *P D(g cm'3) (cm3 cm"3) (cm h"1) (b) (cm3 h"1)

CTW1 1.47 0.468 1.33 6.65 75 1.62±0.09

CTW2 1.39 0.475 0.98 4.71 75.3 1.06±0.16

CTV1 1.44 0.462 1.35 6.75 75.0 1.82±0.23

CTV2 1.47 0.496 0.94 4.80 75.8 1.06±0.14

CNC1 1.45 0.457 1.36 6.80 75.0 0.92 ±0.12

CNC2 1.46 0.464 1.00 5.57 80.7 9.00±0.40

NTW1 1.40 0.509 1.21 6.05 75.0 14.6±2.6

NTW2 1.43 0.495 0.94 5.29 82.13 19.0±1.5

NTV1 1.41 0.507 1.22 6.10 75.0 10.36±2.7

NTV2 1.43 0.503 1.23 6.76 81.0 1.41 ±0.32

NNC1 1.42 0.501 1.23 6.15 75.0 1.53 ±0.12

NNC2 1.40 0.510 1.22 4.90 58.7 0.91 ±0.25

§ CTW, conventional tillage-wheat; CTV, conventional tillage-vetch; CNC, conventional tillage-no cover; NTW, no-till-wheat; NTV, no-till-vetch; NNC, no-till-no cover; p is bulk density; 6 is moisture content; v is pore water velocity; V/V0 is the pulse duration in pore volume; tp is the pulse duration in hours; D is hydrodynamic dispersion coefficient.


107

conventional tillage. This is consistent with results from Singh and Kanwar (1991)

where D values for no-till columns were about 2.5 times higher than average values

from conventional tillage. Wheat as a winter cover crop in no-till (NTW) increased D

compared with conventional tillage and had the highest D value among all tillage

treatments. This is probably because of root channels which existed in the undisturbed

column. Experiments conducted by Gish and Jury (1982,1983) showed that wheat roots

increased D values by 3 times. The dispersion coefficient is related to molecular

diffusion and flow velocity as defined in equation (4.10). Molecular diffusion is much

smaller than dispersion and usually ignored in saturated flow condition with relatively

high velocity. Therefore the dispersion coefficient obtained from tritium BTC can be

used to characterize the solute dispersion. Dispersion is a result of differences in flow

rates in the various portions of the soil pore space. The flow in each portion o f the soil

pore space and its connectiveness to others determine the flow characteristics and the

extent o f dispersion in soils. As discussed in chapter 3, there was preferential flow

paths in soils from no-till. When preferential flow exists, the dispersion coefficient will

increase (Skopp and Gardner, 1992; Li and Ghodrati, 1995).

4.3.5 Alachlor BTCs in Soils from Different Tillage Systems

As shown from Fig. 4.5 to 4.10, there was considerable retardation of alachlor

BTCs compared with tritium. The earlier breakthrough and longer tailing were observed

in soils from no-till compared with that of conventional tillage, where higher peak

concentrations were also observed (Fig 4.5 to Fig 4.10). Alachlor BTCs in soils o f no

till were more asymmetrical than that of soils from conventional tillage. The early


breakthrough and asymmetrical shape of BTCs may be attributed to preferential flow

in macropores and large dispersion (van Genuchten and Cleary, 1979). The early

breakthrough and tailing are characteristics of physical nonequilibrium (Brusseau et al.,

1989). I f one assumes a wide range of pore size distribution in the no-till soil, a solute

BTCs is more likely to be asymmetrical along with large dispersion. The tailing may

also be a result of alachlor in stagnant regions of soil transversely diffused back into

accessible or the mobile phase of the dynamic soil region. The early breakthrough in

soils from no-till compared with that from conventional tillage is consistent with several

reports (Hall et al., 1989; Pivetz and Steenhuis, 1989; Isensee et al., 1990; Gish et al.,

1991; Isensee and Sadeghi, 1994). However, in other studies, there was no apparent

effect of tillage on herbicide nonequilibrium transport (Starr and Glotfelty, 1990;

Steenhuis et al., 1990; Gaynor et al., 1992; Ghodrati and Jury, 1992; Gaynor et al.,

1995). In uniformly packed soil columns, Gaston and Locke (1994) observed a lack of

physical nonequilibrium transport and greater retardation for alachlor in soils o f no-till

than conventional tillage.

4,3 .6 Transport Modeling

The asymmetrical shape of tritium BTCs indicates that there was physical non

equilibrium in the transport of solutes in these undisturbed soil columns. Although the

columns may be considered heterogenous, especially for soils of no-till, this

heterogeneity can be assumed to be random in space, and dispersion may account for

such heterogeneity. As discussed in Chapter 3, there was no improvement using

physical nonequilibrium models to describe the tritium BTCs compared with using the


109

classical CDE. In addition to physical nonequilibrium, Gaston and Locke (1994) found

chemical nonequilibrium as the dominant transport mechanism of alachlor in uniformly

packed soil columns where physical nonequilibrium was assumed absent.

Although nonequilibrium sorption of organic compounds is often formulated in

terms of chemically based retention models, such models have been shown (Nkedi-kizza

et al., 1984) to be equivalent to simplified, physical models based on diffusive mass

transfer (e.g., the two-region model o f van Genuchten and Wierenga, 1976). The

inclusion of a rate coefficient was adequate for characterizing time-dependent sorption

and diffusion-controlled transport (e.g. Rao and Davidson, 1979). The multireaction

transport model (MTRM) o f Selim (1992) with one equilibrium reaction, one kinetic

reversible reaction and one consecutive irreversible reaction was used in describing the

alachlor transport (equation 4.5 - 4.8). The diffusion controlled adsorption can be

accounted for by rate coefficients of different sites in the multireaction model. This

model has been found to be most successful in describing the adsorption-desorption

kinetics of alachlor retention as described in Chapter 2 (Xue and Selim, 1995). Selim

et al. (1976) found that MRTM with one equilibrium and one fully reversible processes

improved predictions o f the excessive tailing o f the desorption or leaching side and the

sharp rise of the adsorption side of BTCs for picloram in comparison to single-reaction

equilibrium or kinetic models.

Attempts were made to describe the transport of alachlor in the various columns

based on the multireaction model (MRTM) described in equation 4.9-4.11. Hence, we

used parameters which were independently obtained from the retention batch results


110

discussed in the previous section as well as in chapter 2. The specific MRTM model

version used for transport prediction included one equilibrium, one kinetic reversible,

and one consecutive irreversible reactions. Specifically model parameters used are given

in Table 4.3. The predicted BTCs are shown in Figs. 4.5a - 4.10b. The correlations

coefficients (r2) and mean square error (MSE) between the prediction and experiments

are listed in Table 4.5 o f this Chapter and Table 2.2 in Chapter 2. These predictions,

in most case, were considered not adequate. The r2 was 0.8917 for the NTW2 to

0.3815 for CTW2; the MSE was 0.1426 for NTW2 to 0.3141 for CTW2.

Unsatisfactory predictions o f alachlor BTCs using parameters from batch data have been

reported in several other studies. There are currently no models available that can a

priori predict leaching processes in soil under a variety of soil conditions without some

calibration or fitting. For example, Gaston and Locke (1994) found there were

considerable discrepancy between predicted and measured alachlor BTCs using two-site

and three site equilibrium/kinetic sorption models. They used parameters which were

independently derived from batch experiments. In some cases, pesticide leaching in the

field was overestimated when parameters based on predictions from laboratory

measurements were used (Rao et al., 1974; Burgard et al., 1993; Yen et al., 1994).

Rao et al., (1974) explained this behavior due to micropore retention through saturated

soil columns. However, in other cases, solutes moved deeper in soil profile than

predicted.

Since the multireaction model was not successful in predicting alachlor BTCs

when independently derived batch parameters were utilized, the capability o f MRTM


I l l

in describing the BTC in a calibration mode was evaluated. Specifically, the MRTM

was used along with a nonlinear parameter optimization scheme of Marquart (1963) to

describe the alachlor BTCs. Calculated (fitted) BTCs are shown in Figs 4 . 6 - 4 . 1 1 . and

the best-fit parameter estimates along with r2 and mean square error for individual

column are given in Table 4 .6 . The multireaction model accurately described the BTCs

in soil columns of different tillage systems. As indicated by Table 4.6, high r2 (0.9836

to 0.9981) and low MSE (0.022 to 0.057) were obtained for all BTCs using the

multireaction model. Best-fit parameters estimated based on nonlinear optimization from

transport experiments (Table 4.6) were different from measured batch parameters

(Table 4.3). The instantaneous distribution coefficient (kg) was significantly lower than

that from batch experiments for all cases. This may be because a significant fraction

of sorption sites did not participate in instantaneous or fast retention reaction in the

undisturbed soil columns. Traditional batch techniques that use high solution : soil

ratios and agitation to obtain equilibrium may not obtain the essential kinetic

information necessary to model solute mobility in undisturbed soils (O’Dell, et al.,

1992). Brusseau et al., (1989) separated the processes responsible for nonequilibrium

transport into two general classes: physical nonequilibrium and sorption-related

nonequilibrium. Physical nonequilibrium results from the existence o f mobile and

immobile domains within the porous medium and affects both sorbing and nonsorbing

solutes. Sorption related nonequilibrium involves only sorbing solutes and arises from

chemical non-equilibrium and intra-sorbent diffusion. Although both physical and

chemical nonequilibrium may be significant in undisturbed soils from different tillage


112

systems, the MRTM in calibration mode successfully described alachlor transport. This

adds the credence to the capability o f MRTM in describing the multiple retention

processes during transport in soils.

4 .3 .7 Retention Parameters of Alachlor During Transport in Undisturbed Soils Columns

As indicated in Table 4.6, the fitted parameters demonstrated the dominance of

the kinetic component o f the multireaction model compared with batch parameters

(Table 4.3 and 4.6). This may be attributed to heterogeneity of the undisturbed

columns. The kinetic reaction is diffusion controlled and is extremely sensitive to flow

or mixing effects. The dependence of kinetic reaction rate coefficients for atrazine

transport in undisturbed soil cores were reported by several investigators (Gaber et al,

1995; Gamerdinger et al., 1991). Gaston and Locke (1994) found that model

parameters for alachlor transport depend on pore water velocity. Specifically, best -fit

parameters for alachlor BTCs not only differed from those determined from their batch

data but also revealed dependency on pore water velocity or residence time. Since the

rate coefficients varied with experimental conditions, an overall retention should be

compared with estimates based on the sum of contributions from both kinetic and

equilibrium retention, i.e ., K = k e + 0k1/pk2 (Selim et al., 1995). The overall retention

K is listed in Table 4.6. Based on Duncan’s multiple range test, there was no significant

difference (p=0.05) for the overall retention parameter K between the no-till and

conventional tillage treatments. The average K was 1.53 for conventional tillage and

1.58 for the no-till. The critical range for significant difference was 0.276 (p=0.05).

This could be due to the effects of heterogeneity in no-till where the retention sites were


C/C

n C

/C,

113

O

Fig.4.5

a) NTW1Alachlor • Measured

Predicted Fitted

Tritium ° Measured — Fitted

b) NTW2 Alachlor • Measured

Predicted Fitted Measured Fitted

Tritium

BTCs of alachlor in soil columns of no-till with wheat as a winter cover crop for (a) replication 1 and (b) replication 2.


114

O

oo 0 .6 ■

0.8

o 0.6oo 0.4

0.2

0

a) CTW1 A lach lo r. Measured

— - Predicted— Fitted

T ritiu m 0 Measured— Fitted

25

b) CTW2Alachlor • Measured


Tritium o Measured— • Fitted

20 25o

Fig.4.6 BTCs of alachlor in soil columns of conventional tillage with wheat as a winter cover crop for (a) replication 1 and (b) replication 2.


c/a c/c

o

Fig.4.7

115

S-

a) NTV1Alachlor • Measured


•T ritium 0 Measured— ■ Fitted

20 25

b) NTV2 Alachlor * Measured

- - Predicted— Fitted

Tritium 0 Measured— • Fitted

2 d Flow Interruption

20 25

BTCs of alachlor in soil columns of no-till with vetch as a winter cover crop for (a) replication 1 and (b) replication 2.


116

0.8" 1

CD•O

0 .,4o° 0.4 •

-»E. 1 11

0.2 1? r•ij

00

Oo

a) CTV1 Alachlor • Measured

Predicted Fitted

Tritium o Measured

b) CTV2Alachlor • Measured


Tritium o Measured— ■ Fitted

-•__ L

20 25

Fig.4.8 BTCs of alachlor in soil columns o f conventional tillage with vetch as a winter cover crop for (a) replication 1 and (b) replication 2.


117

0.8

o 0.6oo 0.4

0.2

0

1

0.8

o 0.6oo 0.4

a) NNC1 Alachlor • Measured

- - Predicted— Fitted

Tritium ° Measured • — Fitted

• • •

20 25

b) NNC2 Alachlor • Measured


Tritium ° Measured— ■ Fitted

j 3 d Flow Interruption

20 25

Fig.4.9 BTCs o f alachlor in soil columns o f no-till with no winter cover crop for (a) replication 1 and (b) replication 2.


118

Oo

o

o

a) CNC1 Alachlor • Measured


Tritium o Measured— Fitted

b) CNC2 Alachlor • Measured

Predicted Fitted

Tritium o Measured

F ig .4 .10 BTCs o f alachlor in soil columns of conventional tillage with no winter cover crop for (a) replication 1 and (b) replication 2.


Peak

H

eigh

t (c

m)

119

5

4

3

2

1

°0 5 10 15Retention Time (minutes)

Fig. 4.11 HPLC chromatogram of effluent in 4 days after alachlor application during transport experiment in soil from no-till with vetch as a winter cover crop.

Alachlor


120

Table 4.5 The alachlor recovery, goodness of prediction for BTCs using kinetic parameters o f batch experiment in soils from different tillages and winter cover crops

Sample6 Recovery

( * )

r2 MSE

CTW1 91.7 0.6363 0.298

CTW2 89.0 0.3818 0.314

CTV1 88.7 0.6994 0.274

CTV2 83.5 0.5444 0.291

CNC1 93.4 0.7276 0.256

CNC2 80.0 0.8982 0.174

NTW1 94.3 0.7909 0.215

NTW2 92.6 0.8917 0.143

NTV1 88.7 0.6994 0.274

NTV2 98.0 0.7815 0.235

NNC1 99.8 0.7749 0.271

NNC2 95.8 0.7800 0.200

* CTW, conventional tillage-wheat; CTV, conventional tillage-vetch; CNC, conventional tillage-no cover; NTW, no-till-wheat; NTV, no-till-vetch; NNC, no-till-no cover.


Table 4.6 Fitted kinetic parameters o f BTCs in soils from different tillages and winter cover crops in transport experiments

Sample8 K k i -(cm 3 g-1)

k2 ^ k3 K rms

CTW1 0.001 ±0.160 4.7701 ±3.6430 0.8993 ±0.6707 0.0070±0.0031 1.7 0.9924 0.045CTW2 0.257±0.802 0.7707±0.6668 0 .1 8 1 7 ± 0 .1008 0.0066±0.0024 1.7 0.9932 0.034CTV1 0.456±0.535 0.4899±0.4626 0.1338±0.0752 0.0092±0.0025 1.6 0.9968 0.030CTV2 0.518±0.164 0 .1982±0.0740 0.0693±0.0192 0.0108±0.0018 1.5 0.9966 0.022CNC1 0.509±0.116 0.1628±0.0517 0.0581 ±0.0155 0.0040±0.0017 1.4 0.9965 0.031CNC2 0.817±0.106 0.0261 ±0.0139 0.0155±0.0196 0.0138±0.0051 1.3 0.9958 0.025NTW1 0.478±0.568 0.5451 ±0.5022 0.1432 ±0.0812 0.0023 ±0.0024 1.9 0.9960 0.030NTW2 0.646±0.325 0.1469±0.1139 0.0524±0.0318 0.0015±0.0029 1.7 0.9933 0.031NTV1 0.511 ±0.600 0.4899±0.4626 0.1338±0.0752 0.0092±0.0025 1.8 0.9969 0.030NTV2 0.676±0.080 0.0929±0.0285 0.0552±0.0143 0.0008±0.0017 1.3 0.9981 0.023NNC1 0.0004±0.02 0.8938±0.2194 0.2110±0.0568 0.0001 ±0.0037 1.5 0.9904 0.057NNC2 0.0004±0.08 0.5163±0.2082 0.1432 ±0.0655 0.0026±0.0058 1.3 0.9836 0.057

8 CTW, conventional tillage-wheat; CTV, conventional tillage-vetch; CNC, conventional tillage-no cover; NTW, no-till- wheat; NTV, no-till-vetch; NNC, no-till-no cover; kg is distribution constant; kl5 k2 and forward and backward rate coefficient for Sls k3 is rate coefficient (h '1) for consecutive irreversible reaction from S! to S2; r2 is adjustable correlation coefficients; rms is the root mean square error, K =kt5+ 0k1/pk2 is overall retention coefficient.

122

less accessible than that of soils from conventional tillage. In addition, for the transport

of adsorbing chemicals, earthworm burrows are different from other macropores in

soils, such as cracks or root channels. Edwards et al, (1992) found a higher attenuation

of atrazine and alachlor in night crawlers burrows than unlined, artificial burrows. They

suspected that organic and microbial lining o f the burrows increased the adsorption of

atrazine. In a transport experiment, dicamba and primisulfuron were more retarded

than expected from batch studies (Stehouwer et al., 1994).

4.3.8 Alachlor Recovery, Degradation and Irreversible Reaction

Gaston and Locke (1994) showed a thin layer chromatogram for 14C recovery

in column effluent as a function o f pore volumes collected. They observed little

degradation of alachlor even after 17 pore volumes (16 days). In this study, an analysis

o f the effluent by HPLC chromatograms (Fig. 4.11) gave no major peaks other than

alachlor, which indicated that there was no metabolites in the column effluent. As

shown in Table 4.6, the fitted transport model indicated that there was a considerable

irreversible reaction in all six soil columns of conventional tillage and in one soil

column of no-till (NTV1). In these columns, the rate coefficients k3 for irreversible

reaction were significantly different from zero (p=0.05) using a t-test. This result is

consistent with mass recovery of alachlor in the effluent as listed in Table 4.5. The

mass recovery of alachlor was varied with a tendency of higher recovery in soils of no

till than conventional tillage. Gaston and Locke, (1996) where they found 88%

recovery o f bentazon in effluent o f no-till whereas 78% in conventional tillage. These

differences may be attributed to the dissolved organic carbon which may activate the


123

alachlor in irreversible sites in undisturbed soils of no-till. This result was further

confirmed by Figs 4.7b and 4.9b where flow interruption was conducted. There was

no alachlor concentration change following flow interruptions during leaching in soil

columns o f no-till with vetch as cover crop (Fig. 4.7b). This could be the result of

compensation between irreversible bonding reaction and reversible kinetic reaction.

Irreversible bonding reaction may cause a decrease whereas reversible kinetic reaction

may increase the alachlor concentration. Selim et al., (1995) reported that there was a

significant concentration decrease o f TNT after flow interruption during leaching when

irreversible reaction or degradation occurred during transport experiments. Alachlor

concentration showed an increase of 0.05 in C/C0 following flow interruption (Fig.

4.9b) during leaching for soil o f no-till with no winter cover crop. This is consistent

with the result o f atrazine in uniformly packed soil columns containing aggregates (Ma

and Selim, 1994). It is an indication o f dominance o f reversible reaction and lack of

irreversible bonding reaction or degradation.

4.3.9 Tillage effect on Alachlor Transport

As discussed in section 4.3.3, no-till increased alachlor retention in batch

experiments of all cover crops. Increased hydrodynamic dispersion resulted in

preferential flow in no-till and led to early alachlor breakthrough and thus reduced

alachlor retardation. Increased heterogeneity o f no-till caused the physical

nonequilibrium transport and reduced values of diffusion-controlled equilibrium and

kinetic parameters. The compensation between increasing adsorption due to high

organic matter contents and reducing adsorption by diffusion-limited mechanisms led


124

to a non-significant difference of overall retention in undisturbed soils between no-till

and conventional tillage. No-till may reduce the bound alachlor residue as a result of

increased dissolved organic carbon (DOC). Although the overall alachlor retention was

not affected by tillage, the no-till reduced the retardation and the shape o f alachlor

BTCs compared with conventional tillage based on alachlor breakthrough in this study.

4.4 Conclusions

No-till increased the retention o f alachlor in completely mixed batch

experiments, especially for instantaneous equilibrium reaction, whereas no-till may

result in preferential flow and physical nonequilibrium transport in undisturbed soils.

Alachlor transport was constrained by both physical and chemical nonequilibrium

processes. Based on breakthrough which is affected by preferential flow or physical

nonequilibrium, no-till reduced alachlor retardation compared with conventional tillage.

Although the physical and chemical nonequilibrium were significant in undisturbed soil

columns from no-till, the MRTM in a calibration model adequately described alachlor

transport in those soils. This would indicate that the MRTM can describe the physical

and chemical nonequilibrium transport o f alachlor in this study where chemical

nonequilibrium (i.e., kinetic reaction parameters) were affected by physical

nonequilibrium (heterogeneity). No-till increased alachlor retention in completely mixed

batch experiments compared with conventional tillage whereas it did not affect the

alachlor overall retention in undisturbed columns based on parameters fitted from

alachlor BTCs in transport experiments. As discussed in Chapter 3, there was more

pronounced physical nonequilibrium solute transport in soils from no-till compared with


125

conventional tillage. This indicated that physical nonequilibrium affected chemical

nonequilibrium and thus reduced the retardation of alachlor during transport. The

difference in rate coefficients may explain the variation o f shape o f alachlor BTCs. The

higher recovery and lack o f irreversible reaction indicated that no-till may increase

dissolved organic carbon (DOC) in soil and thus reduced bond alachlor residue during

transport.


CHAPTER 5

SUMMARY AND CONCLUSIONS

The fate and transport of herbicide have been cited to be affected by tillage

practices. However, there are various viewpoints regarding the overall effect of tillage.

The controversy stems from the observed heterogeneity of the soil and its constituents

as influenced by tillage. The heterogeneity of soils leads to physical and chemical

nonequilibrium transport of herbicides in the soil profile. This study was designed to

quantify and describe the transport of alachlor [2-chloro-N-(2,6-diethylphenyl)-N-

(methoxymethyl) acetamide] as affected by physical and chemical nonequilibrium in a

Gigger soil (Typic Fragiudalf) from no-till and conventional tillage systems.

A batch study was conducted to characterize the chemical nonequilibrium

reactions, i.e ., kinetics of alachlor adsorption and desorption in surface soil samples

collected from long term no-till and conventional tillage plots. A batch method was used

to determine alachlor adsorption with time up to 528 h for a range of input

concentrations from 0.5 to 50 mg/L. Successive dilution was used to quantify

desorption kinetics following 16, 32, 64, 128, 256, and 512 h o f adsorption.

Alachlor time-dependent results indicated initially fast adsorption followed by

slow or kinetic type reactions. Adsorption-desorption results showed extensive

hysteretic behavior resulting from discrepancies between adsorption and desorption

isotherms. Hysteresis was more pronounced with increased reaction time o f adsorption.

Four variations o f a nonlinear multireaction kinetic model successfully described

retention results. However, adsorption and desorption were best described by a model

126


127

version incorporating nonlinear equilibrium, a kinetic reversible mechanism, and a

consecutive irreversible mechanism. The model was successfully used in describing

alachlor adsorption and desorption results. The model predicted the overall alachlor

hysteresis satisfactorily except that, at long times, desorption isotherms were not

considered adequate. The model was also capable of predicting alachlor desorption

kinetics based solely on parameters obtained from adsorption experiments.

Tracer studies were conducted with two dyes (Eosin Y and Brilliant Blue FCF)

and tritium to stain the flow paths and determine physical nonequilibrium characteristics

including soil heterogeneity, as well as the preferential flow in undisturbed soils. The

stained areas and flow paths revealed that solute transport occurred only through a

portion o f the pore space in soil from no-till, whereas the soil was somewhat uniformly

stained and breakthrough curves (BTCs) displayed little preferential flow in soil from

conventional tillage. The higher concentration of dyes in the effluent of no-till than that

of conventional tillage may be explained by the stained flow paths observed.

The tritium breakthrough times was earlier in soil columns of no-till than that

of conventional tillage in all cover crops. To examine the influence o f pulse

characteristics on tritium BTCs, pulses with two durations (long and short pulses) were

applied. Most BTCs of short pulses of the no-till soil showed bimodal peaks when

upward or downward flow was maintained. These bimodal peaks indicated intrinsic

preferential flow paths in the undisturbed soil. All BTCs in columns o f no-till showed

some degree o f tailing of the desorption (elution) side. The excessive tailing and


128

asymmetry indicated possible existence o f nonequilibrium mass exchange between soil

regions having different mobilities.

Tritium BTCs for the conventional tillage were relatively symmetric, which

indicate little or no preferential flow in these soil columns. Flow interruption was

carried out during tritium transport experiments and no concentration change due to

flow interruption was observed.

The BTCs of short pulse were used to predict tritium BTCs of the long pulse

using the principle of superposition. For most soils, the superimposed BTCs were

consistent with observed BTCs of long pulse. These predictions are confirmation of the

intrinsic flow paths in soil columns whereas poor predictions in soil of no-till with

wheat cover crop may be attributed to change in the flow paths during the experiments.

The BTCs in soils o f no-till were distinctly different depending on the flow directions,

which may be due to the structuralization o f soil after no-till practice. The direction of

flow had little effect on tritium breakthrough in soils of conventional tillage.

The convection-dispersion equation (CDE) adequately described tritium BTCs

of soils under conventional tillage. The description was not adequate for the

experimental data o f soils from no-till, especially for the short pulse. The physical

nonequilibrium models including mobile-immobile model (MIM) and stochastic model

(SM) were used to describe the long pulse BTCs of soils from no-till. There was no

improvement in the description of the data. In addition, the CDE, MIM and SM were

not capable of describing the double peaks of BTCs. The partition o f mobile-immobile

fraction may not sufficiently represent the reality of the structure o f these undisturbed


129

soils where preferential flow occurred. Thus at present we are not aware of modeling

approaches that can predict solute in the no-till soils which displayed preferential

characteristics.

Miscible displacement experiments of alachlor were conducted in undisturbed

soil columns o f no-till and conventional tillages to quantify and model its physical and

chemical nonequilibrium transport in soil. The no-till significantly increased the

magnitude of equilibrium but not the kinetic parameters compared with conventional

tillage. This effect may be a result of increasing organic matter content in no-till

compared with conventional tillage. Similar to tritium BTC, an early breakthrough time

and long tailing were observed for alachlor BTCs. This is perhaps due to preferential

flow in all soils of no-till compared with conventional tillage, where higher peak

concentrations, less retention and dispersion were observed. There was a tendency of

increase hydrodynamic dispersion coefficients in soils of no-till compared with that of

conventional tillage.

In an attempt to model the alachlor transport, a multireaction transport model

(MTRM) with one equilibrium reaction, one kinetic reversible reaction and one

consecutive irreversible reaction was used for describing the alachlor transport. Model

predictions in most cases, were considered inadequate when independently measured

model parameters from batch experiments were used. Since the multireaction model was

not successful in alachlor prediction when batch parameters were utilized, the MRTM

in a calibration mode was carried out to describe the BTCs in all soil columns of

different tillage systems. High r2 and low mean square error (MSE) were obtained for


130

all BTCs when a nonlinear parameter optimization scheme was used to describe alachlor

BTCs. The successful description of alachlor BTCs with the MRTM added credence

to its capability to describe both physical and chemical nonequilibrium transport of

herbicides.

The values of fitted parameters from transport experiments indicated a

dominance of kinetic reactions compared with parameters measured in batch

experiments. In contrast, the instantaneous distribution coefficients were significantly

lower than that from batch experiments. These effects may be caused by a significant

fraction of the sorption sites not participating in retention reactions. Due to variation

of kinetic parameters, an overall retention was compared with estimates based on the

sum of contributions from kinetic and equilibrium retention, i.e ., K = k e + 0kj/pk2.

Based on Duncan’s multiple range test, there was no significant difference (p=0.05)

for K between the no-till and conventional tillage treatments. This could be due to the

effects o f heterogeneity in no-till where the retention sites were perhaps less accessible

than that of soils from conventional tillage.

Analysis of the effluent by HPLC chromatograms gave no major peaks other

than alachlor, which indicated that there was no detectable metabolites in the effluent

of the columns and no degradation occurred during transport experiments. Moreover,

parameters from fitted BTCs indicated considerable irreversible reaction in all soil

columns o f conventional tillage and in one no-till soil column. In these columns, the

rate coefficients k3 for irreversible reaction were significantly different from zero

(p=0.05) using a t-test. This result is consistent with mass recovery of alachlor in the


131

effluent where alachlor recovery in the effluent was higher in soils without irreversible

reactions than those with irreversible reactions.

There was a tendency o f higher recovery in soils o f no-till than conventional

tillage. The difference may attribute to the dissolved organic carbon which may activate

alachlor in irreversible sites in undisturbed soils of no-till. This result was further

confirmed by experiments where flow interruption was conducted. There was no

alachlor concentration change after flow interruptions during leaching for the soil of no

till with vetch as a cover crop whereas there was somewhat of an increase o f alachlor

concentration in soil of no-till with no cover crop. Although the no-till increased

alachlor retention in completely mixed batch experiments because o f high organic

matter content compared with conventional tillage, no-till increased the alachlor

transport in term o f breakthrough due to soil heterogeneity and preferential flow.


REFERENCES

Abdul, A .S., T.L. Gibson, and D .N. Rai. 1990. Use of humic acid solution to remove organic contaminants from hydrogeological systems. Envim. Sci. Technol. 24:328-333.

Abraham Tal, B. Rubin, J. Katan and N. Aharonson. 1989. Fate of 14C-EPTC in a soil exhibiting accelerated degradation o f carbamothioate herbicides and its control. Weed Sci. 37:434-439.

Acar, Yalcin B. 1990. Transport of low-concentration contaminants in saturated earthen barriers. Journal of Geotechnical Engineering. 116:1031-1052.

Akratanakul, S., L. Boersma, and O. O. Klock. 1983. Sorption processes in soils as influenced by pore water velocity. II. Experimental results. Soil Sci. 135:331- 341

Amacher, M .C ., H.M . Selim, and I.K.Iskandar. 1988. Kinetics o f chromium and cadmium retention in soils; a nonlinear multireaction model. Soil Sci. Soc. of Am. J. 52:398-408.

Amacher, M .C ., H.M . Selim, and I.K.Iskandar. 1990. Kinetics o f mercuric chloride retention by soils. J. Envim. Qual. 19:382-388.

Amy B. Zins, Donald L. Wyse, and William C. Koskinen. 1991. Effect of alfafa (Medicago sativa) roots on movement of atrazine and alachlor through soil. Weed Sci. 39:262-269.

Allen S. J., G. McKay, and K.Y.H. Khader. 1989. Intraparticle diffusion of a basic dye during adsorption onto sphagnum peat. Environmental Pollution 56:39-50.

Andreini, M. S., and T. S. Steenhuis. 1990. Preferential paths o f flow under conventional and conservation tillage. Geoderma. 46:85-102.

Ball, W .P., M.N. Goltz, and P.V. Roberts. 1991. Comment on "Modeling the transport o f solutes influenced by multiprocesses non-equilibrium" by M.L. Brusseau, R.E. Jessup, and P.S.C, Rao. Water Resource. Res. 27:653-656.

Bear, J. 1969. Hydrodynamic dispersion: flow through porous media (de W iest,R.J.M .,Ed.), Academic Press.

132


133

Beestman, G.B. and J.M . Deming. 1974. Dissipation o f acetanilide herbicides from soils. Agron. J. 66:308-311.

Biggar, J.W ., and M. W. Cheung. 1973. Adsorption o f picloram (4-amino-3,5,6- trichloropicolinic acid) on Panoche, Ephrata, and Palous soils: A thermodynamic approach to the adsorption mechanism. Soil Sci. Soc. Am. Proc. 37:863-868.

Biggar, J.W ., U. Mingelgrin, and M. W. Cheung. 1978. Equilibrium and kinetics of adsorption of picloram and parathion with soils. J. Agric. Food Chem. 26:1306-1312.

Bird R. B., W .E. Steward, and E. D. Lightfoot. 1960. Transport phenomena. John Wiley & Sons, INC.

Blevins, R. L ., G. W. Thomas, M. S. Smith, W. W. Frye and P. L. Cornelius. 1982. Changes in soil properties after 10 year non-tilled and conventionally tilled com. Soil Tillage Res. 2.

Boesten, J.J.T .I. and L J .T . Van der Pas. 1988. Modeling adsorption/desorption kinetics of pesticides in a soil suspension. Soil Sci. 146: 221-231.

Boll, J ., T. S. Steenhuis, and J. S. Selker. 1992. Fiberglass wicks for sampling of water and solutes in the vadose zone. Soil Sci. Am. J. 56:701-707.

Bond W .J., and I.R . Phillips. 1992. Approximate solutions for cation transport during unsteady, unsaturated soil water flow. Water Resour. Res. 26:2195-2205.

Bosetto, M ., P. Arfaioli and P. Fusi. 1993. Interactions of alachlor with homoionic montmorillonites. Soil Sci. 155:105-113.

Bouma, J ., and L.W . Dekker. 1978. A case study on infiltration into dry clay soil. I. Morphological observations. Geoderma 20:27-40.

Bouma, J ., L.W . Dekker, and J.H .M . Wosten. 1978. A case study on infiltration into dry clay soil. II. Physical measurements. Geoderma 20:41-51.

Bouma, J.,A . Jongerius, and D. Schoonderbeek. 1979. Calculation of saturated hydraulic conductivity o f some pedal clay soils using micromorphometeric data. Soil Sci. Am. J. 43:261-264.

Bosetto, M ., P. Arfaioli and P. Fusi. 1993. Interactions o f alachlor with homoionic montmorillonites, Soil Sci. 155:2:105-113.


134

Bowman, B.T. and W .W. Sans. 1985. Partitioning behavior o f insecticides in soil-water systems: II. Desorption hysteresis effects. J. Environ. Qual. 14:270-273.

Brent W. Bean, Fred W. Roeth, Alex R. Martin, and Robert G. Wilson. 1987, Duration o f enhanced soil degradation of EPTC as influenced by herbicide rotation, time, and location. Weed Sci. 36:524-530.

Bresler E. and G. Dagan. 1981. Convective and pore scale dispersive solute transport in unsaturated heterogenous field. Water Resour. Res. 17:1683-1693.

Bresler, E. and R.J. Hanks. 1969. Numerical method for estimating simultaneous flow of water and salt in unsaturated soils. Soil Sci. Soc. Am. Proc. 33:827-832.

Brouwer, W. W. M ., J.J. T .I. Boesten and W. G. Siegers. 1990. Adsorption of transformation products of atrazine by soil. Weed Research. 30:123-128.

Burgard, D .J., W .C. Koskinen, R.H. Dowsy, and H .H. Cheng. 1993. Metolachlor distribution in a sandy soil under irrigated potato production. Weed Sc. 41:648- 655.

Brusseau, M .L ., R.E. Jessup, and P.S.C . Rao. 1989. Modeling the transport o f solutes influenced by multiprocesses nonequilibrium. Water Resour. Res. 25:1971-1988.

Brusseau, M .L ., and P.S .C . 1989. Sorption nonideality during organic contaminant transport in porous media. Cri. Rev. Environ. Control 19:33-99.

Brusseau, M .L ., and P.S .C . 1989. The influence of sobate-organic matter interaction on sorption nonequilibrium. Chemospere 18:1691-1706.

Brusseau, Mark. L. 1993. The influence of solute size, pore water velocity and intraparticle porosity on dispersion and transport in soil. Water Resour. Res. 29:1071-1080.

Butters, G. L ., W. A. Jury, and F .F . Ernst. 1989. Field scale transport of bromide in an unsaturated soil. 2. Dispersion modeling. Water Resour. Res. 25:1583-1589.

Calvet, R. 1980. Adsorption-desorption phenomena. In Interaction between herbicides and the soil. R J . Hance (ed.). Academic Presses, London, pp. 1-30.

Carski, T. H ., and D. L. Sparks. 1985. A modified miscible displacement technique for investigating adsorption-desorption kinetics in soils. Soil Sci. Soc. Am. J. 49:1114-1116


135

Carslaw, H .S ., and J. C. Jaeger. 1959. Conduction o f heat in solids. Oxford University Press. New York.

Chesters, G ., G. V. Simsiman, J. Levy, B. J. Alhajjar, R.J. Fathalla and J. M. Harlan, 1989. Environmental fate of alachlor and metolachlor. Rev. Environ. Con tarn, and toxicol. 110:1-74.

Clay, S. A. and W. C. Koskinen. 1990. Characterization o f alachlor and atrazine desorption from soils. Weed Sci. 38:74-80.

Chen J.S. R.S.Mansell, P. Nkedi-Kizza, and B.A. Burgoa. 1996. Phosphorus transport during transident, unsaturated water flow in an acid sandy soil. Soil Sci. Am. J. 60:42-48.

Chin, Y. P ., W. J. Weber, Jr., and B. J. Eadie. 1990. Estimating the effects of dispersed organic polymers on the sorption of contaminants by natural solids. 2. Sorption in the presence o f humic and other natural macromolecule. Environ. Sci. Technol. 24:873-842.

Chiou, C.T. and M. Manes. 1986. Application of the Flory-Huggins theory to the solubility of solids in glyceryl trioleate. J. Chem. Soc. Faraday Trans. 1.82:243-246.

Chiou, C. T ., D. E. Kile, T. T. Brinton, R.L. Malcolm, J.A . Leeheer, and P. Mac Carthy. 1987. A comparison of water solubility enhancements of organic solutes by aquatic humic materials and commercial humic acids. Environ. Sci. Technol. 21:1231-1234.

Clay, S. A. and W. C. Koskinen. 1990. Adsorption and desorption of atrazine, hydroxyatrazine, and s-glutathione atrazine on two soils. Weed Sci. 38:262-266.

Clay S. A ., D .E. Clay, W .C. Koskinen and G.L. Malzer. 1992. Agrichemical placement impacts on alachlor and nitrate movement through soil in a ridge tillage system. J. Environ. Sci. Health, B27:125-138.

Clay, S. A ., R. R. Allmaras, W. C. Koskinen, and D .L. Wyse, 1988. Desorption of atrazine and cyanazine from soil. J. Environ. Qual. 17:719-723.

Coats, K. H. and Smith, B. D. 1964. Dead-end pore volume and dispersion in porous media. Soc. Pet. Eng. 4:73-84.

Conn, J. S. and Cameron. 1988. Persistence and carry over o f metribuzine and triallate in subarctic soils. Can. J. Soil Sci. 68:827-830.


136

Conservation Tillage Information Center. 1990. National survey of conservation tillage practices. Conserva. Tillage Inf. Center, Fort Wayne, IN.

Cohen, S. Z. 1984. List o f potential ground-water contaminants, Office o f pesticide programs(TS-769C), US EPA, Washington, DC.

Cohen, S., 1992. EPA. Issues phase analysis o f national pesticide survey, Ground Water Monit. Rev. 12: 90-91.

Corey, J.C. 1968. Evaluation o f dyes for tracing water movement in acid soils. Soil Sci. 106:182-187.

Crittenden, J .C ., N .J. Hutzler, D.G. Geyer, J.L. Oravitz, and G. Friedman. 1986. Transport o f organic compounds with saturated groundwater flow:Model development and parameter sensitivity. Water Resour. Res. 22:271-284.

Crosson, P ., M. Hanthom, and M. Duffy. 1986. The economics o f conservation tillage. Edited by Milton A. Sprague, No-tillage and Surface-tillage Agriculture. Tillage revolution. 409-436.

Czapar G. F ., R. Horton, and R.S. Fawcett. 1992. Herbicide and tracer movement in soil columns containing an artificial macropore. J. Environ. Qual. 21:110-115.

Dagan G. 1982. Stochastic modeling of groundwater flow by unconditional and conditional probabilities. 1. conditional simulation and direct problem. Water Resour. Res. 813-833.

Daniel C. Peek, and Arnold. P. Appleby. 1989. Phytotoxicity, adsorption, and mobility of metribuzin and its ethylthio analog as influenced by soil properties. Weed Sci. 37:419-423.

Dao, T. H ., and T. L. Lavy. 1978. Atrazine adsorption on soil as influenced by temperature, moisture content, and electrolyte concentration. Weed Sci. 26:303- 308.

David I Gustafson. 1990. Field Calibration o f surface: A model of agricultural chemicals in surface waters. J. Environ. Sci. Health, B25:665-687.

De Camargo, O. A ., Biggar, J.W ., and Nilsen, D. R. 1979. Transport o f inorganic phosphorous in an Alfisol. Soil Sci. Soc. Am. J. 43: 884-890.

Dennis, R. Pettygrove and Denny V. Naylor. 1985. Metribuzin degradation kinetics in organically amended soil. Weed Sci. 33:267-270.


137

Dick, W .A ., R.J. Roseberg, E.L. McCoy, W .M. Edwards, and F. Haghiri. 1989. Surface hydrologic response o f soils to no-tillage. Soil Sci. Soc. Am. J. 53:1520-1526.

Dubinin, M.M. 1967. Absorption in micropores. J. Colloid. Int. Sci. 23:487-499.

Durbin, F. 1974. Numerical inversion of Laplace transforms: An efficient improvement to Dubner and Abate’s method. Comput. J. 17:371-376.

Edwards, W .M ., M. J. Shiptalo, and L. D. Norton. 1988. Contribution of macroporosity to infiltration into continuous com no-tilled watershed: implications for contaminant movement. J. Contam. Hydrol. 3:193-205.

Edwards, W .M ., M. J. Shiptalo, and S.J. Traina, C.A. Edwards, and L.B. Owens.1992. Role of lumbricus terrestris (L) burrows on quality of infiltration water. Soil Biochem. 24:1555-1561.

Ehlers, W. 1975. Observations on earthworm channels and infiltration on tiled and untilled loess soil. Soil Sci. 119:242-247.

Elrick, D. E ., K. T. Rre and H.K. Krupp. 1966. Applications of miscible displacement techniques to soils. Water Resour. Res. 2:717-727.

Fahim, M. A ., and N. Wakao. 1982. Parameter estimation from tracer response measurements. Chem. Eng. J. 25:1-8.

Fermanich, K.J. and T.C. Daniel. 1991. Pesticide mobility and persistence in microlysimeter soil columns from a tilled and no-tilled plot. J. Environ. Qual. 20:195-202.

Flury, M ., and H. Fliihler. 1994. Brilliant Blue FCF as a dye tracer for solute transport studies-A toxicological overview. J.Environ. Qual. 23:1108-1112.

Flury, M ., and H. Fliihler. 1995. Tracer characteristics of Brilliant Blue FCF. Soil Sci. Soc. Am. J. 59:22-27.

Flury M. 1996. Experimental evidence o f transport of pesticides through field soils - a review. J. Environ Qual. 25:25-45.

Frissl, M .J., and P. Poelstra. 1967. Chromatographic transport through soils. I. Theoretical evaluations. Plant and Soil. 26:285-302.

Froment, G. F ., and K. B. Bischoff. 1979. Chemical reactor analysis and design. John Wiley & Sons, New York, NY.


138

Gaber H .M ., W .P. Inskeep, S.D. Comfort, and J.M . Wraith. 1995. Nonequilibrium transport of atrazine through large intake soil cores. Soil Soc. Am. J. 59:60-67.

Gamerdinger, A .P., A.T. Lemley, and R.J. Wagenet. 1991. Nonequilibrium sorption and degradation of three 2-chloro-s-triazine herbicides in soil-water systems. J. Environ. Qual. 20:815-822.

Gaston, L .A ., and H. M. Selim. 1990. Transport o f exchangeable cations in an aggregated clay soil. Soil Soc. Am. J. 54:31-38.

Gaston L.A. and H.M . Selim. 1991. Predicating cation mobility in kaolinitic media based on exchange selective of kaolinite. Soil Sci. Soc. Am. J. 55:1255-1261.

Gaston L.A. and Martin A. Locke. 1994. Predicting alachlor mobility using batch sorption kinetic data. 158:345-354.

Gaynor J .D ., J.A . Stone, and T.J. Vyn. 1987. Tillage system and Atrazine and alachlor residuals on a poorly drained Soil. Can. J. Soil Sci. 67:959-963.

Gaynor J. D ., D .C. MacTavish, and W.I. Findlay. 1992. Surface and subsurface transport of atrazine and alachlor from a Brookston clay loam under continuous com production. Arch. Environ. Con tarn. Toxicol. 23:240-245.

Gelhar, L .W ., A.L. Gutjahr, and R.L. Naff, 1979. Stochastic analysis of macrodispersion in a stratified aquifer, Water Resour. Res. 15:1387-1397.

Ghodrati, M ., and W .A. Jury. 1992. A field study of the effects of soil structure and irrigation method on preferential flow of pesticides in unsaturated soil. J. Contaminant Hydrol. 11:101-125.

Ghodrati, M ., and W.A. Jury. 1990. A field study using dye to characterize preferential flow of water. Soil Sci. Soc. Am. J. 54:1558-1563.

Gierke, J.S. 1986. Modeling the movement o f volatile organic chemicals through homogeneous, isotropic, unsaturated soils with concurrent air and water flow. M.S. Civil Eng. thesis, Michigan Technol. Univ. Microfilms, Ann Arbor, MI.

Gish, T.J. and W .A. Jury, 1982. Estimating solute travel times through a crop root zone. Soil Sci. 133:124-130.

Gish, T.J. and W .A. Jury. 1983. Effect of plant roots and root channels on solute transport. Trans. ASAE 26SW:440-444.


139

Gish, T. J. Isensee, A.R. Nash, R.G ., Helling. C. S. 1991. Impact o f pesticides on shallow groundwater quality. Transactions o f the ASAE. 344:1745-1753.

Graham-bryce, I.J. 1969. Diffusion of organophosphorus insecticides in soils. J. Sci. Food Agric. 20:489-494.

Green, R. E ., and S. R. Obien. 1969. Herbicide equilibrium in soils in relation to soil water content. Weed. Sci. 17:514-519.

Green, R .E ., and S.W. Karickhoff. 1990. Sorption estimates for modeling, p. 79-101. In H.H. Cheng (ed.) Pesticides in the soil environment: Processes, impacts, and modelling. SSSA Book Seri. 2. SSSA, Madison, WI.

Green, R. E ., and J. C. Corey. 1971. Pesticide measurement by flow equilibrium and subsequent displacement. Soil Sci. Soc. Am. Proc. 35:561-565

Greenland, D .J., R.H. Laby, and J.P. Quirk. 1965. Adsorption o f amino-acids and peptides by montmorillonite and illite. Trans. Faraday Soc. 61:2024-2035.

Gschwend, P .M ., and S.-C. Wu. 1985. On the constancy o f sediment-water partition coefficient of hydrophobic organic pollutes. Environ. Sci. Technol. 19:90-95.

Hall, J.K ., M.R. Murray, and N.L. Hartwig. 1989. Herbicide leaching and distribution in tilled and untilled soil. J. Environ. Qual. 18:439-445.

Hamaker, J. W. 1972. Decomposition: Quantitative aspects. In: organic chemicals in soil environment. V o l.l: p.253. New York: Marcel Dekker, New York, pp. 49- 144.

Helferrich, F. 1962. Ion exchange. McGraw-Hill Book Company, Inc.

Helling, C .S., W. Zhang, T . J. Gish, C. B. Coffman, A. R. Isensee, P.C. Kearney, D.R. Hoagland, and M .D. Woodward. 1988. Persistence and leaching of atrazine, alachlor, and cyanazine under no-tillage practices. Chemosphere. 17:175-187.

Herman W.knoche. 1991. Radioisotopic methods for biological and medical research. Oxford University Press.

Kluitenberg G. J. and R. Horton. 1990. Effect o f solute application method on preferential transport of solutes in soils. Geoderma 46:283-297.

Hall, J .K ., M.R. Murray, and N.L.Hartwig. 1989. Herbicide leaching and distribution in tilled and untilled soil. J. Environ. Qual. 18:439-445.


140

Hatton, T .A ., and E.N. Lightfoot. 1984. Dispersion of tracer solutes in flowing groundwater. Water Resour. Res. 20:1253-1259.

Hodges, S. C ., and G. C. Johnson. 1987. Kinetics o f sulfate adsorption and desorption by Cecil soil using miscible displacement. Soil Sci. Soc. Am. J. 50:323-331

Holden, L. R. and J. A. Graham. 1990. The National alachlor well water survey: Project Survey. Rep. MSL-9633. Vol. 1. Mon-Santo Agric. Co. St. Louis, Mo.

Homberger, G .M ., K.J. Beven, and P. F. Germann. 1990. Inferences about solute transport in macropores forest soils from time series models. Geoderma. 46:249-262.

Horton, R .E ., 1940. An approach toward a physical interpretation o f infiltration- capacity, Soil Sci. Soc. Am. Proc., 5:399-417.

Huang, K. and M. Th. van Genuchten. 1995. An analytical solution for predicting solute transport during ponded infiltration. Soil Sci. 159:217-223.

Huang P.M . 1984. Components and particle size fractions involved in atrazine adsorption by soils. Soil Sci. 138:20-24.

Hutchinson, R.L. R. A. Brown,B. R. Leonard, and C. W. Kennedy. 1993. Effects of tillage systems and winter cover crops on yield and maturity of cotton on a loess soil in northeast louisiana. Proceedings of the 1993 southern conservation tillage conference for sustainable agriculture. Monroe, Louisiana.

Ibl, N. 1983. Comprehensive treatise of electrochemistry, Plenum Press, New York.

International Atomic Energy Agency. 1983.Isotope techniques in the hydrogeological assessment of potential sites for the disposal of high-level radioactive wastes. Technical report Series No. 228.

Isensee, A., and A.M. Sadeghi. 1994. Effects of tillage and rainfall on atrazine residue levels in soil. Weed Sci. 42:462-467.

Isensee, A .R ., R.G. Nash, and C.S. Helling. 1990. Effect o f conventional vs. no-tillage on pesticide leaching to shallow groundwater. J. Environ. Qual. 19:434-440.

Ivany, J.A . Sadler, J.M ., and Kimball, E.R. 1983. Rate o f metribuzin breakdown and residue effects on rotation crops. Can. J. Plant Sci. 63:481-487.


141

Jardine, P. M ., and D. L. Sparks. 1984. Potassium-calcium exchange in a multireactive soil system. I. Kinetics. Soil Sci. Soc. Am. J. 48:39-45.

Jardine, P. M ., Parker, J. C ., and Zelazny, L. W. 1985. Kinetics and mechanisms of aluminum adsorption on kaolinite using a two-site nonequilibrium transport model. Soil Sci. Soc. Am. J. 49:867-873.

Jardine, P. M ., and G. K. Jacobs. 1990. Unsaturated transport o f inorganic cations in undisturbed soil columns. In A. Caughran (ed.) Proceedings o f the radionuclide adsorption workshop. Los Alamos National Laboratory, Los Alamos, NM.

Jardine, P .M ., G. V. Wilson, and R. J. Luxmoore. 1990. Unsaturated solute transport through a forest soil during rain storm events. Geoderma 46:103-118.

Jardine, P .M ., F.M . Dunnivant, H.M . Selim, and J.F. McCarthy. 1992. Comparison of models for describing the transport of dissolved organic carbon in aquifer columns. Soil Sci. Soc. Am. J. 56:393-404.

Jardine, P .M ., G.K. Jacobs, and G.V. Wilson. 1993. Unsaturated transport processes in undisturbed heterogeneous porous media: I. Inorganic contaminants. Soil Sci. Soc. Am. J. 57:945-953.

Jardine, P .M ., and G. V. Wilson, J.F. McCarthy, R.J. Luxmore, D.L. Taylor, and L. W. Zelazny. 1990. Hydrogeochemical processes controlling the transport of dissolved organic carbon through a forested hillslop. J. Contam. Hydrol. 6:3-19.

Jennings A.A. and V. Ravi. 1992. Interface penetration model formulation for analyzing wall-separation-influenced transient hydraulic conductivity data. Soil Sci. Soc. Am. J. 56:1023-1032.

John R. Cantwell, REX A. Liebl, and Fred W. Slife. 1989. Biodegradation characteristics of Imazaquine and Imazethapyr. Weed Sci. 37:815-819.

John D. Weete. 1986. Herbicide analysis by chromatographic techniques.

Johnson J.A ., and W.J. Farmer. 1993. Batch versus column method for determining distribution o f organics between soil and water phases. Soil Sci. 155:92-99.

Jury, W. A ., H. Elabd, and M. Resketo. 1986. Field study o f napropamide movement through unsaturated soil. Water Resour. Res. 22:749-755.

Jury, W. A ., and K. Roth. 1990. Transfer functions and solute movement through soil. Theory and application. Birkhauser Verlag, Basel. Switzerland.


142

Khan, S. U ., 1982. Bound pesticide residues in soil and plants. Residue Rev. 84:1-25.

Kinniburgh, D.G. 1986. General purpose adsorption isotherms. Environ. Sci. Technol. 20:895-904.

Kladivko K. J ., G.E. Van Scoyoc, E.J.Monke, K.M.Oates, and W .Pask, 1991. Pesticide and nutrient movement into subsurface tile drains on a silt loam soil in Indiana. J. Environ. Qual. 20:264-270.

Klute A., and C. Dirksen. 1986. Conductivities and diffusivities o f unsaturated soils, p. 687-734. In A. Klute (ed.) Methods of Soil analysis. Part 1-Physical and mineral Methods. 2nd ed. Agron. Monogr. 9. ASA and SSSA, Madison, WI.

Koch, S. and Fliihler, H ., 1993. Non-reactive solute transport with micropore diffusion in aggregated porous media determined by a flow-interruption method. J. Contam. Hydrol., 14:39-54.

Koch S., and H. Fliihler. 1994. Lateral solute mixing in homogeneous and layered sand. Geoderma 109-121.

Koskinen W. C., L.J. Jardvis, R.H. Dowdy, D .L. Wyse, and D .D. Buhler. 1991. Automation of atrazine and alachlor extraction from soil using a laboratoryrobotic system. Soil Sci. Soc. Am. J. 55:561-562.

Krupp, H.K. and D. E. Elrick. 1968. Miscible displacement in an unsaturated glass bead medium. Water Resour. Res. 4:809-815.

Kumwenda, J.D . T ., D .E. Radcliffe, W .L. Hargrove, and D .C. Bridges. 1993. Reseeding o f crimson clover and com grain yield in a living mulch system. Soil Sci. Soc. Am. J. 57:517-523.

Ladlie, J.S ., W. F. Meggitt, and D. Penner. 1976. Role o f pH on metribuzindissipation in field soil. Weed. Sci. 24:508-511.

Lai, R ., A.A. Mahboubi, and N.R. Fausey. 1994. Long-term tillage and rotation effects on properties o f a central Ohio soil. Soil Soc. Am. J. 58:517-522.

Lapidus, L ., and Amundson, N. R. 1952. Mathematics o f adsorption in beds. The effect of longitudinal diffusion in ion exchange and chromatographic column. J. Phys. Chem. 56:984-988.

Lawruk, T .S ., C. S. Hottenstein, D. P. Herzog, and F. M. Rubio. 1992. Quantification o f alachlor in water by a novel magnetic particle-based ELISA. Bull. Envion. Contam. Toxicol. 48:643-650.


143

Lee, D .Y ., and W.J. Farmer. 1989. Dissolved organic matter interaction with napro pamide and four other nonionic pesticides. J. Environ. Qual. 18:468-474.

Lee D. Y. W. J. Farmer, and Y. Aochi. 1990. Sorption o f napropamide on clay and soil in the presence o f dissolved organic matter. J. Environ. Qual. 19:567-573.

Levanon, D. lodling E. E. Meisinger J. J. and Starr J. L. 1993, Mobility of agrochemicals through soil from two tillage systems. J. Environ. Qual. 22:155- 161.

Li Y. and M. Ghodrati. 1995. Transport o f nitrate in soils as affected by earthworm activities. J. Envim. Qual. 24:432-438.

Lindstrom, F .T ., R. Haque, V.H.Freed, and L. Boersma. 1967. Theory on the movement of some herbicides in soils. Linear diffusion and convection of chemicals in soils. Environ. Sci. Tech. 1:561-565.

Ligon, J. T. V. Wilson, J. F. Allen, and V. P. Singh. 1977. Tracing vertical translocation o f soil moisture. J. Hydraul. Div. Am. Soc. Civ. Eng. 103:1147- 1158.

Lloyd M. Southwick, Guye H. Willis, and H. Magdi Selim. 1992. Leaching of atrazine from sugarcane in southern Louisiana. Journal o f Agricultural and Food Chemistry. 40:1264-1266.

Locke M .A ., and S.S. Harper. 1991. Metribuzin degradation in soil: effects of tillage. Pestic. Sci. 31:239-247.

Locke, M.A. 1992. Sorption-Desorption Kinetics of Alachlor in surface soil from two soybean tillage system. J. Environ. Qual. 21:558-566.

Luxmoore, R .J., P .M ., Jardine, G.V. Wilson, J.R. Jones, and L.W . Zelanzny. 1990. Physical and chemical controls o f preferred path flow through a forested hillslope. Geoderma 46:139-154.

Ma, L ., L. M. Southwick, G. H. Willis, and H.M . Selim, 1993. Hysteresis characteristics of atrazine adsorption-desorption by a sharkey soil. Weed Sci.1993. 41: 627-623.

Ma, L. and H.M . Selim. 1994. Predicting atrazine adsorption-desorption in soils: A modified second-order kinetic model. Water Resour. Res. 30: 447-456.

Madhun, Y .A ., J.L. Young, and V.H. Freed. 1986. Binding of herbicides by water- soluble organic materials from soil. J. Environ. Qual. 15:64-68.


144

Mahboubi A .A ., R.Lal, and N.R.Faussey. 1993. Twenty-eight years o f tillage effects on two soils in Ohio. Soil Sci. Soc. Am. J. 57:506-512.

Marquardt, D.W . 1963. An algorithm for least-squares estimation o f non-linear parameters. J. Soc. Ind. Appl. Math. 11:431-441.

Mansell, R. S. W. J. Bond, and S.A. Bloom. 1993. Simulation cation transport during water flow in soil: two approaches. Soil Sci. Soc. Am J. 57:3-9.

Mansell, R. S., Selim, H.M.,Kanchanasut, P ., Davidson, J. M ., and Fiskell, J. G. A. 1977. Experimental and simulated transport of phosphorous through sandy soils. Water Resource Res. 13:189-194.

Mason, M ., Weaver, W. 1924. The settling of small particles in a fluid. Phys. Rev. 23:412-426.

Matano, C. 1933. On the relation between the diffusion coefficients and concentrations of solid materials(the nickel-copper system). Jpn. J. Phys. 8:108-113.

McMahon, M. A. and G. W. Thomas. 1974. Chloride and tritiated water flow in disturbed and undisturbed soil cores. Soil Sci. Soc. Am. Proc. 38:727-732.

McCarthy, J .F ., and J.M . Zachara. 1989. Subsurface transport o f contaminants. Environ. Sci. Technol. 23:496-502.

Moench, A. F ., and A. Ogata. 1981. Numerical inversion of the Laplace transform solution to radial dispersion in a porous medium, Water Resour. Res. 17:250- 252.

McWhorter, G. C. 1984. Future needs in weed science. Weed Sci. 31:850-852.

Miller, D. M ., M. E. Sumner, and W. P. Miller. 1989. A comparison of batch- and flow-generated anion adsorption isotherms. Soil Sci. Soc. Am. J. 53:373-380

Neuman, S .P ., E.S. Simpson, P.A. Hsieh, J.W . Jones, and C.L.W inter. 1985. Statistical analysis o f hydraulic test data from fractured crystaline rock near Oracle, Arizona, in Proceedings, Memoirs o f the 17th International Congress o f International Association of Hydrologists, Vol.XVII,pp289-301, International association of hydrologists, Tuson, Ariz.

Nielson, D .R ., M. Th. van Genuchten, and J. W. Biggar. 1986. Water flow and solute transport processes in the unsaturated zone, Water Resour. Res. 22: 89S-108S.


145

Nkedi-Kizza, P ., Biggar, J.M ., Selim, H ,M ., van Genuchten, M .Th., Wierenga, P .J., Davidson, J .M ., and Nielsen, D .R ., 1984, On the equivalence of two conceptual models for describing ion exchange during transport through an aggregated soil. Water Resour. Res. 20:1123-1130.

Novick, N .J., R. Mukheiji, and M. Alexander. 1986. Metabolism of alachlor and propachlor in suspensions of pretreated soils and in the samples from ground water aquifers. J. Agric. Food Chem. 34:721-725.

Obien, S.R. and R.E. Green. 1969. Degradation of atrazine in four Hawaiian soils. Weed Sci. 17:509-514.

O ’Dell, J.D ., Jeff D. Wolt, and Philip M. Jardine. 1992. Transport of Imazethapyr in undisturbed soil columns. Soil Soc. Am. J. 56:1711-1715.

Parker, J. C ., and Jardine, P.M . 1986. Effect of heterogeneous adsorption behavior on ion transport. Water Resour. Res. 22:1334-1340.

Parker, J. C ., and M.Th. van Genuchten. 1984. Determining Transport parameters from laboratory and field tracer experiments. Virginia Agricultural Experimental Station, Bulletin 84-3.

Perkins, T. K ., and O. C. Johnston. 1963. A review of diffusion and dispersion in porous media. J. Soc. Petrol. Eng. 3:70-84.

Paul C. C. Feng, Stephen J. Wratten, Eugene W. Logusch, Susan R. Horton, and C. Ray Sharp. 1990. Immunoassay detection methods for alachlor, application to analysis of environmental water samples. Chapter 15.

Philip, J.R ., 1957. The theory of infiltration. Soil Sci. 83:345-357.

Pickering, N. B., T. L. Richard, J.-Y. Parlange, W. J. Waltman, and L. Kromm. 1988. Dye tracer analysis o f preferential flow to groundwater . Dep. Agric. Biol. Eng. Staff Rep. 88-2. Cornell Univ. Ithaca, NY.

Perkins, T.K. and O.C. Johnston. 1963. A review of diffusion and dispersion in porous media. J. Soc. Petrol. Eng. 3:70-84.

Persaud, N ., J. M. Davidson, and P. S. C. Rao. 1983. Miscible displacement of inorganic cations in a discrete homoinoic exchange medium. Soil Sci. 136:269- 278.

Peter, C. J. and J. B. Weber. 1985. Adsorption, mobility, and efficacy of metribuzin as influenced by soil properties. Weed. Sci. 33:868-873.


146Phillips, R. E ., and S.H. Phillips. 1984. No-tillage agriculture, principles and practices.

Van Nostrand Reinhold Company. Inc., New York.

Pivetz, B., and T.S. Steenhuis. 1989. Pesticide, Nitrate, and tracer loss in agricultural tile lines and to ground water under conventional and conservation tillage. ASAE paper 89-2579. Am. Soc. Agric. Eng., St. Joseph. MI.

Pothuluri, J.V ., T.B.Moorman, D.C.Obenhuber, and R.D. Wauchope. 1990. Aerobic and anaerobic degradation o f alachlor in samples from a surface-to groundwater profile. J. Environ. Qual. 19:525-530.

Rao, P .S .C ., R .E. Green, V. Balasubramanian, and Y. Kanehiro. 1974. Field study of solute movement in a highly aggregated Oxisol with intermittent flooding:II. Picloram. J. Environ. Qual. 3:197-202.

Razavi, M .S., B J . McCoy, and R.G. Carbonell. 1978. Moment theory of breakthrough curve for fixed-bed adsorbers and reactors, Chem. Eng. J., 16:211-222.

Ronald E. Jones, JR., Philip a. Banks, and David E. Radcliffe. 1990. Alachlor and metribuzin movement and dissipation in a soil profile as influenced by soil surface condition. Weed Sci. 38:589-597.

Roux, P .H ., R. L. Hall, and R.H. Ross, Jr. 1991. A large-scale retrospective ground water monitoring study for metolachlor, Ground Water Monit. Rev. 11:173-181.

Sanyal, S. K ., S.K. De Datta, P.Y . Chan. 1993. Phosphate sorption-desorption behavior o f some acidic soils of south and southeast asia. Soil Sci. Soc. Am. J. 57:937-945.

SAS Institute, Inc. 1985. SAS/ETS User’s guide: Statistics, Version 5 Edition, SAS Institute, Inc., Cary, NC.

Sauer T .J., and T.C. Daniel. 1986. Effect of tillage on runoff losses of surface-applied pesticides. Soil Sci. Soc. Am. J. 51:410-415.

Sawhney B. L ., and K. Brown. 1989. Reaction and movement of organic chemicals in soils. SSSA special publication: No. 22.

Schiavon, M. 1988. Studies of the movement and the formation o f bound residues of atrazine, o f its chlorinated derivations, and of hydroxyatrazine in soil using 14C ring labeled compounds under outdoor conditions. Ecotox. Environ.Safety. 15:55-61.


147

Schulin, R ., Wierenga, P.J. and Fliihler, H ., 1987. Solute transport through a stony soil. Soil Sci. Soc. Am. J., 51:36-42.

Schweich, D ., M. Sardin, and J.-P . Gaudet. 1983. Measurement o f a cation exchange isotherm from elution curves obtained in a soil column: Preliminary results. Soil. Sci. Soc. Am. J. 47:32-37

Schweich, D ., and M. Sardin. 1981. Adsorption, partition, ion exchange and chemical reaction in batch reactors or in columns. J. Hydrol. 50:1-33.

Selim H. M ., 1978. Transport of reactive solutes during transient unsaturated water flow in multilayered soils. Soil Sci. 126:127-135.

Selim, H .M ., Davidson, J.M ., and Rao, P.S.C . 1977. Transport of reactive solutes through multilayered soils. Soil Sci. Soc. Am. J. 41,3-10.

Selim H. M ., R. Schulin, and H. Fliihler. 1987. Transport and ion exchange of calcium and magnesium in an aggregated soil. Soil Sci. Soc. Am. J. 51:876-884.

Selim, H .M ., M .C. Amacher, and I.K. Iskandar. 1990. Modeling the transport of heavy metals in soils. CRREL monograph 2, 1990-700-057-22040, U.S. Government printing office, p. 250.

Selim, H.M . 1992. Modeling the transport and retention o f inorganic in soils. Advances in Agronomy, 47:331-384.

Selim, H. M ., and M. C. Amacher. 1988. A second order kinetic order approach for modeling solute retention transport in soils. Water Resource Res. 24:2061-2075.

Selim, H. M ., J. M. Davidson, and R. S.Mansell. 1976. Evaluation o f a two-site adsorption desorption model for describing solute transport in soils. In "Summer Computer Simulation Conference, Washington, D. C ., 1976" pp.444-448. Simulation Council Inc ., La Jolla, California.

Selim, H. M. and L. W. Ma. 1995. Transport of reactive solutes in soils: A modified two region approach. Soil Soc. Am. J. 59:75-82.

Selim, H.M . S.K. Xue, and I.K.Iskandar. 1995. Transport o f 2,4,6-trinitrotoluene and hexahydro-l,3,5-trinitro-l,3,5-triazine in soils. Soil Sci. 160:317-327.

Senesi, N ., G. Brunetti, P. L. Cava, and T.M. Miano. 1994. Adsorption of alachlor by humic acids from sewage sludge and amended and non-amended soils. Soil Sci. 157:176-184.


148

Seyfried, M. S., and P. S. C. Rao. 1987. Solute transport in undisturbed columns of an aggregated tropical soil: preferential flow effects. Soil Sci. Soc. Am. J. 51:1434-1444.

Shackelford, David, E. Daniel. 1991a. Diffusion in saturated soil. Background. Journal of Geotechnical Engineering. 117:467-484.

Shackelford, David, E. Daniel, 1991b, Diffusion in saturated soil. Result for compacted clay. Journal o f Geotechnical Engineering. 117:485-506.

Sharom, M. S. and G. R. Stephenson. 1976. Behavior and fate o f metribuzin in eight Ontario soil. Weed Sci. 24:153-160.

Sharon A. Clay and William C. Koskinen. 1990. Characterization of alachlor and atrazine desorption from soil. Weed Sci. 38:74-80.

Sharp, D.B. 1988. Alachlor. In Herbicides-Chemistry, degradation and mode of action, Vol. 3. Kearney P. C. and Kaufman D .D.(eds.). Marcel Dekker, New York, pp.301-333.

Shelton, C.H. F .D . Tompkinks, and D.D. Tyler. 1983. Soil Erosion from five soybean tillage systems. J. Soil and Water Conserv. 38:425-428.

Singh, P. and R. S. Kanwar. 1991. Preferential solute transport through macropores in large undisturbed soil columns. J. Environ, Qual. 20:295-300.

Singh, V. P. Hydrologic systems. Rainfall-runoff modeling. Vol.I. Prentice Hall, Englewood Cliffs, NJ.

Skopp, J ., W. R. Gardner, and E. J. Tyler. 1981. Solute movement in structured soil: two-region model with small interaction. Soil. Sci. Soc. Am. J. 45:837-842.

Skopp,J, and D. McCallister. 1986. Chemical kinetics from thin disc flow system: Theory. Soil Sci. Soc. Am. J. 50:617-623.

Smiles, D .E ., Philip, J.R. 1978. Solute transport during absorption o f soil water by soil: Laboratory studies and their practical implications. Soil Sci. Am. J. 42:537-544.

Smiles, D. E ., Philip, J. R. Knight, J. H ., and Elrick D. E. 1978. Hydrodynamic dispersion during absorption of water by soil. Soil Sci. Soc. Am. J. 42:229-234.

Sparks, D .L. ,L.W. Zelazny, and D.C. Martens. 1980. Kinetics o f potassium desorption in soil using miscible displacement. Soil Sci.Soc. Am. J. 44:1205-1208.


149

Sparks, D .L. 1988. Kinetics of soil chemical processes. Academic Press, Inc., New York, NY.

Selsted, M .E ., and H.W .Becker, 1986. Anal. Biochem. 155:270.

Smedt F. De and P.J. Wierenga. 1978. Approximate analytical solution for solute flow during infiltration and redistribution. Soil Sci. Soc. Am. J. 42:407-412.

Smettem, K .R .J., and N.Collis-George. 1985. Statistical characterization o f soil biopores using a soil peel method. Geoderma 36:27-36.

Suzuki, M ., and J. M. Smith. 1971. Kinetic Studies by chromatography. Chem. Eng. Sci. 26: 221-235.

Steenhuis, Tammo S. Jan Boll, Gil Shalit, John S. Selker, and Ian A. Merwin. 1994. A simple equation for predicting preferential flow solute concentrations. J. Environ. Qual. 23:1058-1064.

Steenhuis, T .S ., W. Staubitz, M.S. Andreini, J. Surface, T .L. Richard, R. Paulsen, N.B. Pickering, and J.R. Hagerman, and L.D. Geohring. 1990. Preferential movement of pesticides and tracers in agricultural soils. J.Irrig. Drain. Eng. 116:50-66.

Starr, J.L . and D.E. Glotfelty. 1990. Atrazine and bromide movement through a silt loam soil. J. Environ. Qual. 19:552-558.

Stearman, G .K ., R. J. Lewis, L. J. Tortorelli, and D. D. Tyler. 1989. Herbicide reactivity of soil organic matter fractions in no-till and tilled cotton. Soil Sci. Soc. Am. J. 53:1690-1694.

Stehouwer, R .C ., W .A. Dick, and S.J. Traina. 1994. Sorption and retention of herbicides in vertically oriented earthworm and artificial burrows. J. Environ. Qual. 23:286-292.

Stevenson, F. J. 1972. Organic matter reactions involving herbicides in soil. J. Environ. Qual. 1:333-343.

Stevenson, F. J. 1982. Humus chemistry, genesis, composition, reactions. John Willey & Sons Inc.

Talbert, R. E. and D. H. Fletchall. 1965. The adsorption o f some s-triazines in soil. Weeds. 46-52.


150

Tan, Y. Bond, W J . and Griffin, D.M . 1992. Transport of bacteria during unsteady unsaturated soil water flow. Soil Sci. Soc. Am. J. 56:1331-1340.

Theis, T. L ., R. Iyer, and L. W. Kaul. 1988. Kinetic studies o f cadmium and ferric yanide adsorption on goethite. Environ. Sci. Technol. 22:1013-1017.

Thibodeaux, L.J. 1979. Chemodynamics, John Wiley and Sons, New York, NY.

Triplett G. B., A. D. Worsham. 1985. Principles of weed management with surface- tillage systems. Edited by Milton A. Sprague, No-tillage and Surface-tillage Agriculture. Tillage revolution. 319-346.

Trundgill, S. T. A.M. Pickles and K.R. J. Smettem. 1983. Soil-water residence time and solute uptake. 2. Dye tracing and preferential flow predictions. J. Hydrol. 62:279-285.

Tsang Y.W . and C.F. Tsang. 1987. Channel Model of flow through fractured media. Water Resour. Res. 23:467-479.

Turner, G.A. 1972. Heat and concentration Waves, Academic, New York.

Tyler, D .D ., and G.W. Thomas. 1977. Lysimeter measurements o f nitrate and chloride losses from soil under conventional and no-tillage com. J. Soil and Water Conserv. 38:374-376.

U. S. Environmental Protection Agency. 1982. Pesticide assessment guidelines, subdivision N; Chemistry: Environmental fate. USEPA Rep. 540/9-82-021. Natl. Technical Information Service, U.S. Dep. of Commerce, Springfield, VA.

U.S. Environmental Protection Agency, 1989a, Health advisory summary atrazine. Office of Drinking water, U.S. EPA. Report#106, 401 M. Street, S. W ., Washington, D. C., 20460, USA, Pl-25.

U.S. Environmental Protection Agency. 1989b. Drinking water heath advisory: Pesticides, Lewis Publishers, Inc., Chelsea, MI.

U .S. Environmental Protection Agency. 1991. Federal Register 56 (20):3526-3597.

United States Environment Protection Agency. 1992. Method for the determination of nonconventional pesticides in municipal and industrial wastewater.

Valocchi, A J . 1985. Validity of the local equilibrium assumption for modeling sorbing solute transport through homogeneous soils. Water Resour. Res. 21:808-820.


151

van Genuchten, M. Th., J.M . Davidson, and P.J. Wierenga. 1974. An evaluation of kinetic and equilibrium equations for the prediction o f pesticide movement through porous media. Soil Sci. Soc. Am. Proc. 38:29-35.

van Genuchten, M. Th. and W.J. Alves. 1982. Analytical solutions of the onedimensional convection-dispersive solute transport equation. U.S. Department o f Agriculture, Tech. Bull. No. 1661. 151p.

van Genuchten, M. Th. 1980. Determining transport parameters from solute displacement experiments. Research Report No. 118, US. Salinity Laboratory, Riverside, California. 37 p.

Van Genuchten, M. Th. 1981. Non-equilibrium transport parameters from miscible displacement experiments. Research Report. No. 119. US Salinity lab. Riverside, CA. (88p).

van Genuchten, M. Th., and Wierenga, P. J. 1976. Mass transfer studies sorbing porous media. Analytical solution. Soil Sci. Soc. Am. J. 40:473-480.

van Genuchten, M. Th. and R. J. Wagenet. 1989. Two-site/two-region models for pesticides transport and degradation: theoretical development and analytical solutions. Soil Sci. Soc. Am. J. 53:1303-1310.

van Ommen, H .C ., L.W. Dekker, R. Diyksma, J. Hulsof, and W .H. van der Molen. 1988. A new technique for evaluating the presence o f preferential flow paths in non-structured soils. Soil Sci. Am. J. 52:1192-1193.

van Ommen, H. C. 1985. The ’mixing cell’ concept applied to transport o f non-reactive and reactive components in soils and groundwater. J. Hydrol. 78:201-213

Villermaux, J. 1974. Deformation of chromatographic peaks under the influence of mass transfer phenomena. J. Chromatog. Sci. 12:822-831.

Walker A ., and Sarrah J. Welch. 1991. Enhanced degradation of some soil-applied herbicides. Weed Research. 31:49-57.

Wallach, R ., and M. Th. Van Genuchten. 1990. A physically based model for predicting solute transfer from soil solution to rainfall-induced runoff water. Water Resour. Res. 2119-2126.

Ward, A. L ., R. G. Kachanoski, and D. E. Elrick. 1994. Laboratory measurements of solute transport using time domain reflectometry. Soil Sci. Soc. Am. J. 58:1031-1039.


152

Warrick. A. W ., J. W. Biggar, and D. R. Nielson. 1971. Simultaneous solute and water transfer for an unsaturated soil. Water Resour. Res. 7:1216-1225.

Warrick, A. W ., J. H. Kichen, and J. L. Thames. 1972. Solutions for miscible displacement o f soil water with time-dependent velocity and dispersion coefficient. 36:863-867.

Watson. K. W. and R. J. Luxmoore. 1986. Estimating macroporosity in a forest watershed by use o f a tension infiltrometer. Soil Sci. Soc. Am. J. 50:578-582.

Wauchope R .D ., Buttler, T. M ., Hornsby, Augustijn Beckers P.W .M . and Burt J. P. 1992. Environmental decision-making. Reviews of Environmental Contamination and Toxicology. 1-156.

Wauchope, R. D. and R. S. Myers. 1985. Adsorption-desorption kinetics o f atrazine and linuron in freshwater-sediment aqueous slurries. J. Environ. Qual. 14:132- 136.

Weber, J. B. and C. J. Peter. 1982. Adsorption, bioactivity, and evaluation o f soil tests for alachlor, acetochlor, and metolachlor. Weed Sci. 30:14-20.

Weber, W. J. Jr., Paul M. McGinley, and Lynn E. Katz. 1992. A distributed reactivity model for sorption by soils and sediments. 1. Conceptual basis and equilibrium assessments. Environ. Sci. Technol. 26:1955-1962.

White, R. E ., G. W. Thomas, and M. S. Smith. 1984. Modeling water flow through undisturbed soil cores using a transfer function model derived from 3H20 and Cl transport. J. Soil Sci. 35:159-168.

Wilson, G. V ., and R.J. Luxmoore. 1988. Infiltration, macroporosity, and mesoporosity distributions on two forested watersheds. Soil Sci. Soc. Am. J. 52:329-335.

Wilson, G .V ., P.M . Jardine, R.J. Luxmoore, L.W.Zelazny, D.A. Lietzke, and D.E. Todd. 1991. Hydrogeochemical processes controlling subsurface transport from an upper subcatchment o f Walker Branch Watershed during storm events: 1. Hydrological transport processes. J. Hydrol. (Amsterdam) 123:297-316.

Wood, L .S ., H .D. Scott, D.B. Marx, and T.L. Lavy. 1987. Variability in sorption coefficients of metolachlor on a capitina silt loam. J. Environ. Qual. 16:251- 256.

Wu, T.L. 1980. Dissipation o f herbicides atrazine and alachlor in a aryland com field. J. Environ. Qual. 9:459-465.


153

Xue S.K ., and H.M . Selim. 1995. Modeling adsorption-desorption kinetics o f alachlor in soils. J. Environ. Qual. 24:896-903.

Yaron, B., Z. Gerstl, and W .F. Spencer. 1985. Behavior o f herbicides in irrigated soils. Adv. Soil Sci. 3:121-211.

Yen, P .Y ., W .C. Koskinen, and E.E. Schweizer. 1994. Dissipation o f alachlor in 4 soils as influenced by degradation and sorption processes. Weed Sci. 42:233- 240.

Zimdahl, R.L. and S.K. Clark. 1982. Degradation o f three acetanilide herbicides in soil. Weed Sci. 30:545-548.


APPENDIX: LETTER OF PERMISSION

L O U I S I A N A S T A T E U N I V E R S I T Y D e p a r t m e n t o f 9 ILP B V.

firalM ta of lo l l I r r o l al u r j l i Noll IMllfm Inu IMIwraity

•oton tmugm. l» 70K0 Nano IOt-lM-2110

S.X.ut

Fax 504—318—1403

American seelity of Agronomy,677 S. Sagoe Rd., Madison, VI 53711.Fax 608-2 3-2021Dear Editor:I am writing to you to request for permission to reproduce the paper "Modeling Adsorption-Disorption Nineties of Alachlor in a Typic Fragiudalf" in Journal of Environmental Quality (24:896-903) by S.K. Xue and H.M. selim for using in ay dissertation.Xt will be greatly appreciated if you can fax me the permission.

SincerelyS.X. Xue

154


VITA

Shikui Xue was bom on June 18, 1963, in The People’s Republic o f China. He

received his Bachelor’s degree o f Science from Huazhong Agricultural University in

1984 in Soil Science and Agrochemistry, and Master of Science degree from Institute

of Soil Science, Academia Sinica in 1987 in Department o f Soil Geography. As a

research associate, he worked in Soil Geography, Institute o f Soil Science, Academia

Sinica from 1987 to 1990. He took the Advanced Course o f irrigation from October,

to December, 1990 at Volcani Center, Agricultural Research Organization, Israel. As

a visiting scholar, he worked in Soil Physics and Irrigation from 1991 to 1992 at

ARAVA Research Station, Israel. As a research assistant, he worked in Soil Physics,

Department o f Agronomy, Louisiana State University for Doctor degree o f Philosophy.

He is the author and coauthor of 14 scientific papers in refereed journals, book chapters

and proceedings on the subject o f Solute Transport, Irrigation, Nutrient Cycling and

Soil Geography.

155


DOCTORAL EXAMINATION AND DISSERTATION REPORT

Candidate:

Major Pield:

Title of Dissertation:

Date of Examination:

March 21; 1996

Shikui Xue

Agronomy

Modeling Physical and Chemical Nonequilibrium Transport of Herbicide in Soils From Different Tillage Systems

Approved:

ijor Professor and Chairman

Graduate SchoolDean o:

EXAMINING COMMITTEE:


Documents

Modeling Physical and Chemical Nonequilibrium Transport of