Agronomy Journal Grow Your Career With ASA
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

Published online 3 January 2006
Published in Agron J 98:8-18 (2006)
DOI: 10.2134/agronj2004.0221
© 2006 American Society of Agronomy
677 S. Segoe Rd., Madison, WI 53711 USA
This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF) Free
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via ISI Web of Science (1)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Weber, J. B.
Right arrow Articles by Wilkerson, G. G.
Right arrow Search for Related Content
PubMed
Right arrow Articles by Weber, J. B.
Right arrow Articles by Wilkerson, G. G.
Agricola
Right arrow Articles by Weber, J. B.
Right arrow Articles by Wilkerson, G. G.
Related Collections
Right arrow Tillage
Right arrow Irrigation
Right arrow Soybean
Right arrow Turfgrass

Soil and Water

Soil and Herbicide Properties Influenced Mobility of Atrazine, Metolachlor, and Primisulfuron-Methyl in Field Lysimeters

Jerome B. Webera,*, K. Allan Taylorb and Gail G. Wilkersona

a North Carolina State Univ., Crop Sci. Dep., Box 7620, Raleigh, NC 27695
b E.I. du Pont de Nemours and Co., Inc., Walker's Mill, Barley Mill Plaza, P.O. Box 80038, Wilmington, DE 19880-0038

* Corresponding author (jerry_weber{at}ncsu.edu)

Received for publication August 24, 2004.

   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 REFERENCES
 
Understanding herbicide mobility in soils is necessary to prevent groundwater contamination. We studied the mass balance distribution of three 14C-labeled herbicides (atrazine, metolachlor, and primisulfuron-methyl) in four soils (Dothan, Portsmouth, Rion, and Wagram) 128 d after application to soil column field lysimeters. Analyses were made of surface soil, subsoil, and leachate samples, and metabolites were identified in surface soil and leachate. Our objectives were to examine, measure, and correlate the leaching patterns of the chemicals and correlate their leaching characteristics with the physicochemical properties of the soils. Metolachlor was the most mobile herbicide as indicated by the retardation factor (Rf) (Rf = 0.35 in 1992 and 0.17 in 1993), followed by atrazine (Rf = 0.19 in 1992 and 0.09 in 1993) and primisulfuron-methyl (Rf = 0.15 in 1992 and 0.12 in 1993). Herbicide mobility (Rf) was related to leachate volume collected from the four soils, herbicide aqueous solubility, and longevity of the chemicals. The herbicides were of greatest mobility in Rion and Wagram soils and of lowest mobility in Portsmouth and Dothan soils. Soil factors affected the weakly basic atrazine differently than the nonionizable metolachlor or the weakly acidic primisulfuron-methyl. Volatility losses of the herbicides were inversely related to longevity (disappearance time in the field (DT50) of the compounds and to humic matter contents of the soils. Carbon-14 herbicide in the subsoil and in the leachate was correlated with herbicide mobility (Rf), total leachate volume, and 50% disappearance time values. Herbicide mobility was in agreement with predictability using a simple decision-aid model.

Abbreviations: 14C, radiolabeled carbon • DAT, days after treatment • DT50, 50% disappearance time (longevity in days) • HM, humic matter • Kd, herbicide/soil distribution coefficient (soil retention index) • Ks, aqueous solubility • LSA, liquid scintillation analyzer • MI, mobility index • OM, organic matter • pKa, –log Ka (ionization constant) • PLP, pesticide leaching potential index • Rf, chromatographic reached (retardation) factor • SLP, soil leaching potential index • TLC, thin-layer chromatography


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 REFERENCES
 
ATRAZINE [6-chloro-N-ethyl-N'-(1-methyl-ethyl)-1,3,5-triazine-2,4-diamine] is a weakly basic, low aqueous soluble herbicide used to control broadleaf and grass weeds when applied postemergence or pre-emergence (Ahrens, 1994) (Table 1). Atrazine retention in soil has been attributed to binding to organic matter (OM) (Gunther and Gunther, 1970; Stehouwer et al., 1994; Weber et al., 1969) and expanding-type clay minerals (Frissel, 1961; Weber, 1966; Weber et al., 1969). Mobility of atrazine in soil has been reported to be less than bromacil [5-bromo-6-methyl-3-(1-methylpropyl)-2,4(1H, 3H)pyrimidinedione], a weakly acidic (pKa = 9.1, where pKa, –log Ka), moderately aqueous soluble (Ks = 815 mg L–1, where Ks = aqueous solubility) herbicide with low soil sorptivity [Kd = 0.2 to 1.8 mL g–1, where Kd = herbicide/soil distribution coefficient (soil retention index)], but more mobile than prometon [6-methoxy-N, N'-bis (1-methylethyl)-1,3,5-triazine-2,4-diamine], a weakly basic (pKa = 4.3), moderately aqueous soluble (Ks = 720 mg L–1) herbicide with low to moderate soil sorptivity (Kd = 0.4 to 1.0 mL g–1) or diuron [N'-(3,4-dichlorophenyl)-N, N-dimethylurea], a nonionizable, low-water-soluble (Ks = 42 mg L–1) herbicide with moderate to high soil sorptivity (Kd = 2.9 to 14.0 mL g–1) (Weber and Whitacre, 1982).


View this table:
[in this window]
[in a new window]
  		Table 1. Herbicide physicochemical properties (20–25 °C) and rates applied to fallow field lysimeters.

 
Metolachlor [2-chloro-N-(2-ethyl-6-methylphenyl)-N-(2-methoxy-1-methylethyl) acetamide] is a nonionizable substituted acetamide herbicide that controls grasses and some broadleaf weeds and sedges when applied pre-emergence or preplant incorporated (Ahrens, 1994) (Table 1). Its retention in soil has been related to OM content (Kozak et al., 1983; Obrigawitch et al., 1981; Patakioutas and Albanis, 2002; Singh et al., 2001) and OM and clay contents (Peter and Weber, 1985; Pusino et al., 1992; Weber and Swain, 1993; Wood et al., 1987; Zhu and Selim, 2000). Soil mobility of metolachlor was inversely related to soil OM and clay content (Alhajjar et al., 1990; Obrigawitch et al., 1981; Singh et al., 2002; Wietersen et al., 1993). In soil leaching columns, metolachlor was more mobile than atrazine (Alhajjar et al., 1990; Bowman, 1988; Keller and Weber, 1995; Seybold and Mersie, 1996), terbuthylazine [6-chloro-N-(1,1-dimethylethyl)-N'-ethyl-1,3,5-triazine-2,4-diamine], a weakly basic (pKa = 1.5), very low Ks (8.5 mg L–1) herbicide with high soil sorptivity (Kd = 10 mL g–1) (Bowman, 1988; Singh et al., 2002) and primisulfuron-methyl {methyl-2-[[[[[4,6-bis(difluoromethoxy)-2-pyrimidinyl]amino]carbonyl]amino]sulfonyl]benzoate} (Table 1) (Keller and Weber, 1995). Metolachlor was less mobile than aldicarb {2-methyl-2-(methylthio)propanal-O-[(methylamino)carbonyl]oxime}, a nonionizable, highly aqueous soluble (Ks = 6000 mg L–1) insecticide with low soil sorptivity (Kd = 0.03 to 0.7 mL g–1) (Bowman, 1988) and carbofuran (2,3-dihydro-2,2-dimethyl-7-benzofuranyl methyl carbamate), a nonionizable, moderately aqueous soluble (Ks = 700 mg L–1) insecticide with low soil sorptivity (Kd = 0.3 mL g–1) (Levanon et al., 1993). It has also been reported to be equal to or greater in soil mobility than alachlor [2-chloro-N-(2,6-diethylphenyl)-N-(methoxymethyl)acetamide], a nonionizable, moderately aqueous soluble (Ks = 242 mg L–1) related herbicide with low to moderate soil sorptivity (Kd = 0.3 to 3.7 mL g–1) (Alhajjar et al., 1990; Vasilakoglou et al., 2001).

Primisulfuron-methyl is a weakly acidic, low aqueous soluble herbicide used to control grasses as well as many broadleaf weeds when applied postemergence (Ahrens, 1994) (Table 1). Soil sorption of many sulfonylurea herbicides, including primisulfuron-methyl, has been reported to be inversely related to pH (Sarmah et al., 1998; Shea, 1986; Ukrainczyk and Ajwa, 1996; Walker and Welch, 1989; Werkheiser and Anderson, 1996) and positively related to clay content and OM content in acid soils but not in alkaline soils (Sarmah et al., 1998; Stehouwer et al., 1994; Ukrainczyk and Ajwa, 1996; Vicari et al., 1996; Walker et al., 1989; Werkheiser and Anderson, 1996) and to extractable Fe and Al in soils low in OM content and pH (Ukrainczyk and Ajwa, 1996; Werkheiser and Anderson, 1996). Strek et al. (1990) reported that sorption of several sulfonylureas was more strongly related to humic matter (HM) content than OM content, probably due to the greater hydrophobic surfaces of HM. Sulfonylurea herbicides were relatively mobile in soil where mobility increased with rate of application (Bergström, 1990) and soil pH (Beyer et al., 1988; Goetz et al., 1986) and decreased with OM content (Beyer et al., 1988; Mercie and Foy, 1986; Shea, 1986). Relative to other herbicides, primisulfuron-methyl was reported to be more mobile than atrazine and metolachlor, respectively, in short-term laboratory studies (Lee and Weber, 1993) but less mobile than the two herbicides in 90-d field lysimeter studies (Keller and Weber, 1995; Lee and Weber, 1993), due to its shorter disappearance time [50% disappearance time (DT50)] and lower rate of application. Regarding other weakly acidic sulfonylureas, primisulfuron-methyl was less mobile than chlorsulfuron {2-chloro-N-[[4-methoxy-6-methyl-1,3,5-triazin-2-yl)amino]carbonyl]benzenesulfonamide}, with pKa = 3.6, high Ks (587 mg L–1), and low soil retention (Kd = 0.69 mL g–1); more mobile than chlorimuron-ethyl {ethyl-2-[[[[(4-chloro-6-methoxy-2-pyrimidinyl)amino]carbonyl]amino]sulfonyl]benzoate}, with pKa = 4.2, low Ks (11 mg L–1), and low soil retention (Kd = 0.55 mL g–1); and of equal mobility with triasulfuron {2-(2-chlorethoxy)-N-[[(4-methoxy-6-methyl-1,3,5-triazin-2-yl)amino]carbonyl]benzenesulfonamide}, with pKa = 4.64, low Ks (32 mg L–1), and low soil retention (Kd = 0.62–1.16 mL g–1) in soil at pH 5–5.4 (Rahman and James, 1989). Mobility differences were no doubt due to differences in physicochemical properties and to differences in abiotic and biotic degradation rates. Chlorsulfuron was also more mobile than triasulfuron in laboratory studies (Weber et al., 1999) and in field studies (Stork, 1995; Walker and Welch, 1989).

The objectives of these studies were to: (i) determine the relative leaching patterns of 14C-labeled atrazine, metolachlor, and primisulfuron-methyl in soil column field lysimeters of four soils with widely differing chemical properties and (ii) compare and evaluate the leachability parameters of the herbicides and soils with physicochemical and biological properties of the chemicals and physiochemical properties of the soils.


   MATERIALS AND METHODS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS AND DISCUSSION
 REFERENCES
 
Installing and Extracting Field Lysimeters
Lysimeters consisted of 0.1-cm-thick steel columns (20 cm i.d. and 96 cm in length), which were driven 91 cm into the soil using a tractor-mounted post driver. The 5-cm lip remaining above the soil prevented surface runoff from entering the lysimeters and radiolabeled herbicide from leaving the treated soil. In February of 1992 and 1993, Dothan loamy sand (fine-loamy, siliceous, thermic, Plinthic Kandiudult) and Wagram loamy sand (fine-loamy, siliceous, thermic, Arenic Kandiudult) columns were obtained from the Central Crops Research Station, Clayton, NC; Portsmouth sandy loam (fine-loamy, over sandy, mixed, thermic, Typic Umbraquult) columns were obtained from the Tide Water Research Station, Plymouth, NC; and Rion sandy loam (fine-loamy, kaolinitic, thermic, Typic Kandiudult) columns were obtained from the Upper Piedmont Research Station, Reidsville, NC. Columns were extracted from the soil at each site by covering the columns, drilling a hole alongside, loosening with a shovel, and hoisting the columns out of the soil using C-clamps and block and tackle. Tops and bottoms of the columns were covered with polyethylene, and the columns were placed upright on thick foam pads and transported to the Central Crops Research Station where they were installed in predrilled holes alongside a 1-m-wide by 2-m-deep trench until soil inside the columns was level with soil outside. A factorial treatment experimental design laid out in a completely randomized block pattern with two replications was used for each soil and herbicide combination plus untreated controls. Soil was excavated from under each column with a 20-cm-diam. auger, and an aluminum-foil-lined funnel and a l-L glass jar completing each lysimeter system was attached to each column.

Herbicide Treatments
On 5 June 1992 and 7 June 1993, the surface 13 cm of soil in each column was tilled and leveled, and application of each herbicide was made uniformly onto the soil in a cross-hatch pattern in 10 mL of water using a pipette. Atrazine treatments consisted of a mixture of AAtrex Nine O (90% atrazine wettable granule, Syngenta Chemical Corp., Greensboro, NC) and 14C-ring-labeled atrazine (specific activity = 1.40 TBq kg–1 in 1992 and 0.53 TBq kg–1 in 1993, 99.5% purity) at rates of 2.80 kg a.i. ha–1 each year and 0.78 MBq in 1992 and 0.57 MBq in 1993, respectively. Metolachlor treatment consisted of a mixture of Dual 8E (960 g a.i. L–1 metolachlor, Syngenta Chemical Corp., Greensboro, NC) and 14C-ring-labeled metolachlor (specific activity = 1.09 TBq kg–1 in 1992 and 1.05 TBq kg–1 in 1993, 98% purity) at rates of 2.24 kg a.i. ha–1 each year and 0.70 MBq in 1992 and 0.56 MBq in 1993, respectively. Primisulfuron-methyl treatments consisted of a mixture of analytical grade (99.9%) primisulfuron-methyl and 14C-phenyl-labeled primisulfuron-methyl (specific activity = 1.94 TBq kg–1 in 1992 and 2.01 TBq kg–1 in 1993, >95% purity) at rates of 0.088 kg a.i. ha–1 each year and 0.45 MBq in 1992 and 0.57 MBq in 1993, respectively. Soil moisture each year was roughly at field capacity (lysimeters drained for 3 d after a rainfall event of at least 25 mm) when herbicide applications were made. Zero-day soil samples for all herbicide–soil treatments were fortified with herbicide solution in the field, mixed, refrigerated to 4°C, and analyzed within 48 h.

Soil Sample Processing and Analysis
In October of each year, at 128 d after treatment (DAT), lysimeters were removed and the soil sectioned every 7.6 cm by depth for a total of 12 soil samples per column. Each sample was placed in a polyethylene bag, sealed, labeled, and frozen at –20°C until analyzed. Untreated soil columns were extracted at the same time and sectioned in the same manner for soil property characterization and background radiation correction. Soil samples were sent to A & L Midwest Agricultural Laboratory, Omaha, NE and to the North Carolina Department of Agriculture Soil and Plant Laboratory, Raleigh, NC to be analyzed for OM and HM content, particle-size analysis, and pH. Organic matter content was determined by chromic acid oxidation colorimetric method (Nelson and Sommers, 1982) and HM by the NaOH–DTPA–alcohol extraction method (Mehlich, 1984). Particle-size analyses, for clay contents, were performed using the hydrometer method (Gee and Bauder, 1986) and pH (1:1 soil:water) using a glass electrode pH meter and standards. Carbon-14 radioactivity for each thoroughly mixed soil section was determined by using four 1-g subsamples, which were combusted in a biological oxidizer (Model OX-300, R.J. Harvey Instrument Co., Hillsdale, NJ; 96% efficiency) and the 14CO2 trapped in 15 mL of Harvey 14C scintillation cocktail. Carbon-14 radioactivity was assayed using a liquid scintillation analyzer (LSA) (Packard Model 2000CA, Packard Instrument Co., Downers Grove, IL) and converted to total 14C activity recovered for each section after correcting for background radiation (CV< 20%). Five hundred grams of the 8-cm surface soil from each column were bagged, labeled, and frozen for herbicide parent and metabolite identification and quantification.

Parent and Metabolite Analysis
Gravimetric water content determinations were performed on 20 to 30 g of soil from the surface 8-cm layers of each column, and 100-g (dry weight basis) samples were combined with the proper solvent in 500-mL flasks maintaining a 1:1 ratio of soil:solvent. Atrazine and metolachlor were extracted with methanol and primisulfuron-methyl with acetonitrile. All samples were shaken for 4 h on a horizontal shaker. Slurries were filtered through glass fiber filter paper (Whatman 5B) under reduced pressure. Filtrates were transferred to 500-mL round bottom flasks and evaporated to <2 mL at 40°C under reduced pressure. Analytes were transferred from boiling flasks into graduated centrifuge tubes containing 15 mL of their respective solvents. Duplicate 0.5-mL samples were added to 15 mL of Fisher Scientific BD Cocktail (Fisher Scientific, Pittsburg, PA) and assayed by LSA to evaluate recoveries. The centrifuge tubes were then placed in a water bath at 40°C and the samples air-evaporated (N2) to <2 mL. Concentrated extracts (150–500 µL) were spotted on normal-phase thin-layer chromatography (TLC) plates (Whatman LK5L Linear TLC, Whatman USA, Hillsboro, OR). The TLC plates were developed in the following solvent systems: for atrazine, 1:1 ethyl acetate:toluene (v/v); for metolachlor, 75:20:4:2 chloroform:methanol:formic acid 90%:water (v/v/v/v); and for primisulfuron-methyl, 75:20:5 toluene:acetone:formic acid 90% (v/v/v). To verify primary solvent system separations, each chemical was spotted on TLC plates and developed in secondary solvent systems. Developed TLC plates were scanned on a BioScan System 400 Imaging Scanner (BioScan Inc., Washington, DC). Reached (retardation) factor (Rf) values were measured on the scans. Recovery of extractable plus nonextractable was >93%. Nonlabeled analytical standards of parent and metabolites (Syngenta Corp., Greensboro, NC) were spotted on TLC plates to identify parents and metabolites using a short-wave ultraviolet lamp system.

Leachate Collection and Analysis
Leachate, collected weekly or according to rainfall intensity, was transported immediately to the laboratory where it was quantified and duplicate 1-mL subsamples assayed by LSA. Selected leachate samples containing >1700 Bq L–1 in >50 mL were frozen in polyethylene bags until analyzed. Frozen leachate samples were extracted with SPE C18 Environmental Plus Sep-pack cartridges (Waters Chromatography Div. of Millipore Corp., Milford, MA), after first being preconditioned with 10 mL of methanol (atrazine and metolachlor) or acetonitrile (primisulfuron-methyl) followed by 10 mL of deionized water. Atrazine and metolachlor analytes were eluted from the Sep-packs with 10 mL of methanol and primisulfuron-methyl with 10 mL of acetonitrile after first being acidified to pH 3 with HCl to increase extraction efficiency. The flow rate ranged from 2 to 5 mL min–1. Two 0.5-mL subsamples of each eluate were assayed by LSA to evaluate recoveries. The remaining eluate volumes were reduced to <1 mL by air (N2) evaporation in a water bath. Total recoveries were >90%. Concentrated analytes were spotted on TLC plates and assayed as described previously.

For each lysimeter, a mobility index (MI = {sum} D x F, where D = depth in cm and F = fraction of chemical present) and Rf (Rf = MI/MImax) values were calculated from the distribution of total 14C recovered in the soil profile, normalized to 100% recovered for all compounds, as described by Weber et al. (1999). The amount volatilized was also calculated using the following equation: percentage volatilized = 100% (applied) – percentage 14C in soil – percentage 14C in leachate. Relative pesticide leaching potential (PLP), soil leaching potential (SLP), and groundwater contamination potential (GWCP) indices were calculated using a simple decision-aid model (Warren andWeber, 1994; Weber, 2005). Analyses of variance (ANOVA) were conducted to evaluate herbicide differences between years, herbicides, soil types, depths, etc. (SAS Inst., 1988). Correlation analyses (r values) were performed and are designated significant at the 1% (**), 5% (*), and 10% ({dagger}) levels. All 14C wastes were disposed of following proper procedures (Dep. of Environ. Health and Hazardous Materials Manage., Life Safety Serv., North Carolina State Univ., 1991).

Climatic Conditions
Precipitation data were recorded at the CCRS Weather Station of the Central Crops Research Station, Clayton, NC (NOAA Site no. 31–1820–07). Weekly precipitation data of the previous 10 yr were used to establish the weekly precipitation average, which if not reached, was added to the lysimeters by irrigation. Air and soil temperatures were recorded at the Smithfield Weather Station, Smithfield, NC (NOAA Site 19 no. 31–7794–07).


   RESULTS AND DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
REFERENCES
 
Soil Properties
Organic matter and HM contents and pH decreased with depth while clay content increased for all soils (Fig. 1A –1D). Mean percentage OM contents for Dothan, Portsmouth, Rion and Wagram soil profiles were 0.6, 1.8, 0.9, and 0.9, respectively, and mean percentage HM contents were 0.2, 1.1, 0.1, and 0.4, respectively. Mean percentage clay contents were 15, 20, 27, and 7, respectively, and mean pH levels were 4.9, 5.2, 5.5, and 5.2, respectively. Based on the properties of the soil profiles, SLP indices of 66, 26, 64, and 71 were calculated for Dothan, Portsmouth, Rion, and Wagram soils, respectively (Warren and Weber, 1994; Weber, 2005).



View larger version (32K):
[in this window]
[in a new window]
  		Fig. 1. Soil property profile of (A) Dothan loamy sand, (B) Portsmouth sandy loam, (C) Rion Sandy loam and (D) Wagram loamy sand (values x 10 for clay). SLP = soil leaching potential, HM = humic matter, and OM = organic matter (Weber, 2005).

 
Climate Conditions
Monthly water input (precipitation plus irrigation) for June, July, August, September, and October was 316, 29, 184, 97, and 18 mm [total = 644 mm (21 L)], respectively, in 1992 and 106, 151, 119, 96, and 43 mm [total = 515 mm (17 L)], respectively, in 1993. Mean air temperatures for the same periods in 1992 were 22.6, 27.5, 23.9, 22.7, and 13.7°C, respectively, and 25.3, 28.0, 24.7, 22.8, and 14.3°C, respectively, for the same periods in 1993. Mean soil temperatures for the same periods in 1992 were 25.8, 28.9, 27.8, 26.9, and 20.3°C, respectively, and 26.0, 31.1, 29.7, 29.7, and 21.4°C, respectively, for the same periods in 1993.

Herbicide Mass Balance Distribution
Analysis of variance indicated that 14C distribution in soil and leachate recoveries were significantly different between 1992 and 1993 growing seasons (1% level); therefore, each year's data were analyzed separately. Factorial ANOVA followed by repeated measurement analysis indicated a significant (5% level) herbicide x soil interaction effect among treatments and within treatment using soil depth as the repeated factor. Comparisons are within year, between herbicides (first letter) [atrazine (A), metolachlor (M), and primisulfuron-ethyl (P)], and across soils (second letter) [Dothan (D), Portsmouth (Po), Rion (R), and Wagram (W)]; i.e., for 1992, percentage 14C volatilized was PR (64.2a) ≥ PD (64.3ab) = PPo (62.1ab) = MD (55.0ab) = AW (54.7ab) = AD (54.0ab) ≥ APo (51.5bc) = AR (51.4bc) ≥ PW (39.6cd) = MW (38.6cd) ≥ MR (30.7d) = MPo (28.0d) (Tables 2 to 4). A mass balance distribution of 14C from each herbicide in each soil, at 128 DAT each year, is presented in Tables 2 to 4.


View this table:
[in this window]
[in a new window]
  		Table 2. Mass balance distribution of 14C recovered from 14C-atrazine-treated fallow field lysimeters at 128 d after treatment.

 

View this table:
[in this window]
[in a new window]
  		Table 3. Mass balance distribution of 14C recovered from 14C-metolachlor-treated fallow field lysimeters at 128 d after treatment.

 

View this table:
[in this window]
[in a new window]
  		Table 4. Mass balance distribution of 14C recovered from 14C primisulfuron-methyl-treated fallow field lysimeters at 128 d after treatment.

 
Carbon-14 volatilization losses over the 128-d period each year ranged from 51 to 61% for atrazine, 28 to 55% for metolachlor, and 40 to 69% for primisulfuron-methyl with few differences (Tables 2 to 4). Analysis of variance for main herbicide effects indicated metolachlor volatilization losses were slightly less each year than those for the other two herbicides, which had equal losses, probably because more 14C metolachlor had leached and was recovered in the subsoil (Table 5). Also, main soil effects on volatilization losses were higher from Dothan than from the other three soils, which had equal losses in 1992 (Table 6). Portsmouth had lower losses than the other three soils, which had equal losses in 1993. The results were in general agreement with losses of 45 to 59% for the three herbicides over a 92-d period (Lee and Weber, 1993) and 43 to 62% losses over a 30-d period from the Dothan soil (Keller and Weber, 1995), 21 to 37% losses over a 30-d period for atrazine (Weber et al., 2002), and 27 to 60% losses of metolachlor over a 90-d period from the same soil (Keller and Weber, 1997; Keller et al., 1998).


View this table:
[in this window]
[in a new window]
  		Table 5. Herbicide and year main effects (means) of 14C distribution (percentage of 14C applied) in fallow soil lysimeters at 128 d after treatment.

 

View this table:
[in this window]
[in a new window]
  		Table 6. Soil and year main effects (means) of 14C distribution (percentage of 14C applied) in fallow soil lysimeters at 128 d after treatment.

 
Carbon-14 herbicide recovered from surface (0 to 7.6 cm) soil at 128 DAT ranged from 18 to 31% in 1992 and 29 to 40% in 1993 for atrazine, 13 to 29% in 1992 and 27 to 44% in 1993 for metolachlor, and 21 to 26% in 1992 and 21 to 40% in 1993 for primisulfuron-methyl (Tables 2 to 4). Herbicide main effects showed that atrazine and primisulfuron-methyl were present in surface soil in equal but greater amounts than metolachlor in 1992, but atrazine and metolachlor were present in equal but greater amounts than primisulfuron-methyl in 1993 (Table 5). Carbon-14 herbicide distribution in surface (0 to 7.6 cm) soil ranged from 13 to 30% in Rion, 13 to 34% in Wagram, 14 to 29% in Dothan, and 21 to 44% in Portsmouth over the 2-yr period (Tables 2 to 4). Soil main effects indicated that Portsmouth retained more of the herbicides in the surface soil both years than the other three soils, which retained similar amounts (Table 6). Portsmouth surface soil also contained two to three times higher OM and 8 to 16 times higher HM levels than the other three soils (Fig. 1A to 1D), and many investigators have shown that OM and HM fractions in soil are highly sorptive for organic chemicals. Higher concentrations of the herbicides remained in the surface soil in the 1993 growing season (mean = 31%) than in 1992 (mean = 21%) likely due to the 28% higher water input in 1992, which moved 10% more of the herbicides down into the subsoil.

Solvent extractable 14C-atrazine, 14C-metolachlor, and 14C-primisulfuron-methyl and their respective metabolites from 0- to 7.6-cm surface soil in the 2-yr study ranged from 0.6 to 3.6% for atrazine, 0.8 to 6.7% for metolachlor, and 1.2 to 5.8% for primisulfuron-methyl (Tables 2 to 4). Herbicide main-effect means indicated that atrazine was less extractable than metolachlor and primisulfuron-methyl, which were equally extractable by their respective solvent systems (Table 5). Main soil effect means indicated that the herbicides were extracted from the Portsmouth soil in greater amounts than from Wagram, Dothan, and Rion soils, respectively, both years (Table 6). The mean portion of each extracted herbicide that was parent compound [(percentage parent)/(percentage extractable) x 100] was 39.5% for atrazine, 72.7% for metolachlor, and 65.8% for primisulfuron-methyl, and the remainder consisted of four to five metabolites (Tables 2 to 4). The majority of each herbicide present in surface soil at 128 DAT each year was in the bound form and ranged from 11 to 37% (mean = 23%).

Carbon-14 in subsoils at 128 DAT ranged from 16 to 27% in 1992 and 6 to 10% in 1993 for atrazine, 28 to 38% in 1992 and 14 to 25% in 1993 for metolachlor, and 11 to 33% in 1992 and 6 to 12% in 1993 for primisulfuron-methyl (Tables 2 to 4). Herbicide main effects indicated that metolachlor leached to subsoils in greater amounts than atrazine and primisulfuron-methyl, which leached equally in 1992 (Table 5). Metolachlor leached in greater amounts than primisulfuron-methyl, which leached in greater amounts than atrazine in 1993. Carbon-14 in specific subsoils ranged from 11 to 28% in 1992 and 10 to 21% in 1993 for Dothan, 15 to 35% in 1992 and 6 to 14% in 1993 for Portsmouth, 11 to 35% in 1992 and 8 to 21% in 1993 for Rion, and 27 to 38% in 1992 and 10 to 25% in 1993 for Wagram (Tables 2 to 4). Soil main effects indicated that 14C in subsoils was greater in Wagram than in Rion, Portsmouth, and Dothan, respectively, in 1992 (Table 6). It was present in equal amounts in Wagram, Rion, and Dothan, but in greater amounts than Portsmouth in 1993.

Total 14C soil recovery from each lysimeter ranged from 43 to 47% in 1992 and 39 to 46% in 1993 for atrazine, 42 to 65% in 1992 and 48 to 58% in 1993 for metolachlor, and 35 to 59% in 1992 and 31 to 46% in 1993 for primisulfuron-methyl (Tables 2 to 4). Herbicide main effects for total 14C soil recovery indicate that more metolachlor was recovered than atrazine or primisulfuron-methyl in each of the 2 yr (Table 5). Total 14C soil recovery in the respective soils ranged from 37 to 45% in 1992 and 38 to 48% in 1993 for Dothan, 36 to 65% in 1992 and 46 to 58% in 1993 for Portsmouth, 35 to 48% in 1992 and 31 to 51% in 1993 for Rion, and 45 to 59% in 1992 and 33 to 53% in 1993 for Wagram (Tables 2 to 4). Soil main effects on total soil 14C recovery indicate that quantities recovered from Portsmouth and Wagram were equal and greater than that recovered from Dothan and Rion, which also had equal amounts in 1992 (Table 6). Greater amounts were recovered from Portsmouth than from the other three soils, which had equal amounts in 1993.

Cumulative 14C recovered in leachate over the 128-d period ranged from 0.2 to 5.5% in 1992 and 0.1 to 1.0% in 1993 for atrazine, 2.7 to 21% in 1992 and 0.4 to 2.8% in 1993 for metolachlor, and 0.2 to 2.3% in 1992 and 0.1 to 0.7% in 1993 for primisulfuron-methyl (Tables 2 to 4). Herbicide main effects on cumulative 14C in leachate indicate that three to nine times more metolachlor was recovered in the leachate in 1992 and 1993, respectively, than atrazine and primisulfuron-methyl, which were recovered in equal amounts both years (Table 5). Cumulative 14C in leachate of the respective soils ranged from 0.2 to 2.7% in 1992 and 0.1 to 0.4% in 1993 for Dothan, 1.6 to 7.2% in 1992 and 0.1 to 0.4% in 1993 for Portsmouth, 0.3 to 21% in 1992 and 0.7 to 2.8% in 1993 for Rion, and 0.2 to 10.8% in 1992 and 0.1 to 0.1% in 1993 for Wagram (Tables 2 to 4). Soil main effects on cumulative 14C recovered in 1992 leachate indicate that more than twice as much 14C was found in Rion than in Wagram or Portsmouth, which had equal amounts that were more than three times that found in Dothan (Table 6). In 1993, seven times more 14C was recovered in the leachate of Rion than in the other three soils, which had equal amounts. Greater leaching of the herbicides in 1992 than in 1993 was due to the greater water input in 1992.

Carbon-14 parent herbicide recovered in leachate over the 128-d period ranged from 0 to 1.7% for atrazine, 0 to 2.3% for metolachlor, and 0% for primisulfuron-methyl in the 2-yr study (Tables 2 to 4). The concentrations of parent herbicide found were small, but it is generally accepted that when amounts reaching 90 cm deep in the soil are greater than 1% of applied compound, there is a fair chance that the material may reach groundwater. Herbicide main effects of 14C parent recovered in leachate indicate that metolachlor was found in three times greater amounts than atrazine or primisulfuron-methyl in 1992, but none of the parent herbicides were found in leachate in 1993 due to the lower water input in 1993 (Table 5). Carbon-14 parent herbicide recovered in leachate of the respective soils over the 128-DAT period ranged from 0 to 0.3% in 1992 and 0% in 1993 for Dothan, 0% in 1992 and 1993 for Portsmouth, 0 to 2.2% in 1992 and 0% in 1993 for Rion, and 0 to 2.3% in 1992 and 0% in 1993 for Wagram (Tables 2 to 4). Soil main effects on 14C parent recovered in the leachate indicated that parent herbicide was more mobile in Rion than in Wagram, Dothan, and Portsmouth, respectively, in 1992 and that no 14C parent was found in leachate from any of the soils in 1993 (Table 6).

Herbicide soil mobility was computed from herbicide distribution in the soil and leachate (added to the bottom section of each column) and presented as MI and Rf values. Only Rf values will be discussed as they are more commonly referred to in the literature. The Rf values ranged from 0.12 to 0.27 in 1992 and 0.08 to 0.13 in 1993 for atrazine, 0.25 to 0.51 in 1992 and 0.11 to 0.23 in 1993 for metolachlor, and 0.09 to 0.22 in 1992 and 0.07 to 0.18 in 1993 for primisulfuron-methyl (Tables 2 to 4). Herbicide Rf main effects indicate that metolachlor was twice as mobile as primisulfuron-methyl and atrazine, which were equal in mobility both years, thus the mobility Rf ranking: metolachlor > atrazine = primisulfuron-methyl (Table 5). The Rf values for the herbicides in the four soils ranged from 0.09 to 0.31 in 1992 and 0.09 to 0.17 in 1993 for Dothan, 0.12 to 0.25 in 1992 and 0.07 to 0.11 in 1993 for Portsmouth, 0.10 to 0.51 in 1992 and 0.13 to 0.23 in 1993 for Rion, and 0.15 to 0.34 in 1992 and 0.08 to 0.17 in 1993 for Wagram (Tables 2 to 4). Soil main effects on Rf values of the herbicides indicate that soil mobility of the chemicals was Rion > Wagram ≥ Dothan ≥ Portsmouth (Table 6).

The total leachate volume collected for the herbicides ranged from 2.3 to12.7 L in 1992 and from 2.1 to 4.3 L in 1993 (Tables 2 to 4). Herbicide main effects on total leachate collected indicated that herbicide type had no effect on the amount of leachate collected (Table 5). Soil main effects on total leachate volume collected were 1.5 times greater and equal in Portsmouth and Wagram soils compared with Rion and Dothan, which also were also equal in 1992 (Table 6). In 1993, Dothan and Wagram had greater and equal amounts of leachate than Rion and Portsmouth, which were also equal, suggesting that columns of the same soil type probably vary from site to site across the landscape. Year differences were largely due to the much greater water input and thus greater amounts of leachate collected in 1992 over that of 1993.

Computed pesticide PLP indices for the three herbicides were in agreement with soil mobility of the chemicals as indicated by quantity of 14C in the subsoil, total soil 14C recovered, quantity of 14C in the leachate, and quantity of 14C parent in leachate, in most cases, and although 14C mobility Rf values indicated metolachlor to be the most mobile compound, it did not distinguish between atrazine and primisulfuron-methyl (Table 5). The PLP rankings for the three herbicides were metolachlor > atrazine > primisulfuron-methyl.

Computed SLP indices for the four soils provided the ranking: Wagram > Rion = Dothan > Portsmouth (Table 6). Highest amounts of herbicide were found in subsoil and lowest amounts in surface soil of Wagram, and highest amounts of herbicide were found in surface soil and lowest amounts in subsoil of Portsmouth, but amounts found in surface and subsoil of Dothan and Rion were inconsistent.

Computed GWCP indices of herbicide/soil combinations indicated that metolachlor was generally the most mobile herbicide in the four soils, followed by atrazine and primisulfuron-methyl, and that the herbicides were least mobile in Portsmouth soil, but it was apparent that both herbicide and soil characteristics were important (Table 6).

Correlation Analysis
Correlations (r values) were performed among herbicide main-effect parameters (Table 5) and soil main-effect parameters (Table 6) and between main-effect parameters (Tables 5 and 6) and physicochemical and biological properties of herbicides (Table 1) and soils (Fig. 1 and Table 6). Carbon-14-volatilized main-effect means of the three herbicides (Table 5) were inversely correlated with 14C in leachate (r = –0.99*), 14C soil mobility (Rf) (r = –0.99*) and total soil 14C recovered (r = –0.99**) (Table 5), and herbicide longevity (DT50) (r = –0.99*) (Table 1).

Carbon-14-in-subsoil main-effect means of the three herbicides (Table 5) were correlated with total soil 14C recovered (r = 0.95**), 14C parent herbicide in leachate (r = 0.99*) and 14C soil mobility (Rf) (r = 0.99*) (Table 5), and with herbicide longevity (DT50) (r = 0.99*) (Table 1). Main-effect herbicide means of total soil 14C recovered (Table 5) were correlated with herbicide longevity (DT50) of the three herbicides (r = 0.99*) (Table 1). Carbon-14-in-leachate main-effect means of the three herbicides (Table 5) were correlated with 14C soil mobility (Rf) of the chemicals (r = 0.99*) (Table 5) and with herbicide longevity (DT50) values (r = 0.98*) (Table 1).

Carbon-14 parent herbicide in leachate main-effect means of the three herbicides (Table 5) were correlated with herbicide longevity (DT50) (r = 0.99*) (Table 1). Carbon-14 herbicide soil mobility (Rf) main-effect means of the three herbicides (Table 5) were correlated with herbicide longevity (DT50) (r = 0.99*) and herbicide Ks of the chemicals (r = 0.99*) (Table 1). Carbon-14 volatilized main-effect means of the four soils (Table 6) were inversely correlated with total 14C recovered from the four soils (r = –0.99**) (Table 6) and mean percentage HM contents of the four soils (r = –0.99**) (Fig. 1).

Surface soil 14C, soil extractable 14C, and total soil 14C recovered main-effect means of the four soils (Table 6) were correlated with mean percentage HM contents of the four soils (r = 0.97,* r = 0.99,* and r = 0.99,** respectively) (Table 1). Carbon-14 in leachate and 14C parent in leachate main-effect means of the four soils (Table 6) were correlated with mean mobility (Rf) values of the herbicides in the soils (r = 0.92{dagger} and r = 0.95,* respectively) (Table 6).

Mean OM contents of the four soil profiles were correlated with their respective mean HM contents (r = 0.94{dagger}) (Fig. 1), and mean OM and HM contents of the soils were inversely correlated with SLP indices of the four soils (r = –0.94* and r = -0.94,* respectively) (Table 6). All of the correlations that occurred were anticipated, and many were previously reported in the literature. Year differences were attributed primarily to differing water inputs although other climatic factors may have also been involved.

In summary, correlation analyses indicated the following: (i) leaching of atrazine, metolachlor, and primisulfuron-methyl out of surface soils over the 128-d period was accompanied by an increase in the herbicides and their metabolites detected in subsoils and in leachate and was directly related to herbicide longevity (DT50) and inversely related to herbicide/soil binding factor (Kd); (ii) soil mobility (Rf) of the three herbicides was related to herbicide Ks, amount of input water, and herbicide longevity (DT50); (iii) loss of the three herbicides through volatilization was inversely related to mean percentage HM contents of the four soils; (iv) amounts of the three herbicides remaining in the surface soil after 128 d were related to the mean percentage HM contents of the four soils and were inversely related to their calculated SLP indices; and (v) calculated groundwater contamination potential (GWCP) indices for the herbicide/soil combinations were related to parent herbicides recovered in the leachates and subsoils and to soil mobility (Rf) values and inversely related to herbicide remaining in the soil surface.

Field lysimeters closely approximate field conditions with the exception that they maximize vertical movement of water, which could be important in soils with high lateral movement, and in our system, no runoff occurred, so all input water, with the exception of capillary water returning to the surface, was directed downward through the soil. This study showed that physicochemical and biological properties of herbicides and physicochemical properties of the soils are both important in predicting herbicide movement and dissipation in soil.


   ACKNOWLEDGMENTS
 
We acknowledge Syngenta Corporation for providing products; L.R. Swain, N. Enders, R.L. Warren, G.B. Clark, N.J. Weber, and P.J. Weber for technical assistance; the North Carolina Agricultural Foundation for financial support; and Judith Abbott-Goodman for secretarial support. This study comprises a portion of K.A.T.'s 1997 Ph.D. thesis.


   REFERENCES
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 REFERENCES
 





This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF) Free
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via ISI Web of Science (1)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Weber, J. B.
Right arrow Articles by Wilkerson, G. G.
Right arrow Search for Related Content
PubMed
Right arrow Articles by Weber, J. B.
Right arrow Articles by Wilkerson, G. G.
Agricola
Right arrow Articles by Weber, J. B.
Right arrow Articles by Wilkerson, G. G.
Related Collections
Right arrow Tillage
Right arrow Irrigation
Right arrow Soybean
Right arrow Turfgrass

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
The SCI Journals Crop Science Vadose Zone Journal
Journal of Natural Resources
and Life Sciences Education		 Soil Science Society of America Journal
Journal of Plant Registrations Journal of
Environmental Quality		 The Plant Genome