| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
a Dep. of Land Resource Science, Univ. of Guelph, Guelph, ON, Canada N1G 2W1
b Dep. of Environmental Biology, Univ. of Guelph, Guelph, ON, Canada N1G 2W1
* Corresponding author (gparkin{at}lrs.uoguelph.ca)
Received for publication May 26, 2000.
 |
ABSTRACT |
|---|
 TOPNOTES ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
Abbreviations: GC–MS, gas chromatography–mass spectrometry • HPLC, high pressure liquid chromatography
|
INTRODUCTION |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
Most researchers have found only very low levels of dicamba in infiltrating water and low total mass of leached dicamba when applied to turf. For lawn or golf green conditions, few of the percolate samples collected between 0.10 to 0.20 m below surface had concentrations exceeding 10 µg L-1 (Gold et al., 1988; Smith and Bridges, 1996; Snyder and Cisar, 1997). In these previous studies applications took place from spring to late fall, with rates ranging from 0.11 to 0.28 kg ha-1. Although soil cores taken by Snyder and Cisar showed noticeable peaks of dicamba concentration in soil and thatch shortly after application, these rapidly declined over the following 2 wk.
In a field study by Harrison et al. (1993), relatively high concentrations of dicamba were measured in run-off and/or percolate on sloped turf grown on clay top-soil. They applied dicamba at 0.28 kg ha-1 in July, August, September, October, and November over 2 yr. The peaks in concentration for leachate samples collected at the 0.15-m depth coincided with the first major irrigation or rainfall within a week after application, although Harrison et al. (1993) considered the irrigation events extreme. The mean concentrations (three replicate samples) of dicamba in leachate following these events were 11, 21, 22, 118, and 57 µg L-1.
Only recently have researchers modeled the transport of solutes applied to turfgrass. The GLEAMS model greatly overpredicted pesticide leaching below turf (Smith et al., 1993). A number of researchers (Pennell et al., 1990; Costa et al., 1994; Brown et al., 1996) indicated that LEACHM (Wagenet and Hutson, 1987) simulates solute transport in unsaturated soil as well or better than other models. However, LEACHM predicted significantly higher dicamba concentrations in leachate and longer travel times for dicamba transport than was found under field conditions using turfgrass (T. Watschke, Pennsylvania State Univ., personal communication, 1997).
Part of the problem with these modeling results may be the lack of site-specific data on soil properties and dicamba attenuation processes. Turfgrass is unique because of the thatch layer, as well as the constant crop cover and the long growing season, which lead to high evapotranspiration rates. In a turf system, pesticides with low volatility, like dicamba, can be sorbed to soil particles or organic matter, taken up by plants, and/or degraded. Baskaran et al. (1997) studied sorption of dicamba to golf green materials (a sand–peat soil mixture and thatch). In their batch-sorption experiment, the adsorption coefficient, Kd, was low (0.54 mL g-1) in soil and moderate (3.29 mL g-1) in thatch. However, in their column leaching experiment, dicamba moved at the same rate as water. Dicamba is rapidly absorbed by plant roots and foliar tissues, and translocated to other parts of the plant (Frear, 1976), after which it is metabolized or exuded via the roots or leaves (Caux et al., 1993). However, Snyder and Cisar (1997) recovered very little applied dicamba in turfgrass clippings from a golf green. It is not clear how important these two processes are in enhancing the dissipation of pesticides applied to turfgrass, compared with degradation.
Dicamba degradation is primarily microbial, as opposed to chemical, and is highly variable. Reported half-life values range from less than 10 d (Burnside and Lavy, 1966; Smith, 1974) to 151 d (Comfort et al., 1992). The main metabolite in soil is 3,6-dichlorosalicylic acid (3,6-DCSA), a less mobile and more persistent compound than dicamba (Smith, 1974; Pearson et al., 1996b). Biodegradation can be influenced by soil properties, including pH and organic matter content, and seasonal climatic factors, such as soil moisture content and temperature (Caux et al., 1993). Therefore, the knowledge of degradation rates of dicamba for different environmental conditions is required to better understand the dissipation of this herbicide in the field. In addition, most degradation studies have only used soil; to date there have been very few direct studies on pesticide degradation in the thatch layer below turf. Increased dissipation has been reported (Hurto and Turgeon, 1979; Branham and Wehner, 1985; Liu and Hsiang, 1996), but for these studies either the authors did not use sterile controls to differentiate between chemical and biological processes, or they did not test soil, but used literature values for comparison. Degradation studies comparing thatch and soil have not been performed for dicamba.
The objectives of this study were to (i) investigate the leaching of dicamba in unsaturated, sandy soil using large field lysimeters topped with Kentucky bluegrass turf, (ii) measure the degradation rate of dicamba in soil and thatch in the laboratory under simulated field conditions, and (iii) test the ability of the model LEACHM (in EXPRES) to simulate the transport and degradation of dicamba in the field. The model EXPRES (Expert System for Pesticide Regulatory Evaluations and Simulations) (Mutch et al., 1993), developed as a management tool for pesticide regulators and applicators, combines the research models PRZM and LEACHM (v.2) (Wagenet and Hutson, 1987) with a user-friendly interface. The model LEACHM is a research-oriented simulation model of water flow and solute transport, requiring a fairly extensive set of input parameters and variables describing site-specific soil, plant, and climatic conditions
|
MATERIALS AND METHODS |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
|
The lysimeters were buried at the Guelph Turfgrass Institute and Environmental Research Centre (GTIERC; 43°32'50''N, 80°13'50''W), Guelph, ON, in the spring of 1995. The lysimeters were placed level with the surrounding ground surface to prevent unnatural runoff or ponding. The entire site (44 lysimeters in two lines, spaced 
1 m apart) was covered with Kentucky bluegrass sod. The grass was cut with shears to a minimum height of 5 cm, with the clippings left on the surface, as is common practice for turf farms.
Banvel (480 g L-1 of dicamba) was obtained from Sandoz Agro Canada (Kitchener, ON). It was applied to four lysimeters at a rate equivalent to 0.6 kg active ingredient ha-1, the maximum recommended rate for turf (Ontario Herbicide Committee, 1996), in 0.56 L of water over 8 m2 on 5 May, 28 Sept., and 27 Nov. 1997. The Banvel was applied using a push-cart sprayer mounted with a four-nozzle boom. Although sod farmers rarely irrigate, each lysimeter received 10 and 6.7 mm of water respectively on 5 Aug. and 28 Sept. 1997, because the summer was extremely dry. The water flow segment of this study, including field and modeling components, is described in Roy et al. (2000).
Water samples were removed from the soil profile or drain using a vacuum pump into Vacutainer vials (Becton Dickenson, Franklin Lakes, NJ) or an Erlenmeyer flask (rinsed with deionized water between lysimeters), respectively. A small portion of the drain sample was transferred to a Vacutainer vial. Sampling was performed every 4 d or less, usually following a precipitation event. During the summer and early autumn the soil was often too dry to get a sample. Sampling was halted late in the year due to the freezing temperatures. Soil water taken from suction samplers was free of sediment; sediment in the soil water taken from the drain was allowed to settle to the bottom of the vial prior to storage. Soil water samples were stored (a maximum of 7 mo) in the Vacutainer vials at -20°C prior to analysis. Dicamba stability was monitored with dissolved 14C-dicamba (>98% purity) obtained from Sigma Chemical (St. Louis, MO); no loss occurred during storage.
Dicamba Analysis
The concentration of dicamba in samples was measured using a competitive indirect enzyme-linked immunosorbent assay (CI-ELISA) (Hall et al., 1989; Johnson and Hall, 1996). Dicamba of >99% purity, used for standards and in the biodegradation study, was obtained from BASF Chemical Corporation (Mississauga, ON). All reagents used for the CI-ELISA were obtained from Sigma Chemical. A brief description of the analysis follows.
Immulon #4 flat-bottomed microtiter plates (Fisher Scientific, Nepean, ON, Canada) were coated (100 µL well-1) with enzyme conjugate (dicamba covalently linked to ovalbumin and diluted to 1:1000 [w/v] in phosphate-buffered saline [PBS]) and incubated at 4°C overnight. Unbound conjugate was removed by washing the plates three times with PBS containing 0.05% (v/v) Tween 20 (a surfactant; Sigma–Aldrich Canada, Oakville, ON, Canada). Unoccupied well sites were blocked with PBS containing 0.1% (w/v) gelatin (200 µL well-1). After a 20-min incubation, the plates were washed as described previously. Equal sample or standard volumes of free dicamba were combined with anti-dicamba antibody sera (diluted to 1:50 [w/v] in PBS) and incubated in glass tubes for 1 h at room temperature. The mixture was added to the plates (100 µL well-1) and incubated at room temperature for 1 h. At this stage dicamba from the sample or standard is competing for the antibodies in solution with the dicamba bound via ovalbumin to the plate surface (i.e., the more dicamba in the sample, the fewer antibodies being bound to the plate). Unbound dicamba and antisera were removed by washing. Goat anti-rabbit horse radish peroxidase (diluted to 1:5000 [w/v] in PBS), a secondary antibody linked with a color indicator, was added (100 µL well-1) to react with antibody–dicamba–ovalbumin bound to the plate. Following a 20-min incubation, unbound antibody was removed by washing. Substrate (approximately 1 mg mL-1 ABTS [2,2'-azino-bis (3-ethylbenzthiazoline 6-solfonic acid) diammonium], and 1 mg mL-1 urea hydrogen peroxide in citrate buffer [pH 5.0]) was then added (100 µL well-1) to each well to induce the color indicator. Light absorbance at 405 nm was measured using a UV-Vis plate reader after 1 h. Color intensity was inversely proportional to the concentration of dicamba in samples and standards.
Absorbance values (A) of the standards and the samples were normalized by dividing them by the absorbance values of the negative controls (wells containing sample water with no dicamba, Ao). The A/Ao values for standards were plotted against the log of dicamba concentrations to construct a standard curve. Concentrations of the samples in water were determined by interpolating from the standard curve. The linear range of the standard curve was from 0.2 to 1.0 mg L-1. There was not enough sample volume to concentrate those samples below 0.2 mg L-1, but those with measured concentrations above the linear range were diluted with deionized water and analyzed again. However, samples with concentrations just below 0.2 mg L-1 were plotted to indicate detection below the quantifiable limit. Although the values fell outside the linear range of the standards, there was an apparent difference compared with the control blanks and nondetect samples.
Biodegradation Study
Experimental Design
The degradation experiment involved eight treatments, including two materials (soil and thatch), two temperatures (4 and 20°C), and two volumetric water contents (approximately 0.28 and 0.21 m3 m-3 for thatch, and 0.25 and 0.18 m3 m-3 for soil). Soil used in the biodegradation study was the A horizon of the Lisbon sandy loam, as was used in the lysimeters. The thatch was collected from Kentucky bluegrass at the Guelph Turfgrass Institute (GTI), University of Guelph, Guelph, ON. It was 2.0 to 2.5 cm thick, with a 5.8% organic carbon content and a pH of 7.2. The thatch may have been exposed to low applications of dicamba before the sod was brought to GTI, more than two and a half years prior to collection for these experiments. Enhanced degradation from prior exposure is unlikely in this case, as it has not been reported for dicamba in the literature. Also, enhanced degradation observed with other pesticides is not long lasting without continued applications (Racke and Coats, 1987; Harvey, 1990; Roeth et al., 1990; Skipper, 1990).
Soil and thatch were collected about a week before the start of the experiment and stored at 4°C prior to sample preparation. The soil was passed through a 2-mm sieve. The thatch was trimmed to remove all green shoots and large, protruding roots. All remaining materials that would come in contact with the herbicide, thatch, or soil were either flame-sterilized or autoclaved (121°C for 55 min) before use to ensure aseptic initial conditions.
The experimental design was modified from Smith (1974). Either 10.0 g of soil or 5.0 g of thatch was added to 20-mL glass vials and each was packed down to an equivalent volume (9 cm3). Controls were sterilized by autoclaving (121°C for 55 min) a total of three times, spaced 2 d apart. A 500 mg L-1 solution was prepared with 250 mg dicamba dissolved in 500 mL deionized water in a volumetric flask. All other dicamba solutions were made by appropriate serial dilutions of this stock solution. To each sample was added 29.5 µg of dicamba, dissolved in either 1.30 or 0.65 mL of deionized–distilled water (two water contents), at an application rate of 0.6 kg ha-1. The herbicide solution was dripped onto the surface of the soil–thatch using a pipettor, and the final mass was determined. The vials were stoppered with glass wool to allow air circulation and then incubated in a sealed black plastic bag at either 4 or 20°C. Wet, sterile sponges were put in the bags to reduce evaporative water loss from the vials. Deionized water was added to the vials periodically to maintain their final mass and therefore, the desired water content conditions. It was found that water evaporated from the samples and condensed in the glass wool at 4°C. Therefore, these samples received more frequent additions of water. Triplicate samples and a control from each treatment set were removed 2, 9, 20, 40, 61, and 100 d after application, capped, and stored in a freezer at -20°C until analysis. Dicamba stability was monitored with dissolved 14C-dicamba; no loss during storage was observed.
All the samples for each sampling date were analyzed simultaneously. In addition, new soil and thatch samples were prepared at the time of analysis as described above and fortified with a known amount of dicamba—3.0, 14.8, and 29.5 µg—representing 0.1, 0.5, and 1.0 of the fraction of dicamba initially applied. These samples, hereafter known as spikes, underwent the same extraction and analysis procedure as the samples from the degradation experiment. The spikes were used to measure recovery and analytical efficiency by comparing the added and measured mass, and served as standards to quantify dicamba levels in the samples. In all cases a control (no dicamba) was also included.
Dicamba Analysis: Gas Chromatography–Mass Spectrometry (GC–MS)
The GC–MS analysis, including extraction, was modified from Clegg (1987) to accommodate the large number of samples and to reduce the volume of solvents used. All solvents used were pesticide grade and were distilled in glass. Each soil sample and each thatch sample received 10.0 and 13.0 mL of 0.1 M NaOH, respectively. The vials containing these samples were capped, shaken manually for 30 s, and allowed to sit overnight. The following day, 1.0 mL of the resultant solution was removed and placed in a scintillation vial. The sample was neutralized with 1.0 mL of 0.1 M HCl, and acidified with two or three drops of 6.0 M HCl, to a pH of 2 or less. Dicamba was partitioned into ethyl ether to prepare for GC–MS analysis by adding 5.0 and 10.0 mL of ethyl ether to soils and thatch, respectively. The vials were capped, sealed with Parafilm (Fisher Scientific), and shaken upright overnight on a rotary shaker.
To 5-mL glass test tubes we added 100 µL of 2,2,4-trimethylpentane (TMP), 20 µL of 10 mg L-1 2,4,5-trichlorophenoxy propionic acid (TP, internal standard), and 1.0 mL of the ethyl ether extracted sample. The solution was mixed for 5 s, and blown down to near dryness (approximately 50 µL) under air. The evaporated extracts were methylated with 1 mL of diazomethane in ethyl ether, and allowed to react for 30 min. Excess diazomethane and ether were evaporated under air to a final volume of 25 µL. 1-methyl-3-nitro-1-nitrosoguanidine (97% purity), used for the preparation of diazomethane, was obtained from Aldrich Chemical (Milwaukee, WI). The methylated extracts were reconstituted with 1.0 mL of TMP, transferred to 1.5-mL glass autosampler vials, and capped.
A Saturn 4D GC–MS–MS (Varian Canada, Mississauga, ON, Canada) was used for mass spectrometric detection of dicamba. A 1.0-µL aliquot of sample was injected onto a DB-1 capillary column (30 m, 0.25 mm i.d. x 0.25 µm thickness; Agilent Technologies, Mississauga, ON, Canada), with helium carrier gas pressure of 82.8 kPa (12 psi) and a 1.0 mL min-1 flow rate. The inlet temperature was 90°C and the detector temperature was 280°C. A temperature gradient was used, starting at 90°C for 1 min, increasing to 150°C at 20°C min-1, and then increasing to 280°C at 5°C min-1. The final temperature was held for 5 min. Residues of dicamba and 2,4,5-trichlorophenoxy propionic acid (TP, internal standard) were determined by reconstruction of selected ions 203 and 196, respectively. The analytical results for dicamba in spikes and samples were standardized based on a single value of the internal standard prior to quantification.
Dicamba Analysis: High Pressure Liquid Chromatography (HPLC)
The analysis of dicamba by gas chromatography, unlike liquid chromatography, does not discriminate between the parent compound and the primary metabolite in soil, 3,6-DCSA. The methylation step used in gas chromatography transforms both species into the same compound, dicamba-methyl ester. However, 3,6-DCSA adsorbs more strongly to soil particles and, as a result, is more difficult to extract. To quantify the amount of 3,6-DCSA contributing to the concentration measured by GC–MS, a small number of samples underwent a comparative GC–MS and HPLC analysis, as outlined below.
A 1.0-mL aliquot of solution from the base-extracted samples was removed and added to between 4 and 5 mL of HCl-acidified water, to a final pH less than 2. The solution was filtered through a 0.45-µm nylon membrane filter using a 5-mL syringe. Often a second filter was required for the thatch samples because of clogging. Following the procedure of Arjmand et al. (1988), the sample was first extracted using C18 solid phase extraction columns (Sigma–Aldrich Canada). The columns were positioned on a vacuum manifold and conditioned by passing 5 mL of methanol, followed by 5 mL of acidified water (pH < 2), through each column, taking care that the columns never completely dried. The acidified sample was subsequently passed through the column, and the C18 cartridges were dried for 15 min by drawing air through them using the vacuum manifold. Adsorbed compounds were eluted from each column with 4.0 mL of methanol and split into two equal portions, one to be analyzed by HPLC and the other by GC–MS. The sample for GC–MS was blown down under air to approximately 50 µL and then methylated and analyzed as described previously. The sample for HPLC was blown down under air to 250 µL, added to a 300-µL microvial insert within a 1.5-mL glass vial, and capped.
A PerkinElmer (Norwalk, CT) liquid chromatography (PE LC) system was used that included a PE LC-95 UV/Visible spectrophotometer detector, a Series 200 PE pump, a Millipore (Bedford, MA) radial-pak cartridge, and a Nova-Pak C18 column (4-µm particle size, 8.0-mm i.d.; Waters Limited, Mississauga, ON, Canada). The detector was set at a wavelength of 210 nm. The mobile phase consisted of methanol and water at 30:70 (v/v) for A and 52:48 (v/v) for B, with the ion-pairing reagent tetrabutylammonium dihydrogen phosphate (98% purity, Sigma Chemical), and was maintained at a concentration of 0.005 M. A gradient solvent program was used. After injection, the solvent balance was held at 50% of A and 50% of B, for 2 min. The balance was changed along a gradient, up to 100% of B through 10 min, and remained at this level for another 20 min. The flow rate was 1.2 mL min-1, and the injection volume was 20 µL.
Degradation Kinetics
The fraction of dicamba remaining in soil and thatch was plotted as a function of sampling time (d). The resulting degradation curves were interpreted and compared by fitting the experimental data with the first-order degradation equation:
 | [1] |
 | [2] |
 | [3] |
Computer Modeling
The computer model EXPRES was used to predict the leaching of dicamba through soil growing turfgrass. Comparisons were then made to measurements made in the lysimeters to determine whether our description of the system and the transport processes were truly representative of the field conditions. In addition, a modeling exercise was performed to test whether using a different degradation rate for the thatch layer would influence dicamba leaching. The degradation rates used were derived from the biodegradation study described above. For one simulation the entire profile was allocated the rate for the soil, while for the other simulation, the 5-cm thatch layer was allocated its own rate. The results could clarify the role of enhanced biodegradation in reducing pesticide leaching below turfgrass compared with bare soil.
The EXPRES model contains the model LEACHM (v. 2) (Wagenet and Hutson, 1987) for simulation of water flow and solute transport. The subroutine for solute (pesticides and conservative tracers) transport, SOLP, solves the one-dimensional convection–dispersion equation (CDE):
 | [4] |
is volumetric water content, 
b is the dry soil bulk density, 
s is a solute source–sink term, z is depth, and t is time. This equation does not account for flow through macropores or other forms of preferential flow. The model EXPRES includes additional routines as well: one to simulate snowmelt based on the mean daily temperature and another for surface runoff and losses of water and pesticide due to erosion. A list of parameters with values required by the model for water flow and dicamba transport are given in Tables 1 and 2. See Wagenet and Hutson (1987) and Mutch et al. (1993) for more details.
|
|
The Arrhenius equation has been used extensively (Walker and Zimdahl, 1981; Parker and Doxtader, 1983; Veeh et al., 1996) to describe the relationship between the degradation rate constant and temperature:
 | [5] |
|
|
RESULTS AND DISCUSSION |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
|
The downward movement of dicamba applied in the spring was limited by the drying of the soil. Further leaching of dicamba in the summer and early autumn was prevented by the low soil water contents (Roy et al., 2000). Degradation probably eliminated most of the dicamba remaining in the soil during these long periods of minimal water infiltration.
The lack of a peak in dicamba concentration at the 10-cm depth, in conjunction with the two large concentrations at the 17- and 29-cm depths detected in two of the lysimeters (Fig. 2) following the September application, may be an indication of preferential flow. Leaching dicamba may have been missed, as only a few water samples were obtained during October because the soil was very dry. Dicamba that remained in the soil matrix near the surface may have been volatilized, as the days following application were windy.
Observed leaching patterns following the late autumn application differed from the earlier applications. Dicamba concentrations remained high throughout the A horizon (Fig. 2). The dicamba concentrations measured in the B horizon soon after the November application (Fig. 2) may have come from the earlier applications or may indicate very fast transport of dicamba applied in November. In either case, high concentrations of dicamba were still detected in the soil profile on 6 Jan. 1998 (Fig. 2). Since degradation generally decreases with depth and temperature (Veeh et al., 1996), it is likely that these high concentrations would not have declined substantially with continued infiltration, suggesting that dicamba could reach the water table under these conditions.
Biodegradation Study
Soil Versus Thatch
The half-lives of dicamba in thatch were 5.9 to 8.4 times shorter than those for soil at the same temperature and moisture conditions (Table 3). Furthermore, the half-lives were shorter than those calculated by other researchers (Smith, 1984; Comfort et al., 1992) for soils under similar conditions (original vegetative cover not reported). These results indicate that the increased pesticide degradation under turf compared with agricultural or bare soils observed by several researchers (Branham et al., 1993; Gold and Groffman, 1993; Horst et al., 1996; Starrett et al., 1996) is related to higher microbial activity.
The decline in the fraction of dicamba remaining in the thatch was rapid (Fig. 3) for all temperature and moisture conditions. In comparison, for a number of the soil treatments, there appears to be an initial acclimation period or slow phase (Fig. 4), where the fraction of dicamba remaining persists at approximately 1.0. This period appears to last at least 40 d for the 4°C treatment and about 10 d for the 20°C treatment and is probably a major cause of the poorer fit (larger RMSE values) between the measured values and the calculated curve (Table 3). Although the first-order kinetics equation (Eq. [1]) does not account for changes in microbial numbers or environmental conditions and, therefore, cannot reproduce an acclimation period, it is more easily incorporated into computer models than more encompassing degradation equations.
|
|
Temperature and Water Content
An increase in temperature from 4 to 20°C resulted in a decrease of the half-life of dicamba ranging from 2.0 to 3.7 times (Table 3), with the greatest difference associated with the low water content (more stressful condition). Numerous research groups (Walker and Zimdahl, 1981; Parker and Doxtader, 1983; Veeh et al., 1996) have reported an increase in the degradation rate of various pesticides with increasing temperature, up to an optimum temperature for biodegradation, before rates decline. In a batch study on dicamba degradation in clay, Comfort et al. (1992) measured half-lives of 23.5 d at 28°C, 38 d at 20°C, and 151 d at 12°C. These values compare well with the half-lives for dicamba in soil measured in this study (Table 3).
The natural log of the degradation rate was plotted against inverse temperature (K) to determine the Arrhenius activation energies, EA. The calculated EA values were 55.0 and 29.8 kJ mol-1 for thatch with low and high water content, respectively. Similar values were obtained for soil: 54.0 kJ mol-1 for the low water content and 39.4 kJ mol-1 for the high water content, which could imply a similarity between the microbial community in each material. The derived values were similar to those reported previously for other pesticides (Walker and Zimdahl, 1981; Parker and Doxtader, 1983; Veeh et al., 1996).
Increasing water content had a less dramatic effect on the rate of degradation than temperature, probably because the difference between water contents was minor. At 4°C, the degradation rate ranged from 1.5 to 2.1 times faster for soil and thatch with higher water content (Table 3). The rates for the 20°C treatments were similar for both the high and low water contents. Slower degradation rates at lower water contents have been observed in previous batch studies (Burnside and Lavy, 1966; Walker and Zimdahl, 1981; Parker and Doxtader, 1983).
Batch Evaluation
Considering the presence of an acclimation period in the soils, the RMSE's for the best-fit of Eq. [1] to the data are fairly low (Table 3). Variability between triplicate samples (Fig. 3 and 4) may be an artifact of the analytical procedure or may reflect the complexity of soil and microbial functions. Concerning the efficiency of the analytical procedure (Table 4), the recovery of dicamba from the soil and thatch samples fortified with herbicide was acceptable, though the recovery from thatch was more variable. The GC–MS derived concentrations were not much higher than those derived from the HPLC, which indicates that there was little, if any, 3,6-DCSA extracted in the samples along with dicamba. Since 3,6-DCSA sorbs more strongly to soil (Smith, 1974), it is likely that the solvent used was not strong enough to extract it. The percent of dicamba in the sterilized controls remained fairly constant, indicating minimal volatilization or chemical degradation losses. The slight decline at later sampling dates (Table 4), which was especially noticeable in the thatch, may be due to incomplete sterilization. Minor turbidity was observed in thatch samples after 21 d during the sterilization test (Trevors, 1996). The decline could also indicate immobilization of some dicamba by organic matter.
|
Larger concentrations of dicamba than predicted were observed at the 10-cm depth after spring application (Fig. 2). It appears that the dicamba moved through the turf layer very quickly and without much spreading. There may also have been less degradation than predicted, since the soil was cooler than the 10°C average used for modeling, with temperatures dropping to freezing some nights the week following application. In addition, the model does not simulate an acclimation period, which probably occurred following the first dicamba application.
Dicamba applied in late autumn appears to have reached the bottom of the A horizon (31 cm) to the middle of the B horizon (43 cm) by early January (Fig. 2), sooner than predicted. The snowfall–snowmelt routine may have delayed the predicted infiltration (Roy et al., 2000) and, therefore, solute transport. The observed concentrations were also greater than predicted. The cold, fluctuating temperatures in December probably impaired microbial activity, reducing degradation losses below what was predicted.
Observations and predictions of dicamba leaching out the bottom of the lysimeters were in agreement, with no leaching losses in either case by the end of the year.
EXPRES Modeling Exercise
Having simulated the leaching of dicamba fairly well, EXPRES was used to test the effect of the higher degradation rate in thatch compared with soil on dicamba leaching. Ignoring the increased degradation within the thatch layer (dashed line) should result in greater dicamba leaching from turfgrass, as illustrated in Fig. 5. Even though the increased degradation only occurs in the top 5 cm of the profile, the effect is still apparent at the 54-cm depth (Fig. 5B). The concentrations are more similar to one another at the 10-cm depth following the late autumn application. It is likely that the water and herbicide were moving quickly through the turf layer at this time, so the enhanced degradation of dicamba had a limited effect on the amount of dicamba leached.
|
|
CONCLUSIONS |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
The more rapid degradation of dicamba in thatch, measured in the batch experiment, may explain why there is less leaching of dicamba from turfgrass than agricultural soils. Higher temperature resulted in increased degradation in both thatch and soil, while water content had less of an effect on the degradation rate. The wide range of half-lives measured in this study illustrates the influence of soil properties and environmental conditions on dicamba biodegradation. The use of small-scale laboratory studies can still provide effective, cost-efficient means of evaluating pesticide behavior while incorporating a number of different environmental conditions into the experimental design.
A modeling exercise with EXPRES indicated that increased degradation in the thatch layer can significantly restrict dicamba leaching, although there appears to be less of an effect when infiltration is rapid. Therefore, it is recommended that degradation rates be measured using local climatic and soil–crop conditions rather than using values obtained from the literature if accurate predictions of herbicide fate are required. The use of models that incorporate the effects of temperature and/or water content would also be beneficial.
Based on samples collected with suction samplers, EXPRES simulated dicamba transport fairly well. There were some signs of preferential flow in the field measurements following the September application, and faster transport of dicamba below the A horizon following the November application. Observed concentrations were higher than predicted, especially at the 10-cm depth, which is probably due to preferential flow through the turf and/or a slower degradation rate than predicted. Differences between observed and predicted values may be due to inaccuracies of the water flow simulation and the inability of the model to modify degradation rates with changing climatic conditions.
|
ACKNOWLEDGMENTS |
|---|
|
NOTES |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
|
REFERENCES |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  L. W. Galagedara, J. D. Redman, G. W. Parkin, A. P. Annan, and A. L. Endres Numerical Modeling of GPR to Determine the Direct Ground Wave Sampling Depth Vadose Zone J., November 11, 2005; 4(4): 1096 - 1106. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
