| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
a LFR Levine·Fricke, 1900 Powell Street, 12th Floor, Emeryville, CA 94608
b Syngenta Crop Protection, P.O. Box 18300, Greensboro, NC 27419
c Monsanto Company, 800 North Lindbergh Boulevard, St. Louis, MO 63167
d Dow AgroSciences LLC, 9330 Zionsville Road, Indianapolis, IN 46268
e Klein & Associates, LLC, 12917 Topping Estates Drive, Town & Country, MO 63131
f Syngenta Ltd., Jealott's Hill Research Station, Bracknell, Berkshire, England RG42 6ET
g LFR Levine·Fricke, 630 Tollgate Road, Suite D, Elgin, IL 60123
h LFR Levine·Fricke, 3382 Capital Circle NE, Tallahassee, FL 32308
* Corresponding author (noel.deguzman{at}lfr.com)
Received for publication November 24, 2003.
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
Abbreviations: ARP, Acetochlor Registration Partnership • DTW, depth to water • ESA, ethanesulfonic acid • GWM, Acetochlor Registration Partnership State Ground Water Monitoring Program • LOD, limit of detection • LOQ, limit of quantitation • MOE, margin of exposure • OXA, oxanilic acid
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
The USEPA granted conditional registration for acetochlor on 11 Mar. 1994 (USEPA, 1994), to the Acetochlor Registration Partnership (ARP), consisting originally of Monsanto and Zeneca, but, as of fall 2000, consisting of Monsanto and Dow AgroSciences. Several unique factors were involved in the registration of acetochlor. The ARP expected, and USEPA required, that market forces following the registration of acetochlor would reduce the total mass of corn herbicides applied in the United States by substituting acetochlor formulations for several already registered corn herbicides such as alachlor, metolachlor, atrazine, and 2,4-D, which tend to be applied at higher rates. The USEPA imposed several other restrictions and conditions on the use of acetochlor to minimize any potential risks to human health and the environment, including monitoring programs designed to ensure the continued protection of ground water and surface water resources.
One of the required monitoring programs was the ARP State Ground Water Monitoring Program (GWM). The regulatory purpose of the GWM was to serve as an early warning of potential ground water impact from acetochlor use. The monitoring program was conducted in seven midwestern states (Illinois, Indiana, Iowa, Kansas, Minnesota, Nebraska, and Wisconsin). A total of 175 sites (25 per state) were monitored for five years. However, in 1999, at the request of USEPA, the program was extended for two more years and expanded to include two soil degradates of acetochlor as target analytes: acetochlor ESA and acetochlor OXA (Hackett et al., 2005). The ARP agreed to automatic cancellation of the acetochlor registration if the USEPA determined that "a pattern of movement" of acetochlor to ground water occurred at 20 or more of the 175 GWM sites. A pattern of movement was defined as detections of acetochlor residues at concentrations equal to or greater than 0.10 µg L–1 in three separate samples from the same well within a 7-mo period. The 0.10 µg L–1 level was not based on a specific human health or ecological concern, but rather, it was selected as an early indication that residues were reaching ground water.
The GWM was unique and unprecedented due to its broad scope and magnitude. There was a high level of input from many scientists and regulators, including the USEPA's Environmental Fate and Effects Division (EFED) experts and the Lead Agency from each state, typically the Department of Agriculture or its equivalent.
This publication describes the methods, results, and interpretation of the GWM. The ARP's surface water monitoring program and the field leaching studies are described in companion publications (Hackett et al., 2005; Newcombe et al., 2005).
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
Site Selection Criteria
The goal of the first phase of the study was to establish a network of 175 monitoring sites in regions of high corn production in each of the seven states representing a range of soil textures typical of corn agriculture in those regions. Each site was expected to have shallow ground water, as defined by each state (Table 1), unprotected by restrictive subsurface layers. A new monitoring well was installed within or closely adjacent to and down-gradient of each site.
|
|
|
Site Characterization and Well Installation
Once a GWM site was confirmed to meet the above criteria, it was visited by ARP personnel to collect additional characterization data. A topographical survey, hydrogeological assessment, soil characterization, and a cooperator interview were conducted. If available, published maps of the site and vicinity were obtained, including county roadmaps, plat maps, USGS 7.5-min quadrangles, NRCS County Soils maps, and aerial photos. Furthermore, a detailed map of each site was produced to identify site-specific features, such as access lanes, study plot location, irrigation and other farming equipment, tile drains, ditches, and other waterway features.
Historical pesticide use, dating back to 1990 (when available), cropping, and other agronomic practices were obtained by interviewing the cooperators. A minimum 4.0-ha portion of the farm was designated as the study plot. Ten soil cores (0–0.15 m) were collected from representative locations in the study plot. These soil cores were composited and a subsample was analyzed (A&L Great Lakes Laboratories, Fort Wayne, IN) for pH, organic carbon and organic matter, cation exchange capacity, USDA texture classification, and bulk density.
Monitoring wells were sited within or closely adjacent to, and down-gradient of the study plot. Various sources of published ground water data were used (for example, the Department of Natural Resources Hydrologic Assessment, the USGS Hydrologic Atlas, and local university data) to assess ground water flow direction for most sites. At sites where published ground water data were not available, trained hydrogeologists evaluated topography in conjunction with surface water drainage features to assess ground water flow direction.
Monitoring wells were installed by licensed commercial drilling contractors under the direct supervision of a professional geologist or hydrogeologist and in compliance with state and local guidelines. Each boring was drilled using a hollow-stem auger advanced by a rotary drill rig. Continuous core soil samples were collected from each boring and lithologic descriptions were recorded using the Unified Soil Classification System (USCS). Each monitoring well was constructed with 0.05-m-i.d. polyvinyl chloride (PVC) casing with flush-threaded joints and 0.254-mm machine-slotted screen. A filter pack of coarse sand to fine gravel was placed in the annular space surrounding and up to approximately 0.6 m above the well screen. The length and position of the well screen was defined by each state (Table 1). A minimum 0.9-m bentonite seal was installed in the annular space above the filter pack. The remaining annular space from 0.6 to 0.9 m below ground surface was sealed using a Portland cement grout or a bentonite grout. The PVC casing extended up to approximately 0.9 m above the surface grade and was protected by a 0.1-m-i.d. steel protective casing and a locking cap. Wooden posts were installed in a square formation 0.2 m from the monitoring well for added protection against farm equipment. Figure 1 illustrates the typical well construction details of the GWM wells. After well installation, each monitoring well was thoroughly developed and equipped with a dedicated bladder pump. Each of the 175 monitoring wells was locked and access was limited to ARP personnel.
Each monitoring well was surrounded by an "acetochlor-free" buffer zone to minimize the chance of direct spray drift contamination of the monitoring wellhead and sampling area. Acceptable buffer zones were defined by each state and ranged from 9.1 to 45.7 m (Table 1).
Each of the 175 wells was given a unique ID, which followed a standard SSnn format where SS reflected the state abbreviation and the nn represented a sequential number within the state (e.g., IL01–IL25). Figure 2 shows the approximate location of the sites. Exact locations of the sites and wells were held confidential to minimize the risk of vandalism or sabotage and to protect the privacy of the cooperator.
|
Well Sampling
Ground water samples were collected monthly for the first five years of the study and quarterly for the final two years of the study. Special procedures were followed to minimize the chance of sample contamination and spurious results: for example, (i) efforts were made to prevent sampling equipment from coming in contact with bare ground; (ii) laboratory-grade detergent, isopropyl alcohol, and deionized water were used to decontaminate the sampling equipment before sampling activities at each site; and (iii) new disposable latex gloves were worn and replaced frequently during the sampling procedures. During the spring chemical application period, additional procedures such as covering and decontaminating the wellhead were employed.
Before purging ground water from a well, field staff measured the depth to water (DTW) in the well using a decontaminated electronic water-level indicator, and calculated the volume of water in the well. Ground water was purged using the dedicated pump, and water-quality parameters (temperature, pH, and specific conductance) were measured.
A ground water sample was collected after at least three well volumes were purged from the well and once the water quality parameters were stable (if the last two sets of measurements fell within ±0.10 pH units, ±0.2°C, and ±10% µmhos). If a well demonstrated inadequate recharge for continued purging, a sample was collected once the ground water recovered to 80% of the original water column height or after two hours, whichever occurred first.
A ground water sample from each well was collected directly into each of two 250-mL Nalgene sample containers (Nalge Nunc, Rochester, NY) and labeled with the well ID, protocol number, a pre-assigned sample ID, and a sampling event identifier. The samples were temporarily stored in an ice-chilled cooler until they were shipped (priority overnight courier) with blue ice to the analytical laboratory at Monsanto in St. Louis, MO, within 120 h after sample collection.
Each state, with the exception of Kansas and Minnesota, requested and received split samples from the associated GWM wells. The split samples were submitted to the state-specified laboratory for confirmatory analysis and other work of interest to that particular state. The split sampling frequency as well as supplemental sample handling procedures followed guidelines provided by the states.
During each sampling event, specific observations were recorded, including: (i) tillage and cultivation; (ii) crop identity and growth stage; (iii) well, well area, and buffer zone condition, including observations of weeds or damage; (iv) evidence of chemical application; (v) evidence of ponding and/or runoff; and (vi) any evidence of active herbicide spraying in the area during the spring season.
Analytical Methodology
Ground water samples were analyzed for parent acetochlor, alachlor, atrazine, and metolachlor during all seven years of the GWM. For the final three years, samples were also analyzed for the tertiary amide soil degradates of acetochlor, alachlor, and metolachlor, specifically tertiary amide sulfonic acid (ESA) and tertiary amide oxanilic acid (OXA). For a complete list of the target compounds, including common name, chemical name, and CAS number, see Table 1 in Hackett et al. (2005).
Gas Chromatography–Mass Spectrometry Method for Parent Herbicides
Parent herbicides were analyzed using stable isotope dilution gas chromatography–mass spectrometry (GC–MS), which was preceded by solid phase extraction for cleanup and concentration. The method involved addition of deuterated analogs of each analyte, as surrogates, to the 200-mL sample before extraction, concentration, and analysis (Hackett et al., 2003). Based on actual prior fortification data, the limit of detection (LOD) and limit of quantitation (LOQ) of this method were determined to be 0.03 and 0.05 µg L–1, respectively, for all nonpolar analytes (Hackett et al., 2003), with the exception of alachlor, whose LOD was 0.05 µg L–1 due to higher background levels of this compound.
Liquid Chromatography Tandem Mass Spectrometry Method for Chloroacetanilide Degradates
The ESA and OXA soil degradates of acetochlor, alachlor, and metolachlor were analyzed by direct aqueous injection reversed-phase liquid chromatography tandem mass spectrometry (LC–MS–MS). The samples were injected directly into the LC–MS–MS (HP1100/Sciex API-3000; Sciex, Concord, ON, Canada) without prior concentration, cleanup, or filtration (Hackett et al., 2003). Based on actual prior fortification data, the LOQ of all six polar degradates was determined to be 0.50 µg L–1. The LOD for the three OXA soil degradates was 0.10 µg L–1, and for the three ESA soil degradates, the LOD was 0.20 µg L–1.
Storage Stability
All ground water samples were stored in a refrigerator at 2 to 10°C on receipt at Monsanto, before extraction or preparation for analysis. Replicate samples were transferred to a freezer at –20 ± 5°C. Sample extracts were either analyzed immediately or stored in a refrigerator at 2 to 10°C before analysis. All reported analytes demonstrated acceptable storage stability under these conditions.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
Management of the Monitoring Network
Table 1 shows the soil texture distribution for sites in the GWM network. All state soil texture targets were met with the exception of two sites in Minnesota. In Minnesota, suitable clay loam sites could not be found due to the presence of restrictive layers or excessive DTW. Therefore, two sites planned to be clay loam were replaced with two loam sites, which were considered to be more vulnerable to leaching.
To comply with all state and federal requirements, the ARP ground water monitoring program resulted in a distribution of sites with extremely shallow water tables, much shallower than is typically used for drinking water. Study data showed that 62% of the wells (112 of 182 wells) had an average DTW less than 7.6 m. Of those 112 wells, 60 wells showed an average DTW less than 3 m and 22 wells had an average DTW less than 1.5 m. Figure 3 shows the minimum, maximum, and mean depths to water of the GWM wells. Of the 58 sites (60 wells) with very shallow water tables (average DTW less than 3 m), 74% (43 sites) were tile-drained for the purpose of lowering seasonally high water tables to allow for critical farm operations such as spring planting. A total of 66 sites (37%) in the study were tile drained.
|
As a result of the modifications made to the monitoring network, a total of 182 wells (180 sites) were installed and sampled during the course of the GWM. The majority (158 sites) remained in the program through the 2-yr extension.
Acetochlor Applications
A total of 775 applications of acetochlor were made over the course of the study, representing an average of 4.3 acetochlor applications per site. While the cooperators were only required to maintain a 4.0-ha study plot, in the majority of the cases, the farmer chose to treat the entire field with the acetochlor product. The majority of sites received four or more acetochlor applications, while 20 sites received three applications, and two sites, due to unanticipated cropping plan changes, received two acetochlor applications.
Sampling Success
Overall, 87% of the total scheduled samples were collected during the study, resulting in 10054 ground water samples collected from the 182 wells. Site inaccessibility due to adverse weather conditions (e.g., unable to access the well, frozen pump discharge lines, safety-related weather issues) proved to be the primary reason for sampling failures. During the later part of the program, sampling failures were typically the result of the water table being below the screened interval of the well.
Analytical Results
Of the 10054 samples analyzed for parent acetochlor, alachlor, atrazine, and metolachlor, 1961 were also analyzed for the tertiary amide sulfonic acid (ESA) and oxanilic acid (OXA) soil degradates of acetochlor, alachlor, and metolachlor. There was no significant variation between the results of the split samples analyzed by the states and those analyzed by the ARP.
Most of the wells (157) never provided a sample showing detectable parent acetochlor (LOD 0.03 µg L–1). As shown in Table 3, the 25 wells with detected acetochlor were distributed by state as follows: Kansas (8), Iowa (7), Illinois (4), and Minnesota and Nebraska (3 each). No detections occurred in Indiana or Wisconsin.
|
Both acetochlor ESA and acetochlor OXA were detected in all seven GWM states, and compared with parent acetochlor, a higher percentage of the wells had soil degradate detections. Acetochlor ESA was detected in 81 wells, while acetochlor OXA was detected in 26 wells. Acetochlor ESA concentrations greater than or equal to the LOD (0.2 µg L–1) were found in 24.5% of the samples (479 of 1961). Of these samples, 294 contained concentrations greater than or equal to the LOQ, with a maximum of 20 µg L–1. Acetochlor OXA residues were observed in approximately 6.5% of the samples (128 of 1961). Of the 128 samples that showed acetochlor OXA residues, 70 were below the LOQ (0.5 µg L–1), while the remaining 58 had concentrations greater than or equal to the LOQ, with a maximum of 19.10 µg L–1.
Figure 4 presents the co-occurrence of acetochlor parent and acetochlor soil degradates observed in the 1961 samples collected in the last 3 yr of the study. The figure shows that acetochlor parent residues were only detected in wells when one or both of the corresponding degradates were also observed. Similarly, acetochlor OXA was only detected in wells that also showed residues of acetochlor parent or acetochlor ESA or both.
|
Similar to the acetochlor soil degradates, alachlor soil degradates were detected at a higher frequency than their parent. Approximately 32.5% (639 of 1961) of samples were found to have detectable alachlor ESA residues. These residues were detected at 76 wells, which were distributed across all seven states. Alachlor OXA was found in 5% of samples (98 of 1961). These alachlor OXA detections came from 16 wells; however, alachlor OXA was not detected in Indiana or Minnesota. Figure 5 shows the co-occurrence of alachlor parent and alachlor soil degradate residues.
|
Metolachlor soil degradates were detected at a higher frequency than metolachlor parent. In fact, metolachlor ESA had the highest detection frequency of all the GWM analytes with metolachlor ESA residues detected in 58% (106) of the 182 wells and in 50% (976 of 1961) of the samples. Detectable metolachlor ESA residues were found in all seven states. Of the metolachlor ESA detections, 25% (248 of 976) were less than the LOQ of 0.5 µg L–1. Metolachlor OXA was detected in 21% of the samples (407 of 1961) at 63 wells across the seven states. Of the 407 samples with metolachlor OXA detections, 66% were less than the LOQ.
Figure 6 presents the co-occurrence of metolachlor parent and its degradates and shows that metolachlor was only detected in wells that also showed detectable concentrations of the two corresponding soil degradates. With one exception, metolachlor OXA was only detected in wells that also showed residues of metolachlor parent or metolachlor ESA or both.
|
Atrazine residues were detected in all seven GWM states and in 55% of wells (101 of 182). The state with the highest number of atrazine detections was Kansas followed, in descending order, by Wisconsin, Nebraska, Iowa, Minnesota, Illinois, and Indiana.
Figure 7 shows that 60% of the wells (110 of 182) had residues of at least one of the four parent analytes. Four wells contained all four herbicides. Furthermore, atrazine was detected in all 25 wells that contained acetochlor parent.
|
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
Monitoring Network Characteristics
While the ARP program was intended to represent the distribution of soils across high-intensity corn counties, several factors weighted the range of selected soil textures toward lighter, sandier, soils, which are more prone to leaching. First, while some states (e.g., Iowa and Illinois) have three to eight times the area in corn production than other states (e.g., Kansas and Wisconsin), the program required 25 sites to be established in each state. As a result, there were a disproportionately high number of sandy loam sites, such as those found in Wisconsin, in the program, and a correspondingly low number of clay loam sites. Second, the selection process assumed that corn was planted with equal likelihood on all soils within the counties of interest, whereas, in the absence of irrigation, corn was less frequently grown on the lighter soils.
Another unique aspect of the GWM is the setting in which the monitoring wells were installed. Unlike other programs, the GWM generally monitored the shallowest ground water aquifers, sometimes sampling the very top of the saturated zone. Such ground water rarely serves as a drinking water source. Furthermore, every GWM well was installed in an agricultural setting, within or immediately adjacent to the study plot (i.e., within 9.1–45.7 m of the chemical use area), whereas other programs typically used existing potable rural wells, which are usually installed in deeper, higher-volume aquifers, and tend to be more remote from the treated fields.
Comparison of Analytical Results
General trends in the GWM were similar to those from comparable studies. For instance, atrazine was detected at a higher frequency than other parent herbicides, polar soil degradates were detected at a higher frequency than their corresponding parent compounds, and the ESA soil degradate was detected at a higher frequency than the corresponding OXA soil degradate.
Table 4 contains a summary of atrazine detections from a range of ground water studies involving several different categories of wells, many of which were sampled because of their potential vulnerability or a history of ground water detections (P. Hendley, P. Hertl, and S. Chen, personal communication, 2003). Similar to the other studies, the GWM shows that the frequency of atrazine detections greater than its maximum contaminant level (MCL) of 3 µg L–1 is very low. More importantly, the GWM results showed a higher detection frequency than all other studies for all categories of atrazine residues, which confirms the early warning nature of the GWM.
|
Acetochlor Detections
Leaching Model Predictions
Simulation models such as the Pesticide Root Zone Model (PRZM) have been used to predict potential environmental occurrence of pesticides with varying success (Jones and Russell, 2001). During the early phase of the GWM, PRZM was used as a conservative, prospective predictor of acetochlor leaching potential at each of the 175 monitoring sites. While a range of scenarios was simulated, the scenario corresponding to a conservative set of parameters (the base case) is described below. The simulations used actual site characterization and well log data such as ground water depth, surface soil properties, and 31 yr of historical local weather data. Conservative inputs were used for acetochlor adsorption and degradation properties in the topsoil and especially in the subsoil. Finally, acetochlor applications to corn were assumed every year at a rate of 2.8 kg ha–1 (i.e., no crop rotation).
The modeling predicted that 17 sites would show acetochlor detections of at least 0.10 µg L–1 before the end of seven years of continuous use. Generally, the characteristics of these 17 sites included a combination of shallow ground water (15 of these 17 sites had DTW < 3 m) and lighter soil textures. Eight of the detections were predicted to occur in Indiana.
In comparison, study results showed that 14 wells (vs. the predicted 17) met or exceeded 0.10 µg L–1 acetochlor in at least one sampling. While there was not a significant discrepancy in the number of sites, none of the 14 wells that had such detections corresponded with the 17 wells identified by modeling. The 14 wells with actual detections were not prone to classical leaching based on currently accepted measures of vulnerability. In fact, only one of these wells was installed in a coarse-grained soil (sandy loam), while 10 wells were located at sites with a heavy surface soil texture (clay loam, silty clay loam, or silt loam). Furthermore, the average DTW for the 14 sites ranged from 1.3 to 15.6 m, with a median of approximately 5 m. These discrepancies suggest that the PRZM model in association with highly conservative parameterization predicted more leaching than actually occurred for sandier soil textures. The modeling predicted residues would not occur before the third year of the program and thus failed to accurately predict observations at sites such as IA07 and IL24, which generated detections within the first two to three months of monitoring. This is not surprising given that subsequent site-specific investigations revealed that these sites, like others among the 14 with detections, were subject to direct interactions between ponded runoff water and the well. PRZM is unable to cope with either this process or preferential flow, the latter of which has been implicated in the only PGW (Prospective Ground Water Monitoring study) where parent acetochlor was detected, albeit only at trace (<0.10 µg L–1) levels, in shallow ground water (Newcombe et al., 2005).
In the case of Indiana, there was a large discrepancy between the number of 0.10 µg L–1 detections predicted by the model (eight sites) and the number of actual detections > 0.10 µg L–1 (none); however, this is at least partly related to the State of Indiana guideline for screening wells at 7.6 to 10.7 m when the depth to water is less than 7.6 m (Table 2).
A model such as PRZM is often used to perform conservative regional-scale assessments of the impact of pesticide leaching on ground water quality, especially where the endpoint is a comparison of similar, potentially vulnerable sites. However, comparison of the PRZM model predictions with actual GWM results raises several questions regarding the relevance of the PRZM model algorithm for leaching in heavy and highly structured soils, as well as the applicability of current USEPA guidance on model parameterization. We suggest that somewhat better results might now be obtained given actual weather data, careful selection of input parameters, and better information on subsoil behavior, but PRZM's inability to deal with preferential flow and other phenomena associated with finer-textured soils would continue to hamper site-by-site comparisons. Nevertheless, for coarse-textured soils recent findings from the FIFRA Environmental Model Validation Task Force (Jones and Russell, 2001) indicate that given significant amounts of site-specific calibration, relatively accurate predictions of pesticide leaching can be made with PRZM.
Investigation of Acetochlor Detections
Several supplemental investigations were conducted to provide a better understanding of the observed acetochlor detections. Many of the findings indicated that surface water features (roadside ditches, runoff channels, intermittent ponding) in close proximity to the well may have led to some acetochlor detections, especially where the ground water at the site was very shallow. For instance, at two sites (IA07 and IL24), a second well was installed at a location further away from the area of potential surface water–ground water interaction and down-gradient of the treated area. While acetochlor residues continued to be observed shortly after each application in the original wells, the second wells remained free of acetochlor residues throughout the program.
Correlation Analyses
The data generated in this study were analyzed using univariate and multivariate techniques, including a principal component analysis (PCA) and partial least squares (PLS) to assess which factors, if any, could be identified as contributing to the occurrence of pesticide residues in ground water. The input parameters included soil texture, irrigation method, tillage method, well screen position, and average DTW. Output parameters included the count of detections by analyte, and the average and maximum concentration of acetochlor, atrazine, and soil degradate detections at each well.
No significant correlations between analyte detections and the predictor variables were observed, with the exception of the irrigation method. Overall, the flood irrigation method (vs. center pivot, travelling gun, and no irrigation) was identified as a potentially significant factor for the sites that showed atrazine detections. These results are generally in line with the factors identified as important considerations relative to predicting leaching potential (Kolpin, 1997). This suggests that the prevalent use of flood irrigation in some states may increase the potential for ground water impact by pesticides, and points to the importance of irrigation management (Spalding et al., 2003).
Human Health Risk Assessment
The monitoring results were used to conduct a conservative human health risk assessment. The methodology was analogous to that used previously by the USEPA for alachlor, a closely related structural analog (USEPA, 1998). The methodology is also identical to that used in the companion paper on the acetochlor surface water monitoring program (Hackett et al., 2005), which contains more details on the underlying assumptions. To produce a risk assessment approximating the 99th percentile of exposure in drinking water wells, the 95th percentile concentrations from the GWM database were used. The conservative nature of this approach is supported by the fact that 95th percentile concentrations in the GWM for alachlor, atrazine, and metolachlor were all higher than the 99th percentile concentrations observed in rural potable wells in NAWWS (Holden et al., 1992).
The two soil degradates were included in the exposure concentrations used for acute and chronic noncancer risk assessment by assuming that both acetochlor ESA and acetochlor OXA possess equivalent toxicity to parent. However, this is a very conservative assumption because both degradates appear to be less toxic than parent to all species tested (Lamb and Clapp, 1999). The acetochlor ESA and acetochlor OXA soil degradates were not included in the total acetochlor residue used for cancer risk assessment, as they exhibit similar toxicological characteristics to alachlor ESA (unpublished data), which was judged by the USEPA as unlikely to be carcinogenic (USEPA, 1998). The USEPA has since determined that acetochlor and alachlor share a common mechanism of toxicity (USEPA, 2001), thus the cumulative risk assessment takes the potential lifetime exposure to both compounds into account.
For acute and chronic noncancer risk assessment, the total acetochlor residue was determined as the sum of parent, acetochlor ESA, and acetochlor OXA. The summed residue was multiplied by a daily water consumption figure and expressed as a percentage of the acceptable daily dose. As shown in Table 5, the exposure to acetochlor residues via drinking water represented only 1.22% of the acceptable chronic daily intake, and only 0.01% of the acceptable acute intake, when using the appropriate time-averaged concentrations.
|
|
|
CONCLUSIONS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
The results show that parent acetochlor is rarely found in ground water, with detections in less than 1.5% of the samples. Among parent herbicides, atrazine was detected most frequently (36% of the samples and 55% of the wells). The degradates of acetochlor, alachlor, and metolachlor were more frequently detected than their respective parent herbicides. These results were broadly consistent with the results of other large-scale studies, such as NAWQA, NAWWS, and NPS, and confirmed the early warning nature of the GWM's shallow well network relative to the previous studies of mainly deeper aquifers.
The study results call into question the use of PRZM modeling to predict pesticide leaching patterns, especially in fine-textured soils. Discrepancies between model predictions and actual study results point to common modeling issues such as preferential flow and subsoil degradation, and they suggest areas where modeling could be improved. Supplemental investigations showed that irrigation method and the interaction between surface and ground water are important factors to consider when studying pesticide transport to shallow ground water.
A conservative human health risk assessment was conducted using the study results. Potential acute and chronic noncancer risks were evaluated using data for acetochlor parent, acetochlor ESA, and acetochlor OXA. Potential cancer risk was determined using data for parent acetochlor and alachlor in a cumulative risk assessment. The results show that acetochlor residues in shallow ground water do not pose a significant risk to human health.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
Related articles in JEQ:
This article has been cited by other articles:
|
|
 |
|
  A. C. Newcombe, D. I. Gustafson, J. D. Fuhrman, I. J. van Wesenbeeck, N. D. Simmons, A. J. Klein, K. Z. Travis, and K. J. Harradine The Acetochlor Registration Partnership: Prospective Ground Water Monitoring Program J. Environ. Qual., May 11, 2005; 34(3): 1004 - 1015. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. G. Hackett, D. I. Gustafson, S. J. Moran, P. Hendley, I. van Wesenbeeck, N. D. Simmons, A. J. Klein, J. M. Kronenberg, J. D. Fuhrman, J. L. Honegger, et al. The Acetochlor Registration Partnership Surface Water Monitoring Program for Four Corn Herbicides J. Environ. Qual., April 20, 2005; 34(3): 877 - 889. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
