For Presentation at the Air & Waste Management Association's 90th Annual Meeting & Exhibition, June 8-13, 1997, Toronto, Ontario, Canada

97-RA122.03

Determination of Atmospheric Atrazine and Alachlor Concentrations By Reverse Phase Liquid Chromatography and Gas Chromatography Mass Spectrometry

Catherine E. Roach and Larry G. Anderson

Center for Environmental Sciences, Campus Box 136, P.O. Box 173364, University of Colorado at Denver, Denver, Colorado 80217­3364

William T. Foreman

U.S. Geological Survey, National Water Quality Laboratory, 5293 Ward Road, Arvada, Colorado 80002

Abstract

Atrazine was detected at 0.036­ to 0.11­mg/g concentrations in leaves collected from trees exhibiting a variety of abnormal phenotypic symptoms at a tall grass prairie restoration site near Baldwin, Iowa. Atmospheric­related transport mechanisms were suspected as the primary routes for herbicide exposure to the trees and include, for example, off­site drift during and after herbicide application and drifting ground fogs that develop or pass over surrounding corn and soybean fields. Interest in atmospheric­related exposure processes led to a separate study to measure atmospheric concentrations of atrazine and alachlor as these surface­applied herbicides were lost from a corn field at a research farm in southeastern Iowa. Low­volume air samples were collected 0.3 m above the field surface using Florisil cartridges, which were eluted with methanol and analyzed by HPLC and by GC/MS for select samples. Twenty­four­hour average air sampling every other day beginning the day of herbicide application found 1130 ng/m3 atrazine and 830 ng/m3 alachlor. Alachlor dropped to 17 ng/m3 and atrazine to non­detectable by the 6th day following application. Rain at 8 days after application resulted in an increase of both atmospheric atrazine and alachlor concentrations to 110 and 160 ng/m3, respectively. The concentrations again decreased to non­detectable by the 12th day, and remained so for the rest of the 62­day monitoring period. The apparent volatilization "peak" following the rain event is comparable to that observed in a number of volatilization studies and further illustrates the importance of soil moisture content on the volatilization process. This paper summarizes the tree leaf findings and details the atmospheric study.

Introduction

Most commonly used pesticides are designed to be acutely toxic to target pests when applied at recommended concentrations and include the majority of chemicals generically classified as insecticides, rodenticides, fungicides, and herbicides such as atrazine and alachlor (Table 1). Herbicides are the most extensively used class of agricultural pesticides in the United States, both in terms of amounts applied and area treated. The triazine chemicals are the most heavily used herbicides in application to agricultural land for control of broadleaf weeds and selected grass species (1). Atrazine is a selective pre­ and post­emergent triazine herbicide used in forestry, grasslands, and grass crops, such as maize, and many other crop and non­crop areas for the control of broadleaf weeds (2). This chemical has been widely applied internationally, especially in North America and Europe, since the late 1950's. Approximately 22.5 million kg of atrazine were applied to corn fields alone in 1993 in major use states, with more than 3 million kg applied to Iowa cornfields (1). Studies have shown that atrazine is frequently detected in surface waters of the Mississippi River basin, primarily as a result of non­point source runoff from agricultural fields (3­7). Atrazine and several primary degradation products also have been commonly detected in ground water surveys throughout the midwest U.S. (8­13). A 1991 study of 303 wells across the midwest U.S. revealed 22.4% detections of atrazine in wells but only 3.3% detections of alachlor at concentrations $ 0.05 mg/L. However, alachlor ethanesulfonic acid (alachlor­ESA) was detected at $0.1 mg/L in 45.8% of the 153 well subset for which this alachlor metabolite was measured (11). A similar study in 1992 revealed atrazine detections in 43% of the 100 monitored wells (12). In a study of 28 agricultural sites in Iowa, atrazine was detected in 97% of ground­water wells sampled with concentrations ranging from 0.5 to 4,600 mg/L and in 100% of soil sampled with an average concentration of 51 mg/g (13). As a consequence of increasing reports of ground water contamination, the U.S. EPA classified both atrazine and alachlor as "restricted use" chemicals. The atrazine use label was revised in 1991 and 1993 in an effort to mitigate ground­ and surface­water contamination. In addition, Iowa implemented an Atrazine Management Plan to help further reduce the potential for ground water contamination (14).

During the summer of 1993 near Baldwin, Iowa, the observation of atypical phenotypic expressions, such as curled brown leaves, light yellow­green leaves (chlorosis), curled and crooked new branch growth (epinasty), or complete defoliation, on non­target hardwood trees was noted at a restoration site. The site is in a Conservation Restoration Program (CRP), a nationally promoted plan by the U.S. Federal government to "idle" agricultural production on designated lands (15). A total of 150 acres of this indigenous tall grass prairie site previously had been planted with ca. 65,000 trees, including red and white oak, green and white ash, box elder, and black walnut. Although the CRP allows use of herbicides to control weeds around planted trees, no pesticides have been applied at this site (15). The adjacent properties are corn and soybean fields where various herbicides are used annually. Samples of tree leaves collected from this site in late August 1993 revealed detectable levels of atrazine, metolachlor, and 2,4­D of the six measured herbicides. Tree leaf samples collected from this site in early August 1994 also contained measurable, although lower, concentrations of atrazine in all trees, including a tree exhibiting phenotypically "normal" expression.

All sampled trees were at a distance of more than 60 m from herbicide treated fields and were not exposed to herbicide inputs from direct surficial runoff of water from surrounding agricultural fields. Since many of the trees were located on hills, exposure to herbicide contaminated ground water that had migrated from the surrounding fields was considered unlikely. However, the subsurface hydrologic gradient at this site was not characterized. Other likely primary exposure routes are related to atmospheric processes (16). Herbicide drift during and immediately following application can be an important route for near off­site exposure (17). Following application, herbicides like atrazine can be lost to the atmosphere both by direct volatilization and by wind erosion of field soil particles to which the herbicide is sorbed. Once airborne, the herbicides can be transported variable distances before being deposited to the Earth's surface. In a study of pesticide deposition in Maryland, atrazine was found to persist in the atmosphere long enough for at least regional­scale transport to occur across several states for distances on the order of at least 1000 km (18). The wide­spread observation of atrazine, alachlor, and other measured herbicides in wet deposition samples collected in the midwest and northeast U.S. in 1990 and 1991 provides further evidence for their ability to undergo long­range atmospheric transport (19). Atmospheric related exposure mechanisms of organic contaminants to trees and other non­target off­field vegetation include uptake of dry gaseous herbicide by the leaf, dry deposition of particle­associated herbicide, and wet deposition related processes (rain, fog, dew) (20­24).

Interest in the possible exposure of trees to gaseous atrazine following volatilization from a field lead to a separate field investigation conducted in 1994 at an agricultural research farm in Crawfordsville, Iowa. This study was designed to measure the concentration of atrazine in the atmosphere during the summer while it volatilized from an application site as a function of time and weather. Both high performance liquid chromatography with ultraviolet detection (HPLC/UV) and gas chromatography with mass spectrometric detection (GC/MS) were used for identification and quantification of atmospheric herbicide concentrations. This report details these atmospheric measurements and summarizes herbicide detections in tree leaf samples collected at the Baldwin site in 1993 and 1994.

Methods

Tree Leaf Collection and Analysis

Leaf samples, consisting of 30­50 leaves per tree, were randomly collected from 3 trees exhibiting atypical phenotypic expression on August 26, 1993 and on August 3, 1994 at a CRP site near Baldwin, Iowa. In 1994, leaves also were collected on­site from one tree categorized as phenotypically "normal" (25). Leaf samples were homogenized, sub­sampled, solvent extracted, and analyzed using GC with either electron capture (ECD) or nitrogen-phosphorous (NPD) detection by the State Chemical Laboratory of the Iowa Department of Agriculture and Land Stewardship, Des Moines. Standard multiple residue methods were used in these analyses (26). Herbicides measured were atrazine, bromoxynil, cyanazine, 2,4­D, dicamba, and metolachlor in 1993, and atrazine and bromoxinil octanoate only in 1994.

Atmospheric Herbicide Measurements

Site Description and Air Sampling

The location for the atmospheric herbicide measurements was at Iowa State University's Southeast Research Farm (SERF), Crawfordsville, Iowa. The SERF encompasses 0.81 km2 of flat agricultural land, 5 km east of Wyman and 22 km north of Mt. Pleasant, Iowa. The on­site sampler was set up on June 10, 1994, to monitor the air over a 4047 m2 corn field test site situated on the east side of the SERF. An off­site sampler was positioned in a southeastern direction at a 20.1 m distance relative to the edge of the monitored plot. This is the minimum distance required between any atrazine treated field and nearby stand pipes or drainage fields that discharge directly into a stream, river, or reservoir. The minimum distance from an atrazine­treated field to a dam, levee or farm pond is 15.2 m, and to a lake or reservoir is 60 m. These guidelines were a result of U.S. EPA required label changes for atrazine applications (14). The southeast placement of the off­site sampler was determined by access to electrical power. Baseline data were collected for one week prior to post­emergent application of herbicides to the test site on the morning of June 16. Meteorological records from the research farm indicate that the early morning wind was 12.4 km/hr (Fig. 1), a little too brisk for pesticide application, so application was postponed until later that day when winds were slightly calmer. The temperature was above average with a high of 32.8 C and a low of 21.7 C, and the mean humidity was 73%. The reported precipitation for June 16 was 0.20 cm without a rain event (Fig. 1); this likely was dew from the early morning hours since the dew point was 22.2 C. The field previously had been tilled, and corn had been planted on May 27 and was about 20 - 24 cm high at the time of herbicide application on June 16. AATREX (atrazine) was applied at 56 mg/m2 (0.5 lb/acre) of active ingredient. Lasso (alachlor) also was applied at a 112 mg/m2 (1 lb/acre) concentration. Additional herbicides chosen to control the many weed species were Basagran (bentazon) at 56 mg/m2, 2,4­D at 27 mg/m2, and Sencor (metribuzin) at 5 mg/m2. All were liquids (except Sencor) applied simultaneously by a tractor mounted sprayer as a tank mixture in a single application. During application, winds were from the south at slightly less than 12.4 km/hr, blowing any herbicide drift toward the north and away from the 20.1 m off­site sampler. Air sampling operations began approximately 3 h after spraying was complete. The plot was not irrigated during the entire study.

Air sampler cartridges were 10 mL syringes packed with 0.5 g of Florisil sandwiched between two porous polyethylene discs. Most samples were collected using one cartridge, although tandem cartridges were used for several days following application and sporadically thereafter to gauge herbicide collection efficiency. No breakthrough of atrazine or alachlor to the backup cartridges was detected. Samplers were regulated by a timer and samples were collected for 23.5 to 24 h periods every other day. Air was drawn through the sample cartridge by the pump at a mean flow rate of 2.0 L/min (range 1.4 to 2.1 L/min) for the on­site sampler and 4.7 L/min (3.8 to 6.6 L/min) for the off­site sampler. The daily total volumes of air sampled for the on­site sampler averaged 2,860 ± 30 L/day and for the off­site sampler, 6,830 ± 70 L/day. The air sampling height was approximately 0.3 m above ground level for both locations. Samples were collected every other day over a two month period beginning pre­application on June 10 and following application of herbicides on June 16 through August 10, 1994. Weekly, unexposed field blank cartridges used as a quality control check revealed no detectable atrazine or alachlor.

Air sample preparation and analysis

A total of 81 air sample cartridges were eluted into 5 mL volumetric flasks with pesticide grade methanol. Aliquots of the cartridge eluant were transferred into two 2 mL glass vials. One set of vials for all samples was used to measure atrazine only by HPLC/UV. The second set of vials was used for GC/MS analysis of atrazine and alachlor for select samples.

HPLC analysis was accomplished using a Varian model 5000 HPLC and 9090 autosampler coupled to a Lambda­Max model 481 LC spectrophotometer. Extract injections of 50 mL were chromatographed on a 150 x 3.6 mm internal diameter C­18 reversed phase column with 5 mm packing. Detection was performed using UV absorption at 230 nm with an absorbance range set to 0.002 and recorded by a strip chart recorder and computer using Labtech chromatographic software. The isocratic mobile phase was 50% pesticide grade methanol in HPLC grade water at a flow rate of 1.5 mL/min. Atrazine standards of 1, 0.5, 0.2, 0.1 and 0.05 ng/mL were used for calibration. Standards were injected prior to sample analysis for initial calibration and to establish atrazine retention time. During sample analysis, every fifth injection was a calibration standard or a lab blank quality control sample. The HPLC detection level for atrazine was estimated at approximately 100 ng/m3 for the on­site sampler and 40 ng/m3 for the off­site sampler.

Samples chosen for GC/MS analysis consisted of 13 samples from a two week period after herbicide application where atrazine was detected by HPLC. Select air sample extracts were reduced in volume to 100 ­ 200 mL in toluene using nitrogen gas evaporation and analyzed on a Hewlett­Packard model 5890 GC equipped with a 5970A mass selective detector. The herbicides were separated on a 25 m x 0.2 mm internal diameter Hewlett­Packard Ultra 2 capillary column having a 33 mm film thickness using He carrier at 2 mL/min. Initially, select samples were chromatographed over a 25 min period using full scan acquisition for ions from 50 amu to 400 amu. Characteristic spectra for atrazine and alachlor were observed in most samples. Metribuzin, another applied herbicide amenable to GC analysis, was not detected. To improve sensitivity, extracts were reanalyzed in selected ion monitoring mode for characteristic ions of atrazine and alachlor only. Calibration standards containing atrazine and alachlor at 0.25, 0.5, 1.0, and 2.0 ng/ L were commercially prepared standards independent of those used during the HPLC analysis. The estimated detection levels by GC/MS were 10­30 ng/m3 for atrazine and 5­15 ng/m3 for alachlor, depending on air and final extract volumes.

Results and Discussion

Herbicides in the Tree Leaves

Herbicides detected in the leaf samples are shown in Table 2. Atrazine was found in all leaf samples collected from both years, with mean concentrations of 0.10 ± 0.01 mg/g in 1993 and 0.042 ± 0.011 mg/g in 1994. The atrazine concentration in the ash tree identified as phenotypically "normal" (0.036 mg/g) was similar to concentrations observed in the obviously stressed trees in 1994. Metolachlor was observed in all leaf samples from 1993 at a mean concentration of 0.64 ± 0.41 mg/g. 2,4­D (method detection limit, MDL, of 0.003 mg/g) was observed in a box elder tree at 0.009 mg/g in 1993. Cyanazine (MDL, 0.01 mg/g), dicamba (MDL, 0.002 mg/g), and bromoxinil (MDL, 0.001 mg/g) were not detected in the 1993 samples. Bromoxinil octanoate (MDL, 0.004 mg/g), the only other herbicide measured along with atrazine in 1994 samples, was not detected.

The higher concentrations of metolachlor relative to atrazine in all leaf samples in 1993 might be a result of one or more factors. Metolachlor could have been used on both the corn and soybean fields surrounding the trees, whereas atrazine is applied only to corn fields. Although specific application amounts on surrounding fields is unavailable, atrazine use on corn for all of Iowa in 1993 was about 3.02 million kg, whereas approximately 4.44 million kg of metolachlor were applied to corn and another 0.23 million kg on soybeans (1). More importantly, there are likely pronounced differences in the relative rates of herbicide loss from the fields and subsequent exposure to and uptake and accumulation by the trees for these two herbicides due to differences in physico­chemical properties (27).

Atrazine concentrations in 1993 leaf samples were about twice as high as observed in 1994. These differences might be a consequence of not sampling the same trees in each year. Alternatively, these concentration differences might reflect different levels of herbicide exposure to the trees each year. State­wide use of atrazine on corn in 1994 (3.39 million kg) was greater than in 1993 (28). However, atrazine usage on fields surrounding the CRP site near Baldwin actually might have been greater in 1993 to control excessive weed growth in an exceptionally wet year. These unusually moist conditions also might have contributed to greater herbicide volatilization from fields (see below) and provided enhanced atmospherically­related herbicide exposure to the trees from increased rain, dew, and especially fog events. Indeed, slowly drifting ground fog events were a common occurrence during summer 1993 at the Baldwin site. Drifting ground fogs may represent an important herbicide exposure mechanism, especially for vegetation growing near agricultural fields (29). Ground fogs forming over agricultural fields can accumulate substantial contaminant concentrations because volatilized pesticide air concentrations are highest immediately above the field surface and decrease rapidly with vertical distance above the surface or horizontal distance away from the field (30). Measurements of pesticides in fogs have revealed concentrations that can be enriched several orders of magnitude above levels predicted from Henry's law (29).

Herbicides in Air at the Southeastern Research Farm

The observation of herbicides in the tree leaves generated further interest (by CER) in understanding aspects of atmospheric exposure mechanisms, specifically the potential for gas­phase drift and related processes. Consequently, a separate study was conducted to measure volatilized herbicides, especially atrazine, in air above a corn field at the SERF following herbicide application.

Atrazine and alachlor were not detected in pre­application air samples collected from June 10 until June 14. Both atrazine and alachlor were detected in five and six, respectively, of the on­site air samples collected following application on June 16 through the sample taken on June 26­27 (Table 3). No herbicides were detected in any of the off­site samples (see below). The maximum atmospheric concentrations for both atrazine and alachlor occurred in the June 16­17 air sample, the collection of which began 3 h following herbicide application. The GC/MS determined concentrations were 1,130 ng/m3 for atrazine and 830 ng/m3 for alachlor. The concentration for atrazine determined by HPLC was 20% higher at 1,360 ng/m3. Similarly, atrazine concentrations obtained by HPLC for 3 other samples were 14­36% higher than determined by GC/MS, providing reasonable agreement considering different analytical techniques and calibration standards, and an added solvent evaporation step prior to GC/MS. Since GC/MS analyses provided more atrazine detections and all of the alachlor data, use of GC/MS data facilitates interpretation.

Within 2 days following application, the atmospheric concentrations of both herbicides had decreased by nearly 80% (June 18­19), and further down to 17 ng/m3 for alachlor and to non­detectable for atrazine by the June 22­23 sample. High air concentrations immediately after application, followed by a rapid decrease is similar to that observed in more detailed volatilization measurements of surface applied atrazine and alachlor (31, 32). On June 24, 8 days after herbicide application, it rained 0.56 cm, the last significant rainfall event measured at Crawfordsville, Iowa during the extremely dry summer sampling period (Fig. 1). The atmospheric herbicide concentration measured on June 24­25 rose to 110 ng/m3 for atrazine and 160 ng/m3 for alachlor. Concentrations then decreased on June 26­27 to 34 ng/m3 for atrazine and 100 ng/m3 for alachlor. The remaining samples collected through August 10 were below the atrazine detection limit by HPLC. The selected few air samples collected after June 26­27 that were tested for data verification by GC/MS also were below detection limits.

The resurgence in atmospheric herbicide concentrations immediately following the rain event on June 24 is an observation comparable to that reported by others. Enhanced volatilization of surface­applied or soil­incorporated pesticides following rain, dew, or irrigation events has been reported for atrazine (31, 32), alachlor (31), and for a number of other herbicides including, for example, simazine (31), prometon (32), DCPA (33), 2,4­D esters (34), triallate and trifluralin (35, 36), as well as for several insecticides (see e.g., 30, 31, 37). Following initial herbicide application, the surface concentrations are high and, if not already moistened, the soil surface typically is remoistened if using water­based tank mixtures. This condition commonly produces high initial volatilization fluxes of the herbicide that decrease rapidly as the soil dries and the herbicide adsorbs to mineral surfaces on the soil particle. Subsequent wetting of the soil particles, especially by rain, irrigation activity, or even by dew events, causes competitive displacement of the herbicide from the soil particle by the much more prevalent, strongly associating water molecules. Herbicides that are displaced from soil particles located at the field surface are thus available for volatilization. Chiou (38) recently has addressed this competitive sorption­desorption processes in relation to theoretical considerations of organic contaminant adsorption and partitioning to soils.

Jury et al. (39) have further postulated, modeled and demonstrated that, for some pesticides, the level of soil moisture content and associated water evapotranspiration are critical factors in moving sub­surface pesticide to the soil surface where volatilization can occur. For example, irrigation of a dry field causes release of atrazine bound to surface soil particles, allowing for subsequent volatilization. As volatilization progresses, the soil surface becomes depleted in atrazine until replenishment occurs from subsurface atrazine. Replenishment results when atrazine adsorbed to dry sub­surface soil particles is displaced by infiltrating irrigation water or precipitation and then is carried back toward the field surface as the soil water moves upward due to a hydrologic gradient driven by water evaporation. Although soil moisture content helps control this replenishment mechanism, note that excessive water inputs via substantial amounts of rains or irrigation potentially would lower overall pesticide volatilization from the field by transporting the herbicide off­site as surface runoff and/or further down into the unsaturated zone or even to the saturated zone. The lack of herbicide detection after June 27 is primarily a result of collecting low air volumes in this study, but also might, in part, be a consequence of the exceptionally dry conditions prevailing throughout the remainder of the study (Fig. 1). These prolonged dry conditions produced up to 1.25 cm cracks in the test plot soil and appeared to inhibit further occurrences of moisture­driven releases of adsorbed herbicide that would allow for detection of "peak" concentrations of volatilized herbicides as observed for the June 24­25 sample.

Atrazine concentrations for the June 16­17 and 18­19 air samples were nearly 40% greater than alachlor concentrations (Table 3). This observation is not consistent with the occurrence of a volatilization process alone, since the solid­phase vapor pressure of alachlor is nearly 36 times higher than atrazine (Table 1) (40). Nearly twice the mass of alachlor was applied versus atrazine and, thus, it seems unlikely that alachlor was rapidly depleted from the soil surface in the ca. 3 h period between end of herbicide application and initiation of the first air sample on June 16. Instead, a substantial portion of the measured atrazine in these samples must have been from wind erosion of soil particles containing atrazine. Glotfelty et al. observed mean losses of surface­applied herbicides of 20 g/ha per day for alachlor versus 1.9 g/ha per day for atrazine (31). Their data suggested that much of the measured atrazine (and simazine) was the result of collecting wind eroded particles and less from true volatilization on days having dry surface soil conditions . They also noted that herbicides applied as wettable powder formulations are more susceptible to wind erosion when the soil is dry. Liquid formulations of atrazine and alachlor were applied to the SERF plot. Interestingly, although the air concentration of alachlor began exceeding that of atrazine by June 20­21, the difference was never more than a factor of 3 through June 26­27. Unfortunately, our study was not designed to distinguish volatilization and wind erosion contributions to the air concentrations.

All the atmospheric herbicide concentrations detected in this study were collected by the sampler located on the herbicide amended corn field. The sampler placed at the 20­m buffer distance from the application site did not collect detectable herbicide concentrations when tested by HPLC, or by GC/MS for a few select air samples. For off­site samples collected in the days immediately following herbicide application when the on­site concentrations were high, this lack of detection may have been due in part to the prevailing winds coming from the south and blowing the herbicide drift away from this off­site sampler placed on the southeast side of the field. In most cases, however, the lack of detection at the off­site sampler was simply of consequence of the low air volumes being insufficient to collect detectable amounts, as occurred for the majority of on­site samples as well. Volatilized herbicide concentrations decrease rapidly as the herbicide is transported away from the field (30). In a USGS study (41) of pesticides in air and rain, weekly composite high­volume air samples collected at an urban location in Iowa City from April through September 1995 revealed ambient concentrations ranging from 0.013 to 2.2 ng/m3 for atrazine and 0.014 to 1.1 ng/m3 for alachlor. Corresponding rural measurements taken at the Cedar Rapids, Iowa airport with an air sampler positioned ca. 20 m from the edge of a corn field exhibited somewhat higher ambient concentrations ranging from 0.018 to 27 ng/m3 for atrazine and 0.012 to 6.1 ng/m3 for alachlor. Nearly all of the measured ambient herbicide concentrations in the 1995 study were one or more orders of magnitude lower than the herbicide detection limits achievable in the present study.

Conclusion

Air measurements 0.3 m above the treated field surface revealed that concentrations of ca. 100 ng/m3 or more of atrazine and alachlor existed for at least 4 days following surface application to corn fields. Substantially higher concentrations occur in the hours immediately following application. These concentrations might be an important source of drift exposure, particularly if ground fogs form over recently treated fields, enrich the herbicide concentrations in the fog water, and subsequently drift to off­site vegetation. The ratio of atrazine to alachlor air concentrations suggested that at least a portion of the measured atrazine (and possibly some alachlor?) in the air samples was from wind eroded soil particles to which the herbicide was sorbed. The study design did not allow for distinction of the volatilization and wind erosion components for these herbicides. Nor was it designed to provide estimates of the amount lost (flux) to the atmosphere in relation to the amount applied to the field. Glotfelty et al. (31) estimated that 19% of applied alachlor was lost over a 3­week period following surface application, but only ca. 2.4% of applied atrazine was lost to the atmosphere. Therefore, atmospheric loss was very important for decreasing alachlor soil residues. Whereas, this loss mechanism appears minimal in relation to other atrazine dissipation processes including uptake by field plants, microbial degradation, and/or other removal processes.

The flux of these herbicides to the atmosphere seems to be controlled at least in part by the soil moisture content, based on the observation of a rise in herbicide air concentrations following the rain event. This finding is consistent with other reports of detailed flux measurements made for these and other herbicides and insecticides.

No herbicides were detected in simultaneous air samples collected 20 m off­site, which was partly a consequence of prevailing winds carrying drift away from this sampler for days immediately following herbicide application. Low sampled air volumes also limited detections for both on­ and off­site samplers.

Observation of herbicide residues in the leaves of severely stressed or dying trees near agricultural fields brings to question the importance of atmospheric processes in facilitating either event higher concentration or chronic lower concentration exposures of herbicides to non­target off­field vegetation. Investigations designed to more fully elucidate these atmospheric exposure routes appear warranted in light of attempts by U.S. Government and State Agencies to create buffer zones in agricultural areas through the Conservation Restoration Program and similar initiatives.

Acknowledgments

We thank Jeff Boon, Center for Environmental Sciences, University of Colorado at Denver, and Roger Bishop at the State Chemical Laboratory of the Iowa Department of Agriculture and Land Stewardship and Kevin Vandee of the Southeast Research Farm. The use of firm, trade, and brand names in this paper is for identification purposes only and does not constitute endorsement by the U.S. Geological Survey.

References

1. Agricultural Chemical Usage: 1993 Field Crops Summary; U.S. Department of Agriculture, National Agricultural Statistics Service: Washington, DC, 1994; Ag Ch 1 (94).

2. The Pesticide Manual: A World Compendium; 8th ed.; Worthing, C. R., Ed.; British Crop Protection Council; Lavenham Press Ltd.: Lavenham, Suffolk, U.K., 1987, p. 36.

3. Pereira, W.E.; Rostad, C.E. Environ. Sci. Technol. 1990, 24, 1400­1406.

4. Pereira, W.E.; Hostettler, F. D. Environ. Sci. Technol. 1993, 27, 1542­1552.

5. Goolsby, D.A.; Battaglin, W.A. In Selected Papers on Agricultural Chemicals in Water Resources of the Midcontinental United States; Goolsby, D.A.; Boyer, L.L.; Mallard, G.E., Cpls.; U.S. Geological Survey Open­file Report 93­418: Denver, 1993, p. 1­25.

6. Battaglin, W.A.; Goolsby, D.A.; Coupe, R.A., In Selected Papers on Agricultural Chemicals in Water Resources of the Midcontinental United States; Goolsby, D.A.; Boyer, L.L.; Mallard, G.E., Cpls.; U.S. Geological Survey Open­file Report 93­418: Denver, 1993, p. 26­40.

7. Huber, W. Environ. Toxicol. Chem. 1993, 12, 1865­1881.

8. Aharonson, N. Pure and Appl. Chem. 1987, 59, 1419­1446.

9. Hamilton, P.A.; Shedlock, R.J. U.S. Geological Survey Circular 1080; U.S. Government Printing Office: Washington, DC, 1992; No. 92­10922.

10. Barbash, J.E.; Resek, E.A. Pesticides in Ground Water: Distribution, Trends, and Governing Factors; Ann Arbor Press: Chelsea, MI, 1996.

11. Kolpin, D.W.; Thurman, E.M.; Goolsby, D.A. Environ. Sci. Technol. 1996, 30, 335­340.

12. Kolpin, D.W.; Goolsby, D.A.; Aga, D.S.; Iverson, J.L.; Thurman, E.M. In Selected Papers on Agricultural Chemicals in Water Resources of the Midcontinental United States; Goolsby, D.A.; Boyer, L.L.; Mallard, G.E., cpls.; U.S. Geological Survey Open­file Report 93­418: Denver, 1993, p. 64­74.

13. Anderson, T.A.; Coats, J.R.; Kruger, E.L. 87th Annual Meeting of the Air and Waste Management Association; Cincinnati, Ohio, 1994; p. 1­8.

14. Hartzler, R.G.; Owen, M.D.K. 1994 Herbicide Manual for Agricultural Professionals; Iowa State University Cooperative Extension Service: Ames, IA, 1993.

15. James R. Tabor, personal communication.

16. Majewski, M.S.; Capel, P.D. Pesticides in the Atmosphere: Distribution, Trends, and Governing Factors; Ann Arbor Press: Chelsea, MI, 1995.

17. 1996 Farm Chemicals Handbook; Meister Publishing Co.: Willoughby, OH, 1996, p. E46­E48.

18. Glotfelty, D.E.; Williams, G.H.; Freeman, H.P.; Leech, M.M. In Long Range Transport of Pesticides and Other Toxics; Kurtz, D.A., ed.; Lewis Publishers, Inc.: Chelsea,MI, 1990; p. 199­222.

19. Goolsby, D.A.; Thurman, E.M.; Pomes, M.L.; Meyer, M.; Battaglin, W.A. In Selected Papers on Agricultural Chemicals in Water Resources of the Midcontinental United States; Goolsby, D.A.; Boyer, L.L.; Mallard, G.E., cpls.; U.S. Geological Survey Open­file Report 93­418: Denver, 1993, p. 75­89.

20. Strachan, W.M.J.; Eriksson, G.; Kylin, H;, Jensen, S. Environ. Toxicol. Chem. 1994, 13, 443­451.

21. Simonich, S.L.; Hites, R.A. Environ. Sci. Technol. 1994, 28, 939­943.

22. Patterson, S.; Mackay, D.; McFarlane, C. Environ. Sci. Technol. 1994, 28, 2259­2266.

23. Polder, M.D.; Hulzebos, E.M.; Jager, D.T. Environ. Toxicol. Chem. 1995, 14, 1615­1623.

24. Wagrowski, D.M.; Hites, R.A. Environ. Sci. Technol. 1997, 31, 279­282.

25. Arnold Webster, consulting arborist, Cedar Falls, IA, personal communication.

26. Pesticide Analytical Manual, Vol. 1. Food and Drug Administration, Department of Health and Human Services, 1994.

27. The Pesticide Properties Database; U.S. Department of Agriculture, Agricultural Research Service, Remote Sensing and Modelling Laboratory: Beltville, MD. Available at website http:/www/arsusda.gov//history.html, March 1995.

28. Agricultural Chemical Usage: 1994 Field Crops Summary; U.S. Department of Agriculture, National Agricultural Statistics Service: Washington, DC, 1995; Ag Ch 1 (95).

29. Glotfelty, D.E.; Seiber, J.N.; Liljedahl, L.A. Nature. 1987, 325, 602­605.

30. Majewski, M.S.; Glotfelty, D.E.; Paw U, K.T.; Seiber, J.N. Environ. Sci. Technol. 1990, 24, 1490­1497.

31. Glotfelty, D.E.; Leech, M.M.; Jersey, J.; Taylor, A.W. J. Agric. Food Chem. 1989, 37, 546­551.

32. Clendening, L.D.; Jury, W.A.; Ernst, F.F. In Long Range Transport of Pesticides and Other Toxics; Kurtz, D.A., ed.; Lewis Publishers, Inc.: Chelsea,MI, 1990; p. 47­60.

33. Majewski, M.S.; McChesney, M.M.; Seiber, J.N. Environ. Toxicol. Chem. 1991, 10, 301­311.

34. Gover, R. In Principles of Health and Safety in Agriculture; Dosman, J.A.; Cockcroft, D.W., eds.; CRC Press, Inc.: Boca Raton, FL, 1989; p. 222­225.

35. Gover, R.; Smith, A.E.; Shewchuk, S.R.; Cessna, A.J.; Hunter, J.H. J. Environ. Qual. 1988, 17, 543­550.

36. Majewski, M.S.; Desjardins, R.; Rochette, P; Pattey, E.; Seiber, J.; Glotfelty, D. Environ. Sci. Technol. 1993, 27, 121­128.

37. Spencer, W.F.; Cliath, M.M. J. Environ. Qual. 1973, 2, 284­289.

38. Chiou, C.T. In Encyclopedia of Environmental Analysis and Remediation, Meyers, R.A., ed.; John Wiley and Sons: New York (in press).

39. Jury, W.A.; Spencer, W.F.; Farmer, W.J. J. Environ. Qual. 1984, 13, 567­586

40. Foreman, W.T.; DeGiacomo, W.G. In U.S. Geological Survey Toxic Substances Hydrology Program­­Proceedings of the technical meeting, Monterey, CA, March 11­15, 1991; Mallard, G. E.; Aronson, D.A., eds.; USGS Water Resource Investigation Report 91­4034; USGS: Reston, VA, 1991, p. 223­226.

41. Foreman, W.T.; Goolsby, D.A.; Majewski, M.S.; Wiebe, F.W.; Battaglin, W.A. Presented at the 17th Annual Meeting of the Society of Environmental Toxicology and Chemistry, Washington, DC, Nov. 1996; Paper 577.

This manuscript contains graphics that must be viewed and printed by the Adobe Acrobat (TM) plug-in included with this program. Please click here to activate Acrobat and view the graphics.