| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
a Univ. of California, 420 S. Wilson Way, Stockton, CA 95215
b Environmental Monitoring Branch, Dep. of Pesticide Regulation, California EPA, 1001 I Street, Sacramento, CA 95812-4015
c California Dep. of Water Resources, P.O. Box 942836, Sacramento, CA 94236-0001
* Corresponding author (tlprichard{at}ucdavis.edu)
Received for publication July 1, 2005.
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONREFERENCES |
|---|
Abbreviations: ANOVA, analysis of variance • CIMIS, California Irrigation Management Information System • ET, evapotranspiration • LEACHM, Leaching Estimation and Chemistry Model • RL, reporting limit
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONREFERENCES |
|---|
These two scenarios are not inclusive of all geographical settings where residues have been detected in California's ground water (Troiano et al., 2000). Pre-emergence herbicide residues were detected in seven wells sampled within a 1554-ha area located near the town of Tracy, CA: atrazine was detected in five wells at 0.16 to 2.8 µg L–1, diuron in one well at 0.06 µg L–1, hexazinone in three wells at 0.051 to 0.11 µg L–1, and simazine (6-chloro-N,N'-diethyl-1,3,5-triazine-2,4-diamine) in one well at 0.098 µg L–1. Tracy is centrally located on the western side of the Central Valley of California (Fig. 1) . The predominant cropping pattern was a rotation of alfalfa with corn (Zea mays L.) and bean (Phaseolus vulgaris L.). The residues were related to agricultural applications, especially since the only reported pesticidal use of hexazinone was on alfalfa. The affected area was surveyed for any obvious potential pathways to ground water. The predominant soil was a clay soil that developed surface cracks on drying. Rapid movement through soil, termed macropore flow, has been identified as a potential pathway for contamination (Bouma et al., 1981: Harris et al., 1994; Lin et al., 1998). Movement of atrazine in cracking-clay soils in another area of California had been investigated (Graham et al., 1992). That study indicated potential movement of herbicides into cracks with some residues detected below the plow layer, but movement to shallow ground water was not confirmed.
|
In response to the detections near Tracy, an investigation was conducted in an alfalfa field to determine fate of diuron and hexazinone herbicide residues by measuring their mass distribution in soil cores, runoff water, and vegetation. Water characteristics of an accompanying water-holding pond were measured by monitoring the changes in pond-water depth over time to determine infiltration characteristics. Water infiltration from the pond was related to changes in elevation of closely associated ground water. An additional objective was to determine potential mitigation of the addition of a surfactant to the spray mixture. Theoretically, the surfactant would have facilitated greater interaction between the pesticide residue and soil matrix reducing potential runoff of residues (Huggenberger et al., 1973).
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONREFERENCES |
|---|
|
Distribution of diuron and hexazinone residues was compared between two treatments. One treatment represented grower standard pre-emergence herbicide treatment of hexazinone and diuron applied at 0.56 and 1.68 kg ha–1, respectively. The second treatment was an addition of surfactant to the herbicide mixture. The soil surfactant treatment was added to enhance infiltration into the soil, thereby reducing surface herbicide concentrations that would readily move off site during a storm or irrigation event. The surfactant was a proprietary blend of nonionic and anionic surfactants designed to enhance water and herbicide infiltration into the soil profile (Garrco Products, Converse, IN). A randomized complete block design was used with four replicate blocks (Fig. 2) . A third treatment was included comparing efficacy of paraquat and trifluralin application to the grower standard, but these data are not reported. Each replicate treatment was applied to one check. The border-check irrigation method can affect distribution of residues because water can redistribute residues as it advances down the length of the check. To measure potential spatial differences due to irrigation, each check was subdivided into thirds, with each third approximately 111.6 m in length. Soil and vegetation samples were taken from each third, representing the head, middle, and tail portions of the check.
|
|
Soil was sampled using a bit that was 33.4 mm in diameter and 102 mm in depth. Samples were obtained from the first depth by inserting the sampler down to the 76-mm depth and collecting all soil from the tube. For deeper depths, the samples were obtained from a central portion of the length of the bit that was calibrated at 60 cm3. This area was isolated from the rest of the core by inserting knives into slots located on either end of the centrally calibrated area. Soil was removed from the ends and the isolated sample collected into a glass mason jar. Before inserting the sampling bit into the borehole, a larger diameter auger that was 64 mm in diameter was used to clear the borehole of soil down to the next appropriate depth for sampling. The smaller soil sampling bit was cleaned between each sample with soapy water, rinsed twice in sequence with deionized water and then methanol, and air-dried before reuse. Excess soil from the sampling was collected in a bucket and removed from the field. The sample borehole was refilled with noncontaminated soil and the sample borehole plugged at the surface with a mixture of soil and bentonite. Samples were placed on dry ice and kept frozen until chemical analysis. At the time of analysis, a portion of each soil sample was used to determine percentage moisture. Bulk density was calculated based on the dry weight of the sample and the volume of the soil sample.
Vegetation samples were taken at the background and at the April soil coring. Alfalfa plants were collected from the areas that encompassed each of the soil cores. The vegetation was placed directly into a 1.1-L glass jar and the jar lid sealed. The samples were transported on wet ice and then stored at 3.3°C until analysis.
Runoff Water Sampling
Runoff water was monitored from two of the four replicates from each treatment. A soil berm was formed at the tail end of those checks to facilitate runoff collection. Water was collected in a basin comprised of a 19-L plastic bucket that was installed at the lowest point inside the bermed check. A battery-powered pump and float switch placed in the collection basin moved runoff water into a 3.8-L sample collection jar. A larger capacity gas-operated pump was utilized to assist the lower flow battery-operated sampling pump during times of higher runoff. Pumps operated until runoff was complete. Both pumps were metered and withdrew water simultaneously from the same collection bucket. Once the initial sample jar was full, the remaining water was pumped through a water meter and then through a proportional sampling device, which collected 2% of the outflow volume in a series of 19-L buckets. A 1-L subsample was collected from each 19-L bucket and it was stored in an amber glass bottle that was sealed with a Teflon-lined cap. This subsample represented the average concentration of herbicide residue moving offsite for the volume indicated by the water meter. The water samples were placed on wet ice until arrival at a storage facility and remained at 3.3°C until analysis.
Holding Pond Soil and Water Sampling and Ground Water Depth Measurement
At the beginning of the study, the holding pond was dry so soil samples were collected from two locations at the bottom of the pond. Soil sampling and handling procedures were the same as described for sampling in the field. Water samples were collected in a 1-L amber glass bottle sealed with a Teflon-lined cap. Samples were placed on wet ice and refrigerated at 3.3°C until chemical analysis. It is important to note that the water entering the pond was the result of runoff from a larger area than the experimental area, for example, the 12 experimental checks were a portion of the total planted area. Although pond-water samples were not a direct measure of loss of herbicide mass from runoff that originated solely in the treatment areas, pond-water concentration was an important indicator of the potential fate of herbicides applied to the field. Water samples were collected after each irrigation when pond inflow had stopped.
A technique was developed to determine pond volume from water depth measurements. First, the volume of the pond was estimated by surveying the three-dimensional measurements of the pond. A relationship was then established between pond depth and estimated volume. This relationship was calibrated by relating measured inflow volume of runoff water to the concomitant increase in pond depth. Inflow was measured using a 200-mm throat broad-crested RCB flume (Plasti-Fab, Tualatin, OR) equipped with a stilling well and pressure transducer to measure head (Clemmens et al., 1984). An additional transducer was placed at the bottom of the pond to measure pond depth. Temporal changes in pond depth were used to estimate pond water volume and, subsequently, infiltrated water according to Eq. [1]:
 | [1] |
Changes in depth to ground water in an area adjacent to the pond were measured with a pressure transducer. The transducer was placed into an open-ended, close-fitting PVC pipe and then affixed in the borehole.
Chemical Analysis
Chemical analysis for diuron and hexazinone were contracted to the Center for Analytical Chemistry (California Department of Food and Agriculture, Sacramento, CA). Soil samples were extracted with a hexane and acetone mixture, vegetation and sediment samples were extracted with acetonitrile, and water samples were processed through a C18 SPE column (Phenomenex, Torrance, CA) that was eluted with methanol. Reconstituted extracts were analyzed with liquid chromatography (model 2690, Waters Corp., Milford, MA) and mass spectrometry methodology (Finnigan LCQ Decca, Thermo Electron Corp., Madison, WI). Specific methods of analysis are available on request. A standard operating procedure had been developed to assure verification of laboratory results (Segawa, 1995). During method development, the laboratory established acceptable warning and control limits, and sample reporting limits (RL). For continuing quality control, a solvent blank and two matrix spikes were analyzed with each extraction set. Extraction sets were analyzed again if the spikes did not fall within the established warning and control limits. Each sample was tracked from initial sampling to final analytical result with an accompanying chain of custody form. Reporting limits for diuron and hexazinone were 8 µg kg–1 for soil, 0.25 µg L–1 for water, 10 µg kg–1 for sediment, and 100 µg kg–1 for vegetation samples.
Data Analysis
Study objectives were to determine fate of herbicides applied to the soil and to compare addition of surfactant on herbicide distribution within the field. With respect to fate of herbicides, the relative potential for downward leaching of residues within the field was compared with movement of residues in runoff water that eventually infiltrated from a holding pond at the edge of the field. Extent of downward movement within the field was measured in soils cores and compared with soil water movement produced from rainfall and irrigation. In addition, expected soil distributions of water content and herbicide residues were predicted using the LEACHM model version 4 (Hutson, 2003). Since LEACHM uses convection–dispersion methodology to model water flow, deviations between observed soil concentrations and predicted values could indicate the extent to which movement was dominated by preferential pathways. This method to deduce preferential flow by comparison of observed to model predictions of soil distribution has been previously reported by Gish et al. (2004) and Tindall and Vencill (1995). Data for physical and chemical properties of diuron and hexazinone used in the model were obtained from the Oregon State University Extension website, which is a compilation of data from two USDA publications (Table 3) (Vogue et al., 1994; Augustijn-Beckers et al., 1994; Wauchope et al., 1992). Values for soil texture and bulk density were taken from measured data (Table 1). Rainfall data was obtained from the Tracy Pestcast weather station, which was located <3 km from the experimental site (Univ. of California, 2002). Estimates for reference evapotranspiration (ETo) were obtained from a CIMIS (California Irrigation Management Information System; California Dep. of Water Resources, 2001) weather station located in Manteca, CA, which was located approximately 22 km east of the experimental site. Surface flux density for water intake was set at 390 mm d–1, which was calculated from the irrigation events. Crop growth was included to provide for loss of water through plant transpiration and the roots were distributed with 60% apportioned in the first 305-mm soil depth. This estimate was determined from a stylized depiction of the root distribution of a mature plant, which indicates that a large portion of active roots are located in the first 305 mm of soil (Weaver, 1926).
|
For surfactant treatment, effects on recovered mass were measured using a split-plot analysis of variance (ANOVA) for a randomized complete block design with surfactant treatment as the main effect and location in the field as the split-plot factor (i.e., head, middle, and tail end location) (SAS Institute, 1988). The effect of field location was measured using orthogonal regression contrasts measuring linear and quadratic spatial effects. Since data from the deposition sheets indicated potential differences in application between surfactant treatments, the deposition data were used as a covariate in ANOVA. In addition, potential for heterogeneity of variance was indicated using Levene's Test for homogeneity of variance so values were transposed to base 10 logarithms before statistical analysis (SAS Institute, 1988).
Recovered pesticide mass was determined as the product of concentration and mass of soil or water. The first two soil cores were contiguous so the concentration for the first 76-mm soil depth was used to directly estimate mass recovered from the first soil segment. To account for discontinuity in sampling between the remaining soil depths, estimates were integrated as the product of the average concentration and bulk density between depths. Herbicide residues were expressed on a kg ha–1 basis to facilitate comparison to application rates.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONREFERENCES |
|---|
|
|
Rainfall after Day 85 was minimal and owing to increasing solar radiation, the accumulation of ETo began to accelerate. The two sharp upward spikes in accumulated rainfall plus irrigation illustrates the contribution of water from the two irrigations (Fig. 3). The irrigations supplied enough water to match the cumulative deficit in ETo. Both irrigation events were similar in terms of total run time and onflow volumes, averaging 175 mm. Calculated runoff depth for the irrigations was 4.3 and 12.4 mm for the first and second irrigations, respectively. Differences between irrigations were caused by small differences in on-flow volumes, run times, and soil infiltration rate influenced by antecedent soil moisture content. The proportion of on-flow water volume occurring as runoff was 2.5% of the first and 7.1% of the second applied irrigation volumes. Runoff from the second irrigation was considered more reflective of typical conditions.
The moisture profile of the soil cores at the third sampling date, which was after two border-check irrigations, was reflective of a distribution caused by the dynamics of border-check irrigation (Fig. 4C). Within checks, there was a significant linear effect (P = 0.02) on total water content of the cores, which was a decrease in water content of soil cores from the head to tail ends—average estimated mass of water in cores was 228 ± 17, 213 ± 16, and 205 ± 12 g for head, middle, and tail end locations, respectively. Higher moisture at the head end of the border check is caused by greater opportunity time for water infiltration. More importantly, water content remained elevated at the lowest sampled depth, indicating that irrigation treatments caused drainage and provided a potential leaching environment.
Soil and Vegetation Sampling within the Alfalfa Field
Vegetation
Hexazinone was not detected in vegetation samples. Diuron was not detected in background samples but it was detected in 7 out of 24 samples obtained before the first irrigation event on 3–6 April. Detections were at or just above the reporting of 100 µg kg–1 with the average for detections at 118 ± 18 µg kg–1 on a dry mass basis. Plant yield measured on 3 April was 2880 kg ha–1, on a fresh mass basis. Assuming vegetative water content of 80%, dry mass would be estimated at 576 kg ha–1, which at the average concentration of diuron would have sequestered diuron at a rate of 0.000068 kg ha–1. Since presence in vegetation was a small portion of the total application rate (0.004%), the following discussion will focus on the fate of residues in soil and water samples.
Soil
Background Sampling.
Diuron residues were detected in all samples centered at the 38-mm depth and only sporadically in the next lower depth centered at 114 mm (Table 4, please note that data at each depth were averaged over all locations). Hexazinone residues were essentially undetected, which was likely due to its lower application rate. Summation of the estimated mass recovered from all depths indicated an average recovery corresponding to 0.12 kg ha–1 diuron, which was approximately an 8% carryover from the previous year's application. There were no significant differences in mass recovered between spatial locations within checks (Table 2).
|
For total mass recovered per core, a test for the main effect of surfactant indicated greater total amount of diuron recovered in treatments with added surfactant, but when deposition data were used as a covariate in the ANOVA, the effect was nonsignificant (Table 2). The test for a linear effect between the head to tail location was significant with or without the addition of the covariate, which appeared to be caused by lower recovery of diuron in the head portion of the minus surfactant plot. When tested within treatments, the linear effect was significant for the minus surfactant treatment but not in the plus surfactant treatment. Although statistical tests were not significant, the pattern for hexazinone appeared similar to diuron where the head end of the minus surfactant treatment appeared lower than the rest of the data, but the residues were more evenly distributed in the plus surfactant treatment. Since deposition data indicated very even application from the head to tail portion of each treatment (Table 2), this pattern could indicate that the surfactant was effective at minimizing redistribution of residues within the field.
Mass of residues recovered from the total soil core length averaged over both treatments was 0.59 kg ha–1 for diuron and 0.11 kg ha–1 for hexazinone. These represented a decrease from the application day values of 65% for diuron and 74% for hexazinone. Based on first-order exponential dissipation and a sampling interval of 106 d, estimated half-life values were 70 and 55 d for diuron and hexazinone, respectively.
After Second Irrigation.
The alfalfa field received two border-check irrigations before soil sampling on 26 June, resulting in a cumulative total application of 498 mm of water. Distribution of diuron and hexazinone residues were similar to the April sampling where diuron residues were mainly located in the first soil segment and with hexazinone residues split nearly evenly between the first two soil segment depths (Table 4). Although the head end of the plots received greater water infiltration and percolation as observed in Fig. 4, soil distribution was the same between locations, indicating similarity in downward movement of residue between head and tail locations (data not shown).
For total mass recovered per core, the overall linear effect for location of checks was significant for both diuron and hexazinone where the concentrations were lowest at the head end when compared with the tail end (Table 2). With border-check irrigation, water flows from the head to the tail and this pattern could indicate movement of residues with the irrigation water. In this case the surfactant was not adequate to minimize redistribution of residues in the irrigation water.
Mass of residue recovered from the total soil core length averaged over both treatments was 0.18 kg ha–1 for diuron and 0.04 kg ha–1 for hexazinone. These represented a decrease from the application day values of 90% for diuron and 91% for hexazinone. Based on first-order exponential dissipation and a sampling interval of 198 d, estimated half-life values were 62 and 56 d for diuron and hexazinone, respectively. These estimates do not include residue lost in tailwater.
Soil LEACHM Model Predictions
LEACHM model predictions for water content and herbicide concentration in the soil profile were compared to observed data at each sampled depth (Table 4). The comparisons were made only to determine whether preferential movement had a great effect on soil distribution of residues in relation to a situation where flow would have been dominated by convection–dispersion theory. At the April soil sampling before the first irrigation, LEACHM predicted slightly lower water content for the deepest soil segment. In contrast, measured water contents indicated much drier soil conditions in the first 76-mm segment and wetter conditions in the remaining profile. The observed data could be reflective of the influence of macropore flow where water preferentially flowed through surface cracks wetting subsurface soil. The pattern was similar at the June sampling, which was after two surface irrigations. In addition to soil structure effects on water flow, plant transpiration is another process that would produce a drier surface soil condition. Although LEACHM modeled approximately 45 mm of water lost to evaporation and 155 mm to plant uptake between the first and second soil coring, relatively high water content was maintained in the first soil segment. LEACHM allows upward movement of water, which may be a cause of the overestimation for this soil.
Surprisingly good agreement was observed between predicted and observed soil concentrations for both pesticides at the April and June sampling dates (Table 4). The agreement between concentrations indicated that the bulk of the residues were confined to the upper layers of soil even though the soil water distribution indicated an effect of macropore flow early in the season. LEACHM predicted deeper movement for hexazinone, an effect that was observed in the field data.
Runoff Water Sampling
A significant difference in diuron concentration was measured between irrigations where the concentration measured in runoff from the first irrigation was approximately twice the concentration of the second irrigation runoff event (Table 5). Although hexazinone concentrations also appeared greater at the first irrigation, the level of probability indicated only a trend (P = 0.07). Significant differences in mass of herbicide leaving the field were not detected between treatments or irrigations. The concentration of diuron herbicide was reduced by 50% from the first irrigation but the runoff volume had tripled in the second irrigation, resulting in no significant differences in the mass leaving the field. The results for hexazinone were similar. Addition of the surfactant did not affect the concentration of herbicides in runoff water (Table 5).
|
|
Comparison of the simultaneous measurements between pond water depth and ground water depth measured in a borehole near the pond indicated that the shallow ground water was relatively constant until the first irrigation, at which time the depth from the surface decreased sharply (Fig. 5B). A similar response was measured for the second irrigation; the rise in ground water elevation corresponded to the loss of water from the pond. The quick response in the ground water level indicated a short travel time for pond water to recharge nearby ground water.
Soil Sampling
Presence of residues in the soil profile of the pond was determined when the pond was dry, which was at the background and before the first irrigation soil sampling dates. For the background sampling, diuron residues at 45 ± 18 µg kg–1 were detected down to the fourth deepest sample, which was centered at 645 mm below the bottom of the pond. Although these soil concentrations were low, they were indicative of movement of diuron residues from the field during the previous season. In addition, the pond was excavated to between 2500 and 3000 mm so residues were much deeper in the soil than measured within the field. Diuron residues were also detected in soil samples taken before the first irrigation and they were located at the third deepest sample centered at 345 mm below the bottom of the pond. The average concentration was slightly greater at 69 ± 49 µg kg–1. Hexazinone was not detected at either sampling date.
Water Sampling
Herbicide residues were measured in filtered pond water samples collected after rain runoff water at 2.2 µg L–1 for diuron and 0.6 µg L–1 for hexazinone. As noted earlier, the pattern of rainfall did not generate significant runoff from the field; therefore, these samples were probably reflective of conditions near the pond and not the entire field. Samples collected after the inflow to the pond had ceased from the first irrigation contained diuron at 12.4 µg L–1 and hexazinone at 1.0 µg L–1. The concentrations in samples collected after the second irrigation were similar to the first irrigation runoff samples with diuron at 11.8 µg L–1 and hexazinone at 0.9 µg L–1. The concentration of diuron in the pond in the first irrigation was about twice that of the average treatment runoff, but hexazinone concentrations in the runoff water were similar. The cause for these variable differences is not apparent but may be due to a combination of variability in spray deposition and in application of irrigation water throughout the entire field: irrigation of the entire field required seven individual irrigation sets applied sequentially. Runoff volumes between sets were variable as measured by pond inflow volumes.
Groundwater Depth and Herbicide Concentration
Samples of ground water obtained from boreholes near the pond on 24 October were located on a transect at 3, 6, 12, and 48 m from the pond. At this time, the last irrigation had occurred >40 d before this measurement; therefore, the pond was nearly empty for about 25 d. Depth to ground water ranged from 3080 to 3140 mm with little hydraulic gradient measured between boreholes. Herbicide residues were detected with the concentration of diuron in the water, decreasing with distance from the pond water; the concentration was near 2.5 µg L–1 near the pond and nondetectable at 12 m and further (Fig. 6)
. The hexazinone concentrations found in the boreholes were similar at all sampling distances at about 0.5 µg L–1. The cause of the difference in pattern of detections is unclear.
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONREFERENCES |
|---|
Deeper soil movement observed for hexazinone was another indication of intimate interaction between pesticide residues and the soil matrix because the difference reflected their respective physical and chemical properties. The Koc for diuron is greater than for hexazinone, which should result in greater soil retention and less potential for downward movement. Dissipation rates are relatively long for both diuron and hexazinone, so soil adsorption would be the major determinant for differences in soil distribution (Jury et al., 1987; Gustafson, 1989). In a previous comparison between diuron and cyanazine, less soil movement of diuron was attributed to greater soil adsorption (Majka and Lavy, 1977). The concurrence between LEACHM modeled and observed soil concentrations of diuron and hexazinone and their relative soil distribution indicated a close association between the pesticide residues and the soil matrix. We are not suggesting that the primary mechanism of water movement was through convection–dispersion, only that the similarity in results between the modeled and observed data indicated that effects of preferential flow in macropores were localized to a shallow layer of the soil. In contrast, Tindall and Vencill (1995) noted similar anomalous high concentrations of atrazine [6-chloro-N-ethyl-N'-(1-methylethyl)-1,3,5-triazine-2,4-diamine],dicamba (3,6-dichloro-2-methoxybenzoic acid), and 2,4-D [(2,4-dichlorophenoxy)acetic acid] deeper in a cracking-soil profile and ascribed the similarity in movement to bypass due to preferential flow. The presence of the alfalfa root system in our study could have provided additional sources for pesticide adsorption, further emphasizing their differences in Koc.
Another factor lessening the influence of soil macropores was the typical pattern of rainfall. Dry conditions at the onset of the study would have promoted surface cracking. Then, winter rains would have filled the cracks causing swelling of the clay soil before the crop irrigations. Hourly rainfall data indicated that 154 h of measured rainfall occurred at the site between the herbicide application and the April soil sampling. The rainfall intensity of 70% of events was 1 mm h–1 or less. For the first 62 d after pesticide application, the maximum hourly intensity measured in a 24-h period was 3.3 mm h–1 and at only six times were intensities >1 mm h–1. After 15 Feb. 2000, a few rains were recorded with a maximum intensity at 7.62 mm h–1. Macropore flow was shown to be a major pathway for solute movement under high intensity rainfall (Júnior et al., 2004). However, the low intensity rain events typical for this site would have tended to swell the clay soil, thereby minimizing effect of larger water applications on rapid downward movement. Lin et al. (1998) reported temporal changes in infiltration rates of Vertisol soils where rainfall events decreased macroporosity, resulting in associated decreases in infiltration rates. Decreased velocity in the downward movement of water would have also contributed to greater interaction between herbicide residues and the soil matrix.
In contrast, data collected for the pond indicated fast movement of water to shallow ground water where water infiltrated from the pond within a few days after collection, elevating the level of nearby shallow ground water. This route provided a direct conduit for residues dissolved in runoff water to enter ground water. The major route of water loss was apparently through the sides of the pond because over time fine soil particles collected and formed a sediment on the bottom of the pond. Even though the herbicide concentrations measured in runoff water were low, they were about an order of magnitude greater than the concentrations measured in nearby domestic wells. Based on the rapid injection of pond water and herbicide residues into the ground water, we concluded that the pond was the most likely source for contamination. It should be noted that the specific volumes of water or herbicide concentrations generated in other years or from other sites could be larger or smaller depending on nuances in irrigation management and subsequent runoff management. As part of the Nebraska Management Systems Evaluation Area study, Spalding et al. (2003) monitored concentrations of pre-emergence herbicides in ground water using multilevel samplers located up-gradient, within, and down-gradient of the experimental site. Seasonal peaks in detections of pre-emergence herbicides in samplers located down-gradient of the plot were attributed to recharge that originated from ditches that collected runoff water from surge irrigation treatments.
Ground water sampled near the pond later in the season indicated similar concentration of hexazinone along a transect that radiated outward from the pond. Diuron concentration, however, decreased with increase in distance from the pond. It may be serendipity but results from the domestic well sampling that resulted in this investigation detected only hexazinone residues in the well located adjacent to this field, mimicking the pattern measured along the transect. Another well situated at the opposite end of this section of land, at around a distance of 1609 m, contained residues of atrazine and bromacil [5-bromo-6-methyl-3-(1-methylpropyl)-2,4(1H,3H)pyrimidinedione]. And as indicated in the introduction, simazine and other combinations of residues had been detected in wells within the original six-section study area. This indicates a complex scenario for pesticide movement to ground water that is determined by a combination of localized cropping patterns, which dictate specific pesticide use, and the specifics of water application and subsequent capture and flow of water from the ponds to ground water.
Lastly, the addition of surfactant to the spray mixture produced mixed results. Before irrigation events, the surfactant appeared to maintain an even distribution of residues throughout the check. After two irrigation events, residues appeared greater at the tail end of the checks, indicating that the volume of water supplied in irrigation washed out potential for surfactant to retain residues. Previous studies with surfactants have indicated variable effects. For example, increased mobility was not observed for lindane (
-1,2,3,4,5,6-hexachlorocyclohexane), diuron, or atrazine (Huggenberger et al., 1973), whereas increased mobility into soil was observed for trifluralin [2,6-dinitro-N,N-dipropyl-4-(trifluoromethyl)benzenamine] and oryzalin [4-(dipropylamino)-3,5-dinitrobenzenesulfonamide] (Koren, 1972) and for lindane (Sanchez-Camazano et al., 2000). Effects of surfactant are complex and depend on the physicochemical properties of pesticides, the properties of the soil, and the modification of the surfactant on both. The lack of a consistent surfactant effect in this study indicated that it would not be an effective mitigation measure for this soil, crop, and irrigation combination. Instead, mitigation effects will focus on management of water infiltration rates observed in the pond.
|
CONCLUSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONREFERENCES |
|---|
|
ACKNOWLEDGMENTS |
|---|
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONREFERENCES |
|---|
Related articles in JEQ:
This article has been cited by other articles:
|
|
 |
|
  H. V. Kazemi, S. H. Anderson, K. W. Goyne, and C. J. Gantzer Atrazine and Alachlor Transport in Claypan Soils as Influenced by Differential Antecedent Soil Water Content J. Environ. Qual., June 23, 2008; 37(4): 1599 - 1607. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||
