Spot Fumigation: Fumigant Gas Dispersion and Emission

Synopsis. The reduction of fumigant pesticide emissions with tree site-specific application for preplant management of orchard replant disease is disc...
1 downloads 11 Views 3MB Size
Download PDF
Environ. Sci. Technol. 2009, 43, 5783–5789

Spot Fumigation: Fumigant Gas Dispersion and Emission Characteristics D . W A N G , * ,† G . B R O W N E , ‡ S . G A O , † B. HANSON,† J. GERIK,† R. QIN,† AND N . T H A R A Y I L †,§ USDA-ARS, Water Management Research Unit, San Joaquin Valley Agricultural Sciences Center, Parlier, California 93648, and USDA-ARS, University of California, Davis, California 95616

Received May 28, 2009. Revised manuscript received June 24, 2009. Accepted June 25, 2009.

Reducing emissions of volatile organic compounds (VOCs) from fumigant pesticides is mandatory in California, especially in “nonattainment areas” like the San Joaquin Valley that do not meet federal air quality standards. A two-year field study was conducted to examine the feasibility of site-specific fumigant application only at future tree sites with dramatically reduced amounts of fumigant chemicals on an orchard basis. Soil gas distribution and atmospheric emission of 1,3dichloropropene and chloropicrin were measured after applying InLine using subsurface drip irrigation. It was predicted that, except in the surface 20 cm of soil, satisfactory pest control could be achieved within a 15 cm radius from the injection point. Also, at radial distances of 15-51 cm from the point of fumigant injection, effective nematode control may be achieved. Cumulative atmospheric emission of the fumigants was estimated to be 18-23% of the applied active ingredients in plots that had been cover cropped with Sudan grass and 2-6% in plots that had remained bare for several months before treatment. Considering the significantly small amount of fumigant used on an orchard basis, the spot drip fumigation may achieve a 10-fold reduction in atmospheric VOCs load from fumigant pesticides.

Introduction The San Joaquin Valley (SJV) of California is a major agricultural area with approximately 3.2 million ha in cultivated annual and perennial crops. Over 0.4 million ha of this cropland is devoted to fruit and nut orchards, which typically are replaced every 15-30 years to maintain economic productivity. Preplant soil fumigation is used in the process of orchard replacement to prevent replant problems. Replant problems typically are biological in origin, involving residual effects of the previous crop on microbial communities, soilborne pathogens, and physical and chemical soil properties (1). As in other parts of the world, the use of soil fumigation in the U.S., especially in the SJV, is faced with several environmental challenges. At an international level, soil * Corresponding author phone: 559-596-2852; e-mail: dong.wang@ ars.usda.gov. † USDA-ARS, Water Management Research Unit. ‡ USDA-ARS, University of California. § Currently at Clemson University, Clemson, South Carolina 29634. 10.1021/es9015662

Not subject to U.S. Copyright. Publ. 2009 Am. Chem. Soc.

Published on Web 07/08/2009

fumigation with methyl bromide (MeBr) is restricted to temporary critical use exemptions (CUE) because of the fumigant’s potential to deplete stratospheric ozone (2). Orchard replanting currently is covered by a CUE for MeBr use, and active research on MeBr alternatives is required for the approval of nominated CUE requests. Stockpiles of MeBr have been depleted, and only a small amount of MeBr is available under negotiated CUE. Orchard growers can only use CUE allocated MeBr if the alternative 1,3-dichloropropene (1,3-D) and chloropicrin (CP) containing fumigants (e.g., Telone II, Telone C35, InLine) is limited by either township caps and buffer zone requirements for protecting direct exposure to humans and the ecosystem or heavy soil conditions that would render 1,3-D ineffective. In California, use of the fumigants 1,3-D and CP are highly regulated not only because of their direct toxicity to humans and the ecosystem but also because of their contributions to volatile organic compounds (VOCs) in the atmosphere. VOCs react with nitrogen oxides in the ambient air under sunlight to form ground level ozone (3), which is a primary smog-forming ingredient. Under the 2007 Ozone State Implementation Plan, the California Department of Pesticide Regulation has been required to reduce agricultural emissions of smog-forming VOCs from soil fumigants and other pesticides in several regions of the state, especially the SJV. It is estimated that, in California, significant reductions in fumigant use will be required to drop VOC concentrations to acceptable levels, especially during the peak ozone period from May 1 to October 31. Additionally, the U.S. EPA recently announced new and more stringent safety measures for soil fumigants to decrease exposure risks for agricultural workers and bystanders and people who live, work, or otherwise spend time near fields that are fumigated (4). These regulations will place further restrictions on uses of 1,3-D and CP, including preplant applications for orchard replanting. Practical and effective emission reduction techniques are sorely needed in orchard replant soil fumigation. Conventionally, soil fumigants are shank injected to approximately a 45 cm (18 in) depth over the entire orchard area with shank knives spaced 51 cm (20 in) apart (the fumigant delivery nozzles are typically spaced 26 or 51 cm apart, depending on whether one or two nozzles are used per shank knife), and the soil surface is not covered with a plastic film after treatment. Because of their high vapor pressure, fumigants quickly vaporize after application and move through the soil by gas-phase diffusion. Once a fumigant reaches the soil surface, its gas phase is emitted to the atmosphere. The rate and amount of atmospheric fumigant emission can be reduced by proper management practices, including the use of plastic films to increase volatilization resistance (5-7), surface-applied organic amendments to accelerate fumigant degradation (8), sprinkler irrigation to reduce effective soil porosity for gas diffusion (9, 10), or the application of fumigant chemicals through drip irrigation (11-13). Because of the large tree spacing, soil fumigation in an orchard replant presents a unique opportunity of reducing total emissions by fumigating only the tree sites with much smaller amounts of fumigant chemicals applied on a field or area basis. It has also been demonstrated that replant disease of almond and other stone fruits can be prevented by preplant spot fumigation (14). Such spot treatments, which use approximately 10-25% of the fumigant needed for strip or broadcast fumigation, offer great potential for significantly reducing fumigant emissions, but this has not been investigated. VOL. 43, NO. 15, 2009 / ENVIRONMENTAL SCIENCE & TECHNOLOGY 9 5783

Although fumigant gas distribution and emission dynamics are well-documented for conventional broadcast shank injection (5, 6), gas distribution and emissions have not been monitored from spot drip applications of fumigants under field conditions. The application method involves a pointsource input into a semi-infinite porous medium (soil), and therefore, estimates of atmospheric emissions of fumigants based on data from linear-source shank application are not applicable. As part of a large project on the development of integrated alternatives to MeBr soil fumigation for almonds and other stone fruits, a two-year replicated field study was conducted to develop and demonstrate site-specific application or drip spot application of soil fumigants only at tree sites with significantly reduced total amounts of fumigant chemicals and treated areas per orchard (14). The specific objectives of the study reported here were to measure 1,3-D and CP soil gas distributions at various radial distances from the subsurface spot drip fumigation and fumigant emissions under two soil conditions, with or without a preplant cover crop treatment. The cover crop treatment was selected for the potential benefit to improve the soil’s physical and chemical properties and to break possible disease cycles in the previous orchard crop.

Experimental Section Field experiments occurred in adjacent field blocks at the USDA-ARS San Joaquin Valley Agricultural Sciences Center located near Parlier, CA (36°36′ N; 119°31′ W). One field experiment was conducted in 2007 and another was conducted in 2008. The soil at the sites was Hanford sandy loam soil (coarse-loamy, mixed, thermic Typic Xerorthents) and had a low organic matter content (1.38% for 0-20 cm and 0.24% for 20-100 cm of soil depth). For both experiments, previously resident orchard trees (plum on Nemaguard rootstock) were removed at least 3 months before soil fumigation, which occurred in late October, and the soil was cross ripped to a 90 cm depth. Within the 3 months before soil fumigation, replicate plots within the experimental fields were randomly assigned to two treatments: (1) a fallow treatment maintained without cover crop and (2) a shortterm rotation with a Sudan grass (Sorghum bicolor var. sudanese) cover crop. There were five replicate plots per treatment in 2007 and three replicate plots per treatment in 2008. The plots measured 22 × 6 m and were arranged in complete blocks. Seeds for the cover crop treatment were planted in July after tree removal and were irrigated to meet evapotranspiration demand. In late September, the grass was mechanically chopped to less than 10 cm pieces and incorporated into the soil with roto tillage to approximately a 15 cm depth at 2 weeks before fumigation treatments were applied. After soil incorporation, there were no visible plant residues at the soil surface. Fumigant Application. In 2007, spot drip fumigation was applied by “chemigation” to the replicate plots, and the treatment was administered through drip emitters (rated at 3.78 L/h [1 gal/h]) installed on polyethylene tubing at 3.6 m (12 ft) intervals (see Figure S1in the Suporting Information). The tubing was aligned with future tree rows such that each emitter was positioned at the site where a tree was to be planted in the orchard to follow. The rows were spaced 6.1 m (20 ft) apart. Each emitter was connected to a 6 mm diameter tube that was inserted into the soil, through a predrilled pilot hole, to deliver the fumigant to a depth of 51 cm (20 in) below the soil surface. Fumigant InLine (60.8% 1,3-D, 33.3% CP, and 5.9% inert ingredients) was applied on 24 October 2007, and the chemigation lasted from 1410 to 2040 h during which 5.6 kg InLine was injected in the irrigation stream with a total of 1544 L of water delivered. An additional 250 5784

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 43, NO. 15, 2009

L of water was applied from 2040 to 2230 h, without the fumigant, to flush the system for residual fumigant chemicals. The targeted fumigation rate was 616 kg ha-1 (550 lb acre-1) over a 152 cm (5 feet) diameter tree site which occupied 8% of the total orchard area. Drip spot fumigation was applied similarly in 2008 as in 2007, except that there was 5.5 m (18 ft) between rows in 2008. Fumigant InLine was applied on 22 October 2008, and the chemigation lasted from 1100 to 1702 h during which 4.0 kg of InLine was injected in the irrigation stream with a total of 720 L of water delivered. An additional 95 L of water was applied from 1702 to 1802 h, without the fumigant, to flush the system for residual fumigant chemicals. Parameters describing the fumigant application are summarized in Table S1 (see the Supporting Information). Soil Gas Measurement. To capture both the vertical and lateral dispersion dynamics when fumigants were applied as a point source (spot fumigation), four sets of soil gas probes were installed at 15, 51, 91, and 152 cm (6, 20, 36, 60 in) from the emitter along the row direction in one of the no cover crop plots. Time-domain reflectometer probes were also installed in another no cover crop plot for measuring changes in soil water content. These probes were installed prior to fumigant application in both 2007 and 2008. On each set of soil gas probes, 10 ports were located at 5, 10, 15, 20, 25, 30, 35, 45, 60, and 90 cm depths from the soil surface. The probes were built with 125 cm long aluminum pipes (3.45 cm O.D., 2.54 cm I.D.) and fitted with 9 soil gas access lines (Teflon tubing, 1.8 mm O.D., 0.71 mm I.D.) inside each pipe with one end of the tubing trimmed flush to the outside edge of the pipe, through a predrilled hole, and the other end mated at the surface with a QuickConnect (Small Parts Inc., Miami Lakes, FL). The pipes were filled with concrete to secure the tubing. The deepest sampling port on these soil gas probes was only 60 cm from the soil surface; therefore, an additional single stainless steel probe was used to extend the sampling depth to 90 cm. The probe was made from 3 sets of stainless steel tubing of 1.6, 3.2, and 6.4 mm O.D. with smaller ones fitted inside larger ones (to provide physical support) and the two ends soldered together exposing the ends of the smallest tubing for sampling. Teflon tubing was used to connect the sampling end of the 90 cm probe to the 10 port QuickConnect. A 10 port soil gas sampler was used for simultaneously withdrawing soil gas samples from the 10 Teflon tubing connected to the sampling probes. To capture sufficient 1,3-D and CP in the soil gas during each sampling event, 50 mL of gas was drawn from each port through polymer-based ORBO 613, XAD 4 80/40 mg tubes (Supelco Inc., Bellefonte, PA). The sample tubes were stored in an ice chest prior to and during transport to the laboratory for analysis. To adequately document fumigant dispersion over time, 10 sampling events were made at various elapsed times after fumigant application. For brevity, six sampling events from each experiment are presented to illustrate the dynamic nature of fumigant gas dispersion in these two spot fumigation experiments. Emission Measurement. Before starting the two year field study, active or flow-through emission flux chambers were constructed using a design similar to that of Gao (15). Briefly, the physical body of the chambers was made from sheet metal and manufactured at a commercial machine shop. Laminar air flow inside the chamber was achieved by a gradual expansion and contraction assisted with five separation baffles on both the inlet and the outlet section of the chamber. The sampling zone was from a 25 × 51 cm (10 × 20 in) opening located between the inlet and outlet transition section of the chamber. Air flow was created by connecting the outlet end of each chamber to a vacuum source (Series 90S ShopVac, Shop-Vac Corp., Williamsport, PA) to draw uncontaminated air into the inlet of the chamber from 3 m

FIGURE 1. Soil gas concentrations of total (cis + trans isomers) 1,3-dichloropropene (1,3-D) and chloropicrin (CP) in 2007 at different elapsed times after fumigant injection and at different radial distances (r) from the injection emitter. No cover crop treatment. Note the differences in scales for concentration among graphs. above the soil surface. To prevent/minimize artificial convective emission, the air flow rate was adjusted to 5 L min-1, and the measurable vacuum buildup in each chamber was less than 1 mm of water height. Subsampling from the total chamber flow was made from a sampling port inserted in the main flow at the chamber outlet, and the flow rate was maintained at 100 mL min-1. To increase reflectance of incident solar radiation and reduce solar heating of the chamber during field applications, the chambers were painted white. A programmable electronic solenoid control system in each chamber enabled automated switching between different sampling events at predetermined (programmed) set times. Electronic flow meters were used to monitor and record (with dataloggers) instantaneous air flow rates of both the main chamber flow and the subsampling flow. To measure 1,3-D and CP emissions, the flux chambers were placed on the soil surface next to the emitter tubing (leading to the subsurface outlet) covering a radial distance of 0-51 cm from the center of the fumigated area at each tree site. A total of 6 chambers was used each year, 3 for the cover crop treatment, 3 for the no cover crop treatment. Air was drawn through the XAD tubes, as in the soil gas measurement, to adsorb 1,3-D and CP from the sampling stream, and the tubes were replaced at different frequencies. Because the emission flux would decrease over time, emission

samples were collected every 3 h in the first 5 days, at 4 h intervals from day 5 to 7 and at 6 h intervals for the remainder of the experiment. The flux chambers were running continuously, and the samples represent cumulative residues for the specified time interval. To minimize risks of fumigant breakthrough, 2 XAD tubes were used in series during the first 3 days of each experiment. The sample tubes were stored in an ice chest prior to and during transport to the laboratory and stored at -80 °C prior to analysis. Chemical Analysis. Fumigant was extracted by breaking the XAD tubes in half, transferring all contents (including resin, glass pieces, and foam plugs) into 10 mL GC headspace vials, adding 5 mL of hexane to each sample, and sealing the vials with aluminum crimp caps fitted with Teflon-faced butyl-rubber septa. The sample vials then underwent 1 h of shaking on a reciprocating shaker at 120 strokes min-1. After settling for 2 h, the hexane extract supernatant was transferred to 2 mL amber GC vials and stored in a freezer (-18 °C) until analysis. Concentrations of 1,3-D and CP in fumigant emission and soil air samples were quantified with a micro electron capture detector on a GC (Agilent 7890A, 5975C-MSD, Wilmington, DE) using a 30 m long × 0.25 mm inside diameter DB VRX GC fused silica column with a 1.4 µm coating (J and W Scientific, Folsom, CA). The micro electron capture detector temperature was 300 °C, and nitrogen was the VOL. 43, NO. 15, 2009 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

5785

FIGURE 2. Soil gas concentrations of total (cis + trans isomers) 1,3-dichloropropene (1,3-D) and chloropicrin (CP) in 2008 at different elapsed times after fumigant injection and at different radial distances (r) from the injection emitter. No cover crop treatment. Note the differences in scales for concentration among graphs. makeup gas with a constant flow of 30 mL min-1. Helium was used as the carrier gas with a constant flow of 2.0 mL min-1. The oven temperature program was 65 °C (without hold) to 85 °C at 2.5 °C min-1. The injection port was at 140 °C in split mode, and the split ratio was 100:1 or 10:1, depending on fumigant concentrations (e.g., low concentrations at the 10:1 ratio). Injection volume was 1 µL, and sample injection was done by an automatic liquid sampler (Agilent 7683B, Wilmington, DE). Under these conditions, the retention time of the cis- and trans-isomer of 1,3-D and of CP was 5.31, 5.95, and 6.74 min, respectively. Quantification of fumigant concentrations was made by comparing the area under respective peaks with that of the standards. The detection limit was 0.01 mg L-1 for all 3 chemicals when an injection volume of 1 µL of solution was used.

Results and Discussion Fumigant Gas Dispersion in Soil. In 2007, soil gas concentrations of total 1,3-D (sum of cis- and trans-isomers) and CP reached a maximum of 16.7 and 9.5 µg cm-3 at the 45 cm depth at a lateral distance of 15 cm from the emitter at 11 h after application (Figure 1). A reasonable explanation for the increase in gas concentrations as time of sampling increased from 2 to 11 h after treatment is that the time was required for diffusion of the gases from the application point to the 5786

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 43, NO. 15, 2009

15 cm distance. After 11 h, a gradual decrease in concentrations of the gases was observed, which followed a typical breakthrough process when the observation point was fixed at one location (e.g., a 15 cm radial distance from the emitter). The magnitude of 1,3-D and CP gas concentrations and their temporal trends were comparable to values reported in the literature where fumigants were applied by broadcast shank injection (10, 16). However, unlike situations in broadcast shank injection where targeted lateral distribution of fumigant gases would be independent of location, both 1,3-D and CP gas concentrations decreased rapidly over lateral distances from the emitter. At the radial distance of 51 cm from the emitter, maximum 1,3-D and CP gas concentrations were less than 1.0 and 0.5 µg cm-3, respectively. Even lower concentrations were measured for radial distances of 91 and 152 cm from the emitter. Gas diffusion (beyond the wetting front) was likely the transport mechanism for 1,3-D and CP to move 91 and 152 cm from the injection point. Within 51 cm, 1,3-D and CP was convectively transported with water, which was confirmed by increases in soil moisture content at this distance (a 60% increase in moisture at 30 cm lateral distance from the source, only a 16% increase at 61 cm). In 2008, spatial and temporal distribution patterns of soil gas-phase 1,3-D and CP were similar to that found in the

FIGURE 3. Emission flux density of total (cis + trans isomers) 1,3-dichloropropene (1,3-D) and chloropicrin (CP) in 2007 at 0-51 cm from the drip emitter measured with active flow chambers. Error bars are standard errors (n ) 3). Note the differences in scales for emission flux between the cover crop treatment. 2007 experiment. Total 1,3-D and CP reached a maximum of 18.9 and 13.8 µg cm-3 at the 45 cm depth at a lateral distance of 15 cm from the emitter at 16 h after application (Figure 2). Both 1,3-D and CP gas concentrations decreased rapidly over lateral distances from the emitter; the maximum 1,3-D and CP gas concentrations were less than 1.0 and 0.5 µg cm-3 at 51 cm from the emitter and less than 0.1 µg cm-3 for radial distances more than 91 cm. Compared to the 2007 data, 1,3-D and CP concentrations at a 90 cm depth were much lower. This was likely caused by reduced convective downward transport with water because only 23 L of water per tree site was applied in 2008 compared to 36 L applied in 2007 (Table S1 in the Supporting Information). Fumigant solute transport modeling may be useful in predicting the distribution of fumigant concentrations as a function of the amount of water applied. Efficacy of soil-borne pest control was inferred by the concentration-time dosage. This was achieved by integrating the concentration values to concentration-time products expressed as total (sum of gas, liquid, solid or adsorbed phases) 1,3-D and CP per kilogram of dry soil within 1 h. In making the conversion, measured average soil bulk density and soil moisture content and literature values of Henry’s constants and adsorption coefficients (17-20) were used. The computed 1,3-D plus CP and 1,3-D concentration-time products were summarized in Figures S2 and S3 (in the Supporting Information), respectively. A laboratory incubation dose response study using InLine (21), after unit conversion, showed that at 20 °C the 90% lethal dose for 1,3-D plus CP was 1743-58 356 mg kg-1 h for controlling a range of weed seeds (Portulaca oleracea, Malvaparviflora,

and Erodium cicutarium) and 138-8176 mg kg-1 h for controlling fungal and oomycete pathogens of plants (Pythium ultimum and Verticillium dahliae). An earlier laboratory study (22) on nematodes showed that a lethal 1,3-D concentration time of 170 mg kg-1 h was required for controlling citrus nematodes (Tylenchulus semipenetrans). When the values in Figures 2S and 3S (see the Supporting Information) are compared, it is apparent that sufficient pest control can be expected in a sphere, centered at the subsurface emitter outlet (51 cm depth), within a radius of 15 cm. Pest control is marginal at less than a 20 cm depth but is effective at depths >20 cm for weeds and some fungal pathogens at the 15 cm radius and for nematodes at the 51 cm radius. A possible method to improve pest control in the surface zone is to cover the treated tree sites with low-permeable plastic films to retain the fumigant gases under the soil surface. Both 2007 and 2008 field sites have been replanted with peach trees as a bioassay for effects of the treatments on pathogen control and tree growth. Results will be reported in the biological literature after substantial tree growth. Fumigant Emission to Atmosphere. In 2007, 1,3-D emissions in the cover crop treatment slowly increased to a maximum of 20 µg m-2 s-1 by 57 h after application (Figure 3). The emission flux density represents an average over a 51 cm radius from the emitter. The emission remained 10-20 µg m-2 s-1 and exhibited a gradual decrease starting 136 h after application. The slow and persistent emission in 2007 was likely caused, at least partially, by the precipitation events that occurred on October 27th and 29th (relatively early in the experiment) for a cumulative of 3.4 mm of rain and lower air temperatures afterward, while in 2008 rainfall occurred VOL. 43, NO. 15, 2009 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

5787

FIGURE 4. Emission flux density of total (cis + trans isomers) 1,3-dichloropropene (1,3-D) and chloropicrin (CP) in 2008 at 0-51 cm from the drip emitter measured with active flow chambers. Error bars are standard errors (n ) 3). Note the differences in scales for emission flux between the cover crop treatment. after October 31 (Figure S4 in the Supporting Information). The daily maximum solar radiation was also lower in 2007 than that in 2008 during the first week of the experiments (Figure S4 in the Supporting Information). Lower solar heating likely implies lower evaporative demand or reduced soil gas and vapor emissions. The relatively large error bars, accounting for differences between the 3 replicated chamber measurements, were likely attributed to variations in emitter and flux chamber placement and soil properties between the 3 measurement locations. Overall, CP emissions were extremely low with a maximum flux density of 0.5 µg m-2 s-1 which occurred at 12 h after application. The total (1,3-D and CP) emission loss by the end of the emission measurement (231 h) accounted for 23% of the applied active ingredient in InLine. The cumulative emission mass loss was calculated by extrapolating the rectangular chamber area to a circular area of 51 cm in radius. Potential biases remain for not counting emissions beyond the 51 cm radius, however, emissions should have fallen off rapidly from beyond the 51 cm radius, as can be deduced from the soil gas measurements (Figures 1 and 2). In the no cover crop treatment, both 1,3-D and CP emissions remained low and the maximum flux density was only 2.9 and 0.4 µg m-2 s-1 for 1,3-D and CP, respectively. Equivalent cumulative emissions by the end of the emission measurement accounted for 2.4% of the applied active ingredient. In 2008, both 1,3-D and CP emissions in the cover crop treatment rapidly reached a maximum of 46 and 11 µg m-2 s-1 only 7.5 h after chemigation ended (Figure 4). The emissions gradually decreased over time, and 1,3-D emissions exhibited a diurnal fluctuation: high emissions in the day and low emissions at night, a phenomenon frequently 5788

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 43, NO. 15, 2009

observed in fumigant emissions under field conditions (23, 24). Similar to the results in the 2007 experiment, CP emissions were consistently lower than those of 1,3-D. Cumulative emissions by the end of the measurements (334 h) accounted for 18% of the applied active ingredient. In the no cover crop treatment, 1,3-D and CP emissions gradually increased to an average maximum of 16 and 4.3 µg m-2 s-1 at 73.5 h after application. Cumulative emissions by 334 h accounted for 6% of the applied active ingredient of InLine. Consistent in both 2007 and 2008, the Sudan grass cover crop pretreatment lead to increased fumigant emissions. While a mechanism for this was not verified in this study, it is likely that the incorporated plant residue (root systems and foliage) increased soil porosity, which in turn enhanced fumigant gas diffusion to the soil surface. The long-term beneficial effect on improving soil physical characteristics such as porosity and infiltration may not be applicable for soils that require immediate fumigation treatment before replanting a new orchard. Compared to 2008, a limitation with the final cumulative emission assessment in 2007 was the ending time of emission measurement where significant flux values were still being recorded (Figure 3). Therefore, it was most likely that the “actual” final cumulative emission exceeded 23% of the applied active ingredient in InLine. Even so, these cumulative totals were comparable to reported values (10-34% of applied) when Telone C35 (61.1% 1,3-D, 34.7% CP, 4.2% inert), a similar formulation to InLine, was applied by broadcast shank injection and the soil surface was treated with a number of emission reduction techniques (9, 10). A significant contribution of the spot fumigation, as compared to the broadcast shank injection, is the dramatic

reduction in total quantity of fumigant chemicals applied to an overall land area. As shown in Table S1 (see the Supporting Information), the equivalent area-based dosage for spot fumigation was 51 and 56 kg ha-1 for 2007 and 2008, whereas a conventional broadcast application would be about 10 times higher. A unique aspect about the low amount of fumigants needed is the support in addressing the township cap limitations in SJV to protect and improve the regional air quality. The township caps limit fumigant use to less than 444 kg km-2 yr-1 (90 250 lbs per 36 sq mile township per year) (25). Furthermore, given the high cost of soil fumigation and the dynamic nature of the regulations, it is important for growers to get a maximum benefit from the amount of fumigant used. If sufficient disease control can be achieved, which was very possible within a sphere of 15 cm in radius from the point of fumigant injection, the method of applying a very small amount of fumigants at only the tree sites would be a viable option for controlling Prunus replant problems and significantly reducing fumigant emissions to protect air quality. Therefore, spot fumigation should be considered by regulators and growers as a practical and economical alternative, among other techniques for reducing fumigant emissions, such as plastic film cover, surface organic amendment, or applying water.

Acknowledgments The authors would like to thank Jim Gartung, Tom Pflaum, Robert Shenk, Allison Kenyon, Nancy Goodell, Curtis Koga, Ricardo Zapien, Matt Gonzales, Stella Zambrzuski, Ashley Torres, Christine Rainbolt, and Patricia Mungur for various aspects in fumigant application, sample collection, and chemical analysis. We also want to acknowledge TriCal Inc. for donating the InLine chemicals and partial grant support from the USDA-ARS Pacific Area-wide Methyl Bromide Alternatives Program.

Supporting Information Available Fumigant application summary (Table S1); schematic of spot fumigation (Figure S1); concentration-time products of 1,3-dichloropropene and chloropicrin (Figure S2); concentration-time products of 1,3-dichloropropene (Figure S3); hourly environmental variables (Figure S4). This material is available free of charge via the Internet at http:// pubs.acs.org.

Literature Cited (1) Browne, G.; Connell, J. H.; Schneider, S. M. Almond replant disease and its management with alternative pre-plant soil fumigation treatments and rootstocks. Plant Dis. 2006, 90, 869– 876. (2) Montzka, S. A.; Butler, J. H.; Hall, B. D.; Mondeel, D. J.; Elkins, J. W. A decline in troposheric organic bromine. Geophys. Res. Lett. 2003, 30, 13:1–4. (3) Carter, W. P.; Luo, D. M.; Malkina, I. L. Investigation of the atmospheric reactions of chloropicrin. Atmos. Environ. 1997, 31, 1425–1439. (4) U.S. EPA. Implementation of risk mitigation measures for soil fumigant pesticides. U.S. Government Printing Office: Washington, DC, 2008, http://www.epa.gov/oppsrrd1/reregistration/ soil_fumigants/ (accessed May 2, 2009). (5) Majewski, M. S.; McChesney, M. M.; Woodrow, J. E.; Prueger, J. H.; Seiber, J. N. Aerodynamic measurements of methyl bromide volatilization from tarped and nontarped fields. J. Environ. Qual. 1995, 24, 742–751.

(6) Yates, S. R.; Ernst, F. F.; Gan, J.; Gao, F.; Yates, M. V. Methyl bromide emissions from a covered field. II. Volatilization. J. Environ. Qual. 1996, 25, 192–202. (7) Wang, D.; Yates, S. R.; Ernst, F. F.; Gan, J.; Jury, W. A. Reducing methyl bromide emission with a high barrier plastic film and reduced dosage. Environ. Sci. Technol. 1997, 31, 3686–3691. (8) Gan, J.; Yates, S. R.; Becker, J. O.; Wang, D. Surface amendment of fertilizer ammonium thiosulfate to reduce methyl bromide emission from soil. Environ. Sci. Technol. 1998, 32, 2438–2441. (9) Gao, S.; Trout, T. J.; Schneider, S. Evaluation of fumigation and surface seal methods on fumigant emissions in an orchard replant field. J. Environ. Qual. 2008, 37, 369–377. (10) Yates, S. R.; Knuteson, J.; Ernst, F. F.; Zheng, W.; Wang, Q. Effect of sequential surface irrigations on field-scale emissions of 1,3dichloropropene. Environ. Sci. Technol. 2008, 42, 8753–8758. (11) Schneider, R. C.; Green, R. C.; Wolt, J. D.; Loh, R. K. H.; Schmitt, D. P.; Sipes, B. S. 1,3-dichloropropene distribution in soil when applied by drip irrigation or injection in pineapple culture. Pestic. Sci. 1995, 43, 97–105. (12) Wang, D.; Ernst, F. F.; Yates, S. R.; Knuteson, J. A. Volatilization of 1,3-dichloropropene under different application methods. Water Air Soil Pollut. 2001, 127, 109–123. (13) Ajwa, H. A.; Trout, T. Drip application of alternative fumigants to methyl bromide for strawberry production. HortScience 2004, 39, 1707–1715. (14) Browne, G.; Lampinen, B.; Holtz, B.; Upadhyaya, S.; Wang, D.; Gao, S.; Doll, D.; Schmidt, L.; Hanson, B.; Goodell, N.; McKenry, M.; Klonsky, K. Integrated pre-plant alternatives to methyl bromide for almonds and other stone fruits. Proceedings Annual International Research Conference on Methyl Bromide Alternatives and Emissions Reductions, San Diego, CA, Oct 29-Nov 1, 2007, 8 (1-3). (15) Gao, F.; Yates, S. R.; Yates, M. V.; Gan, J.; Ernst, F. F. Design, fabrication and application of a dynamic chamber for measuring gas emissions from soil. Environ. Sci. Technol. 1997, 31, 148– 153. (16) Thomas, J. E.; Allen Jr., L. H.; McCormack, L. A.; Vu, J. C.; Dickson, D. W.; Ou, L. T. Diffusion and emission of 1,3-dichloropropene in Florida sandy soil in microplots affected by soil moisture, organic matter, and plastic film. Pest Manage. Sci. 2003, 60, 390–398. (17) Leistra, M. Distribution of 1,3-dichloropropene over the phases in soil. J. Agric. Food Chem. 1970, 18, 1124–1126. (18) Wolt, J. D.; Holbrook, D. L.; Batzer, F. R.; Balcer, J. L.; Peterson, J. R. Dynamics of nematicide distribution and degradation in soil subjected to long-term pineapple culture. Acta Hortic. 1993, 334, 361–371. (19) Lee, H.; Moreno, T. Photohydrolysis of chloropicrin. Bolsa Research Associates: Hollister, CA, 1993 (unpublished study submitted to USEPA). (20) Skinner, W. Soil column leaching of 14C chloropicrin in four soil types. PTRL West, Inc.: Richmond, CA, 1996 (unpublished study submitted to USEPA). (21) Klose, S.; Ajwa, H. A.; Fennimore, S. A.; Martin, F. N.; Browne, G. T.; Subbarao, K. V. Dose response of weed seeds and soilborne pathogens to 1,3-D and chloropicrin. Crop Prot. 2007, 26, 535– 542. (22) Xue, S.; Gan, J.; Yates, S. R.; Becker, J. O. Nematode response to methyl bromide and 1,3-dichloropropene soil fumigation at different temperatures. Pest Manage. Sci. 2000, 56, 737–742. (23) Sullivan, D. A.; Holdsworth, M. T.; Hlinka, D. J. Control of offgassing rates of methyl isothiocyanate from the application of metam sodium by chemigation and shank injection. Atmos. Environ. 2004, 38, 2457–2470. (24) van Wesenbeeck, I. J.; Knuteson, J. A.; Barnekow, D. E.; Phillips, A. M. Measuring flux of soil fumigants using the aerodynamic and dynamic flux chamber methods. J. Environ. Qual. 2007, 36, 613–620. (25) Segawa, R. Descriptions of fumigation methods allowed in nonattainment areas, 2009 http://www.cdpr.ca.gov/docs/ emon/vocs/vocproj/desc_fieldfum_mthd.htm (accessed May 6, 2009).

ES9015662

VOL. 43, NO. 15, 2009 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

5789