Sources and Occurrence of Dacthalin the Canadian AtmosphereY U A N Y A O , † T O M H A R N E R , * , †

J I A N M I N M A , † L U D O V I C T U D U R I , ‡ A N DP I E R R E T T E B L A N C H A R D †

Science and Technology Branch, Environment Canada, 4905Dufferin Street, Toronto, Ontario, M3H 5T4, Canada, andLaboratoire de PhysicoToxicochimie des Systemes Naturels,Equipe Perigourdine de Chimie Appliquee, Site universitaire,24019 Perigueux cedex, France

Dacthal is a herbicide that can undergo long-rangeatmospheric transport. Due to limited use in Canada, itsoccurrence in the Canadian environment is likely associatedwith transboundary flow from the United States where100 times more dacthal is used. To investigate its atmosphericdistribution and possible sources, two sampling strategieswere applied. First, polyurethane foam-disk passive airsamplers were deployed across the country from the springto summer of 2004 and 2005. Results showed highestdacthal concentrations at two sites in southern Ontario(max: 50 pg m-3) and much lower concentrations in otheragricultural regions across Canada. Second, daily high-volume air samples were collected at a field site in northToronto (Downsview). Sampling at this site was triggeredby real-time meteorological forecast models that indicatedwhen air flow to the site originated in the United States.Two such events, one in late June and one in late September2005, were sampled. In both cases, elevated dacthalconcentrations were captured (e.g., up to 319 pg m-3 inthe June event) and they were well correlated withtransboundary flow from the United States. Finally, theoctanol-air partition coefficient (KOA) of dacthal and othercurrent-use pesticides was measured. KOA was used toderive a particle-gas partitioning coefficient (Kp) for dacthal(Log Kp ) -4.1), indicating that this compound shouldexist mainly in the gas phase.

IntroductionDacthal (dimethyl 2,3,5,6-tetrachloroterephthalate), alsoknown as DCPA or chlorthal-dimethyl, is a potent chlorinatedbenzoic acid herbicide. It was first registered as a pesticidein the United States in 1958 as a selective pre-emergenceherbicide for weed control on turf grasses. Dacthal killsgerminating seeds by disrupting microtubule formation inexposed cells and causing abnormal cell division (1). Cur-rently, dacthal is primarily used for vegetable production.Approximately 271 t of dacthal are applied to crops in theUnited States every year, of which 39% is used on onions,28% on broccoli, and the remaining 33% on other types ofvegetables, fruits, and sod (2). In Canada, the annual sales/use quantity of dacthal is less than 1% of U.S. use and about821 kg (2003), of which 453 and 368 kg are utilized in Ontario

and New Brunswick, respectively (3). Despite this moderateuse level in North America compared to some otherpesticides, dacthal has been widely reported in differentenvironmental compartmentssa reflection of its mobilityand persistence.

Dacthal is moderately hydrophobic and persistent. It isslowly hydrolyzed in soils with a half-life between 50 and 100days in most soils (4). It has been widely detected in surfacewater under the U.S. Geological Survey’s National Water-Quality Assessment (NAWQA) program (5) and its di-acidmetabolite, 2,3,5,6-tetrachloroterephthalic acid (TCPA) is themost commonly detected pesticide residue in groundwaterin the United States. Concentrations in fish have beenreported from over 30 states in the United States (6). Dacthalhas a higher vapor pressure (3.33 × 10-4 Pa at 25-30 °C)compared to p,p′-DDT (9.60 × 10-5 Pa at 25-30 °C) (7). Itsvolatilization from soil has been reported to continue up to60 days after application with ∼10% lost to the atmospherein the first 21 days after application (4). James and Hites (8)measured dacthal in air and precipitation samples collectedaround the Great Lakes. Its detection in high-elevation snowsamples from Rocky Mountain National Park, CO (up to 1058pg L-1) is evidence of its ability to undergo long-rangeatmospheric transport (9). Further evidence comes from thework of Rawn and Muir (10) who reported dacthal in a smallagricultural watershed in southern Manitoba, Canada, whereno dacthal was applied, and linked this to transboundaryflow from the United States. Muir et al. (11) assessed theatmospheric transport potential of dacthal compared to othercurrent-use pesticides (including R-endosulfan, alachlor,metolachlor, chlorpyrifos, atrazine, diazinon, and disulfoton).From this group, dacthal had the highest atmospheric half-distance of 3000 km and was identified as a prime candidatefor long-range atmospheric transport. Further, Ma et al. (12)have shown that certain meteorological conditions can resultin substantial high-concentration transport episodes ofpesticides from the United States to Canada that can lastseveral days.

A chemical’s potential to undergo long-range transportdepends on its volatility and persistence in air. The octanol-air partition coefficient, KOA, is recognized as a key parameterdescribing a chemical’s ability to partition between theatmosphere and environmental organic phases such as soil,vegetation, and atmospheric particulate matter. Shoeib andHarner (13) have reported KOA and its temperature depen-dence for several organochlorine pesticides, however no dataare available for dacthal. Freshwater fish acute toxicity studiesusing rainbow trout and bluegill sunfish indicated that dacthalmay have the potential to be slightly toxic to fish (14); it wasthe only compound analyzed for in the Upper Colorado RiverBasin (UCOL) study under the NAWQA program that has notbeen banned from use (15). In addition, the fact that dacthalcan be contaminated by two highly toxic impurities, 2,3,7,8-tetrachlorodibenzo-p-dioxin (2,3,7,8-TCDD) and hexachlo-robenzene (HCB), is of concern (16). Klopman et al. (17)pointed out that the observed rodent carcinogenicity ofdacthal is likely related to its impurities.

This paper addresses questions regarding the occurrence,transport, and phase partitioning of dacthal in the Canadianatmosphere. First, measurements of KOA are made andcompared to some other current-use pesticides and thenused to evaluate partitioning of dacthal to atmosphericparticles. Second, results from a national passive air samplingeffort are used to evaluate the occurrence of dacthal in theCanadian atmosphere. Finally, attempts are made to captureindividual pesticide transport events by coupling daily high-

* Corresponding author phone: + 1 416 739 4837; fax: + 1 416739 4281; e-mail: tom.harner@ec.gc.ca.

† Environment Canada.‡ Laboratoire de PhysicoToxicochimie des Systemes Naturels.

Environ. Sci. Technol. 2007, 41, 688-694

688 9 ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 41, NO. 3, 2007 10.1021/es061725r CCC: $37.00 2007 American Chemical SocietyPublished on Web 12/16/2006

volume air samples with a forecast sampling approach basedon predictions from meteorological models (12). This noveltechnique will allow us to evaluate transboundary transportof dacthal into Canada.

Experimental SectionMeasurement of Octanol-Air Partition Coefficient (KOA).The KOA values for dacthal, chlorothalonil, chlorpyrifos, andtrifluralin were measured at 20 °C using an establishedgenerator column method (13, 18). The other three currentlyused pesticides (CUPs) were chosen due to the strong needto assess their long-range transport (LRT) potentials, andbecause these agrochemicals have been found to exist withhigh concentrations in the agricultural regions across Canada(19). To briefly summarize the method, octanol-saturatedair was passed through a generator column consisting ofglass beads coated with an octanol solution containingdacthal and the other target compounds. These chemicalsvolatized to establish equilibrium concentration in air andwere trapped on C18 adsorbent. Chemicals were eluted fromthe C18 adsorbent with 10 mL of a solvent mixture ofdichloromethane/hexane (50/50, v/v), solvent exchanged intoisooctane, and mirex was added as internal standard for gaschromatography (GC)/mass spectrometry (MS) analysis,described later. Three to five replicate measurements wereconducted. Each pesticide standard (98%) was purchasedfrom AccuStandard (New Haven, CT). Deuterated γ-HCH(d6-γ-HCH) was obtained from Cambridge Isotope Labora-tories (Andover, MA).

Air Sampling. Polyurethane foam (PUF)-disk samplershave been applied successfully in numerous studies toprovide time-integrated concentrations of pesticides andother semi-volatile organic chemicals (20-22). In this study,passive air samples were collected as part of the CanadianAtmospheric Network for Currently Used Pesticides (CAN-CUP) program with the aim of assessing levels of CUPs inthe Canadian Atmosphere (19). PUF-disk samplers weredeployed at agricultural, urban, and receptor regions acrossCanada from the spring to summer of 2004 and 2005 withdeployment periods ranging from 1 to 3 months. For the firstcampaign in 2004, samples were collected at 4 sites: Ab-botsford, BC; Egbert, ON; Vineland, ON; and Kensington,PEI. To increase the resolution of the spatial pattern, 4 moresites were added in 2005: Bratt’s Lake, SK; Downsview, ON;St. Anicet, QC; and Baie St. Francois, QC. Details of thesampling site and periods are given in Table 1.

Episodes of air masses from the southern United Statesto a field site, Downsview, in north Toronto, Canada wereanalyzed from forecasted atmospheric circulation patternsby various numerical weather forecasting models availableat http://weather.unisys.com/. These models produce pre-dicted 6 h-10 d wind vectors, air temperature, and geo-potential height from 1000 to 300 hPa, as well as surface

pressure and precipitation. When the numerical weathermodels forecasted atmospheric circulation patterns to beanalogous to the transport episodes reported by Ma et al.(12), it triggered daily sample collection prior to, during, andafter the event. A high-volume air sampler (PS-1, TischEnvironmental, Inc., Village of Cleves, OH) was employed tocollect these daily samples at a flow rate of approximately200 L min-1 (ca. 280 m3 sample volume). Particles weretrapped with glass fiber filters (102 mm diameter; Pall LifeSciences, Ann Arbor, MI) and gas-phase compounds werecollected with cartridges containing 10 g of XAD-2 resin(Supelpak 2, Supelco, Bellefonte, PA) sandwiched betweenPUF plugs (75 mm × 37 mm, Supelco). All samples werestored at -20 °C and in the dark until extraction. One fieldblank was collected for each sampling period. Two transportepisodes were sampled: June 22-29, 2005 and September21-30, 2005.

Sample Analysis. After adding d6-γ-HCH as a surrogatestandard, high-volume (PUF/XAD/filter) and passive (PUFdisk) samples were Soxhlet extracted with 400 mL of hexane/acetone (50/50 v/v) for 24 h. The extracts were reduced involume by rotary evaporation and exchanged into hexane (1mL). The extracts were then cleaned up using alumina columnchromatography. The alumina column was packed with 1.5g of 6% (w/w) water-deactivated alumina (Aluminum Oxide90, Active neutral, EMD, Germany) and topped with anhy-drous sodium sulfate. After pre-washing with 10 mL of 5%dichloromethane (DCM) in petroleum ether twice, theapplication of sample extract was conducted, and then thecolumn was eluted with 15 mL of 5% DCM in petroleumether. The eluent was concentrated and solvent was ex-changed into isooctane (1 mL). 13C-PCB-105 (24 ng) andmirex (100 ng) were added as internal standards for volumecorrection prior to GC/MS analysis. A 2 µL aliquot of eachconcentrate was injected into a GC (6890, Agilent Technolo-gies, USA)/MS (5973 MSD, Agilent Technologies, USA) systemin splitless mode using an autosampler (7683, AgilentTechnologies, USA). Negative chemical ionization (NCI)mode was used. Helium and methane were utilized as carrierand reagent gases. A DB-5MS capillary column (60 m length× 0.25 mm i.d., 0.25 µm film thickness, J&W Scientific, USA)was employed. The temperature program started at 90 °C for1 min, 15 °C min-1 to 240 °C, 6 °C min-1 to 270 °C, then 25°C min-1 to 290 °C and hold for 6 min. The inlet, transfer line,ion source, and quadrupole temperatures were 265, 250, 150,and 106 °C, respectively. Ions monitored were target/qualifier: dacthal 332/330, d6-γ-HCH 261/263, 13C-PCB-105338/336, and mirex 404. The surrogate recovery ranged from72% to 94% with an average of 82% (n ) 16). The methodrecovery for dacthal is 84% (n ) 8, SD ) 4%). Field blankswere investigated and the detected quantities were less than1% of those measured in samples. No recovery or blankcorrection was applied to the results.

TABLE 1. Description of the Passive Air Sampling Sites and Sampling Periods in 2004 and 2005

sampling site latitude longitude description sampling period

Kensington, PEI 46°25′ N 63°37′ W heart of potato growing country. Jun. 24-Sep. 28, 2004Aug. 8-Sep. 5, 2005

St. Anicet, QC 45°07′ N 74°17′ W rural and agricultural area (corn, pasture land). Jun. 1-Jun. 29, 2005Baie St. Francois, QC 46°05′ N 72°56′ W wetland covered with mixed vegetation. receptor site. Jun. 1-Jun. 29, 2005Egbert, ON 44°14′ N 79°47′ W rural and suburban area, surrounded by fields and

mixed forest.May 17-Aug. 11, 2004May 12-Jul. 7, 2005

Vineland, ON 43°11′ N 79°24′ W intensive agricultural area (fruit, vegetables, wine). May 21-Jul. 23, 2004Jun. 3-Aug. 5, 2005

Downsview, ON 43°40′ N 79°37′ W residential and industrial area Jun. 16-Aug. 11, 2005Bratts Lake, SK 50°16′ N 104°42′ W rural and intensive agricultural area (cereals) with

few trees.May 11-Aug. 3, 2005

Abbotsford, BC 49°01′ N 122°20′ W Fraser Valley, pig and chicken farms and berry crops. Apr. 21-Jun. 1, 2004May 4-Jun. 8, 2005

VOL. 41, NO. 3, 2007 / ENVIRONMENTAL SCIENCE & TECHNOLOGY 9 689

Results and DiscussionKOA and Particle-Gas Partitioning. The KOA values for ourtarget compounds are summarized in Table 2. Measured KOA

values are compared to those calculated as KOA ) KOWRT/H,where R is the ideal gas constant (8.314 J mol-1 K-1), T is theabsolute temperature (K), and H is Henry’s law constant (Pam3 mol-1). The KOW and H values from Mackay et al. (23)were used, except H data for chlorothalonil and dacthal thatwere taken from James and Hites (8) and Michigan Depart-ment of Environmental Quality’s document (24), respectively.

The measured Log KOA value (at 20 °C) for dacthal is 8.51.This is of the same order of magnitude as values measuredfor other CUPsschlorothalonil (8.11), chlorpyrifos (8.88), andtrifluralin (7.93)sand demonstrates the semi-volatile natureof these CUPs. These KOA values are also of the same orderof magnitude as other organochlorine pesticides that areknown for their atmospheric transport potential (Table 2).For instance, HCB (7.38); lindane (7.85); and R-endosulfan(8.64) (13). The value of 8.51 for dacthal compares well withthe calculated value of 8.45 and a value estimated by Muiret al. (11) at 25 °C of 8.28. The same is true for chlorpyrifos(8.88), for which Muir et al. (11) previously estimated a value

of 8.75. Log KOA values typically increase by about 0.48 logunits (factor of 3) for every 10 °C decrease in temperature(13).

It is useful to assess the extent to which pesticides willpartition to atmospheric particulate matter. This will help tointerpret air sampling data which are sometimes separatedinto gas- and particle-phase fractions. Particle-gas parti-tioning also plays a role in the long-range transport potentialof a chemical. Chemicals that are less associated with particleswill generally have a greater potential for transport althoughchemical reactivity and water solubility will also factor intothe calculation (11). Similar to the analysis by Shoeib et al.(13), partitioning to atmospheric particulate matter wasevaluated using the KOA-based model as follows:

where Kp is particle-gas partition coefficient; fom is thefraction of organic matter on particle (default of 0.2 was used);TSP is the concentration of suspended particulate matter(25 µg m-3), and φ is the fraction on particles.

TABLE 2. Octanol-Air Partition Coefficient (KOA) Values and Predicted Particle Percentages (O) of Selected Current-UsePesticides

chemical CAS no.

calculated LogKOA(this study)

(25 °C)

measured LogKOA(this study)

(20 °C)

O(this study)(%, 20 °C)

Dacthal 1861-32-1 8.45 8.51 0.20Chlorpyrifos 2921-88-2 8.28 8.88 0.46Chlorothalonil 1897-45-6 7.69 8.11 0.08Trifluralin 1582-09-8 7.73 7.93 0.05

chemical CAS no.

calculated LogKOA(previous study)a

(25 °C)

measured LogKOA(previous study)a

(25 °C)

O(previous study)a

(%, 25 °C)

HCB 118-74-1 6.78 7.38 0.01Lindane 58-89-9 9.74 7.85 0.04R-Endosulfan 959-98-8 6.00 8.64 0.27Heptachlor 76-44-8 6.12 7.64 0.03trans-Chlordane 5103-74-2 9.98 8.87 0.46cis-Chlordane 5103-71-9 9.86 8.92 0.51Aldrin 309-00-2 4.44 8.08 0.07Dieldrin 60-57-1 8.54 8.90 0.48Endrin 72-20-8 10.1 8.13 0.08

a Values from Shoeib and Harner (13), in which recommended KOW and H values in Mackay et al. (21) were used for KOA ) KOWRT/H calculation.

TABLE 3. Concentrations (pg m-3) of Dacthal in Passive Air Samples Collected in 2004 and 2005

sampling site Abbotsford Egbert Vineland Kensington

date on 21-Apr-04 21-Apr-04 17-May-04 17-May-04 21-May-04 25-Jun-04 24-Jun-04 24-Jun-04date off 1-Jun-04 1-Jun-04 11-Aug-04 11-Aug-04 25-Jun-04 23-Jul-04 28-Sep-04 28-Sep-04conc. 6.0 3.6 8.7 12.8 29.0 28.2 2.5 2.1mean conc. 4.8 10.8 28.6 2.3

sampling site Abbotsford Bratt's Lake Egbert Vineland

date on 4-May-05 4-May-05 11-May-05 11-May-05 12-May-05 12-May-05 3-Jun-05 30-Jun-05date off 8-Jun-05 8-Jun-05 3-Aug-05 3-Aug-05 7-Jul-05 7-Jul-05 5-Aug-05 5-Aug-05conc. 12.7 13.4 16.2 17.6 32.0 32.7 45.8 37.0mean conc. 13.1 16.9 32.4 41.4

sampling site Downsview St. Anicet Baie St. Francois Kensington

date on 16-Jun-05 16-Jun-05 1-Jun-05 1-Jun-05 1-Jun-05 1-Jun-05 8-Aug-05 8-Aug-05date off 11-Aug-05 11-Aug-05 29-Jun-05 29-Jun-05 29-Jun-05 29-Jun-05 5-Sep-05 5-Sep-05conc. 48.3 49.8 13.7 23.7 11.8 6.0 0.4 0.6mean conc. 49.1 18.7 8.9 0.5

Log Kp ) Log KOA + log fom - 11.91 (1)

φ ) KpTSP/(KpTSP + 1) (2)

690 9 ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 41, NO. 3, 2007

Particle percentages for dacthal and other CUPs are listedin Table 2. At 20 °C, dacthal is expected to be mainly in thegas phase, with only 0.20% on particles. It is noteworthy thatJames and Hites (8) reported dacthal in some particle samplescollected at U.S. Integrated Atmospheric Deposition Network(IADN) sites (Sturgeon Point and Eagle Harbor) in 1996, inthe range of 2-10 pg m-3.

Dacthal Spatial Distribution. As determined here, dacthalhas a relatively high KOA value, which allows linear uptakeduring the entire passive sampling period. For chemicalswith lower KOA values (less than 108), the PUF disk mayequilibrate with the air during deployment resulting in anonlinear phase. A linear air sampling rate of 4 m3 day-1 wasused to convert the amount of chemical collected on thePUF disk to a concentration value. This rate is based onresults from previous field deployments where depurationcompounds were used as a check of the sampling rate (25,26). Because of some site-specific variability in sampling rates,passive-sampler-derived air concentrations are believed tobe accurate within a factor of about 2 of the true value. PUF-disk samplers represent mainly the gas phase although veryfine particles will also be captured.

Dacthal was detected in all passive air samples and theresults are shown in Tables 3 and Figure 1a and b. In 2004,the measured dacthal concentrations ranged from 2.1 to 29.0pg m-3 with the highest average concentration observed at

the 2 Ontario sites (Figure 1a). Results for 2005 (Figure 1b)provided better spatial resolution and dacthal concentrationsranged from 0.4 to 49.8 pg m-3 across the country. All threeOntario sites exhibited relatively high concentrations. Thisis consistent with the fact that Ontario is the largest user ofdacthal in Canada, responsible for ∼55% of domestic use(3). The results for Ontario are comparable with totalconcentrations (gas- and particle-phases) of dacthal in airsamples (20-50 pg m-3) taken at U.S. IADN sites around theGreat Lakes including Eagle Harbor, MI; Sleeping Bear Dunes,MI; and Sturgeon Point, NY in 1996 (8).

Episodic Transport of Dacthal. Atmospheric motion inthe U.S./Canada region is essentially zonal (west to east).When a large scale meridional (south to north) air massdevelops across low and high latitude regions, this isconsidered an episodic event that may persist for up to aweek (12). Air samples were collected during two such eventsto investigate if the pesticide air burdens were elevated inthese south to north transboundary flows. As describedpreviously, real-time meteorological forecasting was used toidentify two events in June and September 2005seach lastingseveral days. Results for dacthal are summarized in Table 4and in Figure 2a and b.

June 22-29, 2005 Event. Results from seven daily high-volume air samples are shown in Figure 2a. Concentrationspeaked at 319 pg m-3 at the height of the event and were at

FIGURE 1. (a) Air concentrations of dacthal across Canada during April-September 2004 (VIN, Vineland; EGB, Egbert). (b) Air concentrationsof dacthal across Canada during May-September 2005 (DV, Downsview).

VOL. 41, NO. 3, 2007 / ENVIRONMENTAL SCIENCE & TECHNOLOGY 9 691

a minimum of 25.2 pg m-3 just prior to the event. The lowvalue at the onset of the event is consistent with the passivesampler-derived value (49.1 pg m-3) at the same site, coveringthis time period (see also Figure 1b). No clear correlationbetween dacthal level and daily average temperature wasfound (Figure 2a), indicating that the elevated air concentra-tions observed during the event are not related to temperaturevariation. To relate episodes of high concentrations to sourceregions, three-dimensional air parcel back-trajectories ar-riving at the sampling site at a 10-m altitude were determinedfor all 7 individual air samples. Trajectories (obtained fromthe Centre Meteorologique Canadien) were determined every6 h, resulting in four trajectories for a 24-h sample. The back-trajectories showed that air parcels arriving at the site for thefirst day (June 22-23) mainly came from northern Ontario.During the period of June 23-25, especially June 24-25, airflow could be traced to Ohio, adding the influence from thatsource region where an average use intensity rate wasreported to range from 0.003 to 0.008 lb mi-2 of agriculturalland during 1995-1998 (27). Trajectories for the first andthird day samples (June 22-23 and June 24-25) are shownin Figure 2a.

To further understand the characteristics of the LRT eventand track the potential atmospheric sources, measured windfields at 1500 m (∼850 hPa) were analyzed. Figure 3a (leftpanel) shows the 3-day (June 23-25) images of the regionalversion (24-km horizontal resolution) obtained from the

Global Environmental Multi-Scale Model (GEM)sCanada’snumerical weather forecasting model. A strong wind floworiginating from Texas extended to the sampling site duringthis period. This wind flow was the western branch of ananticyclone wind field centered in Missouri on June 23. Theflow pattern bears a clear resemblance to the atmosphericcirculation pattern occurring in the September 2000 tox-aphene LRT event, as revealed by Ma et al. (12). Consideringthe fact that the back-trajectories described previously onlyshowed the atmospheric transport in the relatively lowatmospheric levels (less than 600 m), our surface-samplingresults may only reflect a portion (e.g., a downdraft) of thefull magnitude of the transport episode.

Although dacthal historical application data are nonexist-ent, Miller and Gomes (28) previously measured dacthal atvery high air concentrations (900 000-8 600 000 pg m-3) inTexas air, indicating higher past usage in the United States.The current U.S. domestic dacthal use quantity is ap-proximately 40% of that (∼678 t y-1) in the mid 1990s. Rostad(29) pointed out that dacthal was widely used in the UpperMississippi River basin. Previous measurements of dacthalin fish indicate that the Mississippi River basin is heavilyimpacted by dacthal inputs (6). During a cruise along theMississippi river in June, 1994, Majewski et al. (30) measuredair concentrations of dacthal as high as 330 pg m-3. Themarked increase of dacthal at the sampling site during June24-25 was very likely subject to strong transboundary

TABLE 4. Concentrations (pg m-3) of Dacthal in High-Volume Air Samples Collected During the Episodic Events in 2005

date on 22-Jun-05 23-Jun-05 24-Jun-05 25-Jun-05 26-Jun-05 27-Jun-05 28-Jun-05date off 23-Jun-05 24-Jun-05 25-Jun-05 26-Jun-05 27-Jun-05 28-Jun-05 29-Jun-05conc. 25.2 190 319 22.7 100 115 123

date on 21-Sep-05 22-Sep-05 23-Sep-05 24-Sep-05 25-Sep-05 26-Sep-05 27-Sep-05 28-Sep-05 29-Sep-05date off 22-Sep-05 23-Sep-05 24-Sep-05 25-Sep-05 26-Sep-05 27-Sep-05 28-Sep-05 29-Sep-05 30-Sep-05conc. 20.4 22.8 5.7 11.9 30.4 12.1 10.3 6.5 14.0

FIGURE 2. (a) Time trend of dacthal air concentration at Downsview during the June 2005 event and comparison with passive samplingresult observed at the same site. (b) Time trend of dacthal air concentration at Downsview during the September 2005 event.

692 9 ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 41, NO. 3, 2007

transport from the southwestern United States to the GreatLakes region.

September 21-30, 2005 Event. Lower dacthal air concen-trations were measured during the September event, rangingfrom 5.7 to 30.4 pg m-3. The much lower concentrations ofdacthal during September are not surprising. Dacthal ismainly used as a pre-emergent herbicide. For example, itsapplication on all major crops (corn, cotton, rice, sorghum,soybeans, and spring wheat) in the southern United Statesis completed by June. Lower air and surface soil temperaturesin late September compared to late June also result in lessvolatilization of the already depleted dacthal soil burden.The observed peak air concentration is related to air massflow from the United States (Figure 2b). Interestingly, thisevent was associated with hurricane Rita, which was the most

intense tropical cyclone ever observed in the Gulf of Mexico.This hurricane formed in the Atlantic Ocean north of theequator, penetrated into the United States on September 24near the Texas-Louisiana border, and then dissipated onSeptember 26 as seen in Figure 3b (right panel). Themeteorology associated with hurricane Rita resulted in astrong southerly wind flow from the southeastern U.S./Gulfof Mexico all the way to the sampling site in Toronto.Effectively, this produced an atmospheric “conveyer belt”delivering dacthal from source regions in the United Statesto the Great Lakes basin.

ImplicationsConsidering that dacthal is a minor-use herbicide in Canada,its wide-scale atmospheric presence appears to be primarily

FIGURE 3. (a) Left panel, vector winds at 850 hPa from June 23-25 2005; (b) right panel, vector winds at 850 hPa from September 24-262005. The lower left corner legend indicates the magnitude of the wind vectors (knots). The red point shows the sampling site, and thelocation of Hurricane Rita is circled on the September 24 and 25 panels.

VOL. 41, NO. 3, 2007 / ENVIRONMENTAL SCIENCE & TECHNOLOGY 9 693

a result of its relatively high volatility, low partitioning toatmospheric particles, and high atmospheric persistenceleading to a high potential for atmospheric LRT. The presentstudy suggests that episodic transboundary input from sourceregions in the United States could be a major contributor toobserved dacthal burdens in the Canadian atmosphere. Givenits strong LRT potential, further studies are necessary to assessdacthal levels in more distant and sensitive ecosystems suchas the Arctic.

The real-time forecast sampling approach that was usedin this study is a promising technique that may help toimprove our understanding of transport episodes for pes-ticides and other chemicals that are prone to atmosphericLRT. Many current monitoring programs that target pesti-cides and other semi-volatile organics sample intermittently(not every day) and thus may not effectively capture theseepisodes.

AcknowledgmentsThe study was supported by Environment Canada’s PesticideScience Fund (PSF) and the Canadian Atmospheric Networkfor Currently Used Pesticides (CANCUP). We are grateful toMahiba Shoeib, Liisa M. M. Jantunen, and Terry F. Bidlemanof Environment Canada for valuable discussions. We alsothank Jacinthe Racine of the Centre Meteorologique Canadienfor assistance with back-trajectory analysis.

Literature Cited(1) Cox, C. DCPA (Dacthal). J. Pestic. Reform 1991, 11, 17-20.(2) National Center for Food and Agricultural Policy. National

Pesticide Use Database, 2004. http://www.ncfap.org/database/national/default.asp.

(3) Brimble, S.; Bacchus, P.; Caux, P. Y. Pesticide Utilization inCanada: A Compilation of Current Sales and Use Data;Environment Canada: Gatineau, QC, 2005.

(4) Ross, L. J.; Nicosia, S.; McChesney, M. M.; Hefner, K. L.; Gonzalez,D. A.; Seiber, J. N. Volatilization, off-site deposition, anddissipation of DCPA in the field. J. Environ. Qual. 1990, 19,715-722.

(5) Larson, S. J.; Gilliom, R. J.; Capel, P. D. Pesticides in Streams ofthe United States: Initial Results from the National Water-QualityAssessment Program; USGS: Sacramento, CA, 1999; http://ca.water.usgs.gov/pnsp/rep/wrir984222/.

(6) Schmitt, C. J.; Zajicek, J. L.; Peterman, P. H. National contaminantbiomonitoring program: Residues of organochlorine chemicalsin U.S. freshwater fish, 1976-1984. Arch. Environ. Contam.Toxicol. 1990, 19, 748-781.

(7) Woodrow, J. E.; Seiber, J. N.; Dary, C. Predicting pesticideemissions and downwind concentrations using correlations withestimated vapor pressures. J. Agric. Food Chem. 2001, 49, 3841-3846.

(8) James, R. R.; Hites, R. A. Chlorothalonil and dacthal in GreatLakes air and precipitation samples. J. Great Lakes Res. 1999,25, 406-411.

(9) Usenko, S.; Hageman, K. J.; Schmedding, D. W.; Wilson, G. R.;Simonich, S. L. Trace analysis of semivolatile organic compoundsin large volume samples of snow, lake water, and groundwater.Environ. Sci. Technol. 2005, 39, 6006-6015.

(10) Rawn, D. F. K.; Muir, D. C. G. Sources of chlorpyrifos and dacthalto a small Canadian prairie watershed. Environ. Sci. Technol.1999, 33, 3317-3323.

(11) Muir, D. C. G.; Teixeira, C.; Wania, F. Empirical and modelingevidence of regional atmospheric transport of current-usepesticides. Environ. Toxicol. Chem. 2004, 23, 2421-2432.

(12) Ma, J.; Venkatesh, S.; Li, Y.-F.; Cao, Z.; Daggupaty, S. Trackingtoxaphene in the North American Great Lakes basin. 2. A strongepisodic long-range transport event. Environ. Sci. Technol. 2005,39, 8132-8141.

(13) Shoeib, M.; Harner, T. Using measured octanol-air partitioncoefficients to explain environmental partitioning of orga-nochlorine pesticides. Environ. Toxicol. Chem. 2002, 21, 984-990.

(14) U.S. Environmental Protection Agency. Reregistration EligibilityDecision (RED); DCPA; EPA 738-R-98-005; U.S. EPA: Wash-ington, DC, 1998.

(15) Stephens, V. C.; Deacon, J. R. Distribution and Concentrationsof Selected Organochlorine Pesticides and PCB’s in StreambedSediment and Whole-Body Fish in the Upper Colorado RiverBasin, 1995-1996; USGS Upper Colorado River Basin NAWQAFact Sheet FS-167-97; USGS, 1998; http://pubs.usgs.gov/fs/fs167-97./.

(16) U.S. Environmental Protection Agency. Guidance for the Reg-istration of Pesticide Products Containing Dimethyl Tetra-Chloroterephthalate (DCPA) as the Active Ingredient; PB88-225263; U.S. EPA: Washington, DC, 1988.

(17) Klopman, G.; Fercu, D.; Rosenkranz, H. S. The carcinogenicpotential of dacthal and its metabolites. Environ. Toxicol. Chem.1996, 15, 80-84.

(18) Harner, T.; Mackay, D. Measurement of octanol-air partitioncoefficients for chlorobenzenes, PCBs, and DDT. Environ. Sci.Technol. 1995, 29, 1599-1606.

(19) Yao, Y.; Tuduri, L.; Harner, T.; Blanchard, P.; Waite, D.; Poissant,L.; Murphy, C.; Belzer, W.; Aulagnier, F.; Li, Y.-F.; Sverko, E.Spatial and temporal distribution of pesticide air concentrationsin Canadian agricultural regions. Atmos. Environ. 2006, 40,4339-4351.

(20) Harner, T.; Shoeib, M.; Diamond, M.; Stern, G.; Rosenberg, B.Using passive air samplers to assess urban-rural trends forpersistent organic pollutants. 1. Polychlorinated biphenyls andorganochlorine pesticides. Environ. Sci. Technol. 2004, 38, 4474-4483.

(21) Motelay-Massei, A.; Harner, T.; Shoeib, M.; Diamond, M.; Stern,G.; Rosenberg, B. Using passive air samplers to assess urban-rural trends for persistent organic pollutants and polycyclicaromatic hydrocarbons. 2. Seasonal trends for PAHs, PCBs, andorganochlorine pesticides. Environ. Sci. Technol. 2005, 39, 5763-5773.

(22) Harner, T.; Pozo, K.; Gouin, T.; Macdonald, A. M.; Hung, H.Global pilot study for persistent organic pollutants (POPs) usingPUF disk passive air samplers. Environ. Pollut. 2006 (in press).

(23) Mackay, D.; Shiu, W.-Y.; Ma, K. C. Illustrated Handbook ofPhysical-Chemical Properties and Environmental Fate for Or-ganic Chemicals; CRC Press: New York, 1997, Vol. 5.

(24) Michigan Department of Environmental Quality. OperationMemorandum No. 1; www.deq.state.mi.us/documents/deq-rrd-part201-rules-Rule752table.pdf.

(25) Pozo, K.; Harner, T.; Shoeib, M,; Urrutia, R.; Barra, R.; Parra, O.;Focardi, S. Passive-sampler derived air concentrations ofpersistent organic pollutants on a north-south transect in Chile.Environ. Sci. Technol. 2004, 38, 6529-6537.

(26) Gouin, T.; Harner, T.; Blanchard, P.; Mackay, D. Passive andactive air samplers as complementary methods for investigatingpersistent organic pollutants in the Great Lakes basin. Environ.Sci. Technol. 2005, 39, 9115-9122.

(27) U.S. Geological Survey. 1997 DCPA Use Map; http://ca.water.usgs.gov/pnsp/pesticide_use_maps/show_map.php?year)97&map)m1872.

(28) Miller, F. M.; Gomes, E. D. Detection of dacthal residues inenvironmental samples. Pestic. Monit. J. 1974, 8, 53-58.

(29) Rostad, C. E. From the 1988 drought to the 1993 flood: Transportof halogenated organic compounds with the Mississippi Riversuspended sediment at Thebes, Illinois. Environ. Sci. Technol.1997, 31, 1308-1312.

(30) Majewski, M. S.; Foreman, W. T.; Goolsby, D. A.; Nakagaki, N.Airborne pesticide residues along the Mississippi River. Environ.Sci. Technol. 1998, 32, 3689-3698.

Received for review July 20, 2006. Revised manuscript re-ceived September 29, 2006. Accepted November 1, 2006.

ES061725R

694 9 ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 41, NO. 3, 2007