Environ. Sci. Technol. 2004, 38, 1368-1374
Organochlorine Pesticides in the Air around the Taihu Lake, China XINGHUA QIU, TONG ZHU,* JING LI, HANSHENG PAN, QUANLIN LI, GUOFANG MIAO, AND JICHENG GONG State Key Joint Laboratory for Environmental Simulation and Pollution Control, College of Environmental Sciences, Peking University, Beijing 100871, China
Organochlorine (OC) pesticides have been used broadly in China’s past, yet very little is known about their atmospheric concentrations and transport. In this work, air samples were collected in the Taihu Lake Region, China, from July 23 to August 11, 2002, to measure concentrations of OC pesticides in air. The average concentrations of R and γ- hexachlorocyclohexane (HCH), hexachlorobenzene (HCB), heptachlor (HEPT), R-endosulfan, p,p′-DDT, p,p′-DDE, p,p′-DDD, and o,p′DDT in the air were 74 and 46, 47, 53, 307, 124, 212, 36, and 767 pg m-3, respectively. It was interesting to note that the concentrations of p,p′-DDT, p,p′-DDE, and o,p′-DDT were all very high, even though the use of technical DDT has been banned in China since 1983. Moreover, the average concentration ratios of o,p′-DDT/p,p′-DDT and p,p′DDE/p,p′-DDT were as high as 6.3 and 1.8. This suggested that there could be an unknown source of DDT-related compounds (DDTs), especially o,p′-DDT and p,p′-DDE. It is very likely that this unknown source was the application of dicofol, an acaricide manufactured from technical DDT and used mainly on cotton fields to treat mites in China. Backward trajectory analysis also provided consistent evidence that the high air concentrations of DDTs were related to trajectories from the area north of the Yangtze River, where cotton fields account for a significant fraction of land use.
Introduction OC pesticides are typical persistent organic compounds (POPs). In the past decades, the “global distillation” or “grasshopper effect” model has successfully illustrated that POPs can migrate to polar regions through evaporation, atmospheric transport, and deposition (1, 2). For this reason, research on the atmospheric concentrations of OC pesticides is very important. Agricultural pesticides are released into the atmosphere by spray drift, postapplication volatilization, and wind erosion of soil. Among these, volatilization can remove large fractions of the pesticide initially applied to a field. In one case, total volatile losses of DDT during about a one month period were more than half of the total amounts originally applied to a cotton field (3). Therefore, high air concentration of an OC pesticide usually means it is currently used in a nearby area or has been transported from another region. China is a large producer and consumer of pesticides. Beginning in the 1950s and continuing until their uses were banned in 1983, technical HCH and DDT were widely used * Corresponding author phone: 86 (10) 62754789; fax: 86 (10) 62751927; e-mail:
[email protected]. 1368
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 38, NO. 5, 2004
in China. From 1952 to 1984, the total production of technical HCH was around 4 million tons (4) and from 1951 to 1983 the production of DDT was 0.27 million tons (5). Even after the ban of technical HCH and DDT, 3200 tons of Lindane (almost pure γ-HCH) was still in use between 1991 and 2000 (4), and DDT production also continues largely thanks to export demand and to dicofol production. However, up to now, in China, very little field monitoring data, especially data concerning air concentrations, is available for the presence of environmental OC pesticide residue. An understanding of the impacts of OC pesticides on China’s environment is therefore very limited. Taihu Lake lies in the Yangtze River Delta. It is a typical shallow freshwater lake, with an area of 2338 km2 and average depth of 1.9 m. The Taihu Lake Region (TLR) is one of the most densely populated areas in China; it accounts for 0.4% of the total area of China and 2.9% of the nation’s population, and it provides more than 14% of the gross domestic production (GDP) (6). Large amounts of OC pesticides were used in this area in the past, and their environmental residues are still a serious problem. In 1992, after the use of technical HCH and DDT had been banned for nearly a decade, the concentrations of DDTs and HCHs in rice field soil in Wuxi, a city located beside Taihu Lake, were still as high as 53 ng g-1 and 32 ng g-1 (7). A study in 1999-2000 also showed that concentrations of DDTs and HCHs in TLR soil were still measurable in most soil samples (8). The Taihu Lake bed is covered with sediment that averages 0.58 m thick (9), and in 2000, the total average concentration of OC pesticides in the sediment was more than 15 ng g-1 (10). Because the lake is shallow, the sediment can be easily stirred by wind or even by fishing activities, releasing OC pesticides into the lake water and subsequently into the atmosphere through air-water gas exchange. One can suspect that the sediment in Taihu Lake might thus be a source of certain OC pesticides in the air. Therefore it is very important to carry out research on the sources and sinks of OC pesticides in this region. However, up to now there has been no report about the concentrations of OC pesticides in atmosphere of the TLR. This paper reports the measurement results of atmospheric concentrations of OC pesticides in the TLR in the summer of 2002, and then discusses possible sources of certain important OC pesticides.
Materials and Methods Sample Collection. A high-volume polyurethane foam (PUF) sampler (Andersen Co.) was used to collect air samples. A glass fiber filter (GFF, diameter 10 cm, Andersen Co.) was used to separate particles before the air reached the PUF plug. GFFs were cleaned at 450 °C for 4 h, and then wrapped in aluminum foil before sampling. The PUF plugs (density 0.02 g cm-3, diameter 8.2 cm, and thickness 7.5 cm, Beijing Longtan Packing Materials Factory) were Soxhlet precleaned with acetone for 22 h, then dried in a vacuum desiccator, and stored in sealed glass bottles until sampling. The sampler was installed on a 10-m2 platform at the yard of the Taihu Laboratory for Lake Ecosystem Research (TLLER) (N 31.42°, E 120.21°), which is located at the entrance of Meiliang Bay in the northern end of Taihu Lake. The platform extended about 100 m offshore, with the sampler inlet situated 2.9 m above the water surface. Before sampling, the temperature, pressure, and flow-rate sensors were calibrated. Daytime (7:00 to 18:00) and nighttime (19:00 to 6:00 the following day) samples of air were then collected daily, except for the sample taken on August 1, which collected 10.1021/es035052d CCC: $27.50
2004 American Chemical Society Published on Web 01/21/2004
TABLE 1. Analytical Conditions, Range of Air Concentrations of OC Pesticides, and Mean Recovery of the Air Samples Taken in the TLR from July 23 to August 11, 2002
target compound
ion selection(m/e) parent (width) daughter
IDL (pg)
blank valuesa (pg m-3)
no. of samples detected (total 21)
range of air concentration (mean) (pg m-3)
mean recoveryb ( SDc (no. of spikes) (%)
HCH R-HCH γ-HCH DDT p,p′-DDT
183(4) 183(4)
145-147 145-147
3.5 4.4
4.2 2.5
21 21
21-164(74) 18-96(46)
89.8 ( 6.5(7) 96.1 ( 5.7(7)
235(4)
6.3
n.d.d
21
34-394(124)
137.8 ( 5.5(7)
p,p′-DDE
318(4)
4.4
3.7
21
55-502(212)
107.3 ( 4.3(7)
p,p′-DDD
235(4)
6.5
n.d.
15
n.d.-75(36)
112.5 ( 5.6(3)
o,p′-DDT
235(4)
4.1
7.5
21
80-2753(767)
103.1 ( 7.3(3)
HCB HEPT R-endosulfan
284(4) 272(4) 195(4)
163-166 199-202 245-246 316-318 163-166 199-202 163-166 199-202 247-249 235-239 157-162
7.9 4.2 38.5
10.4 n.d. n.d.
21 21 16
18-109(47) 18-173(53) n.d.-888(307)
83.1 ( 8.1(7) 103.1 ( 7.3(7) 117.1 ( 9.4(7)
a Each blank value was divided by 308 m3. deviation. d n.d. ) not detected.
b
The amount of spiked standard was 100 ng each and the matrix was blank PUF. c SD ) standard
air for 22 h. The sampling flow rate was 28 m3/h and most samples collected 308 m3 of air. Field blanks were prepared by placing a PUF plug beside the sampler inlet during the sampling period of one sample. One field blank was prepared for about every three samples. In total, nine daytime samples, eleven nighttime samples, one whole day sample, and eight field blanks were collected. The average ambient temperature for the sampling period was between 24.1 and 32.4 °C. After sampling, the samples and field blanks were sealed in glass bottles and stored below 4 °C until analysis. Sample Extraction and Preparation. OC pesticides adsorbed in the PUFs were Soxhlet extracted with 750 mL of mixed solvent (n-hexane/diethyl ether 9:1 V/V) for 22 h. PCB65 was spiked on each PUF plug as recovery surrogate. The extract was concentrated and then cleaned with a column of 5 g of Florisil (PR grade, Supelco Co., pretreated at 130 °C for more than 12 h) and 1-2 cm thick anhydrous sodium sulfate. The OC pesticides were then eluted with 20 mL of acetone/n-hexane (1:9, V/V). The elution was concentrated to 1 mL with a gentle stream of purified nitrogen, and PCB155 was spiked as internal standard before analysis. Diethyl ether (analytical grade, Tianjin Reagent Factory), acetone, and n-hexane (analytical grade, Beijing Second Reagent Factory) were glass-to-glass distilled twice, and a solvent check showed that there was no interference with the target compounds. Standards of R-HCH, β-HCH, γ-HCH, δ-HCH, HCB, HEPT, PCB-65, PCB-155, R-endosulfan, p,p′-DDT, p,p′-DDE, and p,p′-DDD were purchased from Chem Service Inc. (West Chester, PA), and o,p′-DDT was obtained from Dr. Ehrenstorfer GmbH, Germany. Analysis. OC pesticides in all samples and field blanks were analyzed using a gas chromatograph (GC) with an iontrap mass spectrometer (MS) (Finnigan Trace GC/Polaris Q). A fused-silica capillary column (Rtx-5ms, Restek Co.), 30 m × 0.25 mm and 0.25 µm film was used for separation. High-purity helium was used as a carrier gas at a constant flow rate of 1.0 mL min-1. Samples (1 µL) were injected under splitless injection mode. The temperature of the injector was 250 °C and the transfer line was 280 °C. The oven-temperature program was set as follows: 25 °C/min
8 °C/min
100 °C (2 min) 98 170 °C 98 0.7 °C/min
25 °C/min
225 °C 98 235 °C 98 260 °C (2 min) The MS was operated with a 70-eV electron impact (EI) mode. When the MS is using ion-trap as the mass separator,
MS-MS mode can be used to achieve high sensitivity. Some research reported that the instrument detection limit (IDL) of R-HCH and γ-HCH could be as low as 0.8 and 0.7 pg when defined as the amount of standard that generates a peak with a signal/noise ratio of 10 (11). We defined IDL ) 3.36S, where S was the standard deviation of the peak heights from six injections of the standard solutions containing the target compounds, which had concentrations that could produce a signal/noise ratio of about 5. Table 1 lists the selection of parent and daughter ions and the IDLs of some target compounds. The IDLs were from 3.5 to 8 pg for most of the target compounds, except R-endosulfan, which had an IDL of 38.5 pg. Such low IDLs permitted low sample volumes and short sampling times, which enabled us to increase time resolution and study the diurnal variation of atmospheric concentrations of OC pesticides.
Results and Discussion The ranges and mean concentrations of nine OC pesticides sampled from the air around Taihu Lake, from July 23 to August 11, 2002, as well as spiked standards’ recoveries, are listed in Table 1. R-HCH, γ-HCH, p,p′-DDT, p,p′-DDE, o,p′DDT, HEPT, and HCB were detected in all of the 21 samples, whereas p,p′-DDD and R-endosulfan were partially detected. For the eight field blanks, the target compound signals were below IDL or 10% of the samples’ signal, except HCB, which was a little high in the field blanks. Recoveries of the recovery surrogate, PCB65, ranged from 75.0% to 97.3% in all 21 samples. During the sampling period the mean concentration of R-HCH was 74 pg m-3, ranging from 21 to 164 pg m-3. The mean concentration of γ-HCH was 46 pg m-3, ranging from 18 to 96 pg m-3. Table 2 shows our data as they compare with the results of similar studies conducted at the Great Lakes of North America (12, 13), at Russia’s Lake Baikal in summer (14), and in Hong Kong, which was the only data available to us regarding China’s atmospheric OC pesticide concentrations (15). Though the Hong Kong observations were in winter, ambient temperatures were still rather high (ranging from 13.1 to 25.4 °C). Table 2 shows that the air concentration of R-HCH in the TLR was lower than that at the Great Lakes in 1989/1990 and Lake Bakail in 1991, and slightly lower than that in both Hong Kong in 2000 and at Resolute Bay, an Arctic region, in the summer of 1992 (114 ( 16 pg m-3) (16). Also interesting to note is that air concentrations of γ-HCH in the TLR and VOL. 38, NO. 5, 2004 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
1369
TABLE 2. Air Concentrations (pg m-3) of Organochlorine Pesticides in the Taihu Lake Region Measured from July 23 to August 11, 2002 in Comparison with those Reported in the Great Lakes, Lake Baika, and Hong Konga HCHs r-HCH Great Lakes: Green Bay, June 1989 213 ( 46 Great Lakes: meanb, Aug 1990 219 ( 47 Lake Baikal, Lystvianka Hydromet 489 ( 275 Station, June 1991 Hong Kong, Dec and Jan, 2000 92-188 TLR July-Aug 2002 74 ( 45
DDTs r/γ
γ-HCH
21-50 46 ( 24
o,p′-DDT
5.9 ( 2.6
24-29 15-25 22-71 1.0-2.1 124 ( 91 212 ( 134 767 ( 712 b
p,p′-DDE/p,p′-DDT reference
0.91
12 13 14
0.94-3.2
15 this work
Great Lakes mean: Lakes Michigan, Huron, Erie and Ontario, excludes Green
in the Great Lakes were very close. This indicates that even though large amounts of technical HCH have been used in China before, the ban of use nearly twenty years ago has nonetheless significantly reduced HCH residues in the environment. Generally, technical HCH contains isomers in the following percentages (17): R, 55-80%; β, 5-14%; γ, 8-15%; δ,2-16%; and , 3-5%. During the sampling period, the concentration ratios of R-HCH/γ-HCH in the air ranged from 1.0 to 2.1, and were much lower than that of technical HCH. This indicates that the source of γ-HCH in the TLR might be not only the use of technical HCH but also the use of Lindane. Lindane is the commercial name of γ-HCH and it is still used in China and other countries (18). Studies indicate that in the areas where Lindane is used, or in the regions adjacent to such areas, high concentrations of γ-HCH, ranging from several hundred to several thousand pg m-3 (19, 20) can be detected. These concentrations are much higher than those found in the TLR. The low air concentration of γ-HCH around Taihu Lake suggests that, during the sampling period, there was no use of Lindane in this region, and that the atmospheric transport of Lindane from other regions was not significant. It is very likely that the HCHs observed in the air of the TLR came from the environmental residues of technical HCH and Lindane. HCB is not only a fungicide but also an intermediate of some chlorinated industrial processes. HCB is also believed to be extremely persistent in the environment, and due to reaction with hydroxyl radical, it has an atmospheric lifetime of about 80 days (21). The mean air concentration of HCB in the TLR during the sampling period was 47 pg m-3 (ranging from 18 to 109 pg m-3), which was lower than that in both Green Bay, WI (160 ( 87 pg m-3) (22) and near Lake Baikal (194 ( 47 pg m-3 (14), 70-170 pg m-3, (23)). The low observed HCB air concentration around Taihu Lake might be due to its breakthrough when a PUF plug was used as the absorbent (24). Yuan Xuyin et al. reported a high HCB concentration (mean 2.2 ng g-1) in the sediment of Taihu Lake (10). This indicated that, similar to Lake Ontario and Lake Baikal (14, 25), HCB was emitted from Taihu Lake water to the air through water-air gas exchange, and that Taihu Lake might be an important source of air HCB in this region. During the sampling period the mean air concentration of HEPT was 53 pg m-3, and it ranged from 18 to 173 pg m-3. HEPT is not a widely used pesticide in China. The HEPT detected in the air of the TLR might come from the usage of Chlordane, another OC pesticide mainly used to kill termites in south China. Commercial products of Chlordane usually contain a certain fraction of HEPT, so these may also have been a factor. Endosulfan is one of the few OC insecticides that are still in use all over the world, including North America, Europe, and China. The technical formulation contains approximately 70% R-endosulfan and 30% β-endosulfan, and in the environment the R isomer is by far more abundant (26). 9
p,p′-DDE
135 ( 30 1.6 ( 0.5 8.7 ( 5.6 15 ( 8.8 40 ( 14 6.0 ( 1.9 38 ( 72 59 ( 45 105 ( 65 4.6-5.8 8.3 ( 4.4 7.5 ( 4.8
a The errors were one standard deviation of the average value. Bay values.
1370
p,p′-DDT
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 38, NO. 5, 2004
Although GC/MSMS has a relatively high IDL (about 39 pg) for R-endosulfan, the air concentration of R-endosulfan in the TLR was so high that it could be detected in most samples, with an average concentration of 307 pg m-3 and a maximum concentration of 888 pg m-3. R-Endosulfan was not detected in five samples, which might be due to the infusion of a relatively clean Pacific Ocean air mass that occurred during these sampling periods (see below). Although endosulfan is considered less persistent than other OC insecticides, in regions where it is used (for example the Great Lakes area) high air concentrations of R-endosulfan have been shown to persist to some extent, as is evidenced by detected mean concentrations of 350 pg m-3 (maximum 3700 pg m-3) in 1988-1989 (27) and 86 pg m-3 (maximum 890 pg m-3) in 1991-1992 (26). These elevated concentrations have also led to increasing levels of R-endosulfan in the Canadian Arctic region over the period from 1993 to 1998 (28). The term DDTs refers to DDT-related compounds, including p,p′-DDT, o,p′-DDT and their metabolites, DDE and DDD. During the sampling period, p,p′-DDT, p,p′-DDE, o,p′-DDT, and p,p′-DDD were detected in the air samples, with very high concentrations of p,p′-DDT, p,p′-DDE, and o,p′-DDT. The p,p′-DDE and o,p′-DDT in some samples were so high that they could even be identified by MS under full scan mode. The mean concentration of p,p′-DDT in the air was 124 pg m-3 (ranging from 34 to 294 pg m-3), p,p′-DDE 212 pg m-3 (ranging from 55 to 502 pg m-3), and o,p′-DDT 767 pg m-3 (ranging from 80 to 2753 pg m-3). Use of technical DDT was banned in the U.S. in 1972 and in China in 1983. However, after almost the same time span in each country, the summer of 2002 air concentrations of p,p′-DDT and p,p′-DDE near Taihu Lake were much higher than those measured near the Great Lakes in the summer of 1989/1990 (Table 2). The measurement results of DDTs in sediment in Hong Kong and the Pearl River Delta of China also show that there is no apparent sign of declining DDT concentrations as a result of the ban in 1983 (29, 30). Hence, one can suspect that there must be still local or regional emission sources for DDTs in the TLR. It is worth noting that air concentrations of o,p′-DDT around Taihu Lake were abnormally high, with an average concentration of 767 pg m-3 and a maximum of 2753 pg m-3 (Tables 1 and 2). Additionally, the ratio of o,p′-DDT/p,p′DDT was from 2.0 to 15.5. Observations over the East China Sea in 1989 (31) and in Hong Kong in 2000 (15) also showed that o,p′-DDT had higher atmospheric concentrations than p,p′-DDT. These observations contradict the fact that technical DDT contains less o,p′-DDT (15%) than p,p-DDT (85%) (32). Considering the fact that the half-life times of o,p′-DDT and p,p′-DDT due to phytodegradation are similar (33), and that Taihu Lake sediment contains lower o,p′-DDT concentrations (0.35 ng g-1) than p,p′-DDT concentrations (0.66 ng g-1) (10), one can conclude that the o,p′-DDT in the air of the TLR did not come from residue but from a “new” source. Because this source should have higher portions of o,p′-
FIGURE 1. Three types of backward trajectories and their covering area: S(D7/25) is the S type trajectory calculated for the day of July 25; E(D8/2) is the E type trajectory calculated for the day of August 2; and N(N8/10) is the N type trajectory calculated for the night of August 10. DDT than of p,p-DDT, the source in question could therefore not be technical DDT. This conclusion may help to explain the fact that in the Arctic region between 1993 and 1998 o,p′-DDT concentrations have shown an increasing trend, while in most samples p,p′-DDT concentrations were below the method detection limit (MDL) (28). In all air samples, the average ratio of p,p′-DDE/p,p′DDT was 1.8. DDE has no insecticidal uses and in the environment it mainly comes from the degradation of DDT. It follows that the ratio of p,p′-DDE/p,p′-DDT in air can be used as an indicator of the “age” of DDT (34). There are two possible reasons that might explain the high ratio of p,p′DDE/p,p′-DDT in the TLR. The first one is that technical DDT in the environment had already been degrading for a long time. This could not, however, explain that DDT and DDE both had very high concentrations. The second possibility is that there was a new source that either contained DDE or could degrade to DDE in environment, as was the case in California and Texas where abnormal ratios of DDE/ DDT were detected in the bodies of birds even after many years’ ban of technical DDT (35). The high ratios of o,p′-DDT/p,p′-DDT and p,p′-DDE/p,p′DDT led to the suspicion that a new source of o,p′-DDT, p,p′-DDE, and/or p,p′-DDT exists, and this new source is very likely the widely used pesticide dicofol. Dicofol, trade name Kelthane, is a nonsystemic acaricide used extensively for the control of mites. Its acaricidal active ingredient is 2,2,2-trichloro-1,1-bis (4-chlorophenyl) ethanol. At present there are more than sixty registered manufacturers in China that produce dicofol or formulate dicofol products (36). Dicofol is usually manufactured from technical DDT (37). After the synthesis reaction, DDT and the intermediate R-chloro-DDT may remain in the dicofol product as impurities. Because of the hindrance of the ortho-chlorine atom on the benzene ring, it is more difficult for o,p′-DDT than for
p,p′-DDT to be chloridized at the alpha carbon atom. Therefore, in technical dicofol, o,p′-DDT has a higher proportion than p,p′-DDT as impurity. Moreover, the intermediate R-chloro-DDT can convert to DDE (38, 39). This process may occur not only in the manufacturing process, but also when R-chloro-DDT enters the environment, which could lead to elevated environmental DDE levels. Because the concentration of DDTs in technical dicofol is high, the use of dicofol was temporarily banned by the USEPA in 1986. The EC “Prohibition Directive 79/117/EEC” also strictly limited the impurity content of dicofol (40), as dicofol manufactured up until that time had contained very high proportions of DDTs. In some dicofol formulations the content of DDT was up to 575 g per kg of dicofol product, and o,p′-DDT accounted for about 77% of these DDT impurities. Additionally, in some dicofol formulations, the content of R-chloro-DDT reached 54 g kg-1 (41). Di Muccio et al. analyzed samples of dicofol formulations available in Italy and also found that between 0.63% and 23.19% of the dicofol was made up of DDTs (excluding R-chloro-DDT), with an average of 8.28%. p,p′-DDT appeared less abundant among the impurities, whereas p,p-DDE and o,p′-DDT appeared to be more abundant, with o,p′-DDT showing a greater incidence of high values (42). After the EC “Prohibition Directive 79/117/EEC”, DDTs accounted for only up to 7 g kg-1 of the dicofol in the formulation, and in 8 of the 13 samples the level of DDTs was less than 0.1 g kg-1 (41). VOL. 38, NO. 5, 2004 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
1371
FIGURE 2. Air concentrations of OC pesticides in each sample taken from July 23 to August 11, 2002, and their associated types of backward trajectories. Until now there has been no regulation to restrict DDTrelated impurities in dicofol in China. Qualitative analysis of technical dicofol from a Chinese manufacturer showed that DDTs might be a serious problem (43). However, data about the content of DDTs in dicofol is very limited in China. Apparently, further research is required. Backward Trajectory Analysis. Backward trajectories during the sampling period were analyzed to obtain information about the origins of OC pesticides in the atmosphere of the TLR. Hybrid single-particle Lagrangian integrated trajectory (HYSLPIT4), offered by the National Oceanic and 1372
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 38, NO. 5, 2004
Atmospheric Administration (NOAA), was used to calculate trajectories. Meteorological data was obtained from NCEP’s Global Data Assimilation System, the National Weather Service’s National Centers for Environmental Prediction (44, 45). Trajectories reaching 5 days (120 h) back were calculated twice a day, using start times of 0:00 for the nighttime sampling period and 12:00 for the daytime sampling period. The calculation heights were set at 800, 1500, and 3000 m above the ground. The trajectory is influenced by atmospheric circulation systems. The most important system over eastern China is
TABLE 3. Air Concentrations (pg m-3) of Organochlorine Pesticides Measured in the Taihu Lake Region from July 23 to August 11, 2002a HCH r-HCH
S type (N ) 9) E type (N ) 5) N type (N ) 2)
DDT
p,p′-DDT
p,p′-DDE
o,p′-DDT
HEPT
HCB
average concentrations of samples corresponding to different type of trajectories 65 ( 34 41 ( 21 90 ( 51 183 ( 121 721 ( 429 32 ( 8 26 ( 7 56 ( 13 91 ( 22 216 ( 112 133 ( 34 76 ( 18 277 ( 166 448 ( 77 2607 ( 207
44 ( 20 43 ( 13 28 ( 8
54 ( 26 23 ( 6 70 ( 1
average concentrations of daytime and nighttime samples 70 ( 50 43 ( 28 144 ( 123 189 ( 112 777 ( 780 81 ( 43 49 ( 23 114 ( 59 238 ( 155 803 ( 709
51 ( 47 55 ( 42
42 ( 18 53 ( 26
daytime samples (N ) 9) nighttime samples (N ) 11)
γ-HCH
a The concentrations were averaged according to the associated types of backward trajectories and daytime and nighttime sampling. The errors were one standard deviation of the average value.
the Western Pacific high (WPSH) in summer, sometimes combined with tropical cyclones. Their activities caused the trajectories to come from different directions. The WPSH has about a one-week fluctuation cycle, and during the whole sampling period it experienced two eastward retreats and one westward extension. The corresponding trajectories could be divided into three types according to the direction and passed area (Figure 1a-d): S type trajectory, which passed through South China, came from either the IndoChina peninsula or the South China Sea near Taiwan and the Philippines; E type trajectory, with eastern or southeastern direction trajectories, came from the western part of the Pacific Ocean; N type trajectory came from the area north of the Yangtze River. The rest of the trajectories were not regular at the three heights, possibly due to the great difference of weather situations at different altitudes when tropical cyclones occurred (the 9th and 12th typhoons in 2002). As a result these trajectories were not grouped to a special category. Air concentrations of OC pesticides in each sample and their corresponding type of backward trajectory are showed in Figure 2. The average air concentrations of OC pesticides in samples corresponding to S, E, and N type trajectories are listed in Table 3. Figure 2 shows that all the OC pesticides detected had obvious daily variation. Low concentrations were observed mainly during the period from August 1 to 4, and high concentrations were observed from July 25 to 27 and from August 8 to 10. Figure 2 shows clearly that variations in OC pesticide concentration were closely associated with trajectory type. Table 3 shows that in samples with E type trajectory, most OC pesticides exhibited low concentrations, and that R-endosulfan levels were even below the IDL. This might be due to relatively clean air coming from the Pacific Ocean. All OC pesticides (excluding HEPT) detected in samples with N type trajectory had higher concentrations than those with S type trajectory. This indicates that more OC pesticides come from the North than from South China. HEPT was an exception. Its concentrations in samples with S type trajectory were slightly higher than in those with N type. It is likely that HEPT comes from Chlordane, a chemical used mainly in South China to kill termites. Concentrations of DDTs in samples with N type trajectory were more than two times higher than those with S type trajectory, and the average concentration of o,p′-DDT in samples with N type trajectory had an abnormally high value (2607 pg m-3). This suggests that during the sampling period there might have been a “new” source of DDTs, especially o,p′-DDT, in the area north of the Yangtze River. This suspicion is consistent with the hypothesis that technical dicofol use contributed to high concentrations of DDTs, especially o,p′-DDT, in the air of the TLR. Dicofol can be used on many crops and fruit trees, but is used in the greatest proportion on cotton (46). In South China for instance, dicofol is used on fruit trees such as litchi,
longan, and orange, but the amounts used on these fruits are less than those used on cotton. The north area of the middle and lower reaches of the Yangtze River, along with the Huanghe-Huaihe River Plain, have a large area of cotton fields. It is reported that in Jiangsu Province in 1992, the concentration of DDT residue in the soil of a cotton field in Nantong (north of the Yangtze River) was 1230 ng g-1, while the soil of a rice field in Wuxi (south of the Yangtze River) contained a concentration of only 45 ng g-1 (7). This suggests that after many years’ ban on DDTs, there is still severe DDT pollution in cotton fields due to dicofol impurities. Mite-induced cotton damage usually occurs during high temperatures and drought, which are both typical in areas north of the Yangtze River from late July to August. As this study was carried out in the summer, dicofol application coincided with the sampling period. It is therefore very likely that the high air concentrations of DDTs, especially o,p′DDT, observed in the TLR during this study were due to air mass transport from north of the Yangtze River, where dicofol was applied to cotton fields and DDT impurities in the dicofol were emitted. Moreover, the ratios of o,p′-DDT/p,p′-DDT were 8.0, 9.4, and 3.9 corresponding to S, N, and E type trajectories, respectively. The values of S and N type were much higher than that of E type. This suggests that the source of o,p′-DDT was in China. Daytime/Nighttime Difference. Many studies show that the atmospheric concentrations of semivolatile organic compounds (SVOCs), including OC pesticides, have obvious diurnal variation. This variation depends on ambient temperature and observed daytime concentration peaks (4749). Table 3 shows that, except for p,p′-DDT, all the OC pesticides had a higher concentration in nighttime than in daytime, though the differences were within one standard deviation. Many factors, such as atmospheric transport, height of the atmospheric boundary layer, and wind velocity might lead to the daytime/nighttime differences in air concentrations for OC pesticides. At night in the lake area, land-lake circulation breezes might carry relatively high concentrations of OC pesticides from the land. A longer sampling period and higher time resolution would be required to obtain the diurnal variation of OC pesticides in the TLR.
Acknowledgments This study was supported by the National Outstanding Young Scholar Fund (49925513) of the Chinese Natural Science Foundation of China.
Supporting Information Available Map of the location of the Taihu Lake Region (TLR) in China and the sampling site (Figure S1). This material is available free of charge via the Internet at http://pubs.acs.org. VOL. 38, NO. 5, 2004 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
1373
Literature Cited (1) Wania, F.; Mackay, D. Sci. Total Environ. 1995, 160/161, 211232. (2) Wania, F.; Mackay, D. Environ. Sci. Technol. 1996, 30, 390A396A. (3) Willis, G. H.; McDowell, L. L.; Harper, L. A.; Southwick, L. M.; Smith, S. J. Environ. Qual. 1983, 12, 80-85. (4) Li, Y. F.; Cai, D. J., Shan, Z. J.; Zhu, Z. L. Arch. Environ. Contam. Toxicol. 2001, 41, 261-266. (5) Li, Y. F.; Cai, D. J.; Singh, A. Adv. Environ. Res. 1999, 2 (4), 497506. (6) Shen, L.; Lin, G. F.; Tan, J. W., Shen, J. H. Chemosphere 2000, 41, 129-132. (7) Lin, Y. S.; Gong, R. Z.; Zhu, Z. L. In Pesticide and Ecoenvironmental Protection (in Chinese); Chemical & Technologies Publishing House: Beijing, 2000; pp 14-15. (8) Feng, K.; Yu, B. Y.; Ge, D. M.; Wong, M. H.; Wang, X. C.; Cao, Z. H. Chemosphere 2003, 50, 683-687. (9) Yuan, X. Y.; Xu, N. Z.; Tao, Y. X.; Zheng, X. M.; Liu, D. H. Resources Surver & Environment (in Chinese) 2003, 24, 20-28. (10) Yuan, X. Y.; Wang, Y.; Chen, J.; Sun, C.; Xu, N. Z. Environ. Sci. (in Chinese) 2003, 24, 121-125. (11) Lakaschus, S.; Weber, K.; Wania, F.; Bruhn, R.; Schrems, O. Environ. Sci. Technol. 2002, 36, 138-145. (12) McConnell, L. L.; Cotham, W. E.; Bidleman, T. F. Environ. Sci. Technol. 1993, 27, 1304-1311. (13) McConnell, L. L.; Bidleman, T. F.; Cotham, W. E.; Walla, M. D. Environ. Pollut. 1998, 101, 391-399. (14) McConnell, L. L.; Kucklick, J. R.; Bidleman, T. F.; Ivanov, G. P.; Chernyak, S. M. Environ. Sci. Technol. 1996, 30, 2975-2983. (15) Louie, P. K. K.; Sin, D. W. M. Chemosphere 2003, 52, 1397-1403. (16) Falconer, R. L.; Bidleman, T. F.; Gregor, D. J. Sci. Total Environ. 1995, 160/161, 65-74. (17) Metcalf, R. L. Organic insecticides, Their Chemistry and Mode of Action; Interscience: New York. 1955. (18) Li, Y. F.; McMillan, A.; Scholtz, M. T. Environ. Sci. Technol. 1996, 30, 3525-3533. (19) Granier, L. K.; Chevreuil, M. Atmos. Envrion. 1997, 31, 37873802. (20) Haugen, J. E.; Wania, F.; Ritter, N.; Schlabach, M. Environ. Sci. Technol. 1998, 32, 217-224. (21) Bidleman, T. F.; Atlas, E. L.; Atkinson, R.; Bonsang, B.; Burns, K.; Keene, W. C.; Knap, A. H.; Miller, J.; Rudolph, J.; Tanabe, S. In Long-Range Atmospheric Transport of Natural and Contaminant Substances; Knap, A. H., Kaiser, M., Eds.; Kluwer Academic: Boston, 1990; p 281. (22) McConnell, L. L. Ph.D. Dissertation, University of South Carolina, 1992. (23) Iwata, H.; Tanabe, S.; Ueda, K.; Tatsukawa, R. Environ. Sci. Technol. 1995, 29, 792-801. (24) Burdick, N. F.; Bidleman, T. F. Anal. Chem. 1981, 53, 19261929. (25) Hoff, R. M.; Strachan, W. M. J.; Sweet, C. W.; Chan, C. H.; Shackleton, M.; Bidleman, T. F.; Brice, K. A.; Burniston, D. A.; Cussion, S.; Gatz, D. F.; Harlin, K.; Schroeder, W. H. Atmos. Environ. 1996, 30, 3505-3527. (26) Burgoyne, T. W.; Hites, R. A. Environ. Sci. Technol. 1993, 27, 910-914.
1374
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 38, NO. 5, 2004
(27) Hoff, R. M.; Mulr, D. C. G.; Grift, N. P. Environ. Sci. Technol. 1992, 26, 266-275. (28) Hung, H.; Halsall, C. J.; Blanchard, P.; Li, H. H.; Fellin, P.; Stern, G.; Rosenberg, B. Environ. Sci. Technol. 2002, 36, 862-868. (29) Zheng, G. J.; Lam, M. H. W.; Lam, P. K. S.; Richardson, B. J.; Man, B. K. W.; Li, A. M. Y. Mar. Pollut. Bull. 2000, 40, 1210-1214. (30) Zhang, G.; Parker, A.; House, A.; Mai, B. X.; Li, X. D.; Kang, Y. H.; Wang, Z. S. Environ. Sci. Technol. 2002, 36, 3671-3677. (31) Iwata, H.; Tanabe, S.; Sakal, N.; Tatsukawa, R. Environ. Sci. Technol. 1993, 27, 1080-1093. (32) Nunez G, M. A.; Estrada, I.; Calderon-Aranda, E. S. Toxicology 2002, 174, 201-210. (33) Garrison, A. W.; Nzengung, V. A.; Avants, J. K.; Ellington, J. J.; Jones, W. J.; Rennels, D.; Wolfe, N. L. Environ. Sci. Technol. 2000, 34, 1663-1670. (34) Raport, R. A.; Eisenreich, S. J. Atmos. Environ. 1986, 12, 23672379. (35) Hunt, W. G.; Johnson, B. S.; Thelander, C. G.; Walton, B. J.; Risebrough, R. W.; Jarman, W. M.; Springer, A. M.; Monk, J. G.; Walker, W., II Envrion. Toxicol. Chem. 1986, 5, 21-27. (36) Huang, X. Z.; Ji, Y.; Ye, J. M.; Shan, W. L. Jian, Q. Pestic. Sci. Admin. (in Chinese) 2000, 5, 9-12. (37) Tang, C. C.; Li, Y. C.; Chen, B.; Yang, H. Z.; Jin, G. Y. In Pesticide Chemitry; Nankai University Publishing House: 1998; p 230. (38) Brown, M. A.; Ruzo, L. O.; Casida, J. E. Bull. Environ. Contam. Toxicol. 1986, 37, 791-796. (39) Risebrough, R. W.; Jarman, W. M.; Springer, A. M.; Walker, W., II; Hunt, W. G. Environ. Toxicol. Chem. 1986, 5, 13-19. (40) Rasenberg, M. H. C.; van de Plassche, D. J. Dicofol Dossier prepared for the meeting March 17-19 in Norway of the UNECE Ad-hoc Expert Group on POPs, 4L0002.A1/R0011/EVDP, January, 2003. (41) Gillespie, M. J.; Lythgo, C. M.; Plumb, A. D.; Wilkins, J. P. G. Pestic. Sci. 1994, 42, 305-314. (42) Di Muccio, A.; Camoni, I.; Citti, P.; Pontecorvo, D. Ecotoxicol. Environ. Saf. 1988, 16, 129-132. (43) Bi, F. C.; Wang, W. L. Chin. J. Chromatogr. (in Chinese) 1995, 13, 281-283. (44) Draxler, R. R.; Hess, G. D. Description of the Hysplit-4 modeling system; NOAA Technical Memorandum ERL ARL-224; December, 1998. (45) Draxler, R. R.; Hess, G. D. Aust. Meteorol. Mag. 1998, 47, 295308. (46) Domagalski, J. Bull. Environ. Contam. Toxicol. 1996, 57, 284291. (47) Wallace, J. C.; Hites, R. A. Envirion. Sci. Technol. 1996, 30, 444446. (48) Lee, R. G. M.; Hung, H.; Mackay, D.; Jones, K. C. Environ. Sci. Technol. 1998, 32, 2172-2179. (49) Lee, R. G. M.; Burnett, V.; Harner, T.; Jones, K. C. Environ. Sci. Technol. 2000, 34, 393-398.
Received for review September 24, 2003. Revised manuscript received December 3, 2003. Accepted December 9, 2003. ES035052D
