Environ. Sci. Technol. 2008, 42, 1928–1932
Air–Water Gas Exchange of Organochlorine Pesticides in Taihu Lake, China XINGHUA QIU, TONG ZHU,* FENG WANG, AND JIANXIN HU State Key Joint Laboratory for Environmental Simulation and Pollution Control, College of Environmental Sciences and Engineering, Peking University, Beijing 100871, China
Received July 23, 2007. Revised manuscript received December 5, 2007. Accepted December 21, 2007.
Previous research in the Taihu Lake Region (TLR) of China found high levels of atmospheric organochlorine pesticides (OCPs). To understand the sources and the environmental behaviors of these OCPs in the TLR, research on air–water gas exchange was performed in 2004. Hexachlorocyclohexanes (HCHs), DDT related compounds (DDTs), cis-chlordane (CC), trans-chlordane (TC), heptachlor (HEPT), and R-endosulfan in both air and water samples were analyzed, and air–water gas exchange fluxes of these compounds were calculated. The net volatilization flux of R-HCH was 58 ng m-2 day-1, suggesting that the residue of technical HCH in the lake sediment might have been an important source of R-HCH to the air of this region after the ban of technical HCH two decades ago. The main components of technical chlordane, TC, CC, and HEPT, each had net volatilization fluxes >230 ng m-2 day-1, suggesting that waste discharge from manufacturing plants in the upper region was the main source of chlordane to the lake. Unlike R-HCH and chlordane, o,p′-DDT and R-endosulfan had net deposition fluxes, suggesting that these compounds were transported through the atmosphere from land sources and then deposited into the lake. The correlation between air concentrations and ambient air temperature indicated that the current sources of o,p′-DDT and R-endosulfan were from land; R-HCH and chlordane were mainly from the lake.
Introduction China began to produce organochlorine pesticides (OCPs) in the 1950s, and this production continued until they were banned in the early 1980s. However, a number of OCPs are still being produced for nonagriculture purpose, such as technical dichloro-diphenyl-trichloroethane (DDT) for malaria control and for dicofol production (1). Some OCPs were reintroduced later, such as chlordane in 1988 (2), and some were produced for the first time later, such as lindane in the early 1990s (3). Previous studies showed that the environmental levels of OCPs in China were still high (4, 5). It would be interesting to know where these OCPs come from; however, little information on the sources and environmental behaviors of these OCPs was available. Air–water gas exchange research had been used to provide important information about the source and fate of semivolatile organic compounds (6), and * Corresponding author e-mail: 86(10)62754789; fax: 86(10) 62751927. 1928
9
[email protected];
phone:
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 42, NO. 6, 2008
this approach could help us to understand the source and fate of those OCPs that had been observed in the air around Taihu Lake. Taihu Lake is the third largest fresh water lake (with an area of 2338 km2) in China, and it is the main drinking water source of the megalopolises in this basin. These include Shanghai, Wuxi, and Suzhou, all of which have millions of people (see Figure SI-1 in the Supporting Information). Because significant amounts of OCPs were used in this region before 1983 (3, 7), residues of these pesticides might be elevated in the sediment of the lake ((8), see Table SI-1 in the Supporting Information). The lake is relatively shallow (with an average depth of 1.9 m); hence, OCP residues in the sediment could be released into the water, as a result of water turbulence generated by wind and boats, and then volatilized into air. Thus, the lake sediment might be a source of some OCPs. On the other hand, some OCPs could volatilize from the land where they were previously and/or currently used, then enter the lake through atmospheric transport and deposition. Thus, the lake might be the sink of these OCPs. Air–water gas exchange fluxes of these pesticides could indicate their sources. This paper reports the results of OCP concentrations in air and water samples, as well as the calculated fluxes of air–water gas exchange in Taihu Lake. The sources of these compounds and their environmental behaviors are also discussed.
Experimental Section Sample Collection. Air samples were collected at the Taihu Laboratory for Lake Ecosystem Research (TLLER) site (4) on the northern shore of Taihu Lake (see Figure SI-1) from 2004 to early 2005 for four seasons: spring (Mar 29 - Apr 16), summer (Jul 23 - Aug 11), autumn (Oct 22 - Nov 8), and winter (Dec 23 - Jan 4). The winter air samples were collected for 23 h duration (from 17:30 to 16:30 of the following day). For all other seasons, the air was sampled for 11 h (from 7:00 to 18:00 for daytime samples and from 19:00 to 6:00 of the following day for nighttime samples). The sampling flow rate was 28 m3 per hour. Detailed information on the sampling procedures and locations has been given elsewhere (4). Water samples in the winter were collected only in the daytime; however, in the other seasons they were collected in both the daytime and nighttime. The surface water (0–10 cm) under a fixed platform (see Figure SI-1) was collected with a glass bottle and was filtered with stainless steel mesh (50 µm aperture) to remove algae and some suspended particulate matter. The water samples were stored in a stainless steel tank and then passed through a glass XAD-2 adsorptioncolumnpumpedbyaperistalticpump(BT00–600M, Baoding Longer Precision Pump Co., China) at the end of the adsorption column. Each column, with an internal diameter of 2.6 cm, was filled with 100–120 cm3 of XAD-2, and plugged with a thick layer of glass wool (Shanghai Experiment Reagent Co., China, pretreated at 450 °C for 4 h). Water sampling volumes were around 100 L, which were determined by adding the volume of water taken with the glass bottle and deducting the volume of water left in the tank when sampling was completed. After sampling, the glass wool was removed, and the XAD-2 was washed two or three times with 300 mL of deionized water. XAD-2 was stored in a 125 mL glass bottle with a Teflon lined cap, which was sealed in a polythene bag and stored below -10 °C to await extraction. Before sampling, XAD-2 (Amberlite, 20–60 mesh) was washed with deionized water and then Soxhlet extracted 10.1021/es071825c CCC: $40.75
2008 American Chemical Society
Published on Web 02/08/2008
TABLE 1. Air–Water Gas Exchange Fluxes of HCHs, Chlordane, DDTs, and Endosulfan during Four Seasons in 2004 (Mean ± One Standard Deviation, in Units of ng m-2 day-1) and the Estimated Net Volatilization (Positive Value) or Deposition (Negative Value) Rates of These Compounds over the Whole Lake in 2004 Spring wind speed (m s-1)(mean ( SD) R-HCH γ-HCH HEPT TC CC p,p′-DDT o,p′-DDT o,p′-DDD R-endosulfan
Summer
3.0 ( 1.5
3.1 ( 1.1
99 ( 39 9.5 ( 4.5 n.d. c 5.5 ( 3.0 11 ( 10 n.d. -9.2 ( 9.9 1.3 ( 0.6 n.d.
64 ( 21 4.9 ( 3.8 390 ( 190 620 ( 300 400 ( 200 n.d. -120 ( 160 2.7 ( 1.9 -25 ( 77
Autumn
Winter
3.4 ( 1.3
4.4 ( 2.4
29 ( 7 1.9 ( 2.1 81 ( 79 280 ( 260 170 ( 160 -1.2 ( 1.9 -14 ( 14 0.7 ( 0.4 -8.5 ( 17.6
26 ( 14 1.9 ( 1.1 160 ( 120 350 ( 380 190 ( 210 -0.23 ( 0.25 -1.7 ( 0.8 0.05 ( 0.08 n.d.
average of the four seasons a
exchange rate of the whole lake in 2004 (kg)b
3.4 ( 1.6 58 ( 38 4.9 ( 4.5 230 ( 210 340 ( 350 230 ( 210 -0.9 ( 1.6 -39 ( 97 1.3 ( 1.5 -17 ( 56
49 4.1 150 290 200 -0.4 -33 1.1 -7.3
a Annual average value excluded the seasons for which there was no data. b Area of Taihu Lake is 2338 km2. c n.d., target compounds were not detected in water, thus the flux of air–water gas exchange was not calculated.
sequentially with acetone, n-hexane, acetone, and methanol, each for 22 h. The cleaned XAD-2 was stored in 125 mL bottles and loaded in the adsorption column before sampling. Sample Preparation. A previously used preparation method for air samples (4) was used in this research, with the exception that n-hexane replaced the mixed solvent (nhexane and diethyl ether) used for Soxhlet extraction. The XAD-2 used to adsorb the organic compounds from the water samples was Soxhlet extracted with n-hexane:acetone (1:1, v/v) for 22 h. After Soxhlet extraction, samples were rotatory evaporated and then extracted four times with n-hexane in a separatory funnel to remove any residual water. All the other steps were the same as detailed for the air samples (4). Information on solvents and chemical standards can be found in Qiu et al. (4). New standards were added, including o,p′-DDE and o,p′-DDD from Dr. Ehrenstorfer GmbH, Germany; cis-chlordane (CC) and trans-chlordane (TC) from AccuStandard (New Haven, Connecticut). Sample Analysis. A gas chromatograph (GC) with an iontrap mass spectrometer (MS) (Finnigan Trace GC/Polaris Q) was used to analyze the target compounds (4). The mass spectrometer was further optimized, especially for the autumn and winter samples, by adding a damping gas (highly purified helium, 3 mL min-1) (9). In this manner, the instrument detection limit (IDL) was decreased from 3.5-38.5 pg (4) to 0.3–1.3 pg (0.9, 0.9, 0.5, 0.5, 0.6, 0.3, 0.5, 0.4, 0.5, 1.3, and 1.3 pg for R-HCH, γ-HCH, p,p′-DDT, o,p′-DDT, p,p′DDE, o,p′-DDE, o,p′-DDD, TC, CC, HEPT, and R-endosulfan, respectively). The QA/QC methods have been described elsewhere (4), and the results are given in Table SI-2 as Supporting Information.
Results and Discussion Air–water Gas Exchange Calculation. The measurement results, including the concentrations of target compounds in air and water and concentration ratios of related compounds are listed in Table SI-3 of the Supporting Information. Because measurements of p,p′- and o,p′-DDE had some interferences in the GC/MS analysis (1), they are not discussed in this study. The two-film model was used to estimate the flux of air–water gas exchange (6), F ) Kol(Cw - CaRT ⁄ H)
(1)
where R is the gas constant (8.3 Pa m3 mol-1 K-1), T (K) is the absolute temperature, and H (Pa m3 mol-1) is the Henry’s law constant. Cw and Ca are the dissolved and gaseous concentrations of the target compounds, which are considered to be the measured air and water concentrations.
The arithmetical mean of the water concentration in each season was taken; thus, Cw was assumed to be constant during each season. Kol (m s-1) is the overall mass transfer coefficient, which contains contributions from the mass transfer coefficients of the water layer and the air layer, namely kw and ka. The value of Kol is given by eq 2, 1 RT 1 1 ) × + Kol H ka kw
(2)
ka ) 1 × 10-3 + 46.2 × 10-3U/Sc(a)-0.67
(3)
where
kw ) 1 × 10-6 + 144 × 10-4(U/)2.2Sc(w)-0.5 (U* < 0.3 m s-1) (4) kw ) 1 × 10-6 + 34.1 × 10-4U/Sc(w)-0.5 (U* > 0.3 m s-1) (5) U/ ) [(0.61 + 0.063U10) × 10-3]0.5U10
(6)
and Sc(a) and Sc(w) are the air and water phase Schmidt number (dimensionless). The value of Sc(a) was 2.9 for HCHs (10) and 2.0 for the other compounds (11, 12). Sc(w) was set to 1000 for all the target compounds (11, 12). U10 (m s-1) is the wind speed at 10 m height, which was provided by the Taihu Laboratory of Lake Ecosystem Research (TLLER). The wind speed used for each air sample was the arithmetical mean of the hourly wind speed. The Henry’s law constant is a function of temperature, logH ) m ⁄ T + b
(7)
where m is the temperature adjusted slope. The literature sources of the H values and the temperature-adjusted H values are listed in Table SI-4 of the Supporting Information. Table 1 shows the resulting air–water gas exchange fluxes and the estimated volatilization or deposition rates of selected target compounds in Taihu Lake for 2004. The seasonal fluxes of four representative compounds, R-HCH, TC, o,p′-DDT, and R-endosulfan, are also shown in Figure 1. HCHs. Technical HCH had been a widely used pesticide all over the world and contains R-HCH as its main component. Even after the ban of technical HCH 20 years ago, the air concentrations of R-HCH in the Taihu Lake Region (TLR) were 74 pg m-3 in the summer of 2002 (4) and 40 pg m-3 in the summer of 2004. These concentrations were about the same as those observed around the North American Great Lakes in recent years (13). This suggested that, although technical HCH was heavily used in the past in the TLR (3), the atmospheric VOL. 42, NO. 6, 2008 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
1929
FIGURE 1. Air–water gas exchange fluxes of four representative target compounds. Sp, Su, Au, and Wi represent spring, summer, autumn, and winter of 2004, respectively. The error bars represent one standard deviation of the fluxes in each season. A positive value means net volatilization flux from the lake to the air, and a negative value means net deposition flux from the air to the lake. concentrations of R-HCH have decreased rapidly, probably because there are no new sources of technical HCH. Although the concentration of R-HCH in the air was low, its concentration in the lake water was high. These levels were similar to those observed in Lake Ontario from the late 1980s to the early 1990s (14). Because of these high water concentrations, R-HCH had a net volatilization flux out of the lake, with an average value of 58 ng m-2 day-1 in 2004. This flux was higher than that observed in Lake Ontario in 1993, which was close to volatilization-deposition equilibrium (14). If the flux was assumed to be the same over Taihu Lake throughout a year, 49 kg of R-HCH would have volatilized into the air from the lake in 2004 (Table 1). However, the total R-HCH dissolved in the water of the lake was only 8 kg, much less than this volatilization amount. Because there is no known current usage of R-HCH, the residue of technical HCH in the lake sediment should be the source of R-HCH to the lake water. On the other hand, residue of technical HCH in soil might also enter the lake through surface runoff, and thus be the source R-HCH to the lake water. In the environment, R-HCH only comes from the use of technical HCH, whereas γ-HCH comes from both technical HCH and lindane (almost pure γ-HCH). The air concentration ratio of R-HCH to γ-HCH, (R/γ)air, has been used to determine whether γ-HCH in the environment originates from technical HCH or from lindane. The (R/γ)air value was 2.4 in the summer, and 2.9 in all four seasons of 2004. These numbers were lower than the concentration ratio of R-HCH to γ-HCH in technical HCH, which is in the range of 4–7, indicating that γ-HCH in the air over the lake may come from both technical HCH and lindane usage. As discussed above, the residue of technical HCH in the lake sediment might be an important source of R-HCH to the air. Based on the flux values in Table 1, the net volatilization flux ratio of R-HCH/γ-HCH was 12:1. If γ-HCH had the same source as R-HCH, that is, from the lake, then the (R/γ)air should be close to 12. The fact that (R/γ)air was lower than 3 suggested that, in addition to the residue in the lake, γ-HCH could also originate from terrestrial sources of lindane. As our findings from the field investigation in 2003 revealed, lindane was recently used in the wheat production region of North China. Unlike (R/γ)air, in water from Taihu Lake, the concentration ratio of R-HCH to γ-HCH [(R/γ)water] was 3.9, a value close to the concentration ratio of R-HCH to γ-HCH in technical HCH. This observation is different from the southern Baltic Sea, where (R/γ)water was lower than 1.0 because of the use of lindane in the coastland (15). The 1930
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 42, NO. 6, 2008
(R/γ)water value suggests that HCHs, including γ-HCH, in Taihu Lake might be primarily from residuals of technical HCH and that there was no current sources of lindane in the lake water, that is, there are no upstream inputs of lindane into the lake. Because of the ban on the use of technical HCH in 1983, the residues of R-HCH are now low in the TLR, and residues in the sediment have become an important source of R-HCH in this region. γ-HCH might also come from atmospheric transport of currently used lindane outside the TLR. Chlordane. Technical chlordane is a complex mixture of many congeners, including trans-chlordane (TC), cis-chlordane (CC), and heptachlor (HEPT). The historical production of HEPT in China was low, and there has been no production of HEPT at all in recent years. Thus, HEPT in the Chinese environment mainly comes from technical chlordane. The concentration ratio of CC to TC (the CC/TC ratio) has been used to distinguish short- and long-range atmospheric transport of chlordane, because TC is liable to depletion during atmospheric transport (16). The average CC/TC ratio was found to be 0.56 in this research, a value similar to that observed in Columbia, South Carolina (0.58) (16) and in Alabama (0.56) (17), USA. This value was also similar to a value predicted from the composition of technical chlordane based on the vapor pressures of these two components (0.54) (16). Taken together, these observations suggest that chlordane has a local source in the TLR. The water concentrations of chlordane were low in the spring. But they were quite high in the other three seasons, especially in the summer (with the highest concentration of 2100 pg L-1 for TC). The concentration of chlordane in fish from the lake was much higher than those from nearby regions (18), and this might be attributed to the high concentrations of chlordane in Taihu Lake. Because of the high concentrations in water, TC, CC, and HEPT had strong net volatilization fluxes out of the lake. The average net volatilization flux was 340, 230, and 230 ng m-2 day-1 for TC, CC, and HEPT, respectively, in 2004. For HEPT this flux was the average value over three seasons because it was not detected in the water in the spring. The estimated net volatilization flux of TC, CC, and HEPT were 640 kg for the whole Taihu Lake in 2004 (Table 1), indicating that the lake was seriously polluted with technical chlordane. The net volatilization fluxes and low water concentrations in the spring suggested there was a strong and seasonally dependent source for the chlordane in the Taihu Lake water. Given that technical chlordane was mainly used for termite control in residential and commercial buildings in China (19), it would be difficult for chlordane to enter into the lake by way of atmospheric transport and deposition. Thus, chlordane in Taihu Lake water was more likely coming from waste discharged by its manufacturing plants. As of December 2004, there were nine technical chlordane manufacturers in China, and eight of them were located in the TLR, including four located in the upper reaches of the lake (19). These manufacturers were all small chemical plants, and the waste discharged from these plants could easily enter the lake. Moreover, because of the production status and the upstream transport capacity, the seasonal water concentration of chlordane might vary, with the highest concentrations occurring in summer. DDTs. o,p′-DDT is the main DDT impurity in commercial dicofol formulation (the acaricidal activity ingredient of dicofol is 2,2,2-trichloro-1,1-bi(4-chlorophenyl) ethanol). Our previous research had shown that after the ban of technical DDT, dicofol became an important source of DDTs to the environment, especially in air (1, 4). This study showed that the air concentrations of DDTs were high in 2004, especially for o,p′-DDT, which was as high as 380 pg m-3 in the summer. This observation indicated that, even though the Chinese
TABLE 2. ∆Hex Values Observed in Taihu Lake (Mean ± One Standard Deviation) and Reported Values of ∆Hvap, ∆Hoa, and ∆Hwa (in kJ mol-1) four seasons of 2004a detectable sample No.
-∆Hex/R
r2
p
∆Hex
∆Hvapb
∆Hoac
∆Hwa
68 68 58 46 55 60 68 47 49
3970 ( 660 4530 ( 710 6560 ( 660 6680 ( 680 4510 ( 790 7810 ( 800 10700 ( 820 6990 ( 790 9450 ( 1300
0.36 0.38 0.64 0.69 0.38 0.63 0.72 0.63 0.53
