Article pubs.acs.org/est
Monsoon-Driven Transport of Organochlorine Pesticides and Polychlorinated Biphenyls to the Tibetan Plateau: Three Year Atmospheric Monitoring Study Jiujiang Sheng,†,‡ Xiaoping Wang,*,† Ping Gong,† Daniel R. Joswiak,† Lide Tian,† Tandong Yao,† and Kevin C. Jones§ †
Key Laboratory of Tibetan Environment Changes and Land Surface Processes, Institute of Tibetan Plateau Research, Chinese Academy of Sciences, Beijing, 100101, China ‡ University of Chinese Academy of Sciences, Beijing 100049, China § Lancaster Environmental Centre, Lancaster University, Lancaster, LA1 4YQ, United Kingdom S Supporting Information *
ABSTRACT: Due to the influence of the Indian monsoon system, air mass transport in and to the Tibetan Plateau shows obvious seasonality. In order to assess the responses of atmospheric concentrations of persistent organic pollutants (POPs) to the Indian Monsoon fluctuation patterns, a three year air monitoring program (2008−2011) was conducted in an observation station close to the Yarlung Tsangpo Grand Canyon, southeastern Tibetan Plateau. The air concentrations of polychlorinated biphenyls (PCBs) and hexachlorocyclohexanes (HCHs) are generally comparable to those of other remote regions, whereas the concentrations of DDTs are much higher than reported for the polar regions, the North American Rocky Mountains, and the European Alps. The concentrations of DDTs and PCBs were strongly linked to the cyclic patterns of the Indian monsoon, displaying higher values in the monsoon season (May−September) and lower values in the nonmonsoon season (November−March). A “bimodal” pattern was observed for α- and γ-HCH, with higher concentrations in spring and autumn and lower concentrations in the summer (monsoon season). Rain scavenging in the monsoon season likely resulted in the lower HCH concentrations in the atmosphere. This paper sheds lights on the role the Indian monsoon plays on the atmospheric transport of POPs to the Tibetan Plateau.
■
and sea ice cover may influence the transport of POPs.13,14 With the aid of long-term air-monitoring programs, such as the Arctic Monitoring and Assessment Programme (AMAP) and the Integrated Atmospheric Deposition Network (IADN) in the Great Lakes region, climate impact that spans years or even decades has been investigated.15,16 El Niño−Southern Oscillation (ENSO) plays an important role in the air concentration of POPs in both the Great Lakes region and the Arctic.15 The high air temperatures associated with ENSO enhance the volatilization of POPs from the Earth’s surface “reservoirs” (e.g., soils and oceans). POPs are then available for LRAT from source regions to the more pristine regions. Fluctuations of the Arctic Oscilliation (AO), North Atlantic Oscilliation (NAO), and Pacific North American pattern (PNA) can change the contaminant pathways (wind flow) and further change the levels and trends of POPs in air15 and wildlife.17
INTRODUCTION Remote sites, such as the polar regions,1,2 open oceans,3,4 isolated islands,5 and the Tibetan Plateau (TP),6 were previously considered to be pristine environments where there were no direct emissions of pollutants. However, there has been increasing evidence that these areas are contaminated with certain chemicals, particularly persistent organic pollutants (POPs). Long-range atmospheric transport (LRAT) is an important pathway for the transport of POPs to these locations.7 With limited local emissions, the remote areas are considered to be ideal regions for investigating LRAT and the influence of climate change on the fate of POPs. For example, atmospheric pathways of polycyclic aromatic hydrocarbons (PAHs) to the Canadian High Arctic station of Alert were found to be mostly from East Asia, North Europe, and North America;8 King George Island, Antarctic was predominantly affected by emissions from South America;9 POP contamination in the TP was attributed to the emission from southern Asia sources and so forth.6,10−12 Generally, POP concentrations in remote regions vary with the seasons and prevailing winds. The climate-related variables, that is, temperature, wind speed and direction, precipitation, © 2013 American Chemical Society
Received: Revised: Accepted: Published: 3199
December 19, 2012 February 27, 2013 March 1, 2013 March 1, 2013 dx.doi.org/10.1021/es305201s | Environ. Sci. Technol. 2013, 47, 3199−3208
Environmental Science & Technology
Article
Figure 1. General pattern of the atmospheric circulation systems over the TP and the map showing the sampling site (Lulang).
covering consecutive sampling periods from 2008 to 2011, are presented. The relations between the seasonality of atmospheric OCPs and PCBs and the climate patterns (the strength and cycles of the Indian monsoon) are discussed. This paper aims to assess the role of the Indian monsoon on the transport of POPs to the TP. A better understanding of this process will certainly contribute to determining how global air circulation patterns/changes will affect contaminant transport.
Another climate system, the monsoon, operates on a seasonal scale, with seasonally variable wind directions and precipitation patterns. This leads to distinct seasonal differences in both the direction and extent of atmospheric transport of POPs. A previous study demonstrated that monsoon-driven variability exerts a strong influence on the inputs (vertical fluxes) of POPs in the Arabian Sea.18 The TP has an area of over 2.5 × 106 km2 and an average elevation higher than 4000 m. Air masses over the TP are mainly dominated by maritime air from the Indian Ocean (in the summer) and continental air from central Asia (in the winter).19 In the summer, the Indian monsoon brings water vapor from the Indian Ocean and provides moisture and precipitation on the TP,19 which feeds and generates approximately 46 000 glaciers.20 Ice core records clearly showed the features of the pollutants coming from South Asia, as good correlations have been found between the levels of pollutants in the ice strata and the air mass trajectories originating from South Asia.21 So far, the levels of organochlorine pesticides (OCPs) and polychlorinated biphenyls (PCBs) in air of the TP have been reported.6,22−24 Li et al.24 conducted a 10 day air sampling at the area close to Mt. Everest and found that the OCPs in air mainly originated from the India−Nepal−Pakistan region. Gong et al.23 found that the elevated concentrations of DDT in air in south Tibet were associated with the Eastern Asia monsoon system. Xiao et al.22 noted enhanced transport of POPs to the TP at Nam Co lake during the monsoon season by using a monthly integrated flowthrough air sampler (FTS). Wang et al.6 deployed passive air samplers (PAS) across the TP and found that the POPs concentrations from the Indian monsoon impacted regions are higher than other regions of the plateau. The sampling periods of the above studies were normally limited to a few weeks or months; on the other hand, the PAS employed by Wang et al.6 and the FTS utilized by Xiao et al.22 can only provide fewer data from the year-round exposure time. Air mass transport in the TP shows an obvious seasonality. We conducted a continuous ambient air-monitoring program at the Southeast Tibet Observation and Research Station (STORS), operated by the Chinese Academy of Sciences. In this study, the results from the first three years of sampling,
■
MATERIALS AND METHODS Sampling Site and Programs. A low-volume air sampler was deployed on the open plain of the STORS. This station (94.73° E, 29.77° N; 3330 m, Figure 1) is located at Lulang, in southeastern TP, and is 40 km west of the Yarlung Tsangpo Grand Canyon, the deepest and largest canyon on Earth. The climate of this station is mainly controlled by the Indian monsoon. The mean annual air temperature is 5 °C, and the mean annual precipitation is 900 mm, most of which falls during the Indian monsoon season. Due to the relatively lower elevation and the abundant water supply, the southeastern TP contains large areas of forest. Lulang is surrounded by a widespread of pine, spruce, and dwarf shrub (Rhododendron). The sampling period was from November 2008 to September 2011, and the samples were collected every 2 weeks. Polyurethane foam (PUF, 7.5 cm × 6 cm diameter) was used to collect the gas-phase OCPs and PCBs. The particulates were collected on glass fiber filters. The details on sampling program are provided in the Supporting Information (text SI1). Additionally, an automatic weather station at the STORS was recording the meteorological parameters, including the air temperature, wind speed and direction, air pressure, and relative humidity, every 10 min during the sampling period. Details regarding the sampling time and the average temperature for each sampling period are given in Table SI-1 (Supporting Information). Sample Extraction and Analysis. Each sample was transferred to the Soxhlet body and spiked with a mixture of surrogate standards (2,4,5,6-tetrachloro-m-xylene (TCmX), decachlorobiphenyl (PCB 209), 13C12-PCB 138, and 13C12PCB 180). The samples were Soxhlet-extracted using DCM for 3200
dx.doi.org/10.1021/es305201s | Environ. Sci. Technol. 2013, 47, 3199−3208
Environmental Science & Technology
Article
seasonal variations? Do the seasonal variations associate with the cycles of the Indian monsoon? It is critical to obtain the atmospheric levels and long-term trends of the POPs and compare the trends with the seasonally cycled Indian monsoon intensities. General Comments on the Ambient Concentrations. The full data set for the POP concentrations is available as Table SI-6 (Supporting Information). The concentrations of OCPs and PCBs in Lulang were compared with those found from remote polar and high mountain regions (Table 1).6,9,13,14,27−30 In this study, the mean concentration of hexachlorobenzene (HCB) was 7.87 pg m−3, which is lower than the average (52 pg m−3) for the Arctic region (Table 1).14 The total PCB concentrations (sum of PCB 28, 52, 101, 138, 153, and 180) were several picograms per meter cubed, which are comparable to the concentrations reported from the Andes mountains,27 the Arctic,14 and the Antarctic9 but are lower than those for the European mountain regions.28 The dominant PCB congeners were PCB 28, 52, and 101 (Supporting Information, Table SI-6). The heavier congeners with lower vapor pressures (PCB 138, 153, and 180) were always below the detection limits for the entire sampling period. The HCH concentrations ranged from 0.608 to 51.0 pg m−3 and from 0.025 to 7.07 pg m−3 for α-HCH and γ-HCH, respectively. These values are similar to those for other high mountain areas13,27,28 but are higher than those reported for the Antarctic region.9 This may be due to the closer proximity of Lulang to the source regions of India. However, these values are obviously lower than reported for the remote wetland regions of India (Table 1).30 DDTs are the most abundant OCPs in this study. The concentrations of the DDTs ranged (in pg m−3) 0.146−60.5 for o,p′-DDT, 0.146−33.7 for p,p′-DDT, 0.059−4.88 for o,p′-DDE, and 0.084−23.8 for p,p′-DDE. The levels of DDTs present a similar level to the Indian wetlands30 and are apparently higher than those of the polar regions9,14 and other mountain areas (Table 1).13,27,28 This indicates that the atmosphere of southeastern Tibet is likely receiving a considerable contribution of DDTs. It is important to note that data of PAS were used for the above comparison. This may lead to large uncertainties because of the uncertainties of PAS sampling rates and the possible accumulation of particles by PUF PAS. On the basis of the data from active air samplers (AAS) and FTS, which have the relatively accurate air volumes, the direct concentration comparison for different regions of the plateau are presented and discussed in Table SI-7 (Supporting Information). Seasonality of OCPs and PCBs. The Indian subcontinent has experienced extensive use of OCPs, including DDTs and HCHs, in the last decades.30−32 To combat vector-borne diseases, which are particularly prevalent in the monsoon season, India was allowed to continue to use DDT even after its global prohibition.30 Although the use of technical HCH (dominant with α-HCH) had been banned in 1997 in India, the continuous usage of this pesticide in Indian agricultural areas has been reported.30,33,34 India has two dominant agricultural seasons: (1) the peak summer season, where crops are sown at the beginning of the monsoon and harvested at the end of the monsoon; (2) the wheat and barley season, which begins from October to December and ends between February and April.35 Considering such seasonal agricultural practices, the monsoon season is also the season when OCPs are extensively used in India. Apart from OCPs, the improper
24 h. The extracts were concentrated and solvent-exchanged with hexane and were then purified on a chromatography column (from the top to bottom: 1 g of anhydrous sodium sulfate, 2 g of 3% deactivated alumina, and 3 g of 6% deactivated silica gel). The column was eluted with 30 mL of a mixture of DCM and hexane (1:1). The elute was further cleaned using gel-permeation chromatography (GPC, containing 6 g of Biobeads SX3), and the samples were finally solventexchanged and concentrated in 20 μL of dodecane containing a known quantity of pentachloronitrobenzene (PCNB) and PCB 54 as the internal standards. Details on the gas chromatographic temperature and the target compounds are given in text SI-2 (Supporting Information). Quality Assurance/Quality Control (QA/QC). All analytical procedures were monitored using strict QA/QC measures. Laboratory blanks and field blanks were extracted and analyzed in the same way as the samples. All of the chemicals were not detected in the laboratory blanks. p,p′DDE, α-HCH, β-HCH, γ-HCH, δ-HCH, HCB, PCB 28, and PCB 52 were detected in the field blanks. The details of the method detection limits (MDLs) and compound concentrations for each field blank sample are given in Table SI-2 (Supporting Information). The recoveries were between 70% and 130% for TCmX and between 75% and 120% for PCB 209. All of the reported values were field blank-corrected (mean blank concentrations were subtracted) but not corrected for the recovery rates. If the concentration of a compound after blank correction was below the MDL, the concentration was substituted with 1/2 MDL in cases where greater than 70% of data were greater than the MDL. The breakthrough of the OCPs/PCBs was estimated, and the concentrations of most of the chemicals in the second PUF plug were in the range 0−25% of both of the PUF plugs (Supporting Information, Tables SI-3 and SI-4), indicating a good retention capacity of the sampling program. The filter samples containing the particulate phase compounds were also analyzed (Supporting Information, Table SI-5). Due to the lower values, the particulate POP concentrations were negligible, and only the data of the PUF plugs representing the concentrations of gaseous OCPs and PCBs are reported (Supporting Information, Table SI-6).
■
RESULTS AND DISCUSSION Introductory Remarks. Monsoons are marked by seasonally reversing wind systems, and the Indian monsoon approximately initiates in May and fades away in September.25 During this period, the TP is generally regarded as an air pump, which plays an important part in strengthening the Indian monsoon by creating a low-pressure zone that sucks the moisture of Bengal Bay in to the plateau.19,25 The averaged speed and direction of the wind during the monsoon season are shown in Figure SI-1 (Supporting Information). On the southeastern edge of the TP, the Yarlung Tsangpo Grand Canyon cuts through the eastern Himalayas and bends toward the west (Supporting Information, Figure SI-2). This gorge provides a “channel” from the south to the north, which allows the moisture of the Bengal Bay to go directly up into the plateau. Lulang is in the vicinity (40 km west) of the bend of the Yarlung Tsangpo Grand Canyon. Therefore, the climate of this sampling site is closely related to the Indian monsoon variations. Although there are a few studies suggesting the possibility of the monsoon-driven atmospheric transport of contaminants to the TP,6,22,26 direct evidence is still limited. Do the atmospheric POP concentrations in this region show 3201
dx.doi.org/10.1021/es305201s | Environ. Sci. Technol. 2013, 47, 3199−3208
29 28 13 27 30
7.5 32.9 4.5 3.3 6
AAS: active air sampler; PAS: passive air sampler; BDL: below detection limit. b∑ 6 PCBs. c∑ 15 PCBs. d∑ 28 PCBs. e∑ 28 PCBs. f∑ 205 PCBs. g∑ 48 PCBs. h∑ 10 PCBs. I∑ 8 PCBs. The average value and the concentration ranges (in brackets) were listed.
3.2(1.4−5.7) 5 (2−12) 3.3 (0.9−12) 5.3 (1−9)
19.7 (8−47) 3.3 (2−8) 10.0 (0.4−45) 4.0g 5.0 14.1 14.5
238(120−320)d 25(20−31) 61(34−100) 25(9−45) 52(BDL-78) 13(6−19)
2005 AAS 52 0.0 (0.060−120) 4.9 (2.0−38.0)e 13 (5.2−33.0) 1.7 (0.67−3.7) 0.14 (0.10−0.41) 0.22 (0.07−0.86) 0.3 (0.1−1.2) 0.034 (0.022−0.17) 14 2006 PUF PAS
2007−2008 XAD PAS 19.5 5.5c 1.8
2008−2011 AAS 7.87 (0.051−27.1) BDL−16.7b 12.1 (0.606−51.0) BDL−7.07 BDL−33.6 BDL-60.5 6.40 (0.084−23.8) BDL−7.3 this study
sampling site time (year) sampler HCB PCBs α-HCH γ-HCH p,p′-DDT o,p′-DDT p,p′-DDE o,p′-DDE ref
management of ship-breaking activities and electronic waste recycling are the major sources of PCBs in India.30 As discussed above, under the favorable atmospheric transport system (the Indian monsoon), the contaminant levels at Lulang (southeastern TP) might be influenced by these emission sources and might display seasonal variations like the monsoon fluctuation. DDTs. On the basis of the three year monitoring period, the concentrations of DDT class compounds exhibit an obvious seasonal fluctuation, with the higher levels occurring in the monsoon season (May−September) and the lower levels occurring in the cold season (Figures 2a and SI-3, Supporting Information). In some winter samples, the concentrations of p,p′-DDT and o,p′-DDT were even below the detection limit, whereas their concentrations reached up to 33.7 and 60.5 pg m−3, respectively, in the summertime (Table SI-6, Supporting Information, Figures 2a and SI-3, Supporting Information). The variation in the DDTs concentrations was not random but had a clear periodicity. During this three year monitoring period, the concentrations of DDT chemicals (p,p′-DDT, p,p′DDE, o,p′-DDT, and o,p′-DDE) peaked in August and had a minimum in February (Figures 2a and SI-3, Supporting Information). The atmospheric DDT concentrations were plotted versus time, and a sine wave was fitted to the data. The equation for this curve is given by sin(c , t ) = A*sin[π *(t − B)/C ] + D
(1)
where sin(c, t) is the concentration of chemicals at time t (in two weeks), A is the amplitude of the wave, B is the time offset, C is the wave width, and D is the vertical offset. These fitting parameters are given in Table SI-8 (Supporting Information). The fit had an r2 value ranging 0.18−0.52. In terms of the fitting parameters, the B values of DDTs are ∼7 and those for DDEs are ∼5, which may be due to the possible influence of degradation/secondary sources on the seasonalities of DDEs (degradation products). The parameters B and C represent the date when the maximum concentration occurs and the frequency (wave width), respectively. Considering the fitting uncertainty, the B and C values for the DDT class chemicals are analogous. This indicates the cyclic pattern of the different DDT chemicals is similar (Supporting Information, Table SI8). As it is well-known that the Indian monsoon fluctuates periodically during the year, it is reasonable to assume that the atmospheric DDT concentrations in Lulang will fluctuate with the monsoon variation. To verify if this is true, we investigated the variations of the Indian monsoon index (IMI, Supporting Information, Table SI-1) over this three year period.36 Details on the IMI calculation are provided in text SI-3 (Supporting Information). The IMI shows a clear periodicity (Figure 2a) and characterizes the prevailing monsoon episodes. The fitted sine wave for the seasonal IMI is shown in Figure 2a, and the B and C values are listed in Table SI-8 (Supporting Information). When comparing the obtained sine wave parameters, we found that the monsoon wave and the DDT waves had a similar wave width (Supporting Information, Table SI-8). There was no significant difference (t test, significance
