Water pollutant fingerprinting tracks recent industrial transfer from

Transcription

1 Water pollutant fingerprinting tracks recent industrial transfer from coastal to inland China: A case study Weiwei Zheng 1 *, Xia Wang 1 *, Dajun Tian 1 *, Songhui Jiang 1, Melvin E. Andersen 2 Gengsheng He 1, M. James C. Crabbe 3, Yuxin Zheng 4, Yang Zhong 5, 6 & Weidong Qu 1, 7 1 Key Laboratory of Public Health Safety, Ministry of Education, Department of Environmental Health, School of Public Health, Fudan University, Shanghai , China. 2 Institutes for Chemical Safety Sciences, The Hamner Institutes for Health Sciences, Research Triangle Park, North Carolina 27709, USA. 3 Institute of Biomedical and Environmental Science & Technology, Faculty of Creative Arts, Technologies and Science, University of Bedfordshire, Luton LU1 3JU, United Kingdom. 4 National Institute of Occupational Health and Poison Control, Chinese Center for Disease Control & Prevention, Beijing , China. 5 Institute of Biodiversity Science and Institute of High Altitude Medicine, Tibet University, Lhasa , China. 6 School of Life Sciences, Fudan University, Shanghai , China. 7 Center of Global Health, School of Public Health, Fudan University, Shanghai , China. * These authors contributed equally to this work.

2 Correspondence and requests for materials should be addressed to Y. Zheng Y. Zhong and W. Qu

3 1. Supplementary Information 1: Parallels between pollution evolution and life evolution 2. Supplementary Information 2: Industrial export and accept regions 3. Supplementary Information 3: IUR distance matrix 4. Supplementary Information 4: Analytical procedures for whole pollution fingerprints in water 5. Supplementary Information 5: Analytical procedures for target compounds in water 6. Supplementary Tables S1-S9 7. Supplementary Figures S1-S3

4 Supplementary Information 1: Parallels between pollution evolution and life evolution 1.1 Changes of components with pollution evolution resemble the change process of biological sequences with organism evolution. In the process of pollution evolution, the representative changes are the concentration fluctuation of compounds, reflecting the comprehensive actions of hydrographic conditions and pollution release. Over large temporal-spatial scales, the concentration changes of pollution components characterize the combination of random and environmental effects. The concentration range of each component represents steady, sharp fluctuating, emerging (be released or be formed) and disappearing (be purified or be transformed); four kinds of status. Therefore, the composition of the pollution fingerprints can be unchanged, changed, inserted (emerging compound) and deleted (compounds disappeared or degradation by self-purification) with the evolution process. In some degree, this resembles the composition changes of biological sequences in the process of organism evolution. 1.2 Similarities between pollution fingerprints and biological sequences. A biological sequence may consist of thousands or tens of thousands of components (amino acids, nucleotide bases or genes), while water pollution may contain thousands of compounds. Genes from different species may descend from the same or different ancestor, while the compounds from various water bodies may

5 originate from identical or different pollution sources. The relationship between two proteins (genes) can be inferred from computing the sequence identity between them based on the alignment of their amino acid (nucleotide) sequences. Similarly, the relationship between two complex water pollution profiles can be derived from assessing the identical or overlapped compounds and their level differences among their complex components. 1.3 Constructing origin relationships among water pollution is similar to reconstructing evolutionary relationships among species. Because of the complicated release of pollution sources and incessant changes of water flow, the origin of water pollution in different regions and periods is difficult to trace. However, the local water pollution fingerprints within a certain period can also reflect the pollution origin characteristics. Similar pollution fingerprints of water samples can be considered to have similar pollution origins. Therefore, we can construct origin relationships among water pollution in different regions and periods based on comparing similarities or dissimilarities between water pollution fingerprints. Similarly, in essence, reconstructing evolutionary relationships among various species is also based on comparing similarities between biological sequences. Biological sequence compositions can represent the origin characteristics, and the similarities or dissimilarities between sequences denote the divergence between species if the evolutionary rate is constant.

6 To sum up, we can construct a pollution tree based on comparing similarities or dissimilarities between pollution profiles to manifest origin relationships among water pollution in various regions and periods, referring to the tree-building principles and methods which have been widely used in reconstructing relationships among biological sequences.

7 Supplementary Information 2: Industrial export and accept regions In recent decades, owing to their early-development advantages, the Yangtze Delta and Guangdong province (Pearl River Delta) are the two major and important industrial export regions in China. To face the challenges of rising labor and resource costs, industrial structural adjustment and environmental burdens, these two highly developed regions have exported labor-intensive and resource-intensive industries to undeveloped inland regions. Based on graphical proximity, the enterprises in Guangdong province mainly transferred to the undeveloped north regions of Guangdong 1. Yangtze Delta mainly transferred industries to the neighboring central regions; Henan is one of the most important regions for accepting enterprises transferred from Yangtze Delta. Since 2004, Henan has experienced a sharp annual rise in the number of industrial enterprises and gross output values, which respectively has doubled to units and increased 400 billion dollars in 2009 (Supplementary Fig. 1 and Supplementary Fig. 2) 2. The expanding scale of industry in this undeveloped region benefited from transferred investment and industrial acceptance and has caused horrific wastewater discharge increasing to more than 1400 million tons in 2009 (Supplementary Fig. 3) 2. Our selected sampling localities (HPRA, HPRB, HPRC, HPRD and HPRE, see Fig. 1) are in the south of Henan and belong to the Huai River system. Up to now, these regions absorb very low amounts of investment from foreign countries, which take a small proportion of the total foreign investment in the whole Henan province. The main investment for developing local economy is from domestic developed

8 regions. Based on available data from the local Trade and Industry Bureau in these localities, most of the transferred enterprises are from the Yangtze Delta.

9 Supplementary Information 3: IUR distance matrix 3 In our previous research 3, we identified and compared different tree building methods for clustering pollution fingerprints of water samples from different regions during various periods. We showed that employing a neighbor-joining (NJ) method based on the intersection and union ratio (IUR) distance and maximum parsimony (MP) and maximum likelihood (ML) methods a converted binary data matrix can acquire effective and robust source-related clusters for pollution fingerprints of water samples. In this study, IUR-distance-based, MP and ML tree-building approaches were used to reconstruct the pollution origin relationships between water bodies in industrial export and acceptance regions. In similarity comparison and pollution tree construction for pollution origin relationships, the critical process is defining and calculating the distance or similarity matrix of all water fingerprints. To quantitatively represent the similarity or dissimilarity between complex water fingerprints, we defined an IUR distance. For comparison between every two fingerprints, the similarity can be denoted by the ratio of the number in the intersection compound group to the number in the union compound group. The intersection is the number of the common compounds between every two fingerprints. The union is the sum of all detected peaks in two mixtures, subtracted by the number of intersection compounds. The increased ratio of intersection compounds to union compounds reflects increasing similarities between mixtures. However, the differences of common compound levels should also be considered in a similarity comparison. We calculated the coefficient of variation

10 (CV) of each intersection compound between every two mixtures: the IUR was weighted by the CV of each common compound. This adjusted IUR distance reflects not only the similarity in compound composition, but the similarity in compound level between two samples. For instance, if the CV value increases (which indicates the increased deviation in compound level between two samples), the IUR distance value will correspondingly increase, showing that the two samples are less similar. Detailed descriptions and formulae are in Supplementary Table S3 3. On average, between every two fingerprints of water samples, 148 common compounds (ranging from 115 to 192) were matched. The calculated union compound amounts between every two water pollution profiles were The common compounds accounted for about 1/4-1/3 of all union compounds between every two water pollution samples. The IUR distance between every two water pollution fingerprints ranged from to In this study, differences in tree topologies were small between the IUR-distance-based tree and MP/ML tree, showing the similar source-related clusters of water pollution fingerprints (Fig. 2).

11 Supplementary Information 4: Analytical procedure for whole pollution fingerprints in water 3, Preparation of pollutant extracts and concentrates of water Each water sample (40 L) was first filtered through a 0.45 μm glass fiber membrane. The organic compounds were extracted from a water sample by passing water sample through a cartridge containing 10g XAD-2 which was pre-conditioned with methanol and acetone in turn; then washed with ultra-pure water (Milli-Q, Academic A10). The water samples flowed through the cartridge at a rate of 2 ml/min. The cartridge was eluted with 30 ml methanol and followed by 30 ml acetone. The eluent was reduced to <1 ml under a stream of nitrogen gas. Each sample volume was adjusted with methanol so that the organic extract from 1 L of water was equal to 10 μl of methanol. 400 μl organic extract (40 L-eq) was directly adjusted with methanol to the total volume of 1 ml (40,000X concentrate). 4.2 Whole-pollution-fingerprints analysis. The whole pollution fingerprints of extracted water samples were analyzed by gas chromatography-mass spectrum (GC-MS) (GC-MS QP2010, Shimadzu Instruments, Japan) with the 30-m fused silica capillary column (0.25 mm i.d., 0.25 μm film thickness; Agilent, USA). The GC oven program was: 40 ºC for 1min, 30 ºC/min to 130ºC (for 3min), 12 ºC/min to 180 ºC, 7 ºC/min to 240 ºC, 12 ºC/min to 300 ºC (for 5 min). The ion source temperature of MS was 280 ºC and GC interface temperature was 300ºC. A scan range of m/z was used for full scan analysis of

12 samples. All the peaks resolved by GC-MS were included in the whole pollution fingerprints of each water sample. The retention time, peak area and mass information of each peak were collected for subsequent data matrix establishment and pollution-tree reconstruction. The antibiotics and steroids in the pollution fingerprints were identified by liquid chromatography-tandem mass spectrometry (LC-MS/MS). The detailed information for the LC-MS/MS procedures was in Supplementary Information 5.4 and Supplementary Information 5.5.

13 Supplementary Information 5: Analytical procedure for target compounds in water To further identify the industrial-transfer types, we detected phthalate esters (PAEs), polychlorinated biphenyls (PCBs), polybrominated diphenyl ethers (PBDEs), sulfonamide and steroid compounds by their respective methods. 5.1 PAEs detection List of target PAEs 16 kinds of PAEs were detected, including: Dimethy phthalate (DMP), diethyl phthalate (DEP), diisobutyl phthalate (DIBP), di-n-butyl phthalate (DBP), bis(2-methoxyethyl) phthalate (DMEP), bis(4-methoyl-2-pentyl) phthalate (BMPP), bis(2-ethoxyethyl) phthalate (DEEP), diamyl phthalate (DPP), dihexyl phthalate (DHXP), butyl benzyl phthalate (BBP), hexyl 2-ethylhexyl phthalate, bis(2-n-butoxyethyl) phthalate (DBEP), dicyclohexyl phthalate (DCHP), bis(2-ethylhexyl) phthalate (DEHP), di-n-octyl phthalate (DNOP), dinonyl phthalate (DNP) Sample preparation A large-volume (20L) of water was used for each sample to extract target PAEs by passing the sample water through a cartridge containing 10 g XAD-2, which was pre-conditioned with acetone and methanol in turn; then washed with ultra-pure water. The water samples flowed through the cartridge at a rate of 2 ml/min. The cartridge was eluted with 20 ml methanol followed by 20 ml acetone. The eluent was reduced to ml under a stream of nitrogen gas. The extract (20 L-eq) was directly

14 adjusted with methanol to the total volume of 1 ml (20,000X concentrate). The internal standards were added in the extract before GC-MS detection GC-MS analysis Determination of concentrations of PAEs was carried out with GC-MS in the selective ion-monitoring (SIM) mode (the detailed instrumental information and oven program see Supplementary Information 4.2). The GC-MS parameters for the analysis of target PAEs were showed in Supplementary Table S5. The limit of detection (LOD) of all target PAEs ranged from ng/l (Supplementary Table S5) Quality assurance (QA) and quality control (QC) Only glass centrifuge tubes and glass pipets were used. Prior to use, we cleaned the glassware according to methods of United Sates Environmental Agency (U.S.EPA). A method blank (0.1 ml hexane), a spiked blank, and a pair of matrix-spiked samples (1 μg of individual phthalate esters)/duplicates were processed. In method blanks, only DMP, DEP, DIBP, DBP and DEHP were detected with trace levels ( ng/l). These concentrations were considerably lower than those found in samples from highly polluted regions and, therefore, were not subtracted from sample values. The average recoveries of target compounds in spiked matrices and blanks were % and 71.9%-112.6%, respectively (Supplementary Table S6). 5.2 PCBs detection 4, List of target PCBs 39 kinds of PCBs were included and the detailed compound names see

15 Supplementary Table S Sample preparation To detect trace-level PCBs in water samples, a concentration method for large-volume-water was used 4, 5. For each sample, 20 L water, spiked with 10 ng 13 C-PCB as a surrogate and filtered through 0.7 μm glass fiber filters (Whatman, USA), was passed through a glass column packed with 10 g XAD-2 resins, which were preconditioned by methanol: dichloromethane (1:1 in volume). The sequential elution used 200 ml of methanol followed by 200 ml of dichloromethane from bottom to top of the columns with the flow rate of 2 ml/min. The methanol fraction was concentrated by rotary evaporation to half the initial volume and was liquid-liquid extracted with 25 ml hexane three times. The hexane extracts were dried over anhydrous sodium sulphate, combined with the dichloromethane fraction and rotary evaporated for purification on a column filled with 1-2 g of anhydrous sodium sulphate over 3 g of 15% deactivated neutral alumina (aluminium oxide 90, activated at 400 ºC for 12 h). The elution of the column was made with 5 ml of hexane and 12 ml of hexane/dichloromethane (1:2, v/v) and was concentrated to 0.5 ml by vacuum rotary evaporation, transferred to a GC vial with isooctane and evaporated to 100 μl under a nitrogen stream. At this step, 5 ng of the internal standard ( 13 C-PCB 208) was added as internal standard prior to instrumental analysis. Water filters containing the particle phase were freeze-dried for 24 h, weighed, extracted with hexane/methanol (1:2, v/v) for 24 h and rotary evaporated to 2 ml. From here on, the extracts were purified, fractionated and treated in the same manner as for dissolved samples.

16 5.2.3 GC-MS analysis PCBs were detected by GC-MS in SIM mode (for the detailed instrumental information and oven program also see Supplementary Information 4.2). The GC-MS parameters for the analysis of target PCBs are shown in Supplementary Table S7. Quantification was conducted using external standard curves calibrated with internal standard procedure (the concentration gradients and correlation coefficients for 39 congeners are shown in Supplementary Table S7). The LOD of all target PCBs (dissolved and particulate samples) ranged from 2-8 pg/l (Supplementary Table S7) QA/QC All tubes and connections were of glass, stainless steel or PTFE and they were pre-cleaned with acetone prior use in order to avoid contamination. A field blank (distilled water) was set for each sampling campaign. For every batch of 10 samples, both a solvent blank and a procedural blank (both for aqueous and solid samples) were added to ensure that the samples and the analysis process were free of contamination. No quantifiable analytes were detected in the blanks. Mono-PCBs and di-pcbs could not be well calculated because of interference in the chromatography. Thus, mono-pcbs and di-pcbs were not included in the data analysis and integration of results in this study. To estimate the repeatability and accuracy of the analytical method, every sample was spiked with known amounts of surrogate standard mixtures prior to extraction. The average surrogate recoveries for PCBs in the real dissolved and particulate samples ranged from 74.3%-122.1% and 72.1%-127.4%, respectively. Recoveries of all individual PCBs congeners were 72.3%-115.8% (relative standard

17 deviations <20%) in 6 spiked blank samples and were 70.9%-124.4% (relative standard deviations <22%) in 6 spiked matrix samples (dissolved and particulate samples). Reported concentrations were not corrected by surrogate recovery. 5.3 PBDEs detection 6, List of target PBDEs 10 kinds of PBDEs were included and for the detailed compound names see Supplementary Table S Sample preparation Similar to the concentration method for PCBs, large-volume-water (20 L for each sample) was used for extraction. Also filtered through 0.7 μm glass fiber filters and spiked with 13 C PCB 141, each water sample passed through a glass column packed with XAD-2 at a flow rate of 2 ml/min. The resin was eluted three times with methanol, followed by three ultrasonic extractions with a mixture of dichloromethane and methanol (1:1 in volume). The extracts were combined and then liquid liquid extracted using dichloromethane. The dichloromethane extract was concentrated, solvent-exchanged to hexane, and cleaned up. The PBDEs were eluted with a mixture of hexane and dichloromethane (1:1). The eluent was finally reduced to 200 μl under a gentle nitrogen stream. The suspended particulate matter collected on glass fiber filters and sludge samples were freeze-dried. The particulate matter and homogenized sludge samples were accurately weighed before being spiked with 13 C PCB 141 and extracted with a mixture of acetone and hexane (1:1). Activated copper flakes were

18 used to remove elemental sulfur. The extracts were concentrated and cleaned up as described above. 5 ng of 13 C-PCB 208 was added as internal standard prior to instrumental analysis GC-MS analysis We performed the sample analysis with the GC-MS using negative chemical ionization (NCI) in the SIM mode and with the DB-XLB (30 m 0.25 mm i.d., 0.25 μm film thickness; Agilent, U.S.A.) for all PBDE congeners. The GC oven program was: 110 ºC for 1 min, 8 ºC/min to 180 ºC (for 1 min), 2 ºC/min to 240 ºC (for 5 min), and 5 ºC/min to 290 ºC (for 13 min). Methane was used as a chemical ionization moderating gas and helium as the carrier gas. The ion source and interface temperatures were set to 200 C and 280 C, respectively. Ion fragments m/z 79 and 81 ([Br]-) were selected as monitoring ions for target PBDEs. Additionally, m/z 372, 374, and 376 were monitored for 13 C-PCB 141 (surrogate standard) and m/z 474, 476, and 478 were used for 13 C-PCB 208 (internal standard). The internal calibration procedure was carried out for quantification based on six-point calibration curves for individual component. The correlation coefficients of the calibration curves were greater than Peaks were quantified only if the signal/noise >3 and the ratio between two monitored ions was within 15% of the standard value. The LOD (calculated as a signal of 3 times the noise level) ranged from 0.08 to 6 pg/l QA/QC Both field and procedural blank were included for each batch of 10 samples. The average surrogate recoveries for PCBs in the real dissolved and particulate samples

19 were 75.3%-117.4% and 76.8%-105.7%, respectively. Recoveries of 10 individual PBDEs congeners were 71.6%-109.3% (relative standard deviations <25%) in 6 spiked blank samples and were 70.7%-128.9% (relative standard deviations <28%) in 6 spiked matrix samples (dissolved and particulate samples). Reported concentrations were not corrected by surrogate recovery. 5.4 Sulfonamides detection List of target sulfonamides Sulfadimethoxine (SDM), sulfathizole (STZ), sulfadiazine (SDZ), sulfamethazine (SMZ) and sulfamethoxazole (SMX) were selected as the target sulfonamides in this study Sample preparation For each sample, 1 L of water was acidified to ph = 3 by adding HCl, followed by adding 0.2 g Na 2 EDTA and 20 ng of surrogate standards (sulfamethoxazole-d 4 and sulfamethazine-d 4 ). The water samples were extracted using Oasis HLB (200 mg, 6 ml) cartridges which have been sequentially pre-conditioned by 10 ml acetone, 10 ml methanol, and 10 ml 0.1% formic acid and 5 mm ammonium acetate. After loading of samples, the cartridges were washed with 10 ml 0.1% formic acid solution containing 5 mm ammonium acetate and vacuum-dried for 20 min. The dried cartridge was eluted by 10 ml methanol. The volume of elutes reduced to almost dryness under a gentle nitrogen stream and then re-dissolved in 10% methanol solution. The final extract was adjusted to 1 ml and transferred to amber vials for

20 LC-MS/MS analysis LC-MS/MS analysis The target sulfonamides were detected by liquid chromatography-tandem mass spectrometry (LC-MS/MS; Agilent 6460, USA) with electrospray ionization. The chromatographic separation was performed on an ODS C 18 (75 mm 4.6 mm i.d., 3μm film thickness; Phenomenex, USA) column. Methanol was used as mobile phase (A) and water containing 0.1% formic acid solution and 5 mm ammonium acetate was used as mobile phase (B). The gradient program was as follows: the gradient started with 20% A for 2 min, increased to 30% in 2.5 min, 30% to 75% in 7 min, and 75% to 90% in 1 min, held for 6 min, followed by returning to the initial composition in 1 min and held for 6 min. The column temperature was held at room temperature. For MS detection, the instrument was operated in positive electrospray ionization and multiple reactions monitoring (MRM) mode. The source was heated at 600 ºC and capillary voltage was set at 5.5 kv. Other optimized parameters of MS/MS and ion pair are listed in Supplementary Table S8. The concentrations were quantified by external calibration curves using 6 concentrations and the linearity of calibration curves was confirmed (R 2 > 0.99) (Supplementary Table S8). The LOD for 5 sulfonamides (calculated as a signal of 3 times the noise level) ranged from 0.2 to 0.8 ng/l QA/QC No target sulfonamides were detected in either field or procedural blanks, which were included for each batch of 10 samples. The average surrogate recoveries in the

21 real water samples ranged from 75.4% to 127.9%. Recoveries of 5 individual sulfonamide congeners were 82.5%-108.3% (RSD ranged from 12.3% to 19.8%) in 6 spiked blank samples and were 73.1%-112.7% (RSD ranged from 18.5% to 29.3%) in 6 spiked matrix samples. Reported concentrations were not corrected by surrogate recovery. 5.5 Steroids detection List of target steroids Triamcinolone, cortisol, dexamethasone, flumethasone, prednisolone, triamcinolone acetonide were selected as target steroids (see Supplementary Table S4) Sample preparation Each 1 L of water sample was filtered and 50 ng/l of surrogate standard (cortisol-d 4 ) was added in the sample before passing through Oasis HLB (60 mg, 3 ml) cartridges. After extraction, the cartridges were dried by vacuum and sequentially eluted by 10 ml acetonitrile and 5 ml dichloromethane. The elute solvents were evaporated to almost dryness under a gentle nitrogen stream and then re-dissolved in 50% methanol solution. The final extract was adjusted to 1 ml with 50 ng/l of internal standard (testosterone-d 5 ) LC-MS/MS analysis The target steroids were detected by LC-MS/MS (Agilent 6460, USA) with atmospheric pressure chemical ionization (APCI) in negative mode. The ODS C 18 (75

22 mm 4.6 mm i.d., 3μm film thickness; Phenomenex, USA) column was used for chromatographic separation. Mobile phase (A) used 5 mm ammonium acetate containing 0.05% acetic acid (ph = 4) and mobile phase (B) used methanol. The gradient elution started with 50% B, increased to 100% in 10 min, held for 6 min, followed by returning to the initial composition in 1 min. The column temperature was held at room temperature. MRM and negative ACPI mode was performed. Nitrogen gas was used for drying and collision. The capillary voltage was 2.5 kv and capillary current was 3.5 μa. Other optimized parameters of MS/MS and ion pair are listed in Supplementary Table S9. The concentrations were quantified by internal calibration procedure. The LOD for 6 sulfonamides (calculated as a signal of 3 times the noise level) ranged from 0.02 to 0.06 ng/l (Supplementary Table S9) QA/QC In both field and procedural blanks, no target sulfonamides were detected. The average surrogate recoveries in the real water samples ranged from 74.2% to 103.7%. Recoveries of 6 individual sulfonamide congeners were 85.7%-110.4% (RSD ranged from 25.7% to 32.5%) in 6 spiked blank samples and were 82.4%-125.8% (RSD ranged from 36.1% to 49.4%) in 6 spiked matrix samples. Reported concentrations were not corrected by surrogate recovery.

23 Supplementary References 1. Fu, X. S. The trends of industrial transfer in eastern coastal regions: based on the investigation for Zhejiang. Economist 10, (in Chinese) (2011). 2. National Bureau of Statistics of China. China Statistical Yearbook < ( ) (Accessed 6th June 2011). 3. Zheng, W. et al. Pollution trees: Identifying similarities among complex pollutant mixtures in water and correlating them to mutagenicity. Environ. Sci. Technol. 46, (2012). 4. Dachs, J. & Bayona, J. M. Large volume preconcentration of dissolved hydrocarbons and polychlorinated biphenyls from sea water: Intercomparison between C 18 disks and XAD-2 column. Chemosphere 35, (1997). 5. Berrojalbiz, N. et al. Persistent organic pollutants in Mediterranean seawater and processes affecting their accumulation in plankton. Environ. Sci. Technol. 45, (2011). 6. Mai, B. et al. Distribution of Polybrominated diphenyl ethers in sediments of the Pearl River Delta and adjacent South China Sea. Environ. Sci. Technol. 39, (2005). 7. Peng, X. et al. Concentrations, transport, fate, and releases of polybrominated diphenyl ethers in sewage treatment plants in the Pearl River Delta, South China. Environ. Int. 35, (2009). 8. Zhang, D., Lin, L., Luo, Z., Yan, C. & Zhang, X. Occurrence of selected antibiotics in Jiulongjiang River in various seasons, South China. J. Environ.

24 Monit. 13, (2011). 9. Tölgyesi, A., Verebey, Z., Sharma, V. K., Kovacsics, L. & Fekete, J. Simultaneous determination of corticosteroids, androgens, and progesterone in river water by liquid chromatography-tandem mass spectrometry. Chemosphere 78, (2010).

25 Supplementary Table S1: Information of high and low polluted regions in Henan province Region Drainage basin Industrial enterprise a High pollution The tributary of Huai River > 600 enterprises in the sampling drainage areas Low pollution Far from the tributary of Huai River, and no irrigation canals with water source from polluted regions flow through this area < 20 enterprises in the sampling drainage area a Data from local Trade and Industry Bureau covered both existing and suspended enterprises.

26 Supplementary Table S2: The group, region, sampling period, sample size of water samples Sample group a Region Period N Water type Industrial export 1 Shanghai Water sampled in municipal water SC, Zhejiang plants: raw water, water in treatment process and finished water 4 JC, Zhejiang KC, Jiangsu Industrial acceptance High polluted regions, Henan 6 Locality A (HRPA) 3 Surface and ground water 7 Locality B (HRPB) 3 8 Locality C (HPRC) 5 9 Locality D (HPRD) 3 10 Locality E (HPRE) 4 Control regions Low polluted regions, 6 Surface and ground water Henan a The colors illustrate the samples collected in each locality and their corresponding branches in Fig. 2.

27 Supplementary Table S3: The detailed information of IUR distance matrix Matrix Similarity Distance Formula Intersection and union ratio matrix Ratio of intersection compound amounts to union compound 1-similarity d ij 1 m k 1 (1 CV 2 ) M amounts weighted by the CV of common compound levels M n n m i j d i j : distance between water fingerprint samples S i and S j m: the amount of common compounds between S i and S j M: the union compound amount between S i and S j n i or n j : the total number of detected peaks in S i or S j CV: the coefficient of variance of each common compound between S i and S j

28 Supplementary Table S4: Target compounds list Category Plastics Compounds Dimethy phthalate (DMP), diethyl phthalate (DEP), diisobutyl phthalate (DIBP), di-n-butyl phthalate (DBP), bis(2-methoxyethyl) phthalate (DMEP), bis(4-methoyl-2-pentyl) phthalate (BMPP), bis(2-ethoxyethyl) phthalate (DEEP), diamyl phthalate (DPP), dihexyl phthalate (DHXP), butyl benzyl phthalate (BBP), hexyl 2-ethylhexyl phthalate, bis(2-n-butoxyethyl) phthalate (DBEP), dicyclohexyl phthalate (DCHP), bis(2-ethylhexyl) phthalate (DEHP), di-n-octyl phthalate (DNOP), dinonyl phthalate (DNP) PCBs 2-Chlorobiphenyl (PCB 1), 3-chlorobiphenyl (PCB 2),4-chlorobiphenyl (PCB 3), 2,2 -dichlorobiphenyl (PCB 4), 2,3 -dichlorobiphenyl (PCB 6), 2,4 -dichlorobiphenyl (PCB 8), 2,5 -dichlorobiphenyl (PCB 9), 2,2,3 -trichlorobiphenyl (PCB 16), 2,2,5 -trichlorobiphenyl (PCB 18), 2,2,6 -trichlorobiphenyl (PCB 19), 2,3,4 -trichlorobiphenyl (PCB 22), 2,3,4-trichlorobiphenyl (PCB 25), 2,4,4 -trichlorobiphenyl (PCB 28), 2,2,3,5 -tetrachlorobiphenyl (PCB 44), 2,2,5,5 -tetrachlorobiphenyl (PCB 52), 2,3,3,4 -tetrachlorobiphenyl (PCB 56), 2,3,4,4 -tetrachlorobiphenyl (PCB 66), 2,3,4,5-tetrachlorobiphenyl (PCB 67), 2,3,4,6-tetrachlorobiphenyl (PCB 71), 2,4,4,5-tetrachlorobiphenyl (PCB 74), 2,2,3,3,4-pentachlorobiphenyl (PCB 92), 2,2,3,4,5 -pentachlorobiphenyl (PCB 87), 2,2,4,4,5-pentachlorobiphenyl (PCB 99), 2,3,3,4,6-pentachlorobiphenyl (PCB 110), 2,2,3,4,4,5 -hexachlorobiphenyl (PCB 138), 2,2,3,4,5,5 -hexachlorobiphenyl (PCB 146), 2,2,3,4,5,6-hexachlorobiphenyl (PCB 147), 2,2,3,3,4,5,6-heptachlorobiphenyl (PCB 173), 2,2,4,4,5,5-hexachlorobiphenyl (PCB 153), 2,2,3,3,4,5,6 -heptachlorobiphenyl (PCB 174), 2,2,3,3,4,5,6-heptachlorobiphenyl (PCB 177), 2,2,3,3,5,6,6 -heptachlorobiphenyl (PCB 179), 2,2,3,4,4,5,5 -heptachlorobiphenyl (PCB 180), 2,2,3,4,5,5,6-heptachlorobiphenyl (PCB 187), 2,2,3,3,4,4,5,5 -octachlorobiphenyl (PCB 194), 2,2,3,3,4,4,5,6-octachlorobiphenyl (PCB 195), 2,2,3,3,4,5,5,6 -octachlorobiphenyl (PCB 199), 2,2,3,3,4,5,5,6 -octachlorobiphenyl (PCB 203), 2,2,3,3,4,4,5,5,6 -nonachlorobiphenyl (PCB 206)

29 Supplementary Table S4 (continued): Target compounds list Category PBDEs Compounds 2,4,4 -Tribromodiphenyl ether (BDE 28), 2,2,4,4 -tetrabromodiphenyl ether (BDE 47), 2,3,4,4 -tetrabromodiphenyl ether (BDE 66), 2,2,3,4,4 -pentabromodiphenyl ether (BDE 85), 2,2,4,4,5-pentabromodiphenyl ether (BDE 99), 2,2,4,4,6-pentabromodiphenyl ether (BDE 100), 2,2,3,4,4,5 -hexabromodiphenyl ether (BDE 138), 2,2,4,4,5,5 -hexabromodiphenyl ether (BDE 153), 2,2,4,4,5,6 -hexabromodiphenyl ether (BDE 154), 2,2,3,4,4,5,6-heptabromodiphenyl ether (BDE 183) Sulfonamides Steroids Sulfadimethoxine (SDM), sulfathiazole (STZ), sulfadiazine (SDZ), sulfamethazine (SMZ), sulfamethoxazole (SMX) Triamcinolone, cortisol, dexamethasone, flumethasone, prednisolone, triamcinolone acetonide

30 Supplementary Table S5: Compound name, retention time, target and qualifier ions, limit of detection (LOD), concentration range, regression coefficient (R 2 ) and internal standards (IS) of all target PAEs No. Compound a Rention time Target Qualifiter Qualifiter Range R 2 LOD IS (min) (T) ion1 (Q 1 ) ion2 (Q 2 ) (μg/ml) (ng/l) 1 DMP Acenaphthene-d Acenaphthene-d DEP Acenaphthene-d Phenanthrene-d DIBP DBP DMEP BMPP DEEP DPP DHXP BBP HEHP DBEP Phenanthrene-d Phenanthrene-d Phenanthrene-d Phenanthrene-d Phenanthrene-d Chrysene-d Chrysene-d Chrysene-d Chrysene-d Chrysene-d Chrysene-d DCHP Chrysene-d DEHP Chrysene-d DNOP Chrysene-d DNP Chrysene-d 12 a For the full compound name see Supplementary Table S4.

31 Supplementary Table S6: Mean recoveries and relative standard deviation (RSD) of target PAEs in spiked matrices and blanks (n=6) Compound a % Mean recovery (RSD %) % Mean recovery (RSD %) High spiked matrix b Low spiked matrix b High spiked blank b Low spiked blank b DMP 82.9 (7.8) 79.8 (5.6) 76.6 (15.4) 72.5 (17.4) DEP 94.0 (11.6) 82.6 (12.9) 85.1 (19.0) 95.4 (8.6) DIBP (15.3) (8.4) 75.2 (9.5) 87.4 (11.4) DBP (13.3) (9.7) (9.9) (22.5) DMEP 99.8 (8.4) 91.7 (12.6) 74.3 (11.7) 86.4 (5.9) BMPP 95.4 (17.0) 78.6 (9.4) 91.6 (8.7) 96.8 (14.1) DEEP 79.1 (16.4) (11.2) 87.3 (5.2) (19.8) DPP (21.6) 87.0 (4.5) 79.9 (6.3) 72.3 (8.0) DHXP 81.4 (15.7) 98.8 (7.6) 82.1 (15.8) 80.4 (10.2) BBP 79.1 (20.3) (19.2) 75.6 (14.3) 85.0 (7.6) HEHP 89.5 (5.7) 81.4 (8.1) 93.3 (18.6) 71.9 (13.9) DBEP 81.2 (7.4) (16.7) 86.4 (21.2) 77.8 (15.0) DCHP 88.4 (3.1) 95.9 (6.0) 85.1 (4.9) 89.9 (18.3) DEHP (21.5) (15.2) 92.5 (13.1) 73.2 (6.7) DNOP (4.9) (16.3) 88.9 (20.5) (12.2) DNP 84.3 (10.2) (15.8) 95.1 (7.1) 92.6 (16.4) a For the full compound name see Supplementary Table S4. b High spike: 5 μg/l; low spike: 0.5 μg/l.

32 Supplementary Table S7: Compound name, retention time, target and qualifier ions, limit of detection (LOD), concentration range, and regression coefficient (R 2 ) of all target PCBs No. Compound a Rention time Target Qualifiter Qualifiter Range R 2 LOD b (min) (T) ion1 (Q 1 ) ion2 (Q 2 ) (ng/l) (pg/l) 1 PCB PCB PCB PCB PCB PCB PCB PCB PCB PCB PCB PCB PCB PCB PCB PCB PCB PCB PCB a For the full compound name see Supplementary Table S4. b Dissolved and particulate samples.

33 Supplementary Table S7 (continued): Compound name, retention time, target and qualifier ions, limit of detection (LOD), concentration range, and regression coefficient (R 2 ) of all target PCBs No. Compound a Rention time Target Qualifiter Qualifiter Range R 2 LOD b (min) (T) ion1 (Q 1 ) ion2 (Q 2 ) (ng/l) (pg/l) 20 PCB PCB PCB PCB PCB PCB PCB PCB PCB PCB PCB PCB PCB PCB PCB PCB PCB PCB PCB PCB a For the full compound name see Supplementary Table S4. b Dissolved and particulate samples.

34 Supplementary Table S8: Compound name, retention time, precursor and product ions, concentration range, and regression coefficient (R 2 ), dwell time, fragementor and collision energy (CE) of all target sulfonamides No. Compound a Rention time Precursor Product Range R 2 Dwell time Fragmentor CE (min) ion ions (ng/ml) (ms) (V) (ev) 1 SDM , STZ , SDZ , SMZ , SMX , a For the full compound name see Supplementary Table S4.

35 Supplementary Table S9: Compound name, retention time, precursor and product ions, concentration range, and regression coefficient (R 2 ), dwell time, fragementor and collision energy (CE) of all target sulfonamides No. Compound Rention time Precursor Product Range R 2 Dwell time Fragmentor CE (min) ion ions (ng/l) (ms) (V) (ev) 1 Triamcinolone , Cortisol , Dexamethasone , Flumethasone , Prednisolone , Triamcinolone acetonide ,

36 Number of Industrial Enterprises (unit) Year Supplementary Figure S1 The annual number of industrial enterprises in Henan province during The data used was from China Statistical Yearbook

37 Gross industrial output value (billion dollars) Year Supplementary Figure S2 The annual gross industrial output values of enterprises in Henan province during The data used was from China Statistical Yearbook

38 Industrial waste water discharge amount (million tons) Year Supplementary Figure S3 The annual industrial wastewater discharge amount in Henan province during The data used was from China Statistical Yearbook