Chemosphere 134 (2015) 263–271

Contents lists available at ScienceDirect

Chemosphere journal homepage: www.elsevier.com/locate/chemosphere

Source apportionment of PAHs in surface sediments using positive matrix factorization combined with GIS for the estuarine area of the Yangtze River, China Wenwen Yu, Ruimin Liu ⇑, Jiawei Wang, Fei Xu, Zhenyao Shen State Key Laboratory of Water Environment Simulation, School of Environment, Beijing Normal University, No. 19, Xinjiekouwai Street, Beijing 100875, China

h i g h l i g h t s  PMF and geostatistics were used to identify sources of PAHs.  The high molecular weight PAHs with 4–6-ring accounted for major amount.  A 3-factor result from PMF model gave the most satisfactory analysis.  The impacted regions of 3 factors were analyzed by ordinary Kriging.

a r t i c l e

i n f o

Article history: Received 15 September 2014 Received in revised form 16 April 2015 Accepted 17 April 2015 Available online 15 May 2015 Keywords: Polycyclic aromatic hydrocarbons Estuarine sediments Source apportionment Positive matrix factorization Yangtze River Estuary

a b s t r a c t This study used PMF and geostatistics to quantify sources of PAHs based on 30 samples tested for 16 PAHs in surface sediment from the Yangtze River Estuary (YRE) in February 2011. The results demonstrated that the total PAH concentrations varied from 65.07 to 954.52 ng g1 with a mean value of 224.00 ng g1. In the inner estuary, the mean of the total PAH concentrations was 229.89 ng g1, and the high molecular weight of four-to-six-ring PAHs accounted for 51.83% of PAHs. In the adjacent East Sea, the mean value was 218.85 ng g1 and the high molecular weight PAHs accounted for approximately 54% of total PAHs. A three-factor modeling result from PMF provided the most satisfactory analysis of PAH sources. Coke plant emissions and biomass combustion, which contributed 45.64% of the pollution, were the most important sources, and pollutants from these sources were primarily concentrated in the southern branch of the estuary. Gasoline fuel combustion accounted for approximately 40% of the pollution, and the major contaminated area was in the northern region. Petrogenic sources (14.70%) also influenced the estuary, especially in the northeastern region. Water currents and source locations affected the impacted regions of PMF factors; the surrounding natural and artificial influences were also considered. Ó 2015 Elsevier Ltd. All rights reserved.

1. Introduction Polycyclic aromatic hydrocarbons (PAHs) are among most widespread and important classes of environmental persistent organic pollutants (Liu et al., 2009; Dong and Lee, 2009). PAHs spread in the air, water, sediment, and soil (Lohmann et al., 2009; Lee and Dong, 2010; Fernández-Luqueño et al., 2011). PAHs with three or more rings have low solubility in water and low vapor pressure; these PAHs can be widely dispersed by atmospheric transport or through stream pathways, and they eventually accumulate in aquatic sediments once produced (Kolarova et al.,

⇑ Corresponding author. E-mail address: [email protected] (R. Liu). http://dx.doi.org/10.1016/j.chemosphere.2015.04.049 0045-6535/Ó 2015 Elsevier Ltd. All rights reserved.

2005; Jiang et al., 2007). When released into aquatic environments, the PAHs bind to particles easily and are subsequently deposited in sediments because of their high hydrophobicity and weak degradation (Barakat et al., 2011). Thus, sediment often constitutes a pollutant trap and has proven to be an effective monitoring object for identifying environmental impacts (Liu et al., 2011). PAHs have attracted significant attention, as some types of PAHs have been validated as mutagenic and carcinogenic compounds, which are linked to human health problems such as cataracts, kidney, and liver damage (Wang et al., 2011; Chen and Chen, 2011; Vu et al., 2011). Anthropogenic sources are the most important contributors and have been divided into two groups: pyrogenic and petrogenic (Zakaria et al., 2002; Chen et al., 2006; Boonyatumanond et al., 2007). PAHs with more than four rings are generally from pyrogenic sources (Dahle et al., 2003).

264

W. Yu et al. / Chemosphere 134 (2015) 263–271

As PAHs are widespread and potentially toxic, many studies have focused on the source apportionment of PAHs before applying measures to control them. Various studies have applied different qualitative and quantitative methods to analyze sources of PAHs (Zhang et al., 2005; Tobiszewski and Namies´nik, 2012; Yancheshmeh et al., 2014). The diagnostic ratio analysis of PAHs is one of the most widely used qualitative methods of analyzing PAH sources. It has been applied in areas of road-deposited sediments, water sediments, and contaminative soils (Teixeira et al., 2013; Barhoumi et al., 2014; Lv et al., 2014). Some statistical methods, such as Principal Components Analysis (PCA), have been widely used in the source analysis of PAHs and have yielded satisfactory results (Luo et al., 2008; Gupta et al., 2011). However, the application of PCA is limited in its ability to handle noisy data and weighting of uncertainties (Paatero and Hopke, 2003). Additionally, when using PCA, some negative factor loadings are difficult to interpret in terms of positive definite physical parameters such as concentrations, masses, and spectral intensities. Recently, the Positive Matrix Factorization (PMF) approach has been developed and was first adopted as a preferable technique for source apportionment of atmospheric aerosol constituents (Jang et al., 2013). This approach largely overcomes limitations that restrict PCA applications by using experimental uncertainties in the data matrix and by constraining the solutions with non-negative values (Paatero and Tapper, 1994). PMF offers an additional advantage over PCA because of its superior ability to handle missing values and to account for data precision and because it does not rely on previous knowledge of sources by direct measurement or from emission inventories (Vaccaro et al., 2007; Comero et al., 2011). This method has been applied in source analysis of atmospheric pollution in particulate matter, heavy metal pollution in soil and water sediments, and the study of wet deposition (Chueinta et al., 2000; Bzdusek and Christensen, 2006; Keeler et al., 2006; Comero et al., 2012). The estuarine area is a crucial part of every river. Estuarine regions are transition zones between freshwater and seawater, and the hydrological environment and rich biodiversity are always complex in these regions (Huguet et al., 2009; Feely et al., 2010). This essential hydrological and biological environment is sensitive to disruption by human activities (i.e., tidal salt marsh reclamation and urbanization) (Bai et al., 2011). Waste drainage with various pollutants produced by industry and agriculture may destroy the ecological balance of the estuary and threaten human health (Baumard et al., 1998; Silva et al., 2013). Sources of sedimentary pollutants should be determined before controlling measures are established, particularly for the quantitative source identifications (Stout and Graan, 2010; Feng et al., 2014). In past decades, PCA and other methods were widely applied to analyze the heavy metal pollution sources, polychlorinated biphenyl and PAHs in the sediments of estuaries (Pekey and Dogan, 2013; Saba and Su, 2013; De Abreu-Mota et al., 2014; Wang et al., 2014). However, most source analyses of sedimentary PAHs in estuaries were conducted using only qualitative methods; few studies applied PMF in quantitative source apportionment of sedimentary PAHs. In practice, the impacted regions of different pollution sources exhibit various spatial distributions (Vaccaro et al., 2007; Ha et al., 2014). However, most of the previous studies focused on identifying sources and ignored the variations of the impacted regions. A Geographic Information System (GIS) provides an efficient tool for exploring the variety of spatial distributions. Geostatistics, an interpolation technique in GIS, is usually used to estimate distributions based on the data from sampling sites (Jaber et al., 2013). With geostatistics, GIS can also be applied to obtain the regions impacted by sources found in PMF results. This paper focuses on the Yangtze River Estuary (YRE) as the study area. Geostatistics and PMF were used to quantify the

sources of PAHs based on 30 samples tested for 16 PAHs in surface sediment in February 2011. The objectives of this work included (1) to identify the statistical characteristics and compositions of PAHs, (2) to quantitatively apportion the sources of PAHs using PMF and (3) to analyze the impacted regions of source factors. The results could be helpful for PAH prevention and for sustainable development of the YRE. 2. Materials and methods 2.1. Study area and sample collection The Yangtze River Estuary is an important part of the Yangtze River, which is a key Chinese shipping hub and possesses one of China’s most urbanized and industrialized regions (Zhou et al., 2014). It receives more than 240 Mt yr1 of fine sediments from the Yangtze River (Yang et al., 2006; Pan and You, 2010). The study area is divided into two parts: the Inner Yangtze River Estuary (IYRE) and the Adjacent East China Sea (AECS) (Fig. 1). The estuary is ramified into the South Branch and the North Branch by Chongming Island. The South Branch is the main stream of the estuary and receives over 95% of the total estuarine runoff (Chen et al., 2001). The South Branch is further divided by Chongxing Island into the North Bay and South Bay. The rapid economic development along the Yangtze River has resulted in an increased influx of city industrial wastewater and polluting emissions, which contain large quantities of PAHs. Additionally, the construction of dams upstream, especially the Three Gorges Dam, has slowed the river sediment discharge downstream in recent years (Xu and Milliman, 2009). The effect of complex ocean currents and changes in river runoff often cause the mixing of fresh and salt water in the Yangtze River Estuary to occur further upriver (Dong et al., 2012). Thus, the distribution and the source apportionment of PAHs in the estuary may be different from those of other areas and previous studies. In total, 30 surface sediment samples were collected with a grab sampler (Van Veen bodemhappe 2L) in February 2011. Of those, 14 samples were collected in the IYRE and the remaining 16 samples were collected from the AECS. All of the sediments were placed in aluminum containers and kept in a refrigerator at 20 °C for further analysis. 2.2. Chemical analysis A mixture standard of 16 PAHs was used for external calibration. The PAHs were diluted to the concentrations of the working standards, which were then combined to prepare calibration standards for HPLC analyses. All solvents used for sample processing and analysis (dichloromethane, acetone, hexane, and methanol) were of HPLC grade from Tedia Co., Inc. (USA). Deionized water was produced by a Milli-Q system (Millipore Co., USA). Anhydrous sodium sulfate and silicic gel (100–200 mesh) were of analytical grade and activated at 450 °C to remove impurities before use (Li et al., 2012). Sediments were freeze-dried (FD-1A, China) and then sieved through a 100 mesh sieve. An accelerated solvent extractor (ASE300, Dionex, USA) was used for the extraction of PAHs. Approximately 5 g of the sieved sample was mixed with some diatomite and 1 g of activated copper to decrease the sulfur content. The mixture was transferred into a 66 mL extraction cartridge using hexane and acetone (1:1 V/V) for extraction under 1500 psi at 100 °C. Two cycles were conducted in this process, with 5 min of heating followed by 5 min of static extraction, a flush volume of 60%, and an N2 purge time of 60 s. Next, a rotary evaporator (RV 05, IKA, Germany) was used to concentrate the extracted

W. Yu et al. / Chemosphere 134 (2015) 263–271

solution to 1–2 mL. The purification procedure was developed as reported previously by Niu et al. (2003). First, 8 mL of hexane was added to a column, which was then filled with 2 g silica gel (5% deactivated) and 1 g anhydrous sodium sulfate. The concentrations were then transferred into the column and eluted with 10 mL of hexane/dichloromethane (1:1, v:v). The elution containing the PAHs was concentrated to 1 mL; the solvent was then exchanged with methanol and, finally, concentrated to 1 mL for HPLC analysis. Sixteen United States Environment Protection Agency (U.S. EPA) priority PAHs were analyzed using HPLC (Dionex UltiMate 3000) with a UV and fluorescence detector for each of the sediment samples. The analyzed PAHs included naphthane (Nap), acenaphthene (Ace), fluorene (Flu), phenanthrene (Phe), anthracene (Ant), fluoranthene (Fla), pyrene (Pyr), benzo[a]anthracene (BaA), chrysene (Chr), benzo[b]fluoracene (BbF), benzo[k]fluoracene (BkF), benzo[a]pyrene (BaP), dibenzo[a, h]anthracene (DahA), benzo[g, h, i]perylene (BgP), indeno[1,2,3-cd]pyrene (InP), and acenaphthylene (Acy). A C18 column (Varian, 250  4.6 mm, 5 lm) was installed into the analyzer. The injection volume was 20 lL, and the column temperature was 30 °C. The gradation process was performed by the method introduced by Feng et al. (2007): 75% methanol was used, followed by a 17 min linear gradation to 100% methanol for 30 min, initially. A linear gradient (1 min) back to the starting conditions was followed by a 7 min pre-run to achieve equilibrium for each subsequent run. Except for acenaphthylene, which was identified by the UV detector at 229 nm, the other 15 PAHs were quantified by the fluorescence detector. 2.3. Positive matrix factorization model The Positive Matrix Factorization (PMF) model is a multivariate factor analysis tool that decomposes data samples into factor

265

contribution matrices and the factor profiles (Paatero, 2007). These two matrices are then interpreted by an analyst to determine what source types are represented using measured source profile information and emission inventories (Paatero and Tapper, 1993, 1994). In PMF, the factor analytic model is expressed by the following matrix relationship:

X ¼ GF þ E

ð1Þ

or in component form:

xij ¼

P X g ik f kj þ eij

ð2Þ

k¼1

where xij is the concentration of species j measured on sample i, P is the number of factors contributing to the samples, g ik is the relevant contribution of factor k to sample i, f kj is the concentration of species j in factor profile k, and eij is the error of the PMF model for the species j measured in sample i. In Eqs. (1) and (2), the matrix X holds the measured values, containing i rows and j columns. In environmental pollution problems, each row of X consists of measured concentrations of chemical compounds for each individual sample, and every column contains the concentrations of one compound for all of the samples. G and F are the factor matrices to be determined, which only allow nonnegative values, and E is the residuals matrix, the unexplained part of X. In PMF, the solution is approximated by a weighted least squares fit, where the known standard deviations for each value of X are used to determine the weights of the residuals in matrix E. One row of the matrix F would be the ‘‘profile’’ for one source, and the corresponding column of G would indicate the amount of this emission present in the samples. The parameter of p in Eq. (2) is the number of sampling sites.

Fig. 1. Locations of the Yangtze River Estuary (YRE) and 30 sample sites.

266

W. Yu et al. / Chemosphere 134 (2015) 263–271

The values of gik and fkj in Eq. (2) are adjusted until a minimum value of Q for a given p is found. Q is defined as:

Q¼

" N X M X i¼1 j¼1

xij 

! #2 P X g ik f kj =sij

ð3Þ

k¼1

where g ik is the relevant contribution of factor k to sample i, f kj is the concentration of species j in factor profile k, sij is the uncertainty of the jth species concentration in sample i, N is the number of samples, and M is the number of species. Thus, for more uncertain values of xij, larger values of sij are assigned so that such data have less weight. The matrix of uncertainties (sij) corresponding to each entry in the concentration matrix must also be supplied as an input to PMF. The simplest method of obtaining such a matrix is to use the analytical or method uncertainties that correspond to each species concentration value as calculated in the User Guide (Norris et al., 2008). PMF requires that all concentrations and uncertainty values should be positive values; therefore, the data below detection limits (DL) and missing data (MD) must be omitted or replaced with appropriate substitute values (Gonzalez-Macias et al., 2014). In the concentration matrix, values below DL were replaced with DL/2. The MD values were substituted with the mean value of points around the MD point. Values in the uncertainty matrix were calculated by equations in the User Guide (Norris et al., 2008). Four PAH variables were weighted as weak due to any of the following reasons: (1) low Signal-to-Noise ratio (S/N) values, (2) low R2 values, and (3) absolute values of residuals greater than 3. The key to a successful PMF application is to examine the similarity between the source profiles and the specific sets of factor loadings obtained from the PMF model (Zhang et al., 2015). Data analysis in this study was performed with the EPA Positive Matrix Factorization (PMF) 3.0 model. A major consideration in the operation of the PMF is finding the optimum number of factors (p). The p value was determined through the examination of Q values for PMF solutions, resulting in a range of p values (Reff et al., 2007).

3. Results and discussion 3.1. Basic statistical analysis Statistical results of the total concentrations of each PAH from P the 30 sites showed that total PAH concentrations ( PAHs) varied 1 from 65.07 to 954.52 ng g , with a mean value of 224.00 ng g1 (Table 1). The PAH concentration in this study was significantly lower than that during the summer, based on the previous study (Wang et al., 2012). This reduction was probably caused by the lower winter temperature, as temperature influences the absorption efficiency (Vardar et al., 2008; Kim et al., 2012). The sites with P minimum and maximum values of the PAHs were both in the IYRE, with a mean value of 229.89 ng g1, which was higher than the average value of the whole YRE. In the AECS, the values of the P PAHs ranged from 73.94 ng g1 to 668.98 ng g1, with a mean value of 218.85 ng g1, which was lower than the average value of P the whole YRE. However, PAH values for the 16 sampling sites in the AECS displayed less variation than those of the 14 sites in the IYRE, with a lower SD value. The order of all elements in the YRE based on mean concentration was Phe > Fla > Pyr > BbF > BgP > BaP > BaA > Nap > Flu > InP > Acy > Chr > BkF > Ant > DahA > Ace. The order in the IYRE was Phe > Fla > Pyr > BbF > BgP > Nap > BaA > BaP > Flu > InP > Chr > Acy > BkF > Ant > DahA > Ace. The order in the AECS was Phe > Fla > Pyr > BbF > BgP > BaP > BaA > Flu > Acy > Nap > InP > Chr > BkF > Ant > DahA > Ace.

Table 1 Concentrations of 16 PAHs in the samples and the total PAHs concentrations in Inner Yangtze River Estuary and the Adjacent East China Sea. PAHs (ng g1)

Min

Max

Mean

Nap Ace Flu Phe Ant Fla Pyr BaA Chr BbF BkF BaP DahA BgP InP Acy P PAHs of The Inner Estuary P PAHs of Adjacent East Sea

2.81 0.51 4.31 0.11 1.40 4.81 4.51 2.69 1.47 1.27 0.38 0.51 0.16 0.86 0.28 0.57 65.07 73.94

40.37 5.50 35.36 84.44 14.13 96.15 120.63 66.06 41.19 139.25 41.85 107.64 17.92 112.45 94.56 42.35 954.52 668.98

14.38 1.87 14.05 34.17 4.99 26.04 22.74 14.75 8.57 19.21 6.13 15.13 2.90 17.40 11.31 10.37 229.89 218.85

P High-value zones of the PAHs were found around the points D4, D10, and D17 (Fig. 2). D4 and D10 are in the IYRE, and D17 is located in the AECS. A large area with high values around D17 accounted for approximately 1/3 of the entire study area. P Additionally, the highest concentrations of the PAHs and most of the measured PAHs were detected at the D10 site. Anthropogenic inputs might be the main reason that the high concentration values emerged around sites D4, D10 and D17. Finegrained sediments might also contribute to these results (Wang et al., 2015). Different sources and locations of various elements may be the cause of the uneven distributions. 3.2. Composition characteristics The compositions of PAHs in the two study partitions showed that two-to-three-ring PAHs were slightly more prevalent than other PAHs (Fig. 3). The two-to-three-ring, four-ring, and five-tosix-ring PAHs, which represent low, medium, and high molecular weight PAHs (Wang et al., 2012), accounted for 47.26%, 31.87%, and 20.86%, respectively, of total PAHs. In addition, the concentration of the two-to-three-ring PAHs was higher than the summation of the four-ring and five-to-six-ring PAHs in most sampling sites, accounting for more than 50% of the total PAH concentration. In the 14 sites of the IYRE, the two-to-three-ring PAHs accounted for approximately 48.17% of PAHs, while the four-ring PAHs and five-to-six-ring PAHs accounted for 31.29% and 20.54%, respectively. The 16 sites in the AECS observed that the proportion of the two-to-three-ring PAHs was smaller than that of the other types (46.43%); the levels of the four-ring PAHs and the five-tosix-ring PAHs were 32.41% and 21.16%, respectively. The higher aromatic rings (four to six rings) were representative for the total PAH compounds in both the IYRE and the AECS, and more low-molecular-weight PAHs emerged in the IYRE. Two-part compositions should have close correlations with pollutant sources, which are discussed in detail later. 3.3. Quantitative source analysis with PMF A major consideration in the operation of the PMF is finding the optimum number of factors (p). The Qrobust value (257) was closest to the Qtheoretical value (342, i  j  p  (i + j)) for three factors and the fourth of 20 runs was chosen for having the lowest Qtrue value (255.5); the use of three factors provided a good fit, e.g., the R2 for BbF was 0.996.

W. Yu et al. / Chemosphere 134 (2015) 263–271

Each factor was explained by the percentages of all measured elements based on PMF (Fig. 4). Factor 1 was responsible for 39.66% of all factor contributions and was characterized by BbF, Pyr, BgP, and BaP. This group of elements included only highmolecular-weight PAHs with four to six rings. According to the literature, BbF, BgP, and BaP are related to the emission characteristics of PAHs from vehicle emissions and Pyr is mainly derived from the gasoline and fossil fuel combustion of ships (Wang et al., 2010; Cao et al., 2011; Nguyen et al., 2014). Factor 1 might represent a source of gasoline and fossil fuel combustion. Factor 2 strongly reflected Phe variation and was also influenced by Fla and Flu, to some extent. Phe was suggested as an indicator for coke plant emissions (Simcik et al., 1999) and Fla was correlated with biomass combustion, such as grass and wood (Ou et al., 2010). Flu is considered the predominant coal and biomass combustion (low-temperature pyrolysis) profile (Yan et al., 2009). This indicated that Factor 2 might be predominantly sourced from the coke plant emissions and biomass combustion. This factor accounted for 45.64% of the three factor contributions. Factor 3, which contributed for 14.70% of the total PAHs, was explained mostly by the variation of Ace and Flu and slightly impacted by Nap and Phe. Ace and Nap belong to the low-molecular-weight PAHs, which are abundant in petrogenic sources such as carbon and crude oil and mainly caused by petroleum spills (Liu et al., 2009; Ghosal et al., 2013). Thus, Factor 3 might be a petrogenic source in this study. The results of the percentage of variation of every measured element explained by the three factors showed that for Flu, Phe, and Ant, Factor 2 was the main contributor and accounted for more than 80% of the sources of these elements, with Factor 3 and Factor

267

Fig. 3. Triangular diagram of percentage composition for the 16 PAHs in the Inner Yangtze River Estuary and the Adjacent East China Sea.

1 contributing over 10% and less than 10%, respectively (Fig. 5). Factor 2 contributed approximately 86% of the total concentration for Phe. Factor 2 was also the main contributor for Nap (77%) and Ace (65%), whereas approximately 23% of Nap came from Factor 3, and 22% of Ace was from Factor 1. BaA, Chr, and Pyr had very similar source proportions: approximately 50% of these three elements were from Factor 1, approximately 40% came from Factor 2, and the remaining 10% came from Factor 3. The source contribution profiles of BbF, BkF, and BaP were nearly the same, with Factor 1

Fig. 2. Spatial distribution of the total PAHs concentrations in the Yangtze River Estuary based on the ordinary Kriging.

268

W. Yu et al. / Chemosphere 134 (2015) 263–271

Fig. 4. Percentages of 16 PAHs in three factors obtained by the PMF model in the Yangtze River Estuary.

accounting for more than 70%, Factor 2 for 15%, and Factor 3 explaining the remainder. The special elements were Acy and DahA, which were only influenced by two factors. For Acy, approximately 93% of the concentration can be explained by Factor 3; Factor 2 only contributed the remaining 7%. For DahA, Factor 1 was responsible for 64% of the sources and Factor 2 was responsible for the remaining 36%. 3.4. Spatial distribution of factors Factor 1 represented a source of gasoline and fossil fuel combustion, as stated in the above analysis; high values were mainly located in the northern zone of the AECS, and areas around sites D4, D5, D15, D17 and D26 were greatly influenced by this factor (P50%) (Fig. 6a). This distribution conformed to the results of the composition characteristics; samples from AECS were chiefly characterized by the high-molecular-weight PAHs. Many main sea lanes and ships from the Yangtze River cross the north branch via this zone. The heavy sea traffic might lead to serious pollution by high-molecular-weight PAHs. Additionally, the area is dominated by the complex currents of the Taiwan Warm Current and the Yellow Sea Coastal Current, which also affect spatial distributions of suspended sediment that carry more high-molecularweight PAHs (Zakaria et al., 2002; Li et al., 2012). Estuarine runoff flows mainly to the southeast in the winter due to the asymmetric tidal current and vertical current (Bian et al., 2013). Most of the

Fig. 5. Source contribution percentages from three PMF estimated factors to the 16 PAHs in the whole Yangtze River Estuary.

sediments entering the ocean are transported to the east of Jiangsu province (Chen et al., 2003; Han et al., 2006; Bian et al., 2013). Thus, the spatial distribution of Factor 1 was focused on the northeast zone of the AECS. The dilution effect from the

W. Yu et al. / Chemosphere 134 (2015) 263–271

269

Fig. 6. Impacted regions of (a) gasoline and fossil fuel combustion; (b) coke plant emission and biomass combustion; and (c) petrogenic source.

Huangpu River may have weakened the influence of Factor 1 around the D9 and D10 sites. According to the results of PMF, Factor 2 was characterized by a series of low-molecular-weight PAHs (Phe, Flu, Fla). High values of Factor 2 were concentrated in the south branch of the inner estuary around sites D3, D4, D5, D9, D10, D11 and D28, near the Baoshan distract of Shanghai (Fig. 6b). As previously stated, the two-to-three-ring PAHs accounted for a higher proportion of PAHs in the IYRE, and the literature indicated that the two-tothree-ring PAHs were probably derived from oil, fuel spills, and low temperature or incomplete combustion (Wang et al., 2009). The Baoshan district is a very important base of steel production and port container production, which discharges large quantities of wastes with high PAHs (Chen et al., 2011). Additionally, residential living and agricultural activities that use biomass fuel might be other sources of this factor. The distribution of Factor 3 identified that the main area under the influence of this factor was in the northeastern region of the study area, except for a point-source pattern on the point D10 (Fig. 6c). Factor 3 was characterized by Ace, which is regarded as a petrogenic source from the PMF analysis. High values might be caused by the natural crude oil sources in the AECS and water currents which carried pollution that was produced in the IYRE. For the particularly high points of D9 and D10, which were near Shanghai in the south branch of the Yangtze River, oil spills from

docks in these sites and busy ports along the two sides might be the main source. 4. Conclusions The mean concentrations of PAHs at the 30 sampling sites was 224.00 ng g1 with 229.89 ng g1 in the IYRE and 218.85 ng g1 in the AECS. The total PAHs indicated high-value zones around points D4, D10, and D17. High-molecular-weight PAHs with four to six rings were dominant throughout the study area, accounting for 52.73% of PAHs; the proportion of the high-molecular-weight PAHs in the AECS was higher than that in the IYRE. Analytical results of PMF showed that Factor 1 was explained by elements of BbF, Pyr, BgP, and BaP, which might represent a source of gasoline fuel combustion and was responsible for 39.66% of all factor contributions. Factor 2 might be sourced from coke plant emissions and biomass combustion, which was mainly characterized by Phe, Flu, and Fla. This factor accounted for 45.64% of factor contributions. Factor 3, 14.70% of all of the factor contributions, was characterized by Ace, which might be regarded as a petrogenic source. Additionally, different elements had different contribution profiles from the three factors. Spatially, Factor 1 was mainly distributed in the northern zone (around sites D4, D5, D15, D17 and D26) of the study area. Water currents and artificial activities had different effects on the

270

W. Yu et al. / Chemosphere 134 (2015) 263–271

distribution of this factor. Contributions of Factor 2 were mostly found in the south branch of the estuary, near the Baoshan district, which has serious industrial pollution. Factor 3 was the main source influencing the northeastern region of the study area. Oil spills from docks at these sites and busy ports along both sides might be the causes of the particularly high pollution. Acknowledgements The research was funded by the National Basic Research Program of China (973 Project, 2010CB429003), the National Natural Science Foundation of China (Grant No. 41001352). The authors would like to thank the editors and the anonymous reviewers for their valuable comments and suggestions on this paper. References Bai, J., Xiao, R., Cui, B., Zhang, K., Wang, Q., Liu, X., Gao, H., Huang, L., 2011. Assessment: of heavy metal pollution in wetland soils from the young and old reclaimed regions in the Pearl River Estuary, South China. Environ. Pollut. 159, 817–824. Barakat, A.O., Mostafa, A., Wade, T.L., Sweet, S.T., El Sayed, N.B., 2011. Distribution and characteristics of PAHs in sediments from the Mediterranean coastal environment of Egypt. Mar. Pollut. Bull. 62, 1969–1978. Barhoumi, B., LeMenach, K., Devier, M.H., Ben Ameur, W., Etcheber, H., Budzinski, H., Cachot, J., Driss, M.R., 2014. Polycyclic aromatic hydrocarbons (PAHs) in surface sediments from the Bizerte Lagoon, Tunisia: levels, sources, and toxicological significance. Environ. Monit. Assess. 186, 2653–2669. Baumard, P., Budzinski, H., Michon, Q., Garrigues, P., Burgeot, T., Bellocq, J., 1998. Origin and bioavailability of PAHs in the Mediterranean sea from mussel and sediment records. Estuar. Coast. Shelf S. 47, 77–90. Bian, C., Jiang, W., Quan, Q., Wang, T., Greatbatch, R.J., Li, W., 2013. Distributions of suspended sediment concentration in the Yellow Sea and the East China Sea based on field surveys during the four seasons of 2011. J. Mar. Syst. 121, 24–35. Boonyatumanond, R., Murakami, M., Wattayakorn, G., Togo, A., Takada, H., 2007. Sources of polycyclic aromatic hydrocarbons (PAHs) in street dust in a tropical Asian mega-city, Bangkok, Thailand. Sci. Total Environ. 384, 420–432. Bzdusek, P.A., Christensen, E.R., 2006. Comparison of a new variant of PMF with other receptor modeling methods using artificial and real sediment PCB data sets. Environmetrics 17, 387–403. Cao, Q., Wang, H., Chen, G., 2011. Source apportionment of PAHs using two mathematical models for mangrove sediments in Shantou coastal zone, China. Estuar. Coast. 34, 950–960. Chen, C., Chen, C., 2011. Distribution, origin, and potential toxicological significance of polycyclic aromatic hydrocarbons (PAHs) in sediments of Kaohsiung Harbor, Taiwan. Mar. Pollut. Bull. 63, 417–423. Chen, Z., Li, J., Shen, H., Zhang, H.W., 2001. Yangtze River of China: historical analysis of discharge variability and sediment flux. Geomorphol. J. 41, 77–91. Chen, S.L., Zhang, C.A., Yang, S.L., 2003. Temporal and spatial changes of suspended sediment concentration and resuspension in the Yangtze River estuary. J. Geogr. Sci. 13, 498–506. Chen, S., Luo, X., Mai, B., Sheng, G., Fu, J., Zeng, E.Y., 2006. Distribution and mass inventories of polycyclic aromatic hydrocarbons and organochlorine pesticides in sediments of the Pearl River Estuary and the northern South China Sea. Environ. Sci. Technol. 40, 709–714. Chen, Y., Wang, J., Shi, G., Sun, X., Chen, Z., Xu, S., 2011. Human health risk assessment of lead pollution in atmospheric deposition in Baoshan District, Shanghai. Environ. Geochem. Hlth. 33, 515–523. Chueinta, W., Hopke, P.K., Paatero, P., 2000. Investigation of sources of atmospheric aerosol at urban and suburban residential areas in Thailand by positive matrix factorization. Atmos. Environ. 34, 3319–3329. Comero, S., Locoro, G., Free, G., Vaccaro, S., De Capitani, L., Gawlik, B.M., 2011. Characterisation of Alpine lake sediments using multivariate statistical techniques. Chemometr. Intell. Lab. 107, 24–30. Comero, S., Servida, D., De Capitani, L., Gawlik, B.M., 2012. Geochemical characterization of an abandoned mine site: a combined positive matrix factorization and GIS approach compared with principal component analysis. J. Geochem. Explor. 118, 30–37. Dahle, S., Savinov, V.M., Matishov, G.G., Evenset, A., Næs, K., 2003. Polycyclic aromatic hydrocarbons (PAHs) in bottom sediments of the Kara Sea shelf, Gulf of Ob and Yenisei Bay. Sci. Total Environ. 306, 57–71. De Abreu-Mota, M.A., Barboza, C., Bicego, M.C., Martins, C.C., 2014. Sedimentary biomarkers along a contamination gradient in a human-impacted sub-estuary in Southern Brazil: a multi-parameter approach based on spatial and seasonal variability. Chemosphere 103, 156–163. Dong, T.T., Lee, B., 2009. Characteristics, toxicity, and source apportionment of polycylic aromatic hydrocarbons (PAHs) in road dust of Ulsan, Korea. Chemosphere 74, 1245–1253. Dong, Y., Wang, H., Han, G., Ke, C., Zhan, X., Nakano, T., Williams, G.A., 2012. The impact of Yangtze River discharge, ocean currents and historical events on the

biogeographic pattern of Cellana toreuma along the China coast. PLOS ONE 7 (4), e36178. Feely, R.A., Alin, S.R., Newton, J., Sabine, C.L., Warner, M., Devol, A., Krembs, C., Maloy, C., 2010. The combined effects of ocean acidification, mixing, and respiration on pH and carbonate saturation in an urbanized estuary. Estuar. Coast. Shelf S. 88, 442–449. Feng, C., Xia, X., Shen, Z., Zhou, Z., 2007. Distribution and sources of polycyclic aromatic hydrocarbons in Wuhan section of the Yangtze River, China. Environ. Monit. Assess. 133, 447–458. Feng, J., Li, X., Guo, W., Liu, S., Ren, X., Sun, J., 2014. Potential source apportionment of polycyclic aromatic hydrocarbons in surface sediments from the middle and lower reaches of the Yellow River, China. Environ. Sci. Pollut. Res. 21 (19), 11447–11456. Fernández-Luqueño, F., Valenzuela-Encinas, C., Marsch, R., Martínez-Suárez, C., Vázquez-Núñez, E., Dendooven, L., 2011. Microbial communities to mitigate contamination of PAHs in soil – possibilities and challenges, a review. Environ. Sci. Pollut. Res. 18, 12–30. Ghosal, D., Dutta, A., Chakraborty, J., Basu, S., Dutta, T.K., 2013. Characterization of the metabolic pathway involved in assimilation of acenaphthene in Acinetobacter sp strain AGAT-W. Res. Microbiol. 164, 155–163. Gonzalez-Macias, C., Sanchez-Reyna, G., Salazar-Coria, L., Schifter, I., 2014. Application of the positive matrix factorization approach to identify heavy metal sources in sediments. A case study on the Mexican Pacific Coast. Environ. Monit. Assess. 186, 307–324. Gupta, S., Kumar, K., Srivastava, A., Srivastava, A., Jain, V.K., 2011. Size distribution and source apportionment of polycyclic aromatic hydrocarbons (PAHs) in aerosol particle samples from the atmospheric environment of Delhi, India. Sci. Total Environ. 409, 4674–4680. Ha, H., Olson, J.R., Bian, L., Rogerson, P.A., 2014. Analysis of heavy metal sources in soil using kriging interpolation on principal components. Environ. Sci. Technol. 48 (9), 4999–5007. Han, Z., Jin, Y.Q., Yun, C.X., 2006. Suspended sediment concentrations in the Yangtze River estuary retrieved from the CMODIS data. Int. J. Remote Sens. 27 (19), 4329–4336. Huguet, A., Vacher, L., Relexans, S., Saubusse, S., Froidefond, J.M., Parlanti, E., 2009. Properties of fluorescent dissolved organic matter in the Gironde Estuary. Org. Geochem. 40, 706–719. Jaber, S.M., Ibrahim, K.M., Al-Muhtaseb, M., 2013. Comparative evaluation of the most common kriging techniques for measuring mineral resources using Geographic Information Systems. GISci. Remote Sens. 50 (1), 93–111. Jang, E., Alam, M.S., Harrison, R.M., 2013. Source apportionment of polycyclic aromatic hydrocarbons in urban air using positive matrix factorization and spatial distribution analysis. Atmos. Environ. 79, 271–285. Jiang, B., Zheng, H.L., Huang, G.Q., Ding, H., Li, X.G., Suo, H., Li, R., 2007. Characterization and distribution of polycyclic aromatic hydrocarbon in sediments of Haihe River, Tianjin, China. J. Environ. Sci. 19, 306–311. Keeler, G.J., Landis, M.S., Norris, G.A., Christianson, E.M., Dvonch, J.T., 2006. Sources of mercury wet deposition in eastern Ohio, USA. Environ. Sci. Technol. 40, 5874– 5881. Kim, J.Y., Lee, J.Y., Choi, S.D., Kim, Y.P., Ghim, Y.S., 2012. Gaseous and particulate polycyclic aromatic hydrocarbons at the Gosan background site in East Asia. Atmos. Environ. 49, 311–319. Kolarova, J., Svobodova, Z., Zlabek, V., Randak, T., Hajslova, J., Suchan, P., 2005. Organochlorine and PAHs in brown trout(Salmo Trutta fario) Population from Ticha Orlice River due to chemical plant with possible effects to vitellogenin expression. Fresen. Environ. Bull. 14, 1091–1096. Lee, B., Dong, T.T., 2010. Effects of road characteristics on distribution and toxicity of polycyclic aromatic hydrocarbons in urban road dust of Ulsan, Korea. J. Hazard. Mater. 175, 540–550. Li, B., Feng, C.H., Li, X., Chen, Y.X., Niu, J., Shen, Z.Y., 2012. Spatial distribution and source apportionment of PAHs in surficial sediments of the Yangtze Estuary, China. Mar. Pollut. Bull. 64, 636–643. Liu, Y., Chen, L., Huang, Q., Li, W., Tang, Y., Zhao, J., 2009. Source apportionment of polycyclic aromatic hydrocarbons (PAHs) in surface sediments of the Huangpu River, Shanghai, China. Sci. Total Environ. 407, 2931–2938. Liu, J.J., Wang, X.C., Fan, B., 2011. Characteristics of PAHs adsorption on inorganic particles and activated sludge in domestic wastewater treatment. Bioresour. Technol. 102, 5305–5311. Lohmann, R., Gioia, R., Jones, K.C., Nizzetto, L., Temme, C., Xie, Z., Schulz-Bull, D., Hand, I., Morgan, E., Jantunen, L., 2009. Organochlorine pesticides and PAHs in the surface water and atmosphere of the North Atlantic and Arctic Ocean. Environ. Sci. Technol. 43, 5633–5639. Luo, X., Chen, S., Mai, B., Sheng, G., Fu, J., Zeng, E.Y., 2008. Distribution, source apportionment, and transport of PAHs in sediments from the Pearl River Delta and the northern South China Sea. Arch. Environ. Con. Toxicol. 55, 11–20. Lv, J., Xu, J., Guo, C., Zhang, Y., Bai, Y., Meng, W., 2014. Spatial and temporal distribution of polycyclic aromatic hydrocarbons (PAHs) in surface water from Liaohe River Basin, northeast China. Environ. Sci. Pollut. Res.. 21. Nguyen, T.C., Loganathan, P., Nguyen, T.V., Vigneswaran, S., Kandasamy, J., Slee, D., Stevenson, G., Naidu, R., 2014. Polycyclic aromatic hydrocarbons in roaddeposited sediments, water sediments, and soils in Sydney, Australia: comparisons of concentration distribution, sources and potential toxicity. Ecotox. Environ. Safe. 104, 339–348. Niu, J., Chen, J., Martens, D., Quan, X., Yang, F., Kettrup, A., Schramm, K., 2003. Photolysis of polycyclic aromatic hydrocarbons adsorbed on spruce [Picea abies(L.) Karst.] needles under sunlight irradiation. Environ. Pollut. 123, 39–45.

W. Yu et al. / Chemosphere 134 (2015) 263–271 Norris, G., Vedantham, R., Wade, K., Brown, S., Prouty, J., Foley, C., 2008. EPA positive matrix factorization (PMF) 3.0 fundamentals & user guide. Prepared for the US Environmental Protection Agency, Washington, DC, by the National Exposure Research Laboratory, Research Triangle Park. Ou, D., Liu, M., Cheng, S., Hou, L., Xu, S., Wang, L., 2010. Identification of the sources of polycyclic aromatic hydrocarbons based on molecular and isotopic characterization from the Yangtze estuarine and nearby coastal areas. J. Geogr. Sci. 20, 283–294. Paatero, P., 2007. User’s Guide for Positive Matrix Factorization Programs PMF2 and PMF3, Part 1–2: Tutorial. University of Helsinki, Helsinki, Finland. Paatero, P., Hopke, P.K., 2003. Discarding or downweighting high-noise variables in factor analytic models. Anal. Chim. Acta 490, 277–289. Paatero, P., Tapper, U., 1993. Analysis of different modes of factor analysis as least squares fit problems. Chemometr. Intell. Lab. 18, 183–194. Paatero, P., Tapper, U., 1994. Positive matrix factorization: a nonn-egative factor model with optimal utilization of error estimates of data values. Environmetrics 5, 111–126. Pan, G., You, C., 2010. Sediment-water distribution of perfluorooctane sulfonate (PFOS) in Yangtze River Estuary. Environ. Pollut. 158, 1363–1367. Pekey, H., Dogan, G., 2013. Application of positive matrix factorisation for the source apportionment of heavy metals in sediments: a comparison with a previous factor analysis study. Microchem. J. 106, 233–237. Reff, A., Eberly, S.I., Bhave, P.V., 2007. Receptor modeling of ambient particulate matter data using positive matrix factorization: review of existing methods. J. Air Waste Manage. 57, 146–154. Saba, T., Su, S., 2013. Tracking polychlorinated biphenyls (PCBs) congener patterns in Newark Bay surface sediment using principal component analysis (PCA) and positive matrix factorization (PMF). J. Hazard. Mater. 260, 634–643. Silva, T.R., Lopes, S., Sporl, G., Knoppers, B.A., Azevedo, D.A., 2013. Evaluation of anthropogenic inputs of hydrocarbons in sediment cores from a tropical Brazilian estuarine system. Microchem. J. 109, 178–188. Simcik, M.F., Eisenreich, S.J., Lioy, P.J., 1999. Source apportionment and source/sink relationships of PAHs in the coastal atmosphere of Chicago and Lake Michigan. Atmos. Environ. 33, 5071–5079. Stout, S.A., Graan, T.P., 2010. Quantitative source apportionment of PAHs in sediments of little Menomonee River, Wisconsin: weathered creosote versus urban background. Environ. Sci. Technol. 44 (8), 2932–2939. Teixeira, E.C., Mattiuzi, C.D., Agudelo-Castaeda, D.M., Garcia, K.D., Wiegand, F., 2013. Polycyclic aromatic hydrocarbons study in atmospheric fine and coarse particles using diagnostic ratios and receptor model in urban/industrial region. Environ. Monit. Assess. 185, 9587–9602. Tobiszewski, M., Namiesnik, J., 2012. PAH diagnostic ratios for the identification of pollution emission sources. Environ. Pollut. 162, 110–119. Vaccaro, S., Sobiecka, E., Contini, S., Locoro, G., Free, G., Gawlik, B.M., 2007. The application of positive matrix factorization in the analysis, characterisation and detection of contaminated soils. Chemosphere 69, 1055–1063. Vardar, N., Esen, F., Tasdemir, Y., 2008. Seasonal concentrations and partitioning of PAHs in a suburban site of Bursa, Turkey. Environ. Pollut. 155, 298–307. Vu, V., Lee, B., Kim, J., Lee, C., Kim, I., 2011. Assessment of carcinogenic risk due to inhalation of polycyclic aromatic hydrocarbons in PM10 from an industrial city: a Korean case-study. J. Hazard. Mater. 189, 349–356.

271

Wang, D., Yang, M., Jia, H., Zhou, L., Li, Y., 2009. Polycyclic aromatic hydrocarbons in urban street dust and surface soil: comparisons of concentration, profile, and source. Arch. Environ. Contamin. Toxicol. 56, 173–180. Wang, H., Cheng, Z., Liang, P., Shao, D., Kang, Y., Wu, S., Wong Chris, K.C., Wong, M.H., 2010. Characterization of PAHs in surface sediments of aquaculture farms around the Pearl River Delta. Ecotox. Environ. Safe. 73, 900–906. Wang, W., Huang, M., Kang, Y., Wang, H., Leung, A.O., Cheung, K.C., Wong, M.H., 2011. Polycyclic aromatic hydrocarbons (PAHs) in urban surface dust of Guangzhou, China: status, sources and human health risk assessment. Sci. Total Environ. 409, 4519–4527. Wang, Y., Li, X., Li, B.H., Shen, Z.Y., Feng, C.H., Chen, Y.X., 2012. Characterization, sources, and potential risk assessment of PAHs in surface sediments from nearshore and farther shore zones of the Yangtze estuary, China. Environ. Sci. Pollut. Res. 19, 4148–4158. Wang, Z.M., Chen, L.D., Zhang, H.P., Sun, R.H., 2014. Multivariate Statistical Analysis and Risk Assessment of Heavy Metals Monitored in Surface Sediment of the Luan River and its Tributaries, China. Hum. Ecol. Risk Assess. 20, 1521–1537. Wang, H.T., Wang, J.W., Liu, R.M., 2015. Spatial variation, environmental risk and biological hazard assessment of heavy metals in surface sediments of the Yangtze River estuary. Mar. Pollut. Bul. http://dx.doi.org/10.1016/ j.marpolbul.2015.01.026. Xu, K., Milliman, J.D., 2009. Seasonal variations of sediment discharge from the Yangtze River before and after impoundment of the Three Gorges Dam. Geomorphology 104, 276–283. Yan, W., Chi, J., Wang, Z., Huang, W., Zhang, G., 2009. Spatial and temporal distribution of polycyclic aromatic hydrocarbons (PAHs) in sediments from Daya Bay, South China. Environ. Pollut. 157, 1823–1830. Yancheshmeh, R.A., Bakhtiari, A.R., Mortazavi, S., Savabieasfahani, M., 2014. Sediment PAH, contrasting levels in the Caspian Sea and Anzali Wetland. Mar. Pollut. Bull., 84 Yang, Z., Wang, H., Saito, Y., Milliman, J.D., Xu, K., Qiao, S., Shi, G., 2006. Dam impacts on the Changjiang (Yangtze) River sediment discharge to the sea: The past 55 years and after the Three Gorges Dam. Water Resour. Res., 42 Zakaria, M.P., Takada, H., Tsutsumi, S., Ohno, K., Yamada, J., Kouno, E., Kumata, H., 2002. Distribution of polycyclic aromatic hydrocarbons (PAHs) in rivers and estuaries in Malaysia: a widespread input of petrogenic PAHs. Environ. Sci. Technol. 36, 1907–1918. Zhang, X.L., Tao, S., Liu, W.X., Yang, Y., Zuo, Q., Liu, S.Z., 2005. Source diagnostics of polycyclic aromatic hydrocarbons based on species ratios: a multimedia approach. Environ. Sci. Technol. 39, 9109–9114. Zhang, J., Wang, J., Hua, P., Krebs, P., 2015. The qualitative and quantitative source apportionments of polycyclic aromatic hydrocarbons in size dependent road deposited sediment. Sci. Total Environ. 505, 90–101. Zhou, S., Yang, H., Zhang, A., Li, Y., Liu, W., 2014. Distribution of organochlorine pesticides in sediments from Yangtze River Estuary and the adjacent East China Sea, Implication of transport, sources and trends. Chemosphere 114, 26–34.