Source identification of heavy metals in peri-urban agricultural soils of southeast China: An integrated approach

Environmental Pollution 237 (2018) 650e661

Contents lists available at ScienceDirect

Environmental Pollution journal homepage: www.elsevier.com/locate/envpol

Source identification of heavy metals in peri-urban agricultural soils of southeast China: An integrated approach Wenyou Hu a, b, Huifeng Wang a, c, Lurui Dong d, Biao Huang a, *, Ole K. Borggaard b, Hans Christian Bruun Hansen b, e, Yue He f, **, Peter E. Holm b, e a

Key Laboratory of Soil Environment and Pollution Remediation, Institute of Soil Science, Chinese Academy of Sciences, Nanjing, 210008, China University of Copenhagen, Faculty of Science, Department of Plant and Environmental Sciences, Thorvaldsensvej 40, DK-1871, Frederiksberg C, Denmark University of Chinese Academy of Sciences, Beijing, 100049, China d Nanjing Research Institute of Environmental Protection, Nanjing, 210013, China e Sino-Danish Center for Education and Research (SDC), China f Ministry of Environmental Protection of the People’s Republic of China, Nanjing Institute of Environmental Sciences, Nanjing 210042, China b c

a r t i c l e i n f o

a b s t r a c t

Article history: Received 13 September 2017 Received in revised form 9 February 2018 Accepted 22 February 2018

Intensive human activities, in particular agricultural and industrial production have led to heavy metal accumulation in the peri-urban agricultural soils of China threatening soil environmental quality and agricultural product security. A combination of spatial analysis (SA), Pb isotope ratio analysis (IRA), input fluxes analysis (IFA), and positive matrix factorization (PMF) model was successfully used to assess the status and sources of heavy metals in typical peri-urban agricultural soils from a rapidly developing region of China. Mean concentrations of Cd, As, Hg, Pb, Cu, Zn and Cr in surface soils (0e20 cm) were 0.31, 11.2, 0.08, 35.6, 44.8, 119.0 and 97.0 mg kg
1, respectively, exceeding the local background levels except for Hg. Spatial distribution of heavy metals revealed that agricultural activities have significant influence on heavy metal accumulation in the surface soils. Isotope ratio analysis suggested that fertilization along with atmospheric deposition were the major sources of heavy metal accumulation in the soils. Based on the PMF model, the relative contribution rates of the heavy metals due to fertilizer application, atmospheric deposition, industrial emission, and soil parent materials were 30.8%, 33.0%, 25.4% and 10.8%, respectively, demonstrating that anthropogenic activities had significantly higher contribution than natural sources. This study provides a reliable and robust approach for heavy metals source apportionment in this particular peri-urban area with a clear potential for future application in other regions. © 2018 Elsevier Ltd. All rights reserved.

Keywords: Heavy metals Spatial variability Pb isotope ratio Positive matrix factorization (PMF) Input fluxes

1. Introduction There is an increasing public concern about the accumulation of heavy metals in agricultural soils which, in turn, has the potential to restrict the soil's function, cause toxicity to crops and ground water, and hence to threaten human health (Hou et al., 2014; Lu et al., 2015; Qu et al., 2016; Toth et al., 2016). Intensive human activities

* Corresponding author. Key Laboratory of Soil Environment and Pollution Remediation, Institute of Soil Science, Chinese Academy of Sciences, Nanjing, 210008, China. ** Corresponding author. Ministry of Environmental Protection of the People's Republic of China, Nanjing Institute of Environmental Sciences, Nanjing 210042, China. E-mail addresses: [email protected] (W. Hu), [email protected] (B. Huang), [email protected] (Y. He). https://doi.org/10.1016/j.envpol.2018.02.070 0269-7491/© 2018 Elsevier Ltd. All rights reserved.

have led to heavy metal accumulation in peri-urban agricultural soils of China threatening soil environmental quality and food safety (Huang et al., 2006; Luo et al., 2009; Hu et al., 2013; Hu et al., 2017). Heavy metals can enter agro-ecosystems through geogenic sources and anthropogenic activities (Cloquet et al., 2006; Yang et al., 2016). Geogenic sources of heavy metals mainly come from weathering of the parent materials. Anthropogenic activities include inputs of heavy metals through application of fertilizers and organic manures, irrigation, atmospheric deposition, waste disposal, sewage application, and other human activities (Sharma et al., 2008; Hu et al., 2013; Hou et al., 2014; Pan and Wang, 2015). Inputs of heavy metals to soils through agricultural activities have increased within the past decades due to increasing food demands from a rapidly expanding population (Huang et al., 2015; Hu et al., 2017). Heavy metal source apportionment is a crucial step towards

W. Hu et al. / Environmental Pollution 237 (2018) 650e661

prevention or reduction of heavy metal pollution (Huang et al., 2015). Identification of heavy metal sources in agricultural soils is a basis for undertaking appropriate actions to protect soil quality and to develop sustainable management strategies (Lu et al., 2012). In this context, peri-urban agricultural soils are priority areas for research of source apportionment as they are generally located close to multiple pollutant sources such as construction/excavation works, industry, traffic and urban waste disposal (Huang et al., 2015). Despite public concerns, quantitative knowledge of heavy metals in agricultural soils from different sources especially from different anthropogenic sources remains scarce. Discriminating the natural and different anthropogenic sources and their rates of contribution to heavy metal accumulation in soils are crucial for soil environmental protection and food safety (Pan and Wang, 2015). The spatial distribution of heavy metals based on Geographical Information System (GIS) can be used as an aid to identify their possible sources and pollution hot spots (Chai et al., 2015). Identification of soil heavy metal sources and spatial delineation of areas with heavy metal pollution is important for decision makers to develop effective management strategies to improve environmental quality (Zhao et al., 2010; Pan and Wang, 2015). So far, studies on regional input and output fluxes of heavy metals are mostly based on model calculations, statistical yearbooks, and literature data (Luo et al., 2009; Belon et al., 2012; Lofts et al., 2013; Hou et al., 2014). This classical approach to source apportionment can be substantially improved by use of receptor models that are based on application of multivariate statistical methods to identify and quantify apportionment of pollutants to their sources (Wang et al., 2009). Although the positive matrix factorization (PMF) model has been successfully applied for pollutant source identification for atmospheric (Song et al., 2006; Alleman et al., 2010; Gupta et al., 2012; Jang et al., 2013) and sedimentary sources (Chen et al., 2013; Pekey and Dogan, 2013; Gonzalez-Macias et al., 2014), few studies have employed this approach to identify heavy metal sources in soils (Xue et al., 2014; Dong et al., 2015). Furthermore, stable Pb isotope ratio analysis is commonly used to trace the sources of Pb pollution in different environmental compartments at local to global scales (Wong et al., 2003; Cloquet et al., 2006; Komarek et al., 2008; Reimann et al., 2012; Yu et al., 2016). Although previous studies have been conducted to identify the sources of heavy metals in soils using the different individual method, integration of the different methods in the same area to accurately evaluate and validate the source identification results by the different approaches is lacking. Therefore, a combination of spatial analysis (SA), isotope ratio analysis (IRA), input flux analysis (IFA), and positive matrix factorization (PMF) model has been used in this study to identify the status and sources of selected heavy metals (Cd, As, Hg, Pb, Cu, Zn and Cr) in typical peri-urban agricultural soils. The tested soils were from a rapidly developing region in southeast China, where a large number of samples of soils, crops, fertilizers and atmospheric depositions were collected and analysed. This work provides baseline information to develop effective policies and standards to control and reduce heavy metal inputs and long-term accumulation in agricultural soils as well as to provide the theoretical basis and technical support for sustainable agricultural production and management. 2. Materials and methods 2.1. Description of the study area The selected study area is located on an alluvial island of the Yangtze River, a peri-urban area of Nanjing City (32 80 2400 -

651

32130 3700 N, 118 460 2400 -118 490 4700 E), Jiangsu Province, southeast China, with a total land area of 55.6 km2 (Fig. 1). The area is within a subtropical monsoon climate zone with a mean annual temperature, precipitation, and potential evaporation of about 15e16  C, 1100 mm and 1200 mm, respectively. The prevailing wind directions are northwest in winter and southeast in summer. The main soil type in this area is Cambosols (Fluvo-aquic soils) (CRGCST, 2001). The agricultural production is very intensive due to the increasing food demands from urban areas (Hu et al., 2014). Wheat, rice, maize and several types of vegetables are the main crops. The rotation patterns of the agricultural production include vegetablevegetable rotation, crop-vegetable rotation and crop-crop rotation. According to our interviews with farmers and the available information at agricultural centers in the area, these intensive agricultural production practices are characterized by extensive fertilization which may result in heavy metal accumulation in the soils. Intensive industrial activity including iron and steel, chemical, thermo-electric, and medical industries, as well as sewage treatment plants built after 1990s are mainly located on the opposite side of the Yangtze River relative to the study area (Fig. 1). 2.2. Sampling, processing and analysis Eighty-eight surface soil samples were collected in summer 2012 throughout the island based on land use and spatial homogeneity. The sampling sites covered nearly all land uses, including several uncultivated soils along the Yangtze River. All sampling sites were geo-located using a global positioning system receiver (Fig. 1). Each sample consisted of a mixture of five subsamples collected from five spots of an area of about 5 m2. An amount of 1 kg fresh soil samples were collected to provide a representative sample of each soil. All soil subsamples were collected at a depth of 0e20 cm using a stainless steel shovel. Furthermore, three representative soil profile samples were taken at depths of 0e20, 20e40, 40e60, 60e80, and 80e100 cm. A total of nine typical crop samples (3 rice, 3 wheat and 3 leafy vegetable samples) were collected during soil sampling. Twenty eight chemical fertilizers and five commercial organic fertilizers were collected from adjacent agricultural stores or local farmers. For livestock manure, thirteen fresh samples were taken at 1 m depth from five points inside the manure pile using a soil sampler, and were mixed and transferred to the laboratory for analysis. Seven atmospheric wet and dry deposition collection sites were established in the different direction of the study area (Fig. 1), and samples were collected monthly for one year from December 2011 to November 2012 using a conventional wetedry automatic sampler. The wetedry sampler was equipped with a 707 cm2 aperture and a 177 cm2 PUF-based glass bucket in separate containers to sample daily rainfall and monthly particulate dry deposition, respectively (Pan and Wang, 2015). All equipment in contact with the samples were carefully cleaned with 10% HCl solution, and kept in plastic bags until they were used for sample collection (Cui et al., 2014). All solid samples, including soils and potential source samples were ground in a mortar to pass through a 0.149 mm (100mesh) polyethylene sieve and stored at room temperature until analysis. Soil samples were also crushed and passed through a 2mm mesh sieve for soil pH and SOM determination. Soil pH was measured by potentiometry on 1:2.5 (soil: water) paste using a glass electrode pH meter (PHS-3C, Shanghai, China). Soil organic matter (SOM, g kg
1) was determined using the WalkleyeBlack method (Nelson and Sommers, 1996). To measure the contents of Cd, Pb, Cr, Cu and Zn in the soils, fertilizers and atmospheric wet and dry deposition, the 100-mesh sieved samples were digested by a mixture of HNO3 (5 mL)-HClO4 (1 mL)-HF (1 mL) in a poly-tetrafluoroethylene container. The mixture was heated to

652

W. Hu et al. / Environmental Pollution 237 (2018) 650e661

Fig. 1. Distribution of sampling sites in the study area.

180  C for 10 h, cooled to room temperature, and diluted with deionized water to 30 mL. The contents of the heavy metals were determined by inductively coupled plasma mass spectroscopy (X7 Quadrupole ICP-MS, American Thermo Scientific). The contents of As and Hg were measured using atomic fluorescence spectrometry (AFS; Beijing Jitian Instruments Co., Ltd. production, AFS-820). Analytical quality control was conducted using soil certified reference materials (GBW07403 and GBW07404) obtained from the National Standard Detection Research Center, Beijing, China. The recovery rates for Cd, As, Hg, Pb, Cu, Zn and Cr in the standard reference materials were 99e103%, 94e107%, 100e101%, 97e100%, 98e103%, 94e96% and 99e101%, respectively (Table S1). Duplicate samples were analyzed simultaneously for 20% of the soil samples, and the relative standard deviations (RSD) for these replicate samples were less than 5%. The detection limits of ICP-MS for Cd, Pb, Cu, Zn and Cr were 0.01 mg kg
1, 0.1 mg kg
1, 0.1 mg kg
1, 0.1 mg kg
1 and 0.5 mg kg
1, respectively. The detection limits of AFS for As and Hg were 0.8 mg kg
1 and 0.0005 mg kg
1, respectively. Lead isotope ratios in different environmental samples were measured using quadrupole ICP-MS (Agilent 7500a, USA), including 11 surface soil samples, 3 soil profile samples, 9 crop samples, 8 atmospheric deposition samples, and 5 fertilizer samples. The main operating conditions of the ICP-MS for Pb isotope ratios analysis are summarized in Table S2. In consideration of element abundances and isobaric elements, the 206Pb/ 207Pb and 206Pb/ 208Pb isotope ratios were determined. A Tl isotope standard solution (SRM NIST 997) was used to correct the instrumental mass bias and Pb isotopic standard material (SRM NIST 981) was used to correct the isotope mass discrimination effect. Standard solutions of Pb isotopes were measured after each set of five unknown samples and were used to verify the results (Huang et al., 2015). The RSD of the Pb isotopic ratios for five replicate samples were lower than 1%. 2.3. Data processing and statistical analysis 2.3.1. Annual input fluxes of heavy metals from fertilizers Heavy metal inputs to agricultural land from fertilizers were estimated from the annual amounts of fertilizer applied and the contents of heavy metals in the fertilizers as follows (Hou et al., 2014):

IFer ¼

n X

FijCij10
6

(1)

j¼1

Where IFer is the amount of heavy metal (i) input from fertilizers (j) (g ha
1 y
1), Fij is the amount of fertilizer actually applied (g ha
1 y
1), and Cij is the concentration of heavy metal in the fertilizer (mg kg
1). The annual amount of fertilizers was recorded from field surveys in the study area. 2.3.2. Annual input fluxes of heavy metals from atmospheric deposition The input of heavy metals to agricultural soils from atmospheric wet and dry depositions was calculated according to the following equation:

IAt ¼ Cd*Vd*100=S

(2)

Where IAt is the amount of heavy metals input from atmospheric wet and dry depositions (g ha
1 y
1), Cd is the concentration of heavy metals in the wet and dry deposition (mg L
1), Vd is the annual volume of wet and dry deposition from sampling bottle (L), S is the area of sampling bottle (cm2), and 100 is a conversion factor. 2.3.3. Positive matrix factorization (PMF) model Positive matrix factorization (PMF) is a principal component analysis (PCA)-based mathematical receptor model for quantifying the contribution of sources to samples based on the composition or fingerprints of the pollution sources (Paatero, 1997; Wang et al., 2009; Xue et al., 2014). The PMF model decomposes a matrix of speciated sample data into two matrices: factor contributions (G) and factor profiles (F). These factor profiles need to be interpreted by the user to identify the source types that may be contributing to the sample using measured source profile information, and emissions or discharge inventories (USEPA, 2014). A speciated data set can be viewed as a data matrix X of i by j dimensions, in which i number of samples and j chemical species were measured. The goal of the PMF model is to solve the chemical mass balance between measured species concentrations and source profiles, as shown in Equation (3), with number of factors p, the species profile f of each source, and the amount of mass g

W. Hu et al. / Environmental Pollution 237 (2018) 650e661

contributed by each factor to each individual sample (USEPA, 2014):

xij ¼

p X

gik fkj þ eij

(3)

k¼1

Where xij is the matrix of sample concentrations; gik is the contribution of each factor to any given sample; fkj is the matrix of chemical compositions of p sources; eij is the matrix of residual for each sample. Factor contributions and profiles are derived by the PMF model minimizing the objective function Q defined as follows (Paatero, 1997; Wang et al., 2009): n X m X eij Q¼ u ij i¼1 j¼1

!2 (4)

Where Q is the sum of the squares of the difference (i.e., eij) between the original dataset (xij) and the PMF output (gik fkj), weighted by the measurement uncertainties (uij). The PMF model was run using concentration data (including 7 elements in 88 soil samples) and uncertainty data files which encompass errors such as sampling and analytical errors. More details about the PMF receptor model can be found elsewhere (Paatero, 1997; Paatero and Hopke, 2009; USEPA, 2014). 2.3.4. Statistical analysis Statistical analysis of data was performed using SPSS 17.0 and Microsoft Excel 2013. ArcGIS 10.2 software (ESRI, US) was used for mapping the sampling sites. The geostatistical analyses and distribution of heavy metals were also mapped with ArcGIS using ordinary Kriging interpolation analysis and a spherical model. Source apportionment of heavy metals was identified by the positive matrix factorization (PMF) Model (USEPA PMF 5.0). Spearman correlation was used to recognize the correlations among soil pH, OM and heavy metals in soils. Profile distribution figures of soil heavy metals, composition of Pb isotope ratios and annual input fluxes of heavy metals in fertilizers and atmospheric deposition were constructed in Origin 8.5 software. 3. Results and discussion 3.1. Descriptive statistics of soil properties and heavy metals in soils A summary of selected soil physicochemical properties and concentrations of heavy metals are presented in Table 1. The mean soil pH value was 6.78 and the minimum was 3.90 which indicated weak acidity in most soils and moderate to strong acidity at some sites. The soil organic matter (SOM) contents were relatively high (10.2e41.7 g kg
1) with a mean value of 23.1 g kg
1. Low pH and higher SOM in some soils could be related to the intensive vegetable production with high application rates of fertilizers and organic fertilizers and/or acidic pH slowing down mineralization processes (Hu et al., 2017; Yang et al., 2016). Mean concentrations of Cd, As, Hg, Pb, Cu, Zn and Cr were 0.31, 11.2, 0.08, 35.6, 44.8, 119, and 97 g kg
1, respectively. The maximum concentrations of Cd, As, Hg, Pb, Cu, Zn, and Cr in the surface soils were 0.58, 20.6, 0.31, 150, 58.1, 181, and 120 mg kg
1, respectively, exceeding the local background levels by a factor of 1.6e4.8. According to the Chinese soil environment quality standard (MEP, 1995), the maximum Cd concentration level exceeded the 0.3 mg kg
1 limit level by 49% implying higher accumulation of this element than of the other heavy metals (Table 1). In fact, there is a strong focus on Cd in Chinese agricultural soils with intensive monitoring to prevent further accumulation (Luo et al., 2009; Yang

653

et al., 2016). Enrichment factor (EF) can be utilized to evaluate the degree of anthropogenic influence on soil contamination by heavy metals (Luo et al., 2015). For EF < 1, heavy metals mainly origin from natural sources, while EF > 1 is indicative of anthropogenic contamination. The mean EF values of Cd, As, Hg, Pb, Cu, Zn and Cr were 2.12, 1.21, 1.02, 1.52, 1.38, 1.44 and 1.24, respectively. The EF for Cd was the highest value, indicating that the studied soils were relatively higher enriched by this element. The coefficient of variation of the heavy metals ranged from 2.40% to 7.48%, reflecting a low spatial variability, indicating that these metals originate from similar sources (Xue et al., 2014). Fig. 2 shows distributions of pH, SOM and heavy metals in the three soil profiles. Soil pH was lowest in surface soils, and increased significantly with increasing depths. High fertilizer applications and acid atmospheric depositions combined with insufficient liming may have decreased pH and increased the heavy metal content in surface soils (Atafar et al., 2010; Yang et al., 2016). The highnitrogen (N) fertilizer inputs and the uptake and removal of base cations by plants are the most important reasons that soil pH decreases in agricultural top soils (Guo et al., 2010). The top soils had the highest contents of SOM, Cd, As, Hg, Pb, Cu, and Zn, and the contents decreased significantly with increasing soil depth. The vegetable-vegetable rotation soil (P1) had lower pH and higher SOM as well as higher contents of heavy metals compared to the maize-wheat rotation soil (P3) and uncultivated soil (P2). The lower pH and higher SOM and heavy metals in vegetable-vegetable rotation soils were attributed to the higher application of fertilizers during intensive vegetable production. These results are consistent with a previous study (Hu et al., 2013), in which SOM, Cd, Pb, Cu and Zn had strongly accumulated in vegetable soils. 3.2. Spatial variability of soil properties and heavy metals in soils The spatial distribution of pH, SOM and the heavy metals are shown in Fig. 3. Soil pH was lowest in the southeastern part of the study area, while the spatial distribution of SOM was nearly opposite to soil pH. The spatial distributions of Cd, As, Pb, Zn and in particular Cu were similar to SOM with the highest concentrations in the southeastern parts of the study area. The highest contents of Hg occurred in three, widely spaced minor spots. Unlike other elements, the highest contents of Cr occurred in the northwestern and middle parts of the study area. In the southeastern part with the lowest pH and highest SOM and heavy metal contents, land use is dominated by intensive vegetable production with relatively large fertilization rates. According to the previous studies (Huang et al., 2006; Hu et al., 2013), higher application of fertilizers in intensive vegetable production soils not only resulted in the decrease of soil pH, but also leaded to the increase of SOM and heavy metals. In order to further explain the spatial patterns, a correlation analysis was used to recognize the correlations among soil pH, SOM and heavy metals in the soils (Table S3). In general, significant negative correlations were observed among soil pH and concentrations of As, Pb, Cu, Zn and Cr, while significant positive correlations among SOM and the concentrations of Cd, As, Pb, Cu and Zn were observed. The spatial heterogeneity of soil properties, land uses and multiple sources of heavy metals would inevitably lead to patchy patterns of heavy metals in surface soils. 3.3. Heavy metals in fertilizers and atmospheric deposition According to our previous study of Nanjing vegetable plots, heavy metals contribution via irrigation water was minimal and accounted for only 1e3% of total anthropogenic inputs (Hu et al., 2013). According to our field investigation in the study area,

654

W. Hu et al. / Environmental Pollution 237 (2018) 650e661

Table 1 Heavy metal concentrations together with pH and SOM contents in the surface soils. pH

Mean Standard deviation Minimum Maximum C.V. a (%) Skewness Kurtosis Background b EF c Limit levels d (MEP, 1995)

6.78 0.98 3.90 7.83 6.91
1.12 0.27

pH < 6.5 6.5 < pH < 7.5 pH > 7.5

SOM

e

Cd

(g kg
1)

(mg kg
1)

23.1 6.95 10.2 41.7 3.33 0.57 0.09

0.31 0.07 0.19 0.58 4.54 1.59 4.38 0.20 2.12 0.3 0.3 0.6

As

Hg

Pb

Cu

Zn

Cr

11.2 2.18 6.91 20.6 5.11 1.17 4.51 9.24 1.21 40 30 25

0.08 0.03 0.04 0.31 2.42 4.23 24.1 0.08 1.02 0.3 0.5 1

35.6 14.8 21.1 150 2.40 5.69 41.5 31.3 1.52 250 300 350

44.8 7.23 22.4 58.1 6.19
0.98 0.92 31.7 1.38 50 100 100

119 19.7 75.6 181 6.03 0.43 1.41 80.7 1.44 200 250 300

97.0 13.0 56.5 120 7.48
0.49 0.21 77 1.24 150 200 250

a

C.V. means coefficient of variation. Soil heavy metal background value of Nanjing, Jiangsu Province (Li and Zheng, 1989). c EF means enrichment factor. The EF of each heavy metal (HM) relative to the background using Fe as the reference element was calculated by the ratios of element concentration (mg kg
1): EF ¼ [HM/Fe]sample/[HM/Fe]background (Luo et al., 2015). d Grade II limit levels of heavy metals set by the Chinese Environmental Quality Standard for Soils based on soil pH (MEP, 1995). e SOM: soil organic matter. b

Fig. 2. Distribution of pH, SOM and heavy metals with depth in three soil profiles (P1: vegetable-vegetable rotation soil; P2: uncultivated soil; P3: maize-wheat rotation soil).

W. Hu et al. / Environmental Pollution 237 (2018) 650e661

655

Fig. 3. Spatial distribution of pH, SOM and heavy metal contents in the study area.

livestock manures, organic and chemical fertilizers were intensively applied to the soils to increase crop yields. Mean concentrations of the heavy metals in different fertilizers are shown in Table S1. Concentrations of Cd, As, and Hg in chemical fertilizers, Pb and Cr in commercial organic fertilizers, and Cu and Zn in livestock manure were relatively high. Heavy metals in the fertilizers pose an environmental risk to agricultural production if applied in high quantities over a longer time period (Luo et al., 2009). The Cd contents increased in the order: commercial organic fertilizer < livestock manure < chemical fertilizer (Table S4). According to previous studies, chemical fertilizers especially phosphate fertilizers were the major source of Cd due to its presence in the phosphate rocks (Atafar et al., 2010; Lu et al., 2012). Livestock

manures are important sources of soil Cu and Zn pollution because of their use as feed additives to promote animal growth and to control diseases (Nicholson et al., 2003; Lu et al., 2012). Compared to current permissible levels of fertilizers (QSQA, 2009; MA, 2012), mean concentrations of Cd, As, and Cr in some fertilizers exceeded the threshold levels, while the concentrations of other elements seem relatively safe. However, long-term high fertilizer application could inevitably lead to heavy metal accumulation in soils and create environmental problems (Atafar et al., 2010; Hu et al., 2013; Chen et al., 2014; Xue et al., 2014; Hu et al., 2017). Atmospheric deposition has an important influence on urban dust and rural surface soils (Yu et al., 2016). The concentrations of heavy metals in atmospheric dry and wet deposition are shown in

656

W. Hu et al. / Environmental Pollution 237 (2018) 650e661

Table S5. The Cd, As, Hg, Pb, Cu, Zn, and Cr concentrations in the atmospheric deposition were much higher than their concentrations in soils (Table 1). The most abundant heavy metals in atmospheric deposition was Zn, followed by Pb > Cu > Cr > As > Cd > Hg. Comparable results were reported in agricultural soils of China (Luo et al., 2009; Pan and Wang, 2015) and Europe (Nicholson et al., 2003; Belon et al., 2012). The variability of Pb in the atmospheric deposition is particularly high (Table S2), which may be ascribed to industrial activities near the agricultural area (Zhao et al., 2010).

3.4. Pb isotope ratios in different environmental media For further understanding of the origin of heavy metals in the soils, the concentration of Pb and 206Pb/207Pb in different environmental samples were determined (Fig. 4). The 206Pb/207Pb ratio is most commonly used in environmental studies because it can be determined precisely, and the abundances of these isotopes are relatively important (Komarek et al., 2008; Walraven et al., 2013).The results show that the concentration of Pb increases in the following order: crop < fertilizers < deep soils < surface soils < atmospheric deposition (Fig. 4a). In contrast, the 206Pb/207Pb

Fig. 4. Concentrations of Pb and Pb isotope ratios in different environmental media.

W. Hu et al. / Environmental Pollution 237 (2018) 650e661

ratio increases in the order of atmospheric deposition < crop < fertilizers < deep soils < surface soils. Atmospheric deposition has a wider 206Pb/207Pb range which covers the variability range from vehicle exhaust to industrial emissions (Yu et al., 2016). The 206Pb/207Pb of surface and deep soils also covers the wider ratio ranges, demonstrating a wide range of Pb sources in the studied soils. A common approach in the use of Pb isotopes for source identification is to use cross-plots of the isotope ratios, e.g. 206Pb/207Pb versus 208Pb/207Pb, or 206Pb/207Pb versus 208Pb/206Pb (Komarek et al., 2008; Reimann et al., 2012; Walraven et al., 2013). To distinguish the different potential Pb sources, values for the 208 Pb/206Pb ratio were plotted against values for the 206Pb/207Pb ratio (Fig. 4b). The mean 206Pb/207 Pb ratio in fertilizer and atmospheric deposition in the study area were 1.1554 and 1.1434, respectively, which were similar to the results in other peri-urban agricultural soils of China, with 206Pb/207Pb ratio of 1.1592 for fertilizer and 1.1494 for atmospheric deposition (Huang et al., 2015). The 208Pb/206Pb ratios for the top soils were close to ratios for fertilizer and atmospheric deposition (Fig. 4b), further suggesting that fertilization along with atmospheric deposition were the major sources of Pb accumulation in the agricultural top soils. This is consistent with a previous finding in rural soils in the Netherlands, where the Pb isotope composition also attributes a mixture of atmospheric sources and fertilizers (Walraven et al., 2013). Pollution is generally recognized by elevated contents of Pb and/ or other heavy metals in the topmost soil layers (Hansmann and Koppel, 2000). The 206Pb/207Pb ratios obtained for the different soil profiles showed similar trends with increasing ratios at increasing soil depths (Fig. 5). The isotope ratios of 206Pb/207Pb were significantly lower in top soils than deeper layers, implying that the top soils were more close to anthropogenic Pb isotopic composition. Generally, the depth at which anthropogenic Pb was

657

reported generally did not exceed 30 cm (Walraven et al., 2013). The greater the anthropogenic influences, the less isotope composition and the lower 206Pb/207Pb ratios. The result further indicates that accumulation of heavy metals in surface soils are significantly affected by human activities. The isotope ratios of 206Pb/207Pb in deeper layer soils were more close to natural Pb isotopic composition, implying less influence by anthropogenic sources. Similar profiles of Pb decline combined with the increase of 206Pb/207Pb were also observed in the Dutch agricultural soils (Walraven et al., 2013) and the Ulsan Bay sediments, East Sea, Korea (Chae et al., 2014). 3.5. Input fluxes of heavy metals from fertilizers and atmospheric deposition The annual input fluxes of heavy metals from fertilizers and atmospheric deposition are shown in Fig. 6. The fertilizer and atmospheric deposition fluxes exhibited a wide range among the elements. The annual input fluxes of Cd, As, Hg, Pb, Cu, Zn, and Cr in fertilizers were 8.94, 86.29, 0.60, 156, 539, 2917, and 480 g ha
1, respectively. The annual input fluxes of heavy metals with fertilizers increased in the order: Hg < Cd < As < Pb < Cr < Cu < Zn, in agreement with the inputs of heavy metals in agricultural soils in the Yangtze River delta, China (Hou et al., 2014) and in French agricultural soils (Belon et al., 2012). The annual input fluxes of Cd, As, Hg, Pb, Cu, Zn, and Cr in atmospheric deposition were 7.00, 40.53, 0.43, 485, 460, 2480, and 169 g ha
1. The results indicated that the annual fertilizer input fluxes were higher than input fluxes due to atmospheric deposition implying that fertilization is the main anthropogenic source of these elements. Similar results were also found by Luo et al. (Luo et al., 2009) in other agricultural soils of China. For Pb however, the annual Pb flux from atmospheric deposition (460 g ha
1) was about 3 times higher than from

Fig. 5. Lead isotope ratios in the different soil profiles P1, P2 and P3 (P1: vegetable-vegetable rotation soil, P2: uncultivated soil, P3: maize-wheat rotation soil).

658

W. Hu et al. / Environmental Pollution 237 (2018) 650e661

Fig. 6. Annual input fluxes of heavy metals from fertilizers and atmospheric deposition.

fertilization (156 g ha
1) demonstrating that atmospheric deposition is the most important anthropogenic source of Pb in agreement with other studies (Wong et al., 2003; Hou et al., 2014). In order to reduce the input flux of heavy metals from fertilizers and atmospheric deposition, it is necessary to control and optimize fertilization and to decrease the heavy metals emission to the atmosphere by the neighboring industrial production. 3.6. Source apportionment for different heavy metals using the PMF model To further identify and quantify the sources and contributions of heavy metals in soils in the study area, the PMF analysis was conducted (Table 2). The first source factor had high relative concentrations for Cd, Pb, Cu, Zn, and Cr. The source contribution of Cd for the first factor was up to 45.1%. Fertilizers (mineral fertilizers, organic fertilizers and/or manures) have often been highlighted as a major source of Cd, Cu, and Zn in agricultural soils (Belon et al., 2012; Hu et al., 2013). This may also apply in this case as the vegetable production that dominates in the study area is accompanied by high application of fertilizers. The spatial distribution of soil Cd, Cu, and Zn were similar to SOM (Fig. 3), implying an enhanced sorption of Cd, Cu and Zn to SOM. The results also suggested that accumulation of Cd, Cu and Zn was related to fertilizer application, especially organic fertilizers (Hu et al., 2013; Yang et al.,

2016). Therefore, the first source factor in Table 2 represents fertilizers. The second source factor is prominent for As, Pb, Cu, and Zn. For Pb this factor accounts for 32.1% of the sources. Industrial factories emit large amounts of dust containing As, Pb, Cu, and Zn that results in soil pollution with these elements. Previous studies using Pb isotope ratios found that substantial amounts of Pb in the environment came from traffic emissions (Huang et al., 2015). In addition, coal combustion constitutes an important anthropogenic source of soil Pb (Luo et al., 2015). Thus, Pb in the soils may be attributed to transportation and industrial emission sources (Cheng and Hu, 2010; Grezzi et al., 2011; Li et al., 2014). Emission from transportation and coal combustion could be important contributors of As and Cu in atmospheric deposition (Dong et al., 2015; Huang et al., 2015). Similarly, accumulation of Zn in the soils is also strongly associated with transportation. Thus, wearing of automobile tires was found to be significant non-point sources for Zn accumulation in urban surface soils (Wang et al., 2012; Radziemska and Fronczyk, 2015). Therefore, the second source factor is attributed to atmospheric deposition. For the third source factor, Hg and Cr received higher weighting than the other elements (Table 2). Many industrial plants including iron and steel, chemical, thermo-electric and sewage treatment locate surrounding the study area. According to Choppala et al. (2013), the main anthropogenic sources of Cr mostly originate

Table 2 Source contribution for different heavy metals as estimated by the PMF model. Elements

Profile contribution (mg kg
1) Factor 1

Factor 2

Factor 3

Factor 4

Factor 1

Factor 2

Factor 3

Factor 4

Cd As Hg Pb Cu Zn Cr

0.14 2.70 0.00 10.4 13.9 35.4 31.0

0.06 4.70 0.0002 10.7 16.7 42.5 25.7

0.03 2.08 0.03 6.37 7.49 21.6 39.5

0.07 1.54 0.06 5.84 6.32 18.3 0.88

45.1 24.5 0.00 31.2 31.4 30.1 31.9

20.3 42.6 0.15 32.1 37.5 36.1 26.5

10.8 18.9 33.2 19.2 16.9 18.3 40.6

23.8 14.0 66.8 17.6 14.2 15.5 0.91

Percentage contribution (%)

Note: Items of high loadings are emphasized with bold font in each contribution factor.

W. Hu et al. / Environmental Pollution 237 (2018) 650e661

from industrial activities, such as disposal of solid wastes, sewage sludge, spills and leaks from industrial metal processing and other industrial operations. The main source of Hg has been allocated to coal burning (Streets et al., 2005; Zhang et al., 2015). Streets et al. (2005) pointed out that of all Hg emissions in China, approximately 38% comes from coal combustion, 45% from non-ferrous metal smelting, and 17% from miscellaneous activities. The source factorization indicates that the coal-fired plant located to the northwest of the study area causes long-term industrial discharge and eventually leads to Hg accumulation in adjacent agricultural soils. Therefore, the third factor can be considered as a more specific source of industrial emissions which contributes to Cr and Hg accumulation in the surface soils. The fourth source factor was highly related to Hg and Cd, where Hg received higher weighting (66.8%) than the other elements, and thus Hg is the indicator element of this fourth factor. The mean

fatmospheric deposition ¼

3.7. Contribution rates of heavy metals from different sources and its implication for environmental management The overall contribution rates of various pollution sources to the total heavy metal contents in the soils as estimated by the PMF model are presented in Fig. 7. It can be seen that on average, atmospheric deposition has the largest contribution (33.0%), followed by fertilizer application (30.8%) and industrial emission (25.4%). The contribution from soil parent materials was only 10.8% which, in turn, demonstrated that anthropogenic contribution to the accumulation of soil heavy metals was significantly higher than those of natural sources, which was also verified by the results of the SA and IRA analyses aforementioned (Figs. 3 and 6). The contribution percentage (fatmospheric deposition, %) of atmospheric deposition can also be calculated by a mixing model using the following equation (Monna et al., 1997; Bird, 2011):

ð206=207 PbÞsurface soil
ð206=207 PbÞdeep soil ð206=207 PbÞatmospheric deposition
ð206=207 PbÞdeep soil

value of Hg content in the surface soils in the study area was 0.08 mg kg
1 which is close to the local background value (Table 1). Thus, we infer that except for the anthropogenic sources, Hg is mainly influenced by the natural geological background. Therefore, the fourth factor is allocated to soil parent materials. From the discussion above, four sources were apportioned including agricultural activities, atmospheric deposition, industrial emissions, and soil parent materials. Fig. S1 shows the relationship between the measured and the predicted concentrations by PMF of the 88 soil samples. The two highest correlations between the predicted and measured values are seen for Hg and Cr (r2 ¼ 0.999), whereas other elements also show strong correlations (r2 > 0.70) implying that the predicted values by PMF are significantly correlated to the true values (Fig. S1). Thus, it is obvious that the predicted results of the PMF model are in reasonable agreement with the observed values, and the selected factors could very much explain the information contained in the original data. The PMF model can be successfully used to apportion the sources of heavy metals in the soils (Pindado and Perez, 2011; Xue et al., 2014).

Fig. 7. Average mass contribution (%) of each source to total heavy metal contents in the soils as determined by PMF model.

659

 100%

(5)

Referring to Figs. 4 and 5, the mean values of 206Pb/207Pb for surface soils, deep soils and atmospheric deposition were 1.1645, 1.1751 and 1.1434, respectively. Thus, based on the mixing model, the average contribution of atmospheric deposition to Pb in surface soils was 33.4%, which was quite similar to the PMF model result (Table 2, 32.1%). These apportionment results are in accordance with the environmental conditions of the study area. Along the Yangtze River, more than 100 industrial factories are located, including iron and steel, chemical, metal refining, power, and sewage treatment plants which inevitably lead to large amounts of pollutant emissions including heavy metals (Zhao et al., 2010). It was acknowledged that some uncertainties associated with the source identification of heavy metals in the studied peri-urban agricultural soils did exist. The sources of heavy metals in soils are affected by many anthropogenic and natural factors. Each of the source identification methods has strengths and weakness. Firstly, the values of enrichment factor (EF) seemed to be not fitting or in contradiction with the PMF model results. The overlap composition of natural and anthropogenic sources by EF values may overestimate the contribution of the anthropogenic sources. Secondly, the sources from atmospheric deposition and industrial emission were regarded as the two different potential sources, but they cannot be discriminated confidently using the single input fluxes analysis (IFA) or PMF model. Further source investigation of heavy metals from the industrial emission should be conducted in order to obtain a better source apportionment in the future studies. Thirdly, it should be noted that the Pb isotope ratio analysis (IRA) in this study based on limited numbers of fertilizers, atmospheric deposition and soil profile samples. The samples were all collected in a short time period (from several weeks to one year) and might represent only part of temporally changing inputs as the source of fertilizers and atmospheric deposition have changed with time. Therefore, it is better to combine the different methods to quantitatively estimate the impacts of the different factors and improve the reliability of the source apportionment. Future researches should also be conducted to elucidate the relationships and dependencies between the different anthropogenic and natural sources.

660

W. Hu et al. / Environmental Pollution 237 (2018) 650e661

The results of the PMF model and the SA, IRA and IFA methods are mutually complementing and verify each other. The results indicate that soil accumulation of heavy metals in the study area was primarily related to human activities, especially to the atmospheric deposition and fertilization. Currently, integrated and holistic approaches are rare from the perspective of soil heavy metals accumulation and their sources. In China, soil baseline data is lacking for most studied areas, and is also a bottleneck in addressing the relationships between heavy metal sources, accumulation and multi-media distribution due to diverse geological properties and dominant soil forming factors (Lu et al., 2015; Tian et al., 2017). To prevent and control anthropogenic sources of heavy metals to agricultural soils, there is an urgent need for industrial discharge standards and optimization of fertilization to protect soil quality and guarantee safe agricultural products. Meanwhile, regular monitoring and integrated environmental management system to cover the whole agricultural production chain should be gradually established (Lu et al., 2015; Hu et al., 2017). 4. Conclusions An integrated approach consisting of SA, IRA, IFA, and PMF model is an effective method to identify the possible sources of heavy metals in peri-urban agricultural soils. Intensive agricultural production resulted in soil heavy metals accumulation and decrease of soil pH and increase in content of soil organic matter (SOM). According to the local background levels, Cd, As, Pb, Cu, Zn, and Cr appeared at different levels of accumulation in surface soils, especially for Cd with 49% of the sampled soils exceeding the Chinese threshold limit. Spatial variability of heavy metals in surface soils and isotope ratios of 206Pb/ 207Pb in soil profiles indicated that heavy metal accumulation in surface soils were significantly affected by human activities. The results of the IRA, IFA and PMF models suggested that atmospheric deposition along with fertilization were the major sources of heavy metals accumulation in the soils. In view of the significant role of atmospheric deposition and fertilization in heavy metals accumulation in the peri-urban agricultural soils, regular monitoring, source control and integrated environmental management should be implemented to control and reduce heavy metal inputs and guarantee the safety of agricultural production. Acknowledgments This study was supported by the National Natural Science Foundation of China (Grant No. 41101491), the National Sciencetechnology Support Plan Projects (Grant No. 2015BAD05B04), the Key Science and Technology Demonstration Project of Jiangsu Province (Grant No. BE2016812), and the Key Frontier Project of Institute of Soil Science, Chinese Academy of Sciences (Grant No. ISSASIP1629). The authors are also grateful for the support received by Wenyou Hu from the Visiting Scholar Project of China Scholarship Council (CSC Grant No. 201604910075). Appendix ASupplementary data Supplementary data related to this article can be found at https://doi.org/10.1016/j.envpol.2018.02.070. References Alleman, L.Y., Lamaison, L., Perdrix, E., Robache, A., Galloo, J.C., 2010. PM10 metal concentrations and source identification using positive matrix factorization and wind sectoring in a French industrial zone. Atmos. Res. 96, 612e625.

Atafar, Z., Mesdaghinia, A., Nouri, J., Homaee, M., Yunesian, M., Ahmadimoghaddam, M., Mahvi, A.H., 2010. Effect of fertilizer application on soil heavy metal concentration. Environ. Monit. Assess 160, 83e89. Belon, E., Boisson, M., Deportes, I.Z., Eglin, T.K., Feix, I., Bispo, A.O., Galsomies, L., Leblond, S., Guellier, C.R., 2012. An inventory of trace elements inputs to French agricultural soils. Sci. Total. Environ 439, 87e95. Bird, G., 2011. Provenancing anthropogenic Pb within the fluvial environment: developments and challenges in the use of Pb isotopes. Environ. Int 37, 802e819. Chae, J.S., Choi, M.S., Song, Y.H., Um, I.K., Kim, J.G., 2014. Source identification of heavy metal contamination using metal association and Pb isotopes in Ulsan Bay sediments, East Sea, Korea. Mar. Pollut. Bull. 88, 373e382. Chai, Y., Guo, J., Chai, S.L., Cai, J., Xue, L.F., Zhang, Q.W., 2015. Source identification of eight heavy metals in grassland soils by multivariate analysis from the Baicheng-Songyuan area, Jilin Province, Northeast China. Chemosphere 134, 67e75. Chen, H.Y., Teng, Y.G., Wang, J.S., Song, L.T., Zuo, R., 2013. Source apportionment of trace element pollution in surface sediments using positive matrix factorization combined support vector machines: application to the Jinjiang river, China. Biol. Trace Elem. Res. 151, 462e470. Chen, Y., Huang, B., Hu, W.Y., Weindorf, D.C., Liu, X.X., Niedermann, S., 2014. Assessing the risks of trace elements in environmental materials under selected greenhouse vegetable production systems of China. Sci. Total Environ 470, 1140e1150. Cheng, H.F., Hu, Y.A., 2010. Lead (Pb) isotopic fingerprinting and its applications in lead pollution studies in China: a review. Environ. Pollut 158, 1134e1146. Choppala, G., Bolan, N., JH, P., 2013. Chapter two: chromium contamination and its risk assessment in complex environmental settings. Adv. Agron 120, 129e172. Cloquet, C., Carignan, J., Libourel, G., Sterckeman, T., Perdrix, E., 2006. Tracing source pollution in soils using cadmium and lead isotopes. Environ. Sci. Technol 40, 2525e2530. CRGCST, 2001. Cooperative Research Group on Chinese Soil Taxonomy (CRGCST). Chinese Soil Taxonomy. Science Press, Beijing (In Chinese). Cui, J., Zhou, J., Peng, Y., He, Y.Q., Yang, H., Mao, J.D., Zhang, M.L., Wang, Y.H., Wang, S.W., 2014. Atmospheric wet deposition of nitrogen and sulfur in the agroecosystem in developing and developed areas of Southeastern China. Atmos. Environ 89, 102e108. Dong, L., Hu, W., Huang, B., Liu, G., Qu, M., Kuang, R., 2015. Source appointment of heavy metals in suburban farmland soils based on positive matrix factorization. China Environ. Sci. 35 (7), 2103e2111 (In Chinese). 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, 307e324. Grezzi, G., Ayuso, R.A., De Vivo, B., Lima, A., Albanese, S., 2011. Lead isotopes in soils and groundwaters as tracers of the impact of human activities on the surface environment: the Domizio-Flegreo Littoral (Italy) case study. J. Geochem. Explor 109, 51e58. Guo, J.H., Liu, X.J., Zhang, Y., Shen, J.L., Han, W.X., Zhang, W.F., Christie, P., Goulding, K.W.T., Vitousek, P.M., Zhang, F.S., 2010. Significant acidification in major Chinese croplands. Science 327, 1008e1010. Gupta, I., Salunkhe, A., Kumar, R., 2012. Source apportionment of PM10 by positive matrix factorization in urban area of Mumbai, India. Sci World J. https://doi.org/ 10.1100/2012/585791. Hansmann, W., Koppel, V., 2000. Lead-isotopes as tracers of pollutants in soils. Chem. Geol. 171, 123e144. Hou, Q.Y., Yang, Z.F., Ji, J.F., Yu, T., Chen, G.G., Li, J., Xia, X.Q., Zhang, M., Yuan, X.Y., 2014. Annual net input fluxes of heavy metals of the agro-ecosystem in the Yangtze River delta, China. J. Geochem. Explor 139, 68e84. Hu, W.Y., Zhang, Y., Huang, B., Teng, Y., 2017. Soil environmental quality in greenhouse vegetable production systems in eastern China: current status and management strategies. Chemosphere 170, 183e195. Hu, W.Y., Huang, B., Shi, X.Z., Chen, W.P., Zhao, Y.C., Jiao, W.T., 2013. Accumulation and health risk of heavy metals in a plot-scale vegetable production system in a pen-urban vegetable farm near Nanjing, China. Ecotox. Environ. Safe 98, 303e309. Hu, W.Y., Huang, B., Weindorf, D.C., Chen, Y., 2014. Metals analysis of agricultural soils via portable X-ray fluorescence spectrometry. B. Environ. Contam. Tox 92, 420e426. Huang, B., Shi, X.Z., Yu, D.S., Oborn, I., Blomback, K., Pagella, T.F., Wang, H.J., Sun, W.X., Sinclair, F.L., 2006. Environmental. assessment of small-scale vegetable farming systems in peri-urban areas of the Yangtze River Delta Region, China. Agr. Ecosyst. Environ 112, 391e402. Huang, Y., Li, T.Q., Wu, C.X., He, Z.L., Japenga, J., Deng, M., Yang, X., 2015. An integrated approach to assess heavy metal source apportionment in peri-urban agricultural soils. J. Hazard. Mater 299, 540e549. 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, 271e285. Komarek, M., Ettler, V., Chrastny, V., Mihaljevic, M., 2008. Lead isotopes in environmental sciences: a review. Environ. Int 34, 562e577. Li, J., Zheng, C., 1989. The Handbook of the Environmental Background Values. China Environmental Science Press, Beijing (In Chinese). Li, L.J., Holm, P.E., Marcussen, H., Hansen, H.C.B., 2014. Release of cadmium, copper and lead from urban soils of Copenhagen. Environ. Pollut 187, 90e97.

W. Hu et al. / Environmental Pollution 237 (2018) 650e661 Lofts, S., Tipping, E., Lawlor, A.J., Shotbolt, L., 2013. An intermediate complexity dynamic model for predicting accumulation of atmospherically-deposited metals (Ni, Cu, Zn, Cd, Pb) in catchment soils: 1400 to present. Environ. Pollut 180, 236e245. Lu, A.X., Wang, J.H., Qin, X.Y., Wang, K.Y., Han, P., Zhang, S.Z., 2012. Multivariate and geostatistical analyses of the spatial distribution and origin of heavy metals in the agricultural soils in Shunyi, Beijing, China. Sci. Total Environ 425, 66e74. Lu, Y.L., Song, S., Wang, R.S., Liu, Z.Y., Meng, J., Sweetman, A.J., Jenkins, A., Ferrier, R.C., Li, H., Luo, W., Wang, T.Y., 2015. Impacts of soil and water pollution on food safety and health risks in China. Environ. Int 77, 5e15. Luo, L., Ma, Y.B., Zhang, S.Z., Wei, D.P., Zhu, Y.G., 2009. An inventory of trace element inputs to agricultural soils in China. J. Environ. Manage 90, 2524e2530. Luo, X.S., Xue, Y., Wang, Y.L., Cang, L., Xu, B., Ding, J., 2015. Source identification and apportionment of heavy metals in urban soil profiles. Chemosphere 127, 152e157. MA, 2012. Ministry of Agriculture of the People's Republic of China. Agricultural Industry Standards of the People's Republic of China (NY525e2011): Organic Fertilizers Standard (In Chinese). MEP, 1995. Ministry of Environmental Protection of the People's Republic of China. The Soil Environmental Quality Standards of the People's Republic of China (GB15618e1995). China Environmental Science Press, Beijing (In Chinese). Monna, F., Lancelot, J., Croudace, I.W., Cundy, A.B., Lewis, J.T., 1997. Pb isotopic composition of airborne particulate material from France and the southern United Kingdom: implications for Pb pollution sources in urban areas. Environ. Sci. Technol 31, 2277e2286. Nelson, D.W., Sommers, L.E., 1996. Total carbon, organic carbon, and organic matter. In: Sparks, D.L. (Ed.), Methods of Soil Analysis, Part 3, Chemical Methods e SSSA Book Series No. 5, pp. 961e1010. Madison, Wisconsin. Nicholson, F.A., Smith, S.R., Alloway, B.J., Carlton-Smith, C., Chambers, B.J., 2003. An inventory of heavy metals inputs to agricultural soils in England and Wales. Sci. Total Environ 311, 205e219. Paatero, P., 1997. Least squares formulation of robust non-negative factor analysis. Chemometr. Intell. Lab 37, 23e35. Paatero, P., Hopke, P.K., 2009. Rotational tools for factor analytic models. J. Chemometr 23, 91e100. Pan, Y.P., Wang, Y.S., 2015. Atmospheric wet and dry deposition of trace elements at 10 sites in Northern China. Atmos. Chem. Phys. 15, 951e972. 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, 233e237. Pindado, O., Perez, R.M., 2011. Source apportionment of particulate organic compounds in a rural area of Spain by positive matrix factorization. Atmos. Pollut. Res. 2, 492e505. QSQA, 2009. Quality Supervision and Quarantine Administration of the People's Republic of China. National Standards of the People's Republic of China (GBT 23349e2009): Ecological Indicators of Arsenic, Cadmium, lead, Chromium, Mercury in Fertilizers (In Chinese). Qu, C.S., Shi, W., Guo, J., Fang, B.B., Wang, S., Giesy, J.P., Holm, P.E., 2016. China's soil pollution control: choices and challenges. Environ. Sci. Technol 50, 13181e13183. Radziemska, M., Fronczyk, J., 2015. Level and contamination assessment of soil along an expressway in an ecologically valuable area in Central Poland. Int. J.

661

Env. Res. Pub. He 12, 13372e13387. Reimann, C., Flem, B., Fabian, K., Birke, M., Ladenberger, A., Negrel, P., Demetriades, A., Hoogewerff, J., Team, G.P., 2012. Lead and lead isotopes in agricultural soils of Europe - the continental perspective. Appl. Geochem 27, 532e542. Sharma, R.K., Agrawal, M., Marshall, F.M., 2008. Atmospheric deposition of heavy metals (Cu, Zn, Cd and Pb) in Varanasi City, India. Environ. Monit. Assess 142, 269e278. Song, Y., Zhang, Y.H., Xie, S.D., Zeng, L.M., Zheng, M., Salmon, L.G., Shao, M., Slanina, S., 2006. Source apportionment of PM2.5 in Beijing by positive matrix factorization. Atmos. Environ 40, 1526e1537. Streets, D.G., Hao, J.M., Wu, Y., Jiang, J.K., Chan, M., Tian, H.Z., Feng, X.B., 2005. Anthropogenic mercury emissions in China. Atmos. Environ 39, 7789e7806. Tian, K., Huang, B., Xing, Z., Hu, W.Y., 2017. Geochemical baseline establishment and ecological risk evaluation of heavy metals in greenhouse soils from Dongtai, China. Ecol. Indic 72, 510e520. Toth, G., Hermann, T., Da Silva, M.R., Montanarella, L., 2016. Heavy metals in agricultural soils of the European Union with implications for food safety. Environ. Int 88, 299e309. USEPA, 2014. U.S. Environmental protection agency. EPA positive matrix factorization (PMF) 5.0 fundamentals and user guide. Available at: https://www.epa.gov/ sites/production/files/2015-02/documents/pmf_5.0_user_guide.pdf. . Walraven, N., van Gaans, P.F.M., van der Veer, G., van Os, B.J.H., Klaver, G.T., Vriend, S.P., Middelburg, J.J., Davies, G.R., 2013. Tracing diffuse anthropogenic Pb sources in rural soils by means of Pb isotope analysis. Appl. Geochem 37, 242e257. Wang, D.G., Tian, F.L., Yang, M., Liu, C.L., Li, Y.F., 2009. Application of positive matrix factorization to identify potential sources of PAHs in soil of Dalian, China. Environ. Pollut 157, 1559e1564. Wang, M.E., Bai, Y.Y., Chen, W.P., Markert, B., Peng, C., Ouyang, Z.Y., 2012. A GIS technology based potential eco-risk assessment of metals in urban soils in Beijing, China. Environ. Pollut 161, 235e242. Wong, C.S.C., Li, X.D., Zhang, G., Qi, S.H., Peng, X.Z., 2003. Atmospheric deposition of heavy metals in the pearl River delta, China. Atmos. Environ 37, 767e776. Xue, J.L., Zhi, Y.Y., Yang, L.P., Shi, J.C., Zeng, L.Z., Wu, L.S., 2014. Positive matrix factorization as source apportionment of soil lead and cadmium around a battery plant (Changxing County, China). Environ. Sci. Pollut. R 21, 7698e7707. Yang, Y., Chen, W.P., Wang, M.E., Peng, C., 2016. Regional accumulation characteristics of cadmium in vegetables: influencing factors, transfer model and indication of soil threshold content. Environ. Pollut 219, 1036e1043. Yu, Y., Li, Y.X., Li, B., Shen, Z.Y., Stenstrom, M.K., 2016. Metal enrichment and lead isotope analysis for source apportionment in the urban dust and rural surface soil. Environ. Pollut 216, 764e772. Zhang, L., Wang, S.X., Wang, L., Wu, Y., Duan, L., Wu, Q.R., Wang, F.Y., Yang, M., Yang, H., Hao, J.M., Liu, X., 2015. Updated emission inventories for speciated atmospheric mercury from anthropogenic sources in China. Environ. Sci. Technol 49, 3185e3194. Zhao, Y.C., Wang, Z.G., Sun, W.X., Huang, B.A., Shi, X.Z., Ji, J.F., 2010. Spatial interrelations and multi-scale sources of soil heavy metal variability in a typical urban-rural transition area in Yangtze River Delta region of China. Geoderma 156, 216e227.