Combining emission inventory and isotope ratio ...

Chemosphere 204 (2018) 140e147

Contents lists available at ScienceDirect

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

Combining emission inventory and isotope ratio analyses for quantitative source apportionment of heavy metals in agricultural soil Lian Chen, Shenglu Zhou*, Shaohua Wu**, Chunhui Wang, Baojie Li, Yan Li, Junxiao Wang School of Geographic and Oceanographic Sciences, Nanjing University, Nanjing 210046, China

h i g h l i g h t s

g r a p h i c a l a b s t r a c t

Emission inventory and IRA were combined for source apportionment of heavy metals. Raster analysis was used to calculate the input rate for each source type. IsoSource was used to quantify the source contributions in the IRA method. The results obtained using the two methods were similar.

a r t i c l e i n f o

a b s t r a c t

Article history: Received 7 February 2018 Received in revised form 23 March 2018 Accepted 1 April 2018 Available online 5 April 2018

Two quantitative methods (emission inventory and isotope ratio analysis) were combined to apportion source contributions of heavy metals entering agricultural soils in the Lihe River watershed (Taihu region, east China). Source apportionment based on the emission inventory method indicated that for Cd, Cr, Cu, Pb, and Zn, the mean percentage input from atmospheric deposition was highest (62e85%), followed by irrigation (12e27%) and fertilization (1e14%). Thus, the heavy metals were derived mainly from industrial activities and traffic emissions. For Ni the combined percentage input from irrigation and fertilization was approximately 20% higher than that from atmospheric deposition, indicating that Ni was mainly derived from agricultural activities. Based on isotope ratio analysis, atmospheric deposition accounted for 57e93% of Pb entering soil, with the mean value of 69.3%, which indicates that this was the major source of Pb entering soil in the study area. The mean contributions of irrigation and fertilization to Pb pollution of soil ranged from 0% to 10%, indicating that they played only a marginally important role. Overall, the results obtained using the two methods were similar. This study provides a reliable approach for source apportionment of heavy metals entering agricultural soils in the study area, and clearly have potential application for future studies in other regions. © 2018 Elsevier Ltd. All rights reserved.

Handling Editor: T Cutright Keywords: IsoSource Atmospheric deposition Irrigation Fertilization Spatial distribution Lihe river

1. Introduction

* Corresponding author. ** Corresponding author. E-mail addresses: [email protected] (L. Chen), [email protected] (S. Zhou), [email protected] (S. Wu), [email protected] (C. Wang), dg1627010@smail. nju.edu.cn (B. Li), [email protected] (Y. Li), [email protected]. cn (J. Wang). https://doi.org/10.1016/j.chemosphere.2018.04.002 0045-6535/© 2018 Elsevier Ltd. All rights reserved.

Agricultural soil is a long-term sink for potentially toxic elements including cadmium (Cd), chromium (Cr), copper (Cu), nickel (Ni), lead (Pb), and zinc (Zn) (Nicholson et al., 2003; Zhu et al., 2013; Hou et al., 2014; Chen et al., 2015, 2017a; Cao et al., 2016; Ding and Xu, 2016; Jeon et al., 2017; Nedelescu et al., 2017). Heavy metals

L. Chen et al. / Chemosphere 204 (2018) 140e147

including these enter the agro-ecosystem through both natural and anthropogenic processes (Hou et al., 2014; Yang et al., 2016; Jiang et al., 2017; Li et al., 2017; Lin et al., 2017; Xia et al., 2017). Natural sources include weathering, desertification, erosion, and other geological processes. Anthropogenic processes may include inputs of heavy metals through fertilization and irrigation, and atmospheric deposition (Dach and Starmans, 2005; Luo et al., 2009; Sheppard et al., 2009; Yang et al., 2009; Xia et al., 2014; Lin et al., 2017). Atmospheric deposition has been identified as a principal source of metals entering soil (Gray et al., 2003; Loyola et al., 2007; Rothwell et al., 2011; Xia et al., 2014); agricultural activities, including fertilization and irrigation using sewage wastewater are also very important pollution sources (Kachenko and Singh, 2006; Lu et al., 2012). Previous studies have indicated that inputs of heavy metals to soils through agricultural activities have increased in recent decades because of world population expansion (Li et al., 2008; Niu et al., 2013; Huang et al., 2015). As a result of the diverse routes of entry of heavy metals into agricultural soils, there is a growing public concern about heavy metal accumulation in agricultural products, and the consequent human health effects through the agricultural food chain (Luo et al., 2011; Yu et al., 2017; Bi et al., 2018; Han et al., 2018). Action is required to control pollution by heavy metals to reduce the risks of their accumulation. Source apportionment is a crucial step in this process (Lu et al., 2012; Jiang et al., 2017). Methods commonly used for heavy metal source apportionment include multivariate statistical analysis and geographic information system (GIS) mapping (Qu et al., 2013; Huang et al., 2015; Zhou et al., 2016; Li et al., 2017; Lin et al., 2017). Multivariate statistics and GIS mapping are highly subjective and cannot be used for quantitative source apportionment (Chen et al., 2015; Huang et al., 2015; Wang et al., 2016; Hou et al., 2017; Bi et al., 2018; Han et al., 2018); however, emission inventory methods can be used for this purpose. The emission inventory approach involves initial source identification, quantitation of the input of heavy metals to agricultural soil from the sources (including atmospheric deposition, fertilization, and irrigation) using flux observation methods, and calculation of the contribution rate for each source type. Isotope ratio analysis (IRA) is an alternative method for detecting the fingerprint of heavy metal pollution. Comparison of the isotope ratios in the potential sources (environmental samples) with those in affected soils enables the contribution of each source to soil pollution to be estimated (Rosman et al., 1993; Hansmann and Koppel, 2000; DuzgorenAydin, 2007; Foucher et al., 2009; Huang et al., 2015). For technical reasons not all isotope ratios are easy to determine, and only Pb isotope ratio analysis has been widely used (Duzgoren-Aydn and Weiss, 2008; Foucher et al., 2009). Methods for quantitative source apportionment, particularly emission inventory methods, have rarely been applied in research, and the methods for source apportionment have commonly been used in isolation. Generally, only one quantitative method has been used in each study (Phillips and Gregg, 2003; Huang et al., 2015). Therefore, the present study intends to integrate and apply two quantitative analytical methods (an emission inventory method and IRA) at a study site in a verification process, with the aim of obtaining more accurate and reliable results. This study also addresses the issue that agricultural soils merit more attention, because contamination of these soils with heavy metals poses longterm threats to food safety and human health. In addition, ecosystem health may be compromised and result in lower agricultural outputs. These issues highlight that agricultural soils are a priority research area related to source apportionment. The study aimed to develop a combined approach to providing more accurate and reliable source apportionment data. First, heavy metal input fluxes to the agro-ecosystem from atmospheric

141

deposition, fertilization, and irrigation were monitored, and the input rate for each source type was calculated using GIS spatial analysis of raster data. Second, IRA of Pb was used to quantify the source contributions of individual heavy metals using IsoSource, which is a contribution calculation software (Phillips and Gregg, 2003); this was also used to validate the results obtained using the emission inventory method. Third, the advantages and disadvantages of the two methods were assessed, and cost-effective methods proposed for analysis of source apportionment. 2. Material and methods 2.1. Study area The study area, in the Lihe River watershed, is located to the west of Taihu Lake, in the city of Yixing, Jiangsu Province, China (31090 0000 e31200 3100 N, 119 420 0000 e119 560 2000 E); it includes the towns of Hufu and Dingshu. Taihu Lake is the largest lake in the China Eastern Coastal Area. It is located in the lower reaches of the Yangzi River Basin, which is one of the most developed areas and the most populous regions in China. The watershed has a surface area of approximately 260 km2 (Li et al., 2006). The total area of agricultural land is 57.8 km2, of which the dry land is 21.4 km2, and the paddy land is 36.4 km2. Many types of industrial activities occur densely throughout the area, amongst which are ceramics factories, refractory materials plants, and chemical plants. Rice and wheat are cultivated in the agricultural zone. Increasing heavy metal concentrations were observed in the soil of the lower Yangzi River plain in 2004e2014. In particular, the ratio between the median concentrations of Cd (18.2%) and Pb (6.9%) increased during this period (Xia et al., 2017). In addition, previous studies that undertook biomonitoring of heavy metal pollution in soils in the river estuaries of the 24 main rivers flowing into and out of Taihu Lake revealed that the Lihe River estuary was the most strongly contaminated. Moreover, the potential ecological hazard index (RI) for the Lihe River estuary was greater than 220, which indicates that the pollution in this area had reached a serious level and presented a very high ecological risk (Jiao et al., 2010). However, there has been little research into the source apportionment of heavy metals in this high-risk area. 2.2. Sample collection and preparation 2.2.1. Soil Soils in crop fields were sampled at 0e10 cm depth at 32 randomly selected sites throughout the study area during 17e21 May 2016 (Fig. S1, Supplementary Materials). Each sample (0.5e1.0 kg) was divided into 5e9 subsamples. Following collection the samples were air dried at room temperature, ground, and passed through a 2 mm nylon sieve to remove stones and plant roots (Lin et al., 2016). The resulting fine soil powders were stored in polythene zip bags (Li et al., 2012; Lin et al., 2018). 2.2.2. Atmospheric deposition According to the type of land use and urban layout of the study area, 10 wet/dry deposition monitoring points were set up in the form of a cross. One axis incorporated (sequentially) woodland, farmland, suburban area, a town center, suburban area, and farmland from the northwest to the southeast, and the other axis incorporated farmland, a town center, suburban area, and woodland from the northeast to the southwest. Monitoring was performed between 1 September 2016 and 1 September 2017 using a custom-made collecting device (Fig. S2) Three replicate samples were collected at each sampling point. Following collection the

142


samples were evaporated to dryness, and the steamed samples were weighed and analyzed to determine the heavy metal concentrations and the Pb isotope ratios. 2.2.3. Irrigation water Irrigation water samples (32) were collected in fields next to the soil sampling points and at the same time (17e21 May 2016). These were transferred to polyethylene bottles, nitric acid (0.1% v/v) was added as a preservative (Chen et al., 2017b), and the bottles were stored at 20 C until the samples were analyzed. In the laboratory, 100 mL volumes from each water sample were combined and mixed (3200 mL total volume), then filtered through weighed 0.45 mm membrane filters to collect the suspended particulate matter. The filters and attached materials were placed in an oven and dried to a constant weight at 60 C. The mass and heavy metal concentrations of the dried suspended matter were determined. The concentrations of heavy metals in filtered irrigation water were also determined. The concentrations of heavy metals in the particulate matter collected on the filters and in the filtered irrigation water (mg/L) were combined to calculate the total heavy metal concentration (mg/L) in irrigation water samples. 2.2.4. Chemical fertilizers A standardized detailed questionnaire on the types and annual amounts of added fertilizers was developed and implemented as a survey among farmers during 17e21 May 2016. The results indicated that the area could be divided into two sub-areas based on administrative regions. In each sub-area, large amounts of nitrogen, phosphate, potash, compound fertilizers, and organic fertilizers were applied (Table S1, Supplementary Materials). After the investigation, each type of fertilizer used (10 in total) was obtained from selected retail stores for heavy metal analysis. 2.3. Chemical analysis A Milestone ETHOS 1 microwave sample preparation system with temperature control was used to digest the samples, including prepared soils, steamed matter from atmospheric deposition samples, fertilizers, and dried suspended matter from irrigation water samples, for total metal measurement (Hu et al., 2011). Heavy metal concentrations in these samples were determined following chemical extraction, as follows. Approximately 100 mg of each sample was digested in 3 mL of 37% HCl, 1 mL of 65% HNO3, 6 mL of 65% HF, and 0.5 mL of 65% HClO4 (Hu et al., 2011; Lin et al., 2015; Wang et al., 2015). A two-stage digestion program was used, involving initial heating to 200 C over 10 min, then digestion for 15 min at 200 C (Hu et al., 2011). Following cooling the digestion solutions were evaporated to near dryness, and dissolved in 1.0 mL of 65% HNO3, after which deionized water (20 mL) was added. The solutions were stored in 25 mL high density polyethylene vials at 4 C until instrumental analysis (Hu et al., 2011). The concentrations of Cd, Zn, Pb, Cu, Cr, and Ni in each solution were determined using an inductively coupled plasmaemass spectrometer (ICPeMS; PerkinElmer SCIEX, Elan 9000); the operational conditions are listed in Table S2, Supplementary Materials. The standard reference material GBW07405 (National Research Center for Standards, China) was used to verify the analytical accuracy. Reagent blanks and three analytical duplicates were analyzed with each soil sample. The heavy metal concentrations in the filtered irrigation water were analyzed directly using ICPeMS, with no other treatment. The standard water reference material used was GBW (E) 080194. The 208Pb/206Pb and 207Pb/206Pb ratios were determined using ICPeMS. Following digestion, the sample solutions were diluted to a concentration of approximately 30 ng mL1 Pb (Hu et al., 2014). Lead isotope ratios were corrected using standard reference

material SRM981 (National Institute of Standards and Technology, NIST, USA). To calculate the precision and accuracy of the Pb isotope determinations by normalization, NIST SRM981 was analyzed following analysis of every two samples (Hu et al., 2014). The analytical precision of the Pb isotope ratio determinations based on NIST SRM981 at 30 ng/ml Pb were 0.36% for the 208Pb/206Pb ratio and 0.25% for the 207Pb/206Pb ratio. 2.4. Data processing and statistical analysis Pre-processing of the GPS data and the heavy metal concentrations in samples were conducted using Microsoft Excel 2010. During spatial analysis using ArcMap 10.0 the inverse distance weighting method was applied for spatial interpolation. To explore the possible relationship between the mass flux of atmospheric deposition, the density of traffic network, and the density of industrial sites, a point density analysis of the industrial sites and a linear density analysis of the roads were carried out. The data for industry were taken from the local environmental protection bureau and the data for the local traffic network were extracted from a land use map for 2016. Density analysis were performed in the ArcGIS Spatial Analyst module. The grid density method was selected for the density analysis. The raster calculator function in GIS Spatial Analyst was used to predict the spatial distribution pattern of heavy metal input fluxes to agricultural soil from atmospheric deposition, irrigation, and fertilization, and to determine the rates at which these sources contributed to the total input flux to all the agricultural land in the study area. Spatial analysis of raster data in the whole process can provide very accurate results, comparing with simple calculation of the input flux and its percentage at specific sampling points. The input flux of heavy metals to agricultural land from atmospheric deposition was calculated according to the following equation:

D¼CM M¼

m

pR2

(1) (2)

where D is the amount of heavy metal input to agricultural land from atmospheric deposition (mg/m2/a), C is the heavy metal concentration in atmospheric deposition (mg/kg), M is the annual mass flux of atmospheric deposition (kg/m2/a), m is the mass of steamed matter from an atmospheric deposition sample in one settling plastic bucket (kg/a), and R (0.1 m in this study) is the radius of bunghole of the bucket (m). Heavy metal inputs to agricultural land from irrigation were estimated on the basis of the annual volume of irrigation water applied and the heavy metal concentration in the irrigation water, according to the following equation:

I ¼ ðCw þ Cm Mm Þ V 103

(3)

where I is the heavy metal input from irrigation (mg/m2/a), Cw is the concentration of heavy metals in filtered irrigation water (mg/ L), Cm is the concentration of heavy metals in suspended particulate matter in irrigation water (mg/kg), Mm is the mass of suspended particulate matter per liter of irrigation water (kg/L), V is the volume of irrigation water applied (m3/m2/a), and 103 is unit conversion factor. The annual volume of irrigation water used in Jiangsu Province is 0.65 m3/m2 (Jiang and Zhang, 2016). The area of paddy land is 1.7-fold greater than that in the dry agricultural land area, and the amount of water used in paddy land is 4-fold that in the dry land area (Statistical Bureau of Jiangsu Province, 2016). Thus, the


annual volume of water used in paddy land is approximately 0.9 m3/m2, while that in dry agricultural land is 0.225 m3/m2. Heavy metal inputs to agricultural land from fertilization were estimated from the annual amounts of fertilizer applied and the content of heavy metals in the fertilizers, according to the following equation:

I¼

n X

Mi Ci

(4)

i¼1

where IFer is the amount of heavy metals contributed by fertilizer (i) (mg/m2/a), Mi is the amount of fertilizer (i) applied (kg/m2/a), and Ci is the concentration of heavy metal (i) in fertilizer (mg/kg). To identify anthropogenic sources, the results of the IRA analysis were plotted in a coordinate system using Origin 8.5, with different isotope ratios forming the horizontal and vertical axes. The source contributions were calculated using IsoSource. Previous studies have used a three end-member model to calculate source contributions (Fabrice et al., 1997; Li et al., 2011), but this is limited to three sources. IsoSource can be extended to include more than three sources, based on the following equations used in this software (Phillips and Gregg, 2003; Huang et al., 2015):

Rm ¼

n X

Pi Ri

(5)

i¼1

I¼

n X

Ri

(6)

i

where Rm is the isotope ratio of soil, Pi is the percentage contributed from the various sources, and Ri is the isotope ratio of the sources. This is a mathematically underdetermined system of two equations involving three unknowns, for which there is no unique solution. If there are n isotope systems and n þ 1 sources, then we can find certain proportions. However, if there are > n þ 1 sources and n isotope systems, based on the requirement for mass balance conservation it is possible to find multiple combinations of source proportions that give feasible solutions (Phillips and Gregg, 2003). In this study there were two isotope systems and five potential sources, so IsoSource was used.

3. Results and discussion 3.1. Source apportionment using the emission inventory method 3.1.1. Heavy metal input from atmospheric deposition The annual mass fluxes of atmospheric deposition (M, in Equation (1)) for each sampling site in the study area is shown in Fig. S3. The annual mass fluxes of atmospheric deposition varied greatly among sites. Site 10 had the smallest mass flux (0.047 kg/m2/a); this site is located in an area of bamboo forest that is little affected by human activities. Site 8 had the highest mass flux (0.097 kg/m2/a). The annual mass fluxes of atmospheric deposition at sites 3, 7, and 9 were much lower than at the other 5 sampling sites. The mass fluxes of atmospheric deposition along the two line transects showed a spatial distribution trend of low/high/low values, coinciding with rural area/town/rural area land uses, respectively. The spatial distributions of annual mass fluxes of atmospheric deposition for the entire study area are shown in Fig. S4. Point density analysis of industrial sites and linear density analysis of traffic network were carried out and superimposed to create the final density map shown in Fig. S5. It was observed that the spatial distribution of annual mass flux of atmospheric deposition was strongly correlated with

143

the combined density map for industry and the traffic network; this analysis showed low level deposition in the forest zones and high levels at the periphery of the town center. The mean annual heavy metal input to agricultural land from deposition (mg/m2/a) in the study area followed the order: Zn (156.575) > Pb (16.385) > Cu (10.906) > Cr (5.994) > Ni (3.317) > Cd (0.630). This pattern approximates those reported in previous studies (Wong et al., 2002; Blumberg, 2006; Huang et al., 2009; Luo et al., 2009; Hou et al., 2014). In a previous research program, deposition fluxes of 64.7, 20.2, 10.8, 6.1, 5.8 and 0.40 mg/m2/a for Zn, Pb, Cu, Cr, Ni and Cd respectively, were reported for China (Luo et al., 2009). The inputs of Cd and Zn from atmospheric deposition in the study area were higher than the average level for China, and the deposition of Cu is approximately equal the China average, while deposition of Cr, Ni, and Pb was lower than the China average (Luo et al., 2009). It is noteworthy that the atmospheric deposition of Cd, Cu and Zn reflects relatively heavy pollution. The total annual deposition of Cd and Zn found in this study is higher than the deposition fluxes for heavy metals in the Yangtze River Delta reported by Huang et al. (2009), while the levels of deposition of Cr, Cu, Ni, and Pb are lower. However, the annual atmospheric deposition values of Cu (18.6 mg/m2/a) and Ni (8.35 mg/m2/a) in the Pearl River delta region (Wong et al., 2002) were higher than those in the present study, while the levels of the other four metals were lower. The flux of heavy metal inputs from atmospheric deposition was much more variable, probably indicative of local differences in conditions and releases to the atmosphere. The spatial distribution of annual heavy metal input fluxes from atmospheric deposition in the study area is shown in Fig. 1. For all heavy metals high values were detected at the periphery of the town center located in the eastern and middle parts of the study area, and low values were found in the forest zone located in the southwestern part of the study area. These patterns were strongly associated with the density of industrial activities and the traffic network in the study area. 3.1.2. Heavy metal input from irrigation The mean input of heavy metals (mg/m2/a) from irrigation in the study area was in the order of Zn (22.664) > Ni (4.539) > Pb (3.709) > Cu (2.682) > Cr (2.458) > Cd (0.097). In comparison, the mean input fluxes of heavy metals from irrigation in the Yangtze River delta were 0.556, 3.498, 13.730, 2.449, and 24.785 mg/m2/a for Cd, Cr, Cu, Pb, and Zn, respectively (Hou et al., 2014). It is noteworthy that only the input of Pb from irrigation in the study area was higher than the average input level in the Yangtze River delta. The annual spatial distribution pattern of heavy metal input fluxes from irrigation in agricultural land of the study area is shown in Fig. S6. 3.1.3. Heavy metal input from fertilization The values for the heavy metal concentration in fertilizers are listed in Table S3, and indicate that in general the heavy metal concentrations in organic fertilizer were highest, followed by phosphate fertilizer, with potash fertilizer having the lowest concentrations. The mean input fluxes of heavy metals from fertilization (mg/m2/a) in paddy and dry land are listed in Table S4. By comparison, the mean input flux of heavy metals from nitrogenous fertilizer was the highest among the five kinds of fertilizer and that from potash fertilizer was the lowest. The order of the annual mean total input fluxes of heavy metals (mg/m2/a) from five kinds of fertilizer in the study area was Zn (13.313) > Cu (2.376) > Ni (1.113) > Cr (1.066) > Pb (0.635) > Cd (0.011). The input fluxes of Cd, Cr, Cu, and Pb from fertilization in the study area were lower than in the Yangtze River delta, but the flux was higher for Zn (Hou et al., 2014).

144


Fig. 1. The spatial distribution of annual heavy metal flux from atmospheric deposition in the study area (mg/m2/a). For all heavy metals the deeper the colors the higher the input fluxes. (For interpretation of the references to color in this figure legend, the reader is referred to the Web version of this article.)

3.1.4. Comparison of input fluxes The annual percentages of heavy metal input into agricultural soils from atmospheric deposition, fertilization, and irrigation in the study area are listed in Table 1. In general, atmospheric deposition contributed the most, followed by irrigation, with fertilization contributing the least. The mean percentage input from atmospheric deposition for all heavy metals ranged from 37% to 85%. For irrigation, the mean percentage inputs ranged from 12% to 50%, while for fertilizer they ranged from 1% to 14%. The input flux percentages for individual heavy metals also varied greatly. The major source for Cd, Cr, Cu, Pb, and Zn was atmospheric deposition, while for Ni the combined percentage input from irrigation and fertilization was approximately 20% higher than that from atmospheric deposition. These results indicate that for Cd, Cr, Cu, Pb, and Zn, the major pollution sources are industrial activities and traffic emissions, whereas the major source of Ni pollution is agricultural activities. Our results are partially consistent with previous studies (Tang et al., 2007; Cong et al., 2008). In the Beijing area, inputs from atmospheric deposition are highest, followed by irrigation, with fertilization being the lowest (Cong et al., 2008). In the Chengdu economic zone the total input of Cd comprised 89% from atmospheric deposition, 7% from irrigation, and 4% from fertilization (Tang et al., 2007).

used to distinguish between different potential Pb sources, and to verify the results of the emissions inventory method. The values for the 208Pb/206Pb ratio were plotted against values for the 207Pb/206Pb ratio for all groups of possible pollutant sources and polluted soils

3.2. Source apportionment based on IRA

Fig. 2. Pb isotope ratios of soil and heavy metal sources. Symbols of the same color represents the same samples. SM in water: suspended matter in irrigation water. Filtered water: filtered irrigation water. (For interpretation of the references to color in this figure legend, the reader is referred to the Web version of this article.)

To generate more accurate and reliable results, IRA was also

Table 1 The percentage of each type of input flux to soil (%). Industrial activity

Agricultural activity

Atmospheric deposition

Cd Cr Cu Ni Pb Zn

Irrigation

Fertilization

Range

Mean ± S.D

Range

Mean ± S.D

Range

Mean ± S.D

15.26e97.38 39.59e85.51 52.90e80.65 19.36e64.60 56.89e93.72 59.86e96.69

84.84 ± 12.27 62.23 ± 10.13 67.48 ± 7.07 37.79 ± 12.55 76.55 ± 9.07 80.36 ± 6.96

1.59e84.41 7.36e46.60 5.81e33.41 17.43e73.97 4.50e38.64 1.33e32.29

13.56 ± 12.39 26.87 ± 11.11 18.17 ± 8.18 49.96 ± 15.73 20.29 ± 9.13 12.73 ± 6.38

0.33e2.42 6.16e21.41 10.69e24.91 6.67e22.60 1.60e5.05 1.89e13.27

1.61 ± 0.29 10.59 ± 1.95 14.08 ± 1.55 11.5 ± 2.97 3.13 ± 0.47 6.80 ± 1.80


in the study area (Fig. 2). It is noteworthy that the Pb isotope ratios for the compound fertilizer samples (0.769 ± 0.004 for 207Pb/206Pb and 1.863 ± 0.002 for 208Pb/206Pb) were very different from those for the soil samples in the plot; thus, the values for compound fertilizers are not shown in Fig. 2. This phenomenon indicated that compound fertilizer may not contribute Pb to soil. The lead isotope ratios of 32 soil samples and 10 atmospheric deposition samples were relatively close in value, but mostly different from the values for filtered irrigation water and phosphate fertilizer samples. Thus, the Pb isotope ratios in the soil differed significantly from those in filtered irrigation water and phosphate fertilizer samples, indicating that these sources generally did not contribute substantially to Pb pollution of the soil in the study area. In contrast, atmospheric deposition samples, and to a lesser extent those of suspended matter in irrigation water, nitrogen fertilizers, potash fertilizers, and organic fertilizers clustered in the center of the plot near the soil sample cluster, implying that these were more likely to be important contributors to soil Pb pollution. For chemical fertilizers, the plot showed a scattered distribution, which is consistent with chemical fertilizers originating from very diverse sources. The Pb source contribution rates calculated using IsoSource (Table 2) were consistent with the plot (Fig. 2). The contribution of atmospheric deposition to soil Pb pollution ranged from 57% to 93%, with the mean value of 69.3%, which indicates that atmospheric deposition was the major and most quantitatively important source of Pb contamination in the study area. The results derived from the IRA method approximated the results from use of the emission inventory method. However, the mean contributions of irrigation water and fertilizers were all about 10% respectively, indicating that these sources made only marginal contributions to soil pollution with Pb. 3.3. Comparison of the two methods of source apportionment The emission inventory method initially requires input of information on the amount of heavy metals entering agricultural soil from potential sources including atmospheric deposition, chemical fertilizers, and irrigation water; this forms the basis for calculating the contribution rate for each type of source. Although the emission inventory method can achieve high accuracy, it is time consuming, laborious, and costly. Consequently, few studies have used this method for source apportionment of heavy metal pollution of soil and the environment. Some studies have calculated the annual input fluxes of heavy metals to agricultural soil in other study areas (Cong et al., 2008; Luo et al., 2009; Hou et al., 2014). Although the input sources of various heavy metals have shown clear spatial heterogeneity among different study areas, our results are generally consistent with those of previous reports. This includes that the major source of input of Pb is atmospheric deposition, which usually accounts for >80%, suggesting that this source of Pb pollution is similar throughout the country. Industrial plants using coal as fuel,

145

together with traffic emissions, are generally the main source of Pb input to soils and the environment. With the phasing out of leaded petrol the input of Pb to soil has declined (Duzgoren-Aydn and Weiss, 2008), but vehicle brakes and tires continue to emit Pb to the environment, so traffic remains an important source of environmental pollution by Pb (Huang et al., 2015). As there are no regional differences in the types of coal, vehicle brakes, and tires used through the country, these sources of contamination of agricultural soil by Pb are relatively consistent throughout China. However, the sources of other heavy metals varied greatly. For example, Hou et al. (2014) calculated the annual heavy metal input fluxes to the agro-ecosystem in the Yangtze River delta, and found that the major input source for Cd was irrigation (67%), exceeding that from atmospheric deposition (32%) and fertilization (1%). In contrast, Luo et al. (2009) conducted an inventory of average inputs of trace elements to agricultural soils in China, and found that for Cd the major input was from fertilization (63%), while 35% came from atmospheric deposition and only 2% came from irrigation. Cong et al. (2008) estimated the ecological risk of Cd in agricultural soil in 10 districts of Beijing, and found that the Cd input flux from atmospheric deposition accounted for 40e98% of the total inputs, which varies greatly in 10 districts. Tang et al. (2007) reported that in the Chengdu economic zone the source proportions comprising the total input of Cd were 89% from atmospheric deposition, 7% from irrigation, and 4% from fertilization. Thus, the major input sources for other heavy metals vary markedly among different study areas. The IRA method can also provide quantitative information regarding the contribution of various pollutant sources. Diagrammatic plots of three isotopes for each heavy metal enable visual assessment of the distances among clusters in IRA data, providing the most useful tool for identifying heavy metal sources with a high degree of certainty. In the present study, detailed source identification and distributions were assessed using IsoSource to mathematically verify the IRA data. The IsoSource program is based on mass balances, and was applied in a ternary mixing model. Compared with the emission inventory method, the Pb isotope method requires relatively small numbers of samples. The highprecision measurements of isotope ratios allow for a high discriminatory power with minimal statistical manipulation, and it is possible to provide definitive answers from a small number of samples (Cheng and Hu, 2010). However, the application potential of this method is limited. Not all isotope ratios are technically easy to determine, and only Pb isotope ratio analysis has been widely used. Few studies involving Hg and Cd isotope ratios have been reported (Yun et al., 2008; Huang et al., 2015). In addition, the cost of heavy metal isotope ratio analysis limits its routine use. Consequently, few studies have used IRA for heavy metal apportionment in soil and the environment. However, IRA was used in our study to obtain more accurate results, and to validate the results of the emission inventory method.

Table 2 Rates of contribution of Pb sources to soil pollution based on average isotope ratios. Samples

Isotope ratio 207

Pb/

Atmospheric deposition samples Suspended matter in water Nitrogen fertilizers Potash fertilizers Organic fertilizers Phosphate fertilizers Compound fertilizers Filtered irrigation water

206

Pb

0.862 ± 0.002 0.850 0.850 ± 0.001 0.852 ± 0.009 0.839 ± 0.001 0.826 ± 0.001 0.769 ± 0.004 0.886 ± 0.005

Range and mean value of contribution rates (%) 208

Pb/

206

Pb

2.117 ± 0.002 2.078 2.071 ± 0.002 2.074 ± 0.012 2.066 ± 0.002 2.007 ± 0.003 1.863 ± 0.002 2.185 ± 0.020

57e93 (69.3) 0e43 (8.8) 0e37 (7.4) 0e39 (7.8) 0e33 (6.6) 0 0 0

146


3.4. Limitation of the study and future research directions Our study provides a new protocol for heavy metal source analysis that enables accurate results to be obtained under manpower, financial, and time constraints. We performed quantitative sources apportionment of heavy metals in agricultural soil in the following steps. First, heavy metal input fluxes to the agricultural soil from atmospheric deposition, fertilization, and irrigation were monitored, and the input rate for each source type was calculated. Emission inventory is a necessary activity, and ensuring highly accurate results depends on high density spatial sampling. Second, representative soil samples and potential source samples are selected for the IRA procedure. As most analysis situations involve n isotopes and > n þ 1 sources, the source contributions can easily be calculated using IsoSource, which can provide the ranges of proportional contributions from the various sources, rather than a fixed number. Third, we compared the results from the two methods to obtain more precise and reliable analytical results. The application of pesticides, another source of heavy metals in agricultural soils, was not examined in this study. Many previous studies have also failed to consider this source of contamination (Wong et al., 2002; Cong et al., 2008; Hou et al., 2014), or have only investigated the impact of pesticides on soil microbes, soil animals, and soil enzymatic activities (Lu et al., 2009; Nico et al., 2009; Pan et al., 2011). This may have introduced bias into the source apportionment of heavy metals in this context. Various pesticides are known to contain small amounts of heavy metals. Although the use of pesticides in agricultural production mainly introduces organic pollutants, the contribution of heavy metals to agricultural soil by pesticides cannot be ignored if the accuracy of source apportionment is to be ensured. Therefore, in future studies, a standardized detailed questionnaire on the types and annual amounts of pesticides added to soils should be developed and used to survey farmers in the study area. The heavy metal concentrations and isotope ratios of pesticides should also be determined. This will greatly improve the accuracy of the heavy metal source apportionment reported for agro-ecosystems, which will allow more informed policies to be developed for pollution mitigation. 4. Conclusions An approach that combined emission inventory and isotope ratio analyses, which together are more powerful than other individual methods, was used to facilitate heavy metal source apportionment for selected agricultural soils in a representative study area in southeast China. In the study area, among the input source percentages for Cd, atmospheric deposition was highest (84.8%), followed by irrigation (13.6%), with fertilization being the lowest (1.6%); the order of input source percentages for Cr was atmospheric deposition (62.2%) > irrigation (26.9%) > fertilization (10.6%); the major input source for Cu was atmospheric deposition (67.5%), exceeding that from irrigation (18.2%) and fertilization (14.1%); the source proportions comprising the total input of Pb were 76.6% from atmospheric deposition, 20.3% from irrigation, and 3.1% from fertilization; for Zn the major input was from atmospheric deposition (80.4%), while 12.7% came from irrigation and only 6.8% came from fertilization. These results indicate that the source of these five heavy metals was mainly industrial activities and traffic emissions. For Ni, the percentage input from irrigation and fertilization combined was approximately 20% higher than that from atmospheric deposition, which indicates that the major source of Ni was agricultural activities. The source apportionment for Pb obtained using IRA indicated that atmospheric deposition of this pollutant heavy metal to soil was 57e93%, with the mean value of 69.3%; consequently, atmospheric deposition is the major and

most quantitatively important source of Pb contamination in the study area. In contrast, the mean contributions from irrigation and different fertilizers were all about 0e10% respectively, indicating that these sources only marginally contribute to pollution of soil by Pb. The results of the two methods were generally within the same range. To protect the agro-ecosystem of the Lihe River watershed, government regulators should undertake the efficient and targeted management of these anthropogenic sources of different heavy metals, and address the metal pollution of the soil generated by human activities. This study will be useful in both improving the local soil quality and providing a basis for policies that effectively target the protection of soils from long-term heavy metal accumulation. Acknowledgements This work was supported by the Key Technology Support Program of Jiangsu Province (grant number BE2015708), the National Natural Science Foundation of China (grant number 41771243), the Special Fund for Research in the Public Interest of the Ministry of Land and Resources (201511001-03), and the National Key Research and Development Plan (2017YFD0800305). Appendix A. Supplementary data Supplementary data related to this article can be found at https://doi.org/10.1016/j.chemosphere.2018.04.002. References Bi, C., Zhou, Y., Chen, Z., Jia, J., Bao, X., 2018. Heavy metals and lead isotopes in soils, road dust and leafy vegetables and health risks via vegetable consumption in the industrial areas of Shanghai, China. Sci. Total Environ. 619e620, 1349e1357. Blumberg, H., 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. Cao, E.W., Wang, B., Wang, M., Shen, N.N., Zhang, Q.Q., 2016. Migration characteristics of heavy metals pollution in soil and groundwater of regenerate lead industry. Environ. Monitor. Forewarn. Chen, H., Teng, Y., Lu, S., Wang, Y., Wu, J., Wang, J., 2015. Source apportionment and health risk assessment of trace metals in surface soils of Beijing metropolitan, China. Chemosphere 144, 1002. Chen, L., Gao, J., Zhu, Q., Wang, Y., Yang, Y., 2017a. Accumulation and output of heavy metals in Spartina alterniflora in a salt marsh. Pedosphere. Chen, L., Zhou, S., Shi, Y., Wang, C., Li, B., Li, Y., Wu, S., 2017b. Heavy metals in food crops, soil, and water in the Lihe River Watershed of the Taihu region and their potential health risks when ingested. Sci. Total Environ. 615, 141. Cheng, H., Hu, Y., 2010. Lead (Pb) isotopic fingerprinting and its applications in lead pollution studies in China: a review. Environ. Pollut. 158, 1134e1146. Cong, Y., Zheng, P., Chen, Y.L., Hou, Q.Y., 2008. Ecological risk assessments of heavy metals in soils of the farmland ecosystem of Beijing, China. Geol. Bull. China 27, 681e688. Dach, J., Starmans, D., 2005. Heavy metals balance in Polish and Dutch agronomy: actual state and previsions for the future. Agr. Ecosyst. Environ. 107, 309e316. Ding, X., Xu, J., 2016. Characteristics and source apportionment of heavy metal pollution in ambient air in Pudong New Area. Environ. Monitor. Forewarn. Duzgoren-Aydin, N.S., 2007. Sources and characteristics of lead pollution in the urban environment of Guangzhou. Sci. Total Environ. 385, 182. Duzgoren-Aydn, N.S., Weiss, A.L., 2008. Use and abuse of Pb-isotope fingerprinting technique and GIS mapping data to assess lead in environmental studies. Environ. Geochem. Hlth. 30, 577e588. Fabrice, M., Joel, L., Ian, W.C., Andrew, B.C., James, 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. Foucher, D., Ogrinc, N., Hintelmann, H., 2009. Tracing mercury contamination from the Idrija mining region (Slovenia) to the Gulf of Trieste using Hg isotope ratio measurements. Environ. Sci. Technol. 43, 33. Gray, C.W., Mclaren, R.G., Roberts, A.H.C., 2003. Atmospheric accessions of heavy metals to some New Zealand pastoral soils. Sci. Total Environ. 305, 105. Han, W., Gao, G., Geng, J., Li, Y., Wang, Y., 2018. Ecological and health risks assessment and spatial distribution of residual heavy metals in the soil of an e-waste circular economy park in Tianjin, China. Chemosphere 325e335. Hansmann, W., Koppel, V., 2000. Lead-isotopes as tracers of pollutants in soils. Chem. Geol. 171, 123e144.

L. Chen et al. / Chemosphere 204 (2018) 140e147 Hou, D., O'Connor, D., Nathanail, P., Tian, L., Ma, Y., 2017. Integrated GIS and multivariate statistical analysis for regional scale assessment of heavy metal soil contamination: a critical review. Environ. Pollut. 231, 1188e1200. Hou, Q., Yang, Z., Ji, J., Yu, T., Chen, G., Li, J., Xia, X., Zhang, M., Yuan, X., 2014. Annual net input fluxes of heavy metals of the agro-ecosystem in the Yangtze River Delta, China. J. Geochem. Explor. 139, 68e84. Hu, X., Sun, Y., Ding, Z., Zhang, Y., Wu, J., Lian, H., Wang, T., 2014. Lead contamination and transfer in urban environmental compartments analyzed by lead levels and isotopic compositions. Environ. Pollut. 187, 42e48. Hu, X., Zhang, Y., Luo, J., Wang, T., Lian, H., Ding, Z., 2011. Bioaccessibility and health risk of arsenic, mercury and other metals in urban street dusts from a megacity, Nanjing, China. Environ. Pollut. 159, 1215e1221. Huang, S., Tu, J., Liu, H., Hua, M., Liao, Q., Feng, J., Weng, Z., Huang, G., 2009. Multivariate analysis of trace element concentrations in atmospheric deposition in the Yangtze River Delta, East China. Atmos. Environ. 43, 5781e5790. Huang, Y., Li, T., Wu, C., He, Z., 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. Jeon, S.K., Kwon, M.J., Yang, J.S., Lee, S., 2017. Identifying the source of Zn in soils around a Zn smelter using Pb isotope ratios and mineralogical analysis. Sci. Total Environ. 601e602, 66e72. Jiang, H., Zhang, B., 2016. Utilization of water resources and decoupling evaluation with economic development in Jiangsu Province. Jiangsu Agr. Sci. 44, 616e619. Jiang, Y., Chao, S., Liu, J., Yang, Y., Chen, Y., Zhang, A., Cao, H., 2017. Source apportionment and health risk assessment of heavy metals in soil for a township in Jiangsu province, China. Chemosphere 168, 1658e1668. Jiao, W., Lu, S.Y., Li, G.D., Jing, X.C., Yu, H., Cai, M.M., 2010. Heavy metal pollution of main inflow and outflow rivers around the Taihu lake and assessment of its potential ecological risk. Chin. J. Appl. Environ. Biol. 16, 577e580. Kachenko, A.G., Singh, B., 2006. Heavy metals contamination in vegetables grown in urban and metal smelter contaminated sites in Australia. Water Air Soil Poll. 169, 101e123. Li, C., Wang, F., Cao, W., Pan, J., Lv, J., Wu, Q., 2017. Source analysis, spatial distribution and pollution assessment of heavy metals in sewage irrigation area farmland soils of Longkou City. Environ. Sci. 38, 1018e1027. Li, H.B., Yu, S., Li, G.L., Deng, H., Luo, X.S., 2011. Contamination and source differentiation of Pb in park soils along an urban-rural gradient in Shanghai. Environ. Pollut. 159, 3536e3544. Li, H.P., Huang, W.Y., Yang, G.S., Liu, X.M., 2006. Non-point pollutant concentrations for different land uses in Lihe River Watershed of Taihu region. China Environ. Sci. 26, 243e247. Li, Q., Chen, Y., Fu, H., Cui, Z., Shi, L., Wang, L., Liu, Z., 2012. Health risk of heavy metals in food crops grown on reclaimed tidal flat soil in the Pearl River Estuary, China. J. Hazard Mater. 227e228, 148e154. Li, Y., Gou, X., Wang, G., Zhang, Q., Su, Q., Xiao, G., 2008. Heavy metal contamination and source in arid agricultural soil in central Gansu Province, China. J. Environ. Sci. 20, 607e612. Lin, C., Ma, R., Xiong, J., 2018. Can the watershed non-point phosphorus pollution be interpreted by critical soil properties? A new insight of different soil P states. Sci. Total Environ. 628e629, 870e881. Lin, C., Ma, R., Zhu, Q., Li, J., 2015. Using hyper-spectral indices to detect soil phosphorus concentration for various land use patterns. Environ. Monit. Assess. 187, 4130. Lin, C., Wu, Z., Ma, R., Su, Z., 2016. Detection of sensitive soil properties related to non-point phosphorus pollution by integrated models of SEDD and PLOAD. Ecol. Indic. 60, 483e494. Lin, Y., Han, P., Huang, Y., Yuan, G.L., Guo, J.X., Li, J., 2017. Source identification of potentially hazardous elements and their relationships with soil properties in agricultural soil of the Pinggu district of Beijing, China: multivariate statistical analysis and redundancy analysis. J. Geochem. Explor. 173, 110e118. Loyola, J., Quiterio, S.L., Escaleira, V., 2007. Atmospheric concentrations and dry deposition fluxes of particulate trace metals in Salvador, Bahia, Brazil. Atmos. Environ. 41, 7837e7850. Lu, A., Wang, J., Qin, X., Wang, K., Han, P., Zhang, S., 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, 66. Lu, X.Z., Jin, J.H., Hao, J.C., Gao, D.X., Zhang, L.N., Liu, H.F., Zhao, J.N., 2009. Responses of soil enzyme activities in different soil layers to single and combined stress of Hg and Cd. J. Agr. Sci. Cambr. 1844e1848. Luo, L., Ma, Y., Zhang, S., Wei, D., Zhu, Y.G., 2009. An inventory of trace element

147

inputs to agricultural soils in China. J Environ. Manage 90, 2524e2530. Luo, X.S., Yu, S., Li, X.D., 2011. Distribution, availability, and sources of trace metals in different particle size fractions of urban soils in Hong Kong: implications for assessing the risk to human health. Environ. Pollut. 159, 1317. Nedelescu, M., Baconi, D., Neagoe, A., Iordache, V., Stan, M., Constantinescu, P., Ciobanu, A.M., Vardavas, A.I., Vinceti, M., Tsatsakis, A.M., 2017. Environmental metal contamination and health impact assessment in two industrial regions of Romania. Sci. Total Environ. 580, 984e995. Nicholson, F.A., Smith, S.R., Alloway, B.J., Carlton-Smith, C., Chambers, B.J., 2003. An inventory of heavy metal input to agricultural soil in England and Wales. Sci. Total Environ. 311, 205e219. Nico, E., Matthias, K., Stephan, P., Alexandercw, S., Christoph, S., Wolfgangw, W., Stefan, S., 2009. No interactive effects of pesticides and plant diversity on soil microbial biomass and respiration. Appl. Soil Ecol. 42, 31e36. Niu, L., Yang, F., Xu, C., Yang, H., Liu, W., 2013. Status of metal accumulation in farmland soils across China: from distribution to risk assessment. Environ. Pollut. 176, 55e62. Pan, P., Yang, J.C., Deng, S.H., Jiang, H.M., Zhang, J.F., Ling-Ling, L.I., Shen, F., 2011. Proceedings and prospects of pesticides and heavy metals contamination in soil-plant system. J. Agr. Sci. Cambr. 30, 2389e2398. Phillips, D.L., Gregg, J.W., 2003. Source partitioning using stable isotopes: coping with too many sources. Oecologia 136, 261. Qu, M.K., Li, W.D., Zhang, C.R., Wang, S.Q., Yang, Y., He, L.Y., 2013. Source apportionment of heavy metals in soils using multivariate statistics and geostatistics. Pedosphere 23, 437e444. €rlach, U., 1993. Isotopic Rosman, K.J.R., Chisholm, W., Boutron, C.F., Candelone, J.P., Go evidence for the source of lead in Greenland snows since the late 1960s. Nature 362, 333e335. Rothwell, J.J., Taylor, K.G., Evans, M.G., Allott, T.E., 2011. Contrasting controls on arsenic and lead budgets for a degraded peatland catchment in Northern England. Environ. Pollut. 159, 3129. Sheppard, S.C., Grant, C.A., Sheppard, M.I., De, J.R., Long, J., 2009. Risk indicator for agricultural inputs of trace elements to Canadian soils. J. Environ. Qual. 38, 919e932. Statistical bureau of Jiangsu Province, Statistical yearbook of Jiangsu, 2016. Tang, Q.F., Yang, Z.F., Zhang, B.R., Jin, L.X., 2007. Cadmium flux in soils of the agroecosystem in the Chengdu economic region, Sichuan, China. Geol. Bull. China 26, 869e877. Wang, C., Yang, Z., Zhong, C., Ji, J., 2016. Temporal-spatial variation and source apportionment of soil heavy metals in the representative river-alluviation depositional system. Environ. Pollut. 216, 18e26. Wang, Y.Y., Xin-Yan, L.U., Wang, N., Chen, C., Gao, Y., Liu, D., Peng, H., 2015. Comparison between microwave digestion and automatic sample digestion in the determination of heavy metals in soil using ICPeMS. Environ. Monitor. Forewarn. Wong, S.C., Li, X.D., Zhang, G., Qi, S.H., Min, Y.S., 2002. Heavy metals in agricultural soils of the Pearl River Delta, South China. Environ. Pollut. 119, 33e44. Xia, X., Yang, Z., Cui, Y., Li, Y., Hou, Q., Yu, T., 2014. Soil heavy metal concentrations and their typical input and output fluxes on the southern Song-nen Plain, Heilongjiang Province, China. J. Geochem. Explor. 139, 85e96. Xia, X., Yang, Z., Yu, T., Hou, Q., Mutelo, A.M., 2017. Detecting changes of soil environmental parameters by statistics and GIS: A case from the lower Changjiang plain, China. J. Geochem. Explor. Yang, Y., Zhou, Z., Bai, Y., Cai, Y., Chen, W., 2016. Risk assessment of heavy metal pollution in sediments of the Fenghe River by the fuzzy synthetic evaluation model and multivariate statistical methods. Pedosphere 26, 326e334. Yang, Z.P., Lu, W.X., Long, Y.Q., 2009. Atmospheric dry and wet deposition of heavy metals in Changchun City, China. Res. Environ. Sci. 22, 28e34. Yu, L.I., Hong-Guan, L.I., Liu, F.C., 2017. Pollution in the urban soils of Lianyungang, China, evaluated using a pollution index, mobility of heavy metals, and enzymatic activities. Environ. Monit. Assess. 189, 34. Yun, S.G., Jung, G.B., Kim, W.I., Lee, J.S., Kim, M.K., Kim, J.H., Shin, J.D., Lee, D.B., Kim, S.C., 2008. Evaluation on the fate of Cd in soil and plant by using stable isotope methodology. Kor. J. Soil Sci. Fertil. 41. Zhou, J., Feng, K., Pei, Z., Meng, F., Sun, J., 2016. Multivariate analysis combined with GIS to source identification of heavy metals in soils around an abandoned industrial area, Eastern China. Ecotoxicology 25, 380. Zhu, D.D., Zhao, C.P., Zhang, Y., Zhang, J., Hong, C., Fu, J., Zhu, H.L., An, S.Q., 2013. Pollution character and estimation of source of heavy metals in surface sediments of the Jialu River. Environ. Monitor. Forewarn. 5, 41e45.